Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2006-08-16

00:00:47JdGordonamiconn: it needs to use CONTEXT_SETTINGS on archos?
00:01:39amiconnOn Ondio, Left was accept as well as Right. Only Mode was Cancel
00:01:55amiconnNow Left is cancel, that's ugly
00:02:36safetydanbusy irc logs
00:02:43LinusNamiconn: this has been discussed in irc a few times
00:03:21LinusNi tend to like that Left is accept as well
00:03:38LinusNyes, when we have discussed the action patch
00:04:48amiconnThe settings context was/is there...
00:05:20amiconnIt's just that the other change of today makes it unused (for the normal settings)
00:05:48LinusNbecause it suddenly uses the standard/list context?
00:06:32 Quit senab (" HydraIRC -> <- State of the art IRC")
00:06:48 Quit Kohlrabi ("Quit")
00:06:50LinusNhmmm, npt all that easy to solve
00:07:12JdGordonah, forgot about that.. na it should be easy to fix, even clean things up a bit
00:07:20LinusNshould it? how?
00:07:27linuxstbIsn't it an improvement if we have less contexts?
00:07:45amiconnThe settings context is still used, just not for the normal settings
00:07:46LinusNlinuxstb: in general, yes
00:08:18linuxstbWhere is the settings context still used?
00:09:13*JdGordon has to go, ill read the logs and fix shit up during the dya
00:09:21amiconnWhjat about that shortcut in action.c?
00:09:33JdGordonwhich shortcut?
00:09:49JdGordon.. hopefully everything will be back t normal when you all wake up....
00:09:53jsMikeSamiconn: Want to inspect FS #5816 and make sure I'm not ridiculous? :)
00:09:59LinusNJdGordon: this one:
00:10:03LinusN /* Special case to make the keymaps smaller */
00:10:03LinusN if(button & BUTTON_REPEAT)
00:10:03LinusN last_button &= ~BUTTON_REPEAT;
00:10:25JdGordonamiconn doesnt want to get rid of it does he?
00:10:37amiconnI've taken it out, and changed the table lines that really should fire multiple event as having a precondition of BUTTON_NONE
00:10:38LinusNhe does, and i'm inclined to agree
00:11:17amiconnWorks, is less code, allows long press events that fire only once, and the tables have the same size
00:11:30JdGordonoh ok
00:12:24amiconnHmm, but something still doesn't work right
00:13:10 Part Paul_The_Nerd
00:13:19 Quit Lear ("Chatzilla 0.9.74 [Firefox]")
00:13:45amiconnAha, there's another line that needs to be removed
00:14:42safetydanJdGordon has commit access now?
00:15:52JdGordonisnt that scary!!?
00:16:16*JdGordon really gone now
00:16:22safetydannah, just means we don't have to listen to "commit my patch!" requests :)
00:16:27jsMikeSme too...that's terror!
00:16:37safetydanjsMikeS, you poking around the thread model in the sim?
00:16:55peturSlasheri ?
00:16:55amiconnLinusN: If the shortcut gets removed, the check ((items[i].pre_button_code == last_button) || (items[i].button_code == last_button)) has to be changed to (items[i].pre_button_code == last_button)
00:17:02 Join Xerion [0] (
00:17:08jsMikeSsafetydan: I want to...don't know where to start except to redo the SDL layer
00:18:11 Join senab [0] (
00:18:13safetydanjsMikeS, SDL 1.3 should let us fix the threading (if it ever comes out) as it adds a yield function so we can then remove the evil hack
00:18:36jsMikeSWhat's the evil hack?
00:18:38amiconnLinusN: There's something wrong with the restored TREE_EXIT behaviour
00:18:53amiconnThe old "trapped in a sub-dir" behaviour is back...
00:19:32safetydanjsMikeS, yield is currently emulated as "release mutex, sleep(0), grab mutex". it works well enough but is not pretty
00:19:43amiconnHmm, maybe it's sth else...
00:20:25Damme_hmmmm, How do I charge my ipod ? it enters transfermode =&
00:20:29Bagdersafetydan: we need to make sure we lock all other but one thread from running, is that what that new yield() is supposed to do?
00:20:38jsMikeSsafetydan: What so other threads can run amok? Take the mutex and then your thread will block in sleep. Not pretty.
00:21:18linuxstbDamme_: Hold the MENU button as you insert the cable.
00:21:26Damme_linuxstb aah thanks m8
00:21:33jsMikeSI don't know I'd have to think all the variables throught or I'll say dumb stuff. :)
00:21:41linuxstb(or just let it charge in disk mode)
00:21:56safetydanBagder, there's a new yield?
00:21:59Damme_linuxstb yeah, but I wanted to play around ;)
00:22:10Bagdersafetydan: the one you mentioned is in the new SDL I mean
00:22:19safetydanah okay
00:22:30safetydanyes that's what it's supposed to do yes
00:22:49safetydanthe x11 code did it this way as well
00:22:52 Quit mkey (""Welcome to IRC; Where men are men, women are men and little girls are FBI agents!"")
00:23:03Bagderyes I wrote it ;-)
00:23:03 Quit Daishi (Read error: 104 (Connection reset by peer))
00:23:44safetydanSo I can blame you because my poor copy doesn't work then right? :)
00:24:03 Join Daishi [0] (
00:24:24 Quit bluebrother ("Leaving")
00:24:32linuxstbSomeone just write emulators for all our targets...
00:24:35Bagdersafetydan: oh sure, I'm blame-resistant so fire away!
00:24:56*Bagder has his asbestos on
00:25:10 Quit Daishi (Remote closed the connection)
00:25:26safetydanI've always wondered if we could just use the existing threading code in the rockbox kernel though.
00:25:38*jsMikeS just got a new flamethrower for his last birthday
00:26:12jsMikeSsafetydan: With Windows fibers that's exactly, I believe, what could be done.
00:26:22Bagdersafetydan: it would probably be possible to make a sigjmp() thing to simulate it
00:26:25safetydanjsMikeS, that doesn't solve the problem for me on Linux :)
00:26:50Bagderthe target approach is a bit too lowlevel for simulators
00:27:09LinusNwhat problem are we trying to solve here?
00:27:31Bagdertalking about making the simulated treading more similar to the target way
00:27:33 Join Daishi [0] (
00:27:34jsMikeSsafetydan: Linux doesn't have anything you can schedule yourself that automatically saves the CPU context?
00:27:43*amiconn wants a blamethrower
00:27:52jsMikeSCan YOU save the CPU context?
00:27:59LinusNamiconn: to australia?
00:28:23safetydanjsMikeS, don't know
00:28:30*safetydan looks at SCHED_RR...
00:29:07Bagderah, that's the name I was looking for
00:29:24safetydanI wondered if you meant setjmp but sigjmp returned google hits so I stayed quiet
00:30:17LinusNin fact, the target scheduler should be able to use setjmp()/longjmp()
00:30:30jsMikeSsafetydan: I image it can be done but are those sorts of operations privileged?
00:30:47LinusNbut then we have to implement setjmp() and longjmp() on all targets
00:30:58jsMikeSWindows has it
00:31:18LinusNall proper C libraried have it
00:31:25LinusNlibraries even
00:31:29 Quit RedNfire ("blub")
00:31:29BagderCONFORMING TO
00:31:29Bagder POSIX, ISO 9899 (C99
00:31:58jsMikeSlirbarders rrrr
00:32:11jsMikeSNot sure if anything in Windows is proper
00:32:57 Quit Daishi (Remote closed the connection)
00:33:21LinusNamiconn: any progress with the action.c "shortcut"?
00:33:37amiconnFor some reason I cannot leave lists at all now
00:33:44 Join Daishi [0] (
00:33:54amiconns/lists/folders in the tree/
00:33:55 Quit Daishi (Remote closed the connection)
00:34:04amiconnMenus do work
00:34:48 Join Daishi [0] (
00:36:39 Quit Daishi (Client Quit)
00:36:41 Quit Febs ("CGI:IRC (Ping timeout)")
00:37:04linuxstbjsMikeS: BTW, nice work on your X5 asm lcd_yuv_blit() - can you remember the overall improvement in speed compared to the initial implementation?
00:37:56 Quit MadDog011 ()
00:38:06 Quit ender` (" It could have been worse. Imagine you were trying to download pr0n and you end up with some naruto ep.")
00:38:13 Quit senab (" HydraIRC -> <- Leading Edge IRC")
00:38:53XavierGrHiya all!
00:39:06XavierGrLOL I just read todays log...
00:39:24Bagderwe are now officially 59 committers
00:39:32XavierGrI am happy that I didn't update my build today before I left for my village!
00:39:35*safetydan finds some fibre emulating code for Linux...
00:39:56linuxstbBagder: Are all 59 committers active? Can you produce some nice stats?
00:40:07BagderI'll see..
00:40:18LinusNBagder loves stats
00:40:26XavierGrme too
00:40:31*petur is about to get active ;)
00:40:41 Quit Hansmaulwurf ("( :: NoNameScript 3.81 :: )")
00:41:38XavierGrpetur: I see you are progressing well. Are you about to fix the "corrupted-OF-settings" bug?
00:41:53 Join Daishi [0] (
00:41:59safetydanBagder, I think I've asked this before, but have you ever looked at getting FIshEye running on the repository?
00:42:05Bagdergrep author: dump | cut "-d " -f7 | tr -d ';' | sort | uniq -c | sort -rn
00:42:16Bagderand the winner is...
00:42:32Bagder 1459 dave
00:42:36Bagder 1153 amiconn
00:42:40Bagder 851 tucoz
00:42:43Bagder 812 hardeeps
00:42:50Bagder 740 linus
00:42:54Bagder 569 bagder
00:43:01Mikachudoes each commit count as the number of files committed?
00:43:02LinusN5th place.... i suck!
00:43:14*linuxstb can take a month off
00:43:16 Quit mantono ("Don't worry, be happy")
00:43:16Bagderthis count counts each modified file
00:43:27peturXavierGr: read/write the complete eeprom as first step <−− about to commit
00:43:29Bagderso committing 5 files at once would give you a count of 5
00:43:30XavierGrwow I thought amiconn would be first!
00:43:54Bagder44 persons have committed since 2.5
00:43:57XavierGrnice stats though
00:44:00linuxstbAnd remember, quantity counts, not quality.
00:44:05XavierGrbagder are you going to add them to the wiki?
00:44:17XavierGrlinuxstb: lol
00:44:21 Join mantono [0] (
00:44:23peturthe oops commits should be removed from that list ;)
00:44:38LinusNpetur: nooo, then i fall to 6th place
00:45:48*Bagder makes another stats check...
00:45:53Bagder... since day 1
00:46:17linuxstbWhat were those previous stats for? since25?
00:46:47Bagdersince I had a 'cvs log' dump for that lying around
00:47:15*mantono is away: Auto-away after 2 mins idle (gone at 15th Aug, 22:47:14)
00:47:15 Nick mantono is now known as mantono`afk (
00:47:18Bagdercvs log on the whole thing takes a while
00:47:48*LinusN wants to kick mantono for spamming the channel
00:47:56Bagder 1 3323 bagder
00:47:56preglownothing like three instructions to increment a variable in memory
00:48:00Bagder 2 2937 amiconn
00:48:01*preglow does a piledriver on arm
00:48:03Bagder 3 2683 zagor
00:48:07linuxstbIf anyone is interested, I've uploaded a couple of 320x240 videos to
00:48:07Bagder 4 2036 linus
00:48:18Bagder 5 1724 dave
00:48:21LinusNwee, i climbed one place!
00:48:46 Quit mantono`afk (Client Quit)
00:49:00Bagder67 committers
00:50:06Bagderin total 25345 changes
00:50:13 Quit Xerion (" ")
00:50:14XavierGr[01:42] <Bagder> grep author: dump | cut "-d " -f7 | tr -d ';' | sort | uniq -c | sort -rn <−−- Bagder did you managed to invent that command now?
00:50:22Bagderoh yes
00:50:34Bagdershells are your friend
00:51:03Bagderand maybe just maybe I've done some similar command lines in the past ;-)
00:51:06XavierGrdamn! You are magnificent on scripting! :D
00:51:34LinusNkindergarten stuff
00:51:52Bagder | cat -n is useful to at the end
00:52:19LinusNthis is what makes unix so nice
00:52:54preglowis audio shoved in interleaved format to the dac?
00:52:56 Join Jungti1234 [0] (n=jungti12@
00:52:57preglowdoes it vary, perhaps?
00:53:04Jungti1234hi all
00:53:19LinusNthe coldfire dac is interleaved
00:53:27linuxstbpreglow: It is for the ipods, but as you wrote the FIQ, I guess you know that...
00:53:39preglowlinuxstb: long since forgotten, heh
00:53:54XavierGroh forgot to mention my weird experience today. I was listening to some tunes on my H300 (with H100 remote), and suddenly the lcd alignment in the remote was cahnged. The status bar was placed in the middle of the screen without any apparent reason. The H300 main display was okay though
00:54:12preglowafter doing some assembly, i'm going back on wanting to make everything deinterleaved
00:54:18*petur adds 14 to his current commit number
00:54:25XavierGrThough, I couldn't reproduce that again
00:54:27preglowbut having all the codecs output interleaved data is probably not going to be easy
00:55:00Jungti1234it's rockbox ipod bootloader?
00:55:30Bagderpetur: I think that'll make you climb one step to #34! ;-)
00:55:33 Join Paul_The_Nerd [0] (
00:55:34 Quit barrywardell ()
00:55:50 Join CrazyMYKL [0] (n=rumors@
00:55:59*petur excuses himself for being non-active lately
00:56:29*preglow hopes to do another commit soon too
00:56:36amiconnLinusN: What, please, is this:
00:56:38Bagder 12 472 preglow
00:56:41XavierGrpetur: yeah, you own us an explanation why USB OTG isn't ready yet!!! :P
00:56:44amiconnif (tc.dirlevel == 0 && !strcmp(currdir,"/")) if (tc.dirlevel != 0 || strcmp(currdir,"/"))
00:56:49 Join barrywardell [0] (n=barrywar@
00:56:52*petur wonders if he'll need an oops commit and whether that will help him in the list :D
00:56:52amiconnDo you spot a difference?
00:57:12amiconnIt's your 'restoration' of TREE_EXIT behaviour...
00:57:17LinusNheh, wow
00:57:45amiconnAnd I searched for mistakes in my action tables....
00:57:58Jungti1234anyone don't know?
00:58:09amiconnWill you fix?
00:58:46LinusNi didn't see that he had changed the == to !=
00:58:50LinusNi fix
00:59:03LinusNmy bad
00:59:59 Quit Daishi (Read error: 110 (Connection timed out))
01:00:12LinusNthe condition was inverted, but i failed to see it
01:01:03peturhmmmm no extra commit needed, must learn to write more bugs
01:01:21 Quit TeaSeaLancs ("Leaving")
01:01:27Bagderor just plain more ;-)
01:02:18*petur sends real life devils to Bagders' house
01:03:21LinusNBagder: say "aaaaaah"
01:03:48Mikachudid you forget to turn off the precognition?
01:04:12 Quit lodesi ("leaving")
01:04:45 Quit apo` ("Leaving")
01:04:50 Join apo [0] (
01:06:25LinusN"Some button mappings have changed and other things may break" - yeah right :-)
01:07:00petur"other things may keep working" :)
01:07:03 Quit idnar (Nick collision from services.)
01:07:06 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
01:07:10linuxstbPlugins are safe...
01:07:22 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
01:07:28LinusNwe'll hunt them down and destroy them too
01:09:07*petur thinks about committing AGC for iriver only for now - other targets with recording can add it by defining HAVE_AGC
01:09:19LinusNgood idea
01:09:31peturbut for tomorrow...
01:09:42LinusNiaudio x5 should be identical, right?
01:09:56*linuxstb thinks he should try and finish ipod recording and commit that at some point...
01:09:57peturdon't know about its gain steps
01:09:58*preglow shoots himself
01:10:31*LinusN wipes preglow's brain off the floor
01:10:44preglowmy suicide rate went up drastically after i started picking at assembler again
01:11:02barrywardelli have made a keymap file for the H10 if someone has a chance to add it:
01:11:17preglowbarrywardell: how's the rma going?
01:11:20peturjvo wrote a comment that AGC requires 0.5dB gain steps but I fail to see why not that I understand all of the AGC code :/
01:11:25***Saving seen data "./dancer.seen"
01:11:29barrywardellstill waiting
01:11:47barrywardellthinking about ringing or emailing them about it but afraid they might ask too many questions
01:11:53peturanyway, the AGC patch is up to date if anybody cares to try
01:12:12*petur stops talking to himself
01:12:22*LinusN should get some sleep
01:12:34*petur was thinking the same
01:12:57LinusNthat i should get some sleep?
01:13:05peturyeah :)
01:13:11peturand me too ;)
01:13:16*preglow just replaced the crossfeed with a squarewave generator
01:13:18 Quit SereR0kR ()
01:13:21XavierGrbarrywardell: Are you waiting for confirmation or you just wait for a new unit?
01:13:34 Quit Jungti1234 ()
01:13:38LinusNpreglow: coooool
01:13:58preglowand boost factor went significantly down
01:14:00barrywardelli got confirmation that they received it on aug 3rd
01:14:10Mode"#rockbox -o preglow " by preglow (n=thomjoha@rockbox/developer/preglow)
01:14:15XavierGrthe RMA took 2 weeks for me
01:14:20XavierGr(on both occations)
01:14:35XavierGrok maybe 3 on the second attemp
01:14:45preglowfrom 65% boost to 50% boost! woohoo!
01:14:56LinusNpreglow: way to go
01:14:59XavierGririver's RMA used to be good
01:15:03barrywardellah, only took 8 days for my brother
01:15:07XavierGrbut know it is very unreliable
01:15:08barrywardellthat was in may
01:15:19amiconnLinusN, as you have a H300 remote: What does this mean:
01:15:22XavierGrmy experience sucked
01:15:38*petur can write a square wave generator that boosts less :p
01:15:39*barrywardell needs to be more patient
01:15:40amiconnlong Rew should be cancel?
01:15:46LinusNamiconn: odd
01:15:48amiconn(default context for H300 remote)
01:16:31amiconnShould I take that out?
01:16:56amiconnI am currently going through all contexts removing the preconditions for the multi-fire repeat events
01:17:19LinusNthe h300 remote has a stop button, so i guess that should be used
01:17:34amiconnIt is
01:17:39amiconnCancel is mapped 3 times
01:17:56LinusNthen remove rew|repeat
01:18:14amiconnIt's mapped to Stop and Rew
01:18:14preglowamiconn: do you know anything about if interleaving str/ldr with other instructions will increase throughput on arm?
01:18:51amiconnpreglow: I have no idea. That big reference book (4 times as big as the coldfire programming manual) doesn't say anything about execution times
01:19:00LinusNamiconn: is it perhaps to be able to hold Rew to quit all menus in one stroke?
01:19:21amiconnAh, possibly
01:19:42amiconnSO no precondition then
01:19:55preglowamiconn: nah, some docs on the arm site does, though
01:20:01preglowamiconn: more specifically the arm7tdmi specific doc
01:20:20preglowbut it says nothing about that
01:21:02LinusNpreglow: interleaving should be efficient if you run in iram/cache and have a half-decent bus controller
01:21:19Mikachucan i trouble someone to add two lines to the ipod keymap?
01:21:21LinusN(which coldfire doesn't have :-)
01:21:41*LinusN looks away
01:22:41Mikachumaybe you could just remove the prereq instead
01:22:50dionoeawas the keymap thing commited today ?
01:22:56LinusNMikachu: that will be solved with amiconn's upcoming keymap revamping
01:22:58Mikachui don't see how a scroll|repeat could be produced with anything else than a scroll coming before it
01:23:09LinusNdionoea: yes
01:23:15Mikachuthere is a keymap revamp in progress?
01:23:30LinusNyes, which removes the prereq for most repeated events
01:23:35dionoeaneat :) does that mean that the plugins can start using it ?
01:23:42LinusNdionoea: soon
01:24:10amiconnIn fact it removes prerequisites for all events which trigger on repeat and are meant to fire multiple times
01:24:33amiconnIf an event should trigger on long press but should only fire once, the prerequisite is simply kept
01:24:43amiconnAlso simplifies the action code
01:25:19*amiconn hopes that he didn't overlook some of those events
01:25:28 Quit apo (Read error: 104 (Connection reset by peer))
01:25:44amiconnI also made all the tables const, in order to save ram when running from flash
01:26:22 Join apo [0] (
01:26:38amiconnI'll test on h140 and ondio (which also got decent mappins) then commit
01:27:36preglowinteresting... crossfeed crashes with a data abort whenever i try to modify settings
01:27:42preglowgod knows how that is possible
01:28:10*amiconn likes the fact that arm also doesn't like unaligned data
01:28:23amiconnThis way good old archos isn't the only one...
01:28:35preglowmost risc doesn't like unaligned data
01:28:44LinusNwhat would rockbox be if it didn't crash once in a while :-)
01:29:40preglowwhat the hell
01:29:45 Join mirak [0] (
01:29:53preglowit crashes on an address doing a load that should be address invariant
01:30:27LinusNpreglow: i guess you have trashed a register then
01:30:39petursomething I noticed again while debugging: every n times you connect/disconnect usb, my h340 stops being recognized by windows (not even as usb device). A simple rolo or reset of the player solves this. So after usb disconnect, something isn't set corrcetly that is set ok on boot
01:30:40preglowfrom the settings screen???
01:30:51 Join Davide-NYC [0] (
01:31:17mirakI let my fridge unpowered and closed during 1 week ...
01:31:29amiconnpetur: That also happens on H1x0. I guess one thread doesn't reply to th eusb connect message properly
01:31:47preglowmirak: at least it was closed
01:31:51peturthe usb signal comes on screen
01:31:57peturlogo I mean
01:32:22peturah, that doesn't mean everybody acked it...
01:32:31amiconnI know this behaviour. I guess it will also happen on X5
01:32:39preglowlinux is very happy when that happens...
01:33:00amiconnI observed a similar effect on ipod too, even though we don't handle usb in rockbox
01:33:11*petur is off to bed - 'night
01:33:17*LinusN too
01:33:17 Quit petur ("Zzzz")
01:33:18mirakpreglow: in fact after 3 days I realised I forgot bananas in the fruit case inside the fridge, so I called the janitor and asked him to remove them. I am not sure if the horrible moisi appeared after that he opened
01:33:20LinusNnite all
01:33:31amiconnIf it happens, the ipod shows the usb screen like the other targets, and won't get recognised by the pc
01:33:37 Part LinusN
01:33:53mirakit took me one hour and a half to wash it lol
01:34:48 Quit jsMikeS (Read error: 104 (Connection reset by peer))
01:37:28 Join Xerion [0] (
01:40:40Damme_hmm, what is good settings for the ipod scroll wheel ? it jumps really wierd for me then I'm scrolling items longer then one page...
01:40:55 Join jhMikeS [0] (
01:41:00 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
01:42:40preglowi don't get this, the setting doesn't even have to be related...
01:43:13preglowentering any settings screen with a numerical list crashes crossfeed
01:44:41*preglow hopes this is someone else's bug
01:44:50*linuxstb blames jdGordon
01:45:45linuxstbDo you mean entering a settings screen whilst crossfeed is active?
01:48:06preglownot with cvs build, though
01:50:49 Quit midkay ("Leaving")
01:52:56stripwaxDamme_ - I don't know either. I actually turn that off because basically, as far as I can tell, it doesn't work properly
01:53:10stripwax(the scroll wheel acceleration)
01:53:26Damme_I got it working I think
01:53:50Damme_fast : 150 faster 425 fastest 500
01:54:06Damme_it was something likt 50 100 150 before .. that didn't work at all
01:54:32stripwaxDamme_ - so, are the defaults just useless? if so that should be changed
01:54:36 Join roxxor [0] (
01:54:38 Join midkay [0] (n=midkay@rockbox/developer/midkay)
01:55:10roxxorhi.where can i find a exact list of supported archos models? does it work on gmini xs 100?
01:55:12preglowlinuxstb: ahh, caught that bug
01:55:23preglowlinuxstb: turns out i'm dead drunk
01:55:49Damme_stripwax yeah, I'm guessing that, tried that settings in playback and pause-mode, and it seams to work most of the times now
01:56:26jhMikeShow long before changes show up on the main page?
01:57:30linuxstbroxxor: lists the working Archos devices - just the old Player/Recorder/FM Recorder and Ondio. None of the gminis.
01:57:59jhMikeSIt's got a build in progress if that delays that.
01:58:07 Quit barrywardell ()
01:58:20roxxorhm. will it be finished soon?
01:58:22linuxstbI think the main page updates after the build is finished.
01:58:52roxxorCan I expect to use rockbox on a gmini in, e.g. 1 year?
01:59:07linuxstbroxxor: No-one is working on it.
01:59:20roxxorhm, that's bad
01:59:38preglowi'd say that unless jyp returns, there's no chance that port will ever happen
01:59:57roxxorI don't know rockbox that much. The page mostly says it is good, but how can i know?
02:00:02roxxorWhy do you use rockbox?
02:01:00linuxstbEveryone has different reasons.
02:01:02preglowbecause it's heaps better than the usual run of the mill firmware, tons more codecs and gapless playback does it for me
02:01:35jhMikeSwhy...oh why?!?!? Why did my sister's Christmas present have to come to this!
02:01:49linuxstbNow I'm building up a collection of MP3 players, it's also very nice to have the same firmware on them all...
02:02:18roxxorthat makes no sense to me
02:02:19*preglow feels its time for a new mp3 player
02:02:29roxxorwell, thanks anyway
02:02:31roxxorgood night
02:02:34preglowi can find nothing i want, though
02:02:43 Quit roxxor ("blah")
02:03:05linuxstbpreglow: I've had the same problem. I'm not sure my DAB/MP3 player has satisfied me...
02:03:59preglowdab would be a nice feature
02:04:36linuxstbPotentially - if the quality of DAB was better. Any idea what bitrates will be used in Norway?
02:04:43preglowno idea
02:04:50preglowit's still in the initial stages of being rolled out
02:05:17preglowthere's some spotty coverage here and there, but i don't know what they use
02:05:26jhMikeSI take it a new build starts whenever you committ something?
02:06:25preglowthe council that deals with these matters have even managed to recommend both dab and drm
02:06:34preglowdrm being recommended for smaller local stations
02:06:58linuxstbAccording to that page, you've had DAB since February 1999...
02:07:16preglowspotty coverage, like i said
02:07:33preglowyou stumble across it here and there
02:07:34linuxstbAnd sadly, similar 128kbps stations to the UK.
02:07:52linuxstbAlthough a fair amount of higher ones..
02:09:16preglowi think they were closed down
02:09:25preglowdue to broadcasting audio books without permission
02:09:37Mikachusmart business idea
02:09:43amiconnjhMikeS: Basically each commit starts a new build, with a few exceptions and conditions
02:10:10jhMikeSexceptions? conditions?
02:10:21preglowlinuxstb: you know what the highesst permitted bitrate is?
02:10:54amiconn(1) Commits concerning the manual don't trigger. (2) If a build is already running, all commits done while that first build runs will all be included in the next build round
02:11:29jhMikeSWell, I did something and it sure seems like a loooooong 4:58sec
02:11:48amiconn(3) After a build ends, there's some latency until a new one starts. Several minutes; don't know why that is
02:12:20linuxstbpreglow: No, probably whatever the MP2 maximum is. There are some 256kbps stations in France.
02:12:39jhMikeSWhat the hell machine can do a build in 6.6 seconds?
02:12:56amiconnjhMikeS: Hanging build, that
02:13:14amiconnAnd no machine builds an ordinary rockbox build in 6 seconds
02:13:27amiconnBootloaders can be built that fast, with ccache
02:13:36Mikachuaterm builds in 6 seconds :)
02:13:37amiconnThe build system is distributed
02:13:37preglowwoohoo, i got to use the swp instruction!
02:14:24*Mikachu odset ehs aw padcn
02:14:56amiconnEither the build hangs, or it just takes ages because the central server is perhaps heavily loaded
02:16:19preglowi think its stuck
02:16:44*linuxstb goes to bed
02:16:55preglowforget it, wrong box
02:16:56jhMikeSlinuxstb: g'night
02:17:54*preglow does the same
02:18:12jhMikeSnight night
02:18:35amiconnjhMikeS: The number to the far right of the build table rows (with the link) is the number of machines which ran this build round
02:18:47 Join Senphet [0] (
02:19:04amiconnIf you follow the link you'll see which machines built how many of the targets
02:19:40jhMikeSI think I broke it! )
02:22:34jhMikeShmmm...I changed a couple files that compiled fine, I saw it had started a new build and I've been waiting about 1/2 hour
02:23:33Senpheton the playlist menu, what's the difference between queue and insert?
02:23:55safetydanSenphet, queue will add the song to the playlist and then remove it when it's played.
02:24:02safetydaninsert adds it permanently
02:24:18Senphetahh, thanks very much
02:24:23Senphetnice feature
02:25:00jhMikeSIt shows that it was my commit that started it. wtf
02:27:39ShadowdogMUOk, off topic here but I want your opinions
02:27:51ShadowdogMULike? I think I am getting them
02:28:20safetydanumm... there shoes...
02:28:24safetydanthey're even
02:29:17ShadowdogMUI know, I have a shoe fetish
02:29:19ShadowdogMUdo you like them?
02:29:37*safetydan is not a shoe person
02:29:40Paul_The_NerdThere's really only two things that matter with shoes: 1) Are they comfortable enough to warrant the price? 2) Is someone likely to shoot me because they want to steal them? If the answers are yes and no (and it's a given that you like how they look if you're considering them at all) then buy them.
02:30:03 Join webguest58 [0] (
02:30:06*jhMikeS says "way to go" for a first go
02:30:30webguest58eeprom driver for the h3x0 series, cleaned up the h1x0 series driver a bit, added debug entry for h3x0 that allows to write to the eeprom
02:30:43safetydanBuild page really is horked.
02:30:48webguest58hello...what this means for H3xx owners +
02:30:52ShadowdogMUPaul_The_Nerd, Pumas are uber comfy, and those are custom made, $130
02:30:58webguest58eeprom driver for the h3x0 series, cleaned up the h1x0 series driver a bit, added debug entry for h3x0 that allows to write to the eeprom?
02:31:01safetydanwebguest58, not much right now
02:31:07Paul_The_NerdShadowdogMU: Seriously. Get them if you like them. This is entirely the wrong place for asking opinions on shoes
02:31:14Paul_The_Nerdwebguest58: You don't need to repeat yourself.
02:31:18webguest58what is the target for the eeprom driver ?
02:31:21ShadowdogMUHaha, I just saw active people :P
02:31:24webguest58what will it do ?
02:31:56Paul_The_NerdIt will possibly allow running Rockbox from flash, as well as figuring out the bug that can cause Rockbox to change the iRiver firmware's settings.
02:32:37jhMikeSCVS says: "cvs commit: Using deprecated info format strings. Convert your scripts to use the new argument format and remove '1's from you info file format strings."
02:33:47webguest58what about recovering my frezed iriver firmware ?
02:33:54*jhMikeS is gonna get some blueberry pie and hopes somebody will tell him what he horked.
02:34:08webguest58can it write an correct iriver firmware if broken ?
02:34:42webguest58I don't want to run Rockbox from flash if the original firmware get lost forever. that is nutz
02:35:46Paul_The_Nerdwebguest58: How is it nuts? Some people don't use the iRiver firmware, ever.
02:36:14Paul_The_NerdYou don't *have* to to do it.
02:37:36 Quit Noah0504 ("Ex-Chat")
02:37:36webguest58what if Rockbox doesn't get Video support ever ?
02:37:48Mikachutoo late for that
02:37:54kidd3ckzi'll cry
02:38:02webguest58No video again on my H320 since I put a Rockbox build
02:38:11Paul_The_Nerdwebguest58: Not everyone bought a dap for video support. Besides, as I said, you don't have to use it. Also, Rockbox already has the beginning of video support.
02:38:19webguest58Rockbox nuked my iriver firmware
02:38:25Mikachuno, you did
02:38:39webguest58yeah. I know. Just begining. I want more.
02:38:47kidd3ckzso donate
02:38:50webguest58I'm not blasting Rockbox..
02:38:56Paul_The_NerdHow did Rockbox nuke your iRiver firmware?
02:38:57webguest58donate ?
02:39:11webguest58I was thinking Rockbox is open source
02:39:38Paul_The_Nerddonate means "give money and receive nothing in return"
02:39:43Paul_The_NerdWhat does that have to do with open source at all?
02:40:00kidd3ckzu'd donate if u want to help the efforts of the developers
02:40:13kidd3ckzso that they may afford better beer and richer coffee
02:40:58webguest58I say Thanks. and they should appreciate my thanks.
02:41:14webguest58That is enough on many parts of the worls
02:41:26Mikachuyou can't buy food with thanks
02:41:32kidd3ckzi was gonna say beer
02:41:35Paul_The_NerdYou can't replace hardware lost in experimentation with thanks.
02:41:37webguest58but Rockbox is not food..
02:41:40webguest58WTF !!
02:41:59jhMikeSIs noone talking to me now? :( I did everything by the book and would like to know why the build is stuck.
02:42:02Paul_The_NerdDonations are valuable. You aren't required to ever.
02:42:07Paul_The_NerdBut they are nice.
02:42:18MikachujhMikeS: i'm no expert on the build system but i know it hangs like this once in a while for no reason
02:42:28Paul_The_Nerdwebguest58: But you never answered *my* question which was: How did Rockbox nuke your iRiver firmware?
02:42:35Paul_The_NerdjhMikeS: It's probably not your fault.
02:42:51jhMikeSMikachu: It of course has to hang on my first commit after just being granted access. It bumming me out.
02:43:05webguest58paul: installing Kosh build..
02:43:09amiconnjhMikeS: It's stuck because... it's stuck
02:43:15Paul_The_Nerdwebguest58: That's not Rockbox.
02:43:20Paul_The_Nerdwebguest58: That's Kosh's build.
02:43:21amiconnNo special reason depending on what you did
02:43:33amiconnBagder, Zagor or LinusN need to fix it
02:43:37webguest58paul: using normal Rockbox. Still stuck.
02:43:47webguest58Freezed on loading screen.
02:44:01Paul_The_Nerdwebguest58: You expect Rockbox to magically fix errors that something else created if it nuked your original firmware?
02:44:07webguest58HDD make noise and then stop. Stuck on colorful loading screen forever
02:44:22Paul_The_NerdWhat colorful loading screen?
02:44:33webguest58nope, but if it goes to flash I could get a brick from any new build
02:44:37jhMikeSBeginner's bad luck eh...
02:44:50webguest581.29K LOADING...
02:44:51*jhMikeS thinks he should change his name to "Flypaper"
02:44:54Paul_The_Nerdwebguest58: A) You load a build normally before flashing it. You can't flash without already being in the build.
02:45:03webguest58that loading screen
02:45:10webguest58I did that paul
02:45:13webguest58was working fine...
02:45:17Paul_The_NerdDid what?
02:45:26Paul_The_NerdYou can't flash Rockbox yet, only the bootloader.
02:45:34 Join ranunculoid [0] (n=eamon@
02:45:40Paul_The_NerdSo, how can you have "Done it"?
02:45:41webguest58but I installed one or two builds with AA support. Now stuck.
02:45:51webguest58Im using Rockbox right now
02:45:57Paul_The_NerdYes, from disk
02:46:03webguest58Can't boot back to iriver..
02:46:09webguest58yes from disk.
02:46:19webguest58battery ?
02:46:30Paul_The_Nerdwebguest58: Either way, Rockbox didn't cause your problem.
02:46:35Mikachueven if a bot responded to that, what would it do?
02:46:35Paul_The_NerdPlaying around with experimental builds did.
02:46:39ranunculoidhrm, no bot. What's the battery life like on Rockbox for iPods compared to the original firmware?
02:46:46Paul_The_Nerdranunculoid: Check the wiki.
02:46:50Paul_The_NerdIpodRuntime, I believe
02:46:58jhMikeSamiconn: what's the stuff about deprecated format strings?
02:46:58Mikachuranunculoid: less
02:47:16ranunculoidMikachu: It would answer my frequently asked question :)
02:47:20amiconnjhMikeS: Ignore it. Configuration stuff on the cvs server
02:47:28webguest58paul: then what ?
02:47:29webguest58me ?
02:47:44Paul_The_Nerdwebguest58: I said "Playing around with experimental builds did."
02:47:49webguest58BTW, I'm not noob. Following all the instructions on here and
02:47:57Paul_The_NerdMistircriver is entirely unofficial
02:48:01Paul_The_NerdThe experimental builds are unofficial
02:48:17 Join pike^lapt [0] (
02:48:23webguest58thanks for the correction.
02:48:25Paul_The_NerdThey're builds made by other people
02:48:27ranunculoidPaul_The_Nerd: Yeah, I looked there but that hasn't been updated in a while. I was wondering if there'd been any battery life optimisations done in the last few months?
02:48:36webguest58then the mpgplayer plugin is not official right ?
02:48:40Paul_The_Nerdranunculoid: You could run a benchmark and submit it there.
02:48:41*jhMikeS says from the corn: "build it and it will crash"
02:48:44webguest58so, I should not expect video.
02:48:46Paul_The_Nerdwebguest58: No, it's official
02:48:51Paul_The_Nerdwebguest58: We had it before they did even.
02:49:07webguest58why it got removed then ?
02:49:09Paul_The_NerdBecause it was official, and it took them a few hours, or a day, to update their sources with the new code.
02:49:11Paul_The_NerdIt's not removed...
02:49:15safetydanwebguest58, it hasn't been removed
02:49:27webguest58you sure ?
02:49:29ranunculoidPaul_The_Nerd: Ok I'll do that :) I'm glad to contribute in some way since I can't code :)
02:49:34webguest58are you / anyone updating it ?
02:49:42Paul_The_Nerdwebguest58: It's less than a week old.
02:49:50 Quit pike ()
02:49:54Paul_The_Nerdwebguest58: It hasn't been abandoned.
02:49:58webguest58seriously. Ok. I cant wait 1 year
02:50:02webguest58Rockbox is cool.
02:50:03 Join pike [0] (
02:50:27webguest58AA and picture browsing let me forget about battery draining Videos on iriver firmware
02:50:45webguest58I should have said. I CAN WAIT 1 YEAR FOR VIDEO ON RB
02:50:51 Join Febs [0] (
02:50:52jhMikeSranunculoid: I can code but I can't commit...and I added an additional layer of meaning to that today. !
02:51:22*safetydan wonders what AA in rockbox is
02:51:26Mikachualbum art
02:51:29safetydanall I can think of is anti-aliasing
02:51:30ranunculoidjhMikeS: lol :)
02:51:59webguest58Jewel is crack, BTW
02:52:10webguest58my GF play all night with it
02:52:20ranunculoidrockbox will probably never use anti-aliased fonts safetydan :(
02:52:36jhMikeSnever, ever, never never, ever say "never"
02:52:46 Quit stripwax (Connection timed out)
02:52:53ranunculoidI said "probably never" ;)
02:53:01Paul_The_NerdI'm sure it'll show up in some experimental build
02:53:19jhMikeSone of mine I'm thinking
02:53:23Paul_The_NerdAnd then people will ask us why it doesn't get included, and we'll say "Remember how you can't use the equalizer in the experimental build, because the music skips? That's why"
02:54:00ranunculoidI'd prefer aa'd fonts to an eq
02:54:02jhMikeSYou won't even ever see it unless I get it to work right.
02:54:10jhMikeSI mean if I happen to do it
02:54:31 Quit apo (Remote closed the connection)
02:54:59 Join apo [0] (
02:55:01webguest58When 3 is gonna be released ?
02:55:05jhMikeSBetter yet, vector fonts.
02:55:17ranunculoidwebguest58: 2013
02:56:21Paul_The_Nerdranunculoid: I'd prefer battery life to AA'ed fonts.
02:56:31Paul_The_NerdI'd also prefer that your AA'd fonts don't take up MY ram.
02:56:57Paul_The_NerdThe code to deal with them will be in RAM no matter what. So addition of it very slightly hurts everyone who doesn't want it.
02:57:05ranunculoidPaul_The_Nerd: I bet you dont even use pretty WPSs :P
02:57:13Paul_The_NerdIt's an AUDIO player
02:57:18Paul_The_NerdI look at the screen once an hour, maybe.
02:57:36jhMikeSUse bitmap or vector fonts...the bm fonts will work just the same if it's done right.
02:58:06jhMikeSIt should be a very very _pretty_ audio player
02:58:15Mikachuvector fonts even has noticable speed impact on a pc
02:58:16Paul_The_Nerdwebguest58: There are no release dates at all, here. An estimation may be made at some point, if it's ever certain.
02:58:34jhMikeSThen rasterize them in advance
02:58:42 Join finn__ [0] (
02:58:46Mikachuagain with the ram
02:59:06 Part pixelma
02:59:26Paul_The_NerdYup RAM buffer for >1bit fonts, + more ram for code to deal with translucency, + slower text processing than we already have == hit on battery life.
02:59:54ranunculoidMikachu: What's your ram limit? Cause my iPod has 64 megs, isn't that loads?
03:00:14Mikachunot really
03:00:29Mikachuevery piece of mem you use counts towards spinning up the hard drive more often
03:00:32Paul_The_Nerdranunculoid: Remember that any of that RAM being used for code or fonts is NOT being used for audio buffer, and so lessens battery life.
03:00:34webguest58later guyz
03:00:40jhMikeSDon't the BM fonts take any RAM? Except they're monochome...they'd only have to be 5 or 6 bit tops. Less if you don't want every possible level at the edges.
03:00:57Paul_The_NerdjhMikeS: They're currently 1-bit. So, 5-bit is 5 times the space.
03:01:29Mikachueven 2-bit allows 4 levels, which could be enough
03:01:46Paul_The_NerdThat breaks font compatibility between targets, among other things
03:01:57 Part ranunculoid ("Kopete 0.11.92 (0.12 Beta 1) :")
03:02:04 Quit webguest58 ("CGI:IRC (EOF)")
03:02:09Mikachubut then every pixel you draw will be a read + mult + write instead of just write
03:02:15jhMikeSMaybe there's a something spinning in my head
03:02:37jhMikeSDoesn't the bg get redrawn when you draw a string in most cases?
03:03:10Paul_The_NerdIdeally only a portion of it where the string is.
03:03:31Mikachuit's memcpy'd into the framebuffer i think
03:03:34jhMikeSyes, unless you fg draw mode
03:03:46jhMikeS"'s fg..."
03:04:01 Join ranunculoid [0] (n=eamon@
03:04:58amiconnThe fonts can never be memcpy'd except in one situation:
03:05:14Mikachui meant the bg pixmap
03:05:23jhMikeSHow can it be done without deeper font color depths/vectors? I've got a feeling theres a way.
03:05:37amiconnForget about anti aliasing
03:05:45jhMikeSI haven't look yet so tell me: how do they get drawn exactly.
03:05:59amiconnThe unifont is already several megs - that's monochrome!
03:06:03jhMikeSOr kindly give me the file name.
03:06:03Mikachudo subpixel rendering while you're at it
03:06:23jhMikeSIs an entire font loaded in advance?
03:06:33amiconnNo, it uses an lru cache
03:06:50jhMikeSnot mru?
03:06:56amiconnOnly if the whole font fits into the buffer, it is loaded at once
03:07:22ranunculoidThis seems quite difficult. I mean, they're not even anti-aliased in the Apple OS. I's just a 4-bit font that fades to white at the edges. But what do I know :P
03:08:03amiconnjhMikeS: The fonts are essentially monochrome bitmaps. So the various lcd_puts* functions in turn use lcd_mono_bitmap_part()
03:08:17 Quit pike^lapt (Read error: 110 (Connection timed out))
03:08:19jhMikeSIf they're rasterized for say a wps, there's usually not a lot of chars there simultaneously.
03:09:00Mikachumy wps screen is full of text
03:09:15jhMikeSBut a directory listing...well...even that's limited at least in roman, cyrillic, etc.
03:09:31jhMikeSMikachu: How many different characters are on there at once?
03:09:36amiconnBut we support chinese, korean, japanese, arabic...
03:10:52 Join Snyper [0] (
03:11:07jhMikeSMikachu: Many appear more than once so there like a dozen?
03:11:07 Join TCK [0] (
03:11:24Snyperdoes anyone in here use an IDE? (other than dev studio?)
03:11:27***Saving seen data "./dancer.seen"
03:11:36jhMikeSPlus the numbers and such
03:11:39Snypernot necessarily for rockbox, but an IDE in general
03:11:55Mikachubut each new song could have completely new chars
03:11:58Mikachucould make skipping much slower
03:12:08Paul_The_NerdjhMikeS: You can't make the assumption that any character will appear more than once though. I mean, several may or will, but you could have an english WPS with a japanese song playing an an arabic one in the next-song info, leading to a wide variety of characters.
03:12:17jhMikeSTrue, dump the ones not in the new song.
03:12:38amiconnjhMikeS: An important point: We're buffering several tracks in RAM during playback, and we do that for a reason - to avoid spinups. Spinning up the disk to fetch missing glyphs for a new track isn't nice
03:12:49jhMikeSPaul_The_Nerd: No, I'm just playing statistics...
03:13:35jhMikeSIt's spins up for the track. I mean are these things gonna take megabytes or a few 10s or kilobytes?
03:14:01amiconnIt spins up once for severa tracks
03:14:20 Quit finn_ (Read error: 110 (Connection timed out))
03:14:21Paul_The_NerdYou can have 10+ tracks in a buffer, easy.
03:14:44amiconn(depending on the target's ram size and the track length & bitrate of course)
03:14:50Paul_The_NerdOf course
03:14:59Paul_The_NerdI mean, with SID, I could have a couple hundred.
03:15:06Paul_The_NerdExcept I think it's locked to a max of 32?
03:15:06jhMikeSamiconn: I can see that...I've got a good test for the gapless: Fingertips by They Might Be Giants...a bunch of stuff stuck together that's 4-30 sec long w/0 gaps.
03:15:21Mikachuhm, with sid, would it store a copy of the sid codec in front of every song?
03:15:25Mikachuor did i dream that?
03:16:23jhMikeSDo most song played in a row tend to have more similar sets of chars or different? I say similar and extreme variety is the relative rarity.
03:16:46Paul_The_NerdBut you can't rule out that group, just because it's convenient or unlikely
03:18:37 Join finn_ [0] (
03:18:42jhMikeSIt's not ruled out. They're "habits" are potentially more costly to begin with in terms of ram. It shouldn't always be geared to make the worst case ideal at a great expense of the far more likely expected cases.
03:19:33Paul_The_NerdThe question is, will it improve the best case without making the worst case any worse?
03:19:43Paul_The_NerdAnd, how much does it cost?
03:19:49jhMikeSIt's like making half the parking spots handicap spaces...just cause some people have special needs.
03:20:02jhMikeSAbout $100
03:20:15Paul_The_NerdI meant "does it take more RAM, processing power, or both, and how much?"
03:20:43amiconnIf you want to slow down text drawing to a crawl, try antialisaing / alpha blending
03:20:44hotwire_does anyone know which file assigns the keys in the UISimulator?
03:20:47jhMikeSTo really evaluate that, it must be tried...we know 5bit fonts take 5 times as much as 1bit
03:20:52 Join JdG|uni [0] (
03:20:58JdG|unihey all
03:21:01jhMikeSamiconn: ok lmao
03:21:19Paul_The_NerdjhMikeS: The key is that it needs to not be noticeable slower. We already experience problems because of slow text drawing.
03:21:22*JdG|uni is catching up on yesterdays irc log
03:21:39amiconnNo, seriously
03:22:33jhMikeSWould that have to be a feature on EVERY player though? Some it obviously can't be but others? When they're finally powerful enough and a ports is available then I guess such things will be more likely to happen.
03:22:37JdG|uniFebs: you wanting to add the actions stuff straight into the manul? sounds good, and if your lucky you can do a script to convert it straight into tex ?
03:23:15amiconnOn a related note, transposing the framebuffer to exploit HWM would allow for significant speedup of mono bitmap drawing
03:23:24jhMikeSPaul_The_Nerd: Text hauls ass on the x5. The scrolling zips around just fine.
03:23:36amiconnThe scrolling is _slow_ on X5
03:24:13jhMikeSamiconn: True, who's gonna make sure all the plugins that do framebuffer access work too? How much is already done in that regard for other players?
03:24:28jhMikeSamiconn: not now...
03:24:36amiconnIt is
03:24:46jhMikeSIt seems to be a time thing...
03:24:51 Join sharpe [0] (
03:24:51jhMikeSamiconn: since when?
03:25:00Paul_The_NerdjhMikeS: Even if it's not a feature on every player, it either breaks cross-player font compatibility, or it requires extra CPU load to do the antialiasing on the 1-bit fonts.
03:25:13Paul_The_NerdjhMikeS: How many other Rockbox players have you scrolled on?
03:25:36jhMikeSPaul_The_Nerd: No other players. But the OF is a friggin dog.
03:25:47Paul_The_NerdBut we don't compare to the OF.
03:26:00Paul_The_Nerd"Faster than the OF" is not fast.
03:26:01jhMikeSThat's all I got to go by atm.
03:26:03Paul_The_NerdIt's just not as slow as the OF
03:26:07amiconnjhMikeS: Btw, I just found where I found the info that the LCS /CS must be kept low during HWM: The HD66789R datasheet has a timing diagram that shows it this way (page 103)
03:26:21jhMikeSamiconn: I'll check it out.
03:27:01amiconnIt might be that this is only an example and it also works without keeping /CS low
03:27:13amiconnThese datasheets are sometimes not very precise
03:27:15jhMikeSIt's probably 104 for me since it came from datasheetarchive :\
03:27:57jhMikeSI don't have an HD66789R its an HD66773R
03:28:29amiconnYes, but they're very similar
03:29:01 Quit Senphet ()
03:29:37amiconnCompare with HD66773 datasheet page 68
03:30:09jhMikeSThere already.
03:30:23*amiconn says good night
03:30:44jhMikeSThey probably think "why would you unselect the chip if you're gonna write a bunch of stuff to it anyway?"
03:30:48jhMikeSamiconn: gnight
03:31:53 Quit finn__ (Read error: 110 (Connection timed out))
03:32:09 Quit pike (Read error: 110 (Connection timed out))
03:33:25Snyperso yeah, now that its a bit quieter in here...does anyone use an IDE (for development on things other or including rockbox)?
03:34:08Snyper*(other than, or including rckbox ;) )
03:34:17jhMikeSI do
03:34:23Snyperwhat do you use
03:34:37 Quit JdG|uni ("CGI:IRC")
03:34:40SnyperI am now working at a place that mainly uses linux...
03:34:47jhMikeSI've just used DevStudio and Borland C++ way back
03:34:59Snyperyeah, I really like dev studio
03:35:09Snyperand its debugger
03:35:28SnyperI want to find something at least comparable...
03:35:34Mikachuisn't it expensive? ;)
03:35:44jhMikeSmy version OLD...
03:35:58jhMikeSThere is something
03:36:44Snyperyeah, it is expensive, but worth it, I wish they made it for linux....
03:37:30SnyperI dl'd it, went to test its debugger, and didnt have gdb installed, so its installing now...
03:37:33jhMikeSHow come the x5 scroll only got slow in the past couple months? It used to be much faster.
03:38:51jhMikeSI never went to far with it. I just use TextPad.
03:39:48 Join hotwire__ [0] (
03:46:13Mikachuwhat is the word on this one?
03:46:21Mikachu(it doesn't apply to cvs anymore)
03:46:35 Join finn__ [0] (
03:47:23 Quit hotwire_ (Read error: 60 (Operation timed out))
03:48:04Paul_The_NerdMikachu: Catch him and tell him to fix it. ;)
03:48:43Mikachuthat it doesn't apply wasn't the issue
03:48:48jhMikeSSpeaking of excessive refresh: Doesn't the wps update itself entirely every second? I noticed it when messing with the lcd_update(_rect) code. That seems unnescessary.
03:49:16MikachujhMikeS: not only that, every time it draws itself, it parses the wps file :)
03:49:45 Join webguest43 [0] (
03:51:13jhMikeSI noticed it uses a format string to draw numbers too, I believe. Why not just draw the numbers? It's easy.
03:52:24MikachuPaul_The_Nerd: you aren't tracker-empowered, are you?
03:52:34Paul_The_NerdI am in fact Tracker Empowered.
03:52:50Mikachuthen you can close
03:52:52 Join webguest86 [0] (
03:53:10Paul_The_NerdWhat reason should I put for it?
03:53:20Mikachuobsoleted by the action stuff
03:53:22jhMikeSI can
03:53:31jhMikeSor so they say
03:53:55 Quit TCK (Client Quit)
03:54:18Mikachunow i only have ~3 open patches in the tracker
03:54:18 Quit apo (Read error: 104 (Connection reset by peer))
03:55:15jhMikeSI'm gonna go reopen it! hahaha B-]
03:55:33Mikachuevil evil, impolite and eeeevil
03:55:47jhMikeSPower corrupts!
03:55:50Mikachui bet you don't get a lot of kids in the hall references here
03:55:57Paul_The_NerdOh, right. jhMikeS, do you have a forum name? You have commit access, right, which means I need to give you a developer badge.
03:56:19jhMikeSForum Name? I just use MikeS
03:56:38 Join apo [0] (
03:56:40 Join tommeyer [0] (
03:56:41sharpei use 'sharpe'!
03:56:47 Nick finn__ is now known as finn23 (
03:56:47Mikachui don't use the forum!
03:56:47jhMikeSI watched Kids in the Hall back when
03:57:15*Mikachu crushes jhMikeS' head
03:57:37 Nick hotwire__ is now known as hotwire_ (
03:57:39 Quit webguest43 ("CGI:IRC (EOF)")
03:57:51jhMikeSGive me something that says "power" and instills fear and respect in all who gaze upon its glory
03:58:18*jhMikeS 's head is hurting and wonders why he deserved that :(
03:58:44apojhMikeS, it's because you're an evil commie who uses evil commie software :<
03:58:49jhMikeSI have my Castro outfit out of mothballs
03:58:54 Quit webguest86 ("CGI:IRC")
03:59:30Paul_The_NerdDevelopers get a small amount of added power in the Forums, if I recall.
03:59:33*jhMikeS wonders if it was a big heavy piece of irony that caused all the pain
04:00:00Mikachuheeeey, the keyboard still uses hardcoded buttons, is that on purpose?
04:00:22 Quit finn_ (Connection timed out)
04:00:41jhMikeSMikachu: You mean the screen for entering filenames?
04:00:48Mikachuamong other things, yes
04:01:27jhMikeSSometimes it says "Play" for yes but I have to use the joystick instead.
04:01:41jhMikeSapo: Who's commie software?
04:01:49aporockbox :<
04:01:53Paul_The_NerdRockbox is communist
04:02:12 Join hotwire__ [0] (n=hotwire@
04:02:37jhMikeSIt is voluntary so it is not.
04:02:43Paul_The_NerdjhMikeS: You need to read the post
04:03:04Paul_The_NerdAnd especially LinusN's response to it.
04:03:21jhMikeSIs that one of those who thinks anything open souce where people volunteer their time _willingly_ is communist...ok I'll read
04:03:38Paul_The_NerdNo, it's just someone complaining.
04:03:59Paul_The_NerdApparently I wasn't polite enough for him.
04:04:00*Paul_The_Nerd is Llorean in the forums.
04:04:22sharpehey Paul_The_Nerd, have you seen the two posts about the same thing made within 34 seconds of each other, to two different forum categories? :-)
04:04:35Paul_The_Nerdsharpe: The ones Febs merged?
04:05:09jhMikeSThat person should use MP3Tag
04:05:24Paul_The_Nerdsharpe: Ah yes, that one. It was locked by someone handy, ;)
04:05:59jhMikeSHe doesn't know most cd ripping software has an option to put track numbers in front? heh
04:06:15sharpejhMikeS: only communists know of those options.
04:06:20Paul_The_NerdI try not to do the lockings and mergings myself. People already start saying "I wonder where Llorean is" if a question isn't answered quickly enough. ;)
04:06:21jhMikeSWhat was under the [CutOut]
04:06:27Paul_The_NerdJust some cursing
04:06:59sharpeyeah, but this question was answered in five and three minutes, respectively :)
04:07:02jhMikeSI like cursing when it's out of context. It's very revealing.]
04:07:39FebsI'm happy to merge and cut.
04:07:51Paul_The_NerdFebs: And I'm quite happy that you do.
04:07:58*sharpe can't merge and cut.
04:08:20jhMikeSI don't really patrol the forums too much...I don't wanna hear anything else about tagcache or my iPod won't work...maybe I'm just cynical/jaded.
04:08:59jhMikeSMaybe you're a communist 'cause you didn't bill him?
04:09:06sharpeGah. Blue text on red background hurts eyes.
04:09:32*Paul_The_Nerd upgrades Experts.
04:09:51Paul_The_NerdI think it's safe for experts to have move/merge abilities.
04:10:10Paul_The_NerdI don't believe you have "remove" abilities. Primarily just to prevent accidents.
04:10:20sharpehow many Experts are there though?
04:10:27Paul_The_NerdDevelopers get this too though
04:10:44jhMikeSPaul_The_Nerd: Just like the plastic bedding, huh?
04:10:56sharpeBut I never see more than two experts online at a time so i was wondering...
04:11:15Paul_The_Nerdsharpe: That's because a couple of my experts got promoted when I wasn't looking.
04:11:38Paul_The_NerdI don't like the idea of threads being deleted, just because they'll get pushed to the bottom but can still be found in searches. The only time I delete things is when someone modifies their own post, removing the text, and replacing it with "Please erase this"
04:11:38sharpeAh. must've been looking when the time came around for me. Haha.
04:12:42sharpeRemember, they don't ever capitalize 'Please.'
04:12:45Paul_The_NerdActually, we have three Rockbox experts at the moment.
04:12:46 Join Lynx [0] (
04:12:55FebsWell, spam gets removed too.
04:12:57Paul_The_NerdBut Febs is dual classed as Expert/Moderator.
04:13:03Paul_The_NerdWell spam doesn't qualify as "posts" ;)
04:13:03jhMikeSI have a rock box developer badge.
04:13:10*Febs feels special.
04:13:22Paul_The_NerdjhMikeS: I know. I'm quick about that.
04:13:52jhMikeSPaul_The_Nerd: Ok, you did that...I probably wouln't have noticed for awhile.
04:14:15jhMikeSDo other buttons show up when you have whatever access you have?
04:14:29sharpePaul_The_Nerd: You should have a notification-type thing, whenever a post is made without any coherentness to it...
04:14:35Mikachucan you make pictures show for unregistered users?
04:14:40jhMikeSI have control over flyspray but that's about it.
04:14:43Paul_The_NerdMikachu: I can. I'm not sure if I should
04:15:15sharpeOr have it on watch if more than half of the words aren't actually words... woo.
04:15:31Paul_The_NerdjhMikeS: On the forums you can now lock or move threads. The theory being that if you have commit access, we can at least trust you with that, if you feel it's necessary. Not that there's any requirement you ever even visit the forums.
04:15:53Paul_The_NerdMikachu: In theory it keeps the bandwidth use down a wee bit.
04:16:01Mikachuright, it is on the 2mbit server too
04:16:14jhMikeSThey'd have to call me a communist for me to do anything...that's a big insult to a libertarian
04:16:16sharpeWell, apparently i'm trusted enough to have lock/move abilities, and I don't have commit access... only makes sense developers can. :)
04:16:45Paul_The_NerdWell, I no that nobody I've given lock/move to would abuse it. Or else.
04:17:13sharpeThat 'or else' leaves a lot to the imagination.
04:17:20Paul_The_NerdOf course.
04:17:34Paul_The_NerdIf I actually said what I would do, I leave my self open to all sorts of liabiliies.
04:17:36sharpeA man likes to keep his options open...
04:17:54jhMikeSif (!abuse_lockout) {} else { ?? }
04:17:59 Quit hotwire_ (Read error: 110 (Connection timed out))
04:18:57sharpeIt's such a great colorful raindow in the list of online forum goers.
04:19:17jhMikeSif (!abuse_lock_move) { :)))) } else { B^[[[[[ }
04:19:42Paul_The_NerdAnd that's without the Coder and Artist badges.
04:19:49jhMikeSWhat color rainbot?
04:20:10sharpethere's green, red, blue, rockbox-blueish...
04:20:14Mikachuhard to spell word
04:20:24sharpeOr gold.
04:20:30jhMikeSShould I have a can of paint made "Rockbox Blue"?
04:21:02Paul_The_NerdIt's supposed to be that other rockbox color. Kinda
04:21:09sharpeYou should. When it becomes the currency of the communist country, yeah...
04:21:47jhMikeSI'll design a flag
04:22:23jhMikeSWho will be on the RB1 bill?
04:22:32sharpeNah. We can just comission midkay to make one based off the forum tags he's made...
04:23:27sharpeJust go back to your mapping... :D
04:23:37jhMikeSWe'll buy a used oil platform in international waters and declare it Rockboxia
04:24:06sharpejhMikeS: Ever hear of islands?
04:24:06*jhMikeS is very silly at times
04:24:24jhMikeSsharpe: Those are taken already right?
04:24:49sharpeNot if you find an uncharted one...
04:25:11jhMikeSThere's a country call Sealand that an abandoned gun platform about 20mils off the coast of england. It's no joke...they got passports and all.
04:25:30jhMikeSsharpe: Would Gilligan sell it?
04:26:52sharpeI believe that was movie set...
04:27:17jhMikeSThe moon landing was a movie set...not that! That was real.
04:28:20jhMikeSOh boy: that's full of non-sequiturs
04:29:10 Quit Lynx_ (Connection timed out)
04:29:10 Nick Lynx is now known as Lynx_ (
04:31:18*Paul_The_Nerd is half-away
04:31:25*Paul_The_Nerd is beginning the setup for New Forums
04:31:34FebsMaybe we should change the title of that thread. "Warning to those who use Loader 2. If you use it on a player with a bad hard drive, your hard drive might go bad."
04:32:04*jhMikeS dropped his iPod off a 30 story building and now rockbox fails to come up. He thinks the bootloader might not be working properly because the apple firmware did something to it.
04:32:36MikachujhMikeS: that's JdGordon's fault
04:33:39*jhMikeS think JdGordon is a mean guy for providing the 30 story building *sniff*
04:39:34jhMikeSMan, time for bed. Goodnight all.
04:39:42 Quit jhMikeS ("Client exited")
04:39:43safetydanYeesh. Didn't realise my off hand comment about AA would generate so much discussion.
04:40:15*safetydan gently prods Paul_The_Nerd about getting the devloper label on the forums
04:40:48*sharpe points out he also just upgraded the Rockbox Experts group... :)
04:41:00sharpeHe, being Paul.
04:41:28Paul_The_Nerdsafetydan: Wasn't aware you had commit access. I don't track these things. ;) What's your forum nickname?
04:42:12safetydanthe very same
04:42:23safetydansafetydan == dan in the commit logs
04:42:32*safetydan is also my forum nick
04:42:40Paul_The_NerdAlso, would all devs, experts, and moderators please wander over to the forums and look after the "New Ports" category (not the forum, but where the X5 and Toshiba forums are) and see what they thing of the new forum categories. They will be replacing ALL the old ones.
04:42:46Paul_The_Nerdsafetydan: Gotcha. Going to go do it now
04:43:14Paul_The_Nerdsafetydan: 'tis done
04:43:19sharpeI am going to post in the "It's starting: The new forums."...
04:43:29safetydanPaul_The_Nerd, cheers
04:43:48*safetydan feels all powerful and stuff
04:44:19safetydanNow. A question. Does anyone prefer the titles in the dir browser to look like this?
04:44:32safetydanSomehow in all the excitement I missed the fact that there was an existing patch to do this.
04:44:36Mikachudon't know, can't see the image
04:44:58Mikachuso i guess i will just give in and register
04:45:00Mikachubut i won't post!
04:45:02sharpesafetydan: looks like a good idea.
04:45:08Paul_The_Nerdsafetydan: I definitely prefer that screenshot. I haven't seen the 'centered' ones yet, but I don't like centered stuff anyway
04:45:17sharpeMikachu: I barely post any at all...
04:45:43*Paul_The_Nerd discovered that he can arbitrarily change users' post counts.
04:45:49Mikachuooh, can i have 42?
04:46:07Paul_The_NerdI'm trying to decide if post counts are even useful here.
04:46:19sharpeThey don't seem to be.
04:46:49Paul_The_NerdI'm tempted to make posts stop counting, and remove the count-based member groups.
04:46:55Paul_The_NerdI'm no sure I want to do this yet.
04:47:05sharpePlan it on paper.
04:47:16Paul_The_NerdBut every now and then, I see people who seem like they might be posing just to increase their count.
04:47:23Mikachuokay call me stupid, what's the difference in that screenshot?
04:47:29Mikachuit has a little icon in front?
04:47:40sharpeMikachu: the little directory indication at the top.
04:48:04Mikachui would maybe vote against, leaves less space for text and it will only ever show directory names in that space
04:48:08Mikachuso it doesn't tell you anything
04:48:24Soapwould there be any way to create some sort of visual seperation between the current directory "title" and the rest of the screen?
04:48:29sharpeWe could have an option for it. :D
04:48:35Soapmaybe an underline or something?
04:49:19sharpeHmm. How does right-aligning the directory name sound? :)
04:49:27Mikachuwon't help if it needs to scroll
04:49:56Mikachua new low for spam mails, the From: header says "peepeegrowth"
04:49:59safetydanMikachu, icon and left aligned basically
04:50:17Soapsharpe - many of my directory names are long enough that right-alignment wouldn't look different that centered, they would scroll.
04:50:26Mikachui like centered and no icon
04:50:35sharpeI like left aligned...
04:50:54Soapi like centered, scroll it if you need to, and underlined.
04:50:56sharpeAnd with an icon.
04:50:59Mikachui also think that if it's possible, it should start out scrolled to the right edge of the text instead of the left when needing to scroll
04:51:08safetydanSoap, I tried it underlined... it ends up looking too busy
04:51:22Mikachuunderline could break those nice background bmps people have
04:51:48safetydanunderlining is probably better handled in background images
04:51:57Soapsafetydan - I'm just worried it won't stand out as being a different "context" since we can't change the font size, or do a graident titlebar.
04:52:00Mikachubut the underline would look strange when you're in a menu
04:52:13Soapsafetydan - could left (indented) work, and stay indented when scrolling?
04:52:19Mikachuthe scrollbar will end under it
04:52:46safetydanSoap, yes that's possible I believe. Just use an empty icon.
04:52:54Mikachubut i guess you large lcd people don't always have scrollbars
04:53:14Soapindentation would create enough of a visual seperation, IMHO.
04:53:42Soapa single (small) icon the same size of other icons is so easy to gloss over with the eye.
04:53:48Mikachuyou could, in theory, have it in a different color
04:54:10Mikachuyou could have an empty horizontal line under it
04:54:18Mikachuor a couple
04:54:32Mikachuthere is some slack at the bottom of the screen anyway, put that slack between the header and the items
04:55:09midkayMikachu: there's only a bit of slack for some fonts, depending on the font size if it will fill the screen or not..
04:55:10Soapthat would work, something, anything but relying on the presence of a (possible) scroll bar to differentiate between them.
04:55:10Paul_The_NerdAnyone think it would be bad of me to delete posts like this during the re-org?
04:55:17 Quit Criamos`Andy ("( :: NoNameScript 4.02 :: )")
04:55:55midkaya month-old "thanks" post with no replies.. don't we already have some "thank you" thread?
04:55:55SoapPaul_The_Nerd - I'll happily give you a list of posts I think need deletion ;)
04:56:09Soap13 months old
04:56:13sharpeThere is a very large bee flying around here...
04:56:33Paul_The_Nerdmidkay: Yeah, it's 13 months old. ;)
04:57:38Paul_The_NerdI think I'm going to remove one-shot questions that nobody is likely to ever ask again as well.
04:57:48midkayi say go ahead..
04:57:49Paul_The_NerdFor example, someone being told that a page (that no longer exists) is outdated.
04:58:01Soapremove post-whoring posts also???
04:58:04sharpe"When will rockbox be released for the iPods?"
04:58:08sharpethose kind? :)
04:59:08midkay<Soap> remove post-whoring posts also??? −−> it goes without saying that all of BigMac's posts shall be deleted in this case as well :)
05:00:33SoapI was JUST about to post this link
05:00:48midkayare any of his posts not like that?
05:00:56midkayi say clear all of them :)
05:01:19safetydanhotwire__, sim key mappings are in uisimulator/sdl/button.c
05:01:39 Quit tommeyer ("Clap on! , Clap off! clap@#&$NO CARRIER")
05:01:43*Paul_The_Nerd wonders if he should create a locked, historic posts forum for things like this:
05:02:07Paul_The_NerdOr maybe "Announce"
05:02:20sharpeYou could also name it, "Historic Posts."
05:02:45midkayhuh? he already suggested that.
05:02:50 Join Lynx [0] (
05:03:14sharpeHe didn't say the name for it...
05:03:44Paul_The_NerdI'm thinking "Announcements" is better since it allows for future announcements.
05:03:53midkayannouncements is good.
05:04:21midkayPaul_The_Nerd: "Unsupported Builds" forum description has a typo, "on thread per build" should be "one thread per build".
05:04:30midkaysame for New Ports, you say on instead of one.
05:04:34 Join hotwire___ [0] (
05:04:38 Nick hotwire___ is now known as hotwire_ (
05:05:02midkayand the "Cowan" forum should be "Cowon".
05:06:05midkayAnnouncemens -> Announcements ;)
05:06:18Paul_The_Nerdmidkay: I fixed the announcements one already.
05:06:22sharpeWoah. Twice the forums.
05:06:33midkayalso, check your period usage in forum descriptions.
05:06:36*Paul_The_Nerd didn't know how to spell Cowon, so used the spelling he was given by someone else who apparently didn't know
05:06:43Paul_The_NerdI haven't proofread them yet
05:06:52midkaylike every other description ends with a period, you should make up your mind. :)
05:08:27Paul_The_Nerdmidkay: More consistency with period use.
05:08:36Paul_The_NerdThe only ones that shouldn't end with periods are the lists of players.
05:08:40midkaythere we go.
05:08:40Paul_The_NerdUnless I overlooked one.
05:08:49Paul_The_NerdAlso, all spelling errors you pointed out are fixed, I think
05:08:50midkaylooks good.
05:08:56midkayAnnouncements ought to get a little description.
05:09:18midkayit should probably be locked from users creating new topics.. maybe allow them to reply but not create their own threads in there.
05:09:22sharpe"Announcements, Read : announcements."
05:10:28sharpeYes. It's hilarious midkay.
05:10:38midkaysomeone forgot a comma..
05:11:30***Saving seen data "./dancer.seen"
05:12:03 Join Daishi [0] (
05:12:23safetydanFrom misticriver: "How is gapless coming along on rockbox?"
05:12:25Paul_The_Nerdmidkay: I think I may not even let them reply. But right now I haven't set any permissions, all forums are just hidden from everyone outside the badged membergroups
05:12:38midkayoh, cool.
05:12:39dongslol gapless.
05:12:44sharpeI think the 'd' in "Rockbox General discussion" should be capitalized...
05:12:51dongshas the fadein/out pause problem been resolved?
05:13:02midkaysharpe: what? it is.
05:13:11sharpeIt wasn't before.
05:13:12safetydandongs, have you raised a bug about it?
05:13:13 Join pike [0] (
05:13:20dongssafetydan: no, i mentioned it in here.
05:13:21midkaybefore you refreshed maybe 20 minutes ago or what?
05:13:28sharpeTwenty seconds.
05:13:30Paul_The_Nerdsafetydan: Dongs is too good for bug reports.
05:13:35safetydandongs, it'll probably get lost unless you raise a bug report
05:13:40Mikachui like how dongs sometimes comes in and says "lol <random word>"
05:13:41sharpesafetydan: He takes it straight to the developers.
05:13:46Mikachulike lol gapless or lol lossless
05:14:45midkayPaul_The_Nerd: Hardware forum - last word should be plural, in the description.
05:15:13dongsMikachu: both equally useless
05:15:17 Join finn_ [0] (
05:15:22dongswhats that mode called that fades in/out
05:15:25dongson pause
05:15:31Mikachu"fade on pause/stop"
05:15:32dongsi dont use it, and to lazy to find the menu name
05:15:55Mikachumplayer has the best printfs
05:15:58Mikachu printf("vo_xvmc: OSD support by beckend rendering (fast)\n");
05:15:58Mikachu printf("vo_xvmc: Pleace send feedback to configrm that it work,otherwise send bugreport!\n");
05:16:09Paul_The_Nerddongs: You have a funny definition of the word useless. Most people define it as "having no use" but clearly people do use them. May I ask what it means to you?
05:16:19safetydanAh, English as she is spoke.
05:19:33dongsallrighty then
05:19:36dongsplease have this fixed
05:19:52Snypersafetydan: Hey, I was just going over the daily changes. Just wanted to make one correction. I think the peak filters do work. I just didnt test the bandwidth that should be fine.
05:19:56Mikachudongs: heh, you're timecop?
05:20:03dongsyoure surprised?
05:20:07Mikachui guess not
05:20:22sharpeHasn't this been discussed before...?
05:20:28Mikachui use some dockapps you wrote :)
05:20:39 Quit hotwire__ (Read error: 110 (Connection timed out))
05:20:42dongssharpe: of course. and apparently nobody thought it was worth fixing such a horrible UI omission
05:20:45dongsMikachu: lorf
05:21:02dongssharpe: they'd rather work on useless stuff like 'gapless' or 'lossless'
05:21:23safetydanSnyper, it might just have been an earlier patch of mine where they didn't work. I didn't have anyone else to test, so I put the warning in :)
05:22:09 Join psyklops [0] (
05:22:13Snypergotcha, I just downloaded the latest snapshot, built it, and am putting it on my ipod right now....I'll probably test it out just before I got to bed (very shortly after I finish unzipping :)
05:22:13midkaydongs: if it was already decided not worth fixing, why are you mentioning it again?
05:22:19sharpedongs: I believe the argument against immediately changing the display to being paused, was that technically the music is still playing as it is being faded, so having it display a paused icon as it is fading would be misleading...
05:22:19Paul_The_Nerddongs: Again, please define 'useless' or better yet speak common English.
05:22:33dongssharpe: why dont you read my bug report
05:22:50sharpeWhat do you think I have already read?
05:22:52Paul_The_Nerddongs: And of course, if the user has read the manual and knows about the fadeout, and *gasp* LISTENS to the fadeout, they know what's happening.
05:23:10dongsif you think what I described is normal, you need to stop developing software that other, normal people use.
05:23:37dongsPaul_The_Nerd: documenting bugs/features doesnt make them any less in need to be fixed
05:23:41midkaywhy? normal people use it.
05:23:42Mikachui'm with dongs on this one, but i would choose different words
05:23:47Paul_The_NerdHow is it a bug?
05:24:00dongsits a user interface bug, no matter how you look at it.
05:24:03Paul_The_NerdIf it's "working as expected, but not what you want" it's not a bug. Just a feature you don't like it.
05:24:05dongsyou do something, you need instant feedback
05:24:08dongsyou dont need it 5 seconds after
05:24:17Paul_The_NerdYou have instant feedback: The music starts fading.
05:24:18dongsyou will NOT be able to tell the fade has begin
05:24:24sharpedongs: I think you're exxagerating the fade time...
05:24:29midkayisn't it about time dongs was banned? seriously..
05:24:33dongsif the player was in sleep mode, single key press will not start fade
05:24:35dongsit will wake up the lcd.
05:24:39 Join hotwire___ [0] (
05:24:41safetydanBig problem is that it's hard to fix.
05:24:43dongsso you dont know if it actually started fading or not.
05:24:51safetydanor at least on a quick skim of the wps code anyway
05:24:59Paul_The_Nerddongs: Wait, in that case it won't pause either.
05:25:04dongsno shit
05:25:07dongsbut youd know
05:25:09dongsif it started or not
05:25:10Paul_The_NerdSo what does that have to do with anything?
05:25:17Paul_The_NerdI can hear the fade within about 1/3 of a second of it starting.
05:25:23Paul_The_NerdIt's kinda noticeable.
05:25:36Paul_The_NerdAnd I mean, do you regularly not know what your settings are, or something?
05:25:39Mikachui think the fade is annoying, i also disable it
05:25:52Paul_The_NerdI think the fade should default to disabled, yes.
05:26:05dongsfade doesnt have to be annoying but the fact that it deos NOT inform the user its happening absolutely kills whatever positive uses it has
05:26:23midkayby god, all this over a <1 second fade..
05:26:25dongsand like I just said, if player was in sleep, a single tap on pause will not acutalyl start pause at all
05:26:26Paul_The_NerdIs the user deaf?
05:26:43Mikachuwhat if you don't actually have the headphones in?
05:26:49dongsso the user will be sitting there wondering hey did this shit start pausing or is it just now waking up and im gonna try clicking pause again oh shit wait it was already fading fuck, now i have ot wait like 20 seconds while the song fades in/out
05:26:54midkaydongs: why don't you just stop talking? it's not easy to change, there's no reason to change it, we've been over this before at least once.
05:27:02midkayit is not a bug.
05:27:03Paul_The_Nerddongs: It's neither 5 nor 20 seconds.
05:27:45dongsmidkay: yes it is a bug.
05:27:46sharpedongs: You're also the only one, besides whoever it was last time, that has brought it up.
05:27:52dongssharpe: that was me.
05:28:05sharpeSounds like a personal problem then, no?
05:28:19dongsdo you totally not see the problem with the way its done now?
05:28:25dongsi mean forget who youre talking wiht for a second etc
05:28:35dongsand imagine this is a product you paid $200 for
05:28:35Paul_The_Nerddongs: The definition of "bug" is unintended behaviour. If it's working how the designer wants it to work, how is it a bug?
05:28:40dongsand it pauses in this way
05:28:44dongswould you be happy wiht that? y/n
05:28:54dongsPaul_The_Nerd: the designer is wrong.
05:28:59Snypernight everyone
05:29:00 Quit Lynx_ (Connection timed out)
05:29:02Paul_The_NerdThat doesn't make it a bug.
05:29:11Paul_The_NerdThat makes it a feature that you don't like. There is a significant difference.
05:29:11sharpeI really do not see the problem, the audio is still playing as it is faded, which would mean it should still show that audio is playing.
05:29:29dongssharpe: there is no instant feedback about the operation.
05:29:29Mikachuhow about a blinking pause icon!
05:29:32safetydanI kind of agree with dongs that there should be feedback. But I don't like the combativeness.
05:29:41dongssharpe: user is left to guess for '1/3 to 2 seconds' while the shit decides to fade.
05:29:52sharpeOr a blending play to pause icon!
05:29:52dongswhat it he presses pause near hte end of hte song, because hes done listening?
05:29:54Mikachui think there is equal combativeness on both sides here
05:29:54Paul_The_Nerdsafetydan: I agree that feedback would be good. I disagree that it's a bug, and I disagree that it should show "Paused" before the music has stopped playing
05:29:56midkayit's never two seconds.
05:30:01dongswhere the volume is already almost gone?
05:30:07dongshow does he know it acutally started pausing?>
05:30:09Paul_The_NerdFor example, if you want to Pause at 1:17, you should know that it continued for 2 seconds
05:30:20 Join Lynx_ [0] (
05:30:21 Quit psyklops ("Leaving")
05:30:24Paul_The_NerdOr whatever the fade time is.
05:30:36midkayit's entirely dependent upon the current volume..
05:30:40safetydanPaul_The_Nerd, so it's probably a feature request to add another state to the WPS "pausing"
05:30:47Paul_The_Nerdsafetydan: Indeed.
05:30:59sharpeSounds like a feature request to me...
05:31:03 Quit finn23 (Read error: 110 (Connection timed out))
05:31:06Paul_The_NerdIt makes vastly more sense that the pause icon doesn't show until the timer has stopped moving
05:31:10Paul_The_NerdI agree instant feedback is good
05:31:14Paul_The_NerdBut dongs is wrong as to what it should be.
05:31:30Mikachui think pressing play while it is fading out should inverse the fade instantly, not wait for the fadeout
05:31:41Mikachuideally, it's probably hard
05:31:43sharpePaul_The_Nerd: That rhymed.
05:35:15dongsso then
05:36:29midkaydongs: let me sum this up, please... you think it makes more sense to tell the user that "playback has stopped, we're in pause mode" while the music continues to fade out over the course of - as you put it - 5 seconds, rather than having the user find out for themselves whether or not the fade has started within about a third of a second of hitting the button?
05:36:33 Quit hotwire___ ("Chatzilla 0.9.75 [Firefox 2.0b1/2006071020]")
05:37:05Mikachuthat is not exactly right
05:37:10dongsmidkay: it needs instant feedback for whatever. i dont care what it is, but if I press pause, i need to know it actually happened, especially on these players that have sleep mode where a tap would just wake up the LCD without acutally doing anything.
05:37:26*Paul_The_Nerd points out that is also configurable.
05:37:30dongsone solution would be to change pause indicator immediately. another one as Mikachu said, to have a pausing state in wps.
05:37:37 Join hotwire [0] (
05:37:41dongsi dont care what it is, but it needs to be done, to avoid user confusion
05:37:43Mikachu05:26:43 <Mikachu> what if you don't actually have the headphones in?
05:37:49dongs^ that too
05:38:00midkayMikachu: so you'd wait to see if the fade completed.
05:38:01Mikachuif someone approaches you, you pull the headphones out and start talking, then absentmindedly try to pause the player
05:38:03midkaythe pause icon would eventually appear.
05:38:15Mikachuwaiting around for icons is not my favorite pastime
05:38:18midkaydongs: there's a lot more that needs to be done..
05:38:30dongsits small things liek that that piss users off
05:38:32Paul_The_NerdI have first keypress enables backlight disabled, so I press pause once confidently knowing that it'll actually pause.
05:38:33midkayMikachu: nor mine, but it takes around under two seconds at full volume.
05:38:37 Quit hotwire (Client Quit)
05:38:38dongswhy do you think normal people dont use linux?
05:39:03Paul_The_Nerddongs: Because computers don't come with it.
05:39:10Paul_The_NerdNormal users don't install their operating systems. Ever.
05:39:15dongswhen most of your day is spent fighting 'misfeatures' that 'developers' think are done exactly the way they should be, you have a big problem.
05:39:35dongsi doubt preinstalling linux would have any effect on its user base
05:39:48dongseither that, or a lot more people would learn how to install windows XP real quickl.
05:40:16*Paul_The_Nerd shrugs
05:40:23sharpedongs: You're implying all linux users aren't normal, and all non-linux users are. And that's not nice to generalize people.
05:40:23Paul_The_NerdThere's no way to prove either case, so it's relatively pointless even discussing it.
05:40:31 Join hotwire [0] (
05:40:32Mikachu"<benplaut> i put linux on it, cause it was really too old to do much else with it, he gets ahold of a XP cd from another kid and overwrites about 5 hours of work i had done. then he comes and asks me for help with all the inevitable driver problems he's having"
05:41:08dongs5 hours to install linux?
05:41:08Paul_The_Nerddongs: Anyway, you got your feature request written, you got an acceptable compromise. You done? Can we says "Bye bye now"?
05:41:16Mikachuit was an old laptop
05:41:23 Quit hotwire_ (Read error: 110 (Connection timed out))
05:41:28dongsi have a celeron 333 laptop running xp
05:41:39dongsit took 30 minutes at the most to install
05:41:54Paul_The_Nerddongs, Mikachu, please drop this discussion.
05:41:57Mikachuthe point wasn't the install time though
05:41:59Paul_The_NerdThis is not the place. PMSs work.
05:42:13Mikachuapparently there is some PMS going on here :)
05:47:43 Join fatherfork [0] (
05:50:39 Join Lynx- [0] (
05:51:38 Quit Lynx (Connection timed out)
05:52:31safetydanhrm... switching back and forth from tagcache to file view crashes in the sim
05:52:37safetydandoesn't on target though
05:53:58Paul_The_NerdI've heard problems relating to switch to tagcache, where in the past music hasn't worked until it's been rebooted even when it shows up.
05:54:05Paul_The_NerdMaybe there's some stray code in the switch still.
05:54:58sharpeMaybe, if a randomly picked number is even, require another reboot... :)
05:56:45safetydanit looks like the tagcache view doesn't properly restore the filetree
05:57:45sharpeSomeone tell me what the first type of music they think of when they hear, "Party Mode," please...
05:58:11sharpeExactly what I though.
05:58:14safetydanB52's - Love Shack?
05:58:26sharpeRock Lobster?
05:59:07fatherforkDane cook, "THAT'S NOT THE BATHROOM"
05:59:10 Join finn__ [0] (
05:59:21sharpefatherfork: Coats?
05:59:30fatherforkno, "WOOFER"
05:59:42sharpeOh, right. Coats is another thing.
05:59:54fatherforkyeah "someone... shit on the coats"
06:00:17sharpeThen fade back into the crowd...
06:01:45JdGordonhey guys
06:01:53JdGordonhow do i commit a new file?
06:02:06safetydanJdGordon, "cvs add <file>"
06:02:14safetydanthen "cvs commit <file>"
06:02:17 Join evilnick [0] (
06:02:24safetydanThen you type your commit message in to the editor
06:02:39evilnickI've noticed a problem with the cvs builds
06:02:50*safetydan goes back to prodding tagcache
06:03:02evilnickthere's no keymaps "folder" under "apps"
06:03:02sharpeI have to say, I do like the centered menu titles for the options...
06:03:39evilnick(quoted the wrong word there)
06:03:51fatherforklemme get something strait real quick *as I am an hour into a scan*, foobar2000 will only add replaygain data to the ID3 tags correct?
06:04:00evilnickshould have read: there's no "keymaps" folder under "apps"
06:04:04safetydanfatherfork, yes
06:04:16safetydanevilnick, yup looks like it
06:04:50evilnicknot cvs, I'm being a fool. I meant bleeding-edge
06:05:17evilnicki.e. rockbox-bleeding.tar.bz2
06:05:51Paul_The_Nerdevilnick: Ah, but it's there if you use CVS to get your sources?
06:05:59Paul_The_NerdIt sounds like another thing not getting added to FILES and such
06:06:07safetydanPaul_The_Nerd, and you'd be correct
06:06:39joshua_jabup jabup ja get down
06:06:44joshua_whoops, wrong window
06:06:50evilnickPaul_The_Nerd, I think you're right, not being added to FILES
06:07:03Paul_The_NerdIs there a keymaps file or folder?
06:07:14*Paul_The_Nerd doesn't see it.
06:07:23evilnickno folder and hence no files beneath it
06:08:01Paul_The_NerdI meant I don't see any indication that there's supposed to be one.
06:08:01JdGordon.. fixing that
06:08:03sharpeThere isn't a keymaps folder in the releases anyway... CVS or not...
06:08:11JdGordoncvs there is
06:08:13JdGordonunder apps
06:08:21 Join Lynx [0] (
06:08:23Paul_The_NerdJdGordon: It was never added with a commit, that I see...
06:08:30sharpeI meant builds from CVS
06:08:54JdGordonPaul_The_Nerd: ? the keymaps folder was added a week or so ago.. it was populated yesterday
06:09:01 Quit Lynx_ (Connection timed out)
06:09:02 Nick Lynx is now known as Lynx_ (
06:09:07evilnicksharpe: sorry about that, was talking about the bleeding edge source
06:09:23JdGordonwhat make command do i do to check the FILES file?
06:09:37Paul_The_NerdJdGordon: I was just scanning the changelog and didn't see it... I dunno. I didn't check my actual CVS folder.
06:09:37safetydanshould be fixed next time a CVS build happens
06:09:49 Join webguest81 [0] (
06:09:57JdGordonsafetydan: you added it to FILES?
06:10:07JdGordonta, i was about to
06:10:35 Quit webguest81 (Client Quit)
06:11:07safetydanhrmm... apparently is sending spam
06:11:42sharpeOr it's forged as to be from ?
06:11:49 Quit Daishi (Remote closed the connection)
06:13:21 Quit Lynx- (Read error: 60 (Operation timed out))
06:14:20 Quit finn_ (Read error: 110 (Connection timed out))
06:15:49JdGordonhas the build hung?
06:15:58Paul_The_NerdIt's been hung for quite a long while.
06:16:09Paul_The_NerdSince perhaps before I signed into IRC this evening.
06:16:20sharpe... I now hate my friends... :)
06:16:25JdGordonoh, i just looked at the miutes in the time.. not the hour :p
06:16:29safetydanWell snce 23:56 UTC anyway
06:16:34safetydansince even
06:19:54fatherforkwhat do I do if this thing is telling me my nick is owned by someone else? I know this to be highly unlikely.
06:20:18midkayJdGordon: committer now, huh? :)
06:20:24sharpefatherfork: that it is registered, and if it is registered to you, you need to identify yourself...
06:20:44JdGordonmidkay: ye :) no more nagging everyone
06:20:51midkayhaha. thank god.
06:21:13fatherforksharpe, I tried with a few different passwords, none of which worked. I think I remember registering it some time ago. How would I reclaim a lost password?
06:21:47sharpefatherfork: You can wait for it to become unregistered then register it again...
06:22:09midkaysharpe: he doesn't "need" to identify himself.
06:22:31fatherforkmidkay, explain yourself
06:22:39sharpemidkay: Yes, I know. Just like you don't "need" to point that out.
06:22:46fatherforkwell I guess, as long as I'm using it, it doesn't matter
06:22:48 Quit apo (Read error: 110 (Connection timed out))
06:22:48midkayfatherfork: sharpe said "you need to identify yourself". but you don't.
06:23:05midkayit won't kick you off or anything if you don't. i just wanted you to know that it's not urgent.
06:23:28 Join apo [0] (
06:23:43fatherforkI just believe it unlikely that somebody else has registered my nick
06:23:58midkayfatherfork: right, so it might be yours. you might have to wait if you can't figure out the password.
06:24:05midkaybut regardless, if you can't figure it out, no worries.
06:24:24fatherforkthere's no way to have it sent to me? if not, I'll take the no worries option.
06:24:26sharpeI wouldn't have expected 'sharpe' to be unregistered when I chose it... :)
06:24:51midkayfatherfork: not that i know of, no.
06:25:19fatherforkah well, hakuna matata
06:25:20sharpeSince it's only linked to you by a password, there isn't really any other way...
06:26:21safetydanYou usually register an email when you register a nick
06:26:29fatherforkyeah that's what I thought...
06:26:35sharpeI don't recall doing that...
06:27:28Mikachuyou can ask an oper
06:27:52safetydanthe freenode faq doesn't mention how to recover or reset your password though
06:28:35fatherforkcrap, what's the syntax to register?
06:28:49fatherfork"/msg NickServ"... something
06:29:14sharpeit should tell you if you do, "/msg nickserv register"
06:30:43fatherforkyeah, defintely forgot my password
06:30:56sharpemaybe it's 'password'?
06:31:32sharpeWorth a try.
06:33:13Mikachuif you "/stats p", you should get a list of online opers that can help you
06:34:08safetydanRight. That's enough Rockbox coding for today...
06:34:10*safetydan wanders off
06:34:42fatherforkMikachu, you sure? it seems to do nothing
06:34:44sharpeHah. The day hasn't begun for Rockbox coding for me yet.
06:36:24Mikachufatherfork: the list can be empty at times
06:41:37 Join aliask [0] (
06:42:50 Quit midgey34 (Read error: 60 (Operation timed out))
06:44:51fatherforkgah, only an hour left on a 3 and a half hour replaygain scan
06:45:09Mikachuwow, how much music are you scanning?
06:45:27fatherforkeh, about 3500 songs. I've never scanned them before
06:45:32sharpeThe entire music collection of the Eastern Hemisphere.
06:45:55Mikachumy collection is about 6500 songs, but i scanned them incrementally :)
06:46:07fatherforkyeah, it's worse when you do it all at once
06:46:23sharpeI've barely around 1300... I feel so insignificant.
06:46:44fatherforkMikachu, do you listen to all of that?
06:46:52Mikachuthat's what i have in my playlist now
06:47:05fatherforkBecause I can honestly say I would listen to all of my music
06:47:38fatherforkand my new buddy, crossfade, makes it much more awesomer
06:47:45sharpeThat's a band.
06:47:50fatherforkyeah, I know
06:47:56fatherforkI listen to them
06:48:01sharpeI don't...
06:48:26fatherforkAlright then
06:49:05fatherforkdeja vu
06:49:12fatherforkman that was weird
06:49:16sharpeI have it very often.
06:49:27fatherforkit happens when they change something...
06:49:41*Mikachu checks cvs log
06:57:14sharpeSleep time for me. G'night everyone.
06:57:34 Quit sharpe ("Leaving")
06:58:26fatherforkok, so I made the most basic wps ever, only including the current song info, and the equalizer still made mp3s skip on my 5G iPod
06:58:41Mikachutry the hw eq
06:58:51fatherforkI also turned every other enhancer off
06:59:01fatherforkhmm, that's in the daily buid eh?
06:59:07Mikachui believe so
06:59:15Mikachubut i don't have a 5g so i don't know exactly where it is
06:59:30fatherforkyeah I saw where to go
06:59:58 Join finn_ [0] (
06:59:58fatherforkI think I'll be lazy and just wait for senab to release a new patched version.
07:00:54Snyperok, so I am going to be really tired tomorrow, but I couldn't resist...
07:01:09Snyperbad news on the HW eq...the peak filters do not work
07:01:15Mikachuthat is known
07:01:26SnyperI just spent the last 2 hours making sure everything was being sent right
07:01:44Snyperaccording to the datasheet that we actually have...
07:01:52fatherforkwhat time is it there?
07:02:01Snyper1:04 am
07:02:06fatherforksame here
07:02:09SnyperI gotte be up for work at 6:00
07:02:09fatherforkok, go on...
07:02:15Mikachufatherfork: /ctcp <nick> time
07:02:23fatherforkoh cool
07:03:10fatherforkaw boo, I can't use it because I'm "not registered"
07:03:17fatherforkI believe we went over this
07:03:20fatherforkoh well
07:03:25SnyperI tried a few other things just to try...but I am guessing its register numbers...or the full eq is not supported...
07:04:17SnyperI may start just messing with registers, until I find one, but not for a little while....I need sleep....
07:04:26Snyperthank goodness friday is an "off friday"
07:05:21fatherforkfryday is not today
07:05:46fatherforkor tomorrow...
07:09:51fatherforkI remembered my password
07:10:30fatherforkI swear I tried that one
07:11:33***Saving seen data "./dancer.seen"
07:11:47 Join midgey34 [0] (
07:12:22 Quit finn__ (Read error: 110 (Connection timed out))
07:13:17 Quit JdGordon (Read error: 104 (Connection reset by peer))
07:14:15 Join JdGordon [0] (n=jonno@
07:30:00COmputomananyone use oink here
07:30:20fatherforkwell, not me at least
07:30:24COmputomanI used to have an account
07:30:34COmputomangot turned off due to inactivity
07:30:41COmputomanThen i went into their irc chennel
07:30:54COmputomanand said something about how i got booted for inactivity
07:31:03COmputomanthen they permaban me from their server
07:31:10fatherforkyeah they expire
07:31:10 Quit rotator ("zzzzzzzz")
07:31:10COmputomanjust for saying anything related to invites
07:31:28JdGordonhe said invites... ban him !
07:31:30Mikachui think maybe you are retelling this a bit biased
07:31:52COmputomanI sure as hell didnt ask for an invite
07:32:13Paul_The_NerdCOmputoman: But they have no real way of knowing you really had an account.
07:32:48COmputomanI looked on the internet, and apparently lately they have been banning people
07:32:52Paul_The_NerdSo, from their perspective you're either someone who had an account, lost it, and is clueless as to the proper process for getting this fixed, or you're someone trying to get given one.
07:32:53COmputomana hole lot
07:33:13Mikachuyou need to be invited again in any case
07:33:36COmputomanWhich would take a some time in the IRC channel
07:33:41COmputomanI cant think of anywhere else
07:33:55COmputomanthats where i got invited in the first place
07:34:20COmputomanoh well
07:34:24COmputomanI cant change my IP
07:34:30Paul_The_NerdWell maybe if you'd been a more active, and thusly valuable member of the community, things would've gone better.
07:34:33COmputomaneven if i dissconnect my modem for a long time
07:34:48Mikachuban evading isn't likely to get you an invite
07:34:54COmputomanMy internet got cut for while
07:35:24COmputomanWhatever, its true :D
07:35:45Paul_The_NerdSeriously, just accept the consequences of your actions and go on with life.
07:36:22aliaskIt's a bittorrent tracker... what's wrong with any of the other ones?
07:36:58fatherforkreally, something as trivial as oink which requires so much attention seems... trivial
07:39:31COmputomanIts the best
07:39:49COmputomanbut its my fault i guess
07:39:55fatherforkI've never used it, and I'm just fine
07:41:25Paul_The_NerdJdGordon: Out of curiosity, would reversing the list be easier than reversing the keymapping?
07:41:39JdGordonWHICH LIST?
07:41:42JdGordonsoz caps
07:41:51Paul_The_NerdWhichever one that scrolling goes the wrong way on?
07:42:23JdGordoni dont belive any are wrong.. if your refering to the forum thread its not a list thats wrong.. its using the wrong context
07:42:47*Paul_The_Nerd shrugs
07:42:52Paul_The_Nerd'tis a debug screen anyway
07:42:57MikachuJdGordon: i noticed the keyboard screen doesn't use actions, is that on purpose?
07:43:10JdGordonyes it is.. thats gonna be a pin to do
07:43:12 Join muesli__ [0] (i=muesli_t@
07:43:12Paul_The_NerdIs there any value to ever increasing the boost counter more than one?
07:43:19Paul_The_NerdOther than I suppose, seeing if too many subtracts happen.
07:44:02MikachuPaul_The_Nerd: it won't make the cpu run faster
07:44:09Paul_The_NerdOh, I'm quite certain of that
07:44:18Paul_The_NerdSeeing as I beat that over the iPodders heads at least once a weak in the forums.
07:44:28*Paul_The_Nerd seriously cannot spell.
07:44:36Mikachuonce a weak, always a weak
07:45:50Paul_The_NerdWell the ability to set the boost counter higher than one confuses people as-is.
07:45:52 Join finn__ [0] (
07:46:15Paul_The_NerdI was just wondering if maybe the option should just be "Boost" "Normal" and "Idle" and leave the counter outside of user control, which is why I was wondering if there's a real use for increasing it beyond 1.
07:46:16Mikachuwhat do you mean ability? it's only set internally in the code
07:46:27Slasheripetur: #define SCL_HI and_l(~0x00001000, &GPIO_ENABLE); while(!SCL);
07:46:49Slasherino endless loops in the eeprom driver please, that can crash bootloader and brick your player
07:46:50Paul_The_NerdMikachu: You can change the counter in a screen from the debug menu.
07:47:07JdGordonPaul_The_Nerd: its a debug screen... not really all that important
07:47:16Mikachudid they miss the text that says KEEP OUT! ?
07:47:32Paul_The_NerdMikachu: On 4G ipods, being permanently boosted prevents a common freeze
07:47:34Mikachuremoving the counter from the debug screen would be... stupid
07:47:54Paul_The_NerdI was suggesting removing the ability to add more than 1 to the counter, or rather, asking why that ability is there.
07:48:21Mikachuif you increase it to 10 and it ends up at zero, you know some call is unbalanced i suppose
07:48:49Paul_The_NerdThat's about what I said, as the only case I could think of. ;)
07:49:31JdGordonits easier code to just call cpu_boost(true); each time up is pressed.. no need for house-keeping code
07:50:27Mikachuthat is also a very good point
07:50:27fatherforkis not "house keeping code" related to "optimizing"?
07:50:44JdGordonthis is the oposite tho...
07:50:46Mikachuincreasing code size to make the debug screen user friendly seems pointless
07:51:48amiconnSlasheri: That loop isn't endless
07:52:18Slasheriamiconn: if something goes wrong, it might be
07:52:27Slasheriand i don't like that kind of loops in the bootloader
07:52:47Slasheriwhile testing the eeprom driver, i have seen before that this situation might happen
07:53:32*amiconn still thinks that eeprom access and other advanced stuff shouldn't be done at all in the bootloader
07:56:31evilnickC0mputoman, are you still around?
07:56:36JdGordonPaul_The_Nerd: can u delet your msg in the buton thread, it might confuse someone... and if you can be bothered.. all the previous posts which have been address?
07:57:32fatherforkHA, I finally put two and two together.
07:57:37fatherforkGo me.
07:58:58 Quit finn_ (Read error: 110 (Connection timed out))
07:59:20Paul_The_NerdJdGordon: You have commit access now, no? That's your name in the logs, right?
08:00:16amiconnJdGordon: I am missing a rather important option in the new action stuff: If a certain option isn't defined for a target, there should be a way to exclude the associated code
08:00:45JdGordoni.e ?
08:00:56amiconnRight now we're including dead code this way - code for impossible actions
08:01:12 Join Lynx [0] (
08:01:28Paul_The_NerdJdGordon: You should be able to delete replies to your topic yourself now.
08:01:38JdGordonta Paul_The_Nerd
08:01:53JdGordonamiconn: well we can fix that, but it brings back #ifdef'ing...
08:02:37amiconnYes - *some* ifdefing
08:02:51amiconnJust #ifdef <action>
08:03:03amiconncase <action>:
08:03:11hotwireis anybody online any good with SDL?
08:03:28amiconnBut that means defining the actions differently
08:03:38amiconni.e. conditionally
08:03:46JdGordonwhy? we can define them conditionally in the actions.h enum
08:05:19amiconnI am thinking of e.g. the charging splash / shutdown warning for recorder v1 in tree.c you disabled
08:05:52JdGordonive rre-enabled that.. but im not sure if i did it in such a ncie way
08:05:56amiconnIt also seems that thF3 quickscreen won't work anymore
08:06:00JdGordon.. so its not commited yet
08:06:56JdGordonamiconn: can whatever you want to #define out be done with a HAVE_ define? like the quickscreen?
08:07:05JdGordoninstead of action
08:08:04amiconnWhat would be the advantage?
08:08:20JdGordonlooks nicer?
08:08:28 Join Davide-NYC [0] (
08:09:20amiconnWith both methods you would just check one definition where the action is used
08:09:27daurn|laptopis there an ipod linux irc channel?
08:09:42Davide-NYCis anyone here familiar witht he iRiver H100 remotes?
08:09:58Paul_The_NerdFamiliar in what way?
08:10:04Davide-NYCQuestion: how do you navigate UP a level in the file browser?
08:10:07JdGordonamiconn: yes, but.. umm.. i dunno.. i guess it doesnt really make a difference does it
08:10:19 Join theli_ua [0] (
08:10:23aliaskdaurn|laptop: #ipodlinux i think
08:10:27JdGordoncan anyone here fix the build comp? builds have been stalled for the last few hours
08:10:53amiconnAnyone = {Bagder, Zagor, LinusN}
08:10:55Paul_The_NerdDavide-NYC: Don't know anymore. Mine's been broken for a while. Used to be Stop
08:11:00Davide-NYCI am of the opinion that the remote buttons need to be remapped.
08:11:00Mikachui'm afraid it's 8am in sweden
08:11:02amiconnWe have to wait for one of them
08:11:02Bg3rmorning :)
08:11:20Bg3rMikachu: i'm sure amiconn knows this :)
08:11:41Mikachume too
08:11:45Bg3rJdGordon: any news ?
08:11:49amiconnJdGordon: AM I right in thinking that for targets with only non-lcd remotes of one type we could simply add the remote buttons to the main contexts?
08:12:04aliaskLinusN closed a task about an hour ago, so he's awake... might be busy now though...
08:12:08amiconnMikachu: Why 'afraid'? It's the same time here...
08:12:28Bg3rJdGordon: i plan to fix iriver remotes' mappings today if somebody doesn't do it before me
08:12:33Mikachui figured it wouldn't be a time they would be on irc
08:12:39JdGordonamiconn: yes, and just make sure in the get_mapping function you have CONTEXT_BLAA|CONTEXT_REMOtE for ech (or strip _REMOTE)
08:13:02 Join SereR0kR [0] (
08:13:11amiconnYes, I would just switch (context & CONTEXT_REMOTE)
08:13:19 Quit fatherfork ("CGI:IRC (EOF)")
08:13:40JdGordonthat works
08:13:43amiconnI want to bring back archos remote buttons (but need batteries for the remote first)
08:14:40amiconnNah, that wouldn't work. Should be switch (context & ~CONTEXT_REMOTE)
08:15:07Mikachuyou can only have 32 contexts?
08:15:13Davide-NYCI made some good GIFs for the iRiver remotes and the iRiver unit. They couldbe included in the manual. I have no experience with this but hte GIFs are attached to the forum post if anyone has interest.
08:15:28Davide-NYCgood night.
08:15:33JdGordonMikachu: no, the limit is much hiegher than that isnt it?
08:15:42 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
08:15:44Mikachunot if it's a bitmask?
08:15:52JdGordon#define CONTEXT_CUSTOM 0x40000000
08:15:55amiconnNo it's not
08:15:56Bg3ramiconn: the archos remotes are with batteries???
08:16:06Mikachuokay, ignore me then
08:16:08JdGordonpleeent of contexts there
08:16:37Bg3rJdGordon: these context will be enough for everyone
08:16:59amiconnBg3r: Yes. The archos remote is a simple RS232 sender with 2 lithium button cells (CR1616) in it.
08:17:27Bg3r2^28 ....
08:17:36amiconnBg3r: They work for 2..3 years
08:17:44JdGordonamiconn: in the logs u mentioned not needing CONTEXT_SETTINGS any more.. i think it might be possible, but still some things do use that context, but they could be fiddled with to do that
08:18:15amiconnYeah, andf I'd rather want to use CONTEXT_SETTINGS for the new settings as well
08:18:57JdGordonthere still needs to be a difference between settings like the eq screen, and the old-stlye settings..
08:19:48amiconnDid you cvs up recently?
08:19:49safetydanhotwire, what do you need to know about SDL?
08:20:17JdGordonamiconn: ive fixed the debug screens which u pointed out in the log, added h10 and there was one other thing i think but cant remember
08:20:27JdGordonand i read you wrong.. yes
08:21:26hotwiresafetydan: I was thinking of using either the Scroll Lock or Num Lock as the toggle for hold in the UI Simulator, and then use the keyboard LED as the user indicator for it.
08:21:40hotwire(for those ports that have a hold button)
08:22:05amiconnPlease don't use numlock
08:22:18amiconnThat would be badness on laptops
08:22:34amiconnScroll lock or caps lock would be fine
08:22:40 Quit Lynx_ (Connection timed out)
08:22:40 Nick Lynx is now known as Lynx_ (
08:23:34hotwireHow about i initial state the hold boolean as false, and toggle for each touch of Scroll Lock?
08:24:04amiconnHmm, what about remote hold?
08:24:28hotwireright... there are a few ports that support that.
08:25:09hotwirecaps lock for that one then
08:25:15amiconnYes, all targets that have a remote with a hold switch whose state can be read
08:26:11safetydanIs there a reason to use any of the lock keys? The WPS should indicate hold status anyway.
08:26:52Bagder- static const unsigned char *nsec_units[] = { "ns", "ľs", "ms" };
08:26:52Bagder+ static const unsigned char *nsec_units[] = { "ns", "s", "ms" };
08:26:53hotwirecan we safely assume that a wps designer has included it?
08:26:59Bagderwas committed...
08:27:00amiconnOther places which use hold don't show it
08:27:18amiconnBagder: again??
08:27:29JdGordonsorry.. was that me? i gotta fix kate
08:27:32Bagderyes, yday was only a discussion about it I think
08:27:38BagderJdGordon: it was
08:27:58JdGordonye, i would have fixed it.. but the cvs page has been frozen...
08:30:26amiconnJdGordon: Hmm, why does CONTEXT_REMOTE exist at all, btw? Is that just for saving to wade multiple lists everytime?
08:31:13Bg3ramiconn: probably because of the iriver remotes
08:31:35hotwireOkay, suggestions for the buttons to use then?
08:31:56amiconnBg3r: The iriver remotes need multiple, selectable tables. That doesn't necessarily mean that they need to be indicated as a completely separate context
08:32:19Bg3ramiconn: what do u suggest ?
08:32:50amiconnIt's not a suggestion, I just want to know whether I do understand the idea correctly
08:33:11amiconn...before adding more remote mappings
08:33:17amiconn(X5, archos)
08:35:31 Quit theli_ua (Remote closed the connection)
08:35:38amiconnThe archos remote adds such a small number of entries that a separate context is probably unnecessary, but what about the X5 remote?
08:36:12amiconn(the archos remote can only ever send simple button events, i.e. neither repeats nor releases, and there are only 6 buttons)
08:36:15 Join theli_ua [0] (
08:36:29 Quit theli_ua (Remote closed the connection)
08:36:50Bg3rvery simple one ...
08:37:04JdGordonamiconn: umm.. i guess i put _remote in without even thinking about it.. it makes remote selection simple tho
08:38:08amiconnHmm, in fact there might be a reason when there is more that one possible remote
08:39:02amiconnI thought it could be simply possible to chain all _remote[] lists with their respecive _main[] list. Then switchin gthe remote type just means changing the 'entry point'
08:39:19amiconnBut that doesn't work in case the main list is chained itself
08:39:47JdGordonalso, it adds alot of unneeded list searching
08:40:13amiconnYes, but not doing it adds more code.
08:40:28amiconnI guess it has to be decided on a case-by case basis
08:40:59amiconnNo extra context table for remotes with few actions and where code size is important, separate tables otherwise
08:41:28JdGordonyeah, also, like you said, remote contexts arent even needed if the remote doesnt have a screen
08:41:58JdGordonwhich means we can add a #ifdef NB_SCREENS >1 around that part of action.c
08:42:07amiconnIt doesn't depend on the screen least how I understand it now
08:42:37hotwireany objections to using numpad 7 and numpad 1 for hold and remote hold respectively?
08:42:46hotwirein the UI Sim?
08:42:53JdGordonoh.. sorry.. it _can_ depend on the screen... ok, leave it as it
08:43:36amiconnRemote contexts are (1) needed for proper switching if there's more than one possible remote. (2) desirable to keep list searches down if the remote has many possible actions and code size is not a major concern
08:44:12JdGordonyeah, sounds correct
08:45:39JdGordonthere is a bit of wasted code in action.c to deal with it... should we add a #define into the target to enable the contect_remote stuff?
08:46:37JdGordonBagder: is it possible to put a script on the web (passworded of course) so anyone can fix the stuck builds?
08:47:17Bagderwell, the same effort could be put into making such a thing automatically
08:47:42Bagderbut yes it is possible
08:48:20 Join theli_ua [0] (
08:49:23 Quit theli_ua (Remote closed the connection)
08:51:32JdGordonamiconn: do you think thats worth doing?
08:51:51 Join midgey31 [0] (
08:55:00 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:56:03midgey31Anyone know why some of the modified files don't appear under "where" on the Since 2.5 page? (keymaps, mpegplayer, etc)
08:56:13JdGordonbg3r makes a god point.. does the default font have Âľ ? because if not then changing that to u would fix eerything once and for all :)
08:56:37Bg3rmaybe good, but god ... :D
08:56:55Bagdermidgey31: I'll investigate
08:57:22midgey31the commit times differ by two hours as well but that's hardly important
08:57:25Bg3rmidgey31: yeah, i saw that too
08:57:26 Quit ranunculoid (Read error: 104 (Connection reset by peer))
08:57:58 Join einhirn [0] (
08:57:58 Quit einhirn (Client Quit)
08:59:28JdGordonamiconn: ?
09:00:19 Join pondlife [0] (
09:01:36 Join finn_ [0] (
09:01:51 Quit midgey34 (Read error: 110 (Connection timed out))
09:01:54amiconnJdGordon: That looks like soft keylock won't work w/o a remote context
09:02:07amiconnAlso, where would HAVE_REMOTE_CONTEXT be defined?
09:02:27pondlifeIs it ok to attach example audio files to Flyspray reports? I'm thinking copyright restrictions would make this a bad idea.
09:03:02JdGordonsoftlock isnt actually checked untill after the do loop so it should be fine, and i guess in the config-target.h ?
09:04:57amiconnThe code after the loop doesn't eat the action if softlock is enabled and there's no remote context
09:05:26 Quit Landus (Read error: 104 (Connection reset by peer))
09:05:40JdGordonah, that last #endif is 1 line too low then
09:06:53JdGordonthat doesnt look like it works now anyway...
09:07:29JdGordonamiconn: no, it does eat it... the HAVE_REMOTE_CONTEXT is split in two... it gets eaten between them
09:07:46JdGordonno, sorry, im a doofus
09:07:50JdGordonthats where it unlocks
09:08:39LinusNJdGordon: you really need to use a better text editor that doesn't mess up the files
09:09:03amiconnJdGordon: Forget that.. we'll need remote contexts for the archoses too :/
09:09:19amiconnReason: The wps keylock shouldn't lock the remote
09:09:21JdGordonit still removes a bit of code for targets with no remote
09:09:44Bagderok, I fixed both issues on the commits-since-2.5 page
09:10:05hotwirehow does the remote hold usually work? Can the remote be used while the main unit hold is on?
09:10:40LinusNhotwire: yes
09:11:02 Join _FireFly_ [0] (
09:11:35***Saving seen data "./dancer.seen"
09:12:34JdGordonLinusN: does this look good? it works (reimplement the shutdown splash, without adding any knew actions)
09:12:51amiconnJdGordon: Yes. We just don't need HAVE_REMOTE_CONTEXT
09:13:00amiconnAll targets with remotes need remote contexts
09:13:00 Quit Snyper (Connection reset by peer)
09:13:40midgey31so what's the general opinion on plugins that don't use the menu API? there seems to be quite a few of them
09:13:51amiconnWe could do another thing - but that needs discussion: We don't necessarily need to include the softlock stuff for targets with hold switches
09:14:06pondlifeHmm, the new tree view (using the selection widget?) doesn't like some characters...
09:14:34pondlifeI assume if WPS displays a char correctly then the tree view should too?
09:14:43pondlifei.e. it's not a font issue
09:15:26midgey31I'm of the opinion that all the menus should use the standard api unless it's graphics based (ala BrickMania)
09:16:32 Quit finn__ (Read error: 110 (Connection timed out))
09:16:33Bagdermidgey31: I think most of us agree with that
09:16:57 Join ender` [0] (i=null@
09:17:12midgey31hmm, ill look into changing the offending plugins later today then
09:17:59pondlifeWhich source code implements the selection widget? gui/list.c?
09:18:06midgey31Chessclock, Dice, Disktidy, Jewels, Minesweeper, Snake, Solitaire, Star, and Xobox seem to be the ones in need of changing
09:19:35*JdGordon bbs
09:19:40 Quit JdGordon ("Konversation terminated!")
09:20:25*amiconn just had an idea regarding contexts for plugins
09:21:08amiconnContexts common to several plugins could be put in the plugin library. One such context would be context_scroll_zoom[]
09:21:47hotwireLinusN: they are completely independent then?
09:22:52 Join theli_ua [0] (
09:23:06 Join Mikaelh [0] (
09:23:24LinusNhotwire: yes
09:23:33 Quit Mikachu (Read error: 54 (Connection reset by peer))
09:23:47hotwirek, thx
09:25:20 Quit theli_ua (Client Quit)
09:25:31midgey31Badger: anything you can do for this
09:25:51 Join JdGordon [0] (n=jonno@
09:25:57midgey31halfway in bubbles_checklevel() the whole page turns black on firefox
09:26:11midgey31(at least for me) and it works on IE
09:26:21Bagdernot my firefox
09:26:54midgey31hmm, I'm on Win32 here
09:26:58Bagderthere's nothing special in the html at that point
09:27:04midgey31must be their problem
09:27:11BagderI'm on on linux
09:27:13LinusNturns black for me too (win32 firefox)
09:27:27Bg3rhere too
09:27:35Bg3rif you scroll it slowly with the scroll button
09:27:46LinusNi used pgdown
09:27:51aliaskNot halfway, more like 5/6 way down
09:27:55Bg3rit just goes nuts
09:28:10Bagderyes, if I use the scroll wheel I can get it go nuts
09:28:20aliaskIt's been happening for ages, I noticed it maybe 6 months ago.
09:28:31Bg3raliask: yep, confirming
09:29:43Bg3rit seems that this happens in long files (?)
09:30:26 Join theli_ua [0] (
09:30:28aliaskDisabling CSS makes it work.
09:30:33 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
09:30:46 Join petur [0] (
09:30:56aliaskNot like you need it on those pages anyway.
09:34:58 Quit muesli__ (Read error: 104 (Connection reset by peer))
09:35:12 Join muesli__ [0] (i=muesli_t@
09:36:50pondlifeOK, screenshots on - this affects file browser, but not ID3/tagcache view...
09:39:07LinusNpondlife: i think you should display the filename and not the tag in WPS for a fair comparison
09:41:17pondlifeGood point!!
09:41:23pondlifeLeave it with me
09:41:28 Join bluebrother [0] (
09:41:48Paul_The_NerdG'night folks
09:41:51 Part Paul_The_Nerd
09:45:03*midgey31 leaves as well
09:45:13 Part midgey31
09:49:47bluebrotherJdGordon: I just posted some no more working buttons on h100 series to the forum thread
09:50:26 Join ryantrip [0] (
09:50:50bluebrothersome forum admin around?
09:51:40peturone just left
09:51:42ryantriphey, how do i set my scrolling so it doesnt skip a bunch when im browsing alot of files?
09:53:48ryantripdoes anyone know how?
09:53:53bluebrotherthe forum classifies me as "rockbox expert". Is it intended I have "delete", "modify" and "split" buttons there?
09:54:06bluebrotheras I haven't noticed them before.
09:54:20 Quit SereR0kR ()
09:54:34safetydanbluebrother, Paul_The_Nerd upgraded the access for developers
09:54:43safetydanso yes, you get those abilities
09:54:50peturI seem to have gained those as well
09:55:10bluebrotherok. Nice to hear it isn't one of those forum glitches ;-)
09:55:22ryantripdoes anyone have roxbox on an ipod?
09:55:36 Quit Farpenoodle ("Large sacks of cheese")
09:56:06Bagderoops, I'm just a "global moderator" not an actual "administrator"
09:56:08safetydanryantrip, you can't really, it's a performance issue
09:56:57ryantripsafetydan: hmm i read in the forums that its cpu intensive? is that the problem?
09:57:07safetydanryantrip, yes
09:57:47ryantripsafetydan: is it in every version?
09:58:46ryantripso i cant fix it with scroll accelerating?
09:59:02Bagderthere is no current fix
09:59:16Bagderbut you're welcome to work on one! ;-)
09:59:41JdGordonamiconn: shold i commit those remote $ifdefs in action.c?
10:00:10ryantripbadger: lol maybe when i learn c
10:00:23Bagderok, start tomorrow then!
10:00:33ryantripahh rockbox just froze lol
10:00:44ryantripprobobly because im using a custome one for the vista theme
10:00:51ryantripanyone else have it?
10:01:32 Join pixelma [0] (i=pixelma@
10:01:50ryantripbadger: how do i browse though my music without having to skip so much?
10:02:20BagderI guess you need to scroll slower
10:02:51ryantripdang is so slow though :P there should be a plugin invented
10:03:36ryantriphow do you deal with it?
10:05:31safetydanNot use an iPod?
10:05:53ryantripdang, i dont have any other mp3 player
10:07:07ryantripsafetydan: do you theme your box?
10:09:56 Join finn__ [0] (
10:10:08ryantripbagder: are there any mods to use the buttions to move up and down?
10:11:15safetydanryantrip, eh? I just use the iAmp WPS that comes with Rockbox
10:11:19Bg3r haha
10:12:00 Join SereR0kR [0] (
10:12:02ryantripsafetydan: have you seen the vista media play one?
10:12:32safetydanno color on my player
10:12:45ryantripoh that stinks lol
10:13:11safetydanno, that's readable even without the backlight :)
10:13:36ryantripsafetydan: last one on the page
10:16:52 Quit SereR0kR (Client Quit)
10:17:06 Quit JdGordon ("Konversation terminated!")
10:18:41 Join ze__ [0] (i=ze@
10:18:45safetydanway too much eye candy
10:20:24ryantriplol i like it, im using it now
10:21:05ryantripim going to show it off to my friends at school when school starts up. they wont know what the heck i did
10:22:19peturthat eye candy shouldn't slow down much - peakmeters, EQ, crossfeed,... are worse
10:23:01peturSlasheri ?
10:23:05BagderBUTTON_REMOTE isn't defined for non-remote targets, is it?
10:23:54Bg3rBagder: it is 0
10:23:56 Quit finn_ (Read error: 110 (Connection timed out))
10:23:58Bg3rdefined to 0
10:24:24Bg3rand on targets with remote, it's the bitmask of all remote buttons
10:24:27Bagderah yes, I see that now
10:24:41Bg3rhaha for nothing
10:24:53 Join mkey [0] (
10:26:00*Bagder is more or less just a Rockbox code reviewer these days
10:26:35peturcode review is good
10:27:09Bagderyes I agree
10:27:32ryantripBagder: then why dont you write a less cpu intesive scroling system for the ipods?
10:27:33Bagderalthough it feels a bit like back seat driving after a while
10:27:44Bagderryantrip: please do
10:27:56Bagderthat would be tomorrow
10:29:05Bagder"why dont you"-questions really can't be answered in any other way
10:29:13Bagderof course we want everything perfect always
10:29:37ryantripBagder: you have an ipod correct?
10:29:45ryantripoh ok
10:29:46 Quit ze (Read error: 110 (Connection timed out))
10:29:47 Nick ze__ is now known as ze (i=ze@
10:29:59ryantripmy music keeps on freezing after 20 minuts
10:30:03Bagderarchos recorder, iriver h1x10, iaudio x5 and sansa e260 only ;-P
10:30:57ryantriplast time i got an ipod :P
10:31:15ryantripunless the next ones are outstanding
10:31:22daurn|laptopBagder: would you consider getting a new archos player?
10:31:36daurn|laptop(402, 404 or 604?)
10:31:44Bagdernot really
10:31:54 Join mirak2 [0] (
10:31:54Bagderbut that's because I don't need any other player
10:31:55daurn|laptopno one knows anything about the 404 & 604
10:32:04daurn|laptop402 i should probably work on a rockbox port
10:32:11Bagderand I have a player without rockbox already to work on
10:32:13pondlifeAnother broken keymap...? On H300 sim, selecting Reset Settings displays the confirmation screen briefly then auto-cancels...
10:32:49Bagderdaurn|laptop: no one knows what? aren't they tms320-based too?
10:33:19 Join finn_ [0] (
10:33:30daurn|laptopno one even knows the arch
10:33:31Bagderdaurn|laptop: obviously no gmini/av owner ever wanted a Rockbox port badly enough
10:33:45Bagderthen they are just not interested enough
10:33:45daurn|laptopno scans - no chip specs -nothing
10:33:58daurn|laptop - they came out a week ago
10:34:09Bagderbut I would be surprised if they aren't tms320
10:34:24daurn|laptopthey're older models were dm320
10:34:34daurn|laptopdunno if they would have kept the chip
10:34:53BagderI'd guess they do
10:35:12daurn|laptopif you guided me, i could make a gmini 402 port
10:35:12ryantripso, which is the best player from rockbox?
10:35:16Bagdersince it's powerful enough to do quite a lot
10:35:43daurn|laptopBagder: could you help me with a new port?
10:35:43Bagderdaurn|laptop: we're a whole team who'd help guiding you, but you'd need to drive it since you have the hw
10:36:03BagderI'd love to
10:36:19daurn|laptopi'll have a go
10:36:34daurn|laptopbut, first, i have a HUGE urge to eat chips, gotta go down to the store
10:36:56*Bagder tries to understand the inside of the PCI maze of the Linux kernel
10:36:56 Quit mirak2 ("CGI:IRC (Ping timeout)")
10:36:58 Quit apo (Read error: 110 (Connection timed out))
10:38:08 Join webguest48 [0] (
10:38:33 Join senab [0] (
10:39:33 Quit webguest48 (Client Quit)
10:39:54 Join webguest84 [0] (
10:40:10senabi was thinking today, why is 'Show ID3 info' called that?
10:40:28senabsurely 'Show Tag Info' is better considering not every format uses ID3
10:41:51midkayunrelated-ish: do we actually support non-id3 tags of any kind?
10:42:23Bagderyes on non-mp3 files
10:42:41midkaywhich? ogg, right?
10:43:13senabflac aswell
10:43:26midkayk :)
10:43:31 Join JdGordon [0] (
10:43:44senabsame as ID3 Database when selecting TagCache
10:43:55senabmaybe Tab Database is better
10:44:15BagderI've never liked the name 'tagcache'
10:44:20BagderI find it grossly misleading
10:44:25dwihnoLots of keyboard changes lately in cvs
10:44:27LinusNme too
10:44:44LinusNi still fail to see what makes it a cache
10:44:53LinusNas opposed to a database
10:44:56senabyeh, it doesn't really give an insight into what it does
10:44:58 Quit webguest84 ("CGI:IRC (Ping timeout)")
10:46:34amiconnBg3r: It should be (and was before) ľs (Microseconds) not us (U-seconds??)
10:46:50LinusNsenab: to answer your question why it is called "ID3 info": it
10:47:07JdGordonamiconn: does the default font even shown the u sign properly?
10:47:09LinusNis simply because rockbox was MP3-only in the old archos days
10:47:18Bg3ramiconn: yes, i know, but it gets converted to utf-8 by my editor and i'm sure that me and JdGordon are not alone
10:47:22amiconnJdGordon: Yes. ľ is plain iso8859-1
10:47:40senablinus: aaahhh ok
10:47:49amiconnBg3r: Then *imo* it's a bad editor
10:48:09Bg3rit's just vim with unicode support ...
10:48:17 Quit finn__ (Read error: 110 (Connection timed out))
10:48:29ZagorTrolltech announces Qtopia Greenphone, a linux-based phone with open (programmable) firmware.
10:48:46markunBg3r: vim only converts if I want to
10:49:11markunDo you have some option set to autoconvert to UTF-8?
10:49:26amiconnBg3r: The ľ has been in debug_menu for years, and it never was a problem before Why is it now?
10:49:26aliaskIs there any reason Sound settings and Manage Settings (and even the settings part of Playlist Options) aren't moved into General Settings?
10:49:39Bg3roh, oh, okay
10:49:49BagderZagor: yeah, a cool looking phone I wanna get my hands on and hack
10:49:55midkaysound settings ought to stay main menu, but i think manage and playlist could be merged..
10:50:25senabzagor: it's not the first phone based on linux
10:50:25Bg3rmarkun: ":set encoding=iso-8859" ?
10:50:32aliaskmidkay: Why?
10:50:41Zagorsenab: of course not, but it's the first that allows you to write your own firmware
10:51:09midkayaliask: because sound settings are more important and frequently accessed and thus should stay at main level. just my opinion.
10:51:11ryantripcan you play videos in rockbox?
10:51:33markunBg3r: I use "set fileencoding=latin1"
10:53:09midkayaliask: sound settings are entirely different from other UI/behavior settings and should stay independent (also IMO).. you shouldn't lump things together too generally. sound settings, e.g. settings pertaining to what you hear, and general settings, all the other misc. rockbox settings.
10:53:13senabryantrip: theres a mpeg2 player in *very early* development
10:53:34LinusNimnsho, any editor that decides to automatically change the contents of a file is a bad editor
10:53:55ryantripsenab: is it on the daily builds?
10:54:05LinusNyou should be able to load a file and save it again without it being altered
10:54:39senabyes, it's only able to play m2v elementary streams at the mo (no audio) and can be opened as a viewer
10:55:06kidd3ckzcan ipod record sound yet?
10:55:11kidd3ckzor will it be able to?
10:55:22aliaskmidkay: It's just always been a big question to me, it seems like it should go in there for me. I never change anything in them - is it different on other targets (to H300)?
10:55:34midkay"in them"? sound settings?
10:55:34peturkidd3ckz: linuxstb is working on it
10:55:40aliaskYes, sorry.
10:55:56kidd3ckzsweet :)
10:55:59amiconnJdGordon: In case you're planning to work on converting more screens to the action system: Some keypad-*.c files will need ifdefing of some context tables
10:56:10 Quit eGen_ ("mýdlo končí ! ... ručník končí !")
10:56:14senabpetur: isn't recording through the headphone socket mono only?
10:56:15midkaya lot of people like to tweak.. *shrugs* i think of it as General Settings are rather set-it-and-forget-it "i want things to behave this way" but sound settings are much more subjective and they can often require a lot of tweaking.
10:56:22JdGordonamiconn: im not adding any more screens untill this lot works 100%
10:56:37JdGordonand why?
10:56:38amiconn(e.g. the recorder keypad will only have fm radio mappings for fm/v2, but not for v1
10:56:49Bg3rmarkun: do you happen to know how to comment lines in .vimrc ?
10:56:49JdGordonah, righto
10:57:00JdGordonBg3r: # i would guess
10:57:03amiconn..and the ondio keypad will only have mappings for fm radio and recordning for the ondio fm, but not the sp
10:57:04petursenab: there's also a true line-in via the dock connector iirc
10:57:07Bg3rJdGordon: it isnt ...
10:57:21amiconnSame will probably apply to the ipods when ipod recording gets committed
10:57:39preglowi wonder how much that phone'll cost
10:57:40midkayaliask: for example: display settings pertain to how rockbox works. playback settings as well. system settings change rockbox functionality. playlist options change rockbox behavior. those can all be lumped together under "general" settings.. sound settings alone affect the output of sound and they rather stand out from the other general settings.. to me..
10:57:50 Join loco [0] (
10:58:15Bagderpreglow: apparently they don't sell it stand-alone but together with a full development kit, so I'd guess it won't be cheap
10:58:32locoJdGordon: the text_editor plugin is yours, right?
10:58:33preglowoh well
10:58:47JdGordoncan anyone please test the pitch screen? bluebrother says it doesnt work in h300, i need to know about other targets also..
10:58:48senabpeter: forgot the dock has a line in/out
10:58:48preglowas long as it's not cheap enough to kill my hopes of not getting another phone until this one breaks :>
10:58:49JdGordonloco: yes
10:59:01aliaskI'm going to create a forum thread about it, I'm doing a bit of a rework to the menu code, and I'd like as much input as possible.
10:59:21midkayaliask: sure. i agree with the manage settings/playlist settings, in any case.
10:59:27locoJdGordon: it crashes for me, is that expected? ;)
10:59:40JdGordononly if it has long-ish lines in it
10:59:42senabi recently got a nokia n80, it's so good having a wifi connection in my house
11:00:03locoJdGordon: it also crashes when starting it without opening a file.
11:00:33locoJdGordon: btw, it's iriver h1xx, both sim and real player. segfaulting
11:00:48JdGordonok.. ill keep it in mind
11:01:12JdGordonthe bug is actually in the list code i tinhk.. not the plugin.. it doesnt like long lines (even though it shold cope with them)
11:01:41locoJdGordon: well, it crashed also if i just open the editor and move the joystick.
11:02:02locoJdGordon: i get 0x0808a3fc in utf8length (utf8=0x8 <Address 0x8 out of bounds>) at common/unicode.c:240
11:02:04JdGordonOH... this is after updating?
11:02:14preglowdoes that phone have a minijack out port? :>
11:02:34Bg3rGreenphone will be available mid-September in limited quantities only
11:02:43JdGordonignore that... ok, can u put up a bg in FS? im concentrating on the new button stuff first
11:03:03senabpreglow: u need an adaptor but yes u can use 3.5mm
11:03:17locoJdGordon: yeah, i'll do that. thanks
11:03:21 Quit illogic-al (Remote closed the connection)
11:03:29senabi'm listening to the stream on it right now :)
11:03:32locoJdGordon: so the backtrace isn't of any use?
11:03:43JdGordonit might be
11:03:46JdGordonput it in...
11:04:04daurn|laptopbtw, in other news, medios got sound working on the archos gmini 402 last night
11:04:24Bagderdaurn|laptop: that should indeed help your rockbox port!
11:05:27 Join Landus [0] (
11:05:50JdGordondoes anyone know how to check what speed usb is runing at in linux?
11:06:13BagderJdGordon: I use dmesg and check the message from when I insert the cable
11:06:20safetydanJdGordon, dmesg | grep USB
11:06:38 Part loco
11:07:06JdGordon[42951168.110000] usb 1-1: new high speed USB device using ehci_hcd and address 5 <- does that mean it should be 2.0?
11:07:25senabhigh speed is 1.1
11:07:36senab12mbits iirc
11:07:51JdGordonok, damn, that would explain my speeds then
11:07:55bluebrotherwasn't that full speed?
11:08:18senabinfact yes ignore me, hi-speed is usb2
11:08:24LinusNhigh speed is 2.0 iirc
11:08:40aliaskI love the irony - high speed > full speed.
11:08:42Bagderyeah, high is more than full (in this situation)
11:08:58bluebrotherstandard speed: 1.5MBit, full speed: 12MBit, high speed: 480MBit IIRC
11:09:04senabu beat me to it petur
11:09:12dstrekultra speed: ohsnapMBit
11:09:18bluebrotherI love this webchat lag :o
11:09:49peturgood thing the webclient is no longer banned
11:10:49JdGordonhmm... ok, im reading of the disk at 30MB/sec (usb attached IDE hdd), which is faster than my IDE drives.. so i guess theres the problem...
11:10:59dstrekso is there anything started by anyone for the e10 ?
11:11:37***Saving seen data "./dancer.seen"
11:11:40LinusNnot that i know of
11:11:53dstreknot that I can really see a reason for one
11:12:02Bagderthere's always reasons for rockbox
11:12:10dstrektrue enough
11:12:12aliask <−− input appreciated
11:12:15senabjdgordon: i don't get anywhere near that, more 15MB/s
11:12:39*JdGordon stops complaining!
11:12:45midkayaliask: thanks for voicing my view for me. :)
11:13:00aliaskNo problems :)
11:13:19senabthats on the
11:13:41senab*thats on when writing though, for reading its more like 25mb/s
11:14:08JdGordonjdgordon, safetydan, senab, LinusN, midkay, aliask and 1 Guest are viewing this topic. <- lol
11:14:25aliaskYay people looking :D
11:14:25midkayhaha. always such a rush when someone links to a forum post.
11:14:48aliaskYeah, I'd like to see the stats on how often a page is accessed before and after it gets posted in IRC...
11:15:09amiconnmidka, aliask: 'Manage Settings' should imo stay separate. It's a different thing than the settings itself
11:15:32amiconn'Playlist Options' could be moved under 'General Settings'
11:15:46aliaskI think the other way round is more logical :/
11:16:02senabamiconn, i beleive manage settings should be in general settings
11:16:28LinusN"Manage settings" applies to both general and sound settings
11:16:39JdGordonaliask: you should write a revised menu tree in the wiki so we can all pick at it
11:16:46midkayamiconn: i wouldn't mind that in the least..
11:16:48LinusNJdGordon: amen to that
11:16:48aliaskGood idea actually.
11:17:02markunpetur: I was looking at the uboot source and perhaps we could use their usb code
11:17:02amiconn'Manage settings' doesn't set settings itself. It allows to store and load settings - and that includes sound settings, which also should stay separate because of frequent access
11:17:16*JdGordon goes back to wondering why the pitchscreen wont open...
11:17:21 Join TCK [0] (
11:17:28amiconn(although the latter is debatable since the introduction of wps context menu)
11:17:52aliaskThere was a bit of work done earlier about this, I'm just linking this for reference - it's a bit dated
11:19:24senabi like that aliask
11:20:32senabalthough time & date should be in general settings imo
11:21:35 Quit petur ("CGI:IRC (EOF)")
11:21:37 Join p3tur [0] (
11:21:50aliaskYes, it's not perfect.
11:21:51 Nick p3tur is now known as petur (
11:22:31JdGordoncan someone with the ipod please check if the you can access the pitchscreen from the wps (menu hold)
11:22:51 Join p3tur [0] (
11:22:51 Quit petur (Client Quit)
11:22:58 Nick p3tur is now known as petur (
11:22:59JdGordonwoops.. sorry.. ignore that.. ipod doesnt have a mapping for pitchscreen from wps
11:23:17 Quit senab (" Like's GUI? Then try HydraIRC -> <-")
11:23:35*petur kicks the webclient where it doesn't like it
11:24:02*LinusN suggests using a real client instead
11:25:39peturfirewall doesn't allow it
11:25:46petur(at work)
11:26:02*petur gets reminded he should be programming
11:27:18 Quit ryantrip ()
11:28:39LinusNpetur: use a tunnel
11:28:54 Quit finn_ ("""""" leaving """"""")
11:29:23LinusNpetur: does the fw allow ssh?
11:30:00peturI'm the windows fanboy, remember? I go for fancy gui apps :p
11:30:22LinusNthere are gui apps for ssh tunneling, for example putty
11:30:22amiconnpetur: rdp?
11:30:32amiconnThat's what I am using. Very convenient imo
11:30:43peturthe firewall only all allows what is needed - no idea until I try
11:30:48LinusNyou just use ssh to set up a tunnel, then you can use any irc client
11:31:07peturI have no experience at all in this domain
11:31:29*LinusN goes to eat lunch with bagder
11:31:44 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
11:32:17tucozbluebrother, around?
11:33:05aliaskI use putty at school to bypass the school's filter and credit system :D
11:33:22aliaskAnyway: I've created a wikipage for people to fiddle with -
11:33:40*aliask goes to eat dinner
11:36:13 Join barrywardell [0] (
11:39:03bluebrothertucoz: a bit.
11:39:36tucozgood. Do you have an idea how we could use certain commands in a different context?
11:40:13tucozAs we would want to include all commands (some with the same name) to the entire manual
11:40:23JdGordonare you guys talking abou the manual? if your lucky (and your good with perl...) you should be able to take the keymap files for the target and get it to dump it straight to tex for you
11:40:44tucozJdGordon, I do it the hard way :D
11:40:53JdGordonok :)
11:41:00tucozat least as a test
11:41:46bluebrotherI don't think we could simply "dump" the keymap file.
11:42:07bluebrotherMaybe we could convert them with a little perl script but I'm not sure.
11:42:42bluebrothertucoz: what do you mean exactly?
11:42:45tucozI think that is doable. But, it will not be too much work as long as we have set up one action map one target
11:43:43tucozok, lets say we have two ActionTreePgLeft, where one is \ButtonLeft and one is \ButtonRight (silly example).
11:44:04tucozone is used under a tree context, and the other is used in a settings context
11:44:11bluebrotherah, that case. I looked into it and was also thinking about it.
11:44:35bluebrotherit is?
11:44:51tucozno, that was just a (silly) example
11:45:23bluebrotherbut as far as I understood the actions are always the same buttons. Aren't they?
11:45:27amiconnJdGordon: WPS keylock doesn't work on Ondio, and according to the code also doesn't work on Recorders
11:45:49JdGordonis this after the recent change? or always?
11:45:55bluebrotherso an action "page up" is always the same button in a defined context.
11:45:58tucozbluebrother, no I think I found places where that was not the case
11:46:04tucozhold on
11:46:15bluebrotherso we could just add the context to the name, like ActionWpsNext
11:46:25tucozyes, that is done
11:46:35JdGordonamiconn: ? and by not work does it lock and not unlock? or not lock at all?
11:46:56amiconnJdGordon: Doesn't work since actions, and not at all
11:47:06amiconnThere's no definition for the lock combo
11:47:23amiconnAnd we need a way to unlock programmatically
11:47:24tucozbluebrother, ACTION_STD_CANCEL is normally BUTTON_LEFT or BUTTON_OFF but in the quickscreen it is BUTTON_MODE
11:47:51JdGordonhmm... im sure i put them in... im actually starting to think the wrong version of the patch got commited..
11:48:08JdGordonamiconn: programtic unlock? its easy, but why?
11:48:33bluebrotheroh. Why isn't that calles ACTION?QS?CANCEL?
11:48:42bluebrotherreplace ? with _
11:48:57bluebrothermaybe I should stick to only one keyboard layout ;-)
11:49:02amiconnYou might lock buttons in the wps. Then e.g. playback stops, and you're dropped to the browser. The buttons were unlocked automatically before. Now they aren't...
11:49:03tucozbluebrother, it is. Forget me
11:49:14tucozno, it isnt'
11:49:21JdGordonamiconn: haha, ok, good reason :)
11:49:33tucozit is ACTION_STD_CANCEL
11:49:43bluebrothertucoz: you're right, I looked into the code.
11:49:44amiconnIt might even be that in this case you can't unlock at all now, since the new context doesn't define a lock combo
11:49:57bluebrotherBut we could simply call it ActionQsCancel
11:50:09tucozJdGordon, why is ACTION_STD_CANCEL not ACTION_QS_CANCEL in the quickscreen context?
11:50:52bluebrothersimilar in the bookmark context, it uses ACTION_STD_OK
11:50:55tucozbluebrother, yes. So, we simply use the scheme ActionContextWhatAction for the
11:51:08JdGordontucoz: umm.. does the std key work in all targets?
11:51:22JdGordonyour proably right tjo
11:51:30bluebrotherthat was my idea ... hadn't noticed that out of context mappings before.
11:51:48tucozno idea. Was just wondering about the naming as the rest of the qs stuff has a QS in the middle
11:52:04bluebrotheralso, I thought maybe it would be nice to start with converting only STD and WPS contexts as a first step.
11:52:23tucozsure. That is done for the h1xx and h300
11:52:25amiconnMaybe it's even better to keep the buttons locked, but then the unlock combo needs to be stored in a way that the action code completely ignores context as long as the buttons are locked, and waits for the exact same combo which was used to lock buttons for unlocking
11:52:44*amiconn hopes that was understandable :/
11:52:56JdGordontucoz: oh sorry, misread you... what i should have said was hopefully the std cancel button would work in the QS on all targets so a QS_EXIT isnt needed
11:53:19tucozI see
11:53:24 Join jhMikeS [0] (
11:53:32JdGordonamiconn: i tinhk that is doable
11:53:54jhMikeSI see the build system is back up :)
11:55:53*JdGordon is considering throwing his h300 at the freeking wall.. bloody on button combos are not working :'(
11:55:55bluebrotherJdGordon: but isn't that a bit ... inconsistent? At least you have an entry in the keymap table, so does this save anything?
11:56:56jhMikeSWhat went wrong with it after my commit? I'm curious if anyone here knows.
11:57:29JdGordonbluebrother: na, its more consistent because the std off does exit (off in h300), but the addition of the menu button is to add convienicne
11:57:30amiconnjhMikeS: It has nothing to do with your commit in particular
11:57:48jhMikeSamiconn: I'm still curious
11:57:51 Join Arathis [0] (
11:58:16amiconnIt just hangs from time to time. Someone with central server access still needs to figure out why
11:58:30barrywardellJdGordon: thanks for adding the H10 patch. Unfortunately it got broken by amiconn's commit so I'll fix it up and make a new one later today when I have the chance
12:00:15amiconnbarrywardell: Just 'const' all the button mapping structs and remove the preconditions from multi-fire 'repeat' actions
12:00:16*jhMikeS made a joke last night about JdGordon providing the 30 story building that he dropped his iPos off and now notices JdGordon wants to throw an H300 at the wall.
12:00:27*JdGordon saw
12:01:11JdGordoni tihnk i found the problem.. gwps.c doesnt include "config.h" !
12:01:51bluebrotherJdGordon: I still don't get it. Wouldn't calling it ACTION_QS_CANCEL better as it clearly indicates it's a quickscreen action?
12:02:11*jhMikeS hopes JdGordon didn't think it inapropriate...just kidding
12:02:28barrywardellamiconn: so things that fire multiple times when the button is held down get no preconditions and things that fire just once on a long hold get preconditions?
12:02:44 Join MadDog011 [0] (n=MadDog01@212-200-212-95.adsl.sezampro.yu)
12:02:50JdGordonthe idea of actions is that there are standard combinations, and the screens shuold always use them, but if a certain target wants to extend a screen (with its buttons) it can..
12:02:53 Join eGen_ [0] (
12:03:07JdGordonjhMikeS: I have a sense of humour.. so i didnt care
12:03:27JdGordonbesides.. you couldnt outdo amiconn's award he gave me :D
12:03:34bluebrotherhmm. But aren't you kindof redefining a standard action that way?
12:03:40jhMikeSJdGordon: What's that?
12:04:09JdGordonjhMikeS: cant rememebr the wording.. something like "bigest badness award" or sometthing
12:04:19JdGordonbluebrother: no, not redefinng... extending..
12:04:40bluebrothermeaning it's an additional key for that action?
12:05:18bluebrotheranyway, I guess I need to grab some time this evening reading the action stuff completely and not only snippets ;-)
12:05:48JdGordonoh bloody hell... HAVE_PITCHSCREEN is not defined in the config-h300.h .. no wonder it doesnt work!
12:06:37jhMikeSSomeone didn't let me remove my own task related to the commit. I left things alone after the hang and went to do it this morning and it's gone.
12:07:21*jhMikeS gonna go make coffee
12:08:41preglownothing you could do about it anyway
12:09:58 Quit mkey (""Welcome to IRC; Where men are men, women are men and little girls are FBI agents!"")
12:10:28preglowis there some way to stop using the settings list scheme when dealing with large lists over integers?
12:10:32preglowit just looks silly
12:11:10safetydannot without adding more code
12:11:40safetydanI got used to it in the end. The consistancy is nice.
12:11:54safetydanAnd it's basically the same as having a slider for the numeric value.
12:12:42amiconnbarrywardell: correct.
12:13:00preglowit just looks silly with a great giant list of values from 0% to 200%, heh
12:13:15preglowi'll survive, though
12:13:26JdGordonthe idea is to move away from #ifdefing on KEYPAD_<target> or screen or something and move to HAVE_SOME_SCREEN right?
12:13:56 Quit tucoz ("Leaving")
12:13:58amiconnpreglow: I agree that it looks a bit odd, but it also has an advantage: You can immediately see the step size
12:14:47 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
12:14:59 Quit dan_a (Read error: 110 (Connection timed out))
12:15:11amiconnsafetydan: One thing that still annoys me with the new settings method is that Left is 'cancel' instead of 'accept' on some (most?) targets
12:15:27tucozbluebrother, now the irivers and ipods are finished. I'll wait with the ondios and recorders til later
12:15:38safetydanamiconn, yeah that's still driving me batty too
12:15:55amiconnJdGordon: HAVE_SOME_SCREEN is a good idea imho
12:16:26JdGordonhopefully ill finish this pitchscreen fixing before house starts in 20min :p
12:17:11 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
12:18:39daurn|laptopBagder back yet
12:18:46JdGordondamn the pitch screen is hard to control!
12:18:55 Quit amiconn (Nick collision from services.)
12:18:55 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
12:24:33 Join PyromancerX [0] (
12:25:19JdGordonwhat is the point of l/r in the pitch screen only being +-2%? can anyone actually tell the difference with such a small change?
12:27:20*midkay ..
12:27:30midkayhm, that's supposed to be like "me..." but..
12:27:37midkayit doesn't really come across as it, i guess.
12:28:54tucozJdGordon, like a nudge on a record player?
12:29:13tucozfor beat matching I guess
12:29:38midkayit's very noticeable going from 100% to 102 or 98, but not maybe from like 168 to 170 or 166...
12:30:04midkayit'd be nice if it were more exponential.. 1% adjustments for the first 10% or something, then go up to 2%, later 5% maybe..
12:30:54tucozI agree. That sounds like a good idea.
12:30:59preglowJdGordon: sure i can
12:31:13preglowit's not even hard
12:31:17JdGordon.. ok.. i get it :D
12:31:28preglowdepends heavily on the material as well
12:31:35daurn|laptopmidkay: thats what a percentage IS
12:31:46preglowdaurn|laptop: no, it's not...
12:31:54midkayit's easy to hear the slightest change when they're absolutely even output levels, but when you get to the point where one is much louder than the other it's considerably harder.
12:32:00midkaydaurn|laptop: that doesn't make any sense.. what?
12:32:05daurn|laptop168 + 2% = 171.36
12:32:05midkaywhat is what a percentage is?
12:32:12*JdGordon back in an hour
12:32:17preglowmidkay: people are used to a linear scale there
12:32:18midkaythat's not what we're talking about...
12:32:32safetydanJdGordon, it's not on yet
12:32:33midkaypreglow: i don't mind, i'm just saying. it's harder to notice when it's drastically different.
12:34:13*jhMikeS would like to make the pitch control only alter pitch and not tempo
12:34:42jhMikeSThe pitch adjustment should move in "cents" or 1/100ths of a semitone. Thats how it is in the musical worlds
12:35:16jhMikeS2^(1/1200) steps in other words
12:36:06preglowjhMikeS: you forget that sometimes what you want is to change the tempo
12:36:28jhMikeSNo I didn't...that's why there should be tempo controll too
12:36:33preglowas a matter of fact, that's, despite the name of the screen, probably the primary reason for it being there
12:36:51preglowanyway, it's on my very, very distant todo list
12:37:00preglowcoding a good time stretcher isn't easy on such a restricted platform
12:37:13safetydanpreglow, in your eqgraph.c file you call fcos(), is that different from the fsincos() in the eq code?
12:37:30preglowah, shouldn't be, no
12:37:37preglowit just does cos, of course
12:37:38jhMikeSCan you convert to frequency domain first and shift that and go back to time domain?
12:37:48safetydanokay, thought you might have been using a lower precision cos function for speed
12:37:53preglowjhMikeS: in a way
12:38:07preglowjhMikeS: that's kinda how the phase vocoder method of pitch shifting works
12:38:15preglowsafetydan: hell no, not there, heh
12:38:35jhMikeSpreglow: I'm very green at anything audio so don't laught at me! :)
12:39:01preglowjhMikeS: this is a wee bit beyond your run off the mill audio algorithm as well, heh
12:39:11preglowjhMikeS: but anyway, an fft based approach will very probably be too expensive for us
12:39:22preglowwe'll probably have to use some granular synthesis based approach
12:39:29preglowand they sound very crap
12:39:33jhMikeSpreglow: Don't want to lost audio quality.
12:39:48preglowwell, then i'd say forget it
12:39:52preglowwe don't have enough cpu power
12:40:00preglowtime stretching/pitch shifting isn't a trivial thing to do
12:40:33jhMikeSI can tell as it's a load even on a good pentium.
12:43:00 Quit Pyromancer (Read error: 110 (Connection timed out))
12:43:11jhMikeSpreglow: Is granular synthesis that method where the audio is chopped into little bits and then respliced?
12:43:58*jhMikeS hates the way that sounds except on halloween
12:44:04aliaskI'm pretty sure that's how the iPod firmware does it. I don't think it sounds that bad, until you start speeding up too much
12:44:21 Quit MadDog011 (Remote closed the connection)
12:44:22barrywardellJdGordon: if you see this when you get back, here's the updated H10 keymaps file:
12:44:53 Part barrywardell
12:44:56preglowi think the ipod firmware does some seriously simple-ass time stretching
12:45:14jhMikeSIf you stretch out time you can hear the repeating splices.
12:45:41preglowyes, that's a bit too simple
12:46:36jhMikeSnot fft (freq), not granular (time), what's left then?
12:46:53preglowgranular's pretty much the only choice we have
12:47:09preglowwith a highly optimised fft, we might have a chance
12:47:16jhMikeSugh...everythings gonna sound like darth varder
12:47:25preglowbut it'll suck battery like mad, and you wouldn't be able to pitch stuff up too much
12:47:32preglowno, that's a vocoder :>
12:47:54preglowno, it's not
12:47:57jhMikeSpreglow: I don't but that granular thing can sound awfully like it
12:48:00preglowyou're right! :D
12:49:16jhMikeSMaybe the best place to use such a thing isn't when you're at the beach but if you're transposing for music or doing karaoke or something where the unit can be plugged...I dunno
12:49:23preglowthat would be my first shot at it
12:49:27preglowunfortunately, i don't much care about it, heh
12:50:17jhMikeSpreglow: It's a task that would teach me something in its own sake
12:50:55preglowlots of links there
12:51:05preglowthere's a lot of info on the web about this
12:53:17LinusNpreglow: but most info ends up in "and you can buy the closed-source implementation from us"
12:54:37LinusNvery little documentation with actual algorithms described
12:55:04xorAxAxhave you looked into audacity's etc. code?
12:55:10safetydanhrmm... for this graphing thing to work the dsp will have to expose the eq coefs
12:55:24 Join keilj [0] (
12:55:37keiljhello all
12:56:21 Join mantono [0] (
12:56:25jhMikeSTo start with maybe I should make the pitch shift work in cents or if some is working on the screen, that would be a more sensible scale.
12:57:23LinusNprobably, but i believe most people use it to speed up the playback rather than transpose it
12:57:41midkaysafetydan: can i ask what you're up to? eqgraph.c/"graphing thing".. sounds interesting. :)
12:57:53LinusNxorAxAx: no, i haven't
12:58:04safetydanmidkay, just attempting to hook up preglow's eq graphing code
12:58:31midkayeq graphing code? you mean hooking the graphical EQ up to the hardware eq?
12:58:35jhMikeSThe pitch and tempo aren't percieved linearly so there's too much resolution near the 100% mark and not enought at the bottom/top
12:59:16safetydanmidkay, something like this
12:59:17preglowsafetydan: got it to work at all?
12:59:29midkaysafetydan: ah, i see. cool. :)
12:59:37safetydanpreglow, well it compiles, but no not actually attempted to feed it data yet
12:59:50preglowsafetydan: i've got a test program somewhere around here too
12:59:52preglowi can see if i can find it
12:59:55LinusNjhMikeS: agreed
12:59:57safetydanthat'd be handy
13:00:03jhMikeSWith 2 cents per step youd have 1200 levels instead of 1000 for -/+ 1 octave, just change from % to semitones
13:00:13preglowLinusN: yeah, but there is source code hanging around
13:00:21preglowthe phase vocoder method is pretty well described
13:00:34 Quit mantono (Client Quit)
13:01:06LinusNi guess you can use a pretty simple algorithm just to be able to speed up audio books
13:01:17preglowyou can indeed
13:01:20preglowpsola is good for that
13:01:34preglowwhich is granular synthesis with pitch estimation
13:01:38xorAxAxcan the ipod slowdown audio books?
13:01:58preglowsafetydan: i've found some code, but it's a bloody mess...
13:01:59tucozMaybe there is a open source filter for this. Like buzz machines or a free vst filter or something like that.
13:02:26preglowthere is
13:02:33preglowcsound probably has algos
13:02:36safetydanpreglow, still probably worth a look
13:03:44preglowthe dsp community isn't too big on sharing source code for some reason
13:03:51 Quit keilj ()
13:04:06tucozIt would be interesting to see it work on an actual target. I have often been astonished by how well stuff works on rockbox.
13:04:51LinusNpreglow: exactly, people are very protective with their knowledge
13:05:15preglowmost dsp people think they'll make it big with their knowledge, or something
13:05:18preglowwhich they mostly don't
13:05:21preglowit's a very competetive market
13:05:37preglowtons of independent people selling stuff
13:05:38tucozMaybe there is some articles available. The academic community usually are quite good at sharing their results
13:05:54preglowtons of articles, but the best ones aren't always available
13:06:01preglowthe aes ones in particular have to be bought
13:06:15tucozaes, like in crypto?
13:06:23preglowaudio engineering society
13:06:33safetydanpreglow, do I all eq_magplot once per band and sum the resulting graphs?
13:06:37safetydancall even
13:06:40preglowsafetydan: yes
13:07:03safetydanman this is going to be slow
13:07:24preglowit needs SDL and SDL_gfx
13:07:31preglowsafetydan: indeed
13:07:47preglowsafetydan: i didn't mean it to be used in a very realtime like fashion
13:08:09safetydanno, I can't imagine it's easy to be realtime with this sort of thing
13:08:48safetydanAnyway, I can't get that URL. I get 403 Forbidden.
13:08:49preglowsafetydan: there
13:09:22preglowbut yes, it'll be slow
13:09:42preglowi can also derive expressions that are hardwired to the eq filters, i guess, but i don't know if it'll benefit us much
13:09:59preglowthe current routine works for all second order iir filters
13:10:14preglowwhich i also gathered is a nice thing to have in case one wants to plot filters other places
13:10:43safetydanHave you tried it on a target?
13:11:28 Quit theli_ua (Remote closed the connection)
13:11:40***Saving seen data "./dancer.seen"
13:12:11preglowthere are two ways to deal with the plot mixing: either you multiply the linear spectrums of each band together, then calculate the db values
13:12:21preglowor you calculate db values for each spectrum, then add them together
13:12:25preglowthe first one should be cheaper by far
13:13:57safetydanI'm going to pretend I understood that and press on regardless :)
13:14:40amiconnsafetydan: Will you use the cording sine/cosine routines instead of the table based one?
13:14:45preglowgist of it: you need to combine the plots by multiplying them
13:14:52preglowamiconn: yes
13:14:56preglowthe table based one is gone
13:15:52amiconnlog/exp could also use another algorithm (the one I use in the grayscale lib for gamma correction)
13:16:05safetydanlog/exp can be done with cordic as well
13:16:12safetydanbut whether you want that is another issue
13:16:23tucozOT: does anyone know how to get remove the apt lock? Adept updater crashed and I killed all adept processes, but the lock is still there.
13:17:23tucozhmm, there were still a stray process that I killed and it works now.
13:17:30aliaskNever mind then :)
13:18:49*amiconn wonders whether it would be manageable to do the grescale -> bitpattern conversion in realtime (grayscale lib)
13:19:19amiconnIt would have several advantages to do it this way
13:20:16 Join theli_ua [0] (
13:20:21aliaskamiconn: Keep answering the questions as in my head before I ask them :)
13:20:39 Join SereR0kR [0] (
13:20:39tucozpreglow, what do you think about this?
13:21:54preglowit's good
13:21:56preglowbut expensive
13:22:08tucozthe source is available for download
13:22:18preglowafaik, it's based on jos' resampling algorithm
13:22:32preglowwhich is basically fir filtering by a sinc function you calculate as you go
13:22:35preglowfrom a lookup table
13:22:59preglowit's worth trying at some point to see how well we'll be able to hold up, but it'll be more expensive than what we have no by a very large margin
13:23:18tucozahh, expensive as in computing cost. I thought you mean $$$
13:23:26preglowheh, no
13:23:30preglowsrc is used all over the place
13:23:56preglowthe code can't be used as is anyway
13:24:10preglowor shouldn't, i guess
13:25:10tucozthere is also this:
13:25:29tucozthe pitch shifting using the fourier transform source code, that is
13:25:41preglowsrc doesn't do pitch shifting, if that's what you think
13:25:47preglowit just does ordinary high quality resampling
13:26:06tucozaha, I was looking for time streching
13:26:09 Quit muesli__ (Read error: 104 (Connection reset by peer))
13:26:34 Join muesli__ [0] (i=muesli_t@
13:28:09tucoznot really into the audio lingo
13:29:18 Join Rob2222_ [0] (
13:34:44*JdGordon back
13:34:58jhMikeSI guess the dspdimension thing does was I was thinking but it's loaded with floating point :(
13:35:03safetydanah, House must be finished
13:35:58Bg3rfinished as destroyed ? :D
13:36:04JdGordonwhats the funny?
13:36:16aliaskHe knew what was going on :)
13:37:25aliaskTV show
13:37:32*preglow dropkicks television
13:37:36JdGordonabout the only decent thing on tv atm
13:37:52aliaskYep, which is why I use bittorrent >_>
13:38:12 Quit eGen_ ("mýdlo končí ! ... ručník končí !")
13:38:34aliaskStill waiting for september 5th - new season of house :D
13:39:15 Join scott666_ [0] (
13:40:23aliaskDoes someone with commit access want to look at this: ?
13:41:54preglowamiconn: IBSS is zeroed out, yes?
13:42:26JdGordonnoone has a problem with chaning the pitchscreen so you have to have #define HAVE_PITCHSCREEN in the target config.h for it to be used?
13:42:59JdGordon.. replacing a check for !KEYPAD_PLAYER
13:44:04JdGordonamiconn: Linu*: Bagder: ? silense means no...
13:44:22BagderI agree to everything I haven't read
13:44:30JdGordonsounds good
13:44:57 Quit daurn|laptop (Connection timed out)
13:45:49 Quit Rob2222 (Read error: 110 (Connection timed out))
13:46:07 Join daurn|laptop [0] (n=quae@
13:46:15JdGordonhmm.. no.. this isnt going to work for ipod :'(
13:46:31daurn|laptop Bagder?
13:46:52*safetydan has angered the SIGFPE gods
13:46:55 Join ed__ [0] (
13:46:56bluebrotherjust discovered accidential next skips when playing files using "right" from the file browser
13:47:29JdGordondidnt accidental next do that before anyway?
13:47:33ed__hotwire: your using my registered nick, please change it :)
13:47:58bluebrotherJdGordon: no, as I usually play new directories using the right button.
13:48:07JdGordoned__: /msg nickserv ghost <nickname > <password>
13:48:24ed__JdGordon: i know, i'm was just trying to be polite :)
13:48:27JdGordonput u dont purposly accidnelty press right again
13:48:30preglowsafetydan: sigfpe? you're using the fpu? :>
13:48:40 Quit hotwire (Nick collision from services.)
13:48:47 Nick ed__ is now known as Hotwire (
13:48:53safetydanapparently it's related to divide by zero
13:49:01safetydanthere's no float code that I'm using anyway
13:49:04LinusNbluebrother: which platform?
13:49:14 Join _hotwire_ [0] (
13:49:17bluebrotherLinusN: h120
13:49:18Bagderdaurn|laptop: yes?
13:49:49bluebrotherbut it only occurs when playback is stopped.
13:50:55JdGordonhow should i rename the quickscreen context so it obviously applies to the pitch screen also? seen as the 2 scren have the same layouts
13:51:31 Join eGen_ [0] (
13:52:14LinusNJdGordon: but they don't have anything else in common, so why should they share the context?
13:52:32JdGordonbecause the screen layout is identical
13:52:56LinusNdoes the pitch screen have 3 option lists? i think not?
13:53:03 Part Hotwire ("Konversation terminated!")
13:53:18JdGordonok, it has 1 extra button.. which a few of the targets use in the qs anyway
13:53:25JdGordondown inverse
13:54:00LinusNmy point is that the context is about what the screen does, not what keys they happen to use
13:54:07jhMikeSNow when I select "right" to delete a file it shows the screen for a second and then it disappears.
13:54:28jhMikeSTHe screen not the file. I have to use select now
13:54:29JdGordonhow about CONTEXT_COMPASSPOINTS? and have n/s/e/w (n being up) that is used on these 2 screens
13:55:05 Join webguest33 [0] (
13:55:18webguest33just a question
13:55:20JdGordonjhMikeS: ok, that should be easy to fix...
13:55:34jhMikeSThe keys are acting wierd
13:56:04webguest33what's the use of the eeprom driver? Is there any benefit from flashing rocbox in epprom instead of bootloader/disk?
13:56:05jhMikeSWhy does it say I have to push "Play" to do the action but I really have to push "select"
13:56:19JdGordonwhich action?
13:56:23LinusNjhMikeS: historical reasons
13:56:30jhMikeSFor one: delete a file.
13:56:38LinusNand the language files haven't been updated
13:57:24JdGordonLinusN: conext_compasspoints is workable?
13:57:51jhMikeSWhat historical reason is there to tell people to use the wrong button? Or are you saying you changed it from play at some point and didn't update the lang files? Maybe the lang file should have the button names as separate words? If it does, ignore that part. :)
13:58:07 Join MadDog011 [0] (n=pera@212-200-212-95.adsl.sezampro.yu)
13:58:13LinusNJdGordon: my opinion is that they should be two contexts
13:58:29JdGordonthe lang file didnt always allow diffeent string for different targets
13:58:54LinusNjhMikeS: until not long ago, all targets shared the same strings
13:59:02jhMikeSLinusN: It does now?
13:59:18JdGordonLinusN: i think its a waste, it also goes aslightly against the making proting easier and consistancy....
13:59:25JdGordonbut if you still say no then ill live
13:59:36LinusNJdGordon: i only state my opinion
13:59:40 Quit muesli__ (Read error: 104 (Connection reset by peer))
13:59:56jhMikeSnever're saying different strings for different targets now?
13:59:58*JdGordon likes pleasing those higher in the food chain :p
14:00:03 Join muesli__ [0] (i=muesli_t@
14:00:06safetydanfear my... errr... preglow's eq graphing power
14:00:10Bagderthere's food? where!
14:00:32LinusNjhMikeS: yes, you can now have target specific strings
14:00:51jhMikeSHave the button name strings separated from the other text. The screen can then select the right string.
14:00:56*LinusN ph333rs safetydan's powers
14:01:39safetydannow if I could just figure out how to combine each band's graph...
14:01:55LinusNjhMikeS: there is a general "confirm with Play" string used by several screens
14:02:00jhMikeSsafetydan: eq graphing?
14:02:24LinusNall these screens use the same button
14:02:31safetydanjhMikeS, graph the effect of the equalizer settings basically
14:02:31jhMikeSLinusN: That's yucky and should be avoided IMO.
14:02:46LinusNno, that's called "consistency" in the ui world
14:02:54jhMikeSLinusN: But the tell tell me to use the _wrong_ button :D
14:03:02LinusNall you have to do is change the string, goddammit!
14:03:05preglowsafetydan: you add them
14:03:05jhMikeSThat's even worse.
14:03:09preglowsafetydan: ehh
14:03:11preglowsafetydan: multiply them
14:03:17 Quit pondlife ("byebye")
14:03:28preglowif the graph is still in linear amplitude format, at least
14:03:34jhMikeSLinusN: What if there's a screen that still needs "confirm with Play" :)
14:03:37LinusNjhMikeS: apps/lang/english.lang
14:03:51LinusNjhMikeS: then we use a different string for that
14:03:53safetydanpreglow, multiply and clamp?
14:03:55 Quit webguest33 ("CGI:IRC (EOF)")
14:04:05LinusNor change the ui so it uses the same button as the other screens
14:04:12jhMikeSI'm trying not to conflict with something that actually needs that string as it is. Ok fine I'll change it.
14:04:23*jhMikeS is a bugger sometimes.
14:04:44LinusNjhMikeS: check how it's done for the h100 and h300 targets
14:04:49preglowsafetydan: clamp shouldn't be necessary, afaik
14:04:50preglowlemme think
14:04:58safetydanpatch as is is here (note extreme roughness and hackishness)
14:05:13LinusNthere is also a patch in the tracker that adds more target specific strings for the irivers
14:05:16jhMikeSsafetydan: I had the idea to do that too BTW.
14:05:39safetydanjhMikeS, yeah the idea has been kicking around since the equalizer was implemented
14:05:44jhMikeSsafetydan: Is that part of the EQ screen or something separate?
14:06:08safetydanjhMikeS, at the moment it's just tacked on
14:06:18jhMikeSOH...we have an mp3 encoder why not use that for recording?
14:06:24jhMikeSAdapt it, rather
14:06:35BagderjhMikeS: we should, and there's a patch that does/did it
14:06:36LinusNjhMikeS: there is a half-ready patch for that too in the tracker
14:06:53preglowsafetydan: what, you've stuffed it in the eq screen already?
14:06:57jhMikeSHmmm...what's the other half?
14:07:18peturjhMikeS: many people who use rockbox for recording do it because it allows lossless recording
14:07:25LinusNwell, half-ready might have been a bad choice of words
14:07:35Bagderthe next half is probably making sure we can have proper encoding codecs
14:07:43jhMikeSsafetydan: I thought "there's no damn room on the eq screen for peak meters for each band".
14:08:13preglowsafetydan: ehh, now i notice that my code has a fixed shift in the DIV. have you removed that?
14:08:23jhMikeSpetur: Of course that option should exist too...encode to FLAC or is that too heavy?
14:08:24JdGordonwhats a good name for the left/right actions in the pitch screen?
14:08:48safetydangotta go unfortunately
14:08:49 Quit safetydan ("Ex-Chat")
14:08:59JdGordonnudge will do
14:09:07LinusNjhMikeS: the first thing we need to do (and i believe the patch addresses that) is to create a framework for encoders
14:09:55jhMikeSJdGordon: How bout +1/-1 Semitone? :))
14:09:56LinusNJdGordon: one argument against sharing contexts is that it makes it harder to customize the buttons
14:10:00 Join lodesi [0] (
14:10:23JdGordonLinusN: ye, i just realised it wouldnt work "as nicley" by sharing...
14:10:46JdGordonpitchs screen has 8 actions.. quick has 4
14:10:57 Quit Febs (Read error: 104 (Connection reset by peer))
14:11:00*juxtap sees the word "semitone"
14:11:34jhMikeSLinusN: I'm quite fuzzy on the audio portion right now but it's just the inverse of a decoder, send it PCM samples and get compressed samples out...does the framework have to be radically different?
14:11:50juxtaphow's the slow-me-down feature coming? anyone?
14:12:10LinusNjhMikeS: no, but it must be created, with dynamic loading of encoders etc
14:12:28LinusNit's not difficult or anything, it just needs to be done
14:12:59preglowit is partially done
14:13:02LinusNjuxtap: all these feature requests are slowing me down for sure
14:13:06jhMikeSYou'd need to set params before starting of course. But say with mp3 doesn't it need to know when things are starting/ending?
14:13:49preglowjhMikeS: look at that
14:13:59juxtapLinusN, i'm not nagging :p
14:14:12peturthe encode framework will also allow to write a 'mono' wav writer
14:14:17jhMikeSI think I should be able to specify the record to directory in my settings as well, more that just "use current"
14:14:19 Join Febs [0] (
14:17:00LinusNjhMikeS: i think many users (and devs) will thank you if you update and commit that patch
14:17:49peturwhat patch?
14:18:00LinusNthe encoder patch
14:18:21preglowshould we allow an encoder and a decoder do be active simultaneously?
14:18:38LinusNwould be nice for transcoding
14:18:47preglowwould also be a memory drain
14:19:04LinusNlet's not do that for now
14:19:05preglowanother plugin buffer and another malloc buffer
14:19:13jhMikeSLinusN: the record directory one you mean?
14:19:14Bagderlet's not go there yet
14:19:25JdGordonwhy a memory drain? couldnt the encode buffer use the plugin buffer and decode stays where it is?
14:19:30LinusNBagder: it's a silly place
14:19:38Bagderits only a model!
14:19:48LinusNjhMikeS: no, the encoding framework
14:20:08peturjust be carefull with touching the recording code - we're just getting *some* credibility amongst tapers, any instability will blow that to pieces...
14:20:18 Quit theli_ua (Remote closed the connection)
14:20:27peturmust be tested well before commit
14:20:28LinusNpetur: speaking of that, when will you commit the REC
14:20:33LinusNREP i mean
14:20:41peturtonight maybe
14:20:52jhMikeSI have tested everything before even show it to anyone.
14:20:58preglowpetur: taper's should really stick to one version they know will work
14:21:01preglowtapers <-
14:21:02peturwell only AGC for now
14:21:05LinusNaliask: recording enhancement pack
14:21:11 Quit Kitar|st (Read error: 104 (Connection reset by peer))
14:21:30aliaskI look forward to it :)
14:21:49peturthe AGC patch is commit-ready I think, should be easy to use on other targets too I hope
14:22:07LinusNtapers should use tape recorders, or call themselves something else :-)
14:22:18 Join Kitar|st [0] (
14:22:31peturMmmm can commit the rest of the REP ;)
14:23:09 Join theli_ua [0] (
14:23:59preglowi'm comfortable with calling them tapirs
14:24:12BagderI like tapirs too
14:24:28*petur is tapir wannabee
14:24:47*preglow envisions petur with a long snout
14:25:15 Quit Rob2222_ ()
14:25:44*aliask had to wikipedia that one...
14:25:56Bg3rhaha this word means something like "idiot" in the bulgarian slang :D
14:26:14Bagderwe should speak more bulgarian here!
14:26:23 Join mantono [0] (
14:26:29*Bagder stops being silly and goes blinking leds
14:26:34 Quit eGen_ ("mýdlo končí ! ... ručník končí !")
14:27:15jhMikeSLinusN: I should use that patch to get a clue about how things should work? Dumb Q: Where the heck is the decoder loading done?
14:28:02LinusNjhMikeS: i think that patch is more or less complete
14:28:24 Quit muesli__ (Read error: 110 (Connection timed out))
14:29:11jhMikeSok. but if less I'll come back here and discuss it. :)
14:29:13LinusNJdGordon: shouldn't we add an action_signalscreenchange() call before entering the WPS?
14:29:32Bg3ri think we should
14:29:33 Join Criamos [0] (
14:29:48JdGordonprobably... adding unneeded calls does nothing.. so better too many than not enough
14:29:57JdGordonyes/no screen needs on on entry also
14:30:07amiconnJdGordon: While the quickscreens and the pitchscreen share the layout, the buttons work quite different in both
14:30:11LinusNwell, it is a screen change, and we have reports of accidental track skips
14:30:48JdGordonLinusN: i meant adding calls doesnt effect it negativly, so its better to have them and not need them.... i was agreeing :p
14:31:00LinusNi know
14:31:18JdGordonamiconn: ye, addded a brand new context for pitchscreen, 6(7?) more keymaps files to fix :'(
14:31:32LinusNenjoy! :-)
14:32:06LinusNhmmm, maybe each screen should start with calling action_signalscreenchange()?
14:33:11Bagdersir yes sir!
14:33:15JdGordonLinusN: probably
14:33:20*Bagder straightens up
14:33:28daurn|laptophow should i do this port of mine
14:33:33daurn|laptopi don't want to brick
14:33:44daurn|laptopand i don't know how to do a bootloader
14:33:58LinusNdaurn|laptop: then you're in trouble
14:34:24Bagderso how do you run code on the unit now?
14:34:44daurn|laptopexploit in avi code
14:34:54daurn|laptop - use the official file browser
14:34:58daurn|laptop - pick an app
14:35:02daurn|laptopit launhes it
14:35:26Bagderand that's how you always start?
14:35:37Bagdercan you start that avi automatically on boot?
14:35:38jhMikeSLinusN: It's a big patch. Do I just need encoder04.diff ?
14:35:54LinusNjhMikeS: yes i think so
14:35:59daurn|laptop4 clicks to start it
14:36:08Bagderdaurn|laptop: anyway, I'd suggest you start your rockbox bootloader that way to start with then
14:36:10daurn|laptopright right down ok ok
14:36:40daurn|laptop(for first in list)
14:37:41*preglow does his first commit in ages
14:37:55*LinusN does the norwegian commit dance
14:38:05Mikaelhdoes it fix all present and future bugs?
14:38:19*petur goes to see how many points preglow scores
14:38:38JdGordonget a coffee first.. takes 5 min :p
14:39:10preglowMikaelh: of course
14:39:18*preglow creeps back under a bush in slottsparken
14:39:19 Nick Mikaelh is now known as Mikachu (
14:39:46LinusNMikachu blew his cover
14:40:39amiconnJdGordon: I doubt it's 6 or 7 keymaps to fix. The player has no pitch screen
14:40:44*petur thinks the commit of preglow should come with a free mirror (to watch your ears)
14:40:45preglowwhat cpu is in the gigabeat again? arm9?
14:40:51JdGordonamiconn: ok, 5 or 6
14:40:55Bagderpreglow: arm9 core yes
14:41:07preglowi wrote watch?
14:41:10preglowi meant 'mind' :)
14:41:23JdGordonexcluding the h10 and newtargte there are 6 keymap files
14:42:14jhMikeSLinus: The NULLs in the codec_api are entrypoins into the decoder to be filled in at init?
14:42:45preglowdoubt it, there's just one entry point
14:42:53preglowthe codec calls back to rockbox, not the other way
14:43:07LinusNjhMikeS: when the encoder is loaded, yes
14:43:35jhMikeSpreglow: how would you tell it to do anything without that? :)
14:44:09preglowthrough variables in the api struct, mainly
14:44:12LinusNwell, those are callbacks
14:46:09 Join Rob2222 [0] (
14:47:56LinusNthe encoder is actively pulling the pcm data and encodes it
14:48:18 Join perplexity [0] (
14:48:43LinusNand rockbox provides the enc_get_wav_data() callback to provide the pcm data to the encoder
14:49:21jhMikeSencoder = pull, decoder = push?
14:49:30 Join muesli__ [0] (n=muesli_t@
14:49:36aliaskEh, did anyone else get the CVS mail for preglow's latest commit?
14:49:55LinusNjhMikeS: i believe so, yes
14:49:57 Join dan_a [0] (
14:49:59 Join eGen_ [0] (
14:50:01Mikachuhasn't appeared on the mainpage either
14:50:06preglowit's still being compiled
14:50:13peturcan take some time
14:50:15preglowno :/
14:50:19aliaskIt's in the buildpage though...
14:50:20Mikachuthe cvs page says it's done
14:50:28*jhMikeS 's got steep learning curve here but will prevail.
14:50:33preglowwell, the new spotty build table certainly looks good
14:50:33*LinusN sees it on the main page
14:50:36preglowsomeone get rid of those warnings
14:50:40perplexity<beg> Updated bootloader for h3x0 </beg> ;)
14:50:56LinusNperplexity: updated with what?
14:50:59Bagderthose warnings really should be fixed
14:51:24peturh300 usb power patch should come first ;)
14:51:25JdGordonjust pull his comp from the sdl builds?
14:51:35perplexityWell, the remote autodetection now works.. the current bootloader won't start on the h300 non-lcd remote.. and also when you plug the charger in it still boots into the iriver firmware..
14:52:01BagderJdGordon: its all sim builds with gcc 4.1.x
14:52:13JdGordonoh righto
14:52:19Bagder... imho it should be fixed in the source code
14:52:47Bagderor possibly some odd gcc option if there's any
14:52:51 Quit Febs (Read error: 104 (Connection reset by peer))
14:53:04 Quit mantono ("Don't worry, be happy")
14:53:06 Join Febs [0] (
14:53:17Mikachuif it's only in doom/, you could add the right -W-no-...?
14:53:37preglowsource code fix sounds nicest
14:53:42preglowbut i don't really know what's wrong, heh
14:53:48 Join mantono [0] (
14:54:00Mikachujust click on a yellow box!
14:54:01LinusNperplexity: ah yes
14:55:08perplexityactually, I also still have issues with bootloader usbdisk mode, but that is something I don't find at all annoying.. whereas starting my car and having it boot into the iriver software coz I forgot to unplug it is driving me nuts ;)
14:55:22 Quit MadDog011 (Read error: 54 (Connection reset by peer))
14:56:19perplexityI actually attempted to build the bootloader myself, fully prepared to build a wiggler if I did something silly, but the code it builds seems to be too large for the space allocated in the firmware builder.. so I gave up before I broke something badly..
14:57:54JdGordonis there no config-x5.h ?
14:58:01 Quit Criamos ("( :: NoNameScript 4.02 :: )")
14:58:19JdGordongrr.. named config-iaudiox5.h
14:59:05 Quit mantono (Read error: 54 (Connection reset by peer))
14:59:06LinusNperplexity: yes, the bootloader has been too big
14:59:28LinusNi have a fix for that at home
14:59:32Bagderwhoa, those libmad warnings are frightening
14:59:52perplexityOh, which negates my next question then :)
15:00:10preglowBagder: they're bloody annoying
15:00:16preglowBagder: gcc 4.1 complains by default
15:00:20Bagderyou checked the code?
15:00:21preglowBagder: if you fix it, then gcc 4.0 complaions
15:00:29Bagderhow annoying
15:00:40preglowi belive it's the two dimensional array pointer thing
15:03:33Bagderof course we could do some weird #ifdef on gcc version
15:04:42Bagderthere was one warning that stood out
15:04:47Bagderd_main.c: In function 'D_PostEvent':
15:04:48Bagderd_main.c:128: warning: value computed is not used
15:05:11*JdGordon is gonna be anoying for a sec.. how did u access the pitch screen from the wps in all targets except iriver?
15:05:28LinusNON+up/down methinks
15:05:39Mikachuit's in the menus on ipods
15:06:20JdGordonoh gwps.h stil has the old key maps... dw
15:07:13JdGordon... or not... there is no WPS_PITCH or sometihng :p
15:08:07jhMikeS eng: "Reset Colours"
15:08:08jhMikeS voice:
15:08:08jhMikeS new:
15:08:08DBUGEnqueued KICK jhMikeS
15:08:08jhMikeS+ id: VOICE_KBIT_PER_SEC
15:08:08***Alert Mode level 1
15:08:08jhMikeS+ desc: spoken only, for file extension
15:08:09***Alert Mode level 2
15:08:09jhMikeS+ eng: ""
15:08:11jhMikeS+ voice: "kilobits per second"
15:08:13 Join Lear [0] (
15:08:13jhMikeS+ new:
15:08:15jhMikeSis that some kind of old lang format?
15:09:33aliaskG'night all
15:09:39 Quit aliask ("sleep")
15:10:48jhMikeS id: VOICE_KBIT_PER_SEC
15:10:48jhMikeS desc: spoken only, for file extension
15:10:48***Alert Mode level 3
15:10:48jhMikeS user:
15:10:48***Alert Mode level 4
15:10:48jhMikeS <source>
15:10:48***Alert Mode level 5
15:10:48jhMikeS *: ""
15:10:50jhMikeS </source>
15:10:52jhMikeS <dest>
15:10:54jhMikeS *: ""
15:10:56jhMikeS </dest>
15:10:58jhMikeS <voice>
15:11:00jhMikeS *: "kilobits per second"
15:11:02jhMikeS </voice>
15:11:06jhMikeSis that what it should be now?
15:11:18Bagderlooks correct, yes
15:11:43***Saving seen data "./dancer.seen"
15:12:00markunjhMikeS: there is also a conversion script in tools to do it for you
15:12:20jhMikeSthanks but there was only one string
15:13:20JdGordonBagder: can you please add a make target so it stops compiling after the rockbox.elf has been built?
15:13:43markunDoes anyone here have experience with TLB tables for ARM?
15:13:50Bagderless warnings now
15:14:08JdGordonLinusN: yes, but that becomes tedious..a nd i cant automate it
15:14:10Bagderwe should fix these last ones
15:14:30LinusNJdGordon: make rockbox.elf?
15:15:10LinusNtry "make apps/rockbox.elf"
15:15:13JdGordonbut stop just after its copied to rockbox.iriver..
15:15:29JdGordonoh its there already?
15:15:44LinusNor even "make rockboc.iriver"
15:15:52LinusN"make rockbox.iriver"
15:16:03LinusNshould work too
15:16:22jhMikeSI don't like the names wav_2_* I think *_enc would be better IMO
15:16:23JdGordonnothing to be done
15:16:27Bagderoh man is that doom code weird
15:16:43LinusNJdGordon: is there anything to be done then?
15:16:51Mikachumake: *** No rule to make target `rockbox.ipod'. Stop.
15:17:24JdGordonsorry, so rule
15:18:20 Join dpassen2 [0] (
15:18:32LinusNlooks like it's not that simple
15:19:07 Join mantono [0] (
15:19:23JdGordonalso, error: firmware image is 228274 bytes while max size is 204800! for recorder... is that my gcc version?
15:20:20JdGordondoes anyone want to review my changes so pitchscreen works again?
15:20:49***Alert Mode OFF
15:21:53JdGordon.. guess not..
15:22:10Bg3rJdGordon: link ?
15:22:30JdGordon:( really? i was hopgin no one would so i didnt have to upload it :D
15:22:53JdGordonive got a clean source tree and it compiles 100% on all targets so it should be safe to commit it
15:23:44peturbut does it work? :p
15:23:59Bg3rJdGordon: commit it
15:24:01JdGordon.. of.. course.. it.. does :p
15:24:08Bg3rbut don't run away after that :D
15:24:30Bg3r == .. of .. course.. it.. should ?
15:24:42Bagderanyone wanna proof-read my doom fix
15:24:43JdGordoncan any target enter the picth screen from the wps other than irivier?
15:25:33amiconnYes, archos recorders
15:25:36Bagder(I don't want to test it, just get rid of that warning)
15:26:08JdGordonamiconn: im 10sec too impatiant
15:26:24JdGordonwhats the combo ill add it with the next lot of fixes..
15:26:29preglowmarkun: it's got an mmu as well?
15:27:09JdGordonBagder: "1. auto-detect and auto-use ccache if installed and present in $PATH" <- awesome
15:27:23Bagderyeah, I forget to use ccache all the time
15:27:40Mikachuisn't the return superfluous?
15:27:48Bagderyes it is
15:27:54Bagderonly there for clarity
15:28:32Bagderand I didn't feel like reversing the condition
15:28:42preglowmarkun: doesn't the arm arm pdf describe how the mmu works?
15:30:03amiconnJdGordon: On+Up and On+Down the same way as on irivers
15:30:30LinusNJdGordon: i have added "make bin" now
15:30:38LinusNupdate your cvs and reconfigure
15:30:46CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:30:46*amiconn wonders why JdGordon doesn't simply check the old .h files for the button mappings
15:31:03JdGordonamiconn: i checked gwps.h and its not there...
15:31:11JdGordoncheers LinusN
15:32:29JdGordonthanx LinusN, works perfectly
15:33:49 Join nls [0] (
15:33:51Mikachuthe cvs builds page doesn't show pending commits done after the currently building ones?
15:35:42markunpreglow: yes and yes
15:35:43Bg3ranyone with h100 remote here to tell me whether SOURCE or BITRATE is on the left ?
15:35:43amiconnJdGordon: That mapping was hard-wired in gwps.c
15:35:59JdGordonamiconn: oh ok then
15:36:02Bg3rah, nevermind
15:36:05amiconn || CONFIG_KEYPAD == IRIVER_H300_PAD
15:36:05amiconn case BUTTON_ON | BUTTON_UP:
15:36:05DBUGEnqueued KICK amiconn
15:36:05amiconn case BUTTON_ON | BUTTON_DOWN:
15:36:17JdGordonis it still there?
15:36:23markunpreglow: but it says the TLB has "up to" 4096 entries and I want to know how to make it less than 4096
15:36:29MikachuJdGordon: could you be consistent with the _ in WPS_AB_? :)
15:37:10amiconnnope, but you can always read old revisions in viewcvs
15:37:54JdGordonMikachu: naa.. this way is more fun...
15:38:23tucozA clip from a Rockbox concert :)
15:38:44MikachuThis video may contain content that is inappropriate for some users <- eh? :)
15:38:54Mikachu(this means you must register to see it)
15:39:31nlsJdGordon, the set time / date screen doesn't work...
15:39:31markunpreglow: I think I figured it out. Add an invalid entry at the end. It should never reach it anyway.
15:39:35tucozyes, wouldn't know why it got flagged as inappropriate
15:40:00JdGordonnls: can u add it to the thread on the forums please? im gonig to bed soon and i might forget
15:40:41nlsugh you're making me reg on the forums? ;-)
15:41:17 Quit Lex (Remote closed the connection)
15:41:52CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:41:52*JdGordon apologises...
15:42:18 Quit TCK (Read error: 104 (Connection reset by peer))
15:42:25 Part LinusN
15:43:58bluebrotherJdGordon: the pitch screen still has bugs on h120
15:44:15JdGordonmost recent build?
15:44:18bluebrotherholding up / down stops increasing / decreasing the pitch
15:44:40bluebrotheralso, I can't leave that screen
15:44:50bluebrotheryes, just downloaded.
15:45:13JdGordonodd.. it works fine on my h300
15:45:16bluebrother(but holding Stop for a longer time shuts down the player)
15:45:35JdGordonpressing stop doesnt quit the screen?
15:46:07bluebrotherno. Also, now pressing Down increases the pitch ...
15:46:45bluebrothernudging with left also doesn't work.
15:46:59bluebrotherI get a speed increase by 2% upon release.
15:47:19bluebrotherrebooted, up and down are still swapped.
15:48:02bluebrotherdamn, do I have to shutdown the player again to leave that screen?
15:48:44JdGordonbluebrother: woopsies... im not sure how i broke that..
15:50:31 Join mkey [0] (
15:51:01preglowmarkun: so, how's progress?
15:52:26JdGordonbluebrother: im gonna blame the underpants gnomes for this one :D
15:53:16JdGordonno.. im gonna blame amiconn for removing the button_repeat special case :D
15:54:11peturtsk tsk tsk
15:54:33bluebrotheranyway ... I'll just refrain from using that screen until I see a suitable commit message ;-)
15:54:33nlsJdGordon, ok added to the forum thread
15:54:49JdGordon10min hopefully
15:56:20tucozare the recorder keys still wrong?
15:56:52JdGordonshouldnt be
15:57:06tucozok, good
15:57:16tucozah, missed your recent commit.
15:59:27amiconnJdGordon: I hope to be able to have a look at the recorder button actions tonight. I suspect the quickscreen operation to be very wrong
16:00:00JdGordonqs looks fine.. up/down/left/right
16:00:11JdGordonplay is cancel tho which is proabbly wrong?
16:00:13amiconnAlso, the F3 quickscreen probably doesn't work at all. We'll probably need a new action for it
16:00:32amiconnJdGordon: The quickscreen could be operated in 2 ways on recorders before
16:00:50JdGordonoh, ok, that proably wont work then..
16:01:13JdGordonwhen the repeat special case was removed, how do we do an action repeatadly on a repeat event?
16:01:32amiconn(1) The slow way: Press the quickscreen button (F2 or F3). Quickscreen pops up. Adjust the value you want, then press the same button again (F2 or F3 depending on the quickscreen) to leave it again
16:02:00amiconnJdGordon: Just with BUTTON_NONE as a prerequisite. That's how volume, seek etc work now
16:02:37JdGordonfeels wrong tho.. but ok
16:02:51amiconnIt's not wrong at all
16:03:09amiconnWith this scheme, you can have both repeating and non-repeating events on long press
16:03:22amiconnWith the previous one, this wasn't possible
16:04:25amiconn(2) The real quick way (and the reason why these things are called _quick_screens): Press _and hold_ the quickscreen button, and, while still holding it, adjust the respective value (with left/down/right). If you do it this way, releasing the quickscreen button leaves the quickscreen again
16:05:22amiconnUnfortunately this isn't possible on iriver because the hardware can't read both the quickscreen button and left/down/right at the same time
16:05:56JdGordonit shouldnt be too hard to get working for recorder.. just need more mappings in the list
16:06:20amiconnI have an idea how to implement that. It would require one additional event, and a couple of extra mappings
16:06:55amiconnAn additional problem is that the quickscreen button is different for the F2 and F3 quickscreens
16:08:09JdGordonno problem.. just add a new context for the F2 screen (the one that isnt QS) and bobs your uncle :p
16:08:15tucozIsn't there a menu in wps on the ondio?
16:09:27 Quit SereR0kR ()
16:09:31tucozthat is ACTION_WPS_MENU? the comment says 'we can't have that'
16:10:16amiconntucoz: The main menu isn't directly reachable from wps on Ondio - too few buttons
16:10:35JdGordonnonesence... combinations can be used...
16:10:55daurn|laptopwho here can help me out with a new port?
16:10:55tucozok. thanks.
16:11:32amiconnJdGordon: Ahahahah. The combos are all used up, at least those usable without being excessively fast, or without making things highly inconsistent or illogical
16:12:00daurn|laptopamiconn: all 2 buttons combos?
16:12:07amiconnInstead, the menu combo (Long Mode) brings up the wps context menu. Pressing Long Mode again then switches to the main menu
16:12:08daurn|laptopor, do you have 3/4/5 button combos left
16:12:56Mikachuamiconn: can't you also go to the file browser and press menu there?
16:13:10Mikachuor you have to go via context there too?
16:14:17amiconn3 button combos are electronically possible (Mode+Off+ one of the directions), but not mechanically. Have a look at the ondio keypad then tell me how to press that...
16:14:25amiconnMikachu: Yes that's possible
16:14:44JdGordonbluebrother: all fixed
16:14:44amiconn...but it seems to be wrong...
16:15:24daurn|laptopwho here can help me out with a new port?
16:15:52tucozwhat player are you porting?
16:15:57daurn|laptopgmini 402
16:16:15tucozmaybe the guys over at archopen can help?
16:16:30daurn|laptopi am the guy at archopen
16:16:41daurn|laptopwell - sort of
16:16:46tucozaha, hehe
16:17:01midkayhaha. the guy.
16:17:07daurn|laptopi'm the only person who speaks english who actually HAS a gmini 402
16:17:16daurn|laptopmidkay: shushen
16:17:28midkayand who simply asks others to compile cool stuff for him, right ;)
16:18:00nlsJdGordon, I found out why the "set time/date" screen doesn't work. It uses ACTION_STD_PREV/ACTION_STD_NEXT _AND_ ACTION_SETTINGS_INC/ACTION_SETTINGS_DEC which are mapped to the same buttons for irivers
16:18:02daurn|laptopmidkay: not my fault i don''t have a toolchain installed at every internet cafe
16:18:15JdGordonnls: im in the thread now...
16:18:29midkayhow do you get on irc at an internet cafe without, say, a personal computer?
16:18:34nlsoh :-)
16:18:41daurn|laptopcgi applet
16:19:01midkay.. without a *personal computer*? or did you mean "at work" versus "at an internet cafe"?
16:19:19daurn|laptop? - you open it in a webbrowser
16:19:29 Join BockBilbo [0] (
16:19:41BockBilboanyone knows if rockbox is available on ipod nano?
16:19:43 Join fatherfork [0] (
16:19:45midkayyou need a computer to have a web browser.. :)
16:19:53daurn|laptop.. if you goto net cafes without browsers, you've got a problem
16:19:56midkayBockBilbo: it is available, it should be like the second sentence on the front page.
16:20:00*JdGordon wonders why \n isnt translated into <br> in the web scripts
16:20:11JdGordon...for commit messages
16:20:13midkaydaurn|laptop: if it's your computer, then, why shouldn't you have a toolchain installed everywhere?
16:20:21BockBilbomidkay, im asking this because its not working for me on the senab experimental build
16:20:39midkayBockBilbo: we don't offer support for experimental builds.. did you install the bootloader?
16:20:51BockBilbomidkay, yeah
16:20:58 Quit bluebrother ("CGI:IRC")
16:21:07BockBilboi followed the instructions on the wiki for linux
16:21:19midkayBockBilbo: does a standard Rockbox build work?
16:21:31BockBilbothat's what i was about to try right now
16:21:43midkaygood idea.. :)
16:21:54JdGordonbed time for me, leave a message in the thread if any more problesmw ith buttons pops up
16:21:58 Quit JdGordon ("Konversation terminated!")
16:22:17fatherforkhaha, dang, I just woke up
16:22:40 Quit muesli__ ("ich will KEe!!!")
16:22:41tucoznls, hi. What do you think I should call the long press keys. Like ActionWpsSeekFwd will be 'Long Right' for instance.
16:23:18tucozThat is, should I have a leading capital letter in Long, or should I call it Hold?
16:23:37BockBilbomidkay it doesnt
16:23:51BockBilbosee.. this is what i do.. i have a folder called games with 3 .gbc files on it
16:23:54tucozif we use these macros outside of tables, a leading capital letter might look wrong. Not sure.
16:24:05 Quit Lear ("Chatzilla 0.9.74 [Firefox]")
16:24:43BockBilboi remember when using rockboy on my 60gb ipod that the procedure to run rockboy is just selection the gbc file
16:24:45nlstucoz, somebody apparently got confused with the Hol button vs. Hod-swithc
16:24:47BockBilboam i wrong?
16:24:52 Join Criamos [0] (
16:25:10 Quit petur ("later")
16:25:12tucoznls, aha. So we should write Long Right then?
16:25:39nlsI'm not sure really...
16:25:48tucozme neither.
16:26:13tucozFor the tables it's ok, but for the in text key presses it's hard to do this clever.
16:27:01 Join SereR0kR [0] (
16:27:07Bagderpreglow: I've saved the last two warnings for you ;-)
16:27:12nlsyes and as you say it will look strange with Hold/Long with a capital letter before the button in text.
16:28:09tucozI'll just prepare the platform files, and we can decide this later
16:28:38nlsConverting the buttonmaps to the actions will not be fun... espescially plugins in the future...
16:30:15tucozNo, but it will be worth it when we add new targets.
16:30:41tucozhowever, the plugins are the worst part with regards to tables
16:31:06midkayBockBilbo: rockbox isn't working, or rockBOY isn't working?..
16:31:07nlsyes probably and at the rate rockbox changes hte manual will never be near up to date :-)
16:31:26tucozhehe, no. Probably not. Unless we get a new feature freeze.
16:31:41*nls wants another 3 months of feature freeze NOW :-)
16:31:45Mikachui'm going for the ugly code award here,
16:31:52BockBilbomidkay, rockBOY
16:31:56BockBilborockbox works fine
16:32:20midkayyou asked if "rockbox is available on ipod nano".
16:32:29midkayrock*boy* is not available on the nano.
16:32:43BockBilboohhh my bad..
16:32:51BockBilbooh.. so its not.. :(
16:33:04BockBilbosorry for the confussion
16:33:11nlstucoz I started looking at the config file options tabe in the appendix but it made my head hurt so I din't get far.
16:33:50tucoznls, I know. And I am not even sure that table is complete. The wiki version certianly wasn't
16:34:23nlsIt's a monster both code wise and it will need hundred of opts for different targets and is very outdated and changes frequently :-(
16:36:17 Quit mkey (Read error: 60 (Operation timed out))
16:36:45nlstucoz, how do tou feel about a HAVE_LCD_REMOTE in the platform files?
16:36:56tucozgood idea.
16:37:17nlsI think I'm gonna add that
16:37:23tucozI am working on the platform files atm, so I can add that as well.
16:37:30nlsok good
16:37:47tucozonly LCD_REMOTE? or HAVE_REMOTE in general?
16:38:43amiconnThe config-*.h files use HAVE_REMOTE_LCD
16:39:26 Join barrywardell [0] (
16:39:36tucozok. I am just thinking what we should do if we want to document a target with a remote, but without LCD
16:39:47nlsonly recorders have non lcd remotes (only) right?
16:40:16tucozok, that works for now. Wonder about new targets.
16:41:07nlsthat's a later problem ;-)
16:42:07tucozyes :)
16:42:41tucozh1xx, h300, and the x5 have remotes right?
16:42:46BockBilbothanks so much midkay
16:42:48BockBilboi gotta go
16:42:50 Quit BockBilbo ("Ex-Chat")
16:43:01amiconnLCD remotes: H1x0, H300, X5
16:43:18tucozok. good
16:43:24amiconnNon-LCD remotes: Archos player, recorder; Iriver H1x0, H300
16:43:50amiconnPlus ipods (at least some) if we or the ipl guys can figure out the serial port
16:43:51tucozBut I do not think we should care about the non-lcd remotes for the irivers
16:44:15tucozif they have a lcd-remote, we can document the non-lcd remote in the same section.
16:44:58 Join A_M [0] (
16:44:59*amiconn actually has a simple remote for his mini2g
16:45:11amiconnJust 5 buttons...
16:47:52A_MI could use some help here; I can't make rockbox because there is "No rule to make target 'bla bla bla/menu_select.o' needed by 'bla bla bla/rockboxui.exe'".
16:48:04perplexitythe H300 also has a non-lcd remote.. why should it not be documented also ?
16:49:21A_Mthe trouble is menu_select.[ch] are source files I used earlier but don't need anymore, I've deleted them from the source dirs as well as all source files that included them, yet make still tries to make them... :S
16:49:43fatherforkI have a wireless iPod remote
16:50:17Bagderbarrywardell: I sent you a mail you might like
16:50:40tucozperplexity, it will be documented. But, not in a separate h300 non-lcd manual
16:51:02perplexityAhh, gotcha.. sorry, misunderstood..
16:51:43tucozwe want to document the common features as much as possible. But, there will be parts that are very target specific.
16:51:53barrywardellBagder: yay, thanks. just finished reading it now
16:58:01jhMikeSI need to know where a particular chunk should go now as there seems to be no place in recording.c for it.
16:58:57jhMikeS lcd_puts(0, 7, buf);
16:58:57jhMikeS+#if CONFIG_CODEC == SWCODEC /* Info: CPU speed, bitrate, encoder type */
16:58:57jhMikeS+ snprintf(buf, 32, "CPU:%3dMHz %4dkB %s", cpu_frequency/1000000,
16:58:57***Alert Mode level 1
16:58:57jhMikeS+ pcm_rec_current_bitrate(), info[global_settings.rec_quality]);
16:58:59jhMikeS+ lcd_puts(0, 8, buf);
16:59:05jhMikeS gui_syncstatusbar_draw(&statusbars, true);
16:59:07jhMikeS peak_meter_draw(0, 8 + h*2, LCD_WIDTH, h);
16:59:33 Quit theli_ua ("Leaving")
17:00:51tucoznls, I have made actions for the standard and wps contexts now. I'll talk to bluebrother later on and maybe commit that and the remote option.
17:02:58 Join lex [0] (
17:05:21 Quit A_M ("CGI:IRC")
17:07:01 Quit Zagor ("Client exiting")
17:08:58***Alert Mode OFF
17:09:25nlstucoz, ok, I started with the config file stuff again adding remote options, and making corrections.
17:09:38Bagderand we are now 60 committers
17:10:13Mikachuif you want to avoid awkwardness feel free to add me when it would have been 69
17:10:26jhMikeSWhere is the recording screen drawn with the peakmeters and all?
17:11:47***Saving seen data "./dancer.seen"
17:12:06 Join apo [0] (
17:12:06jhMikeSHelp! I'm kinda stuck here for a sec! :) Bagder?
17:12:23BagderI'm not really familiar with the recording area
17:14:20tucoznls, you are brave :)
17:15:12tucozHmm, should I just commit these actions perhaps. As a start. They don't break anything, and we can decide on the long press stuff later on?
17:19:30tucozhmm, in the crossfade section of the manual should I call it track instead of song?
17:20:12jhMikeSok I found it...that part changed a lot
17:20:42nlsidk what do we usually call it?
17:20:52nlsbut imo track is better
17:21:29BagderjhMikeS: it might change even more soon as petur has some recording stuff pending commit
17:21:46tucozyes, I think so too. I am rewriting the cross fade section and add a figure on how the different modes work.
17:24:19 Join Kohlrabi [0] (
17:26:52 Quit linuxstb (Read error: 110 (Connection timed out))
17:27:27 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
17:31:25 Quit _FireFly_ ("Leaving")
17:31:29 Quit fatherfork ("CGI:IRC (EOF)")
17:34:27amiconnjhMikeS: What's your question wrt recording? I can't figure it out from the code snippet...
17:34:51 Join CriamosAndy [0] (
17:35:44 Join A_M [0] (
17:38:00nlswhat do I call the unit of hexadecimal rgb (colors in the config files)
17:39:24 Join Caliban [0] (
17:39:24amiconncolour specification?
17:40:47 Quit A_M ("CGI:IRC (Ping timeout)")
17:43:09tucozhex number?
17:43:41nlsarbitrary 6 digit hex :-)
17:45:14tucozmaybe something like hh:mm:ss for a timer
17:45:25tucozthat is, in that way
17:45:36tucozif you understand
17:46:39nlsyou mean like rrggbb
17:49:30jhMikeSamiconn: trying to find the place where the recording parameters are written to the screen during record but I finally tracked it down
17:52:55 Quit Criamos (Read error: 110 (Connection timed out))
17:56:36 Quit secleinteer ()
18:01:35 Join webguest80 [0] (
18:03:08 Quit apo (Remote closed the connection)
18:03:31webguest80is there an option now to disable horizontal screen scrolling ?, i find it very irratating to navigate with that feature, if one pauses just a fraction to long with a right click (h140) the darn thing goes into Horizontal creen mode when i just want it to play a song, very irritating
18:04:52 Join secleinteer [0] (n=secleint@
18:06:12preglowi definitely think it's a bad thing after having used it a while
18:06:35preglowthe list feels so "sticky" now
18:06:43preglowsince it's reacting to button depresses, not presses
18:07:32nlsand to top it off the setting is on by default and also hidden :-)
18:09:12 Join apo [0] (
18:09:45Mikachumake it default to off but please keep the option
18:09:59preglowscrolling itself is nice
18:10:12preglowbut i don't like having it depend on keeping a button pressed for so and so long a time
18:10:21preglowmakes it very hard to navigate when drunk!!
18:10:28Mikachua common problem for you
18:11:06preglowdespite the fact that i like my beer, i'm very seldom drunk, as a matter of fact
18:11:55webguest80default to off is a good idea, navigation is very clumsy with that feature
18:12:45Mikachui hardly use left/right to navigate
18:12:55Mikachuonly to exit dirs, and i always know where i'm going :P
18:16:58preglowso am i, but i don't like having to time my presses
18:19:53 Join madNeSs [0] (
18:20:59 Part madNeSs
18:21:18 Join excitatory_ [0] (
18:29:19 Join bluebrother [0] (
18:36:33 Nick apo is now known as two-mbit-apo (
18:37:48 Join mkey [0] (
18:40:09 Join Gnelik [0] (n=Gnelik@
18:41:17 Quit dan_a (Read error: 110 (Connection timed out))
18:41:29 Quit Arathis ()
18:44:25COmputomanHey is there a version of the apple firmware that is hacked?
18:44:56 Join fatherfork [0] (
18:47:24fatherforkWould replaygain cause my iPod music playback to glitch and require restart more often than normal?
18:47:48fatherforkusually, I have to restart is about ever hour or so of playback
18:48:12fatherforkbut now that I'm using replaygain, I've had to restart 3 times in the last hour
18:48:12 Join sharpe [0] (
18:48:36fatherforkaw sharpe, you just missed my question...
18:48:44 Part webguest80
18:49:05sharpeAnd to think I've been up the past thirty minutes.
18:49:27fatherforkI asked if replaygain would cause my iPod audio playback to glich and require restart more often than normal
18:50:27sharpeAh, well, good thing I missed it, I know very little about replaygain. But I can tell you it wouldn't require more restarts.
18:51:35fatherforkbecuase usually I have to restart my iPod about every hour, but now that I have replaygain on, I've had to restart 3 times in the last hour
18:52:22fatherforkok, well how about another question...
18:52:23sharpeWhich iPod do you have?
18:52:31fatherfork5G, 60GB
18:52:55sharpeStrange. My 5g rarely ever has to restart...
18:53:04fatherforkhmm, using any patches?
18:53:39fatherforkI must conceed that that's most likely where my problem lies
18:54:18sharpeMakes sense.
18:54:37fatherforkyeah, I guess in that case, I'll just deal
18:56:14 Join theli_ua [0] (
18:57:58mirakwho did most of the coldfire asm optimisations for the codecs ?
18:58:59mirakI did 68000 assembly when I was young and mips and x86 asm at univ, I am interested in optmising the idct in xvid and libmpeg2
18:59:11Bagdermidkay: mr preglow did
18:59:16mirakI am wondering if there are some things to now
18:59:20Bagdermirak even
18:59:42mirakBagder: mmm ?
18:59:59fatherforkhe said "midkay", meant "mirak
18:59:59mirakboth codecs uses almost the same idct
19:01:37 Quit bilbravo (Remote closed the connection)
19:02:01mirakpreglow: are you there ?
19:03:37 Quit mantono ("Don't worry, be happy")
19:04:19 Quit theli_ua ("quit")
19:04:54 Join jaczehack [0] (
19:07:51 Quit fatherfork ("CGI:IRC (EOF)")
19:09:33 Join adiamas [0] (
19:11:48***Saving seen data "./dancer.seen"
19:14:50 Join dan_a [0] (n=dan-mirc@
19:15:03 Join bilbravo [0] (
19:17:41 Quit barrywardell ("CGI:IRC (EOF)")
19:20:24 Quit Kohlrabi ("Quit")
19:21:16preglowmirak: somewhat
19:22:23 Join hartigan [0] (
19:26:37 Join Ribs [0] (
19:27:37 Quit Rob2222 (Read error: 104 (Connection reset by peer))
19:28:07 Nick two-mbit-apo is now known as apo` (
19:31:44hartiganHi there! I just registered for the Twiki and I was wondering if there was anyone here who could provide me with permissions to edit. I have no intention of modifying content but, I have noticed some spelling and grammar fixes that I could help out with. Thanks!!
19:32:06 Join Lynx [0] (
19:33:51COmputomanHey what font files are good
19:34:03mirakpreglow: somewhat not ? :)
19:34:42 Quit _hotwire_ (Read error: 110 (Connection timed out))
19:34:44 Join _hotwire_ [0] (
19:34:45hartiganI typically stick with what comes with the WPS's that I've downloaded.
19:34:52hartiganfont-wise that is.
19:35:09COmputomanThe font is just so little i cant read anything
19:35:15COmputomanthe default font that is
19:35:23COmputomanIs there a way to increase the size?
19:35:34COmputomansomeone in this channel told me no yesterday
19:35:36mirakpreglow: I watched once an idct for vorbis I think, and you seemed to avoid loops
19:35:50preglowvorbis doesn't use idct
19:35:52preglowit uses imdct
19:36:08mirakm like what ?
19:36:10hartiganI've never changed a font size or found need to.
19:36:31hartigantry Black Glass WPS it's very clean and very readable.
19:36:37preglowit's not very easy to make an imdct faster by doing loops
19:36:38mirakpreglow: I can find the coldfire docs on the wiki ?
19:36:43preglowmirak: yes
19:36:46hartiganjust look in the Custom WPS page on the Wiki
19:36:59mirakpreglow: ok
19:37:22sharpeCOmputoman: no,
19:37:33hartiganthanks sharpe
19:37:41mirakpreglow: in wich case the asm was faster than with gcc ?
19:37:53 Quit excitatory (Nick collision from services.)
19:38:10 Nick excitatory_ is now known as excitatory (
19:38:18preglowmirak: asm is almost always faster than what gcc doers
19:38:45mirakpreglow: nice to know
19:39:35mirakpreglow: are there enough registers or is it really tricky ? Do you store intermediate variable when you don't have enough registers in iram or in slow ram?
19:39:43 Quit Ribs ("eh eh ehhhh!")