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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2008-03-27

00:00:49jhMikeSNico_P: I take it the event log is heavily filtered.
00:01:04Nico_Pit only displays playback events
00:01:22*gevaerts looks around for people who know how hardware works to comment on his conclusions
00:01:33jhMikeSwhy is it jsut pause events?
00:01:49 Quit DerPapst (Nick collision from services.)
00:01:52 Join DerPapst [0] (
00:02:04 Join CyBergRind|w [0] (n=cbr@
00:02:17Nico_PjhMikeS: there are others too (Q_AUDIO_FILL_BUFFER, Q_AUDIO_TRACK_CHANGED)
00:03:25linuxstb_kugel: No. But if you want to parse a viewport struct, you should adapt the parse_viewport() function in apps/gui/wps_parser.c, rather than use parse_list directly.
00:03:51 Quit mirak (Client Quit)
00:04:51kugellinuxstb_: No, I don't want parse a vp struct directly. Thanks
00:04:57jhMikeSyeah but what's with the 4 Q_AUDIO_PAUSEs being sent with no other playback events
00:05:00 Join stripwax [0] (
00:05:15jhMikeSI mean like Q_AUDIO_PLAY in between
00:05:28GameJerkthere a way to have the music keep playing when i hit the menu button instead of it fading out and pausing?
00:05:47XavierGrgevaerts: these are the results for high or full speed?
00:05:54Nico_PjhMikeS: that's because I was pausing and unpausing. unpause is done by sending Q_AUDIO_PAUSE with false as a vaule
00:06:02gevaertsXavierGr: high speed
00:06:10XavierGrgevaerts: Nice!
00:06:14scorche|shGameJerk: read the manual please...the last few questions are answerable in there
00:06:33XavierGrHopefully I will have a c240 the next days so I will be able to help in testing (if it isn't a v2)
00:06:35Nico_PjhMikeS: the logf message is actually different, but there isn't enough space in the logf buffer for the latter part
00:06:36 Quit toffe82 (Read error: 104 (Connection reset by peer))
00:06:51gevaertsXavierGr: I'm not sure if I agree. Having to disable _all_ interrupts is not very nic
00:07:20XavierGrwell at least you have a slight idea what is to blame for not working properly
00:07:24jhMikeSNico_P: yeah, was gonna say. now I see.
00:07:55gevaertsYes, but it seems to rule out any kind of creative use of usb (audio,...)
00:08:18 Quit bertrik ("bye")
00:08:23XavierGrgevaerts: ah yes, indeed, forgot that part
00:08:45gevaertsUnless some hardware-wizard can think of a solution
00:09:17Nico_PjhMikeS: I can't try your patch right now but I will ASAP
00:10:07 Quit CaptainKewl ("( :: NoNameScript 4.02 :: )")
00:11:03 Quit haemmy ("I was raided by the FBI and all I got to keep was this lousy quit message!")
00:13:52 Quit petur ("*plop*")
00:14:34amiconngevaerts: How about just disabling interrupts during the actual block transfer?
00:15:15*amiconn still doubts that this is the root of the problem though
00:15:30amiconnIt might still be an effect of something we're doing wrong
00:15:38GameJerkscorche: i found the fade out on pause/stop option but i can't find anything referring to being able to have the music continue playing when i hit the menu button
00:15:48gevaertsThat could work. The problem is that in theory a transfer can take a long time (seconds...) if the bus is very busy
00:17:05amiconnHmm, isn't a block always transferred in one piece?
00:17:16 Join toffe82__ [0] (
00:17:25 Nick toffe82__ is now known as toffe82 (
00:17:45 Quit davina ("GNU/Linux the free alternative to Windows")
00:17:58markunI want to simplify the #ifdefs in firmware/SOURCES a bit. Do you think it's better to have a NO_ATA or a HAVE_ATA define?
00:18:14markun(or maybe something else)
00:18:26 Quit cbr|w (Read error: 110 (Connection timed out))
00:18:39gevaertsNo (unless you use 512 byte blocks). And even if it was, we have no control over when it is actually sent. We just tell the controller that this block is ready in case the host asks for it
00:19:00amiconnAren't the block on the bus always 512 byte (or less)?
00:19:22gevaertsOn the bus the packets are 512 bytes (for high speed)
00:20:18amiconnAh, yes, packets I mean
00:21:06gevaertsbut this controller is pretty high level, so the software can't directly touch packets
00:22:22amiconnI sincerely doubt the apple OF (or any other PP OF) disables all interrupts permanently during usb connection
00:22:49 Quit roolku (Read error: 110 (Connection timed out))
00:23:24 Quit dantje_ ("Ex-Chat")
00:24:25gevaertsActually, I think (but have not verified) that in our current situation you can basically do anything as long as you only touch IRAM and leave the external RAM to the USB controller.
00:24:30amiconnInterrupts or not, the CPU core(s) is/are running in parallel to the usb controller. And when they're running, they're reading code andreading and writing data
00:26:21amiconnBut maybe a cache line miss (and hence fetch) is what disturbs the controller, so it's not the fact that interrupts are running, but rather the fact that interrupts routines are more likely to cause cache misses? (speculating...)
00:27:33amiconnjhMikeS: I vaguely remember having seen something about priorities, so maybe there are configuration bits to configure cpu vs. usb controller (vs. dma engine??) priority for ram accesses?
00:28:48 Join countrymonkey [0] (n=4b05639a@gateway/web/cgi-irc/
00:30:06countrymonkeyI guess patches 8722, 8737, and 8761 are ready for commit?
00:30:29 Quit gregzx ("ChatZilla 0.9.81 [Firefox]")
00:32:41 Join roolku [0] (
00:32:41 Quit toffe82 (Read error: 104 (Connection reset by peer))
00:32:55*bluebrother wonders what "changes were made" this bookshare guy is talking about
00:33:23countrymonkeywhat task? I'm bookshare
00:33:29 Join toffe82__ [0] (
00:33:49roolkubluebrother: I aggree that the play button should resume, but never got around changing it
00:33:59 Quit toffe82__ (Read error: 104 (Connection reset by peer))
00:34:51bluebrotherroolku: using this display button for switching to the file browser from wps is also confusing −− all my other targets use select instead.
00:35:13bluebrotherI might give it a try once I find the time. No big deal ...
00:35:31 Join corevette [0] (
00:36:06countrymonkeyIf you mean task 8722 than the changes are 2 deprecated strings
00:36:15roolkubluebrother: as for lineout - it is not wired to the dock connector in the player
00:36:56 Quit robin0800 (" HydraIRC -> <- Po-ta-to, boil em, mash em, stick em in a stew.")
00:37:03roolkubluebrother: I thought about a hardware mod, but I am not really using the player...
00:37:29 Join cool_walking_ [0] (
00:38:03bluebrotherstrange. Don't they have enough pins on that connector? Oh well.
00:38:19 Join toffe82__ [0] (
00:38:27 Nick toffe82__ is now known as toffe82 (
00:38:43bluebrothercountrymonkey: yep, I was referring to FS #8722. This guy could really be a bit more descriptive IMO.
00:39:31 Quit Mouser_X ()
00:40:23 Quit ol_schoola ()
00:40:24countrymonkeyI'm bookshare. I'm sorry. I said made today because there was a date stamp on the comment and the strings were depricated that day.
00:41:48roolkubluebrother: I think they couldn't spare the space for the required capacitors on the circuit board - there are plenty of spare pins
00:42:01*bluebrother would really appreciate if people won't use different nicks in all places
00:42:15 Join miepchen^schlaf_ [0] (
00:42:28*linuxstb_ apologises...
00:42:38countrymonkeyWhat's wrong with that?
00:43:28*linuxstb_ retracts that apology (he thinks...)
00:43:58*gevaerts can't remember what nick he uses on the tracker
00:44:22linuxstb_Bagder: Missing arm compiler on ...
00:44:29countrymonkeyYou use fg
00:44:35Nico_Plinuxstb_: you managed to fix the printf warnings in checkwps? congratulations!
00:44:45Nico_Phow did you do it?
00:44:47gevaertsSo I have two nicks and one wikiname
00:46:17linuxstb_Nico_P: I forget now - there were so many warnings to fix (fixing one introduced another...)
00:46:26bluebrotherwell, in your case it's kinda obvious. But non-obvious cases can be quite confusing (and therefore annoying)
00:47:09 Join nas [0] (
00:47:29 Quit matsl (Remote closed the connection)
00:47:47countrymonkeySo it can be committed?
00:47:57countrymonkeyI mean, so they can be committed?
00:49:36 Join emeraldd [0] (
00:49:37 Quit toffe82 (Read error: 104 (Connection reset by peer))
00:50:11 Join toffe82__ [0] (
00:50:19 Nick toffe82__ is now known as toffe82 (
00:50:47*linuxstb_ guesses the broken build server belongs to saratoga
00:50:47jhMikeSamiconn: I just randomly named a register CACHE_PRIORITY with no real idea what it is.
00:51:17gevaertsNice random name :)
00:52:36*bluebrother is annoyed by rbutil creating files on $HOME without asking
00:52:36 Quit toffe82 (Read error: 104 (Connection reset by peer))
00:53:16linuxstb_bluebrother: Didn't you write rbutil?
00:53:55bluebrothernot that part of it.
00:53:59 Join toffe82__ [0] (
00:54:03 Join barrywardell [0] (n=barrywar@rockbox/developer/barrywardell)
00:54:08 Nick toffe82__ is now known as toffe82 (
00:55:00 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
00:55:03bluebrotherand there's a bug in bootloader installed detection.
00:55:55linuxstb_What files get created in $HOME?
00:56:38 Quit stripwax (Read error: 110 (Connection timed out))
00:57:35 Quit miepchen^schlaf (Connection timed out)
00:57:40bluebrotherit created a folder "Olympus mrobe100 Original Firmware Backup" and puts the mi4 file into it
00:58:08gevaertsSo wa
00:58:16*gevaerts blames his keyboard
00:58:18bluebrotherit should at least ask me about it (or tell me that it has done this)
00:58:28gevaertsSo what if you have two mrobe100s ?
00:58:42bluebrotherit will overwrite the first backup
00:58:45linuxstb_Shouldn't it just rename to OF.mi4 on the device?
00:59:19bluebrotheryes. But after some discussion Domonoky added this "host backup".
00:59:47linuxstb_What does the bootloader install actually do?
00:59:48bluebrotherI disliked that idea from the beginning and now I got hit by it.
01:00:08linuxstb_(on the mrobe100)?
01:00:19bluebrothermove PP5020.mi4 to OF.mi4, download PP5020.mi4 and put it in place (all in the System folder on the player)
01:00:29jhMikeSbtw, any yielding takes control of the interrupt level since there really no safe way to mask and restore them anyway when giving control to other code
01:00:54bluebrotherI was skimming the bootloader installation code and wondering if there could be some stuff consolidated
01:01:23*Nico_P has been reading about LLVM and it looks really promising
01:01:28 Quit OlivierBorowski_ (Read error: 113 (No route to host))
01:01:29 Quit csc` ("Powering Off")
01:01:43bluebrotheroriginally I was searching why rbutil told me the bootloader is already installed (though it wasn't)
01:02:07***Saving seen data "./dancer.seen"
01:02:16bluebrotherthat code is also not too robust. Another item on my todo list ...
01:02:33*bluebrother starts to feel like a grumpy old man
01:04:26preglowanyone tried that mikmod codec?
01:04:37 Quit GameJerk (Read error: 110 (Connection timed out))
01:05:25linuxstb_amiconn: Regarding the (real) red on iriver - it's because I removed adc.h from lcd.h.... I can fix if you want.
01:05:38preglowah, it's a plugin
01:05:40kugellinuxstb_: parse_list again. If I have 12|13|, and I only need the last one, what should I do? NULL,&b didn't work
01:05:47amiconnlinuxstb: Please do.
01:06:09 Quit DerPapst (Read error: 110 (Connection timed out))
01:06:15linuxstb_kugel: You have to use all the parameters - why have them there otherwise?
01:06:17amiconnAlthough, the remote type detection doesn't belong where it currently is (lcd driver)
01:07:09 Quit emeraldd ("using sirc version 2.211+KSIRC/1.3.12")
01:07:15kugelbecause I need for one part the complete list, and for another part only the last part
01:07:18Nico_PjhMikeS: I just tried your OS stacks screen patch and I don't see any weird value
01:07:39barrywardellamiconn: did you see my question earlier about the H10 lcd driver?
01:07:50*amiconn scrolls back
01:08:23*countrymonkey needs to eat dinner
01:08:55bluebrothersomeone with an account at sandisks forums around?
01:08:56 Quit countrymonkey ("CGI:IRC (EOF)")
01:09:09Lloreanbluebrother: Aye.
01:09:28bluebrotherLlorean: you're aware of this thread?
01:10:01bluebrotherthe admins telling that all support questions should go to is a bit ... well, annoying, as the thread is about an unofficial installer
01:10:10LloreanQuite. I've emailed the moderator who started it and asked him to either remove the suggestion that people ask us for help, or change the installer mentioned to be the official one.
01:10:29barrywardellamiconn: basically, setting clcd_clock_src to PLL breaks OF loading in the H10 bootloader and I was wondering if you thought it was okay to define it out for the bootloader? and if it should also be done for smal H10?
01:11:14bluebrotherLlorean: nice. I'm sometimes wondering why we are writing an official installer if people frequently come up with unofficial and broken ones.
01:11:22amiconnYou could define it out. Doing so means that the LCD runs a lot slower especially at high cpu clock though
01:11:54barrywardellbut only in the bootloader which probably doesn't matter too much?
01:11:59Lloreanbluebrother: Well, that installer might predate a working RBUtil on Sansa.
01:12:12 Join kushal_12_27_200 [0] (n=kushal@
01:12:46amiconnSetting bits 30 and 31 to 1 means the LCD controller is clocked from the PLL, i.e. it runs at 80 MHz when the cpu is boosted. With both bits set to 0 it runs from the base clock (24MHz)
01:13:08amiconnI really wonder why that setting breaks the OF.
01:13:26amiconnHmm, wait
01:13:35barrywardellamiconn: it's necessary for the (big) H10 bootloader. without it, the OF doesn't start up properly
01:14:19 Part toffe82
01:14:51amiconnbarrywardell: I think I know why that happens. It's probably unnecessary to comment out for small H10
01:14:53 Quit miepchen^schlaf_ ()
01:15:00amiconnRemember the speed setting registers?
01:15:02bluebrotherhmm, good point.
01:15:40*linuxstb_ wonders if the build system has died, or just resting...
01:15:51amiconnWe still didn't do a complete series of measurements in order to use the fast fifo-ed transfer on big H10
01:15:56barrywardellamiconn: ah, yes.
01:15:58amiconnOn small H10, it just works
01:16:11amiconnSo I guess the OF will also work
01:16:19barrywardellbig H10 ended up too fast for the controller IIRC
01:16:32amiconnYes, but the speed can be programmed
01:16:50 Quit roolku ("sleep")
01:16:51 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
01:17:00amiconn...which the big H10 iriver loader doesn't do, but the small H10 iriver loader does
01:17:03 Quit saratoga (Client Quit)
01:17:08linuxstb_saratoga: Hi... bah...
01:17:09 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
01:17:31*amiconn should really do a series of experiments with that
01:17:36linuxstb_saratoga: Is your build server?
01:17:46 Join saratoga222 [0] (n=9803c6dd@gateway/web/cgi-irc/
01:17:47amiconnWould be nice to compare results with a big H10 then...
01:17:55 Quit saratoga222 (Client Quit)
01:17:56 Join saratoga22222 [0] (n=9803c6dd@gateway/web/cgi-irc/
01:18:02 Quit bluebrother ("sleep")
01:18:25amiconnbarrywardell: I'd say disable it for big H10 bootloader for now, putting a big comment that lcd controller speed settings need work...
01:18:35barrywardellI did a seried of experiments with my (big) H10 previously. there's a link somewhere in the IRC logs
01:18:41 Part feindbild ("")
01:18:43 Join miepchen^schlaf [0] (
01:18:44barrywardellamiconn: OK, will do
01:18:50amiconnIs that link target still available
01:19:02*amiconn needs to find that link as well
01:19:19barrywardellI think it was pastebin, available indefinitely but can't remember exactly
01:19:26linuxstb_saratoga22222: Is your build server?
01:19:32*amiconn fetches the latest bunch of logs now that is back
01:21:28*amiconn usually sets pastebin contents to limited lifetime
01:21:35amiconnIf I don't forget it, that is
01:21:44*gevaerts is going to sleep
01:22:09 Quit saratoga ("CGI:IRC (Ping timeout)")
01:22:10*linuxstb_ is still trying to see if his commits were green...
01:22:53 Quit gevaerts ("ZZzz..")
01:23:16 Quit Nico_P (Remote closed the connection)
01:26:01 Nick fxb is now known as fxb__ (
01:26:03saratoga22222linuxstb_: yes and i've just corrected that
01:27:08linuxstb_saratoga22222: Thanks.
01:28:53barrywardellamiconn: found my original records
01:29:23amiconnI've found the correct log. 2007-10-12
01:29:49 Join OlivierBorowski_ [0] (
01:32:33linuxstb_saratoga22222: I've just made another commit if you want to wait to see if your server's OK now.
01:33:13 Quit amiconn (Nick collision from services.)
01:33:19 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
01:35:15*amiconn wants the buildmaster script to become a bit more intelligent :\
01:35:26saratoga22222linuxstb: is there someway I can tell besides watching top?
01:35:50saratoga22222linuxstb_: you too
01:35:54 Quit techhelper1 ()
01:37:26linuxstb_saratoga22222: Just check the build table in a few minutes. But I guess if arm-elf-gcc is running, it's OK.
01:38:02 Join Matt65 [0] (n=189e0b72@gateway/web/cgi-irc/
01:39:37saratoga22222is there someway on the table to see what machine built what?
01:40:06Matt65I just tried installing rockbox, following the instructions word for word on a Sansa e260. NOT v2. After it reset it brought up the sandisk loading screen, and froze there
01:40:06amiconnHover over the table cell...
01:40:14saratoga22222oh never mind
01:41:46saratoga22222I don't understand this
01:41:47saratoga2222273 Warning: couldn't get zip from
01:41:47saratoga2222273 Copied in 3 seconds, 725193 bytes/second
01:41:54saratoga22222did it get the zip or not?
01:42:14linuxstb_saratoga22222: Check the build logs for any of the red builds - arm-elf-gcc isn't in yout path
01:43:53Matt65can anyone point me in the right direction atleast?
01:44:32linuxstb_Matt65: Which instructions did you follow? manual or automatic install?
01:44:33saratoga22222rbclient@biosgroup-desktop:~$ arm-elf-gcc
01:44:33saratoga22222arm-elf-gcc: no input files
01:44:35saratoga22222rbclient@biosgroup-desktop:~$ arm-elf-gcc
01:44:35saratoga22222arm-elf-gcc: no input files
01:45:15linuxstb_saratoga22222: Are you ssh'ing to the machine to login as rbclient?
01:45:22saratoga22222linuxstb_: yes
01:46:14linuxstb_Is the path set in ?
01:46:37saratoga22222linuxstb: I set it in .profile
01:47:10 Join |AhIoRoS| [0] (n=ahioros@
01:48:20 Join techhelper1 [0] (
01:48:57saratoga22222ah redefines the path
01:50:26linuxstb_It doesn't redefine it, but it can add extra directories to the path
01:50:36saratoga22222hmm that shouldn't matter
01:51:29saratoga22222can i invoke that script on my own to see whats happening?
01:51:38linuxstb_I would guess that the build script is invoking a non-interactive shell, hence .profile isn't read - just .bashrc
01:52:37linuxstb_That script takes a lot of complex parameters - it's not straightforward to run outside the build system
01:53:03linuxstb_If I was you, I would simply fix the path in - that's what that line is for.
01:53:16saratoga22222i don't understand what it should be
01:53:23saratoga22222the default seems to point to the binary
01:53:27saratoga22222but shouldn't it be the path?
01:53:48linuxstb_What does the PATH line contain in your copy of
01:54:17linuxstb_Mine is $ENV{'PATH'}.=":/usr/local/sh-1/bin:/usr/local/coldfire/bin:/usr/local/arm-elf/bin";
01:54:47saratoga22222my appears to have vanished somehow
01:57:34saratoga22222made a new one
01:57:43saratoga22222so i have to wait for a commit to see if its working?
01:58:45linuxstb_I wonder if something/someone deleted that file because of the errors... The latest build was clean.
01:59:42linuxstb_But yes, you need to wait for a commit to see if it's working. You could just commit a trivial change somewhere though...
02:00:03*linuxstb_ has run out of commits
02:01:18saratoga22222i'll do that
02:02:33linuxstb_Matt65: That's obviously unexpected behaviour... What install instructions did you follow? (manual or automatic)?
02:04:09 Quit Matt65 ("CGI:IRC (EOF)")
02:08:19saratoga22222hmm my machine didn't build anything
02:08:50 Quit kushal_12_27_200 ("This computer has gone to sleep")
02:09:33linuxstb_Yes, the latest set of builds happened when your was missing.
02:10:01*linuxstb_ needs to sleep, goodnight.
02:14:28Beta2KNot allowed
02:14:33Beta2KDevelopers can't sleep
02:19:10 Join gromit` [0] (
02:22:57 Quit Thundercloud (Remote closed the connection)
02:25:52 Quit XavierGr ()
02:27:53 Join argumentD [0] (
02:30:56 Quit nas ("Leaving")
02:33:34 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
02:40:52 Join Shaid3 [0] (
02:40:59 Join HP_Administrator [0] (
02:41:38 Nick HP_Administrator is now known as [FS] (
02:41:42 Quit [FS] (Client Quit)
02:42:35 Join ol_schoola [0] (
02:50:29 Quit saratoga22222 ("CGI:IRC (EOF)")
02:50:54 Join robin0800 [0] (
02:51:13 Part pixelma
02:52:43 Quit argumentD (Read error: 110 (Connection timed out))
02:56:07 Quit robin0800 (Read error: 104 (Connection reset by peer))
02:57:49 Quit Shaid (Read error: 110 (Connection timed out))
02:57:49 Nick Shaid3 is now known as Shaid (
02:58:01*kugel has a custom list version with %V-like syntax ready
03:02:10***Saving seen data "./dancer.seen"
03:07:36 Quit |AhIoRoS| ("Abandonando, see you")
03:11:51 Join shnee [0] (
03:14:26 Part shnee ("Konversation terminated!")
03:29:55 Quit framo ("moo")
03:35:09 Quit DerDome (Nick collision from services.)
03:35:10 Join DerDom1 [0] (
03:35:20 Nick DerDom1 is now known as DerDome (
03:39:08 Join shnee [0] (
03:41:13 Join joshin_ [0] (n=joshin@VDSL-130-13-184-242.PHNX.QWEST.NET)
03:41:19kugelJdGordon: ping
03:43:28kugelJdGordon: What happened to your idea of "viewportifying" quickscreen? Was a bit quite after your (horrible :) ) commit some weeks ago. Even though, quickscreen with viewports would be nice (escpecially since I want to make the quickscreen consistent with the (custom) list, see FS #8799)
03:46:04 Join ctaylorr [0] (
03:46:29 Join kushal_12_27_200 [0] (n=kushal@
03:52:30 Quit joshin (Read error: 110 (Connection timed out))
03:53:59 Join divadevidonut [0] (i=divadevi@gateway/tor/x-c505ec10fa5261f1)
03:54:39 Quit BobShield (Read error: 104 (Connection reset by peer))
03:59:56 Join piga [0] (
04:05:13 Join miepchen^schlaf_ [0] (
04:10:18 Quit ctaylorr (Read error: 110 (Connection timed out))
04:11:05 Join MarcGuay [0] (n=d8ef54e2@gateway/web/cgi-irc/
04:14:06 Quit basti (Remote closed the connection)
04:15:26 Join basti [0] (i=bnc@
04:16:31 Quit MarcGuay ("CGI:IRC (Ping timeout)")
04:16:45 Join MarcGuay [0] (n=d8ef54e2@gateway/web/cgi-irc/
04:17:34 Quit MarcGuay (Client Quit)
04:17:36 Join MarcGuay [0] (n=d8ef54e2@gateway/web/cgi-irc/
04:19:08 Quit MarcGuay (Client Quit)
04:20:22 Join MarcGuayArg [0] (n=d8ef54e2@gateway/web/cgi-irc/
04:21:07 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:25:40 Quit divadevidonut ("Leaving")
04:25:50 Quit MarcGuayArg ("CGI:IRC (Ping timeout)")
04:45:05 Join digger [0] (n=4ab678c1@gateway/web/cgi-irc/
04:45:19diggerhey can someone help me
04:45:29diggercan someone help me
04:46:26scorchenot if you dont ask a question...
04:46:34diggeri have a ipod nano 2cd gen
04:46:40diggerand i cant seem to get it to work
04:46:56scorchedid you look at the front page?
04:47:17scorcheyou didnt even look at
04:47:36diggeryea at the builds but nano 2cd gen isnt on there
04:47:46scorchelook at ...
04:48:23diggerso it dont work with 2cd gen nanos
04:48:44diggeris there any other things like rockbox that do
04:48:53diggerthat bullshit
04:49:04diggerumm are they wokring on 2cd gen nanos?
04:49:28scorcheno "they" doesnt quite work like that
04:49:45diggerwell thanks for the help ill try and trade with my sister she has a first gen
04:49:56scorchepeople with the device and the ability come forward and do the port
04:50:29 Quit digger ("CGI:IRC (EOF)")
04:56:17 Quit piga ("Leaving")
05:02:13***Saving seen data "./dancer.seen"
05:05:25 Quit JdGordon (Read error: 110 (Connection timed out))
05:06:10 Join JdGordon [0] (
05:13:59 Join Bl2 [0] (
05:14:11 Part Bl2
05:19:42 Join quitte_ [0] (
05:33:15 Join herbi [0] (
05:33:17herbihi guys
05:33:27herbiis there a bpm counter plugin for rockbox?
05:37:37 Quit quitte (Read error: 110 (Connection timed out))
05:38:42 Quit Horscht ("")
05:55:19 Quit kugel ("ChatZilla 0.9.81 [Firefox]")
06:06:00 Quit techhelper1 (Remote closed the connection)
06:08:16 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
06:08:16 Quit jhulst (Read error: 113 (No route to host))
06:09:17 Join Insectoid [0] (
06:10:42InsectoidiPOD video 5g, empty configuration file with voice. Whenever I attempt adding a directory to playlist (context menu, add...) the player locks up and when I check it in windows, the .playlist_control file is empty.
06:11:21InsectoidThe drive continues to hum, but no sound of reading, just the high-pitched drive powered sound
06:12:45InsectoidSo far I've run chkdsk /R on the device, and also jkdefrag... I didn't expect anything but I figured it wouldn't hurt.
06:19:36 Join csc` [0] (n=csc@archlinux/user/csc)
06:30:24 Join csc`` [0] (n=csc@archlinux/user/csc)
06:33:27 Quit csc` (Nick collision from services.)
06:33:29 Nick csc`` is now known as csc` (n=csc@archlinux/user/csc)
06:43:42 Join toffe82 [0] (
06:55:23 Quit csc` ("Powering Off")
06:56:50 Join csc` [0] (n=csc@archlinux/user/csc)
06:57:03 Quit Insectoid ()
07:02:15***Saving seen data "./dancer.seen"
07:04:04 Nick fxb__ is now known as fxb (
07:08:22 Nick miepchen^schlaf_ is now known as miepchen^schlaf (
07:15:21 Join linuxstb__ [0] (
07:21:52 Quit corevette (Read error: 104 (Connection reset by peer))
07:22:12 Quit midgey ()
07:22:39 Join corevette [0] (
07:26:21 Quit linuxstb (Read error: 110 (Connection timed out))
07:26:40 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
07:29:04 Quit linuxstb_ (Read error: 110 (Connection timed out))
07:32:51 Part toffe82
07:32:52 Quit Zom (Remote closed the connection)
07:33:03 Join Zom [0] (
07:33:06 Join daurnimator [0] (i=daurn@unaffiliated/daurnimator)
07:43:52 Nick fxb is now known as fxb__ (
07:51:59 Quit Shaid (Read error: 110 (Connection timed out))
08:04:44 Quit miepchen^schlaf ()
08:07:28 Quit BigBambi (Remote closed the connection)
08:08:17 Join corevette_ [0] (
08:09:11 Quit perrikwp (" ajax IRC Client")
08:10:27 Quit corevette (Read error: 104 (Connection reset by peer))
08:27:54 Join Rob2223 [0] (
08:33:22 Quit corevette_ (Read error: 110 (Connection timed out))
08:37:59 Quit shnee ("Konversation terminated!")
08:43:54 Quit DerDome (Read error: 110 (Connection timed out))
08:45:44 Join davina [0] (
08:46:32 Quit Rob2222 (Read error: 110 (Connection timed out))
08:50:25 Join Buschel [0] (
08:51:26 Join ender` [0] (
08:51:43 Quit chrisjs169 (Read error: 113 (No route to host))
08:59:17 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:00:01 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
09:01:58 Quit cool_walking_ (Remote closed the connection)
09:02:20***Saving seen data "./dancer.seen"
09:03:00 Join Mathiasdm [0] (
09:07:52 Join corevette [0] (
09:22:20 Join petur [50] (n=petur@rockbox/developer/petur)
09:22:34 Join disorganizer [0] (n=c2785409@gateway/web/cgi-irc/
09:26:23markungood morning
09:27:28*petur is still a bit pissed for riding at work without music - r16830 crashed on WPS loading. Current svn seems to have fixed it...
09:27:47LinusNrockbox sucks!
09:28:11markunLinusN: it sure does!
09:28:16peturI didn't even manage to make it sound like a bag of shit
09:28:39markunLinusN: are you going to skip the meizu fun?
09:28:58LinusNi'm seriously considering buying one
09:29:11LinusNif i only could buy me some time as well :-)
09:29:59*petur would love to have a peek at the iriver E100 if he had any time :/
09:30:13markunaccording to kgb2008 there are 2 versions of the M6SL with different LCD modules:
09:30:45markunI couldn't find the datasheet for the S6D0154 yet
09:31:44markunthe strange thing is that it seems to be for a 320x240 panel, while the S6D0129/139 is for a 240x320 panel
09:34:01LinusNis that philips?
09:34:26 Quit Buschel ()
09:37:19markunthe lcd driver is samsung
09:39:04 Quit GodEater (Remote closed the connection)
09:40:37 Nick linuxstb__ is now known as linuxstb_ (
09:45:47 Join ch4os_ [0] (n=ch4os@unaffiliated/ch4os/x-059673)
09:46:51 Quit disorganizer ("CGI:IRC")
09:47:37markunLinusN: at least I could find the LCD init for that chip in the firmware:
09:48:49LinusNi just requested a datasheet from samsung
09:48:51markunthe other 3 are here:
09:49:31markunI have the one of the S6D0129
09:56:50 Join GodEater [0] (i=c2cbc962@rockbox/staff/GodEater)
09:58:09 Join disorganizer [0] (n=57b2711f@gateway/web/cgi-irc/
09:59:33 Quit csc` ("Powering Off")
10:07:12 Join tvelocity [0] (
10:08:18 Join gevaerts [0] (
10:10:39markunLinusN: does the LCD init sequence for the M6TP look familiar?
10:12:56LinusNnot really
10:16:08 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
10:25:35amiconnmarkun: Why is that strange? Using an lcd controller + panel "rotated" seems to be a common thing looking at our targets...
10:25:58markunamiconn: but using 2 different orientations in the same firmware..
10:26:20LinusNgevaerts: there?
10:26:24amiconnThe ipod color also has 2 very different lcd controllers
10:26:57amiconnPlus, using different orientations doesn't necessarily mean a different framebuffer layout. Most newer lcd controllers have switchable update directions
10:27:36markunI didn't find such an option in the S6D0129 datasheet, but I might have missed it
10:27:55markunor do you mean the LCD controller part in the SoC?
10:28:26gevaertsLinusN: yes
10:28:40LinusNregarding the memory address contention theory
10:29:14*gevaerts listens, hoping to gain new insights
10:29:28LinusNi don't know an awful lot about the portallayer, but is there a separate code and data cache?
10:29:46LinusNit's unified, isn't it?
10:30:43gevaertsI'm not sure, but I didn't find any mention of separation in MrH's memory controller document
10:31:02LinusNi was thinking what would cause the contention
10:31:25amiconn8KB unified cache, per core
10:31:32LinusNfirst of all, we could have the wrong dma priority
10:31:50*pondlife wonders if any decision on date/location of DevCon2008 (Euro) will occur.
10:31:59*LinusN too
10:32:03pondlifeCalling all Swedish Dictators....
10:32:24LinusNi was under the impression that the Berlin maffia was investigating
10:32:51pondlifeDate is more critical for me, really.
10:32:53LinusNif not, i don't see any problem with having it in stockholm once again
10:33:01pondlifeI can do either Berlin or Stockholm
10:33:31LinusNgevaerts: do you have a clue to what code is executed the most in usb mode?
10:33:32pondlifeWould slightly prefer Stockholm as (a) I've never been and (b) it's the Home Of Rockbox
10:33:58 Join tvelocity_ [0] (
10:34:01markunpondlife: don't expect to see much of stockholm during the devcon :)
10:34:02LinusN"Stockholm - the home of Rockbox" - has a certain ring to it
10:34:17LinusNi'll call the mayor
10:34:28pondlifemarkun: Haha, that's not a problem... I just like seeing new airports.
10:34:30gevaertsLinusN: from what I can see, only the tick handlers and (if the code uses double buffering) the ata driver
10:34:41 Quit disorganizer ("CGI:IRC (Ping timeout)")
10:35:05markunI would prefer Berlin I think
10:35:05LinusNgevaerts: what if the scheduler ran in IRAM?
10:35:33pondlifeHmm, looking at - shouldn't "Show Files - Supported" include video files?
10:36:03LinusNgevaerts: it could be that the ATA register accesses are synchronized, and has a bad impact on the bus
10:37:12*petur could go by car to Berlin and pick up markun on the way ;)
10:37:30markunthat would be even better :)
10:37:34gevaertsLinusN: ATA (or SD for sansa) does contribute to the problem, but it's not the only cause. And while putting the scheduler in IRAM might help, I doubt if it's a good solution in the long term
10:37:38markunor we can go to stockholm by car ;)
10:38:24LinusNgevaerts: so maybe dma priority is the issue then
10:38:28*petur found tickets for Stockholm for 10 euro :)
10:38:31*gevaerts volunteers to share that car as well if the date is possible
10:38:53markunrockbox roadtrip
10:39:04*petur adds taxes and ends up at 98,57 euro - still not bad
10:39:21peturmarkun: trunk full of beer ;)
10:39:21 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
10:39:23 Join chrisjs169 [0] (
10:39:36LinusNRockbox Euro Tour 2008
10:39:51gevaertsLinusN: I think it's something like that, but I have no clue on how to solve it (or even where to look). The USB controller doesn't seem to use the portalplayer "sidma" engine (or at least we don't configure it - maybe that's the problem)
10:39:55*pondlife probably can't make any date before July 19th now... :/
10:40:29pondlifeEither that or 2nd May :)
10:40:46pondlifeI'm assuming it's a Friday-Sunday thing, as normal.
10:44:24pixelmathere was a change by amiconn to the DevCon2008 page yesterday...
10:44:27 Quit chrisjs169 (Read error: 113 (No route to host))
10:44:57*gevaerts runs to the wiki to have a look
10:45:37pondlife"Have free hosting location"
10:48:08*gevaerts promises to buy everyone at the DevCon a beer if the date is decided this week
10:48:33*pondlife decides on a date
10:48:39pondlifeCan I have my beer please?
10:49:03gevaertspondlife: not _a_ date, _the_ date
10:49:06*petur gets overloaded by triggers
10:49:13pondlifegevaerts: Pedant :p
10:49:59*pondlife would hope petur has a suitable sound associated with his beer trigger
10:51:02 Quit tvelocity (Read error: 110 (Connection timed out))
10:51:03 Join handmadematters [0] (
10:51:30markunBagderr: did you get any thinking done since last may? :)
10:52:31Bagderrthat's fallen into a large black hole
10:53:10Bagderrbut yeah, I should pick up that issue again and see what I can do about it
10:53:12 Quit HellDragon (Connection timed out)
10:53:48pixelmaGodEater: was my little rant about using "not supported" instead of "doesn't run" in the "ipodpatcher.exe..." thread too subtle? ;\
10:53:55LinusNlike my project manager said, "It's on my list of priorities"
10:54:05*pondlife adds a potentially controversial discussion point to the wiki
10:54:23handmademattersIsnt RbUtil running on OSX 10.4 ??? As far as i remember i saw something like min version 10.4 in the Subversion log
10:54:43Bagderr*more* playlist-centric?
10:54:51*LinusN revokes pondlife's wiki write access
10:55:56GodEaterpixelma: where did I screw up ?
10:58:17pixelma (I think it' important)
10:58:54GodEaterI didn't post there did I?
10:59:26pixelmano but you used it
10:59:27*gevaerts proposes to change the playlist into an "action list", so e.g. changing the volume will be done by inserting a "change volume action" at the appropriate point in this list. This will enormously simplify the GUI
11:00:11GodEaterpixelma: right - but there I was quoting one of the previous posters...
11:00:19pixelmaGodEater: or do you disagree?
11:00:29 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
11:00:37*gevaerts puts on his "don't take me too seriously" shirt
11:01:52*GodEater clarifies his post.
11:02:13pondlifegevaerts: Playlists should ultimately support videos, why not also cfg files?
11:02:24***Saving seen data "./dancer.seen"
11:02:59GodEaterpixelma: is that better ?
11:03:02 Join Thundercloud [0] (
11:03:06gevaertsActually, now that you mention it, that might be useful in some circumstances (e.g. <insert far-fetched use case here>)
11:03:35pondlifegevaerts: Switching shuffle on and off...?
11:03:56pondlifeImagine a playlist that tells itself to shuffle... ;)
11:04:07pondlifeOn second thoughts, don't.
11:04:19pixelmaGodEater: yes, thanks
11:04:31barrywardellhandmadematters: it should run on any version since 10.4
11:04:48*gevaerts was thinking more along the lines of changing the WPS and/or backdrop to suit the currently playing music. Much more powerful than this 'album art' thing
11:05:42pondlifeEither way, the current UI has evolved such that the playlist viewer is rather peripheral. Stuff like it not working with iTunes filenames are important I think.
11:06:22pondlifeWhich comes back to how we deal with playlists generally - generation of big playlists could be speedier I hope.
11:09:25 Join DerPapst [0] (
11:09:55*DerPapst wonders when gevaerts idea makes it into flyspray as a request...
11:11:59*Bagderr wonders how much longer we will have a feature-request tracker
11:13:24Bagderrcan anyone think of a recent feature request being filed that was useful to us?
11:13:30*DerPapst suggest to sut it down it and add all requests to the gsoc08 page, removing the red on top of the page.
11:13:41DerPapst*shut even
11:15:34linuxstbBagderr: I agree the feature tracker has become useless, but I wouldn't like to lose it completely - e.g. if someone mentions (e.g. in the forums, mailing lists or IRC) a good idea, we should be able to keep track of it somehow.
11:15:52linuxstbWould limiting feature requests to devs be a bad idea?
11:16:06*linuxstb runs
11:16:18 Quit linuxstb_ (Remote closed the connection)
11:16:31DerPapstpeople will keep making requests. even with that limitation. Either as bug or as patch
11:17:10 Join Shaid [0] (
11:17:14gevaertsCan we add a "feature requests" section to the forum that is invisible to the devs ?
11:20:05pondlifeYep, keep the FR option so people don't clog up the bug tracker.
11:20:25 Join disorganizer [0] (n=c2785409@gateway/web/cgi-irc/
11:20:42DerPapstmaybe auto-reject all feature requests after a week? :-P
11:21:10pondlifeFRs are easy to ignore.... ;)
11:21:18disorganizerjust make a bot closing them stating "this features is not work the effort and binary size"
11:21:20BagderrI think we already pretty much ignore them
11:22:24BagderrI can see the tracker being useful for devs adding ideas so that they aren't forgotten, I see no use at all for Joe User to submit random wild thoughts
11:22:38disorganizerbut to be honest: why not run a script once a day closing every feature request which did not get an answer within a week?
11:22:52pondlifedisorganizer: Because good ideas would get lost?
11:23:03pondlifeIt's for long-term storage of ideas
11:23:04disorganizerno, you can also see closed threads
11:23:15BagderrI think limiting it to allow devs only to submit them would be good enough
11:23:38pondlifeAlthough that might result in more invalid bug reports being filed...
11:24:00pondlife..which I guess we can deal with, but not say "log an FR".
11:24:12disorganizerbut there also good ideas may be lost just because someone is a non-dev. propably having feature-request being filed by users via the forum and devs being able to put them on the tracker if seen fit would be the best idea?
11:24:12gevaertsI see two reasons : (1) if a dev adds an idea just to remember it, it's pretty likely that it won't get a comment soon, and (2) that would be asking for duplicates. At least now some people search the list and add a comment to an existing entry
11:24:49Bagderrauto-closing seems pointless to me
11:25:12*gevaerts was talking about why not to auto-close, in case that wasn't clear
11:25:24pondlifeDoes Flyspray know who's a dev?
11:25:42*JdGordon rudly barges in on the convo :p
11:25:45 Nick disorganizer is now known as dis_afk (n=c2785409@gateway/web/cgi-irc/
11:25:59Bagderrthere are "groups" in flyspray and each user has one assigned to them
11:26:00pondlifeJdGordon: That wasn't rude, that was useful
11:26:09*pixelma wanted to just comment on a feature request but then decides to close the task...
11:26:10Bagderronly devs are allowed to close tasks etc
11:27:10*gevaerts points out that the dev groups on forum, svn and fs are not synchronised automatically
11:27:21JdGordonor at all?
11:27:30markunBagderr: what do you think about adding a "HAVE_ATA" instead of checking for model numbers in firmware/SOURCES?
11:27:55Lloreanpondlife: On a note relating to something you said earlier, I really would kinda like to see playlists allowing any "supported" file type as long as it didn't break playlist auto-creation (auto-creation should still be restricted to things for which there are .codecs I think, or possibly have a mode setting between the two).
11:27:56Bagderrsounds like an improvement to me
11:28:17pondlifeLlorean: That was gevaerts' idea, I should point out!
11:28:24pondlifeBut yes, I agree with you
11:28:27Bagderr(that was for markun)
11:28:41*gevaerts puts rockbox.mi4 in his playlist
11:28:53Lloreanpondlife: Well, he mentioned WPSes in it. You mentioned videos.
11:29:07LloreanAt least, that I saw in the logs on my tiny phone screen.
11:29:15pondlifeLong term, I'd hope videos would become codecs too.
11:29:49 Join DerPapst060 [0] (
11:29:49JdGordonLlorean: .txt and .mu3 in a playlist?
11:29:59 Quit DerPapst (Nick collision from services.)
11:30:05LloreanJdGordon: If users want to, sure. I can't think of a real use for .txt, but I wouldn't mind .m3u
11:30:09 Nick DerPapst060 is now known as DerPapst (
11:30:23pondlifem3u with recursion...?
11:30:35gevaertsThat could get tricky
11:30:36LloreanYou can already "insert" a .m3u, and get all the songs within it.
11:30:58pondlifeWhich is enough, IMHO.
11:31:36LloreanWell, there's a feature request for including folders, which i suppose would be dynamically expanded on playlist load.
11:31:36gevaertsWhat if this second .m3u contains the first one ?
11:32:07LloreanI'd say including a .m3u should include expansion on load as well. (And should not be able to be recursed, so remember filenames of .m3us loaded)
11:32:40LloreanThough not expanding them on load, would be another means of handling shuffling albums without using database...
11:34:45Lloreanpondlife: I'll settle for directories and extm3u support though. :-P
11:34:55gevaertssub-playlists need to be thought about, but I think that it would be a good thing to allow basically any supported file type to be included in a playlist. Whether the UI allows them all is a different question
11:36:16*gevaerts dreams about automated bug finding through playlist-triggered RoLo
11:37:19pondlifeLlorean: This is actually a distraction from my point. I think the UI is too browser-centric where it should be playlist-centric. This is a personal opinion, of course; I wouldn't want to prevent the existing setup from being used.
11:37:36Lloreanpondlife: More playlist centric in what way?
11:38:10pondlifeI'd like to be able to have the playlist viewer as my main screen, and also think it should handle tags (not rely on filenames)
11:38:19LloreanHandling tags is a difficulty.
11:38:24LloreanWhy not EXTM3U?
11:38:33pondlifeIt seems a hidden, unloved cousin at the moment
11:38:40LloreanWhen Rockbox creates a playlist, it can embed tags as EXTM3U data anyway.
11:38:59pondlifeEXTM3U would be a solution.
11:39:26pondlifeThe playlist creation is rather slow at the moment anyway, so it might all need rework
11:39:45markunI don't see IAUDIO_7 defined anywhere
11:40:53Lloreanpondlife: A couple things I'd like to see about playlists, while we're at it: It marks an item as "the currently playing item" when playback is stopped, I think. This is bad (especially since it doesn't mark the resume point anyway).
11:40:55barrywardellsince cabbiev2 is the default theme now, do people think the default background/foreground colours should be black/white?
11:41:45pondlifeLlorean: Stoped or not should make minimal difference to the playlist
11:41:49Lloreanpondlife: And if the playlist is going to be still visible while stopped, the "Insert" option should actually *insert* in the visible playlist, and "Play Now" should be the new first option on the insert menu to preserve old behaviour.
11:42:02Lloreanpondlife: Well, used to be there was no "playlist" while stopped.
11:42:07LloreanNow there is.
11:42:12LloreanOr rather, there sorta is.
11:42:13pondlifeThere definitely should bne
11:42:13JdGordonbarrywardell: na, leave them as is
11:42:22Bagderrmarkun: that port hasn't gotten anywhere so it could very well still have basic flaws
11:42:33Lloreanpondlife: I don't see how there's a "Current" playlist while stopped.
11:42:35barrywardellJdGordon: why?
11:42:41pondlifeLlorean: Why not?
11:42:42JdGordonfor the purists :p
11:42:43LloreanI'd like the option to be renamed "Previous Playlist"
11:42:53pondlifeStill current
11:42:54LloreanBecause right now, "Insert" inserts in an empty playlist.
11:43:06LloreanSo the actual behaviour of the OS conflicts with the current designation.
11:43:13pondlifeThat would change - we discussed before
11:43:15LloreanBasically, things need to make up their mind one way or another.
11:43:29LloreanBut is there a "Current" song when stopped then?
11:43:38pondlifeIt's just not playing
11:43:43pondlifePress play and it resumes
11:43:56pondlifeSo there's a current song, and song position at all times.
11:44:12pondlifeIF a playlist ends, then there is no current song
11:44:37pondlifeCurrently it's a bit confused
11:44:45LloreanThat seems kinda silly to me. You can "resume" a bookmark too, but it's not "current"
11:45:18pondlifeResuming a bookmark just sets up the corresponding playlist and position
11:45:27LloreanSo does resuming by way of the play button.
11:45:37pondlifeYes, into the current playlist
11:45:43LloreanInto the last used playlist.
11:45:56pondlifesemantics... call it current...
11:46:16pondlifeStop doesn't destroy a playlist though
11:46:21LloreanThis whole discussion was *on* semantics in the first place, I asked if "Current" was a valid term.
11:46:22pondlifeStarting a new one does
11:46:24LloreanNo, stop shuts down playback.
11:46:36LloreanIf you close Winamp, do you still have a current playlist?
11:46:46pondlifeNo idea
11:46:49pondlifeI don't use winamp
11:46:57 Quit Nevtus (Read error: 110 (Connection timed out))
11:47:03LloreanIf you close media player X, do you still have a current playlist when it's not running?
11:47:04pondlifeBut I would like to start it up and have it resume where I last wa
11:47:16pondlifeLlorean: Yes, I do
11:47:22GodEaterpondlife: winamp does that I think
11:47:41 Join Nevtus [0] (n=Nevtus@unaffiliated/nevtus)
11:47:51pondlifeI dion't see a need to clear the current playlist until you start a new one.
11:48:02GodEaternor me
11:48:05pondlifee.g. by selecting in a browser or loading a new one
11:48:45LloreanThere's not a "need" to misname it though, either.
11:48:58LloreanIf you call it "Previous", people know it's not running right now. It makes it clear playback is stopped.
11:49:04markunBagderr: ok to commit?
11:49:06pondlifeNo, that's the point
11:49:12LloreanI'm not saying it shouldn't be cleared, I'm discussing the nomenclature.
11:49:14pondlifeIt is still current, but stopped
11:49:20LloreanIt's also Previous.
11:49:24pondlifePrevious could refer to the one before that
11:49:27LloreanIt is not active.
11:49:38pondlifeCurrent, but not active.
11:49:39LloreanOkay, if I press stop, no song is playing.
11:49:40Bagderrmarkun: looks fine, shoot!
11:49:47LloreanSo, the song that was previously playing is the "previously playing song"
11:49:48GodEateruntil it gets cleared, I'd call it current
11:49:57LloreanAnd the playlist that was previously running was the "previously running playlist"
11:50:00BagderrI modified the source tarball script, I need a commit to see how it works ;-)
11:50:04pondlifeYes, but the playlist is still active and valid
11:50:09 Join webguest78 [0] (n=1485000d@gateway/web/cgi-irc/
11:50:14LloreanIt's valid, but not active.
11:50:22GodEaterit is playback which has become inactive
11:50:23pondlifeI think current song and current playlist are easier terms
11:50:24GodEaternot the playlist
11:50:29markunBagderr: I just wandered if MMC and ATA could exist in the same device. The #else makes it impossible.
11:50:35markunwondered even
11:50:53 Quit webguest78 (Client Quit)
11:50:58Lloreanpondlife: In what way. While we remember that they're there, they are not in use, right now. That's kinda what current means: right now, active.
11:51:00Bagderrmarkun: the only mmc devices are the ondios afaik
11:51:13LloreanThey are not in use.
11:51:18LloreanThat's explicitly what stop DOES, it stops using them.
11:51:19GodEaterit's still there
11:51:23LloreanPause leaves it in use.
11:51:31LloreanGodEater: It's always there, on disk, somewhere.
11:51:39GodEaterit's also still in memory
11:51:43markunBagderr: don't we have players with sdcard slots?
11:51:45barrywardellJdGordon: by "for the purists" you mean nostalgic reasons? In that case I'll leave it
11:52:05GodEatermarkun: we think the Tatung Elio is the only player with an sdcard slot
11:52:08pondlifeStop and pause should behave the same in this regard. Although the UI stays on the WPS for pause.
11:52:17LloreanGodEater: Only in a temporary sense. If you "Insert" a single song, it clears.
11:52:28Lloreanpondlife: No, "Stop" needs to clear the playback buffer for other features to work.
11:52:53LloreanUsers should know that there are distinct differences between "Stopped" and "Paused", and distinct naming between things helps illuminate this.
11:52:56pondlifeYes, but that will be changed into 2 options "Insert into new playlist" (first) and "Insert into current playlist"
11:53:01GodEaterLlorean: but if I choose to save it, it doesn't get cleared, and is therefore still current
11:53:40pondlifeThere is generally always a current playlist - it might be dynamic or saved, but it's always there. Except when you just installed Rockbox.
11:53:43LloreanGodEater: It's inactive until you resume it.
11:54:15Bagderrmarkun: still, I don't think your patch makes that any worse, it still clarifies the situation
11:54:19GodEaterno, playback is inactive
11:54:23Lloreanpondlife: Okay, I've explained how "Current" can confuse users. Explain how "Previous" could confuse users?
11:54:28GodEatersorry, I completely agree with pondlife here...
11:54:42pondlifeI think of previous as the playlist I was listening to before the current one
11:54:47LloreanPeople commonly ask how to clear their playlist. If they knew it was actually inactive while stopped, it would help.
11:54:52pondlifeprevious...current... next
11:55:15LloreanIf you've stopped playback, the one you were previously listening to, as you ARE NOT listening right now, was what was "Current" when you were listening.
11:55:16pondlifeAssuming you mean on insert, if the option read "Insert into new playlist" it would be clear
11:55:25LloreanIt's called "time", and it doesn't pause when you hit "stop"
11:55:35pondlifeLlorean: I might stop for a phone call..
11:55:44LloreanThen hit the pause button...
11:55:57pondlifePause doesn't work with voice
11:56:02 Quit DerPapst (Nick collision from services.)
11:56:11Lloreanpondlife: That's a problem that needs fixing, not a reason for naming things.
11:56:18 Quit barrywardell ()
11:56:19LloreanHow can "In the past" be "Current"
11:56:21 Part handmadematters
11:56:22LloreanIt's a contradiction in terms.
11:56:24pondlifeIt's not in the past
11:56:27pondlifeIt's still there
11:56:29LloreanYou've shut down the playback engine.
11:56:36GodEaterbut the playlist hasn't gone anywhere
11:56:36 Join DerPapst [0] (
11:56:37LloreanThe playlist is a definition of the files the playback engine goes through
11:56:42LloreanIt is not used by the rest of the OS
11:56:57LloreanThe Playlist Viewer has the capability of showing you and allowing you to manipulate the previous playlist if playback is stopped.
11:57:03pondlifeYes, and those files are still there. pressing play shouldn't even need to read the disk.
11:57:16LloreanGodEater: It's *gone*. Just because you bring can bring it back instantly doesn't mean it isn't gone until you do.
11:57:24pondlifeIf you think of it as the current playlist, it makes much more sense
11:57:35GodEaterit's not gone, it's still there in the playlist viewer
11:57:49pondlifeWe're not switching back and forth here, we're dealing with the same entity.
11:58:00pondlifeIt's still current until you start a new one
11:58:15Lloreanpondlife: It has to reload from disk when you hit resume, so I don't understand your "shouldn't"
11:58:36 Join PaulJam [0] (
11:58:38pondlifeIt shouldn't need to, that's just a weakness in the current code
11:58:45Lloreanpondlife: By that logic, any text file I close is "the one I'm currently working on" until I start a new one, be it weeks later. It's silly
11:59:11pondlifeSTOP != CLOSE
11:59:22LloreanYou're closing the playback system
11:59:30LloreanMy whole point is that users need to *see* that it's being closed.
11:59:43pondlifeBut that makes Rockbox less flexible
11:59:51LloreanNAMING makes rockbox less flexible?
12:00:00LloreanBecause other things depend on playback being closed down, so the distinction between stop and pause needs to be illustrated, not swept under the rug.
12:00:06 Join handmadematters [0] (
12:00:16pondlifeWhat other things?
12:00:19 Part handmadematters
12:00:22LloreanAnything that needs the audio buffer.
12:00:38LloreanSeveral plugins, changing settings that require claiming more RAM, etc.
12:00:43pondlifeYes, but that's just a coding thing. The playlist remains.
12:01:08LloreanSo, there's nothing currently playing, but there is a current playlist?
12:01:20pondlifeNaming is just to make things clearer
12:01:26LloreanAnd calling it the "previously playing playlist" is too confusing for users?
12:01:27GodEaterI can do the same thing in any number of media players
12:01:34GodEaterhitting stop does no clear the playlist, it's still there.
12:01:47LloreanLet's take it to the users. "If you press stop, is the list you were listening to the current playlist, or the previous one?"
12:02:14GodEaterI don't see how you can HAVE a previous one until you have a new current one.
12:02:19LloreanFind out which is least confusing to the masses, because I think your naming makes the distinction less obvious, and fuzzes things.
12:02:32pondlifeThe changeover point is when you start a new playlist, surely.
12:02:38LloreanGodEater: Why do you need a new current file playing to have a previous file playing?
12:02:39pondlifeNot when you press stop
12:02:53LloreanBy that rule, when the playlist ends naturally, it's still "Current"
12:02:59pondlifeIndeed it is
12:03:00GodEateragain yes
12:03:12LloreanSo even though there's "Nothing to Resume" there's a current playlist?
12:03:13*Bagderr added the size of the source tarball on the listing
12:03:13LloreanThat's absurd.
12:03:18GodEaterno it's not
12:03:27GodEaterI can choose to step back through the playlist without changing it
12:03:27pondlifeNo. The playlist is current, but the current song is off the end of the playlist.
12:03:42LloreanOkay, take THAT to our user base and see how many agree with you.
12:03:45pondlifeLlorean: How much do you use the playlist viewer?
12:03:48LloreanThe current song is not in the current playlist?
12:03:55pondlifeIt is
12:04:03Llorean"The playlist is current, but the current song is off the end of the playlist."
12:04:09Llorean"off the end" sounds like "outside" or "not in" to me.
12:04:17pondlifeI mean the current song is nothing at that point
12:04:22pondlifeBut not outside the playlist
12:04:35*Llorean couldn't possibly have known you meant that.
12:04:41GodEateryes he could
12:04:53GodEaterit was obvious to me at any rate
12:04:54LloreanGodEater: How was I supposed to know that his words meant other than what they say?
12:05:21pondlifeI just think of the current track and offset as being based in the current playlist.
12:05:34pondlifeSo I may have assumed you thought the same.
12:05:34GodEaterwhich they can't help but be
12:05:58LloreanSo the current song is entry 101 in a 100 entry playlist.
12:06:22LloreanWhich means the playlist is still active, the song is in the playlist, despite not actually being defined within or by the playlist, and the playlist is current despite the fact that you're not in it?
12:06:33GodEaterthere's nothing playing
12:06:41LloreanSo there's no current song.
12:06:44GodEaterthere doesnt' have to be a current song for the playlist to no longer remain current
12:06:52pondlifeBut the playlist is still current
12:07:03pondlifeYou could view it and skip backwards to track 100
12:07:07LloreanEven though you're no longer IN the playlist, nor is it active, nor have you a playlist index?
12:07:17Lloreanpondlife: You can view *any* playlist by selecting it.
12:07:32pondlifeI'm referring to the current one
12:07:33GodEateronly by loading it from disk
12:08:10LloreanSo the fact that the previous playlist is buffered in RAM makes it always "current" even if there's no obvious sign that it's current to the user, up to and including "Nothing to Resume" messages?
12:08:22LloreanNot even "Go back in the playlist to resume" but actually "Nothing"
12:08:47*markun doesn't have a clue what these guys are discussing about
12:08:48pondlifeLike I said, the UI isn't too playlist-centric at the moment, but it should be more.
12:09:10pondlifeStart Menu > Current Playlist or similar would help
12:09:14 Join BigBambi [0] (n=86ceaf40@rockbox/staff/BigBambi)
12:09:17BigBambiAlright chaps
12:09:27amiconnmarkun: You forgot to adjust the #endif comment
12:09:39markunamiconn: yes, I did. I'll change it
12:09:44pondlifeCurrently it's hidden a bit under Playlists
12:09:46Lloreanpondlfe: Seriously, if we're going to do this, we should see which makes more sense to users. If I'm wrong, and users like "Current" better fine, but I suggest we do ask them "If you hit 'Stop', and music is no longer playing, is what you listened to the Current or Previous playlist" and see which is less confusing (more popular) to them.
12:09:52markunamiconn: I might as well use #elif defined(...)
12:09:57GodEaterLlorean: you're still wrong - I've just had a playlist finish on my player. I got to the playlist menu, and choose "View current", and "oh wow - the playlist is STILL there"
12:09:58 Join MethoS [0] (
12:10:06BigBambiFor me, it is current
12:10:20*DerPapst votes for "curret" too and hides
12:10:25LloreanGodEater: I never said you couldn't do that. What am I wrong about?
12:10:44BigBambiI think this comes from cassettes for me - the cassette may have stopped but it is still the current one on the machine. Until I play something else, it is current
12:10:52pondlifeLlorean: Put a poll up if you wish, but you won't convince me it's inconsistent at the moment.
12:11:07GodEaternor me
12:11:09pondlifeBigBambi: Cassettes, WMP...
12:11:14GodEaterCD players
12:11:20Lloreanpondlife: It's not your opinion I care about. It's "I think users would understand how Rockbox works a little better if it was named this way"
12:11:31Lloreanpondlife: I understand fully your point of view on the matter.
12:11:39LloreanAnd I'm not saying you're wrong, in a purely technical sense.
12:11:44pondlifeI agree with you, but think "Current" is clearer
12:11:52LloreanI'm saying "you're wrong, in a sense of what functionality is exposed to the user in each state"
12:11:54GodEaterLlorean: well I disagree with you - I think calling something "Previous" when there's nothing replaced it is ridiculous
12:11:59BigBambipondlife, GodEater: Sure, I just said tapes as that was my earliest form of music
12:12:10pondlifeThe point is that "Previous" makes the user think they can't get it back easily
12:12:14LloreanGodEater: Well tell my "current" girlfriend she's not spending enough time with me then
12:12:17pondlifeOr that they have to do something to get it back
12:12:18BigBambiI don't think the technical sense matters to users
12:12:19LloreanGodEater: Seriously, that statement on its own, is absurd.
12:12:35BigBambiIt is what people use in everyday language, technically correct or not
12:12:41Lloreanpondlife: So if the button said "View Previous Playlist" it would make them think that they'd need to do something other than click it to get there?
12:13:06LloreanBigBambi: The idea is to make it clear that the playlist is not actually "in use" in the "playback" sense, right now.
12:13:07pondlifeIt would make me think of some kind of playlist history
12:13:17amiconnLlorean: The list is always current until it's replaced by a new one, regardless of playback state
12:13:21Lloreanpondlife: How about "View Inactive Playlist"
12:13:32pondlifeThe playlist isn't inactive
12:13:35pondlifeThe playback engine is
12:13:37GodEatersorry the g/f comparison is spurious. You've "cleared" your playlist there.
12:13:37BigBambiLlorean: I understnqd that, I just don't think that users really care
12:14:03BigBambiI know they should care from the point of view that without freeing buffers, certain things won't work
12:14:11LloreanGodEater: Well, "Queue" and "Insert" functions all treat the playlist as having already been cleared, rather than acting like they do on a populated playlist.
12:14:33Lloreanpondlife: So how will "Inactive" confuse users?
12:14:40pondlifeIt's not relevant
12:14:49pondlifeView playlist is fine, surely
12:14:52BigBambiBut lets face it, most people couldn't care less how it works, they just want it to work, and we aren't going to change people's use of English to make a technical distinction like that
12:14:56 Join bughunter2 [0] (
12:15:09Lloreanpondlife: My point is to differentiate more clearly between "Playback Stopped" and "Playback Paused"
12:15:20pondlifeI want to reduce that distinction
12:15:33Lloreanpondlife: When playback is stopped the playlist is Present but Inactive. It is not being used, by or for anything, until you run another program that accesses it such as the viewer.
12:15:34pondlifeOr eliminate it
12:15:37LloreanPresence != Activity
12:15:43 Part Bagderr
12:15:47Lloreanpondlife: What are you going to do about the playback buffer then?
12:15:55pondlifeNot a lot
12:15:56BigBambiIgnoring technicalites, from a users point of view I don't think the difference between pause and stop should matter at all
12:16:01GodEaternor me
12:16:02Lloreanpondlife: Then you can't remove that distinction.
12:16:03pondlifeIt will rebuffer if it needs to
12:16:13pondlifeIt will not rebuffer if the data is present
12:16:16pondlifeThat's it.
12:16:24pondlifeThe user shouldn't worry
12:16:29GodEateror care
12:16:33LloreanThen there's a distinction between Pause and Stop, since Pause will prevent the actions.
12:16:37LloreanOr will "Pause" still allow the actions?
12:16:45LloreanAnd you can end up needing to rebuffer after Paused?
12:16:45pondlifePause still allows
12:16:50pondlifeYes, probably
12:17:15LloreanSo basically, your proposal is "Playback is always active, no matter what"
12:17:18pondlifePause is only of use if you need a fast resume AFAICT.
12:17:27pondlifeLlorean: No, not that radical
12:17:32LloreanThat's what you've just described.
12:17:42BigBambiEverything should 'pause', unless a subsequent action requires a 'stop', in which case a 'stop' is performed
12:17:43amiconnPause != Stop makes very much sense to me...
12:17:43LloreanYou can Pause, and you can Stop, but they do the same thing: Pause.
12:17:44GodEaterI don't see why that would be a bad thing
12:17:53GodEaterbeing as how Rockbox is for audio playback primarily
12:18:09pondlifeThis is a red herring, not my point at all
12:18:34pondlifeMy point is "We should consistently treat a playlist as valid, even after stopping playback. "
12:18:50Lloreanpondlife: There's a difference between "Inactive" and "Invalid" and it's very distinct.
12:18:54pondlifei.e. the current playlist continues to live until replaced.
12:19:21pondlifeI don't think we need to pop words like "Inactive" up to tell the user that playback is stopped.
12:19:51BigBambiThing is, from a users point of view, I don't want to have to care about Rockbox internals
12:20:01LloreanBigBambi: But unfortunately you DO have to care right now.
12:20:04*amiconn thinks the current behaviour is perfectly okay and shouldn't be changed
12:20:11LloreanAnd as long as you do, we shouldn't go out of the way to *hide* that state from you
12:20:29pondlifeIt's almost ok, but should be totally consistent
12:20:32BigBambiWe should go out of the way to make it clear though
12:20:40LloreanBigBambi: Yes, that's what I've been saying...
12:20:45LinusNit would probably be more obvious if rockbox stayed in the wps when the current playlist has ended
12:20:52BigBambiBut clear as in everyday language, not technical special cases
12:20:56LinusNor when you stop for that matter
12:21:06pondlifeNot sure I like that idea much
12:21:30pondlifeThe WPS is a bit of a boring screen with no track
12:21:33 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:21:33LloreanBigBambi: Hiding any distinction between a playlist that is actively being played, and not being played (so the buffer is free) though?
12:21:40LinusNi probably wouldn't like it either, i'm just saying that it would be more obviuos
12:21:48LloreanBigBambi: If you feel better language is needed than "Current" vs "Inactive", fine, suggest a replacement.
12:21:53BigBambiLlorean: Sure.
12:22:16LinusNpondlife: and i think it wouldn't be empty, it would display the last track with 100% played
12:22:20pondlifeAh, ok
12:22:30BigBambiWhy should a user care if the buffer is free or not? If it isn't, and it needs to be, clear it. If not, playback resumes faster
12:22:31pondlifeMight look like it's crashed
12:22:33LinusN(like many other commercial players do, btw)
12:23:21amiconnpondlife: I really don't understand what problem you're trying to solve
12:24:26amiconnThere is a current playlist. If it's playing, 'Queue' and 'Insert' act on that one. If it's not playing, these commands start a new playlist. Convenient, and consistent.
12:24:28BigBambiLlorean: Besides, I came in half way through and probably missed the point. I don't have any major issues with the current behaviour myself, I'm just trying to see it from a plain users point of view
12:24:34LinusNamiconn: i think he tries to address the confusion many users seem to have regarding dynamic playlists
12:24:54pondlifeamiconn: The bugs are elsewhere - try stopping and using View Current Playlist
12:25:14pondlife(With a big playlist generated from the database.)
12:25:19LloreanBigBambi: My only point is that "View Current Playlist" could be renamed when playback is stopped to something that makes it clear to the user that things are handled differently in this state, because it's not active (or in-use) at the moment, simply "Available"
12:25:38pondlifeLlorean: I don't think that distinction needs to be made at all.
12:25:43pondlifeIt's still Current!
12:25:47pondlifePop up your poll
12:25:51amiconnYes, of course it's current
12:25:51LinusNi agree with pondlife
12:25:57*BigBambi too
12:25:59 Join roolku [0] (
12:26:13LloreanIn a technological sense, YES.
12:26:13amiconnIt's exactly what I would expect - I see that list that will start playing when I resume
12:26:20LloreanAn a "the way Rockbox handles things from a user perspective" not so much
12:26:24LinusNamiconn: exactly
12:26:40amiconnLinusN: ?? I think you agree with pondlife?
12:26:47GodEaterLlorean: but the point pondlife is making is about changing the behaviour to match the language
12:26:51pondlifeamiconn: I get it loading and displaying a list, but with the marker at the wrong place
12:26:51GodEateryou're going the other way round
12:26:58LloreanLinusN: You can't resume the "Previous" list, once you've stopped?
12:27:16amiconnLlorean: There is *no* previous list
12:27:24LinusNi can resume the current list
12:27:29LloreanGodEater: I'd accept changing the behaviour to match the language. But changing the language to match the behaviour can be done much more quickly, and easily, before then.
12:27:39LinusNsince i can resume it, it is still current
12:27:51pondlifeI guess my main problem is that a lack of clarity is making it tricky for me to know what is a bug and what is not...
12:28:04LloreanLinusN: So if you resume a movie you'd been watching previously, it's still the movie you're currently watching, even when not watching it?
12:28:15GodEateryes again
12:28:21BigBambiLlorean: What I think is that it should be called current, but that inserting shouldn't always wipe the playlist - perhaps insert into current and insert into new options would be the answer
12:28:22pondlifeWhat about a book?
12:28:27amiconnNo, it's the movie that's on the current list
12:28:37pondlifeIt's still my current read, even if I stop reading igt.
12:28:40amiconnLlorean: Current list != current playing item
12:28:40LinusNi don't use the dynamic playlists much, but i would expect the "insert" operations to insert into the current list even when stopped
12:28:54amiconnLinusN: They do
12:28:55BigBambiLinusN: Exactly
12:29:17BigBambiamiconn: If you stop playback, then do insert, you don't get a new playlist?
12:29:20LloreanHow is the list NOT the currently playing item?
12:29:24LloreanYou click on it, and it launches.
12:29:25 Join MethoS- [0] (
12:29:36amiconnThese operations always operate on the current list (because there is only one list in ram at all times)
12:29:41 Quit Shaid (Read error: 110 (Connection timed out))
12:29:43Lloreaneach song is a sub-item, but the "list" defines what is played.
12:30:19Lloreanamiconn: Currently, "View Current Playlist" shows you a list, but "Insert" while stopped plays a single song.
12:30:21amiconnIt's just that these operations clear the list before inserting when stopped, and don't do that when playing (or paused)
12:30:51amiconnLlorean: Of course, because "Insert" changes the contents of the list
12:30:52LinusNamiconn: and that is not what i would expect
12:30:52pondlifeI think there should be obvious options "Insert into new playlist"
12:30:58GodEaternor me
12:31:01BigBambiNor me
12:31:09*Llorean already said that part.
12:31:11GodEaterpondlife: how about just one option "clear list"
12:31:20GodEaterand then the insert/queue options don't need to be changed
12:31:27amiconnIt's something you can grasp quickly. And it's a very convenient thing for me
12:31:29pondlifeI prefer the "Insert into new "
12:31:35BigBambiI would have too options - insert into current, and create new (or insert into new)
12:31:44pondlifeQuickerm, and maintains current functionalilty
12:32:08amiconnIt saves having an extra entry in the context menu, or needing extra steps to start a fresh list => convenience
12:32:16pondlifeI agree
12:32:17BigBambiBut confusion
12:32:22LloreanYou need a way to clear the list.
12:32:29LloreanOtherwise you'd need duplicated of several options
12:32:42Llorean"Insert" "Insert Shuffled" "Queue" and "Queue Shuffled" all need an "Into New" option
12:32:59pondlifeOnly "Insert" and "Queue" do...
12:33:05LloreanNo, the shuffled ones are needed
12:33:13pondlifeInsert Shuffled doesn't appear with a new playlist
12:33:16BigBambiRight, insert into new starts a new one with that song, and all others function the same with whatever is the current - either what you have just listened to, or the 'new' one
12:33:18pondlifeAh, for a folder
12:33:27LloreanYou can't insert a single song shuffled into an empty playlist
12:33:31LloreanIt's very difficult on the playback engine.
12:33:38*GodEater votes again for one extra entry on the playlist menu just to clear the list
12:33:49pondlifeIf I stop, I don't get the Queue options.
12:34:13GodEaterbecause the resume point is lost
12:34:15Lloreanpondlife: Then just two more, though I think Queue should be available from stop.
12:34:23*GodEater does too
12:34:29pondlifeExactly, it's currently inconsistent
12:34:36*GodEater find himself agreeing with Llorean for the first time in several minutes ;)
12:34:36pondlifeAll I want is consistency
12:34:36 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
12:34:51pondlife(and a working View Playlist)
12:35:02LloreanGodEater: The single, one, *only* point we've actually disagreed on is if something you're not using right now is "what I'm currently using" or "what i was previously using"
12:35:22GodEateryes but we disagreed a lot ;)
12:35:24pondlifeLlorean: What about my book analogy?
12:35:39pondlifeDoes that work for you - "my current read"
12:35:54pondlifeEven when the book is in the bedside drawer
12:36:07Lloreanpondlife: I'm in the middle of "On Basilisk Station." I was previously reading it (at dinner), but am not currently reading it, as it is not being read, and no reading is taking place (discounting the words on the screen right now, which are not relevant)
12:36:19LloreanI have been reading it.
12:36:24LloreanBut am not doing so in the present tense.
12:36:35pondlifeIt's not your previous read though.
12:36:37LloreanIf someone looked at me, they would see that I am not, in fact, reading the book.
12:36:49pondlifeIt's your current book, when you are reading.
12:37:07pondlifeIn general parlance.
12:37:08LloreanI intend to finish it, yes.
12:37:16GodEateryou'd confuse the hell out of me if you told me your previous read was a book that you then proceeded some hours from now to pick up again and continue reading.
12:37:21BigBambiIt is the book you are currently reading in an ongoing sense
12:37:24roolkuBagder: the download mirrors haven't picked up the m:robe bootloader - it seems to be the same permission problem as last time
12:37:33LloreanGodEater: I would say "I've been reading <book>"
12:37:52BigBambiWhich suggests you are no longer reading said book
12:37:57LloreanGodEater: I wouldn't say "I am reading <book>" unless you could look over at me and see an open book. At least, not intentionally say that.
12:38:02BigBambiI would
12:38:11pondlifeperfect vs imperfect
12:38:25pondlifeI was reading, I have been reading
12:38:29LloreanBigBambi: If it's not in my hands, open, then in the present tense, I am not actually reading the book.
12:38:32pondlifeI read
12:38:43pondlifeBut it remains your current book
12:38:45BigBambiLlorean: You are in an ongoing sense
12:38:53LloreanI have been in the past.
12:38:54pondlifeYou are not currently reading
12:38:58LloreanI am not currenlty
12:38:59BigBambiAnd you will again
12:39:01LloreanI intend to in the future.
12:39:14GodEaterLlorean: that's because you're odd. *Most* people would say that they were currently reading it.
12:39:22LloreanGodEater: Perhaps.
12:39:24XavierGrwow English class lessons :)
12:39:26BigBambiEverybody I know would say they are reading it
12:39:38LloreanAnd I agree, in common parlance, if you're not being very precise, "I am reading Book" works.
12:39:44amiconn'Queue' isn't available when stopped, because playlist manipulation when playback is stopped clears the playlist before doing the actual operation. And you can't queue something in an empty list, so it's consistent as well
12:39:56BigBambiAnd it is also what the vast majority of the population would say
12:39:57LloreanBut my whole point was that the playback system functions oddly, and we could introduce a distinction in terms to make it clearer, since it's not something likely to be cleared up soon.
12:40:04GodEaterLlorean: the problem with our users is that they are "most people" and are not keen on precision.
12:40:08LloreanUsing "common terms" to *hide* quirks of the system, seems silly to me
12:40:19BigBambiLlorean: In fact, you are the first person I have met who wouldn't say that
12:40:21LloreanIf you use the "common term" for it, they'll expect it to actually work like that.
12:40:22pondlifeWe should also get rid of the quirks
12:40:24LloreanIt doesn't, not yet.
12:40:38LloreanBigBambi: I know. I'm actually diagnosed with something.
12:40:44LloreanBut my point isn't "I expect people to talk my way."
12:40:57LloreanIt was "with precise wording we can make the quirks surprise people less"
12:41:03pondlifeMy point neither. My point is that Rockbox should be less quirky
12:41:12Lloreanpondlife: Yes, but why should we hide them until it is?
12:41:13BigBambiBut people won't get the nuances in the precise wording
12:41:31LloreanBigBambi: Thats why you use something as blatant as "Inactive" so people know they can't Insert into it.
12:41:46BigBambiI don't think people expect a DAP to require that level of subtelty of language
12:41:47pondlifeBecause it's better just to decide what it should do, then make it do that.
12:41:49GodEaterI wouldn't even do that
12:41:51 Quit Rick (Read error: 104 (Connection reset by peer))
12:41:52amiconnLlorean: Imo there is no quirk in the current playlist behaviour at all.
12:41:53*BigBambi apologises for that spelling
12:42:08GodEaterif it's inactive, the options to insert shouldn't be made available.
12:42:15Lloreanamiconn: The quirk is that people don't expect to Insert to clear, since it doesn't say it does.
12:42:31 Join Rick [0] (
12:42:35pondlifeIndeed. That should be reworded.
12:42:42LloreanBigBambi: I don't see "Current" (or "Active") vs "Inactive" as particularly subtle. :-P
12:43:05 Quit jhulst_ (No route to host)
12:43:32*BigBambi sticks his tongue out at Llorean
12:43:35Lloreanpondlife: How long do you think it'll be, though, before the rest of the core doesn't care about the difference between stop and pause? Seriously, it's not like the word choice can't be reverted later.
12:43:37amiconnThen the item name should change when stopped
12:43:51pondlifeLlorean: I'm not calling for that
12:44:20amiconn...and not more
12:44:24LloreanAs long as there's a difference between whether the contents of the playlist are in RAM or not, I'd say there's a distinction that should be made clear, somehow, to the user.
12:44:36LloreanDenoting the "current" playlist as Active or Inactive helps do this.
12:45:11BigBambiBut here to me, inactive just suggests that it isn't right at this isntant playing, not that it'll disappear if I try to insert something into it
12:45:18pondlifeThe playlist is always still valid after playback is stopped, no need to put extra words in to explain that.
12:45:26LloreanBigBambi: Yes, but the "Insert" options still need to be changed.
12:45:41LloreanBigBambi: If it's not, at this instant, playing, that does clue you that things that require the playback buffer can probably use it, right?
12:45:41BigBambiLlorean: Yep :)
12:45:45pondlifeThe insert option wording should reflect what it's going to do.
12:45:55 Quit MethoS (Read error: 110 (Connection timed out))
12:46:01pondlifei.e. When not playing, display it as "Insert into new playlist"
12:46:17BigBambiLlorean: Me yes, users no - most probably have no idea what the playback buffer is, or why it is clear or otherwise
12:46:24LinusNmarkun: "Button Code Definitions for the toshiba gigabeat target"
12:46:31Lloreanpondlife: Yes, but there's no reason to intentionally leave one word out that could make the actual state slightly more clear to the user
12:46:50pondlifeBut then users will also want "Insert into current playlist", rather than the less-than-obvious action of resuming playback then doing the insertion.
12:47:03Lloreanpondlife: Don't see why they can't have that.
12:47:09pondlifeMe neither
12:47:40LloreanBigBambi: Well, it'll still help some people. Do you think it could add confusion somewhere?
12:48:19BigBambiLlorean: No, it's a start :)
12:48:24LinusNmy opinion is that the playlist manipulation functions should work regardless of playback state, and that there should be a "clear playlist" function
12:48:26markunLinusN: ouch..
12:48:45markundidn't want to commit the meizu stuff
12:48:55LloreanLinusN: Works for me, but I still think the Active/Inactive thing is a valid "clue" for people in regard to knowing whether the RAM is claimed.
12:49:06BigBambiI agree with LinusN!
12:49:37LinusNi don't think the user cares about any claimed ram
12:50:03LloreanSome users don't know there's a difference between Stop and Pause.
12:50:13GodEaterand nor should they have to
12:50:14BigBambiLlorean: It can't hurt, but I think the number of people you are helping is very small. Most people couldn't give a monkeys about the state of the RAM, or even that the player has any
12:50:16pondlifeThey shouldn't really need to care
12:50:22LloreanBut they *do* need to right now...
12:50:30LinusNand i'm not sure there should be a difference
12:50:34LloreanWhy shouldn't we fix the UI hints until they don't need to care any more.
12:52:09LinusNwhat are the exact differences between Stop and Pause anyway? There is a bug that prevents voice playback in paused mode, but what else?
12:52:29pondlifeWhen paused the audio buffer can't be used by plugins
12:52:52LinusNthe difference is even more unclear on players without a stop button
12:53:10GodEaterand the "stop" button is a pain to use on the ipod imo
12:53:19LloreanI don't understand why we can have "Resume Playback" vs "Now Playing" but we can't distinguish in the same way between playlists...
12:53:20GodEaterI often end up shutting down by accident
12:53:21pondlifePause resumes instantly as PCM is kept filled
12:53:29PaulJamthere is also a difference between stop and pause when using crossfade and selecting a new song.
12:53:36pondlifeLlorean: Because there's no distinction
12:53:47Lloreanpondlife: There's the exact same distinction as between those.
12:53:48LinusNLlorean: are you sure that the "current" vs "precious" naming is a real issue?
12:53:50pondlifeStopped or playing, the playlist remains
12:54:02LloreanLinusN: I'm trying for "Active" vs "Inactive" now.
12:54:10pondlifeNo need for any distinction
12:54:22LinusNi think there shouldn't be a distiction
12:54:24LloreanLinusN: My point is to distinguish that playback is not going, at the moment. Much like the "Resume Playback" vs "Currently Playing" entry in the main menu.
12:54:35LloreanWe already distinguish.
12:54:41pondlifeYes, but that option will change playback state
12:54:49pondlifeViewing a playlist does not
12:55:09LinusNLlorean: we distinguish the menu action, not the playlist
12:55:12 Quit DerPapst (Connection timed out)
12:55:25LinusNthe playlist is the same
12:55:29LinusNthe playback state is not
12:55:34LloreanThe playlist isn't being used by playback.
12:55:37LloreanIt's not active.
12:55:49pondlifeNo, the playback is inactive, not the playlistr
12:55:53LinusNit's being used by the user
12:56:03Lloreanpondlife: by that argument, the song is active, but not the playback.
12:56:30*markun is making a complete mess today :(
12:56:44LinusNmarkun is having a bad svn day
12:56:59markunbad git day even
12:57:03pondlife"b7bd17ce1715c3a8acbf74a63694c28b6075da5b" ??
12:57:07Lloreanpondlife: Then why not permanently make the list entry "Currently Playing"?
12:57:23Lloreanpondlife: Since, by your standards, when Paused or even Stopped, it's Currently Playing
12:57:34pondlifeCurrent != currently
12:57:52pondlifeThe point is that selecting that option will start playback
12:58:00LloreanWell you said I'm currently reading the book.
12:58:07Lloreanpondlife: It won't if you're Paused.
12:58:09LinusNLlorean: the playback state is one thing, the playlist is another
12:58:11LloreanIt only resumes playback if you're Stopped
12:58:19pondlifeI said the book is your current read
12:58:33pondlifeAlthough you are not currently reading it
12:59:24LloreanWhat makes it current?
12:59:30pondlifeIf I'm paused, the menu option reads "Now Playing", it could also read "Currently Playing"
12:59:34LloreanThat it was the one I most recently used?
12:59:41LinusNLlorean: you are still using it
12:59:46pondlifeThe English language, mainly
12:59:47LloreanLinusN: Not according to pondlife.
13:00:08LinusNyou can for example insert into it, or view it, or resume playback
13:00:15pondlifeLlorean: Please stop putting words in my mouth
13:00:20Lloreanpondlife: You said I'm not reading it.
13:00:25LloreanWhat use am I making of it, if not reading?
13:00:43pondlifeI said.... you are not currently reading it.
13:00:50pondlifeIt is your current read, however
13:00:57markunLlorean: maybe holding it in your hands while you think about other things before continuing to read again
13:00:57pondlifeEnglish is not logical here
13:01:15LloreanAnd I asked "what makes it current", LinusN said "that you are using it" and I said "not according to Pondlife." Am I using it (the only use for it i have is reading) or not?
13:01:17LinusNenglish is not my native language, but i do understand pondlife's book analogy perfectly
13:01:35pondlifeYou are not currently using it
13:01:39LloreanThank you
13:01:42LloreanThen I put no words into your mouth.
13:02:06pondlifeYou are reading it however
13:02:15pondlifeBut not currently
13:02:15LinusNnow this discussion is getting silly
13:02:26***Saving seen data "./dancer.seen"
13:02:31pondlifeIt got silly some hours back
13:02:57Llorean"(7:01:34 AM) pondlife: But not currently" <−− I thought i was currently doing some sort of verbing with it...
13:03:22LloreanAnyway, ignore verbs or tenses.
13:03:30LloreanMy point is this: We distinguish between Stop and Pause in the main menu.
13:03:32LinusNas for the playlist, i think it's simple: the playlist is always the current one, even if you aren't playing it
13:03:33pondlifeThe point is the use of the term current in English.
13:03:37Llorean"Now Playing" while Pause, "Resume Playback" while stopped.
13:03:50pondlifeIndeed, we do. I'm not arguing for a change
13:04:03LloreanThis creates a distinct difference between "Active" and "Inactive" playback, and whether the playlist is _in use_, or merely _current_
13:04:03 Join qwedsa [0] (
13:04:22pondlifeThe action of selecting that option when stopped resumes playback. Otherwise it does not. Hence different wording.
13:04:26LloreanI'm not reading the book, but it is my "current read", to use your terms.
13:04:45LloreanI'm not listening to the playlist, right now, but it is my current playlist.
13:04:56LloreanWhy not distinguish between whether it's a Paused playlist or a Stopped playlist then?
13:05:03pondlifeWhy bother?
13:05:10LloreanWhy NOT bother?
13:05:21LloreanWhat harm is done by adding a small amount more clarity?
13:05:29LloreanWhat simplicity is lost?
13:05:31pondlifeMore words often reduce clarity
13:05:41LloreanWhat clarity is lost, then?
13:06:09pondlifeWhy does the user need to have terms like "Stopped playlist" that make no sense.
13:06:24pondlifeA playlist is a playlist, stopped or not
13:08:02LloreanPlayback is simply the iteration through the playlist. If playback is stopped, the playlist is not in use for playback.
13:08:13LloreanIf you aren't viewing it, then nothing is using the playlist. It simply exists, in ram.
13:08:23pondlifeIt remains the current playlist.
13:08:29LloreanStopped or not, you can distinguish between Active and Inactive.
13:08:33pondlifeNo need
13:08:40pondlifeI'm tired of this discussion
13:09:08LloreanYou can't just dismiss it as 'no need'
13:09:19markunLlorean: stop it already
13:09:20LloreanIt's clear that people don't realize there's a difference of state between "Stop" and "Pause"
13:09:21pondlifeI can. I don't see any need at all.
13:09:25LloreanYOU don't see.
13:09:29pondlifeI don't
13:09:33LloreanI do.
13:09:38pondlifeExplain why it's important then
13:09:41LloreanUsers do not see a distinction between "Stop" and "Pause"
13:09:57pondlifeSo, what's that got to do with the price of fish?
13:10:00LloreanAs long as that distinction exists, more clues in the parts of Rockbox that users see (not the manual) helps make it clear the system is in a different state.
13:10:12Llorean"Active" and "Inactive" provide such a clue.
13:10:17pondlifeThe status bar makes that distinction clear anyway
13:10:27GodEaterno, they provide more terms to confuse users with
13:10:40GodEaterI predict *many* questions from users if such a term is introduced.
13:10:45LloreanGodEater: Nobody's presented me with how it's confusing yet, and I've asked more than once.
13:11:10pondlifeI don't see a need or an explanation for one
13:11:14GodEaterLlorean: you're addressing a bunch of technical people who understand the point you're trying to make.
13:11:22GodEaterwe all still think it won't gel with the users
13:11:26GodEateryou're in a minority of one
13:11:30BigBambiPeople just don' t care about the technical minutiae, and nor should they have to.
13:11:44pondlifeI don't think the users should need to know about the internal state - it should just work.
13:11:52LloreanGodEater: Being a minority doesn't mean you get to say "shut up, you're wrong, and I'm not telling you why"
13:11:59GodEaterI haven't said shut up
13:12:09GodEaternor have I said you're wrong
13:12:11BigBambiBecause of common use of language, technically correct or not
13:12:23preglowcan someone give me a timestamp for the discussed issue?
13:12:39GodEateryou can choose to use the english language in any way you please - but we're telling you that most people don't use it the way you do.
13:12:51LloreanGodEater: It's no longer an issue of using the English language.
13:13:00GodEaterwhether you like that or not is neither here nor there
13:13:24pondlifepreglow: 10:40 GMT-ish
13:13:29GodEaterit IS an issue of language
13:13:31*BigBambi is off to a french lesson - hopefully this will have finished when I get back in 4 hours
13:13:42pondlifeI need to work
13:13:43LloreanGodEater: How so?
13:13:49GodEateryou want to use it to try to describe some inconsistent behaviour in Rockbox
13:13:55GodEaterwe'd rather fix the behaviour
13:14:06LloreanI want it to describe the behaviour until it's fixed.
13:14:15GodEaterI don't
13:14:18LloreanWhy not?
13:14:21GodEaterI'd leave it alone until it's fixed
13:14:32GodEaterbecause it doesn't cause enough grief to warrant the change
13:14:35LloreanIt's been three years, I don't think this is being fixed soon.
13:14:52 Quit Zarggg (Connection reset by peer)
13:15:01GodEaterthis is the first time I've seen it come up
13:15:03LloreanGodEater: Do you think this would honestly confuse users more?
13:15:05pondlifeAs long as we generally agree that a playlist remains valid after STOP, I can do the fix
13:15:16 Quit BigBambi ("CGI:IRC")
13:15:23pondlifeI was unsure of the intended policy until now
13:15:33Lloreanpondlife: Including all the RAM issues?
13:15:43pondlifeThere are no RAM issues
13:15:57pondlifeThe playlist remains valid.
13:15:58GodEaterLlorean: I bow to your greater length of service time in the forums and mailing lists here of course, but I personally haven't seen many issues with this crop up amongst the users.
13:16:20GodEaterif we were getting daily or even weekly "I don't get it" questions from users, I'd see the need to make the change
13:16:23GodEaterbut I don't see them
13:16:40GodEaterso to quote an old adage "if it aint' broke, don't fix it"
13:16:52pondlifeIt's more for the clarity among developers that I asked... wish I hadn't now...:)
13:16:56GodEaterand by that, I merely mean the terms, not the behaviour
13:17:01LloreanGodEater: What about the concept that, if it helps more than it hurts then it's beneficial?
13:17:28GodEaterLlorean: we have no metric to measure that with unless we make the change, and find out one way or the other
13:17:35pondlifeWe disgree, but I think adding "Active"/"Inactive" hurts more than helps
13:17:44Lloreanpondlife: But you refuse to tell me in what way.
13:17:49LloreanOr have thus far.
13:17:51GodEatermy opinion, whether it matters or not, is to leave it alone
13:17:53pondlifeI think it complicates matters
13:17:58GodEaterI do too
13:17:58LloreanComplicates how?
13:18:09GodEaterI don't understand the concept of an active or inactive playlist
13:18:16GodEateryou either have one, or you don't
13:18:22LloreanGodEater: Stopped vs Paused. "In use by the playback engine"
13:18:33pondlifeWhy is that useful?
13:18:38GodEaterLlorean: yes, *I* as an involved person, understand that.
13:18:44LloreanWhy don't we have a "Resume Playback" button when Paused then?
13:18:48GodEaterI predict that is the question you will get asked a LOT
13:18:52GodEaterif you choose to use those terms
13:18:54Lloreanif it's unuseful to distinguish, why do it anywhere else right now?
13:18:59pondlifeLlorean: I already answered that twice
13:19:05pondlifeRead the logs
13:19:06Lloreanpondlife: No, you didn't.
13:19:19Lloreanpondlife: You answered "why are they named different things." I just asked "why do they DO different things"
13:19:51LloreanWe could just as easily have made the button "Resume Playback" whether stopped or paused, rather than "Now Playing" for the latter.
13:20:12pondlife"Why don't we have a "Resume Playback" button when Paused then" - we could, in fact we probabaly should
13:20:16preglowis there some kind of a core to this discussion? i can't be bothered to wade through hours of talk...
13:20:26pondlifepreglow: Not really
13:20:53preglowseems to me you're all babbling on about semantics
13:21:02pondlifeIn summary - does a playlist continue to exist after you press STOP. Is it fair to call it the Current Playlist.
13:21:03Lloreanpreglow: I suggest that until such time as there is no difference in behaviour while Rockbox is paused vs stopped, we should distinguish the playlist in some way (active vs inactive) much like we do the WPS menu entry
13:21:13Lloreanpondlife: I never denied that it continued to exist.
13:21:21pondlifeThis paused/stopped thing is a red herring
13:21:28preglowinactive/active playlist???
13:21:35preglowwhy do we need that distinction?
13:21:41pondlifeWe don't
13:21:43markunpreglow: ask Llorean
13:21:57LloreanBecause users don't realize there's a difference between Stopped and Paused, so more changes in the UI between the two might highlight it...
13:22:00preglowthere's no point in revealing any kind of internal state to the user unless the user benefits
13:22:14preglowand in this case, i don't think there's even any difference in internal state
13:22:25Lloreanpreglow: Users ask why voice doesn't work while "stopped". Users don't get why some plugins won't launch until they restart their player. Etc.
13:22:32pondlifeLlorean: The user needs to know about stopped/paused, because of the differences re voice/plugins...
13:22:35preglowLlorean: they don't? i get it easily, pause pauses, stop stops, it feels like on a cd player
13:22:38preglowif i pause, it resumes faster
13:22:57pondlifeLlorean: They don't need to care about "inactive" playlists.
13:23:12pondlifeYou are comparing apples and elephantss
13:23:48*pondlife has had the longest lunch hour ever and remains hungry...
13:23:51pondlifeback in a bit
13:23:53preglowi agree it would be nice to make it easier for the user to understand the difference in state between paused and stopped, but i hardly think this is a good solutionm
13:23:59preglowit sounds confusing even to me
13:23:59Lloreanpondlife: If the playback engine is not using the playlist, then the playlist is not being used by the playlist engine. I don't see how you could imagine it is in use. It's just available.
13:24:18Lloreaner, playback engine.
13:24:51pondlifeI don't imagine anything. I don't see that stopping playback makes any difference to the playlist.
13:25:04preglowme neither
13:25:09pondlifeA playlist is a playlist is a playlist.
13:25:23markunLlorean: did you ever use a player like winamp or foobar2000?
13:25:23LloreanYes, but it can be in use or not.
13:25:29Lloreanmarkun: I've used both.
13:25:41preglowdepends how you see the playlist being used, it's just an abstraction
13:25:54pondlifeLlorean: Not really.. Playback can be in use or not, the playlist just sits there.
13:26:03Lloreanpondlife: It is used *by* the playlist.
13:26:08LloreanEr playback engine
13:26:11preglowi don't see the playback engine taking hold of the playlist and releasing it when done, i just see it picking entries from it when needed
13:26:13LloreanIt being the playlist.
13:26:13pondlifeAnd the UI
13:26:15preglowit's never active or inactive
13:26:33markunLlorean: they also have playlists which keep being active even if playback is stopped
13:26:33pondlifeYou can do a lot of playlist work without using the playback engine
13:26:48Lloreanmarkun: They keep being present when playback is stopped.
13:26:49pondlifeThe database uses it
13:27:02markunLlorean: yes, which I would like in rockbox too
13:27:10Lloreanmarkun: They already stay present like that.
13:27:44pondlifemarkun: That was my initial point, I was unsure if this was intended originally, or had just migrated to working that way
13:27:57Lloreanpondlife: No, it was intentional that they stick around.
13:28:05*Llorean would've answered that one, if asked.
13:28:30LloreanI saw when it was added. I just don't like that it was done without fixing the other issues. It made the UI less consistent, for the time being.
13:28:40pondlifeIndeed, my point
13:28:51pondlifeThanks, you did answer earlier, too
13:28:58preglowLlorean: but ok, how does this concept of active/inactive playlist make the user easier understand the paused/stopped issue?
13:29:05pondlifeIt doesn't
13:29:16*dis_afk wonders if this discussion in worth the binary size of the log
13:29:21preglowit isn't
13:29:27Lloreanpreglow: It depends on whether you see the playlist as "a list of files" with emphasis on list, or "a list of files" as in, all the files, in a row.
13:29:32GodEaterit clearly doesn't, because it's confusing the developers too
13:29:36 Nick dis_afk is now known as disorganizer (n=c2785409@gateway/web/cgi-irc/
13:29:56*Nico_P chose not to follow and is glad he did :)
13:30:08pondlifeMove along, move along... :)
13:30:09preglowLlorean: if we need that distinction to make sense of this concept, then it's flawed from a user's perspective
13:30:26preglowin being too complex
13:30:39Lloreanpreglow: My point is that I think for some users, it will make the state more clear. For other users, I don't think they'll case. But I don't think it'll cause any harm.
13:30:41*GodEater thinks we've done most of our only students work so far with this discussion ;)
13:30:46*disorganizer as user would prefer STOP to actually empty the playlist and clear the entires in "view playlist" to be honest :-)
13:30:49Llorean*don't think they'll care
13:30:57 Quit MethoS- (Read error: 113 (No route to host))
13:31:01Lloreandisorganizer: It used to.
13:31:06preglowLlorean: i think plugins could just stop the playback engine if it happens to be paused and they want the playback buffer or any other resource used by playback
13:31:11Lloreandisorganizer: Right now it's at kinda a halfway point, and needs to make up its mind. :-P
13:31:12preglowthat'd fix that particular problem
13:31:18Lloreanpreglow: I agree.
13:31:40Lloreanpreglow: I've only suggested this as a temporary fix, since the whole engine has worked this way for so long, with people suggesting fixing it for a while.
13:31:44disorganizerllorean: it should again. this is the only logical behaviour users without technical knowledge can reproduce without looking in the manual ;-)
13:31:57preglowif it's playing, the plug should spout a warning
13:32:21Lloreanpreglow: It should sprout a prompt asking the user to stop playback.
13:32:32preglowwould also work
13:32:45preglowwould make for a nice pluginlib addition
13:32:49disorganizerwhy not: if you press stop it pauses, and if you press stop a second time within 2 seconds it clears the playlist and goes to "stop" state
13:33:10pondlifedisorganizer: No way should stop clear the playlist. I do lots of playlist preparation on target and don't want to have to have playback started.
13:33:21preglowno such hidden actions on stop, please
13:33:23preglowand that's that
13:33:34pondlifeIt may be that the iPod keymap makes it seem different to H300..
13:33:37preglowbuttons should do what they say, and mostly only that
13:33:43amiconnHours of discussion about minor UI string ambiguity. Just rename 'Insert' to 'Insert into new playlist' when stopped, and all is well
13:33:45preglowespecially buttons marked as clearly as "stop"
13:33:49pondlifeamiconn: Right
13:33:53disorganizerwell, me as a user would expect stop to empty the playlist. then i can prepare a new one and hit play to start playback. that would be logical behaviour
13:33:56*amiconn thinks there are way more important things to fix
13:34:08pondlifeLike my lunch.
13:34:17GodEateramen to that
13:34:29amiconnLike playback engine bugs in general, or e.g. non-working voice when paused
13:34:53markunamiconn: I would like any bugs to be fixed, also unimportant ones
13:35:21markunor at least less important that others
13:35:29amiconnmarkun: The #elif might not be that good an idea. It works for now, but wil need changing in case the Elio port will ever be picked up again
13:35:30preglowwe should rather discuss what to rename those playlist actions, so they actually describe what they do properly
13:35:39Nico_Pamiconn: I'm precisely thinking of improvements to the playback engine :)
13:35:48amiconn(the Elio is hdd based *and* has an sd slot)
13:35:54Lloreanpreglow: They should keep the same names, never clear on their own, and the bottom option should be "Clear Playlist"
13:36:04LloreanWell, "Play Next" can keep its clear. :-P
13:36:06markunamiconn: I'm done with committing for today..
13:36:09*amiconn disagrees with Llorean
13:36:39*GodEater agrees with Llorean
13:36:44amiconnNo separate 'clear playlist' option, please
13:37:00Lloreanamiconn: If we're going to show people a "Current Playlist" in the viewer, you HAVE to be able to insert into said 'current' playlist, or all reasonable argument that it's actually current goes out the window. :-P
13:37:03amiconnIf you then forget to use it, your playlist will grow and grow and...
13:37:38amiconnLlorean: It *is* current, even if it's cleared when inserting from stopped state
13:38:01LloreanNot from a common sense perspective, though
13:38:03disorganizerso why not clear it on stop anyways?
13:38:11LloreanIt's current from a "it's in RAM until it's cleared" perspective.
13:38:16amiconndisorganizer: Then you could *never* resume
13:38:29markunamiconn: so you need to start playback if you don't want to lose your playlist when you want to insert some songs?
13:38:31LloreanBut common sense dictates that if it's currently your playlist, and you choose an action to be performed upon your current playlist, it merely does what that action explicitly says it does: Insert
13:38:58disorganizeryou can resume on pause. try pressing stop on a cd-player
13:39:16disorganizerhit pause => you can resume, hit stop=> it starts from the beginning
13:39:24Bagderin other news: the gigabeat build is now named Gigabeat F/X on the site
13:39:27GodEaterdisorganizer: yes, but the playlist is still there
13:39:36GodEaterthe cd is still in the drive
13:39:39Lloreandisorganizer: There's reasons why you *need* to stop, but still want Resuming (such as shutting down, for example)
13:39:47GodEateryou want Stop to be similar to eject
13:39:53GodEaterwhich we have no equivalent of
13:39:56amiconnLlorean: No, it's always in ram. You change its contents, not the fact that it's the current list
13:40:07LloreanGodEater: Well, if we're mimicing CD state, we should restart at entry 1 when stopped. ;)\
13:40:09markunBagder: great
13:40:15amiconnAnd the inconsistency can be resolved by changing the item name in the context menu when stopped
13:40:32*Bagder also reminds people that the www dir is in fact in use now...
13:40:34amiconnThat's all that should be done, imo
13:40:37disorganizerin fact if i press stop imho the playlist will be my new playlist (just that it holds the content of my old playlist which i previously played)
13:40:47Lloreanamiconn: Not from user perspective. If you load a .m3u, then for the user, that *file* is the current playlist. If you then stop, then insert, from a user perspective that's a new playlist, not the same current playlist, re-populated.
13:41:16*amiconn thinks that is the root cause of the misunderstanding
13:41:26amiconnThe current playlist is what's in ram, at all times
13:42:05LloreanBut the idea is to make things more common parlance relevant, rather than technically relevant, apparently.
13:42:41LloreanIt should make sense from a user perspective.
13:42:58LloreanThey shouldn't need to care what's going on behind the scenes. Which is perhaps the one thing all of us arguing before agreed on.
13:43:46*petur thinks this is a nice question to ask our usability study student to see how well he'll cope with the task :)
13:44:05amiconnI'm a user too, and the current behaviour makes very much sense to me. A cumbersome separate clear option woould piss me off
13:44:28amiconnEspecially since 99% of my playlists are dynamic lists started by playing a file in a dir
13:44:38Lloreanamiconn: What if we put the "Play Next" option at the top when stopped, so the current behaviour is preserved in terms of button presses?
13:44:39 Quit CyBergRind|w ("!")
13:45:00*Llorean thinks: or better yet, add a "Play Now" option.
13:46:23*amiconn almost always either plays an album by clicking at the first file in a folder, or a various songs collection by using 'Insert shuffled' on a folder
13:50:26*petur too
13:57:55roolkuBagder: did you see me comment re: m:robe bootloader?
13:58:19roolkuBagder: the download mirrors haven't picked up the m:robe bootloader - it seems to be the same permission problem as last time
13:59:02Bagderah indeed. I do suck!
13:59:04roolkupoor bluebrother picked up an old bootloader and was confused...
13:59:04Bagderfixed now
14:02:09 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
14:02:52 Join [CBR]Unspoken|w [0] (n=cbr@
14:02:57 Quit kushal_12_27_200 ("This computer has gone to sleep")
14:05:29 Join fehmicans [0] (n=canavar@
14:07:05*gevaerts notices that our gsoc applicant has an updated application
14:07:14 Join erik006 [0] (
14:13:31GodEatergevaerts: imo that application still needs a lot of work
14:13:44 Join erik006-2 [0] (
14:14:47gevaertsGodEater: I agree. I also think it shows a lack of knowledge about rockbox
14:15:36 Join leftright [0] (n=d9e1f911@gateway/web/cgi-irc/
14:19:07 Join daven [0] (
14:19:45GodEaterI'm not really sure where to start critiquing his modified application without either pulling it apart completely, or going the other way and spoon feeding him with questions we'd want answers to
14:20:03leftrightas a simple user, only now do I understand the difference between stop and paused with regards to playlist actions due to the past discussion, I must admit that I stumbled upon Insert when stopped creates a new playlist, and that insert when paused will add to playlist, but creating a new playlist on the fly is very vague
14:20:23davenHi there, I'm currently re-installing rockbox on an iPod Mini G2 that I've accidentally wiped the disk of, is there anywhere I can get the original disk image from?
14:20:40davenI'm running Linux, so don't have access to iTunes to recreate the installation
14:21:10gevaertsdaven: have you seen ?
14:21:15pondlifeGodEater: Obviously, I've not seen the application, but I'd avoid any spoon-feeding.
14:21:24davengevaerts: Nope, thanks for pointing me at it.
14:21:37 Quit GodEater (Remote closed the connection)
14:21:58 Join GodEater [0] (i=c2cbc962@rockbox/staff/GodEater)
14:22:50GodEaterpondlife: PM
14:24:17leftrightPlay Next is not very descriptive of a function which will replace the current playlist, "Replace" is more descriptive
14:24:39leftrightor New playlist
14:24:47 Join erik006-2_ [0] (
14:24:52Lloreanleftright: "Play Next" already exists. It plays next, quite literally. It waits until the current song ends.
14:25:16leftrightyes, but its not very descriptive of what it actually does
14:26:08leftrightdoes it clear the playist as well as play selected next
14:26:29LloreanThat's why it's not "Insert"
14:26:51leftrightso it doesnt clear the playlist
14:28:31davenhm, this is interesting, rockbox seems to start, it gets past the checksum stage, reports "Rockbox Loaded" and then hangs...
14:28:33LloreanIt does clear the playlist.
14:28:53LloreanIf it were "Play next without clearing" it would be "insert" or "queue".
14:28:55 Quit erik006 (Read error: 110 (Connection timed out))
14:28:56leftrightthats my point, Play Next is not decriptive enough
14:29:07LloreanProcess of elimination helps a good deal there.
14:29:23leftrightthats user friendly ?
14:29:53LloreanIt's no more confusing than a mere "Replace" is.
14:30:08leftrightReplace playlist would work too
14:30:18LloreanNow, or when the song ends?
14:30:19 Join jgarvey [0] (
14:30:40leftrightbetter than Play Next
14:30:55 Join MethoS [0] (
14:31:04disorganizerplay next as a new playlist :-)
14:31:06PaulJam"replace playlist" sounds to me as if the new playlist would start playing immediately.
14:31:14LloreanPaulJam: That was more or less my point. :)
14:31:33Lloreanleftright: "Play Next" leaves a lot less room for doubt, when the rest of the options are included, than "Replace" or "Replace Playlist"
14:31:38leftrightplay next means play selected next, how is the user supposed to deduce that it means clear current playlist and then play selected next
14:31:52*disorganizer wonders if we really need so many methods of adding songs to a playlist. and how such a basic thing can trigger such a long dispute
14:31:55 Quit erik006-2 (Connection timed out)
14:31:58LloreanBecause the option wouldn't exist if its behaviour were exactly identical to "Insert Next"
14:32:02 Nick erik006-2_ is now known as erik006 (
14:32:38leftrightas a user I'm tellojg you what I see, I see play next, that means play selected next,
14:32:51LloreanAnd that's what it does.
14:33:04Llorean"Play", as in "It becomes what is playing"
14:33:11leftrightnever mind
14:33:17 Quit leftright ("CGI:IRC")
14:33:54*dionoea wonders why people would need the bible on their DAP (see forums).
14:34:05GodEaterdionoea: I wondered the same ;)
14:34:18GodEaterbut then I'm an atheist, so I don't see why anyone would need it at *all* ;)
14:34:41LloreanThere's one verse in it, I frequently wish I could remember.
14:34:46JdGordonGodEater: so you have proof there is no god?
14:34:52LloreanBut I mean, I suppose it's as useful (or not) as wikipedia or dictionaries.
14:35:27GodEaterbeing an aetheist doesn't require proof
14:35:40LloreanOff-topic, guys.
14:36:37 Quit ColinT (Remote closed the connection)
14:45:12 Join ColinT [0] (
14:47:53 Join kugel [0] (n=kugel@unaffiliated/kugel)
14:48:05kugellinuxstb: Thanks for your answers
14:50:07kugellinuxstb: But I didn't ignore remoetes. I've added comments in settings.h and settings_list.c regarding that. It's a TODO
14:50:28preglowi've got a wavpack case here where sample rate is reported as zero...
14:51:14 Quit Nico_P (Remote closed the connection)
14:51:44 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
14:56:04*kugel 's getting this again...
14:57:39 Join BobShield [0] (
14:58:03 Quit fehmicans (Remote closed the connection)
15:02:30***Saving seen data "./dancer.seen"
15:02:40 Join BobShiel1 [0] (
15:02:40 Quit BobShield (Read error: 104 (Connection reset by peer))
15:03:21amiconnmarkun: What was that strange 'xargs' required for? I think it was the cause of all the red...
15:06:01 Quit BobShiel1 (Remote closed the connection)
15:10:24 Join MethoS- [0] (
15:12:19 Join erik006-2 [0] (
15:13:42 Join jcollie [0] (
15:18:54 Join erik006-2_ [0] (
15:20:24 Quit erik006 (Read error: 104 (Connection reset by peer))
15:20:49 Nick erik006-2_ is now known as erik006 (
15:24:48 Join TMM [0] (
15:26:54 Join mikus [0] (n=5316f86b@gateway/web/cgi-irc/
15:28:32 Quit MethoS (Read error: 113 (No route to host))
15:32:24 Join qwedsa_ [0] (
15:33:05quitte_what does IDA mean?
15:33:16 Quit m0f0x (Remote closed the connection)
15:34:00 Quit erik006-2 (Read error: 110 (Connection timed out))
15:35:57JdGordonInter Dimensional Animal?
15:36:58LinusNInteractive DisAssembler, i believe
15:37:30LinusNif you mean IDA Pro
15:38:32quitte_something about a ida database in the forum. probably the assembler,yes
15:40:06markunwow, I come back and you guys have already stopped the inactive-playlist discussions!
15:41:46disorganizermarkun: no, i think we just paused it. its still the active discussion though :-)
15:42:01pondlifemarkun: We can resume if you really want
15:42:14markunI knew it :)
15:42:42gevaertsBut there have been other discussions in the mean time, so it can't be the current one
15:43:01disorganizerwell the other discussion was just a new discussion :-)
15:43:24markungevaerts: well, maybe it's not currently the discussion but still the current one!
15:43:38amiconnmarkun: I know that reversal, but what problem did it actually try to solve?
15:44:05*gevaerts thinks of applying a current to participants in that discussion
15:44:15markunamiconn: I had a missing crt0.S in the meizu built which caused make to not create sysfont.h for some reason
15:44:38pondlifegevaerts: Can I have a currant instead?
15:44:48 Join toffe82 [0] (
15:44:58gevaertspondlife: no :)
15:45:32 Quit Mathiasdm ("Yuuw!")
15:47:28amiconnmarkun: I am wondering about the xargs thing, not about the meizu port in general
15:47:43markunamiconn: yes, I'm also talking about htat
15:48:19 Quit qwedsa (Read error: 110 (Connection timed out))
15:48:20 Quit mikus ("CGI:IRC (EOF)")
15:49:21 Quit Nevtus ("Gone")
15:49:49markunamiconn: but better ask Bagder about it
15:56:17ThundercloudBTW, question, is the FLAC+CUE support in amarok 2 going to be improved when compared to 1.4?
15:57:34gevaertsThundercloud: maybe you should ask that in an amarok channel
15:58:36Thundercloudgevaerts: Hahahaha whoops
15:58:37ThundercloudWrong channel :)
15:59:37 Join av7 [0] (n=4ea75b78@gateway/web/cgi-irc/
16:00:07 Quit av7 (Client Quit)
16:02:02 Join argumentD [0] (
16:04:19 Join Mathiasdm [0] (
16:04:59*scorche|sh sees he missed a large argument about semantics :(
16:09:24 Join R1ckdu [0] (n=5085f6f9@gateway/web/cgi-irc/
16:09:28disorganizerits still the current discussion though
16:09:30*scorche|sh thinks that if one is applying to a project, one should at least read the manual to see that we *do* have a delete function...
16:10:30 Join Rincewind [0] (
16:10:34*gevaerts leaves replying to scorche|sh this time
16:10:52 Quit corevette (Read error: 110 (Connection timed out))
16:10:55scorche|shi think i am going to let it be
16:12:00gevaertsYes. I don't think this application has much chance anyway, however much the text is cleaned up
16:12:41*gevaerts looks around to see if there are any other students who are just about to submit their application
16:13:08 Join Massa_ [0] (
16:13:28 Nick Massa_ is now known as Massa (
16:13:53scorche|shthere are a few who have expressed interest, but...
16:14:20MassaHello everybody!
16:14:27 Quit R1ckdu ("CGI:IRC (Ping timeout)")
16:14:34markunhi Massa
16:14:38 Join R1ckdu [0] (n=5085f6f9@gateway/web/cgi-irc/
16:15:21MassaI'm just a bit playing around with viewport usage in WPS - but I've some problems (maybe I don't understand correctly)
16:15:45markunMassa: are you the same Massa from years ago?
16:16:03MassaYes, I think so - I'm still alive :)
16:16:42MassaBut nowadays I don't have much time for coding, so I just play around as "user" ;)
16:16:51R1ckduhey everybody, i have a toshiba gigabeat and im interested in running a version of rockbox on it, obviously, there seems to be a solution to this on the rockbox wiki, however i am not pocessing the knowledge to implement rockbox, is there anyone willing to help me?
16:16:51Rincewindgevaerts: you have some advice for me, maybe?
16:17:11GodEaterR1ckdu: what sort of gigabeat do you have ?
16:17:39MassaBack to viewports and WPS: do I understand correctly, that I always have to set a foreground and background colour for a viewport?
16:17:54 Join RoC_MasterMind [0] (
16:18:29MassaIsn't it possible to use a "transparent" colour?
16:18:33R1ckdugigabeat s, otherwise the solution should have been easier as i understand?
16:18:35gevaertsRincewind: not really. Just trying to remind you to submit it in time ;)
16:18:44 Quit elinenbe (" HydraIRC -> <- Leading Edge IRC")
16:18:53disorganizermassa: yes, but the bg-color gets ignored when using a backdrop, afaik. and no, not that i know of.
16:19:08LambdaCalculus37R1ckdu: There's code in SVN for the Gigabeat S; just don't expect a fully working version soon.
16:19:29LambdaCalculus37You're free to download the source code and have a little fun with it.
16:19:34RincewindI hope to be ready to submit friday evening. This would leave the weekend to discuss things and clean it up
16:19:36GodEaterR1ckdu: there's little point putting it on the S as yet
16:19:36markunR1ckdu: I think the progress on the Gigabeat S port is a bit halted because nobody is working on enabling USB on it as I understand.
16:19:57MassaI thought, when I use "FF00FF" as background (or foreground colour) that it would be "transparent"... (as in BMPs)
16:20:09*scorche|sh sees gevaerts hide behind the couch
16:20:29disorganizermassa: not that i know off. if yes thats undocumented
16:21:57Massadisorganizer: no, it does not work - but that's what I thought (would be logical for me)
16:22:13R1ckduOk. Are there predictions on how long one could expect it to take till there are advancements, as to have full compability?
16:22:22LambdaCalculus37R1ckdu: Not really.
16:22:32LambdaCalculus37It depends on when people want to come and finish it off.
16:22:38scorche|shR1ckdu: we gont give predictions, because then murphy's law would force us to break them
16:22:40Massadisorganizer: so it's not possible to use different background colours in different viewports when using a backdrop?
16:23:09LambdaCalculus37R1ckdu: If you really want to run Rockbox on a Gigabeat, I suggest hitting eBay and getting a Gigabeat F, which is supported.
16:23:20markunor wait
16:23:29LambdaCalculus37Or help work on it.
16:23:35gevaertsRincewind: I would suggest to try and submit your application soon enough, so that you have some time to adapt it to comments. Of course, since I'm pretty new to the project (wasn't around last year), disregarding my opinion is entirely allowed and, if other developers have different opinions, even advisable
16:23:54disorganizermassa: need to check, because now as you say it i think the behaviour changed during commit, but as far as i remember this was true. checking in a few minutes if you want
16:24:55GodEatergevaerts: I think submitting earlier is a good thing too
16:25:05Rincewindgevaerts: I try to submit it as soon as possible, but I have exams at the moment, so I can't use every waking minute to work on it. I discussed things with domonoky a little bit (since he is german, too)
16:25:35gevaertsRincewind: sure. Just don't submit three minutes before the deadline ;)
16:25:40R1ckduWell thank you all for the information, it was helpful. (lambdacalculus37: i dont think i could go with that comfortably... my budget is too limited & i do not have any experience with code other than a 2 week basic java introduction, otherwise i'd be more than glad to help)
16:26:03Massadisorganizer: would be nice - where do you check? (I'm no longer familiar with the source code structure nowadays ;) )
16:26:25*gevaerts would have mixed feelings about candidates who put their gsoc application before their exams
16:26:31GodEaterRincewind: what sort of shape is it in at the moment ?
16:26:35LambdaCalculus37R1ckdu: You can get a Gigabeat F for less than $100, but if your budget is limited then I can't twist your arm about it.
16:26:56RincewindGodEater: I have finished the abstract and worked most of the rest out in my head
16:27:52 Join DaCapn [0] (
16:28:34Rincewindgevaerts: would be bad if the student isn't a student anymore when gsoc is happening, because of failed exams, right?
16:28:49GodEaterRincewind: very much so!
16:29:31R1ckduLambdaCalculus37: If that were true, how would I then treat the dislodged poor S Series?
16:29:48MassaSomething else: Which device can you buy currently which is similar to the Sansa E200 on which Rockbox works? (I know this question has been asked a lot times, but I assume the answers are always changing)
16:30:06GodEaterMassa: you can't buy anything new
16:30:12GodEateryou have to get 'em from ebay
16:30:44disorganizermassa: as soon as you use a bmp as backdrop of your wps the bg-color is ignored and the bmp will instead be: transparent :-)
16:30:45LambdaCalculus37R1ckdu: Like you always did. It'll just have a companion. :)
16:30:51LambdaCalculus37But I digress.
16:30:56MassaGodEater: that's bad :( - I don't like eBay anymore...
16:31:28disorganizermassa: and i checked with the ui emulator. but keep in mind viewports must not overlap.
16:31:49Massadisorganizer: you meant "the bg-color will instead be: transparent" (instead of "bmp...")?
16:32:27Massadisorganizer: how to overlap viewports would have been my next question - so what happens if you overlap them?
16:32:29disorganizermassa: well if you use the %X tag to display a wps backdrop your viewports wont use bg-coloring for the text which is displayed. so the bmp color is shown
16:32:58GodEaterBloody hell - that was quick ;)
16:33:10disorganizermassa: if you overlap them something very bad happens :-) the result is unpredictable and its not supported
16:33:11gevaertsRincewind: exactly :)
16:34:41 Quit R1ckdu ("CGI:IRC (Ping timeout)")
16:34:52 Join R1ckdu [0] (n=5085f6f9@gateway/web/cgi-irc/
16:35:15MassaGodEater: if the link was for me - I'm not a student (it has been more than 15 years ago that I was studying :D)
16:35:30GodEaterMassa: it was for the channel in general
16:35:54GodEaterI only submitted it to boingboing about an hour ago - I was impressed with how fast they posted it
16:36:02 Join csc` [0] (n=csc@archlinux/user/csc)
16:36:41Massadisorganizer: so how do you write text in one viewport which conditionally overlaps another one (e.g. when no album art is available)?
16:36:42R1ckduLambdaCalculus37: Well I guess i wont leave here without having tried to get something working on that big old jukebox. Is there a way you could talk/walk me through the patch process? Or would that be too much distraction?
16:37:26LambdaCalculus37R1ckdu: Clarify "big old jukebox". Are you talking about an Archos MP3 player?
16:37:39markunGodEater: nice replies to that post
16:37:48GodEatermarkun: I thought so too ;)
16:38:27LambdaCalculus37GodEater: Especially the reply by zikzak. He sounds like a fan, all right. :)
16:38:39 Quit herbi ("poop")
16:38:43markuna big contrast to the statement that rockbox is the worst open source project ever
16:39:07GodEaterand sounds like a bag of shit
16:39:08MassaGodEater: ahh, I thought because you're talking about iPods and Sansas in your post it would be a request for me to e.g. port rockbox to a newer device :)
16:39:26GodEaterMassa: well if you wouldn't mind... ;)
16:39:28disorganizermassa: at the moment thats a problem. hopefully some day we get conditional viewports. also remember that the x/y positions of bmp's are relative to their viewport, not the screen ;-) i dont know for sure whether you can use a conditional inside the viewport for text or AA though. never tested it
16:39:53Massamarkun: where did you read that statement?
16:40:02R1ckduLambdaCalculus37: No. I'm sorry about that I'm actually talking about the gigabeat S , with reference to the wiki post: .
16:40:27GodEaterthe Gigabeat S is not a "big old jukebox"
16:40:41GodEaterit's bloody tiny in comparison to some!
16:40:52GodEaterand it's super light too
16:40:57LambdaCalculus37R1ckdu: When you say "big old jukebox", that makes us think of the old Archos devices.
16:41:08LambdaCalculus37R1ckdu: But if you're really serious:
16:41:10GodEaterR1ckdu: you're aware there's no sound on the S at the moment aren't you ?
16:41:25Massadisorganizer: hmm, I think I continue gambling around with viewports... (e.g. what happens if I put my albumart (with conditional) at the "initial" viewport and the rest in defined ones?)
16:42:46*LambdaCalculus37 has seen a Gigabeat S next to his iPod video... it *is* a little smaller
16:42:56R1ckduGodEater: Well by today's standarts, where many youngsters get their players every year and complain about their size of an ipod nano, one of that category would certainly call it big and old, but if you ask me it still is a great device and has not let me down too often(apart from formatting randomly sometimes, which everyone using the basic firmware knows about by now)
16:43:00Massamatun: of course this had been posted by "anonymous" - I'm sure it's somebody which never really tried to use rockbox...
16:43:13disorganizermassa: discouraged. you should not use the "initial" viewport. it may be that someday this will be skipped . use viewports for everything.
16:43:33disorganizerneed to go home now . see you all later
16:43:34markunMassa: of course he didn't, because rockbox is such a POS that he couldn't even install it :)
16:43:52GodEaterLambdaCalculus37: and it's definitely lighter
16:44:06markunMassa: no, I guess I misunderstood "basic setup"
16:44:06LambdaCalculus37GodEater: Yes, definitely.
16:44:51MassaGodEater: it seems that the cowon D2 would be a nice device - but it's too expensive and currently only partially supported by rockbox
16:45:23disorganizermassa: a last comment: viewportification is easier than it looks. and it will be expanded to be even more usefull some day :-)
16:45:36 Quit disorganizer ("CGI:IRC")
16:46:02R1ckduGodEater: No, I haven't read any of that yet. But that makes the situation a complete different one. To bad I have no ability to work on that. In such, I guess there really is no use in porting rockbox onto my device yet. But I will be checking your site to see if there is any developement. Where does the information about the sound come from by the way?
16:46:15Massamarkun: maybe "basic setup" is the installation for him/her? Would explain a lot ;)
16:46:38GodEaterR1ckdu: I'm not sure if it's on the wiki or not, let me see
16:47:18*gevaerts might try to start a worse open source project if he finds the time
16:47:28GodEaterThe Audio progress bit
16:47:38GodEateri.e., it doesn't do anything functional yet ;)
16:48:06LambdaCalculus37GodEater: We need the I2C driver still, I see.
16:48:27GodEaterLambdaCalculus37: we need quite a bit :)
16:49:05 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
16:49:13 Join midgey [0] (
16:50:30R1ckduGodEater: So you mean keeping this as a bookmark would give me a good overview on how things work out in the future?
16:51:15LambdaCalculus37R1ckdu: You can just bookmark at the front page:
16:52:10pondlifeI like "...recently bought an iPod Classic (the new one) and have to say that it might be a nice piece of eye candy, but compared to a device running Rockbox its simply retarded."
16:54:41R1ckduLambdaCalculus37: Well that would be a solution too ;-) . Anyways, it was nice to talk to you, GodEater / LambdaCalculus37, wish you a good day and good luck with your codes. Have a good one, bye!
16:56:00LambdaCalculus37pondlife: Golden Quote?
16:56:10pondlifeSilver Quote
16:56:12GodEaterR1ckdu: bye :)
16:57:13Massapondlife: Yes, it's a nice comment - but it would be much better if it could sound like "... recently bought an iPod Classic and directly had to put rockbox on it because of the weird original firmware" ;)
16:57:28pondlifeMassa: You know what to do....
16:58:17pondlifeHmm, my database is acting very odd
16:59:30Massapondlife: Was just a joke - I don't own any iPod, I still use my iRiver H340... (and my son's Sansa) - as I read, the newer iPod firmware is so good encrypted, that it's nearly impossible to inject rockbox boot code!?
16:59:49pondlifeMassa: Mine was a joke too, I ran out of smileys... ;-)
16:59:57 Quit R1ckdu ("CGI:IRC (Ping timeout)")
16:59:58pondlife(found one in the sofa there)
17:00:42pondlifeIf I go into Database > Track and select the 5th one down ("Heroes"), it does the searching countdown from 16xxx to 0, but then plays another track. The generated playlist has blank entries for 1-3, then is a directory-sequenced, not alphabetical.
17:01:01pondlifeTime for a reboot
17:01:26pondlifeAha, dircache is rebuilding...
17:01:49pondlifeMaybe this is why my playlist viewer has been doing buggy things all day.
17:02:25 Part LinusN
17:02:30pondlifeYep, works now I've rebooted
17:02:34***Saving seen data "./dancer.seen"
17:03:01kugellinuxstb, linuxstb: ping
17:03:05kugellinuxstb, linuxstb_: ping
17:03:15pondlife3 linuxstbs?
17:03:34kugelNo ;)
17:05:43markuncan I rename audiohw_set_headphone_vol to audiohw_set_master_vol in tlv320.c?
17:05:49*linuxstb_ appears
17:05:59GodEaterin a piff of logic ?
17:06:02GodEateror a puff
17:06:09*GodEater goes to buy a dictionary
17:06:38*linuxstb_ has just come back from a very long lunch, so hopes kugel's questions are easy
17:07:16pondlifeGodEater: PIFF ={E814163C-01B5-4DBF-B0AB-750DCA1D1BA8}
17:07:18*GodEater likes boingboing's headline for the article ;)
17:07:40kugellinuxstb_: In theory, I like your idea of having a unified viewport_parse function, which both list and wps (and more later) would use.
17:07:45GodEaterpondlife: sounds dreadful
17:08:06 Quit TMM (Client Quit)
17:08:55linuxstb_kugel: But only in theory?
17:08:57kugellinuxstb_: But, I'm not sure if it fits my needs for the list. In the list, I we have more than one viewport. I can't simply use the |..|..|..|..| from the settings, and use it for a viewport in the list.
17:09:47kugels/I //
17:11:50linuxstb_I thought the list was all contained within one viewport - the parent?
17:12:14kugellinuxstb_: That's true, but in the parent, there are more viewports
17:12:46kugeland editting the parent leads to problems (i.e. not updating parts of the screen which show parts of the wps)
17:13:06 Nick fxb__ is now known as fxb (
17:14:25Massakugel, linuxstb_: I don't want to disturb your discussion - but I hope you could easily answer a related question ;)
17:14:28 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
17:14:29*scorche|sh looks at boingboing, and then looks at "Bryan"
17:15:00kugelMassa: ask it, quickly :)
17:15:19Massacurrently the background colour definition of a viewport will be ignored (=transparent) when using a backdrop.
17:15:28linuxstb_Massa: Yes.
17:15:31 Join EskimoSpy [0] (n=cd9cbcfe@gateway/web/cgi-irc/
17:15:36kugelI though it gets automatically ignorend
17:15:48kugelah, missread :/
17:16:17GodEaterscorche|sh: read back in community
17:16:24MassaIs this a general limitation to have different background colours or is this just the current implementation?
17:17:10 Nick fxb is now known as birthday`fxb (
17:17:12MassaWouldn't it be nicer to allow different background colours for different viewports and to define the FF00FF colour as transparent (as it's done in BMPs)?
17:18:24linuxstb_If you don't have backdrops, then the background colour works, if you do have a backdrop, make the backdrop contain the colour...
17:19:15Massalinuxstb_: that's how it is currently - but with my suggestion it would be able to have both...
17:19:19*kugel thinks that Massa's idea would be indeed a nice feature, but doesn't know about limitations
17:19:38linuxstb_Massa: What's the advantage though?
17:20:13amiconnWell, in case of colour targets, it should be possible to implement this idea (in the viewport logic) without slowdown
17:21:02scorche|shGodEater: ah...i forgot to read the backlog today :)
17:21:16amiconnI can even imagine 2 possible ways to implement this. Either fiddle with drawmodes, or have a flag per viewport that specifies whether the backdrop is used or not
17:21:18Massalinuxstb_: you could have some text (in one viewport) which will be highlighted by a background colour and other which shows the backdrop...
17:21:54Massaamiconn: so there's no limitation for changing the background colour when using backdrops?
17:23:22kugelMassa: I'd wonder if there's a limitation. How would the bar line selectors be possible? :)
17:23:51 Quit XavierGr (Nick collision from services.)
17:24:02 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:24:16Massakugel: you're right
17:28:16kugellinuxstb_: could we get back to my issue?
17:28:47 Join DGMurdockIII [0] (
17:29:38 Quit MethoS- (Remote closed the connection)
17:29:51kugellinuxstb_: The only idea I have is to have an extra vp between the parent and the list vps, which is only there to contain the coustomlist information (and to be a valid target for a viewport_parse function)
17:29:53DGMurdockIIIis there anymore progress on the Creative Zen Vision
17:29:57Massaamiconn: I like your first suggestion (drawmodes) more - it allows more flexibility...
17:30:01pixelmais setting foreground/background colours in each backdrop really needed? What if someone could just leave it blank and then the global settings will be used (and a (WPS) backdrop would have priority over the background colour like it is in the menus)?
17:30:05scorche|shDGMurdockIII: look in the forum thread
17:30:23kugellinuxstb_: or a viewport-like struct
17:30:58 Quit csc` (Read error: 110 (Connection timed out))
17:31:31 Join csc` [0] (n=csc@archlinux/user/csc)
17:31:53kugellinuxstb_: Another idea would to split the list head and the actual list completely, and allowing the user to set the list head independantely from the list
17:34:20 Join kushal_12_27_200 [0] (n=kushal@
17:34:39*kugel awaits an answer eagerly
17:36:21*GodEater thinks linuxstb_ has succumbed to his large lunch and has fallen asleep
17:36:35 Nick birthday`fxb is now known as fxb__ (
17:36:56kugelother people are welcome to answer too though
17:37:37GodEaterI don't even understand the problem....
17:39:44Massapixelma: I don't completely understand what you meant. What do you want to say with "in each backdrop"? We were talking about setting the background colour even when a (=one) backdrop is used...
17:40:08pixelmaI meant "each viewport"... sorry
17:40:50Massakugel: I'm currently also unable to understand your problem...
17:41:32Massapixelma: if you don't have a backdrop, you are already able to set the background colour for each viewport - but with a backdrop it'll be always "transparent2
17:42:28 Quit OlivierBorowski_ (Read error: 110 (Connection timed out))
17:42:49kugelIt's about a function to parse a |..|...|...| structure (i.e. those in %V tags) to a viewport. linuxstb_ had the idea to implement this as a generic function, to be used for wps and lists. But in the list we have more than one viewport, so we cannot directly parse the |...|...|..| to a viewport
17:42:57 Join maraz_ [0] (
17:43:01 Quit maraz_ (Client Quit)
17:45:44pixelmaI think I didn't make myself clear. Currently it seems mandatory to define background and foreground colour for each viewport you want to use. I was thinking that it could be nice that - if you don't fill out this definition (so use "%V|x|y|width|height|font|||") this could be translated to "use the global foreground or background colours"...
17:47:26pixelmamaybe then leaving the bgshade part blank could mean "use the backdrop" and if you fill it out the background colour would be used
17:49:32pixelmaif you don't have any backdrop set, leaving the bgshade part blank could mean "use global background colour"
17:50:19pixelmawas just an idea, not very well thought through before speaking (and trying to explain)
17:54:03Massakugel: you mean store the parse result of the "|...|...|...|" definition (x,y,width,height,font,...) to the viewport?
17:54:13 Join mikus [0] (n=5316e543@gateway/web/cgi-irc/
17:55:03Massapixelma: this could be an addition to my idea of using the "FF00FF" colour as transparent colour (which would use the global colour when no backdrop is used)
17:56:51Massakugel: sorry, I still don't get it - doesn't the viewport structure have the x, y, width, ... as members? So what's the problem?
17:57:35 Quit DaCapn (Read error: 110 (Connection timed out))
17:57:40kugelMassa: In lists there's more than 1 viewport.
17:59:54Massakugel: And? Every viewport still has its dimension and font... :?
18:00:20*Massa is currently a bit slow on the uptake
18:01:31MassaAbout the bg-colour: I just had a look at the sources - and must admit, that my knowledge about the source structure is almost gone :(
18:01:55MassaWhere will the wps_data objects be drawn to the screen?
18:02:42Nico_Pkugel: I don't understand your problem either. that generic list parsing func is meant for WPS use btw
18:03:11kugelNico_P: it's not.
18:03:25Nico_Pkugel: how do you envision using it?
18:04:26kugelin my customlist patch. I allready implemented a %V-like setting (list viewport: x|y...)
18:07:25MassaNico_P: It's been a while since I looked at the rockbox sources. Can you tell me in which source file the drawing of the wps_data objects to the screen is done?
18:07:39Nico_PMassa: gwps-common.c
18:08:23Nico_Pnice to see you back btw :)
18:09:21MassaNico_P: that's where I'm already in - but I don't find it :( ... I have to dig around a bit :)
18:10:00Nico_PMassa: most of it is done in gui_wps_refresh()
18:10:45 Quit gevaerts ("work->home")
18:11:40MassaNico_P: I was never totally away - just had no time to code. I was still reading from time to time (and always had a look at the commits ;) )
18:11:45 Join Nevtus [0] (n=Nevtus@unaffiliated/nevtus)
18:12:06 Join mf0102 [0] (n=michi@
18:13:40MassaNico_P: Wow! The method looks totally different than last time I had a look at it :)
18:13:53RincewindA quick english question: is it "plugin" or "plug-in", "backend" or "back-end" or "back end"?
18:14:11Nico_PMassa: maybe because of the viewports commit... unless last time you looked at it was before the WPS tokenizer ;)
18:14:41 Part DGMurdockIII
18:15:01MassaNico_P: indeed, it was before the tokenizer ;)
18:20:46 Join [FS]_Rage [0] (
18:20:47 Nick [FS]_Rage is now known as [FS] (
18:21:09 Nick [FS] is now known as Rrage (
18:21:15MassaNico_P: as far as I understand, write_line actually writes the current line to the screen. But where will they get the colour?
18:21:54Nico_PMassa: it's written within the current viewport, so with that vp's fg colour
18:22:24RrageDoes anyone know a good tut for ipod loader? i dont understand the wiki
18:22:35 Quit petur ("work->home")
18:23:22MassaNico_P: where is that done? I don't see it...
18:23:51scorche|shRrage: wiki? should eb following the manual
18:24:15Rragethe ipod loader 2
18:24:15LambdaCalculus37Rrage: You can use Rockbox Utility:
18:24:28LambdaCalculus37Rrage: Ask the iPodLinux guys, not us.
18:24:39Nico_PMassa: setting the current viewport is probably done when encountering a vp tag (I'm not familiar with that code), and then all viewport drawing is handled at the lcd driver level
18:25:10 Quit Rrage (Client Quit)
18:25:54 Quit ch4os_ ("Lost terminal")
18:29:50 Join waldo [0] (
18:35:18 Quit ColinT (Remote closed the connection)
18:38:01 Join R1ckdu [0] (n=5085f6f9@gateway/web/cgi-irc/
18:40:55 Join ChuckMcKnight [0] (n=9e9ef0e7@gateway/web/cgi-irc/
18:41:00MassaNico_P: I'm just searching for the place where background colour of the viewport is used (without backdrop) or ignored (with backdrop) - but I don't find where the colours are used...
18:41:02pondlifeRincewind: I'd go for plugin and back-end
18:41:13pondlifeInconsistent, I know... :)
18:41:17Nico_PMassa: that'd be in the lcd drivers code
18:41:28pondlifeplug-in would be perhaps better, but we always use plugin
18:41:43pondlifeaka ploojin
18:41:52 Quit ChuckMcKnight (Client Quit)
18:41:54Nico_PMassa: but linuxstb_ / linuxstb is the person to ask in this case
18:42:28Rincewindpondlife, yes, ispell would like plug-in, too.
18:43:29Rincewindbut I think I better use plugin to be consistent with rockbox style
18:43:39 Join ColinT [0] (
18:44:02Nico_Ppondlife: I'm currently implementing an FSM approach for the "buffering" in playback.c
18:44:22*pondlife likes FSMs
18:44:43pondlifeWhat states do we need?
18:45:00pondlifeNot many, I'd hope
18:45:29*Rincewind doesn't like FSMs very much but has to use them in vhdl all the time
18:45:36 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:45:41 Join gevaerts [0] (
18:45:58Nico_Ppondlife: I have 5 right now (idle, waiting, filling, full, finished)
18:46:09pondlifefinished != idle?
18:46:12Nico_Psome of them might overlap thoough
18:46:27Nico_Pfinished is when we've reached the end of the playlist
18:46:41pondlifeAh, ok.
18:47:31R1ckdu Lambda Calculus37 : Hey, maybe you remember me from about an hour ago. I just really got myself into trouble with my gigabeat S, my firmware crashed and i cannot reload it(not even with the tool that is supplied on the page on your wiki[ the page concerning gigabeatsinfo]). The tool connects, the player reacts and then the util, says there has been no connection. What can this be?
18:48:54 Part linuxstb_ ("Leaving")
18:49:21Nico_PR1ckdu: what OS are you using?
18:49:23MassaNico_P: thanks - I think I found it...
18:51:02*Nico_P asks for an authorization to use goto
18:51:26Nico_Pit's for a sane purpose: avoid code duplication for a return case
18:52:01R1ckduI am using Vista.
18:52:11*gevaerts hands Nico_P a ticket saying "Allow one goto"
18:52:39 Join MethoS [0] (
18:53:29Nico_Pgevaerts: thanks :)
18:58:08 Quit R1ckdu ("CGI:IRC (Ping timeout)")
19:00:56 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
19:02:01 Quit mikus ("CGI:IRC")
19:02:07 Join superfly_za [0] (
19:02:38***Saving seen data "./dancer.seen"
19:05:23 Join erik006-2 [0] (n=erik006@
19:08:00 Nick Massa is now known as Massa_ (
19:08:14 Quit EspeonEefi ("さよなら")
19:08:16 Join miepchen^schlaf [0] (
19:09:06 Quit miepchen^schlaf (Client Quit)
19:09:51 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:10:22 Join miepchen^schlaf [0] (
19:14:21 Join Arathis [0] (
19:14:40 Part superfly_za ("Leaving")
19:15:00 Quit bughunter2 (Read error: 104 (Connection reset by peer))
19:15:21 Join DerPapst [0] (
19:15:22 Join bughunter2 [0] (
19:17:01toffe82high intensity light
19:17:03toffe82sorry :)
19:17:37toffe82#rockbox is not google yet :)
19:17:38amiconnNico_P: Imho goto is okay when used sparsely, and reasonably.
19:17:47EskimoSpyIs it possible in Rockbox to choose music by database, then by genre....then by album title from there, rather than just being shown artists?
19:18:28EskimoSpyDoesn't make much sense as it is for people who tend to have more full albums than hand-picked songs here and there from various albums
19:19:00Nico_Pamiconn: that's what I think too. I'm pretty sure it's acceptable the way I used it (and the patch will be up for review)
19:19:02 Join MethoS- [0] (n=clemens@
19:19:02gevaertsEskimoSpy: it is possible. I believe it's described somewhere in the manual or on the wiki, but I have never done it
19:19:33 Part pondlife
19:20:01EskimoSpyneato, I'll have to look around I guess
19:20:11EskimoSpyI'll probably be back here soon though
19:20:12amiconnNico_P: Check e.g. ata_mmc.c. That one also has a few gotos, to avoid having write the same code over and over for error returns
19:20:30Nico_Pthat's exactly how I plan to use it
19:20:38Nico_Phow I am using it actually
19:22:38amiconnJdGordon: I was now able to test the button bar fix on target a bit. It's nice to have the lists back to normal, but there's a minor nitpick: The list position in the browser doesn't adjust to the new list height until you move the cursor. Also affects the scroll bar
19:22:55 Join alazar [0] (
19:24:12EskimoSpyWell, it's not in the manual and there's no sane way of browsing the wiki. Anybody know how to choose music via Database, then sort by genre, and THEN sort by album rather than artist?
19:25:07bluebrotherEskimoSpy: you need to adjust your tagnavi.config. The wiki tells you how ...
19:25:14bluebrother(try the search box ;-)
19:25:43 Quit Thundercloud (Remote closed the connection)
19:25:52EskimoSpyAh, okay
19:25:59bluebrotherfor your convenience:
19:26:01EskimoSpyI told you I'd be back here for more info
19:26:05Nico_Pto anyone interested, here is my preliminary work:
19:26:24Nico_PIt seems to work nicely on the sim. Now I'm off to do some shopping.
19:27:48EskimoSpyget me some pizza rolls
19:30:06 Quit PaulJam (Read error: 113 (No route to host))
19:30:19 Quit erik006 (Connection timed out)
19:30:40alazarso, can I just delete the whole doom directory without some kind of bad effect? Or is there more to removal of a package?
19:30:52EskimoSpybluebrother, about says to edit "the" tagnavi_custom.config....should this exist already, or should it be created as a copy of tagnavi.config, or should it be created as a new, empty file?
19:31:21bluebrotherhmm, some buttons are missing for recording from the manual. That's the reason why it doesn't build
19:31:52bluebrotherEskimoSpy: tagnavi.config will include that file if it's present. You could also edit tagnavi.config directly −− but that will get overwritten if you update Rockbox
19:32:04BigBambialazar: Delete it as you wish
19:32:18alazarvery nice, thank you
19:32:25bluebrotherI haven't played around with that at all but I think you should just create that file.
19:32:52 Quit MethoS (Read error: 110 (Connection timed out))
19:35:05EskimoSpyhmm, alright
19:35:17 Part alazar
19:35:24EskimoSpyI was wondering about it including it though....which file's contents take precedence over the other?
19:36:03bluebrotherI think it's added as a new submenu
19:37:48EskimoSpyI'll watch for that, then...thanks cap'n
19:38:26preglowanyone here tried to compile the mikmod patch for coldfire?
19:38:32 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:39:28 Nick fxb__ is now known as birthday`fxb (
19:39:55LambdaCalculus37preglow: Which patch is that?
19:41:14 Nick birthday`fxb is now known as fxb (
19:41:33EskimoSpyI want a damn mikmod patch :P
19:42:17LambdaCalculus37preglow: Nope, haven't seen it till now. I used FS #8680 on my iPods.
19:42:26LambdaCalculus37Why, is it giving you trouble?
19:43:06preglowgcc pukes all over the place
19:43:43DerPapstdid the not licences
19:44:03*DerPapst curses his return button
19:44:19DerPapstdid the non-licence free patch work?
19:44:49LambdaCalculus37DerPapst: Yes, it works wonderfully.
19:45:20DerPapsti mean for coldfire
19:45:26preglowyes, that patch should probably be commited
19:45:38preglowi haven't got too many .mod, thugh, mostly s3m, it and xm
19:46:08pixelmait doesn't work completely "wonderfully" for me, it has a few more minor problems than the one before (but one big plus)
19:46:17LambdaCalculus37I've got a lot of .S3M, .IT, and .XM files I'd like to listen to as well.
19:46:27LambdaCalculus37pixelma: Which DAP did you try it on?
19:46:29preglowpixelma: what doesn't work?
19:46:38 Join MethoS-- [0] (n=clemens@
19:46:45*DerPapst has mikmodule and hides :-P
19:46:51*amiconn got a few freezes with the old mod patch as well
19:47:12LambdaCalculus37amiconn: You mean FS #2634?
19:47:14 Join PaulJam [0] (
19:47:42*LambdaCalculus37 has not had any issues playing MOD files on his iPods
19:48:18pixelmatried it on my c200. I got one file that was just skipped by the old patch and the new one tries to play it what either causes "undefined instructions" at varying addresses or just sounds _completely_ wrong
19:48:40 Quit [omni] (Remote closed the connection)
19:49:10amiconnLambdaCalculus37: No, that's the feature request. The patch I am (still) using is fs #5241
19:49:40LambdaCalculus37amiconn: Oops, my mistake. :)
19:49:51 Join EspeonEefi [0] (i=espeonee@STRATTON-FIVE-FOURTY-TWO.MIT.EDU)
19:50:15amiconnSome mods sound wrong with it, it skips some mods which should be small enough for its limit, and it causes occasional freezes
19:50:17pixelmawhat bothers me most is that the new one plays (I think) all files at least two times if not infinitely (wasn't patient enough to wait)
19:50:21 Join GodEater_ [50] (n=bryan@rockbox/staff/GodEater)
19:50:21EskimoSpybluebrother, thanks a ton for the tagnavi tip, exactly what I was looking for and works fine now
19:50:31bluebrotherEskimoSpy: you're welcome
19:50:55pixelmathe old one did this with some of my files.
19:51:17*LambdaCalculus37 hasn't used the older patch, so he can't comment on it
19:51:28*pixelma should just remember (to comment on in) the tracker
19:51:52*DerPapst wonders why no one tried to port libmikmod which uses LGPL and performes very well on iPods.
19:52:14*GodEater_ 's C240 has turned up, and is also a V1
19:52:41LambdaCalculus37GodEater_: You know what to do now. :)
19:53:18LambdaCalculus37DerPapst: It's a mystery of life which no one will understand. ;)
19:53:23DerPapsttry to brick it and ask gevaerts to help fixing it?
19:53:39GodEater_DerPapst: roger that, I'll get right on it ;)
19:54:28pixelmathe old mod patch showed "current pattern/all patterns" in the place of "current playing time/track time". The new one shows "current pattern/2:00"
19:54:40preglowthe mikmod plugin seems to not use interpolation
19:56:40LambdaCalculus37pixelma: I think we should leave a comment regarding that in the tracker.
19:57:30pixelmaI always tend to forget the tracker though... ;\
19:57:33*bluebrother has manuals building again
19:58:04bluebrotherthere are buttons missing for Ipods and H10, namely Pause, Exit and Menu for the recording screen.
19:58:28bluebrotherdoes someone know those buttons so I can add them while I'm at it?
19:58:51*amiconn would prefer if the mod codec would display the true times
19:59:15*GodEater_ appears to own a Rockbox'd C240 :)
19:59:42LambdaCalculus37GodEater: How many players does that make your collection now?
20:00:13amiconnUsing the pattern as the "time" isn't nice. It leads to wrong playtimes in the db, and it also confuses "caption backlight"
20:00:24GodEater_LambdaCalculus37: 5 now
20:00:32amiconnGodEater: all different ones?
20:00:37GodEater_amiconn: yep
20:01:02*GodEater_ will now go on a hunt for a cheapish sdhc card
20:01:05amiconnWell, still some to go ;)
20:02:09*DerPapst owns "only" 4 rockboxed, 3 different types.
20:02:37GodEater_amiconn: to get a complete collection ? Heh - yeah, *lots* to go
20:02:54*amiconn wonders whether someone has more different targets than he has
20:03:12 Join Chronon [0] (
20:03:12GodEater_Llorean has quite a few doesn't he ?
20:03:22preglowmikmod plugin doesn't seem to work too stellar in the sim either
20:03:25preglowseems to collide with native headers
20:03:41amiconnLinus probably has more targets than me, but not that many different ones
20:03:45 Quit MethoS- (Read error: 113 (No route to host))
20:04:18*amiconn has 12 rockbox targets, all different
20:04:36 Quit DerPapst (Nick collision from services.)
20:04:40*amiconn must be crazy ;)
20:04:46 Join DerPapst [0] (
20:06:01pixelmabluebrother: almost got a m:robe100.svg ready (just needs the text) :)
20:06:18amiconnPlus 2 broken Ondios, and an iPod 2nd Gen that lacks a HDD but would be working otherwise
20:06:47 Quit MethoS-- (Read error: 110 (Connection timed out))
20:07:28domonokyamiconn: how many left till full rockbox ownage ? :-)
20:07:28bluebrotherpixelma: nice :)
20:08:00*bluebrother feels bad with only 3 targets :(
20:08:27*domonoky also only has 3 ... goes looking at ebay :-)
20:08:31*pixelma also has only three (but a nice exotic variety)
20:08:39amiconn13 or so (to get essential coverage, i.e. leaving out variants with really tiny differences)
20:08:44DerPapstdomonoky: according to the current build page still a lot ;-)
20:09:11amiconnThose variants are H100 vs. H1x0 (x >= 2), and FM Recorder vs. Recorder v2
20:09:20domonokyamiconn: so your are halfway through :-)
20:09:36amiconnOh, I was only counting the supported ones
20:09:47 Join gregzx [0] (
20:10:21*LambdaCalculus37 has 4 targets
20:10:48*amiconn has an even architecture distribution (4x SH1, 4x coldfire, 4x arm), but a rather uneven lcd type distribution (9 b/w or mono vs. 3 colour)
20:13:49Rincewindoh my god, at first I thought that I would never write enough in my gsoc proposal to reach the character limit and now I am half way done and used 4000 chars already :(
20:13:59*domonoky just bought an m:robe100 on ebay, and still no player with color display.. but then i have all archs (1 sh, 1 coldfire, 1 arm) :-)
20:14:37 Join mikus [0] (n=5316bf7c@gateway/web/cgi-irc/
20:14:40*LambdaCalculus37 has a bad distribution of archs... 3 PP and 1 SH
20:16:06*Rincewind has only two players and people call him crazy already
20:16:12*GodEater_ has included his gigabeat S in his count
20:16:24pixelmamy 3 targets are all nice differently (all archs, all display "types")
20:16:53HorschtOk, as of recently, I am experiencing "track repeats". sometimes, after a song has finished, it plays a second time.
20:17:24Horschtunfortunately, I was not able to reproduce it controlled
20:17:44Horschtor find out what causes it, it just happens
20:17:51pixelmaHorscht: me too, it's known (at least pondlife is aware of it, I think Nico_P too)
20:17:52GodEater_buffering badness
20:18:16LambdaCalculus37Horscht: What SVN revision?
20:18:19Horschtah, ok. thanks pixelma
20:18:41Horschtthe rev I am using is 6 days old
20:18:54Horschtright now my Ipod is charging, so I can't tell the exact rev
20:18:59pixelmanaughty... ;)
20:19:02 Join piga [0] (n=leonardo@
20:19:10scorche|shpixelma: all display types?...excluding charcell, i assume...
20:19:45pixelmascorche|sh: ok, you're right
20:20:17 Quit piga (Client Quit)
20:20:21Horschtrev 16784
20:20:45Horscht3 days old
20:21:02LambdaCalculus37Naughty... :P
20:21:11Horschtwhy is that naughty?
20:21:23Horschtjust because I am not using the very latest rev?
20:21:32GodEater_Horscht: Thou shalt always report bugs on the latest revision
20:21:42DerPapstyou're not allowed to be more then 10 revisions behind.
20:22:13LambdaCalculus37Horscht: Now go to your room, and think about how you offended the Swedes. :P
20:22:20*amiconn has been imprecise regarding display types as well ;) 1x charcell, 3x mono, 5x greyscale, 3x colour
20:22:39HorschtLambdaCalculus37, who cares about the swedes, this channel is about me :p
20:22:46GodEater_it's all about me!
20:23:03Horschtmy right fist says something different!
20:23:10*DerPapst hides
20:23:25*LambdaCalculus37 considers having the topic changed to #rockbox-casa-de-godeater
20:23:38*gevaerts only has four players
20:23:45*Horscht only has one
20:24:03 Join petur [50] (n=petur@rockbox/developer/petur)
20:24:13Horschtbut only because you lazy punks haven't ported rockbox to my very useless Thomson Lyra :p
20:24:34gevaertsHorscht: you know who does the work on new ports ;)
20:24:44*GodEater_ yawns to show how lazy he is
20:25:48*Horscht does absolutely nothing
20:26:18*gevaerts accuses Horscht of typing
20:26:28amiconnHorscht: Then port it!
20:26:37*amiconn did that for half of his targets
20:27:01Horschti ported it, but then a code monster eated it
20:27:07*markun hopefully soon for 2/3 of his targets :)
20:27:21*gevaerts thinks that amiconn cheats. It's easy to have lots of rockbox players if you just buy random players and then do the port !
20:27:29markunHorscht: what hardware is inside?
20:27:42*BigBambi does the V1 dance
20:27:51HorschtSD Readerr :p
20:27:59markunHorscht: I mean things like the CPU
20:28:01 Join framo [0] (
20:28:04bluebrotherpixelma: do you know any of the missing buttons in ?
20:28:07Horschtmarkun, to make it clear, I was joking
20:28:08BigBambiI also join the 5 targets clubs (counting the beast)
20:28:25*bluebrother still wants to get a gigabeat
20:28:39*LambdaCalculus37 will be joining the 5 targets club when he geats a Gigabeast
20:28:48BigBambibluebrother: gigabeat or gigabeast? :)
20:28:51HorschtI don't want, expect, or need a rockbox port to my thomson lyra. That's my old player that was abandoned in favour of a rockboxable Ipod
20:29:07Horschtit only has a whimsy 256MB + an SD slot
20:29:20pixelmabluebrother: I could only look it up in the keymap files myself
20:30:13bluebrothertoo bad :/
20:38:22markunshouldn't the bootloaders also use the lds files in the target tree?
20:39:54amiconnThe bootloaders often need to be linked differently
20:39:54 Quit mikus ("CGI:IRC (EOF)")
20:40:11 Quit kushal_12_27_200 (Read error: 113 (No route to host))
20:40:19markunamiconn: I know, but neither do they need to be linked the same as the other targets
20:40:38markunnow it's all in 1 big file
20:41:12 Join jborn_ [0] (
20:41:30 Nick jborn_ is now known as JoeyBorn (
20:41:31amiconnYes, that could be split out into individual linker files in the target tree
20:42:03 Join Arathis2 [0] (
20:42:15amiconnBut either they need to stay separate files, or the target specific linker script needs to be sprinkled with #ifdef BOOTLOADER
20:44:06 Quit rasher (Read error: 110 (Connection timed out))
20:44:52*DerPapst wonders.... Would something like an awesome battleship plugin with AI (different strenghes), changeable fieldsize, highscore and built-in help using own graphics on various bitmap targets be accepted as GSoC project for someone not being leet enough to mess with the core?
20:45:26*GodEater_ doesn't see why not
20:45:50*DerPapst thinks about applying...
20:46:08*GodEater_ wasn't aware DerPapst was still in college
20:47:00gevaertsDerPapst: think about it this way : the worst that can happen is that it doesn't get accepted
20:47:03GodEater_well it sounds like a better application than the one we have already
20:47:12 Join rasher [50] (n=rasher@rockbox/developer/rasher)
20:48:04RincewindDerPapst: if it isn't enough work you can think about a games collection. A Mastermind clone would be nice, too.
20:48:25DerPapstthough i would have start coding before May 26 then make my exams and continue after them.
20:48:38RincewindDerPapst: same here
20:48:48DerPapstmastermind is nice too.
20:48:52 Quit GodEater_ ("fooooooooooooooood")
20:49:09DerPapsti could do that after finishing battleship ;-)
20:50:18RincewindI read a blog post about a mastermind ai that can solve every game in astonishingly few steps (don't remember how few)
20:50:38*DerPapst has to read a bit more about how gsoc works... the rules and stuff.
20:51:39scorche|shargh...i almost typed !faq for that =/
20:51:45*bluebrother thinks about closing FS #8773
20:52:15*DerPapst is reading the faqs already :-P
20:52:46Rincewindscorche|sh: are you some kind of über-mentor for gsoc, or just a regular at #gsoc?
20:53:44scorche|shRincewind: eh...i guess you could say i am just helpful and a trusted friend...
20:54:10Nico_Pthere is an algorithm to solve mastermind quickly
20:54:21bluebrotherhmm ... can anyone tell me which button leaves the recording screen on Ipods?
20:54:38*DerPapst tries...
20:54:40DerPapstwait a sec
20:54:52Nico_Pit can ben solved in less than 5 guesses
20:55:11 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
20:55:25scorche|shRincewind: actually, why did you say that?...i dont even see you in there..
20:55:26kugellinuxstb: I have a problem with the vp parse function
20:55:51RincewindI was in there a few times
20:55:54kugellinuxstb: it works when I call them in the lists, but not from wps_parser
20:56:16Rincewind#gsoc isn't in may auto-join list yet
20:56:23Nico_Pkugel: got some code to showN
20:56:54DerPapstbluebrother: menu leaves the recording screen on my 5.5G
20:57:17Rincewindscorche|sh: and I got !faq-ed there :)
20:57:31bluebrotherDerPapst: hmm, and what button invokes the recording menu?
20:57:54 Join disorganizer [0] (
20:58:05amiconnI think we should apply the modified target tree scheme for arm targets to those not using it yet (i.e. shuffle around and rename a bunch of files)
20:58:14bluebrotherI had the impression Menu did that
20:58:25amiconnThis will ease a number of things, and help to avoid unnecessary code duplication
20:58:26DerPapsti can't fire a menu up.
20:58:46 Quit Arathis (Read error: 110 (Connection timed out))
20:58:54DerPapstno long-presses or button combinations.
20:59:22kugelNico_P: I hope the diff pleases you
20:59:35amiconnThe modified scheme is arm/<sub-architecture>/<manufacturer_or_brand>-<model> instead of arm/<manufacturer_or_brand>/<model>
20:59:44kugelThat's the complete current customlist patch
21:00:37amiconnGigabeats, iFP, logixdax, cowon d2 and m:robe 500 already use that scheme
21:00:49disorganizerkugel: you said the vp-list needs more than 1 vp. how many? not that i run into the 15 vp limit some day :-)
21:01:05amiconnIt's mostly the PP targets that need cleaup
21:01:12bluebrotherDerPapst: what does the select button?
21:01:19Nico_Pkugel: and so what is it that isn't working?
21:01:29kugeldisorganizer: several, though I added only 2
21:01:40DerPapstscorche|sh: btw i was reading the gsoc faqs before you wanted to do the !faq thingy :-P
21:01:50scorche|shDerPapst: good man :)
21:01:52kugelNico_P: viewport_parse_viewport when called from wps_parser.c
21:02:03DerPapstbluebrother: nothing.
21:02:13disorganizerkugel: so we propably need to "raise the limit" so we dont run into the max-vp problem.
21:02:23kugeldisorganizer: maybe
21:02:39***Saving seen data "./dancer.seen"
21:03:26disorganizerkugel: almost definitely. my wps=13 + 2 = 15 :-) + possible statubar vp ..... :-( what happens if we run into the limit? in the old pre-commit version of the wps-vp patch it crashed.
21:04:05*bluebrother will leave a few fixmes then
21:04:12kugeldisorganizer: the list w/o my patch allready uses 6 or 7 vp
21:04:51kugeldisorganizer: I also the think the MAX_VIEWPORT thing only applies for the wps
21:05:21*DerPapst notices the gsoc faqs have some typos...
21:05:36kugelNico_P: Any ideas?
21:06:14Nico_Pkugel: reading
21:06:17disorganizerlinuxstb: does MAX_VIEWPORT apply only to the wps or is this a limit for all vp at any time?
21:06:26kugelSorry for impatience
21:06:28gevaertsDerPapst: mention them on #gsoc ?
21:07:09disorganizerkugel: are the wps-vp's deleted / erased from memory when the list comes up?
21:07:40DerPapsti'm unsure... maybe they're intended...
21:08:06scorche|shthey make sure you are paying attention! ;)
21:08:14gevaertsStill, mentioning faq typos shows them that you really are reading them
21:08:40kugeldisorganizer: I'm not totally sure, but I think the wps is only initialized once, so no
21:09:03 Nick Massa_ is now known as Massa (
21:09:14*disorganizer obviously does not understand how the viewport system works
21:09:17DerPapste.g. "We'll be funding a few more more student projects in 2008." more more? inteded or not... that is the question ;-)
21:09:19MassaNow I'm back again :)
21:09:57gevaertsDerPapst: that looks like a clear error to me
21:10:13Rincewindit's nice that the faq has errors, then my proposal can have errrors, too ;)
21:10:32scorche|shgevaerts: well, the people you want to impress are the org mentors...not google =P
21:10:35MassaDerPapst: I also know some more student projects: "bringing jpeg code to core?" or "calling plugins from WPS" ;)
21:11:14disorganizerkugel: wouldnt it make sense to remove the wps vp-structs from memory if they are not needed?
21:11:45kugeldisorganizer: would it make sense to recalculate the whole wps everytime it's entered?
21:12:15Massaamiconn: now I have an almost working version of the viewport colouring. But I've two small problems.
21:12:51*gevaerts warns Rincewind that we will go over the application with a battery of spell checkers, grammar references and dictionaries
21:13:20MassaFirst problem is that I have to define a separate "initial" viewport with whole screen resolution and background "colour" FF00FF
21:13:38*DerPapst thinks about not applying anymore.
21:13:47 Join DerDome [0] (
21:13:57gevaertsDerPapst: I put "application" in singular ;)
21:14:07bluebrotherDerPapst: just because of the spell checkers?
21:14:14kugelMassa: there's allready a initial viewport in the wps
21:14:27kugelMassa: can't you edit this one?
21:14:40MassaIf I don't define it, the backdrop is not visible in most part of the screen.
21:14:41kugelDerPapst: What's your project?
21:14:55*Rincewind has to do a special application for gevaerts spell checker then, and his normal one
21:15:08DerPapstnone yet. but i'm thinking of a battleship plugin.
21:15:39DerPapstwith ai n stuff.
21:15:45Massakugel: No, no. I just changed some small behaviour which we discussed earlier this day: it's how the background colour definition is handled for viewports.
21:15:47 Join piga [0] (n=leonardo@
21:16:20kugelMassa: I know I know. That's why I asked if you can't give the allready existing initial viewport FF00FF as bg color
21:16:22*gevaerts wonders if Rincewind heard any news recently from Lavaeolus
21:16:24 Join shnee [0] (
21:17:20Nico_Pkugel: are you sure you're giving it a pointer to the correct location? I'd set a breakpoint on viewport_parse_viewport and see what it's getting
21:18:34*disorganizer thinks if we dont minimize vp-overhead we will get slapped with the memory waster price of the year
21:18:37Massakugel: what do you mean? Currently the bg-colour will be ignored and is always transparent if using a backdrop. With my changes, it'll also work when using a backdrop. If you want to be transparent, you've to define the bg-colour "FF00FF" (as it's currently in BMPs)
21:18:42kugelNico_P: What's a breakpoint? :) Anyway, I'm doing exactly the same as in list.c (giving a viewport *vp and a *ptr as arguments
21:19:35 Join Soap [50] (n=Soap@rockbox/staff/soap)
21:19:42kugelMassa: You talked about adding a initial vp with bg color FF00FF. I thought you could assign FF00FF to the EXISTING (there's an initial viewport atm) initial viewport
21:19:49Nico_Pkugel: you should start learning how to use gdb :)
21:20:00amiconnkugel: Setting the initial viewport to 0xff00ff wouldn't be wise. Imagine how it would look without a backdrop...
21:20:08Nico_Pmaybe the pointer's not pointing where you think it is
21:20:37kugelamiconn: hmm true
21:20:39Massakugel: that's how it should work - but currently my code does not do that ;) - I first have to find the original "initial" viewport definition...
21:21:05*Nico_P makes supper
21:21:44kugelNico_P: but doesn't it look correct to you?
21:21:54disorganizerwouldnt it be cleaner to use drawmodes and let the vp-definition decide to be transparent etc with an additional parameter?
21:21:57Nico_PI haven't looked in enough detail to say
21:22:28Massaamiconn: your suggestion directly leads to my second problem :) - I decided that a FF00FF in a viewport definition without backdrop should lead to the system defined background colour.
21:22:41 Quit DerDome (Nick collision from services.)
21:22:42 Join DerDom1 [0] (
21:22:52 Nick DerDom1 is now known as DerDome (
21:23:18kugelNico_P: Look in wps_parser.c. After the "#if 1" I have the old function (parse_list...) after the "#else". this parse_list piece is exactly the same with is written in viewport.c. And the one in wps_parser.c works
21:23:19Massaamiconn: but how can I access it in the lcd driver?
21:23:29DerPapst#gsoc is noisy...
21:23:42scorche|shit is actually quite quiet atm...
21:23:43amiconnThe default background colour *is* the one set in the default viewport
21:24:09amiconnYou cannot access global_setting from firmware/, that's by design
21:24:30kugelMassa: lcd-xx.h.c in firmware/driver
21:24:36Massaamiconn: you mean the one in "default_vp"?
21:24:39 Quit PaulJam (".")
21:24:47 Quit JoeyBorn ("okey dokey")
21:25:47Massaamiconn: hmm, so what would you suggest? Totally ignoring FF00FF without backdrop? Or always setting it to LCD_DEFAULT_BG?
21:26:01amiconnNeither, I think
21:26:24amiconnThere's the global background colour, and the backdrop path resp. no backdrop path
21:26:32Nico_Pkugel: couldn't the error be in the size validation?
21:26:35amiconnThe settings code could handle this
21:26:48amiconnHmm, actually not that good an idea
21:27:20*Massa don't want a purple colour for FF00FF when not using a backdrop ;)
21:27:24amiconnMassa: I think the default viewport needs a special case, i.e. it can never have its transparency flag set
21:27:31kugelNico_P: I don't think so. the wps %V is fine. And the size validation works (I can see that since my list viewport is fine)
21:27:39 Join kushal_12_27_200 [0] (n=kushal@
21:28:14disorganizerso why dont we just add a new parameter to the %V definition and let the user decide whether to use transparency or coloring?
21:28:17kugelNico_P: Also, the size validation is copied 1 by 1 from wps_parser
21:28:36kugeldisorganizer: I don't like that idea actually
21:28:48disorganizerwhy not? personal preference?
21:29:04amiconnlcd_set_background() would set the transparency flag if colour = LCD_RGBPACK(0xff, 0x00, 0xff), but not if the default viewport is active
21:29:12kugelI'm against having mass vp parameters
21:29:24 Quit Arathis2 ("Bye, bye")
21:29:25amiconnThere's one important question left though - how do we handle this on greyscale?
21:29:28Massawhat are the lss_pattern, lse_pattern, lst_pattern members of the viewport struct?
21:30:09amiconnThose are the colour for the gradient selector bar (nasty stuff if you ask me)
21:30:20disorganizerkugel: so we just also drop the bg color, as you can always use a backdrop to color the area :-P
21:30:39Massaamiconn: so not used for wps?
21:30:48kugeldisorganizer: and if we don't have a backdrop?
21:31:10disorganizerkugel: system defined bg color. that logical and easy to understand
21:31:11Massaamiconn: how is FF00FF handled for greyscale when using BMPs?
21:31:22amiconnNot at all yet
21:31:29 Quit quitte_ ("Changing server")
21:31:43*disorganizer wonders if any user will understand that FF00FF makes transparency
21:31:51amiconnBut I have plans how to handle it: calculate a mask on load (and at compile time for built-in bitmaps)
21:31:54 Join quitte [0] (
21:32:29Massadisorganizer: the problem here is that I don't have access to the system defined bg-color (in lcd-XX.c)
21:33:08Massadisorganizer: WPS designer already have to understand FF00FF because this is the colour which is used for transparency in bitmaps...
21:33:20disorganizercould we solve this with the parser? on ||| for the colors the system colors could be used
21:33:42 Nick disorganizer is now known as dis-afk (
21:35:06Massaamiconn: BTW, the lcd_get_background methods will always return the viewport background colour, even if it's not used (because a backdrop is defined) - is this correct behaviour?
21:35:06 Quit DerDome (Read error: 104 (Connection reset by peer))
21:36:29Massaamiconn: shouldn't it return the currently used colour?
21:36:37 Join MethoS-- [0] (
21:37:02 Join DerDome [0] (
21:37:06 Quit perrikwp (" ajax IRC Client")
21:37:06 Quit GodEater (" ajax IRC Client")
21:37:38 Quit DerDome (Client Quit)
21:38:35kugelit returns the bg color of the currently active vp
21:40:23 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
21:40:35 Quit jhMikeS ()
21:43:10 Quit framo ("moo")
21:46:54 Nick erik006-2 is now known as erik006 (n=erik006@
21:49:01 Nick dis-afk is now known as disorganizer (
21:50:24disorganizeragain to my proposal: how about using the %V definition with empty colors to let the wps parset set the systemFG and BG colors as vp colors?
21:54:30kugelNico_P: Hmm, I think I found the problem
21:54:50 Join SoapHotel [0] (n=Soap@
21:54:54kugelThe part " /* Check for trailing | */" doesn't work in my version of viewport parsing
21:55:10kugelI commented that out and it works
21:56:00 Quit nplus ("Leaving")
21:57:04*DerPapst is registered at gsoc o.O
21:57:19Nico_PDerPapst: have you submitted an application yet?
21:57:28 Join FOAD_ [0] (
21:57:43DerPapstnope. i decided ~2 hours ago to apply :-P
21:57:49scorche|shpfft...leave him alone till we get it =P
21:58:27*midgey notices we now have a LocalizablePlugins page
21:59:46kugelNico_P: Indeed, weird
21:59:47 Quit rasher (Read error: 131 (Connection reset by peer))
22:00:32DerPapstNico_P: i'm planing a battleship clone plugin with AI and 2 player mode. i'll whipe something up and pastebin it. but probably not today anmore.
22:00:34kugelI'm not entirely sure why the trailing pipe is needed anyway#
22:00:38DerPapsthave to get up eraly :-/
22:00:54 Join rasher [0] (
22:01:43Nico_PDerPapst: quite a nice project idea :)
22:02:25*disorganizer wonders if we could get a nethack clone on rockbox :-)
22:02:36DerPapstscorche|sh: the faqs said not to blindly submit any applications without talking to at least one of the mentors :-) (or similar)
22:02:41 Quit kushal_12_27_200 ("This computer has gone to sleep")
22:02:50Massadisorganizer: for me this is the second (optimization) step - useful defaults when no colours are defined...
22:02:57*scorche|sh pats DerPapst's head
22:03:08DerPapstawww :-)
22:03:32*DerPapst just wanted to proove he read big parts of the faqs ;-)
22:03:33Rincewinddisorganizer: nethack without a keyboard??
22:03:40disorganizermassa: the question if whether we really need transparency without having a backdrop. as overlapping viewports are not allowed, why transparency?
22:03:55disorganizerrincewind: selectable actions, like with psp or nds
22:04:33disorganizerrincewind +on
22:04:41Rincewinddisorganizer: there was a thread once in the forum about this
22:05:07*disorganizer must think of searching the logs and forums everytime he gets an idea
22:05:33disorganizergot to go to my little one for a short while. be back in a few seconds
22:05:36 Nick disorganizer is now known as dis-afk (
22:05:46Rincewinddisorganizer: I don't want to discourage you, nethack would certainly be cool.
22:06:56Massadisorganizer: there is no transparency without backdrop - it'll currently will just use the bg-color of the default viewport. When later on overlapping will be allowed we could also set the colour of the underlying viewport...
22:09:04 Nick dis-afk is now known as disorganizer (
22:09:24disorganizermassa: what would be the sense of having transparency when there is no backdrop?
22:09:39disorganizerwhat would you see "through" the viewport's background?
22:10:40 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
22:10:52Massadisorganizer: the colour of the default viewport (that's the system defined background colour).
22:11:07*DerPapst wonders if overlapping viewports are wanted..
22:11:27disorganizerbut if we use ||| to tell the wps parser the use FG and BG of the system, that would have the same effect.
22:11:48*Massa would really love overlapping viewports!
22:12:06kugelof which target with remote can I compile the sim with arm-elf installed?
22:12:21disorganizerlinuxstb: why are overlapping viewports not allowed?
22:12:27 Quit Soap (Read error: 110 (Connection timed out))
22:12:58*disorganizer would also love overlapping viewports, and transparency would make sense then. but imho not with a strange transparency coding like with bitmaps. what if i like the color FF00FF in my wps?
22:13:01Massadisorganizer: but only if you don't use a backdrop - so if you add a backdrop to your wps, you have to change the viewport definitions...
22:13:34disorganizerno, because then you want to see the backdrop, which should pre-color the areas the viewports use :-) if needed.
22:13:54Massadisorganizer: then you have to use FF00FE instead (which IMHO does not make any visible difference)...
22:14:09 Quit FOAD (Read error: 110 (Connection timed out))
22:14:10 Nick FOAD_ is now known as FOAD (
22:14:17*disorganizer hates having codes instead of parameters
22:15:21disorganizerwhat about a byte with flags for the viewport's behaviour as parameter?
22:15:56Massadisorganizer: I don't care if it's FF00FF or an additional flag - but for me it was a logical behaviour because FF00FF is already used as transparency colour in BMPs...
22:15:58disorganizerone bit toggles transparency, ...
22:16:00 Quit bughunter2 ("Leaving.")
22:16:24*disorganizer thinks not every wps - clicker does use transparency in his/her bitmaps
22:17:20kugelFF00FF is fine, no need for an extra parameter imo
22:17:21disorganizerbut as said, transparency imho only makes sense when we have overlapping bitmaps.
22:17:41kugel1) hardly anyone uses FF00FF anyway, 2) bitmaps do it in the same way
22:18:10disorganizer-bitmaps +viewports
22:18:10DerPapstFF00FE would look almost similar on all colour targets.
22:18:27kugelalmost as ugly as FF00FF*
22:18:35disorganizerwhat about adding half-transparency ?
22:18:47disorganizerlike a transparency % value as parameter :-)
22:19:25kugeldoesn't sound bad too me, probably worth an existing parameter. But I doubt the code work is worth it
22:19:43 Quit XavierGr (Nick collision from services.)
22:19:54 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:20:07SoapHotelwould FF00FE get dithered on a 16 bit screen while FF00FF not?
22:20:22disorganizernote to myself: use sarcasm tags more often
22:20:51 Nick SoapHotel is now known as Soap (n=Soap@
22:21:26MassaI'll just implement it as is (with FF00FF as transparency colour), when it's finished I'll create a patch tracker entry and then the discussion can take place and the decision could be made be the one who commits it (or not)...
22:21:43 Join erik006-2 [0] (
22:22:07 Quit kugel ("ChatZilla 0.9.81 [Firefox]")
22:22:19 Quit linuxstb (Read error: 113 (No route to host))
22:23:04*disorganizer thinks one of the findings of a usability study will propably be that we use to many things which are logical to programmers but for noone else :-P
22:23:51disorganizeranyways i dont care, i just wanted to get the point across that imho we need to get wps's more readable.
22:24:39Massadisorganizer: when implementing partial transparency, you also have to add additional parameters to bitmap definitions (and to make it clean: also a parameter which defines which colour is used as transparency colour) - but I'm sure most users (not programmes) will be scared of more parameters
22:24:43DerPapstwhich would break backwards compatibility (or however that is spelled)
22:25:30Rincewinddisorganizer: if my gsoc proposal gets accepted we will habe a gui wps editor by the end of the summer
22:25:35disorganizerwhich is not too important, as the next weeks will break wps syntax anyways because of the development (%v tags propably need to get changed anyways if you followed the irc logs)
22:25:49*disorganizer hates guis
22:26:13Massadisorganizer: what we need is a really good "clickable" tool which reads WPS files and displays it graphically and also creates wps files :)
22:26:14disorganizerthey never do what you expect them to do
22:26:24MassaRincewind: that's what I meant!
22:26:31Soapto be rolled into RBUtil, Rincewind?
22:26:50RincewindSoap: possibly. I plan to use Qt
22:27:21MassaDerPapst: of course it would break compatibility - it's just to round up the discussion (I personally don't think partial transparency is needed).
22:28:46 Quit linuxstb_ (Read error: 110 (Connection timed out))
22:28:52disorganizerthe only important thing: we need to do the wps-compatibility breaking things before the new wps gallery is in production :-)
22:29:03 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
22:29:09disorganizerlinuxstb: problems?
22:29:35DerPapstbad isp... like me
22:29:36 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
22:31:40 Join n1s [0] (
22:36:53 Quit erik006 (Read error: 113 (No route to host))
22:38:17 Quit jgarvey ("Leaving")
22:45:05 Quit Mathiasdm ("Yuuw!")
22:46:06 Join Shaid [0] (
22:47:22 Join joey1 [0] (n=cbdc7d1f@gateway/web/cgi-irc/
22:51:21 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
22:52:32*DerPapst 's proposal for his gsoc application (draft version).
22:52:51*scorche|sh is too busy refreshing woot to look
22:53:16*gevaerts accuses scorche|sh of lying. He's also busy with other things ;)
22:53:36scorche|shwork is secondary to bags of crap
22:54:45joey1Hi guys, Does anyone know how to use VMware to compile rockbox? Ive followed the steps on the wiki page but get stuck on the make zip part. It says: "cp: cannot stat '/home/user/rockbox/wps//cabbiev2/lockclosed-128x128x16.bmp': no such file or directory" what is it that im doing wrong?
22:55:16gevaertsDerPapst: s/playfiled/playfield/
22:56:01bluebrotherDerPapst: how about writing a sentence why we need your plugin?
22:56:20bluebrotherjoey1: what player are you building for? Ipod Video?
22:56:30gevaertsDerPapst: also, maybe elaborate a bit on what you want to do with the remote
22:56:38bluebrothermake zip should continue (it only fails on some theme file, not Rockbox itself)
22:57:07joey1no im build for the iriver h10
22:57:58joey1so my zip should be created?
22:58:12DerPapstnoted. thanks :-)
22:58:30 Join phinze [0] (
22:58:33bluebrotherit should. We had a image file for the default theme missing a while back (on other targets)
22:58:51*bluebrother tries building for h10
22:59:12*petur too
22:59:14pixelmajoey1: how did you retrieve the source?
22:59:30 Join mikus [0] (n=5316ccd1@gateway/web/cgi-irc/
23:00:19 Nick erik006-2 is now known as erik006 (
23:00:44pixelmaah, there really is a mistake in the .wps file
23:00:50joey1i typed svn co cvn://svn.... into the Eterm box
23:00:58*pixelma blames linuxstb_ ;)
23:01:44pixelmaI'll fix (but will take a bit to testcompile)
23:01:58 Join Naked [0] (
23:02:43***Saving seen data "./dancer.seen"
23:02:55 Quit swimmer (Read error: 60 (Operation timed out))
23:02:55 Quit Hadaka (Read error: 60 (Operation timed out))
23:02:58 Nick Naked is now known as Hadaka (
23:03:44 Join swimmer [0] (
23:03:56joey1also what folder do i browse to on my computer to find the zip that i compiled
23:04:37 Quit petur ("*plop*")
23:05:00 Quit disorganizer ("ChatZilla 0.9.81 [Firefox]")
23:05:07 Quit ol_schoola ()
23:07:02 Quit piga (Remote closed the connection)
23:07:43 Quit mikus ("CGI:IRC")
23:08:50RincewindDerPapst: two player mode: one player main, the other player remote screen
23:09:21Nico_PDerPapst: have you seen the application template?
23:09:42DerPapstNico_P: yes. but i forget about it :-) thanks for reminging
23:09:58DerPapstRincewind: exactly
23:11:03RincewindDerPapst: you need a detailed project plan with milestones and estimates when you complete them
23:11:13*Rincewind hates that bit
23:12:08*gevaerts put that bit in the template ;)
23:12:26 Quit mf0102 ("Verlassend")
23:12:27DerPapstYep. that'll come later. I'm working on the "need" part, that you didn't have to do.
23:12:40*Rincewind hates gevaerts by proxy
23:12:46scorche|shit can be rough, but we want at least an idea and a way to measure your progress :)
23:12:58RincewindDerPapst: Need: Games are fun!
23:13:05Bagderto know when to bring out the whip!
23:13:25DerPapstand "nice if you have to kill some time"
23:13:28 Join matsl [0] (
23:13:43gevaertsRincewind: saying that you hate various mentors isn't going to help ;)
23:13:49Rincewindand "nice if you have to kill an enemy"
23:14:38*Rincewind might buy gevaerts a box of beer, too.
23:14:57*gevaerts suddenly likes Rincewind's proposal a lot
23:14:58*domonoky thinks beer surely helps :-)
23:15:17*DerPapst wonders when beer shows up
23:15:24DerPapsterr... s/beer/petur
23:15:49bluebrothergrrr ... we have capitalization issues at least on m:robe bootloader installation in rbutil
23:16:04gevaertsDerPapst: he will soon, now that you've spoken the magic word
23:16:23*DerPapst hides... just in case
23:16:40roolkubluebrother: ?
23:17:24bluebrotherrbutil calls the folder "SYSTEM" while on my player it's "System"
23:17:34bluebrotherand linux cares about the capitalization ...
23:17:36*domonoky will take the blame.. :-)
23:18:06roolkubluebrother: but surely not on FAT, how could it
23:18:26markuncan anyone build a iriver h1xx and/or h3xx bootloader for me and verify something?
23:19:52 Quit linuxstb_ (Remote closed the connection)
23:20:00 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
23:20:24markunlinuxstb: I split up, but need to do some testing before I can commit
23:20:29markundo you want to take a look?
23:20:30bluebrotherroolku: it does on FAT as that can distinguish cases.
23:21:21 Join linuxstb_ [0] (
23:21:27roolkubluebrother: yes, you are right
23:21:29 Quit domonoky (Read error: 104 (Connection reset by peer))
23:21:31 Quit linuxstb (Nick collision from services.)
23:21:54bluebrotherno issue on windows as that doesn't care about case. Still, we need to make rbutil a bit smarter in that domain.
23:21:57DerPapstGSoCApplicationTemplate2008 isn't of any help yet since i'm still busy with the abstract part. Though the page should mention what bluebrother said if the idea isn't on the Ideas List.
23:22:08 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
23:22:16linuxstbmarkun: Sure
23:22:43markunlinuxstb: I have a diff of every against the old one and a patch
23:23:03*linuxstb wonders what pixelma is blaming him for
23:24:06DerPapstrandomly breaking wpses
23:24:27RincewindDerPapst: there is one item on this in the middle of the template page
23:24:29pixelmalinuxstb: I hope to commit in a few minutes then you'll see ;)
23:24:30gevaertsDerPapst: it says 'Briefly explain why your project is useful for rockbox'.
23:24:35*Nico_P has a playback patch that shouldn't make a difference but makes things cleaner
23:24:45*linuxstb tries to look innocent
23:24:56*DerPapst clears his glases
23:25:21*gevaerts denies having added that to the wiki one minute ago
23:25:34scorche|shDerPapst: and you were doing so good with readong the FAQs ;)
23:26:10DerPapstwell... it doesn't say that for the abstract part ;-)
23:26:18DerPapstbut what ever. :-P
23:26:51*gevaerts thinks that's as it should be. The abstract should just describe the project, not try to sell it.
23:27:16ChrononI would second that, gevaerts.
23:28:09*amiconn is looking at an arbitrarily chosen manual and scratches head
23:28:10DerPapstgevaerts: so should be enough for the abstract part?
23:29:36 Join MethoS- [0] (
23:29:42 Quit Galois (Remote closed the connection)
23:29:51*Nico_P summons pondlife
23:29:56 Join Galois [0] (
23:29:59Nico_PSlasheri: here?
23:30:01gevaertsI think it's a good abstract.
23:31:04DerPapstmkay. thanks.
23:31:21*DerPapst will do the rest tomorrow.
23:31:55gevaertsBut I'm only the guy who wrote half a template on the wiki. There are other mentors...
23:31:56RincewindDerPapst: you used 989 chars, you have 1511 left for your abstract
23:32:01 Join Zarggg [0] (
23:32:01pixelmajoey1: update your source tree (run "svn up") and then try to make the zip again
23:32:28 Join erik006-2 [0] (
23:33:02*Nico_P would like comments on his FSM approach to playback.c buffering:
23:33:09gevaertsRincewind: you don't _have_ to bludgeon us to death with lots of text you know ;)
23:33:25DerPapstRincewind: heh. quality over quantity ;-)
23:33:41*Rincewind wanted to use his fancy emacs character count function again
23:33:48*DerPapst could add some random ahmms and such
23:35:27DerPapstgevaerts: i'll bugger some other mentors next days ;-) But the opinion of one mentor is good for the beginning :-)
23:36:33dionoeaHow many applicants does rockbox have?
23:36:56gevaertsDerPapst: sure, just keep in mind that I joined rockbox less than two months ago ;)
23:37:11DerPapsti know. :-)
23:37:57gevaertsYou should. You answered first when I asked my first question here :)
23:38:20DerPapstIf DevCon2008 actually happens in berlin and while gsoc takes place i might get my menor a beer ;-)
23:38:27pixelmalinuxstb: hmm... the commit message is a bit misleading but if you look at the file changed (and the dev)...
23:38:41pixelmaerr s/dev/diff
23:38:55RincewindBerlin would be nice, might be possible for me to attend, too.
23:39:07DerPapstpixelma commited an entire dev?
23:39:30*gevaerts wonders if anyone will actually do anything else than drinking beer at the DevCon
23:39:40DerPapstummm.... no.
23:39:44pixelmaDerPapst: and that from someone who knows menors ;P
23:39:45*bluebrother wonders when devcon will actually happen
23:39:49Bagdergevaerts: build towers
23:39:54DerPapstmeh :-P
23:40:10bluebrotherand drink beer afterwards? ;-)
23:40:15*gevaerts reminds people of his beer offer if the date is decided this week
23:40:38 Join SoapHotel [0] (n=Soap@
23:40:46*bluebrother missed that offer
23:41:07gevaertsBagder: before or after drinking beer ? That might have an effect on the height of the towers
23:41:34Bagderonly for the weak! ;-P
23:41:38DerPapstand the broken daps afterwards :-P
23:42:44*gevaerts repeats his promise to buy everyone at the DevCon a beer if the date is decided this week
23:42:46bluebrothergiven enought daps the tower will still be of some reasonable height ;-)
23:43:02*bluebrother would like to have a date soon too
23:43:44DerPapstwell, still 3 days to go.
23:43:50 Quit MethoS-- (Read error: 110 (Connection timed out))
23:44:02bluebrotherhow about july, 4th - 6th?
23:44:29scorche|shgevaerts: you can see what happened at past devcons...outlines are on
23:44:42scorche|shbluebrother: would make travel expensive because of US holiday
23:44:48DerPapstany weekend would be good for me.
23:45:10BagderI'm also fine with basically any date
23:45:31pixelmame too, just want to know soon enough to organise
23:45:55*gevaerts can't make any date between 2008-07-10 and 2008-08-10, but most other weekends should be fine
23:45:56DerPapstpreferrably after 1st half of june though
23:45:57bluebrothersame here −− and I have the first two weekends in june blocked
23:46:05pixelmaDerPapst: you don't have to travel far, do you?
23:46:12BagderI think it is more a matter of the organizers to pick one that seems fine and then the rest have to adjust
23:46:18bluebrotherhave we decided a location already?
23:46:24DerPapst530 whule UNI takes place
23:46:33DerPapst*while and 530km
23:46:46pixelmaI think Berlin and Stockholm are the candidates now
23:47:10scorche|shi would be fine with either, since the cost is so high for either one now =/
23:47:12*DerPapst wouldn't attend if Stockholm is chosen
23:47:17 Quit erik006 (Success)
23:48:03bluebrotherBerlin would be much shorter for me (but Stockholm would be interesting ;-)
23:48:10n1sbluebrother: thanks for fixing those button commands for the manual
23:48:25scorche|shstockholm sounds quite interesting according to GodEater's comments...
23:48:45pixelmabluebrother: I didn't see much of Stockholm last year... (though I've been there before)
23:48:56bluebrothern1s: if you know the correct buttons please fill them in ...
23:49:12DerPapstthough i doubt you'll be able to do much site seeing while DevCon
23:49:16scorche|shwherever it is, i plan on making it a week thing...2 weekends, so if others want to stick around with me and do something...
23:49:51scorche|shit just isnt worth flying out there from the US and staying just one weekend
23:50:00n1sbluebrother: the next step of crazy automation is probably to auto generate those button commands from the source .c files...
23:50:43bluebrothern1s: would be nice, but after looking at the button maps I wasn't really sure which button exactly gets used
23:50:44DerPapsta self writing manual. lovely.
23:50:53 Quit roolku ()
23:51:05*scorche|sh heads out to do some work
23:51:07bluebrotherI guess it would be quite hard to achieve this. But if you manage to do it: great ;-)
23:51:26DerPapstmaybe gsoc2014
23:51:49Rincewindthis might be a way to have customisable buttons? If the custom buttons update the manual accordingly....
23:51:59n1sbluebrother: it's currently at the very bottom of my list of ideas/interesting things :)
23:52:15 Quit SoapHotel (Read error: 104 (Connection reset by peer))
23:52:43*DerPapst hears "customisable buttons" and hides
23:52:46 Join SoapHotel [0] (n=Soap@
23:52:54bluebrotherRincewind: no ...
23:53:03Rincewindrelax, I'm just joking
23:53:12 Quit Soap (Nick collision from services.)
23:53:14 Nick SoapHotel is now known as Soap (n=Soap@
23:53:15bluebrotherunless we can get the manual to recompile once the user changes a setting ;-)
23:53:43DerPapstmanual on target, reading the config file
23:53:51ChrononAnd then support on the forums will really boil down to RTFM
23:54:18n1sWell since the manual suffers from a lack of love and care all the automation we can get is nice IMHO
23:54:20bluebrotherRead the fine machine?
23:54:52bluebrotherit should be RTFD in that case −− read the fine DAP ;-)
23:55:52 Quit tvelocity_ (Remote closed the connection)
23:55:59*gevaerts doesn't look forward to reading the manual on a c200
23:56:19bluebrotherjust think of reading it on a charchell player ;)
23:56:39DerPapstsounds like endless fun
23:56:56DerPapstways better then battleship ;-)
23:57:02Rincewindthere could be podcast of the manual
23:57:06gevaertsendless anyway. Not sure about the fun
23:57:24bluebrotherdoes anybody know how I can color the output of a Makefile running on windows cmd.exe?
23:58:23DerPapsti don't think that is possible. i haven't ssen something like that on windows at least.
23:58:39 Quit joey1 ("CGI:IRC")
23:58:44DerPapst /ignore typos

Previous day | Next day