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

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

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

#rockbox log for 2006-04-19

00:00:01Bagdert0mas: LinusN's and it is probably just switched off
00:00:08linuxstb_preglow: I don't suppose you tried removing the long-call option when you disabled IRAM?
00:00:09lostlogicthat punk.
00:01:01preglowlinuxstb_: no, no i did not
00:01:07 Join Rondom_ [0] (
00:01:34SereR0KRwhy does rockbox rule? :/
00:01:52Bagderit does?
00:01:58SereR0KRmore than IPL
00:01:58t0maswell, now I'm officially 18
00:02:00t0masit's bed time :P
00:02:04t0masas I still have to work tomorrow
00:02:11SereR0KRit's your birthday?
00:02:12yobesoomgo buy a pack of ciggies and some pr0n
00:02:16t0mas19th is
00:02:22t0masand it's 0:02 here ;)
00:02:24Bagdernow now, all kids off to bed!
00:02:27SereR0KRhappy birthday to you
00:02:32lostlogicpreglow: does auto-dir-change backwards work for you, or break?
00:02:33Bagderhappy birthday t0mas!
00:02:35SereR0KRyes it's 0:02 here too :>
00:02:37t0masBagder: not a kid anymore :P :D :D
00:02:43t0masand tnx
00:02:47amiconnlostlogic: auto-dir-change backwards???
00:02:54lostlogicis there any such thing?
00:02:56 Quit webguest13 ("CGI:IRC (EOF)")
00:03:04amiconnOf course not
00:03:14lostlogicI have no idea, an irritation on the forums said that skipping backwards from one dir to another broke for him, I'll ignore it.
00:03:18amiconn(unless you implemented backwards playback ;) )
00:03:35amiconnSkipping backwards is not auto-change
00:04:10preglowlostlogic: it hung me completely
00:04:16preglowahh, no, can shut down
00:05:45 Quit Kohlrabi ("Fast alle Menschen sind Regenw├╝rmer")
00:05:57 Quit ProgramZeta (Connection timed out)
00:06:42*t0mas turns off his phone and now really goes to bed
00:07:22linuxstb_Unsurprisingly, removing the long-calls option works when disabling IRAM. My rockbox.ipod is now 321716 bytes (4g color).
00:07:52lostlogicpreglow: hmm... wonder what that's about.
00:07:55amiconnlinuxstb: compared to?
00:07:58preglowlinuxstb_: as compared to?
00:08:22linuxstb_374966 for the latest cvs build.
00:08:38preglowas i suspected
00:08:50preglowit's eating in
00:09:16 Quit midkay ("Leaving")
00:09:42linuxstb_Now to test it...
00:12:38*amiconn whispers 'thumb code' ...
00:13:00 Join mikolas_ [0] (
00:13:24Bagderyou're ruining the moment!
00:13:56 Quit goffa_ (Read error: 104 (Connection reset by peer))
00:14:00linuxstb_It's definitely slower. q7 vorbis is at almost 100% boost. I'm sure it was better than that before.
00:14:15twisted`# just my imagination #
00:14:35preglowamiconn: thumb code is another wasps nest
00:14:52 Quit ScootScat (Read error: 110 (Connection timed out))
00:14:54preglowis that the wasps buzzing????
00:14:55 Join ap0calypse [0] (
00:15:54*amiconn is currently limited in doing Really Useful Things :(
00:15:56 Join JBGood [0] (
00:16:02ap0calypsehi, i have a question. as i try to install the bootloader i cant even find my ipod (i.e ipodpatcher 1.2.3)
00:16:07ap0calypseno numbers seem to work
00:16:10amiconnI have no idea how long my ups survives...
00:16:16lostlogic*ponders* wtf is different about trying to skip backwards at an auto-dir-change boundary than trying to skip backwards past the start of a playlist!?
00:17:36Paul_The_Nerdap0calypse: How many drives do you have, what drive letter is the iPod currently, and how many numbers (sequentially) did you try?
00:18:09ap0calypsei tried up to 7
00:18:39ap0calypsedrive letter Q
00:18:56Paul_The_NerdWell you're gonna probably need to try higher then.
00:19:20 Quit Rondom (Nick collision from services.)
00:19:23 Nick Rondom_ is now known as Rondom (
00:20:36amiconnap0calypse: Check the disk management. It will tell you the device number
00:20:37preglow50kb wasted in calls
00:20:41preglowthat's bad
00:20:52*lostlogic beats GCC with a clue stick
00:21:04preglowmost of that should be able to be reclaimed
00:21:10amiconnAh, power is back...
00:21:15linuxstb_My first test was disabling all iram usage. I'm now trying it with only the icode disabled.
00:21:25lostlogiclinuxstb_: that's a good think.
00:21:58linuxstb_321840 bytes now...
00:22:28amiconnThat's insignificant
00:22:51linuxstb_Yep, I think we can live with that.
00:23:10 Quit linuxstb_ ("CGI:IRC")
00:23:25lostlogicwow, reenabling idata killed the poor guy's IRC connection.
00:24:01*amiconn is happy that his ups survived the whole power outage
00:24:03linuxstbJust moving back to my main computer...
00:28:06linuxstbWith data in IRAM, my q7 vorbis test track is boosting at around 76%, compared to around 90% without any IRAM usage.
00:28:06 Quit mikolas (Read error: 110 (Connection timed out))
00:28:21preglowso we know there's an impact, then
00:28:22 Join ProgramZeta [0] (
00:28:32Bagderindeed significant
00:28:40linuxstbI'll download a cvs build now and see what happens.
00:30:09 Nick mikolas_ is now known as mikolas (
00:30:12preglowi'd expect data iram to pay more than code iram
00:30:19 Quit ap0calypse (Read error: 104 (Connection reset by peer))
00:30:29amiconnWhom should I tell that debian dependencies are broken?
00:30:43Bagderamiconn: file a bug report on the package with the broken deps
00:30:53preglowi wonder if the pp chip cache iram as well...
00:31:09amiconnThat would make no sense
00:31:13preglowof course
00:31:19preglowbut neither would wait-state iram
00:31:19amiconnImagine the PP5002 doing that...
00:31:49amiconnpreglow: Iiuc, PP5002 iram has no waitstates, but the cache has...
00:32:20 Quit jbauman (No route to host)
00:32:25preglowright, right
00:32:42linuxstbOK, a cvs build is settling at around 75%.
00:32:59preglowlinuxstb: my, i'm tempted...
00:33:01linuxstbSo it does seem to make a tiny impact.
00:33:29preglowlet's just disable code iram
00:33:38preglowand say "HOORAY" then forget we ever used it for code
00:33:49 Join ap0calypse [0] (
00:34:19Paul_The_NerdWhy are we cheering?
00:34:34*Paul_The_Nerd hopes at least *one* person laughs at that.
00:34:35preglowi don't know
00:34:43Bagderwe forgot
00:34:48preglowi say we don't dig too much in that
00:34:52preglowit's a happy occasion after all!
00:35:05linuxstbApart from 50KB of code size, remind me why long-calls are bad?
00:35:07*sharpe has very small rubber bands
00:35:18sharpethey're too long? :)
00:35:28preglowlinuxstb: wastes a register, more code, more memory use
00:35:46 Quit Paul_The_Nerd ("Leaving.")
00:35:49ap0calypseok , after a few seconds itunes opens up and then the drive disappears from my computer
00:36:08sharpeenable disk mode in the itunes properties
00:36:14ap0calypseit is.
00:36:50sharpethen it doesn't like you.
00:36:53linuxstbUninstall itunes?
00:36:55sharpehave you disabled itunes starting up when your ipod is inserted?
00:37:05ap0calypsebut it goes back to normal
00:37:14ap0calypseand the options erase
00:37:18ap0calypseso everytime i connect it
00:37:22preglowlinuxstb: it's fairly self evident that the extra memory reference and two extra instructions will amount to something
00:37:22ap0calypseit pops up
00:37:26sharpeyou mean, your settings don't take effect?
00:37:32preglowlinuxstb: even though it might not be the most cpu time sucking beast ever seen
00:37:46sharpetry pressing enter after you set them...
00:38:17 Join jumpfroggy [0] (
00:38:18 Join speacial_ed [0] (
00:38:24ap0calypseive installed this before on a differnt ipod but i ddint have this trouble
00:38:47linuxstbBTW, I've only done tests on my 4g (PP5020). I don't know if it's worthwhile to test on the PP5021 and PP5022.
00:38:51sharpeheh, try what linuxstb suggested :D
00:38:59jumpfroggyThose are the 3g, right?
00:39:19linuxstb5021 is the Nano and 5g/Video. 5022 is the 2nd Generation mini.
00:39:29jumpfroggyahh, got it.
00:39:36linuxstbThe 3g is the 5002.
00:40:13jumpfroggyyeah, just looked it up again.
00:40:23 Join MusiFreq [0] (
00:40:29jumpfroggyI saw on the lists that no one has been testing the 3G recently
00:40:38amiconnGah! Why does debian use all those difference names for their versions?
00:40:49Bagderno-one knows
00:40:49*amiconn doesn't know whether sid == testing
00:41:03Bagderthey do that to confuse the audience I think
00:41:23linuxstbI wonder how this affects the iFP7xx port... I know Tomasz had the same long-calls issues.
00:41:24*amiconn wades through 'reportbug'
00:42:24 Quit MusiFreq (Client Quit)
00:42:58 Join MusiFreq [0] (
00:43:19amiconnpreglow: Disabling .icode will impact 3g and lower
00:43:43Bagderit could of course be made per-target
00:43:45preglowamiconn: i know
00:43:50preglowamiconn: but we can do it only for 4g and up
00:44:09preglowh1x0 is the only swcodec player with spdif, yas
00:44:17amiconnSo far, yes
00:44:48preglowjust making sure i just add SPDIF for that target for now
00:44:58preglowit seems line-in and mic aren't translated
00:45:07 Join dj-fu [0] (
00:45:17preglowthey're stored verbatim in an array in pcm_recording.c
00:45:41amiconnThen something is wrong on swcodec
00:45:41preglowhave a look
00:45:43preglowline 64
00:45:48amiconnThey are translated on archos
00:45:56preglowdoesn't look too good
00:46:13dj-fuso uh, what's the best DAP?
00:46:30preglowlooks like the entire screen is in english
00:46:46amiconnHmm, the debian uninstallable-package bug is already reported twice :/
00:47:50preglowwhat the hell is the pcm_recording.c file?
00:47:54preglowlooks like a bloody debug screen
00:49:02linuxstbI wonder why Tomasz had long-call problems. Looking at for the iFP7xx, IRAM is at 0x400000 and DRAM is at 0xC00000. Which I calculate to be 8MB apart, with DRAM being 1MB in total.
00:49:31preglowgcc being retarded
00:49:33preglowthat's my guess
00:49:51preglowthe moment it sees something sectioned, it shits its pants in anticipation and starts making wrong calls
00:50:02linuxstbOK, who broke the ifp build?
00:50:13 Quit ender` (" You can kill the enemy as easily with a rifle as an airstrike, so use the airstrike: cluster bombing from B-52s is very, very")
00:50:46preglowwhere's it broken?
00:50:57preglowthen not me
00:51:13linuxstbSo far...
00:53:41preglowcan anyone tell me why pcm_recording.c is in cvs still?
00:53:45preglowit's not bloody used
00:53:52preglowthrew me off my trail, it did
00:55:37 Join midkay [0] (n=midkay@rockbox/developer/midkay)
00:56:58 Quit xen ("( :: NoNameScript 4.02 :: )")
00:57:50 Join mphill [0] (
00:58:30mikearthurlinuxstb: is there a known iriver hanging bug atm?
00:59:41 Quit Zagor ("Client exiting")
01:00:05linuxstbmikearthur: Don't ask me...
01:00:15 Join piroko [0] (
01:00:29 Join [TCK] [0] (
01:00:39mikearthuryou just normally know everything :D
01:00:42pirokois anyone aware of the lock-up-after-an-hour-or-so bug?
01:00:57mikearthurpiroko: me
01:01:00preglowcomplete lockup?
01:01:03preglowor just playback lockup?
01:01:09linuxstbpiroko: How has your debugging gone? Do you still get it without cpu scaling?
01:01:12mikearthurcomplete lockup
01:01:21pirokolinuxstb: cpu scaling seems not to be the culprit
01:01:27preglowlinuxstb: there are still cpu scaling issues???
01:01:41linuxstbApparently not... But people keep blaming it for everything :)
01:01:44pirokoif it's plugged in, it'll restart and then freeze at the rockbox logo, and if it's not plugged in, it'll just lock up
01:01:54pirokolinuxstb: that was just me, sorry. someone suggested it to me so I went with it
01:01:56linuxstbOne thing I am aware of is wrapping issues with USEC_TIMER
01:02:06linuxstb... and udelay()
01:02:36preglowthat's a valid concern
01:02:40preglowperhaps we should add wrapping protection
01:02:43pirokothe main thing that made me think it was at least a processor problem was that the lockups only happen after an hour or so of playback
01:02:46linuxstbUSEC_TIMER is definitely unsigned, so it will wrap after (I think) about 73 minutes.
01:02:56pirokothat sounds about right
01:02:59mikearthuraye, for me too
01:03:03preglowthen we should most definitely add wrapping protection
01:03:19pirokothat has to be it
01:03:28pirokoit never crashes before an hour
01:03:33linuxstbI think all we need to do is use the existing TIME_BEFORE() and TIME_AFTER() macros.
01:03:56pirokonow would this affect all players/recorders or just ipods?
01:04:03preglowlinuxstb: 71 minutes, according to my ti-89
01:04:06preglowwhich is NEVER MISTAKEN
01:04:08amiconnJust ipods
01:04:09linuxstbJust ipods - it's an ipod-specific timer.
01:04:21pirokothat would be why it isn't as widely known then, i guess
01:04:42mikearthurmine is an iriver h120
01:04:55pirokoand you're getting lockups too?
01:04:59 Join midkay_ [0] (
01:05:09preglowwe should still fix udelay
01:05:09mikearthurand that time sounds about right
01:05:21 Join solexx_ [0] (
01:05:37pirokohmm... why would it be doing the same thing on the iriver too if udelay is only for the ipods?
01:06:24pirokopreglow: wouldn't a stopwatch have sufficed? ;)
01:07:46 Part larcher
01:10:50pirokoand the room goes silent as many programmers fire up their editors and dive into oceans of code...
01:11:02pirokosorry, had to make it somewhat dramatic :P
01:13:03*piroko wonders if everyone has spontaneously combusted
01:13:52warthawgi never spontaneously combust on the first date
01:14:04pirokoTHE first date?
01:14:41warthawgi think i would go wild on the second date, if i ever have one
01:15:00*piroko has had one girlfriend his entire life
01:15:07pirokobut I am still with her :)
01:15:35mikearthurgood man
01:15:37mikearthurhow long?
01:15:42piroko3 years
01:15:45pirokobest friends too!
01:16:22pirokowe should make a place on to upload our photos so we can all see how ugly we are
01:16:41preglowi just need to take a sufficiently ugly one first, then
01:16:41mikearthurI'm incredibly good looking actually
01:16:50midkay_wiki page!
01:16:53 Join VoltageX [0] (
01:16:56pirokoi agree
01:17:01 Quit midkay (Nick collision from services.)
01:17:04 Nick midkay_ is now known as midkay (
01:17:05pirokoit'd be cool to know what the people look like who we are talking to on a daily basis
01:17:12midkayit actually rather would..
01:17:13warthawgi agree has a picture of mee
01:17:26 Quit solexx (Read error: 110 (Connection timed out))
01:17:55preglowyou can probably find pictures of me acting the fool around the internet
01:17:58warthawgsorry, mike, you are not ugly enough to play this game :)
01:18:05 Join e0f [0] (
01:18:07pirokobut I am!
01:18:13mikearthurif its any consolation
01:18:15warthawgme too!
01:18:18mikearthurI'm ugly on the inside
01:18:24mikearthurI eat babies
01:18:27e0fsup guys :■
01:18:34sharpemedically, we're all ugly on the inside.
01:18:38pirokomy picture:
01:18:47 Quit piroko (Remote closed the connection)
01:19:05mikearthurmyface on myspace
01:19:06 Join piroko [0] (
01:19:08sharpei'm shocked!
01:19:15pirokosharpe: at what?
01:19:39pirokoall of my programs are still spontaneously crashing at the same time...
01:19:56pirokoI too am often shocked at nothing
01:19:58warthawghere is my g/f:
01:20:01midkaypiroko, damn, you look scarily familiar.. :)
01:20:16pirokomidkay: do you have any pictures of you online?
01:20:25pirokomaybe we've met in a past life
01:20:52Bagderpics =>
01:21:07Bagderof at least 8 rockbox devs
01:21:14Bagderuh, 7
01:21:33mikearthurwho is who
01:21:42midkayhm, i have one here, but it's..
01:21:56midkayrather.. creepy.
01:22:07pirokolet's have it!
01:22:12midkaylet me upload it..
01:22:38midkayi wish i had a better one. sigh. if there was a wiki page i might...
01:22:44midkayalso available - a picture of my broken toe!
01:22:54midkayif anybody wants that. which i know they do.
01:23:05preglowin the same kind of closeup? hell yeah!
01:23:09mphilldo you still use itunes to transfer mp3s with rockbox?
01:23:14preglowmphill: no
01:23:18sharpeyou're right, that is rather creepy.
01:23:20pirokomidkay: you don't have any more identifiable pictures?
01:23:24***Saving seen data "./dancer.seen"
01:23:31midkaypiroko, not really.. or.. w8w8w8w8.
01:23:48mphillpreglow: do you drop them right into the FS?
01:23:51speacial_edmphill: you can if you bott the ipod into the original firmware then connect it
01:23:54preglowmphill: yes">,_Rosie_and_Phoebe_s_Party/Me_and_Drew.jpg
01:23:55*Bagder runs off to bed
01:23:58mikearthurthat picture of me pwns
01:24:08preglowmphill: even though you can use itunes as well, if you want, but then you need to use our tag database to find your files
01:24:09mphillpreglow: what directory? :)
01:24:12mikearthurcaptain planet baby
01:24:14preglowmphill: wherever you like
01:24:26pirokomidkay: don't you have a digital camera?
01:24:33midkaypiroko, yes..
01:24:36twisted`this is sooo wrong
01:24:39pirokocaptain planet... he's our hero... gonna take pollution down to zero...
01:24:43mphillok, does rockbox support FLAC?
01:24:49midkaymikearthur, ew ew what the hell ew.
01:24:52twisted`I wanted to find a quote Carmack once made... about game stories are the same as porn stories?
01:24:54midkaycreeeeeped out.
01:24:54mikearthurmidkay: captain planet
01:25:07 Quit ProgramZeta (Read error: 110 (Connection timed out))
01:25:07twisted`'n I looked up at the Google Search thing... 'n it says: Carmack porn
01:25:10twisted`that's so... wrong...
01:25:20preglowmphill: yes
01:25:22 Join whatboutbob [0] (
01:25:32pirokomidkay: have you ever been in illinois or ohio?
01:25:52preglowperhaps someone just posted his picture there
01:25:58midkaypiroko, no. :)
01:26:21pirokomidkay: ah. chances are good we've not met then, but who knows.
01:26:32whatboutbobpreglow: thanks for tackling optical recording
01:26:45preglowwhatboutbob: np
01:26:57preglowwith a little luck it'll be done tomorrow
01:27:18midkaypiroko, i doubt it :) here's one that's not so recent but slightly calmer. :) - must be about .. hm.. half a year old..
01:27:30sharpemidkay, have i told you, i've got keyboard input working well for the emulator?
01:27:39midkaysharpe, you have not, cool :)
01:27:51VoltageXsharpe: on-screen keyboard?
01:27:51pirokomidkay: yeah, no go... :(
01:27:58sharpeand it emulates at a decent speed
01:28:09sharpeVoltageX: nope
01:28:16pirokonot sure if it'll work...
01:28:20midkaypiroko, as expected :) you look like about three different people i know in seperate photos :)
01:28:22whatboutbobamiconn mentioned earlier that 'a gain setting doesn't make sense for s/pdif in' i reading that correctly that we then wouldn't be abel to control the gain for optical recording?
01:28:30VoltageXhow do I plug in a keyboard to my iRiver
01:29:00midkaypiroko, i've never seen such a 'varyingly familiar face'.. :)
01:29:04linuxstbsharpe: Are we going to see a C64 emulator patch soon?
01:29:06 Quit obo (Read error: 110 (Connection timed out))
01:29:12linuxstbAnd did you choose a name?
01:29:21pirokomidkay: lol. my real name is jeremy, if that does anything freaky for you...
01:29:28sharpekind of, but i haven't put it anywhere
01:29:36midkayooooooooooooooooooOoOOOOOhohhhhhh ;)
01:29:45sharpei think it was c64box
01:29:46piroko:O A C64 emulator...?
01:30:12VoltageXI want teh keyboard for iRiver!!!
01:30:53linuxstbsharpe: Have you done any proper speed tests yet? i.e. what % of realtime is it, and how many frames per second can you display?
01:31:43sharpe15fps, seems to be okay
01:31:55sharpei may do a little more tweaking to see how fast i can get it
01:32:07linuxstbThat's on a 5g?
01:32:34linuxstbNot bad at all then.
01:32:55sharpekeyboard input scrolls thru the glyphs fine
01:33:06VoltageXsharpe: where/how do you plug a keyboard in?
01:33:16sharpevoltagex: you don't...
01:33:33sharpebut that's an interesting idea.
01:33:34linuxstbWith quite a lot of effort, I could only make pacbox do 20fps on the 5g. That's emulating a 3MHz Z80.
01:33:44 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
01:33:53VoltageX<VoltageX> sharpe: on-screen keyboard?...<sharpe> VoltageX: nope
01:34:01sharpethis is emulating a ~1MHz 6502/6510... :)
01:34:12linuxstbWell, I didn't write the emulator...
01:34:30sharpewell, if you consider it displaying what key it will enter, then yeah, it's an onscreen keyboard...
01:34:35pirokois this emulator readily available right now?
01:34:38sharpei didn't write this cpu emulator...
01:34:52midkaylinuxstb, what does pacbox currently run at on the 5g, do you know off the top of your head?
01:34:57VoltageXsharpe: steal the onscreen keyboard from RockWord :)
01:35:08linuxstbmidkay: As I just said, it just about manages 20fps.
01:35:13sharpei'll be back in a few minutes
01:35:31midkaylinuxstb, oh, misread, i thought you meant with a lot of optimizations you managed 20, but never finished or committed them..
01:35:31linuxstbThat's displaying 1 in 3 of Pacman's 60fps, so it's running in realtime.
01:35:40lostlogicmikearthur: iriver freezing −− during playback? after some time period randomly, or when you do something in particular
01:35:41pirokoback later
01:35:41 Quit piroko ("telinit 0")
01:35:53linuxstbmidkay: No, everything I've done is committed.
01:35:59midkaylinuxstb, kuhl :)
01:36:07mikearthurlostlogic: I think its just over an hour
01:36:13mikearthurand during playback, without being touched
01:36:27lostlogicmikearthur: then no, nobody knows anything about any such bug
01:36:33lostlogicstarted recently? still happens on latest CVS build?
01:36:45VoltageXdo you think that 25fps video will ever be possible on the H340? (that's assuming iRiver intentionally crippled the official firmware)
01:37:04linuxstbHighly unlikely - unless you want uncompressed video.
01:37:26 Quit YouCeyE (Read error: 104 (Connection reset by peer))
01:37:54VoltageXlinuxstb: faster than 10fps then?
01:38:00whatboutbobpreglow: have you confirmed that 20/96 is possible w/ optical recording?
01:38:12linuxstbVoltageX: Possibly - using a simpler codec.
01:38:25VoltageXhmm ok
01:38:35 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
01:38:41VoltageXthe ogg video format would be interesting
01:38:43linuxstbI don't think anyone thinks iriver did a bad job with their video codec.
01:39:20VoltageXlinuxstb: no, I'm definately not saying they did a bad job, just a little suspicious since they also want to sell the PMP series
01:39:55midkayVoltageX, iriver wouldn't slow it down intentionally, i just bet they didn't want to spend a year on it. "oh, hey, 10fps video runs on the H340. well, let's do some testing and release it.".. instead of "oh, we can run 1280x720 H.264 video @ 60fps, but.. nah. give them less".. :)
01:40:09 Quit YouCeyE (Client Quit)
01:40:33midkayVoltageX, it probably can't handle much more..
01:40:36VoltageXmidkay: lol. I keep thinking that iRiver act like a normal company
01:40:37 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
01:40:54mikearthurlostlogic: been happening the last couple of days
01:40:57linuxstbNo-one has seriously attempted to get video working yet. There are a couple of "first-attempt" ports of video decoders to Rockbox in the patch tracker, without any attempt at optimisation, and they both work at about 1-2fps without audio.
01:41:00mikearthurI update theCVS most days
01:41:03midkayVoltageX, haha. they seem quite generous compared to most..
01:41:18mikearthurlostlogic: file a bug?
01:41:19lostlogicmikearthur: anything particular about your tracks, or the tracks it's playing when it dies?
01:41:22VoltageXbut I think the latest iMP250 firmware was late 2004 year, at least 3 years after the product release
01:41:27amiconnMeh, why does a song get named 'beer bottles'?
01:41:39VoltageXnow, RockBox on iMP250 would be cool.
01:41:41*amiconn goes fetching one
01:41:45mikearthurlostlogic: nope, worked perfectly fine before, seems time-related
01:42:00midkaylinuxstb, huh? i haven't seen any patches like that..
01:42:19lostlogicabout an hour... would be twice around the compressed audio buffer... but twice around is weird... there was a time when once around caused problems...
01:42:25linuxstbmidkay: There's an xvid decoder and mpeg-2 decoder IIRC.
01:42:54linuxstbI'm pretty sure they both have issues on the ipod though - data aborts and/or endian problems.
01:43:03 Quit XavierGr (Read error: 104 (Connection reset by peer))
01:43:14linuxstbThe person who ported them has a H3x0.
01:43:36VoltageXlinuxstb: yay iriver.
01:44:11VoltageXthey haven't even dropped support for the iMP series yet
01:44:23midkaylinuxstb, i see, interesting
01:44:27amiconnlostlogic:An hour is more than twice around the buffer... at least with my usual bitrates
01:44:33 Join solexx [0] (
01:44:47sharpe'kay, i'm back.
01:44:48TeaSeaVoltageX: Yeah. But were more than willing to drop support for the H3xx, their best HD DAP.
01:44:54TeaSeaOr at least, HD DAP with a colour screen.
01:44:55lostlogicamiconn: aye, but it could be twice around at 128kbps
01:45:06amiconnAh, yes
01:45:15*linuxstb hasn't heard of the iMP series...
01:45:21*amiconn is currently test-driving playback on his mini2g
01:45:29 Quit muesli__ ("ich will KŘhe!!!")
01:45:36midkaylinuxstb, MP3-CD players.
01:45:47 Join BHSPitLappy [0] (
01:45:49amiconnIf preglow is right, 31 minutes to go
01:45:49 Join Paul_The_Nerd [0] (
01:46:07lostlogicpreglow: I'm now going to make changes to the behavior of track skipping. haven't decided what changes yet though. Something to do with audio stopping immediately on press.
01:46:35VoltageXTeaSea: it still appears on the site, you can still download firmware, you can stil get support for it.
01:46:47TeaSeaVoltageX: Well meh, they've PRACTICALLY dropped it.
01:46:58sharpewith respect to current_tick, HZ is about one second, isn't it?
01:47:12lostlogicsharpe: should be exactly one second
01:47:14linuxstbHZ should be exactly one second.
01:47:21lostlogicas that's what HZ means
01:47:51lostlogicor the unit "1/s"
01:47:54sharpepirate rockbox!
01:48:20 Quit [TCK] (Read error: 104 (Connection reset by peer))
01:48:24midkayhertz (h├╗rts) n., pl. hertz. (Abbr. Hz) A unit of frequency equal to one cycle per second.
01:48:28midkayinteresting. :)
01:48:51VoltageXAVAST ye! kthxbye.
01:48:54jumpfroggyyeah. 1ghz processor = 1 billion cycles per second
01:48:57sharpeshouldn't it be one cycle per second?
01:50:23sharpeArr, matey. Thar be plugins off the port bow, and codecs off the stern. Fasten down the hatches, I see ipodlinux comming...
01:52:19VoltageXData abort! Arrrr we be sinkin'
01:52:46 Join ashridah [0] (
01:52:48amiconnWhat are you sinking about? ;)
01:53:01VoltageXamiconn: you've seen that video too?
01:53:14lostlogicamiconn: ahahahahahahaha
01:53:18amiconnThat was a self-runner at devcon...
01:53:59amiconnHmm, perhaps not the right word...
01:54:21preglownever heard it before, at least :P
01:54:24VoltageXsomething that was looped over and over on a display?
01:54:26preglowstaple ? :P
01:54:48amiconnrunning gag...
01:54:55sharpea neccessity?
01:55:51preglowi've seen it
01:56:18sharpeonly twelve more hours and i'll have the project gutenberg dvd downloaded.
01:56:26preglowwhatboutbob: no, i haven't confirmed that
01:56:31preglownor will i probably be able to, but we'll see
01:56:42VoltageXsharpe: every gutenberg text??!
01:56:51lostlogicwould this be considered odd coding?
01:57:05*twisted` checks
01:57:06 Join jdahlin [0] (
01:57:36preglowdo while 0 ???
01:57:46sharpeno, just a few thousand of them i think.
01:57:46sharpe"the best of"
01:57:47lostlogicI'll take that for a yes...
01:57:54lostlogicpreglow: how should I write that?
01:58:11twisted`I got no idea what ur tryin there
01:58:11sharpe/* do not do this */ ?
01:58:29 Quit solexx_ (Read error: 110 (Connection timed out))
01:58:31VoltageXwhile false?
01:58:33Paul_The_NerdDo this once. Maybe.
01:58:41preglowisn't do { } while(0) equivalent to just dropping the entire construct?
01:58:47*twisted` should probably start coding again
01:58:55twisted`preglow: I thought so too
01:58:59 Quit speacial_ed ("Bye")
01:58:59preglowit gets done once, then never agai
01:59:00lostlogicpreglow: apparently, continue didn't work. best make it a while(1) break instead.
01:59:09sharpedo this, only if the laws of nature are violated, and all false statements are true.
01:59:11 Quit lostnihilist ("Leaving")
01:59:21preglowthat is nasty
01:59:23amiconnlostlogic: Are these queues public or private?
01:59:27VoltageXsharpe: or if it's opposite day.
01:59:38twisted`I just went from pasting bash quotes to a hot chick I know
01:59:41lostlogicamiconn: this particular one private.
01:59:43twisted`to... geeky code talk on irc
01:59:47*twisted` is doing something wrong
02:00:02twisted`lostlogic: I blame YOU
02:00:22*amiconn wonders what that code snipped should achieve
02:00:35jdahlinare there any video game emulators ported to rockbox?
02:00:37amiconnsnippet even
02:00:58preglowamiconn: you know why pcm_recording.c is still in cvs?
02:00:59preglowit's dead code
02:01:03VoltageXjdahlin: there's doom, gameboy, chip8
02:01:07Paul_The_Nerdjdahlin: Gameboy / GBC, and Pacman (the original Arcade game), and Chip8
02:01:07lostlogicamiconn: just posted a fixed version
02:01:15jdahlinWould be nice to have NES
02:01:20amiconnpreglow: Why should I know?
02:01:20*Paul_The_Nerd wouldn't consider "Doom" an Emulator.
02:01:20*twisted` should read the rockbox manual
02:01:28VoltageXjdahlin: get coding :)
02:01:29amiconnIf it's dead, experimental code, it should go away
02:01:30lostlogictwisted`: you and me both.
02:01:40jdahlinVoltageX: I'm considering
02:01:43VoltageXPaul_The_Nerd: ok then, it's an engine
02:02:01jdahlinVoltageX: however, the buttons on my X5 are not placed well
02:02:04Paul_The_NerdVoltageX: Not even an engine. The Doom code was released under GPL, so it's really a genuine source port.
02:02:04twisted`lostlogic: howso?
02:02:16lostlogictwisted`: I've not read it, I hear it's quite a nice work, and I'd probably learn some things.
02:02:19VoltageXjdahlin: although if you understand emulators, a few optimisations on RockBoy would be appreciated
02:02:21midkayanyone know how it's decided which plugins run on which platform for the manual?
02:02:38jdahlinVoltageX: optimisations such as?
02:02:38twisted`lostlogic: I just gotta figure out howto get back from the WPS to the section where I can pick... DIFFERENT music...
02:02:46VoltageXjdahlin: speed
02:02:52preglowamiconn: i've no particular opinion about that
02:02:55lostlogictwisted`: ipod? just hit select
02:02:56preglowjust wondering if you do
02:03:02jdahlinVoltageX: right, but where and how?
02:03:10Paul_The_NerdVoltageX: Many of the optimizations it needs are going to require knowledge of CPU-specific assembly as well, though.
02:03:13amiconnI know next to nothing about swcodec recording
02:03:21amiconnI never used it yet
02:03:23jdahlinwhat's the current bottlenecks?
02:03:25VoltageXPaul_The_Nerd: how did I know this was gonna be evil
02:03:26*lostlogic knows nothing about swcodec playback.
02:03:44VoltageXjdahlin: sound doesn't run at 100%, gbc games barely
02:03:54sharpepatrick from spongebob squarepants is left handed.
02:03:59twisted`lostlogic: like the middle button?!
02:04:01twisted`lostlogic: OMG
02:04:05*twisted` tries
02:04:24twisted`I totally dig rockbox' WPS vs. original firmware
02:04:26twisted`I love dark-geek
02:04:36linuxstbjdahlin: Which device are you running Rockbox on?
02:04:49*VoltageX yells at the VisualBoyAdvance developers to get the hell over here
02:04:56jdahlinlinuxstb: X5
02:05:16linuxstbI don't think rockboy works at all on the X5 yet. The LCD code probably needs adapting.
02:05:44twisted`how do I clear the current playlist?
02:06:24lostlogicwtf. how'd my ipod's partition table get stupid?
02:06:57sharpeit got down syndrome?
02:07:13linuxstbpreglow: If a file isn't currently being used, I would say "cvs remove" it. It can always be retrieved from CVS if it's needed in the future.
02:07:18twisted`the hell
02:07:22twisted`it won't play what I pick!
02:07:30twisted`it just continues... to play what I played before I rebooted...
02:07:58preglowlinuxstb: yeah, i'll see about it
02:08:08preglowdoesn't look like it's even linked in right now
02:08:13preglowor referenced any places
02:08:55twisted`I read that I'm able to make screenshots, how can I do that? (on the ipod, duh)
02:09:26lostlogictwisted`: you are hitting play on a file instead of hitting select
02:09:30lostlogicyou do need to read the manual.
02:09:53twisted`cus I don't want to enter the artists album screen... I just want to PLAY everything they did
02:10:24midkaygod, how confusing - one more try, does anyone know how plugins are defined to work on certain platforms in the manual?
02:11:30Paul_The_Nerdtwisted`: You can modify the tagnavi.config and create a few new categories, like "Genre All" genres : songs, or "Artist All" artist : songs, that then just show all songs from a given artist, instead of breaking it down by Album
02:12:22twisted`Paul_The_Nerd: u are a nerd indeed
02:12:31 Quit VoltageX ()
02:12:56Paul_The_Nerdtwisted`: I take it that would solve your problems. :)
02:13:20sharpei'm guessing %f isn't supported as a format specifier in snprintf...
02:13:52*preglow bed
02:14:07linuxstbPaul_The_Nerd: Is tagnavi.config documented anywhere?
02:14:38twisted`Paul_The_Nerd: it was a compliment yeah
02:14:47linuxstbsharpe: No - Check firmware/common/sprintf.c
02:15:16sharpeshall use %d then.
02:15:18linuxstbmidkay: You're asking how the LaTeX source works?
02:15:40Paul_The_Nerdlinuxstb: Not really, unfortunately. I've only tinkered very slightly myself, but the two "Example" lines in it seem to show the extent of its features so far.
02:15:41midkaylinuxstb, i figured it out, i kept looking for a plugin-related main.tex and found one finally..
02:16:39*lostlogic is dumb, but at least I didn't commit the dumbness before realizing it this time.
02:19:27 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
02:19:31 Join Spiker611 [0] (
02:19:42Spiker611hello i have a stupid question about rockbox
02:19:50amiconnHmm. No lockup after 71 minutes....
02:20:15Spiker611how do i install patches on my ipod via windows?
02:21:11Paul_The_NerdSpiker611: Patches are actually applied to the source code. You'll need to set up a development environment of some sort
02:21:31Paul_The_NerdSpiker611: For ease of setup and use, I would recommend
02:21:42Spiker611oh ok thank you :)
02:22:09 Quit Spiker611 (Client Quit)
02:25:03 Quit Febs ("CGI:IRC (Ping timeout)")
02:25:40jumpfroggyWhat's the benefit of using VMware vs. setting up cygwin?
02:25:53jumpfroggyEasier for the newbie? Safer (sandboxed)?
02:26:14lostlogicjumpfroggy: yes, yes, and faster.
02:27:21 Quit jumpfroggy ("CGI:IRC (EOF)")
02:27:46 Join jumpfroggy [0] (
02:27:58midkayjumpfroggy, you can compile in like 1/3 or 1/4th the time..
02:27:59lostlogicyes, vmware seems to compile faster than cygwin
02:28:26sharpei just use cygwin...
02:28:32midkaysharpe, looooser :)
02:28:37lostlogicI ... jus tuse linux...?
02:28:50scottderdoes EQ effect the line-out?
02:28:50midkaylostlogic, even bigger looooser ;)
02:28:59lostlogicscottder: yes
02:29:01sharpei just use... windows?
02:29:10lostlogicscottder: it's digitally applied prior to sending to audio out to the DAC
02:29:17lostlogicsharpe: get a real OS :)
02:29:32sharpeyou're right. i'll program my own
02:29:35scottderlostlogic: ok, thanks...I wasn't 100% sure
02:29:44sharpein assembly
02:29:47sharpewith a magnet
02:29:48lostlogicsharpe: :)
02:29:56scottderguess I shouldn't mess with differnt tips on my IEMs and EQ at the same time :)
02:30:06sharpeand a microscope with a fine pair of tweasers :)
02:30:18sharpesomething doesn't seem to be right...
02:30:26*lostlogic needs a pizza.
02:30:57*dj-fu smashes sharpe in the face with a openbsd cd
02:32:01lostlogic*giggle* my last OS install I didn't even use a CD... just netbooted into a rescue image :)
02:32:58sharpenegative instructions per second, probably isn't too good either.
02:33:05sharpeneither is negative frames per second
02:33:09lostlogicsharpe: :(
02:33:17sharpeor "-(fps"
02:34:22jumpfroggymid & lost: I never would have thought vmware would have been faster, what with emulation and all. Why is that?
02:34:36 Join solexx_ [0] (
02:34:37amiconnmidkay: Don't exaggerate
02:34:58lostlogicmidkay: because it's very good very low level emulation vs. running a second set of libraries on top of windows... it just ends up being faster
02:35:24jumpfroggytrue, I guess it's emulated either way.
02:35:46midkayamiconn, when did i exaggerate?
02:35:51jumpfroggyHey, while we're on the subject of vmware emulation, anyone know where the iPod 3G status is at currently? (yeah, nice segue)
02:36:04midkayamiconn, went from 4-5mins for me to .. what was it.. 1m30s?.. seems like about a 3x improvement for everyone..
02:36:09amiconnmidkay: [02:28:10] <midkay> jumpfroggy, you can compile in like 1/3 or 1/4th the time..
02:36:19midkayamiconn, apologies if your PC isn
02:36:20amiconnYes, 3x is likely, but 1/4...
02:36:22midkayisn't as fast as mine :)
02:36:37amiconn7min->2.5min for me
02:36:41amiconnThis is a laptop
02:36:43midkaysry. 1/3.62
02:36:49 Quit dj-fu ("Lost terminal")
02:36:52Paul_The_Nerdjumpfroggy: It boots, audio doesn't work, we don't have a 3G dev to work on fixing that and it's really not something that can be done very well blind.
02:37:19amiconnI get 1 min 10s on my desktop pc (running linux natively)
02:37:34lostlogicreal 0m41.924s
02:37:35amiconn(w/o ccache)
02:37:36lostlogicipod video
02:37:38 Part jdahlin ("Leaving")
02:38:11sharpefor some reason, i just don't think it's emulating the c64 at over one billion instructions per second.
02:38:16lostlogictoo bad this machine's not on a good enough pipe to help as a CVS build server
02:38:19sharpei'll have to fix that
02:38:28lostlogicsharpe: yeah, get up to over 1 billion instructions
02:38:32lostlogicthat's how to fix it right?
02:38:40lostlogicreal 0m34.978s
02:38:52lostlogicdamn, ipod is statistically significantly slower to build.
02:38:55midkayamiconn, well, desktop pc != laptop?
02:39:01sharpeso like, super cool the arm cpu, and screw with the fsb and voltage levels
02:39:12lostlogicsharpe: :)
02:39:17sharpeover one billion instructions per second for an emulated c64
02:39:40jumpfroggyPaul: Hmm... I'm going to try it out, just setup cygwin again. I'd love to get into dev work, just don't know if I have time.
02:40:13Paul_The_Nerdjumpfroggy: I would *definitely* suggest our VMWare image over Cygwin. While Cygwin makes it *slightly* easier to directly work with the files, VMWare is an improvement in basically every other area.
02:40:34jumpfroggyreally? I'm used to cygwin, never tried vmware. I'll give it a look.
02:41:07 Join solexx__ [0] (
02:41:44sharpeis it just me, or... should a variable not sporatically change ?
02:43:03lostlogicsharpe: well that depends what you want it to do...
02:43:13lostlogicI mean sometimes if variables weren't cahnging I'd get quite upset
02:43:17 Quit SereR0KR ("XChat Aqua")
02:43:24sharpebut, it's not supposed to change.
02:43:32sharpeit's set to 0, and isn't changed.
02:43:33lostlogicthen it probably shouldn't do that
02:43:46sharpehowever, i get negatives, positives, and "-(" values...
02:44:22Paul_The_NerdWell, if it's not supposed to ever change, why not use either a constant, or make it static?
02:44:39sharpei mean, it's not supposed to, i just haven't used it yet
02:44:49sharpeso it shouldn't change for the time being
02:44:58Paul_The_NerdWell, still, if it's not supposed to, then making it static will cause your compiler to yell at you wherever you ARE changing it.
02:45:07Paul_The_NerdThen you can undo that once you've resolved the oops
02:45:09jumpfroggyBasically, you want it to change only when you change it, not on it's own...
02:45:19lostlogicmake it const, ya mean?
02:45:25Paul_The_NerdThat's right
02:45:28sharpeokay. let me try to re-explain this...
02:45:39sharpei have a variable, set to 0.
02:45:45Paul_The_NerdAnd it should not YET be changing
02:45:51sharpei removed the code that changes it.
02:45:55sharpedue to problems
02:45:58Paul_The_NerdYet it still changes.
02:46:08lostlogicsounds like you have a bug :)
02:46:09sharpedisplaying the value of it, which should be '0' gives random numbers...
02:46:13 Quit solexx__ (Remote closed the connection)
02:46:16 Join solexx__ [0] (
02:46:18Paul_The_NerdTherefor you should make it impossible to change, and see where the compiler complains about you changing it.
02:46:19jumpfroggylostlogic: I agree. Bug.
02:46:21sharpeit could be a random number generator
02:46:25lostlogicwe're so helpful
02:46:36Paul_The_NerdThen simply undo that once you've fixed the offending code.
02:46:40jumpfroggysharpe: it really sounds like an intialization error / deallocation error
02:46:42lostlogiccould also be accidentally writing over its memory area
02:46:51Paul_The_NerdCould be
02:46:53jumpfroggyprint it out right after you set it, make sure it's actually being set to 0
02:47:05jumpfroggythat gives you assurance that it was set to 0 at some point...
02:47:25jumpfroggythen make sure you're not setting it somewher else. Easier if it's a local var, only in 1 function (vs. used by lots of stuff... maybe you forgot to comment something out?)
02:47:53jumpfroggyThen, check if you're accidentally deallocating it, or using it after it's out of scope, something like that. Maybe it's been freed (FREE THE BEAST!) and then you used it afterwards.
02:47:59 Join ProgramZeta [0] (
02:48:02lostlogiccan't free in C
02:48:09*lostlogic hugs C
02:48:10jumpfroggymalloc, etc?
02:48:14 Quit warthawg ("my work here is done")
02:48:14lostlogicno malloc in rockbox
02:48:35jumpfroggyhave you tried the hidden free function?
02:48:48lostlogic::blink blink::
02:48:56*jumpfroggy points to the side
02:48:58jumpfroggyLook over there!
02:49:03*jumpfroggy changes subject.
02:49:11jumpfroggySo how about them 3G ipods? buggy, eh?
02:49:39jumpfroggysharpe: is this var used only in a small place, like within one function?
02:49:57lostlogicjumpfroggy: none of th edevs has one to play wiht
02:50:46 Quit solexx (Read error: 110 (Connection timed out))
02:50:47Paul_The_NerdShort for "Elite Developers"
02:50:52Paul_The_NerdDon't you know _anything_? :-P
02:51:05linuxstbWhat a good idea - port Elite to Rockbox...
02:51:07ashridahah, was half afraid we'd gone down the apple path, and we were turning into iRockbox
02:51:10scottderl337 d3v3l0p3rz
02:51:20 Quit solexx__ (Remote closed the connection)
02:51:37lostlogic*punches something*
02:52:32 Quit sharpe (Read error: 104 (Connection reset by peer))
02:52:44lostlogicugh, voice remains a very sure way to break things on ipod.
02:52:45 Quit solexx_ (Read error: 110 (Connection timed out))
02:53:06linuxstbJust on the ipod?
02:53:08 Join sharpe [0] (
02:53:15sharpei love my router.
02:53:16lostlogicany swcodec target
02:53:40twisted`edevs are people who do E coding :P
02:53:40lostlogicwhich is why I am now going to try to make it suck less, because playback itself is down to minor broken stuff.
02:53:56lostlogicno, it's people who hand out ecoli
02:54:30sharpethat's a bacteria...
02:54:59jumpfroggylost: for which target?
02:55:19lostlogicjumpfroggy: for which target what? voice for swcodec => iriver, ipod
02:55:32scottdermakes me think of the dot com boom
02:55:37scottderput E in fornt of anything
02:55:41lostlogicjumpfroggy: for what target who?
02:55:46scottderand you could get teuckfulls of VC money
02:55:58jumpfroggynotice how tab only remembers the last char you've typed? annoying
02:55:58lostlogicas long as we're on typo picking.
02:56:04Paul_The_NerdWhy is it considered "necessary" for voice to be merged/combined/whatever into the audiostream, instead of just preempting it? When the voice plays, I personally would prefer not to have music in the background making it harder to hear anyway.
02:56:05jumpfroggytype lost, then tab... goes through the T's
02:56:13lostlogicjumpfroggy: you client blows
02:56:15BHSPitLappyjumpfroggy: then you have a ghey client
02:56:29jumpfroggyyeah, web IRC
02:56:30lostlogicPaul_The_Nerd: not my idea
02:56:38BHSPitLappyjumpfroggy: luser!
02:56:50scorchejumpfroggy: CGI:IRC?
02:56:55lostlogicPaul_The_Nerd: maybe I'll make it configurable, although then we'd need a separate pcm buffer of some smallish size for voicing
02:57:01jumpfroggySCO:RCHE, y:es.
02:57:18scorchedo a space, then part of the name, then tab
02:57:22Paul_The_Nerdlostlogic: Well, that's why the message didn't start with lostlogic: :) I dunno, it seems like it'd make the voice easier to hear at least if playback either paused while the voice played, or if the voice clip was 2 seconds, when you came back it was 2 seconds further up the audio thread.
02:57:25jumpfroggyhaven't had a chance to try out rockbox yet, since I have the ill-fated ipod 3g. What is the voice used for?
02:57:40BHSPitLappyjumpfroggy: blind people.
02:57:49jumpfroggyscorche: whoa, that works. WHy?
02:57:57lostlogicPaul_The_Nerd: could also just lwoer the audio further, that'd be easy :-P
02:57:57midkayor users who just want voiced menus?
02:57:59scorchejumpfroggy: ;)
02:57:59linuxstbBHSPitLappy: And other uses - e.g. in a car.
02:58:01jumpfroggyBHSPitLappy : yeah, but what does it say?
02:58:05midkaye.g. on the go or in a pocket..
02:58:17lostlogicjumpfroggy: it reads the menus to you
02:58:18midkaymenu entries, etc..
02:58:20lostlogicor spells file names
02:58:29jumpfroggydoesn't pronounce file names?
02:58:30lostlogictells you wtf you're doing
02:58:35 Join ScootScat [0] (
02:58:37Paul_The_Nerdlostlogic: Well, I was just wondering if it could kill two birds with one stone if you weren't trying to decode two things at once.
02:58:44lostlogicjumpfroggy: no, we don't have a full text2speech engine running on the damned target
02:58:48jumpfroggy"You are browsing by artist"
02:58:54jumpfroggy"You are browsing by name"
02:59:00lostlogicPaul_The_Nerd: actually that's easier than having two pcm buffers to swap, I think... maybe not.
02:59:02BHSPitLappylinuxstb: does it say, "For the love of pete, stop playing with your DAP and watch the effing road!"
02:59:03jumpfroggy"You're in your pajamas and eating day old cake", etc
02:59:08Paul_The_Nerdjumpfroggy: It's all pre-rendered audio, but there's a method to create .talk clips for your filenames so that it CAN read them to you, it just requires use of a PC.
02:59:14 Join midgey34 [0] (
02:59:19jumpfroggyI see, makes sense.
02:59:55BHSPitLappyspeech2text seems far more practical ;)
03:00:13Paul_The_Nerdlostlogic: As I've stated before, since I don't really know the playback code at all I feel I should just offer whatever stupid idea comes to my head and let the experts either shoot it down or see if it's something they've overlooked. :)
03:00:24lostlogicPaul_The_Nerd: understood
03:00:27 Quit TeaSea (Remote closed the connection)
03:00:35lostlogicPaul_The_Nerd: I wonder what blind people would say on the matter
03:00:41sharpe"your life is slowing degrading into an abysmal nonexistance, in which you fade into oblivion. by the way, you are browing by artists."
03:00:41lostlogic*pings the mailing list*
03:00:41Paul_The_NerdBHSPitLappy: IBM is supposed to be making a DAP that does that. Record seminars, then have the ability to do a text search on that recording.
03:00:42BHSPitLappythey've been here
03:00:50linuxstbBut presumably voice pre-empts music on the hwcodec platforms?
03:01:07 Join TeaSea [0] (
03:01:14BHSPitLappyPaul_The_Nerd: totally not what I meant, but okay
03:01:21BHSPitLappyI meant voice commands
03:01:26 Join solexx [0] (
03:01:49Paul_The_NerdBHSPitLappy: Well, that's still qualifying as a mobile device with a 'Text2speech' engine on it. :)
03:01:56Paul_The_NerdErr Speech2text
03:01:58*Paul_The_Nerd cannot type
03:02:03BHSPitLappywhat's "pre-empting" exactly in the audio context
03:02:22jumpfroggysharpe : hah!
03:02:25linuxstbpausing the music in order to play a voice clip.
03:02:40BHSPitLappysharpe: emoPod?
03:02:58jumpfroggyhow about marvin the paranoid androPod
03:02:59amiconnlinuxstb: No, music preempts voice on hwcodec
03:03:04amiconn..i.e. no voice during playback
03:03:16amiconn(and neither during recording)
03:03:24linuxstbAh, so you need to manually pause playback in order to use the menus?
03:03:36amiconnlinuxstb: Stop, not pause
03:03:52BHSPitLappyducking would be nice
03:03:54sharpe, "everyone hates you when you browse your genres."
03:04:15amiconnPause just sample-exact pauses, and that doesn't allow to feed a different bitstream to the mas
03:04:27midkaysharpe, haha.
03:04:35midkaysharpe, little insults thrown in, that's good :)
03:04:57linuxstbamiconn: I'm guessing the voice clips are not permanently in RAM then?
03:05:08sharpe"you are currently at the main menu while everyone is staring at you in class."
03:05:17amiconnThat would be another problem on archos
03:05:24midkay"keep using the artists browser and you'll die young, seriously."
03:05:43sharpethat's not really paranoia inducing...
03:05:45amiconnThe voice file is loaded at the time the first voice clip is needed (when playback is stopped)
03:05:47midkayhey, i got a good idea..
03:05:49sharpemore like fear inducing
03:05:59 Join earHertz [0] (
03:06:01midkaylet's .. discontinue rockbox for the archos models.
03:06:05amiconnIt loads the whole file at once on the HD-based archoses, and only the header on Ondio
03:06:15amiconn...then loads the clips on demand
03:06:32earHertzmidkay: does rockbox have a projedt cloak for IRC?
03:06:36amiconnThe latter is because reading the MMC is slow on Ondio, but otoh there's no spinup time
03:07:09amiconnReading the whole voice file at one on Ondio would take ~5 seconds
03:07:12scorchemidkay: why?
03:07:18midkayearHertz, whois would indicate yes. :)
03:07:21midkayscorche, why what?
03:07:32midkayearHertz, if i think i know what you're talking about..
03:07:58earHertzmidkay: how can I get one?
03:08:09 Join omega21 [0] (
03:08:18lostlogicamiconn: so opinions about the mixing feature of swcodec? Is it better subjectively to mix or to interrupt?
03:08:22twisted`this is seriously... the first channel... I've been in that actually... discusses the project the IRC channel is for...
03:08:37earHertztwisted`: this is teh gerbal channel
03:08:49amiconntwisted`: That's what the channel is meant for...
03:09:05 Join ravon_ [0] (
03:09:09midkayearHertz, technically i think you're supposed to be a 'dev' e.g. have cvs commit access, but you can probably snag one anyways, not sure ;) there was a post on the rockbox-dev list like a month ago..
03:09:43earHertzmidkay: damnit, I've been torturing myself reading theru platlist.c
03:09:47 Join VoltageX [0] (
03:09:48amiconnlostlogic: I'm not sure. Personally I don't use the menus during playback, but then I'm not blind, and mainly use directory clips for browsing in the car, and not the menus
03:10:00omega21sorry to interrupt, i just have a quick question about the iPod port
03:10:07linuxstblostlogic: Would it simplify things if we interrupted playback?
03:10:09midkayearHertz, ah, sympathy ensues ;)
03:10:22ashridahomega21: fire away
03:10:24lostlogiclinuxstb: would be a different can of worms, probably a little bit simpler, but take a bit more RAM
03:10:33omega21does anything in the install procedure delete the songs on my iPod?
03:10:50lostlogic(probably need 512kB or so)
03:10:51amiconnI *think* voice-over is better than interrupting
03:10:54ashridahomega21: only if you have to reformat it with fat32, from my vague understanding
03:10:55linuxstbIt would also make browsing menus very odd - a voice clip, a tiny bit of music, then another voice clip...
03:10:56omega21I just need to know if I should do a backup before installing
03:11:04lostlogichmm, indeed
03:11:07omega21its aleady fat32
03:11:07earHertzmidkay: I write this whole generioc starte machine (I posted it on teh list) only to find that while it did what I wanrted, it didn't so it where I needed it
03:11:18ashridahbut backups aren't a bad idea, the restore procedure nukes the entire thing if you go back to standard ipod, doesn't it?
03:11:19earHertzstate machine, rather
03:11:24lostlogicwell I'll work with the assumption that the functionality as coded is the desired functionality unless there is an outcry to change it then.
03:11:36Paul_The_Nerdomega21: As long as you follow the instructions as they are typed, and keep a backup of the original bootpartition.bin that you extract, no further backup *should* be necessary
03:11:58Paul_The_Nerdomega21: That being said, if you ever feel the need to ask "Should I make a backup" it's probably a good idea to do it just in case.
03:12:01midkayearHertz, a _what_? where? :)
03:12:13omega21and rockbox can deal with the strange iTunes file structure?
03:12:24twisted`earHertz: gerbal?
03:12:26VoltageXomega21: no
03:12:30twisted`earHertz: those things that run round in cages 'n shit?
03:12:31ashridahomega21: if you do make one and don't need it, you waste time. if you don't make it and need it, you're screwed. :)
03:12:31linuxstbomega21: yes
03:12:34Paul_The_Nerdomega21: We have something called "TagCache" that can generate a database out of it.
03:12:41amiconnVoltageX: It can, using the tagcache
03:12:55VoltageXamiconn: but it can't read the itunes DB
03:13:05earHertzmidkay okay, imagine a simple pattern-matcheer. like a regex. It looks for the letetr "a" followed by "b" or "c"
03:13:06linuxstbBut it can deal with the files.
03:13:08amiconnNo, but then it doesn't need to
03:13:16omega21i think i get it. :)
03:13:20VoltageXis the tagcache format described in the wiki?
03:13:22omega21ill give it a shot.
03:13:25midkayearHertz, yes... :)
03:13:41earHertzasa a stream of inopurt is examione, the state machine TRANSITIONS from one state to another
03:13:42omega21also, (in case you were wondering) I called Apple and installing rockbox doesn't void any warranties.
03:13:59earHertzif it has read "a", then it's in a state where it wants to read a "b" or a "c"
03:14:10midkayearHertz, got it..
03:14:12linuxstbomega21: Did they know what Rockbox is?
03:14:24omega21i told them it was an open source firmware replacement
03:14:28earHertzif it reads something else, it transfers back to tteh initial state, where it is waiting for an "a"
03:14:43midkayearHertz, mmhmm..
03:14:47linuxstbomega21: Where in the world are you?
03:15:04midkayomega21, well, that'd be really impressive if they seriously meant that, but it sounds like they may have been confused/whatever :)
03:15:17amiconnlostlogic: I wonder whether it would be better (for voice) and simpler to never stop pcm
03:15:26earHertzSo I wrote a generic state machine to do patetrn amtching on striongs. Generic in that a programmer can modify teh pattern just by passing in his own data structure
03:15:41earHertzit's like a simple regex
03:15:44lostlogicamiconn: major battery waste?
03:15:52omega21well, i should get going now. thanks for the help!!
03:15:52amiconn...just mixing/feeding digital zero if there's no real data
03:15:57linuxstbmidkay: I don't know if it's the same thing, but Apple have twice repaired my iBook under warranty when I gave it to them with Linux installed. So they seem tolerant of that sort of thing.
03:16:15amiconnlostlogic: Do you think it would waste battery?
03:16:17 Quit omega21 ("Leaving")
03:16:28midkaylinuxstb, that's pretty damn cool, they seem rather laid-back about warranties/replacements..
03:16:39amiconnpcm uses dma, and should work even at the lowest cpu clock
03:16:44midkayearHertz, cool :) did you say you posted it to the list? where?
03:16:47amiconnWould be interesting to measure
03:16:55Paul_The_Nerdlinuxstb: My view on the matter is "Either you can remove rockbox to the point where it'd be too much hassle for them to track that it was installed, or your drive is so screwed up they probably won't bother to go through the effort to find what's on it anyway"
03:16:58earHertzum, somewhere in march
03:17:04linuxstbamiconn: There is no DMA on the ipod - we have the FIQ handler running 5000 times a second.
03:18:18amiconnlinuxstb: I though we do use dma, unlike ipl??
03:18:35linuxstbNo. Neither project uses DMA.
03:18:54amiconnlostlogic: Another idea - let pcm run as long as there's at least one non-zero pcm block
03:18:55earHertzmidkay: it's written to be small on code, not understandable. ;(
03:19:11midkayearHertz, well, it looks very complicated, but sounds quite cool :)
03:19:27amiconn...pause if all blocks are zero, and unpause at the first non-zero block
03:19:50amiconnWe'll probably need a block type identifier later anyway
03:19:57earHertzit's actually not complicated. It just shows my tendency to write code that implements algorithms in data structures rather than in teh code/ ;)
03:20:01lostlogicamiconn: pausing instead of every stopping might be interesting.
03:20:07sharpeokay, it seems to be emulating at ~ 110000 instructions per second...
03:20:11sharpeat 11 fps
03:20:16midkayearHertz, haha.
03:20:52earHertzmidkay: what I did with teh ID3 tags too. It also shows that I'm object orreiented even in C. ;)
03:21:06lostlogicamiconn: but we don't want to turn i ton as soon as we have data, as that would defeat the purpose of having a buffer
03:21:07earHertzI keep writing C++ code in C. ;)
03:21:15midkayearHertz, um, are you submitting me your resume over IRC? :)
03:21:48earHertzmidkay: somebody HERE us hiring???
03:21:53earHertzer, is
03:22:02amiconnHmm. I thought it would just start playing the zero-blocks. Maybe the delay would be too long
03:22:37midkayearHertz, you can work for me if you want. i won't pay you except in a load of "thanks", and you just do whatever i want whenever i want. sort of like... hm. charity.
03:22:38 Join jnc [0] (
03:22:50amiconnIt wouldn't defeat the purpose of a buffer this way
03:22:51sharpeor slavery
03:23:06earHertzmidkay: Ok, Jeff Davis
03:23:06midkaysharpe, shhhhhhh!aA@#O%.
03:23:16amiconnMaybe the pause idea is better, although it still doesn't solve the voice-during-pause problem
03:23:23lostlogicamiconn: ahhh... I see what you are saying −− let the DMA read past the insertion point, so you get kinda a jitter effect when it runs low instead of a long pause
03:23:26***Saving seen data "./dancer.seen"
03:23:45midkayearHertz, i'll be whoever you want me to be.
03:23:49amiconnNo that's not what I mean
03:24:01earHertzmidkay: ewwww
03:24:15earHertzthis ain't the pickup channel ;)
03:24:26midkayearHertz, WHAT. you're mentally .. perverted.
03:24:35amiconnlostlogic: Does 'pause' always pause on a pcm block boundary?
03:24:53earHertzmidkay: Yeah, I acidentally wandered into a furry chat room once. Scared me for life
03:24:54lostlogicamiconn: no, it pauses the DMA right were it is.
03:25:09midkayearHertz, haha. ah. forgiven.
03:25:24lostlogicamiconn: changing to pause at the next boundary might be a way to do voice during pause
03:25:29amiconnThe only solution I can see for voice during pause is then to use extra pcm buffers
03:25:47sharpeokay, it seems to be about 30% real-time emulation of the c64.
03:25:48earHertzSo the tagcache now works asa db?
03:25:52amiconn...but it would cause a glitch when unpausing while the voice is babbling
03:25:58lostlogicI'm going to be working on the playback.c part of voice to help it correctly play and stop clips... then it'll be on to the buffer side of it.
03:26:06Paul_The_NerdearHertz: It works rather decently, yes. :)
03:26:23 Quit ravon (Read error: 113 (No route to host))
03:26:47amiconnI wonder whether preglow would scream if I suggest mixing audio & voice at the 16bit pcm level
03:26:59sharpehe'd shriek.
03:27:03earHertzPaul_The_Nerd: I did some research into using Garsia_Wachs encoding of teh string data, to see if I could make it smaller. But it would only save about 30% of teh data space.
03:27:04amiconnI mean, voice isn't exactly hifi stuff
03:27:04sharpenot scream.
03:28:32Paul_The_NerdearHertz: Well, considering how large the storage capacity tends to be on most players, the primary concern of our TagCache (in my opinion) should first be speed in searching, then speed in generation, then probably space (especially when loaded in ram). But these are simply my views on it as a user. I'm just an observer here, for the most part.
03:28:40sharpedoes anyone think i shouldn't have an option for the number of frames per second allowable?
03:28:52earHertzPaul_The_Nerd: I'm just an observer too
03:29:02amiconnPaul_The_Nerd: Plus code size
03:29:26earHertzPaul_The_Nerd: the nice thing about garsia-wchs is that it reatins teh albateical order of the source, but comprresses like a huffman encoding
03:29:47Paul_The_Nerdamiconn: Oh right, since it's on disk, it's usable on archos. So yes, *DEFINITELY* code size.
03:30:15amiconnearHertz: If it's as demanding as huffman, forget it
03:30:36amiconnIt would slow down tagcache like hell
03:30:50 Quit powr-toc ("Leaving")
03:31:32earHertzamiconn: huffman decoding doesn't need to be demanding, if you can sacrifice space for lookup tables. But that's the rub: on my dataset (all my sonmg id3 info), teh g-w only gives 30% savings. The huffmna lookup would eat into that.
03:32:13earHertzsince g-w retains albetical ordering, you'd only decode on display to the human user
03:32:26amiconnHuffman *is* demanding on our targets
03:32:33amiconn...even with lokup table
03:32:43amiconnDid you try the jpeg viewer?
03:32:45earHertzamiconn: for the mp3 encoding, right?
03:33:19earHertzthe mp3 huffman uses somethinglike 40 diffeeent huffman encodings
03:33:33 Quit gursikh ()
03:33:33amiconnOn coldfire & arm it's bearable, but still ~30% of the whole decoding time
03:33:40lostlogicamiconn: we do
03:33:45lostlogicamiconn: mix at 16bit
03:33:52amiconnOn SH1, huffman is pathetic (no shift-by-n instruction)
03:34:16earHertzamiconn: OH!
03:34:16sharpeokay i've a correct calculation of ips/fps for the c64 emulator now.
03:34:29VoltageXI think I've messed up here, where do I put MSMike to get him to speak for me?
03:35:33amiconnearHertz: Don't forget that our target cpus are by orders of magnitude slower than a desktop cpu
03:36:30*ashridah looks at his soon to be athlon64/x2 system
03:36:33ashridahseveral orders
03:36:35*ashridah nods
03:36:35amiconnA rough comparison would be archos SH1 ~ early '386, and iriver coldfire and ipod ~ '486DX2 or DX4
03:37:04amiconnWithout '387 / '487 of course
03:37:11Galoisdual core 75mhz arm ~ DX4 ??
03:37:20amiconnInsofar 486DX2/4 is wrong, should be SX2/4
03:37:58earHertzamiconn: well, I've got an Ipod, which is as fast as some desktops I had, years ago
03:38:38amiconnGalois: *rough* comparison
03:39:09amiconnIt's definitely not much more. I'm not couting the second core, since we don't use it (yet)
03:39:17Galoisall of a sudden video looks out of reach
03:41:31 Quit TeaSea (Read error: 104 (Connection reset by peer))
03:41:54*lostlogic remembers his dx2 66 couldn't decode 128kbps mp3 without skips unless everything else was shutdown
03:42:03 Join TeaSea [0] (
03:42:10lostlogicon good old winplay
03:42:10amiconnWell, the dsp-like instructions might be somewhat more efficient, but they have limited scope
03:42:42amiconnGood for decoding audio or video, but no use for general things
03:43:06amiconnlostlogic: Really?
03:43:17lostlogicamiconn: that's how I remember it
03:43:56amiconnMy Amiga (68060 @ 50MHz) can decode mp3 realtime at 50% CPU load (320kbps) ...
03:44:24amiconnOkay, that's superscalar...
03:44:39amiconn...maybe an early pentium would be a better comparison
03:44:54amiconn~ P100
03:45:17sharpeanyone know a ips/mips rating for the 6502 or 6510 processor?
03:45:29lostlogicnight amiconn
03:47:42 Quit DewDude (Read error: 110 (Connection timed out))
03:50:06 Join cellardoor [0] (
03:50:34 Join daysaway [0] (
03:50:35 Quit cellardoor (Client Quit)
03:50:58daysawaystupid mIRC won't work
03:51:20daysawaycan anyone help me out with rock box and playing video files on an h120?
03:51:46Paul_The_NerdRockbox doesn't officially play video on the H120
03:52:09daysawaybut some crazy kids managed to get it
03:52:22twisted`if I set Tag Cache to Ram, it won't delete it when I shut down right...
03:52:23daysawayor shouldn't i be talking about this in here
03:52:36Paul_The_NerdAs far as I know all that've been done has been to make the RVF playing plugin compile for H120
03:52:52daysawayi've got the video to work
03:52:56Paul_The_Nerdtwisted`: Yeah, that just means it keeps it in memory, like dircache, so you don't have to spin up the disk to browse.
03:52:58daysawaywith a firmware someone made
03:53:15daysawaybut its only on a small portion of the screen
03:53:19 Quit VoltageX ()
03:53:41Paul_The_Nerddaysaway: That's about all you're gonna get unless you do some more drastic programming. The plugin was built for a very specific sized screen.
03:54:15daysawayThere's a video online of some guy playing shrek full screen on an h120
03:54:38daysawayI was just wondering if anyone in here had made that specific firmware
03:55:06Paul_The_Nerddaysaway: Well, most people in here would've actually put it as a patch on the tracker if they adapted a feature to work on other targets as well. :)
03:55:08 Quit TeaSea (Remote closed the connection)
03:55:40sharpelets hope the data sheets give a ips rating...
03:55:45 Join TeaSea [0] (
03:56:10daysawayThe forums didn't help too much so I was just wondering.
03:58:57Paul_The_NerdIt sounds like this person was more interested in showing off than being helpful. =/
04:00:35sharpei'll just use 300000
04:00:41 Quit daysaway ("CGI:IRC")
04:06:01sharpeat 15fps it's 37% that of realtime...
04:06:09twisted`Paul_The_Nerd: ah k
04:06:10twisted`Paul_The_Nerd: nice
04:07:11sharpe10fps @ 48% realtime...
04:08:10Paul_The_NerdSounds like there's some optimization work to be done then
04:08:20*BHSPitLappy graphs that trend...
04:08:30sharpetis what i was going to do...
04:08:34sharpegraphing it i mean
04:09:33sharpe5fps is 56% realtime...
04:09:54BHSPitLappyso 100% realtime is... -14fps, ouch
04:10:22Paul_The_NerdSomehow I don't quite think it's linear. :-P
04:13:19 Join Aditya|Nap [0] (
04:13:27sharpe1fps : 63% realtime
04:14:00 Join maeck [0] (
04:14:57 Part Aditya|Nap
04:15:17sharpewell, it's slightly linear...
04:15:20BHSPitLappyit isn't a linear graph
04:15:29sharpeno, it isn't
04:15:29BHSPitLappyit would have a horizontal asymptote along the x axis
04:16:02sharpewell, tis more quadratic i suppose...
04:16:11BHSPitLappycould be that too
04:16:11sharpei mean, it give a better approximation
04:16:42BHSPitLappyfrom a positive-values-only standpoint, it's hard to tell the difference between those two possibilities
04:18:16 Quit mikearthur (Connection timed out)
04:18:41sharpeoh noes, my router still doesn't like me
04:19:13*BHSPitLappy is with the router.
04:19:55sharpetime to see what it's at, at 20fps..
04:20:21BHSPitLappywhich values are real?
04:20:24BHSPitLappyand which are projected
04:20:38sharpethat was the real
04:20:57BHSPitLappywere all of them real?
04:20:58sharpeprojected is 25%, tis off by 6%
04:21:24sharpei need to go change the batteries in my calculator...
04:22:07 Quit carini (Remote closed the connection)
04:24:02BHSPitLappyso I project 1.43fps as the realtime mark
04:25:02BHSPitLappyor maybe .49
04:25:10BHSPitLappywe need more data!
04:25:17sharperemember, 1fps was 63%
04:26:55sharpe3fps is 60% realtime
04:27:21BHSPitLappythis one formula looks pretty good
04:27:33BHSPitLappyunfortunately it doesn't have the asymptote like it needs to
04:28:04BHSPitLappyso it says -9.74fps is the key
04:28:47sharpeheh, let me try at .5fps...
04:28:49BHSPitLappysharpe: can you test with decimals
04:28:51BHSPitLappytry .10
04:29:02BHSPitLappyI'm curious about .10
04:29:20sharpewait, atoi doesn't handle decimals, does it?
04:29:26*BHSPitLappy shrugs
04:30:01BHSPitLappy-make- it.
04:30:11sharpei'll just modify the code...
04:30:24sharpeinstead of getting atoi to handle decimals
04:31:05sharpewhich, it shouldn't
04:31:44twisted`maeck: yo
04:31:54sharpe.1 is 64% realtime...
04:32:08maeckhow is the drawing going?
04:32:33twisted`ppfff... went insane on it
04:32:53Paul_The_Nerdsharpe: I was gonna say, just see what the maximum you can get is by disabling it.
04:32:58Paul_The_Nerdsharpe: 0fps
04:33:01twisted`maeck: mainly cause I don't know what kind of style
04:33:31twisted`maeck: cause, well, don't get me wrong, but the things you described... sound so... "standard"
04:33:35BHSPitLappysharpe: doesn't sound like it's doing it right...
04:33:36maeckpick a style. One that is fairly clean and easy (for now).
04:33:50sharpethe upper limit seems to be aronud 65%
04:34:23*twisted` kicks xpdf
04:36:03maeckstandard should work for now. Better gfx can be put in later.
04:37:11 Join lostnihilist [0] (
04:38:02twisted`hmm k
04:38:27sharpei don't think disabling it would get it much past 65%
04:38:31twisted`maybe... I can convice a friend of mine to do the art for it... at least the final art then... she draws sooo exclusively pretty...
04:38:44twisted`it's mind blowing...
04:39:00Paul_The_Nerdsharpe: It was just a suggestion so you could know the upper bound for now.
04:39:03maeckIs that really due to het art or het appearance?
04:39:26maeckhet == her
04:40:06twisted`maeck: her art, it's insane
04:40:23twisted`maeck: tho... would be... new for her tho, she useually draws girls, but I can always ask :P
04:40:31sharpefor icons.
04:40:58sharpeokay... i need to figure out how i'm going to optimize it... :)
04:42:45twisted`maeck: somehow... with the descriptuin you gave, I can't stop thinkin bout 'gnome' style icons
04:42:48twisted`and that's really bad :P
04:43:43twisted`maeck: hoe is 't daar trouwens vanuit nederlandse standpunten gezien?
04:47:15sharpedisabling the blinking gives about a 5% speed increase...
04:49:04 Join ts-x [0] (
04:52:39ts-xPaul_The_Nerd: So I finally set up vmware as you suggested a few weeks back, and am having some issues. I can check out the source, but can't figure out how to patch. Does it create some sort of a virtual drive I can copy/paste to? Sorry, I'm a total noob at this...
04:53:13Paul_The_Nerdts-x: When the VMWare image is running, it should essentially be a virtual PC accessible across the network, as \\Debian
04:53:38Paul_The_Nerdts-x: So, to get .patch or .diff files to it, just browse to \\Debian\User and make a folder, and put them there, or wherever you'd like 'em
04:55:24ts-xI saw that in the forums...I can't seem to find \\Debian. When I enter it, it says 'Network path not found'
04:56:29Paul_The_Nerdts-x: I'm assuming the VM is running while you're doing this. I believe it puts itself in the Workgroup workgroup, so if your computer is elsewhere, you may need to actually do the whole "Browse your Network Neighborhood" and manually track down where the computer is
04:58:28sharpewhat about if i...
04:58:38ts-xIt is running...let me keep looking. Wonder if my firewall is causing issues...
04:58:41sharperender the screen directly into the frame buffer...?
04:59:08Paul_The_Nerdsharpe: I believe that's a technique used elsewhere with decent results.
04:59:35sharpewell, i suppose we know what i'm going to be doing...
04:59:37 Join speacial_ed [0] (
05:00:07 Join solexx_ [0] (
05:00:23 Quit hardeep ("BitchX: shaken, not stirred")
05:00:52 Quit pixelma (" rock on...")
05:01:23 Join Rob2222 [0] (
05:03:27 Join dj-fu [0] (
05:03:41 Quit ts-x ("CGI:IRC (Ping timeout)")
05:05:11 Quit solexx_ (Remote closed the connection)
05:05:13 Join solexx_ [0] (
05:09:50 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
05:10:18 Quit solexx_ (Remote closed the connection)
05:10:25 Join solexx_ [0] (
05:10:55 Join ts-x [0] (
05:13:19ts-xPaul_The_Nerd: It was my firewall...ZoneAlarm and vmware apparantly don't play well together
05:14:23Paul_The_Nerdts-x: Ah. Well that makes sense then. :)
05:15:22 Quit TeaSea ("Leaving")
05:19:08sharpethe battle cry of the frame buffer platypus.
05:21:12ts-xPaul_The_Nerd: All right, so I patched and am attempting to make but keep getting 'Warning: File 'common/crc32.c' has modification time in the future ?
05:21:16 Quit solexx (Read error: 110 (Connection timed out))
05:21:46 Join ravve [0] (
05:22:08Paul_The_Nerdts-x: The vmware image sometimes has the clock fall out of sync. You may need to use the "data" and "hwclock" commands to set the clocks properly.
05:22:19maeckMan, I cannot stand those frame buffer platypussies...
05:22:21Paul_The_Nerdts-x: You *probably* need to "su" first
05:22:21 Join Arrogant [0] (
05:22:57maeckFlat snout and all that mumbo jumbo
05:23:19ts-xSo how do I stop it from repeating that message...the vmware equivalent of cntr-break so to speak?
05:23:28Paul_The_Nerdts-x: Ctrl+C
05:23:29***Saving seen data "./dancer.seen"
05:24:13sharpebut, platypuses are cool.
05:25:06sharpeit's the mammal that lays eggs!
05:26:23sharpealso the echidna lays eggs
05:26:56maeckwhat about your frame buffer, should I put the frying pan on the fire?
05:27:00sharpeand is classified as a mammal
05:27:38sharpenot the frame buffer platty!
05:28:05maeckNative to austalia, eh..!?
05:28:44sharpebut with the powah of google i am.
05:29:48maeckThe platypus, not you... you havent been using words like Mate, Sheila and Bloke often enough...
05:30:25sharpewait, try to construct a sentence with those three words...
05:30:53sharpethat sheila mated with that bloke?
05:30:56maeckI told you, that Sheila is a Bloke, Mate!
05:31:09maeckLike mine better
05:31:29sharpethe cosby show is on!
05:31:41maeckWorked with a couple of ausies a while back... Twists your mind
05:32:27sharpeapplying a physical action to a intangible thing.
05:32:51ts-xPaul_The_Nerd: It's working! Thanks for your help. Holy *s* vmware is fast compared to cygwin :)
05:33:50sharpei figure if i can get ~ 90% real time i'll be okay...
05:34:28ts-xcygwin used to take 90+ minutes, vmware took 2!
05:34:56sharpeto compile what?
05:35:12ts-xJust a regular build
05:36:20sharpecygwin takes me around five minutes or so... for a completely new build...
05:36:28 Quit ravon_ (Read error: 113 (No route to host))
05:37:30ts-xYeah I think something was wrong - the install took like four tries. Man, now that I've used vmware, forget cygwin
05:37:43 Part midgey34
05:38:32Paul_The_NerdVMWare is just a lot better overall once you get it working.
05:38:56Paul_The_NerdI also like the fact that I can put the vmware images on a portable drive, and any computer I can put a VMWare Player on is ready for me to work on, with all my resources etc already
05:39:19sharpethink if instead of doing each character sequencially... i should do all the pixels in a row at once?
05:39:28sharpeer, anyone...
05:40:58sharpeso, 8x8 blocks of pixels, or 200 lines of 320 pixels...
05:41:30Paul_The_NerdDepends on how you're rendering it to the screen and everything. I mean, I don't know how your code ends up working out, so I don't know which method would work faster for you.
05:42:11sharpeeh, can you think of an alternative to each pixel at a time?
05:43:03Paul_The_NerdNot really
05:43:15sharpeneither can i...
05:45:03sharpehmm... three levels of for loops, or like, two larger ones?
05:45:40 Join psycho_maniac [0] (
05:49:08 Quit ashridah ("Leaving")
05:53:46 Join shadou [0] (
05:55:56ts-xPaul_The_Nerd: Thanks again for your help
05:55:59ts-xgnite all
05:56:20Paul_The_NerdNo problem
05:56:20 Quit ts-x ("CGI:IRC (EOF)")
05:56:31Incand openbsd
05:56:41sharpewell, i discovered i was setting the foreground color eight times instead of just once...
05:57:49Paul_The_NerdThat's probably not the best idea. :)
05:58:51 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
05:59:04sharpethat's about a 2% in emulation speed.
06:00:41 Quit psycho_maniac ("CGI:IRC (EOF)")
06:00:51sharpewell, it's running at 20/15 frames per second...
06:00:55sharpesomethings wrong with that...
06:01:25BHSPitLappyjust over 1fps, right
06:01:57sharpeno, the cap is 15 frames per second...
06:02:44sharpewell, how about this for an idea...
06:03:06sharpei have a seperate thread to handle the video output...?
06:03:58 Join ashridah [0] (
06:04:03sharpeprobably not the best idea...
06:05:16 Quit shadou ("Leaving")
06:05:20sharpewith a 26 frame per second max, it's doing 50 fps...
06:07:54scottder/nsm/snort2.6/lib/snort_dynamicpreprocessor// undefined symbol: FTPPBounceInit
06:08:00scottderOpps sorry wrong channel
06:08:40sharpeif you want, i can give you a completely unhelpful comment(tm)...
06:08:40BHSPitLappyyou're only in #rockbox, lol
06:08:56BHSPitLappyhe's in some secret cult channel
06:09:04BHSPitLappy(OTHER than rockbox)
06:09:04scottderumm I am in a few channels :)
06:09:12BHSPitLappy(on THIS network)
06:09:22scottder@#snort-gui #rockbox #snort #honeynets
06:09:37scottderSeeee :)
06:09:46BHSPitLappy"−−- [scottder] #RockBox " <−− seee, secret cults!
06:09:50scottderI am only a little crazy...not a lot crazy
06:09:59BHSPitLappyfrom what I can see, you like to snort honey while using rockbox.
06:10:06BHSPitLappyyou CRAZY lunatic.
06:10:07*scottder uses snort to packet sniff what porn sites BHSPitLappy goes to
06:10:46scottderewww...that's just gross dude
06:11:02 Quit Daishi ("Client exiting...")
06:12:01sharpewhat about if i have it based off of frames, instead of instructions?
06:12:02*BHSPitLappy quickly closes out the Windows support windows
06:12:14BHSPitLappyo noes... I'm caught...
06:12:17sharpeyou sick, sick person.
06:13:11sharpethat would mean figuring out how many instructions to do per frame...
06:13:39sharpeto have 300000 instructions per second...
06:14:41sharpethat wouldn't help me at all.
06:16:03 Quit amiconn (Read error: 110 (Connection timed out))
06:16:04 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
06:18:05sharpeyay my incessant ramblings of my problems.
06:21:18sharpehow can i be sure of running a single function 300000 times per second... hmm...
06:21:24maeckme repeatedly hitting head against wall...
06:22:19sharpethat's not good for your brain, you know.
06:24:31maecksucks to be me
06:24:38sharpeeh, me too.
06:24:48sharpeand it's not just with what i'm coding either.
06:26:08 Quit arf-arf4242 (Read error: 110 (Connection timed out))
06:27:33sharpewhat about... running x number of cycles needed per frame... then displaying the frame...
06:28:09maeckJust realized need to change my code to make it work in the menu... Well, it works as the main menu, it is just not possible to return the selected icon asyet...
06:28:20maeckso that will be something for tomorrow then, I guess
06:28:50maeckPretty menu that cannot select is kinda pointless, although you can spin icons around...
06:28:59sharpeeye candy
06:29:01maeckcan do that for hours...
06:29:30sharpeespecially for the people who regularly use amphetamines... it'll be kick ass for them.
06:29:50maeckBTW, another dude inserted plugin code for GFXMENU on the Nano
06:30:24maeckVery theraputic
06:30:31sharpeyou mean, someone else make a plugin for the same thing?
06:30:36sharpenot make
06:30:42sharpewonderful typos.
06:30:58 Quit speacial_ed ("Bye")
06:31:42maeckno, just took my code and shriked the whole stuff down for the wee Nano. Saves me some work
06:32:15maeckand... shriked === shrinked
06:32:31sharpewell, to achieve 10fps at full emulation speed... i need to execute 30000 instructions per frame
06:33:11maeckAre you talking about your C64 emu? A 90mhz ARM isnt capable of running the emu on 25fps?
06:33:31sharpefor right now, it isn't capable.
06:33:33maeckthat 6502 used to run, on what 2mhz?
06:34:03sharpei'm sure i can emulate the cpu by itself at that speed, fully.
06:34:08Paul_The_Nerdmaeck: 75mhz arm. That's where they're set.
06:34:15Paul_The_NerdAssuming that sharpe is telling the cpu to boost, that is.
06:34:51sharpei is tellin tat cpu to boost itself.
06:34:57Paul_The_NerdGood, good.
06:35:17Paul_The_NerdI've seen it be forgotten.
06:35:19maeckPaul: ah, th e90Mhz was a guess, I guess
06:35:25 Quit MusiFreq (Read error: 104 (Connection reset by peer))
06:35:50sharpei keep having random moments of deja vu, and slight short term memory loss.
06:37:43sharpeokay, for 25fps, it needs 12000 instructions executed.
06:37:50maeckWell, I'll put on one more song and go to sleep. I think 'The End' by the Doors will be appropriate
06:37:52sharpeevery frame...
06:38:03 Join john_ [0] (
06:38:21 Nick john_ is now known as billy05 (
06:39:22sharpei wonder if there'd be any preformance gain if i declared everything that i only use in a single function with the register keyword...
06:41:26sharpewhat should i use, a while loop or a for loop?!
06:43:13 Quit quobl ("Leaving")
06:46:20 Quit dj-fu ("leaving")
06:49:06 Join MusiFreq [0] (
06:51:13 Quit billy05 ("Lost terminal")
06:53:19 Quit maeck ("Chatzilla 0.9.65 [Mozilla rv:1.7.12/20060202]")
06:54:52 Quit vmx (Read error: 110 (Connection timed out))
06:54:58 Quit RotAtoR ("zzzzzzzzz")
06:55:15 Join vmx [0] (
06:56:12sharpeheh, i'll also have a minimum emulation rate....
06:59:31 Join ninow [0] (
06:59:42ninowerm i'm having more problems with the rockbox
06:59:48sharpewhat be those?
07:00:01ninowalright well i wiped my ipod clean, and reinstalled the firmware
07:00:04ninowand put the songs back on
07:00:11ninowhowever, it isn't displaying all of the songs
07:00:21sharpeipod firmware or rockbox?
07:00:22ninowI went to the "limits" and changed it to the maximum
07:00:31ninowbut it still isn't showing all of the songs that i added
07:00:57ninowbut when i plug it into usb, and look on my ipod, it shows in my explorer window that all of the songs are there hehe
07:01:21sharpeart thou usingeth thine tagcache?
07:01:24Paul_The_Nerdninow: Did you actually restart the ipod with a proper shutdown after changing the limits?
07:01:39 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:01:44 Part LinusN
07:02:05 Join LinusN [0] (
07:02:21ninowyeah, held menu and select
07:02:26ninowsharpe, i believe i am
07:02:29Paul_The_NerdThat's not a proper shut down
07:02:35Paul_The_NerdThat's a hard-reset
07:02:51ninowk i restarted and now it's showing the songs it wasn't showing before, but it isn't showing the ones that it was
07:02:56Paul_The_NerdIf you check now, are the limits what you set them, or back to the original values?
07:03:02ninowwhat i set them as
07:03:11sharpeif you're using tagcache, try foring an update...
07:03:17Paul_The_NerdYou got through the menus to check them in 4 seconds?
07:03:39ninowi checked before you asked
07:04:20Paul_The_Nerdninow: Okay, and the files were added in disk mode. Are you browsing the disk, or using the Album/Artist/Genre view?
07:04:30ninowtrying that again sharpe
07:04:47 Join arf-arf [0] (
07:04:55 Part Paul_The_Nerd
07:04:56ninowpaul, through album/artist, etc.
07:05:14sharpeheh, it's funny when you see gcc complain about seven unused variables
07:06:19midkayhahahahahahahaha, that is so hilarious, yeah!
07:06:30sharpeout of nowhere.
07:06:36midkayerm, hey.
07:06:41ninowhmm sharpe, i did that, shut it down
07:06:51ninowand now it's showing the other half of files ;p
07:06:54 Join B4gder [0] (
07:07:12sharpetry it again, heh...
07:08:35sharpei always forget to add rb->lcd_update(); ...
07:08:39 Quit aegray (Remote closed the connection)
07:09:15 Join aegray [0] (
07:09:51ninowsharpe: now it shows the other half again
07:09:59ninowand doesnt show the half i saw before
07:10:08sharpethat's kind of funny :)
07:10:17ninowmoreso frustrating
07:11:16sharpei can't remember if it's safe to do so or not, but you can try deleting the tag cache files and the rebuilding them
07:12:00ninowhow would i do that
07:12:22sharpewhen it's connected to your computer, delete the .tcd files in the .rockbox directory...
07:15:50ninowi hope this works
07:15:53ninowotherwise i give up
07:16:41 Join hardeep [0] (i=hardeeps@SDF.LONESTAR.ORG)
07:17:00 Quit earHertz (Success)
07:18:11ninowk it works
07:18:15ninowthank you very much sir
07:18:22sharpeyou're welcome
07:18:26ninowgood night
07:23:31***Saving seen data "./dancer.seen"
07:25:54sharpethe emulation speed is around 70% now
07:26:59B4gderon what target?
07:27:05sharpeipod 5g
07:27:14sharpei feel like i've done something.
07:27:33midkaysharpe, don't worry, you haven't. ;)
07:27:33sharpeit was running at ~30-40% for 15 fps
07:27:46sharpebut now it's ~70% for 25fps
07:28:08midkaysharpe, very cool.. and how did you achieve that? :)
07:28:28sharpedifferent method of timing
07:28:49sharpethat's based off the number of frames required
07:29:43sharpeexecutes as many instructions it needs to, while the tick count is less than what is need to be by the time the next frame is drawn
07:32:32sharpeactually, that's 70% for 15 fps
07:32:42sharpeor something
07:32:55midkayhaha. either way. nice. :)
07:33:33B4gderI don't quite understand the measure
07:33:43B4gderwhere does "fps" get into the picture?
07:34:17sharpe300000 instructions per second, 25 frames per second, 12000 instructions per frame...
07:35:28sharpe300000 is how many instructions the cpu should emulate...
07:35:32B4gderthe C64 was 1MHz right? Each instruction took about 2-3 cycles. It would make it run at about 400000 instructions per second, wouldn't it?
07:35:50sharpeone sec..
07:36:18sharpethey take about 2-6 cycles actually...
07:36:32B4gderyes, but in average I think it is more to the 2-3
07:36:41B4gderof the ones most commonly used
07:37:07sharpei'm not very sure about the instructions per second though.
07:37:29sharpeall the sources i've look at say something like 300000 or near that...
07:37:34B4gdermidkay: prolly a freenode-thing
07:37:46midkayB4gder, yeah, probably.. never heard of it..
07:38:16B4gdersharpe: ah, ok so 300000 is the goal and you're at about 70% of that?
07:38:36sharpeunless my math is horribly off, i'm at 70%
07:40:28B4gdervery cool indeed it is anyway
07:41:21sharpewell, i'm not sure... right now, apparently 0 instruction aren't executed...
07:42:08sharpeso i may be close to 100%, but i don't think so...
07:42:59 Quit ninow ()
07:43:34sharpeHZ/framerate should be the interval that each frame can take...
07:45:32midkaynite all
07:46:18midkaygood luck on that, sharpe :) *gone!*
07:49:08 Quit sharpe ()
07:49:12 Join sharpe [0] (
07:49:45B4gdermidkay: ah it sounds like +R is registered-users-only can talk
07:50:13 Join infamis [0] (
07:52:57Mode"#RockBox +o B4gder " by ChanServ (ChanServ@services.)
07:53:12sharpeB4gder, what do you think of the c64 emulator idea?
07:53:25B4gderI like it a lot
07:53:56TiMiDwhat do you think about a multiplayer tetris ? :p
07:54:10sharpein general, or with rockbox? :D
07:54:23B4gderwell, in general its hardly a new idea ;-)
07:54:29Mode"#RockBox +o LinusN " by ChanServ (ChanServ@services.)
07:54:45TiMiDI meant this one :)
07:55:11B4gderTiMiD: I have no idea, I haven't checked it out
07:55:50TiMiDI thought that for 3.0, 2 tetris plugins would sound weird
07:56:08sharpeyeah, you need 3 tetris plugins to keep up with the version number
07:57:19sharpejust like we need 3 plugins that emulate a console/system for version 3.0
07:57:19TiMiDanyway, this one handles both vertical and horizontal screens, so it could replace the 2 others easily
07:57:49sharpeconsidering that there's rockboy, and pacbox...
07:57:57TiMiDand chip-8
07:58:08TiMiD(or was it removed ?)
07:58:08sharpeand my c64 emulator...
07:58:25sharpewell, we're all ready for version 4.0 then... :D
07:58:56sharpei have so much code randomly commented out :)
07:59:11Bg3rmorning :)
07:59:11 Quit cismo (Read error: 104 (Connection reset by peer))
07:59:30t0mashm? we have channel ops?
07:59:46t0masI tought we never used them
07:59:51t0massome irritating user passed by?
08:00:03TiMiDBg3r: 'afternoon
08:00:35B4gdert0mas: just the fact that lilo talks about the current state of abuse
08:01:02t0mashm? what state of abuse?
08:01:21t0masthey want us to have a channelop online all day or something?
08:01:23B4gder" Once again, we're still having problems with a kiddie attacking channels"
08:01:26ashridahthere's been some recent bots attacking various channels and spouting propaganda about various events from several years ago
08:01:34sharpetrollbots, annoyancebot, script kiddies oh my!
08:01:49t0masah ok
08:02:02Mode"#RockBox +o t0mas " by ChanServ (ChanServ@services.)
08:02:05*t0mas joins the party
08:02:28*sharpe is the person walking in the rain and notices the party he wasn't invited to...
08:03:01t0maswell I can invite you to my party tonight... but I don't know if you'll like my friends.. and I live in the Netherlands
08:03:03 Join sucka [0] (
08:03:13 Join cismo [0] (
08:03:20sharpethat's okay.
08:04:14sharpethat's around a six to seven hour timezone difference.
08:04:30 Quit ravve ("Leaving")
08:04:45t0masgmt + 2 here atm
08:04:55sharpeseven hours it is then.
08:05:23t0masghehe, I just got up (08:05) so it's pretty late there?
08:05:38sharpetwo in the morning...
08:05:59t0masthat's - 6 hours ;)
08:06:25sharpegmt-5... gmt+2...
08:06:38t0masgmt + 1 + saylight saving actually
08:06:43B4gdermath is hard at night
08:06:45t0mas*daylight saving
08:07:24sharpeespecially when you're trying to figure out how to pass variables as pointers to calculate how many frames are displayed per second. that seems so simple too. and it is.
08:08:13sharpeand gcc is complaining about a variable not being defined when it's the argument...
08:09:10TiMiDhmm night is not a good time to program
08:09:23sharpetechnically it's morning for me
08:09:27sharpeand god, i just fixed it.
08:10:09 Join mtnbkr_ [0] (i=mtnbkr@tor/session/direct/x-0b06def844b06142)
08:10:14sharpei think i may go to sleep around three...
08:10:15t0masok, time for a fast breakfast... and then running off to work..
08:10:28TiMiDhah and tomorrow you'll look at your code and say "how stupid I was I spent 2 hours on a stupid error thike that"
08:10:40 Join Paul_The_Nerd [0] (
08:11:04TiMiDthat's night programming
08:11:11TiMiDor drunk programming
08:11:11sharpetechnically, i'd be looking at it today
08:11:33TiMiDtechnically It will be tomorrow for me :p
08:12:23TiMiDdepends at the hour you'll awake after your coding night
08:12:30TiMiD(or mornig if you wish)
08:13:15 Join ender` [0] (i=ychat@
08:13:56 Quit mtnbkr (Remote closed the connection)
08:14:00 Quit lostnihilist (Read error: 110 (Connection timed out))
08:17:06 Part jnc
08:18:13infamishad a hard time with playback today (seeking & manual track skips) w/ cvs <= 4/18 19:08
08:18:44infamishaven't updated to the new stuff yet though
08:19:10infamishad to reboot like 15 times over the course of 2 hours
08:19:19t0mas[08:11:36] <sharpe> +technically, i'd be looking at it today <−− this afternoon ;)
08:19:32sharpetrue, but it's still today
08:19:43 Join dj-fu [0] (
08:19:48sharpethis afternoon is included in the today struct...
08:20:28 Quit actionshrimp (Read error: 110 (Connection timed out))
08:20:31 Join lostnihilist [0] (
08:20:40Paul_The_Nerd"Today" is always defined (for me) as "Until I sleep"
08:22:04infamisThen what is tomorrow? when you wake up
08:22:10sharpeso, while (awake) { today = day_awaken; } today = this_day; ?
08:22:26Paul_The_NerdBut that's just for things like "I'll do it tomorrow" or "Don't worry, it'll get done today"
08:22:26infamiswhat if you take naps throughout the day?
08:22:52Paul_The_Nerdinfamis: Then my 24-hour day is really a linked list of sub-days
08:22:53t0massorry to say it but... eh... NERDS!
08:23:16sharpelook who's talking! hah!
08:23:16 Join ravon [0] (n=ravon@
08:23:25ashridahpot, meet kettle!
08:23:34sharpewhat am i?!
08:23:35t0masmy day is just 24 hours... from 0:00 to 23:59... and I normally sleep from 0:00 to 7:30 :P
08:23:56infamisnerds count in 24 hour-mode :)
08:24:04t0mas(and yes, pot meets kettle ;))
08:24:17t0masall europeans count in 24 hour mode...
08:24:30infamisthat's odd
08:24:37B4gderno, its even!
08:24:45t0masit is... 24 is even ;)
08:24:46*Paul_The_Nerd grins.
08:24:52Paul_The_NerdNow that was awesome.
08:25:47t0maslate night IRC-ing in the morning
08:25:53infamiseh, europeans
08:26:17infamisdo europeans type with an accent too?
08:26:41infamisaccents are funny
08:26:48sharpei really do not believe i'm getting 1760 frames per second.
08:27:09 Join jnc [0] (n=jnc@
08:27:09t0masinfamis: little big brother guess: you're from chicago right?
08:27:26infamishow'd you know?
08:27:27jncyay chicago
08:27:37t0masI have my sources :P
08:27:52infamisdon't tell me you did a geographical traceroute or whatever
08:28:03t0masone of them being and guessing chi1 = chicago net 1
08:28:20infamisso who you callin' a nerd?
08:28:23sharpequick, someone look at my hostname!
08:28:54amiconngood morning
08:29:03amiconnSo many ops today??
08:29:17t0masyeah, that was what I tought when there were just 2 of em
08:29:18TiMiDt0mas: can you guess from where I 'm ?
08:29:19infamiswell it's only 1:27AM, so good morning
08:29:26B4gderto prepare for many "ops" commits ;-)
08:29:35*B4gder runs to hide
08:29:39t0masTiMiD: nope, but I would guess USA...
08:29:50t0maseh nope
08:29:56sharpei know!
08:30:00sharpeearth! he's from earth.
08:30:03TiMiDow with my crapy english that's unprobably
08:30:09TiMiDyes earth :p
08:30:13TiMiDnot france
08:30:16Paul_The_NerdTiMiD: Since I doubt you are an ancient Norse God, hmm... Israel?
08:30:35TiMiDneither israel
08:30:53Paul_The_NerdWrong side of the world then.
08:30:55infamissharpe, USA
08:30:59Paul_The_NerdFrom where we're guessing.
08:31:00TiMiDbut connected from a french ip
08:31:00t0mas"Guess the country part 1, please SMS your answere to 123-ROCKBX and win this great price!"
08:31:16*Paul_The_Nerd notes that it's *very* easy to figure out where he is.
08:31:20sharpeactually, it'd be RCKBX
08:31:30sharpefor me anyway.
08:31:41infamisI used to know a guy who had mindspring
08:31:47infamisso I say usa
08:32:02*t0mas notes he's getting late... bye!
08:32:51infamis...wouldn't wanna be ya
08:33:02sharpei think i can get it to 100% emulation...
08:33:45HClt0mas: can you take a look at my host cloaking thing sometime?
08:33:56HClstill not working :/
08:35:29TiMiDso about the tetris plugin, should I commit it ?
08:36:32Paul_The_Nerd"Feature Freeze"
08:36:45sharpeokay, at 11fps, it seems to be at 100% emulation...
08:36:59 Part dj-fu
08:37:09sharpeand at 25fps, it seems to be 72%...
08:37:21TiMiDit's not really a feature
08:37:23Galoissome say plugins aren't part of the core, so they're immune to the feature freeze
08:37:42TiMiDand it would be great not to have 2 plugins to do the same thing
08:38:34Paul_The_NerdGalois: Some plugins do modify part of the core though I doubt this one does.
08:40:02TiMiDI think this one is bug free
08:40:17TiMiD(but of course it isn't :D)
08:40:21hardeepfamous last words
08:40:27 Join nave7693 [0] (
08:40:51Paul_The_NerdWell the keys would be "does it modify any files outside of \apps\plugins" and such
08:40:55 Join Ninpo [0] (
08:40:56 Quit mikolas (".")
08:41:05 Part nave7693
08:41:21infamisthere should be a pickup line plugin....list of nouns; "Baby, your %1 is like %2"
08:41:25Ninpohey peeps, quick question...just got rockbox onto my ipod, this may sound dumb but how do I put my music onto it? Just copy the files to the root?
08:41:42TiMiDNinpo: yes
08:41:46Ninpooh fantastic
08:41:52Ninpoits really that easy
08:42:00Galoisanywhere except the iPod_Control folder
08:42:07Ninposweet :D
08:42:07Paul_The_NerdSlasheri: Are you around?
08:42:10Ninpotime to test some oggs
08:42:25TiMiDPaul_The_Nerd yes, it adds some fn to the plugin api
08:42:46TiMiDbut thhat's not a big deal
08:43:02TiMiD(multiscreen api)
08:43:08Paul_The_NerdTiMiD: Yeah, but the whole point of the feature freeze is moot if you start saying "Things below X size don't count"
08:43:29TiMiDhmm it depends what you call a feature
08:43:55hardeephmmm, playback code doesn't set status to "playing" immediately when you call audio_play() which is different then archos behaviour
08:44:00TiMiDI don't care if it's not included, but I just thought 2 tetris plugins were confusing
08:44:15hardeepi know playlist viewer assumes it does, and maybe other parts of code too
08:44:22hardeep(hoping lostlogic reads irc logs)
08:45:19Paul_The_NerdTiMiD: I do think two tetris plugins is a confusing thing. And I really don't have any say on anything, but if they aren't adding the sid codec (which again isn't *really* core, as it's a similar case to a plugin) I kinda have doubts as to whether a plugin is a good idea.
08:47:21TiMiDyes, this must be asked to a core dev
08:47:28TiMiDof course
08:47:46TiMiDbut maybe codecs are a more critical part of the code
08:48:30B4gderimho, adding features to a tetris plugin is not good freeze stuff
08:49:00B4gderit is better to stay put and not start adding things just because it might be considered not the core or whatever
08:49:06 Join Ribs [0] (
08:50:25TiMiDthen maybe addinf a defin to compile the archos tetris plugin on archos only
08:50:37NinpoThank you all o/
08:50:44 Part Ninpo ("rockbox is fuckin awesome")
08:51:03B4gderTiMiD: that might be an idea, yes, I don't know enough to comment on that
08:51:36TiMiDas for multiscreen work one core, I'm thinking about restarting to port some screens to the remote, would this be considered as "new features" ?
08:52:30B4gder"remote support" is listed for 3.0
08:52:52B4gderso I'd say we want it to cover as much as possible
08:53:15SlasheriPaul_The_Nerd: yes, hi
08:53:23infamisahh I hate loosely typed languages...compiles with error but keeps going, doesn't tell you there's an error, and now the program doesn't work and you don't know why!
08:53:38infamisnot an error but a 'warning'
08:58:47amiconnTiMiD: Tetrox is running on archos as well
08:58:47TiMiDwhich language ?
08:58:54TiMiDamiconn: yes
08:59:08TiMiDbut it's not horizontal
08:59:13TiMiDjust mentioned
08:59:13amiconnIt's indeed overkill to have 2 tetris plugins
08:59:29Paul_The_NerdSlasheri: I figured it out. I was trying to search for a string in the Title tag in tagcache, such as song ? song ~ "blah" and didn't realize I should use "Song" instead of "Title"
08:59:59Paul_The_NerdSlasheri: I completely forgot to look at the real searches.
09:00:24Slasherihehe :)
09:00:36amiconnWhat's odd, and bad, is that both I and my sister consider rockblox quite hard, while tetrox is too easy because it's so slow, even though the blocks are smaller due to the vertical aspect
09:01:03TiMiDhmm even with the level increase ?
09:01:10amiconn*Imho* we should drop rockblox, and keep tetrox but make it a bit faster...
09:01:42TiMiDon my plugin there is an option to select the speed at start
09:01:49SlasheriPaul_The_Nerd: hmm.. maybe it would be more clear to add also a "title" that has the same meaning as song
09:02:26Paul_The_NerdSlasheri: I think that would be a very good idea, so that people understand it's explicitly searching in the Title tag.
09:02:35LinusN"title" is way better than "song" imho
09:02:40TiMiDsince I never tested rockblox on archos, I can't compare the speed
09:02:52Slasheriok, i will add both :)
09:03:13Paul_The_NerdLinusN: "song" makes sense in the original use of it, which is for the last level where it lists songs.
09:03:37amiconnHonestly, I think one is sufficient
09:03:48LinusNare audio books and podcasts songs too?
09:03:55B4gderI think one is sufficient too
09:03:56Paul_The_NerdGood point
09:04:01LinusN"song" implies music
09:04:10B4gderotherwise you can just start thinking "title" too and then it'll never end
09:04:14LinusNand vocals
09:04:25TiMiDI think title is better tahn song
09:04:29TiMiDfor the name
09:05:57Slasherihmm, maybe it would be more clear to remove the song entirely then
09:06:26Paul_The_NerdSeems the best way, yes.
09:06:32Slasheriat least no duplicate entries then
09:06:33amiconnWhat about the using 'Search by track' and 'All tracks'?
09:06:47amiconnThat'd be neutral (wrt audio books etc)
09:07:10LinusN"track" sounds good
09:07:12amiconn"Songs" is only displayed at the top level, and that's indeed all tracks
09:07:12Slasherigood point, i will update that also
09:08:10Paul_The_NerdI think "Title" is better than "Track"
09:08:15amiconnPerhaps these strings should get their <> back.
09:08:24Paul_The_NerdSince specifically you're searching within the Metadata tag called "Title"
09:08:48amiconnPaul_The_Nerd: Okay, 'Search by title' but 'All tracks'
09:09:16Paul_The_NerdYeah, the top level "All Tracks" seems fine.
09:09:55LinusNsounds ok
09:10:18Slasheribtw, do you think the toplevel should be "Artists" (as it's currently) or "All artists"?
09:11:08amiconnThe <> would make them distrinct from real track names, which would be useful when they eventually get reused at lower levels
09:11:09Paul_The_NerdI think that part's fine now, in my opinion.
09:11:47hardeepI personally like "Artist" (no plural)
09:11:52hardeepsame with "Album"
09:12:08Slasheriah, hmm
09:12:13Paul_The_Nerdamiconn: If you have it displaying icons, it shows a folder icon next to Artists, and the song icon next to actual music.
09:12:29Paul_The_Nerdamiconn: In that regard, it's not really different from the directory tree.
09:12:33amiconnI know
09:13:24amiconnStill, you get both folder icons for these special folders and folders pertaining to actual albums or artists
09:13:36LinusNthat, combined with the "The" removal feature, will make it hard to browse albums by Prince :-P
09:14:42Slasherii am about to add "<All>" entry now to make it easier to create larger playlists
09:14:58amiconnPaul_The_Nerd: The point is that if you eg browse by artist, there could be an additional 'All tracks' directory in addition to the album folders
09:15:10Paul_The_NerdSlasheri: I think rather than <all> adding the ability to do "Or" as opposed to just "And" solves it.
09:15:30amiconn...and '<All tracks>' would be more distinct from the album names
09:15:35 Quit JBGood (Read error: 110 (Connection timed out))
09:15:51Paul_The_NerdSlasheri: Then you could do Genres : Albums OR Titles, I just don't know how you'd define the level after Albums
09:16:03SlasheriPaul_The_Nerd: hmm, but that would show as the first entry in the browser and it would list all tracks
09:16:28Paul_The_NerdSlasheri: I suppose a hard-coded "All songs that me conditions up to now" style option makes sense.
09:16:59 Join petur [0] (
09:17:08SlasheriPaul_The_Nerd: yep, that is what i was about to do :)
09:18:52Paul_The_Nerdamiconn: The thing is that the Album Artist Genre things are just top level. Do you want a hard-coded solution to differentiate strings typed in tagnavi.config from strings like actual tag names provided from the cache, or is it just those top level ones you'd like to see changed?
09:19:40amiconnThe problem with the strings in tagnavi.config is that they can't be localised
09:20:28hardeepSlasheri: I had a thought on improving playback responsiveness from tagcache for large playlists, especially on non-disk cache systems
09:20:39hardeepinsert the track you want to play first and immediately begin playing
09:20:40amiconnThe strings needed for the tagcache (tagdb) top level are already present in the .lang files
09:20:44hardeepinsert the rest in the background
09:20:50Slasheriamiconn: i think that just hard-coding those common searches would be one solution for that
09:21:15amiconnI'm not sure about the best way to keep both the flexibility of tagnavi.config *and* use the localised strings :/
09:21:30Paul_The_NerdIf that's the case, I think the option to hide the hard-coded searches should be offered, if there's ever the ability to create branched tagnavi.config lists.
09:21:33Slasheriamiconn: yep, i know..
09:22:01Slasherihardeep: hmm, sounds like a good idea. Is there some special tricks with the playlist engine to do that?
09:22:47amiconnPaul_The_Nerd: It would also be possible to detect the standard searches when they're defined in tagnavi.config, and redirect the strings to the localised ones. But that means more code...
09:22:47hardeepSlasheri: only think that would need to be handled is using "insert_shuffled" in shuffle mode, otherwise it should work (in theory)
09:23:06hardeepi was going to look into it −− can you think of any problems on the tagcache side?
09:23:34***Saving seen data "./dancer.seen"
09:23:40Slasherihardeep: i think tagcache shouldn't mind at all doing that
09:23:51amiconnhardeep: I don't think that's a good idea...
09:23:51Slasheriplease implement it if you would like :)
09:24:10hardeepamiconn: why not?
09:24:28Paul_The_Nerdamiconn: Yeah. *Ideally* I'd like the option to be able to do Artists, then "A-F" "G-blah" etc, which is why a hard-coded search doesn't seem optimal unless I'd have two artists options visible. (Not that you can do that sort of search with the current tagnavi, but I'm sure in the rather long term it'll become as terrifyingly extensible as our WPS is getting to)
09:24:38amiconnWe don't do that when playing from directory mode either, or playlists
09:25:17hardeepyeah, but they're fast operations =)
09:25:34amiconnYes, and so is playing from tagcache
09:25:44Paul_The_NerdWell the tagcache playlist generation seems about as quick, if not quicker, than clicking on a folder and choosing insert.
09:25:52 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
09:25:53hardeepit's really fast with disk cache
09:26:06hardeephow is it on an archos player though? (i haven't had a chance to try)
09:26:19amiconnIt's also fast without dircache (which is what I always use), and even on archos
09:26:30hardeepoh, really?
09:26:38hardeepeven for large playlists?
09:26:48hardeepi'm thinking of the "All tracks" type playlists
09:29:17amiconnHitting Play on an arbitrary track in 'Songs' on my archos recorder takes 27 seconds (for >3200 songs)
09:29:30amiconnHowever, there is a bug
09:29:55amiconnAlthough the tagcache engine counts all songs, only the first 500 get actually added to the dynamic playlist
09:30:13amiconnI have set 'Max files per directory' to 500
09:30:33Paul_The_Nerdamiconn: With TagCache in memory and DirCache, ~1700 songs takes 2-4 secondsish.
09:30:43Paul_The_NerdSo there's quite a large disparity in speeds.
09:30:50amiconnThere's not
09:30:50 Quit whatboutbob ("CGI:IRC")
09:31:01amiconnTake into consideration that this is an 11MHz CPU
09:31:04Paul_The_NerdOh, right
09:31:57*Paul_The_Nerd decides to test tagcache without dircache/memory.
09:32:01hardeepif it's only ~25s for 3000 songs then I agree that the change is unnecessary... I was expecting it to be a lot more (minutes...)
09:32:13amiconnhardeep: Generating 'root.m3u' from the menu takes 45 seconds.
09:32:22amiconnNow figure what is the faster operation...
09:33:12amiconnThe bug should be fixed though. Not sure how complex that would be...
09:35:13hardeepamiconn: that's definitely faster then I was expecting
09:35:25hardeepthe old tag database was really slow as i recall
09:35:31amiconnYes, the tagcache engine *is* fast
09:35:44amiconnI still need to test the improved commit speed
09:36:08amiconnSlasheri: Btw, I think we should use the faster crc32 for the swcodec platforms, and the space saver only for archos
09:36:41amiconnThere is an other crc32 routine in one of the flash preparation tools which is drop-in compatible to the 4-bit-at-a-time one
09:37:28amiconnSlasheri: Ah, yes, in flash/make_firmware/make_firmware.c
09:37:39amiconn(that one needs some code policing though)
09:38:07Slasheriamiconn: hmm, good idea
09:38:17amiconnThis one adds 1KB worth of table data, so no good for archos
09:38:49Slasheriamiconn: the commit is still not as fast as it could be.. with little more memory usage, it can be made still significantly faster
09:38:55Slasherii will look into that later
09:40:34amiconnhardeep: Some comparison data for you: The initial tagcache scan took ~17 minutes on my archos. The commit took 5 minutes (without the crc32 method added yesterday)
09:41:13amiconnI should run another test...
09:42:01hardeepi just ran one on my h120... i didn't time the scan but the commit took about 2 mins for me
09:43:20Slasheriyep.. currently the slowest function is tempbuf_locate in tagcache.c
09:43:53sharpewas it gargamel on the smurfs that was the human?
09:43:58sharpeor was that the cat?
09:44:04Slasherii think i will convert it to use direct memory lookup instead of searching entries one-by-one.. That would increase the complexity from O(n^2) to O(n)
09:44:30petursharpe: gargamel is the human
09:45:12sharpei'm writing a large incoherent sentence which just so happened to be discussing the hidden socialistic views in the smurfs show.
09:45:41peturdon't forget to mention they're *Belgian* :D
09:49:53sharpeokay, the sentence ends with the statement about being belgian...
09:50:47sharpeand gains coherence around the first appearance of the patriarchal type of gov't in the smurfs
09:51:33B4gderand the *red* guy is in charge which means communism
09:51:38B4gderjust like Rockbox
09:51:49*Paul_The_Nerd grins
09:51:58B4gdernow who's our Gargamel?
09:52:05*petur didn't know swedes were red
09:52:47peturB4gder: gl ?
09:52:50sharpevanity is the effeminate smurf
09:53:05sharpehe's the modern 'metrosexual' of the smurfs.
09:53:17Paul_The_NerdThe Gargamel of the moment would have to be the guy from Popcap I think.
09:53:25Paul_The_NerdHoping to turn us into gold, but never really going to succeed
09:54:29peturI can understand that guy... the graphics do look similar to the casual user
09:54:43sharpesubstantial evidence shows that frequently used rambling of the third kind result in unemployed, ethical, belligerent notes that sometime show signs of intelligence due to the lack of oxygen in the cookie cutter of love which is wielded by the chef that teaches in italy which is home to the italian culture which includes some people who are happy, some who are depressed, and even some who are completely oblivious to the network of disesta
09:54:56Paul_The_NerdI dunno. They looked pretty different even at first glance to me.
09:55:00 Quit hardeep ("BitchX: good to the last drop!")
09:55:04Paul_The_NerdY'know, the two screenshots that were linked.
09:55:48Paul_The_NerdI would think that if it's a product of his company, and he's asserting that the graphics were ripped (which I think is the word he used) a little more than casual observation would be made. :)
09:55:50peturMy point is: if I show you one, then the other, you won't be able to tell which one is ours.
09:56:12sharpeokay, i'm going to sleep now
09:56:17sharpeg'night everyone.
09:56:23sharpei'll be back today.
09:56:53Paul_The_Nerdpetur: If you showed me just the sets of jewels, side by side, I could tell you which were ours.
09:56:56B4gderpetur: well, they wouldn't even have to be very similar for that to be the case
09:57:15 Quit jumpfroggy ("CGI:IRC (EOF)")
09:57:17B4gderfor a casual user
09:57:26B4gderit doesn't give him right though
09:58:18peturI guess if he had asked it politely the reactions would have been different...
09:58:33B4gderI'm sure they would've been
09:58:59Paul_The_NerdI'd say the Green and Orange ones are *very* similar
09:59:29*Paul_The_Nerd shrugs
10:00:30peturthe gpl ones were created to make them look like the originals. I probably don't know enough about the law, but I wouldn't risk it going to court
10:06:46ravonAre there any eq presets available somewhere? I'm an audio retard and just can't get it right :/
10:07:36Zagorravon: flat is preset :-)
10:07:50peturan excelent one!
10:07:57B4gdermy favourite!
10:08:00peturI use it every day
10:08:10Paul_The_Nerdravon: It's not like there's a "right" anyway. It's more of a "Tinker until you like the way it sounds." I'd recommend using the simpler modes, or just changing the gain, not the Q or the frequency.
10:08:11B4gderI've crafted my own flat
10:08:21*Paul_The_Nerd is quite the fan of flat.
10:08:47ravonPaul_The_Nerd: Yeah, I've tried. It either gets no bass, gets distorted or lacks treble. I fail.
10:09:13ravonI usually use the player's different presets (rock, pop, ska etc...) until I find one that sounds good enough.
10:09:25Zagorravon: start from the beginning. what do you find lacking when flat?
10:09:39ravonZagor: Bass
10:09:40peturravon: the tip is to use small gains
10:09:50Zagorravon: what headphones do you use?
10:10:36ravonZagor: JVC HA-G77
10:11:31ravonThe Cowon firmware can make the head shake with those, so they can produce bass.
10:15:29 Join webguest44 [0] (
10:15:32 Part webguest44
10:15:48 Join leftright [0] (
10:16:25Zagorravon: well I'm not an eq specialist since I am very happy with flat settings.
10:16:27Paul_The_Nerdleftright: Regarding your audio problem, nobody seems to be able to reproduce it. Are you sure it's not some form of hardware failure?
10:16:42amiconnSlasheri: Full scan: 17:50, commit: 3:30 (on my recorder v1)
10:17:20Paul_The_Nerdravon: Try setting the first band to 100hz, same Q as it had, upping the gain to an arbitrary level (+10?) and then lowering the precut by whatever amount you picked to raise that.
10:17:36ravonPaul_The_Nerd: Testing
10:17:57petur10 may be a bit much, no?
10:18:26Paul_The_Nerdpetur: He was talking about head shaking, so I was thinking "aim high"
10:18:31leftrightI did a clean install and a reset of all settings, this seems to have fix the bugginess on my players. It seems that its quite important to do a clean istall or at least a system reset before shouting "bug"
10:18:46peturhe was also talking about distortion
10:18:59Paul_The_Nerdpetur: Yeah, but I imagine he just meant the clipping that is *oh* so easy to cause
10:19:12Paul_The_Nerdleftright: Aaah, a good idea yes. I thought you said it was happening on two h140s though
10:19:15ravonPaul_The_Nerd: Better, thanks. Still doesn't match up against the fancy Cowon boosts, but it's way better now.
10:19:16peturby using gains of 10 :P
10:20:07Paul_The_Nerdravon: Well, raise that value more, and then adjust the precut to compensate. Also, you may want to change it to 120hz, or maybe a little more. I don't know the range that Cowon desides is "Bass"
10:20:18leftrightPaul it was, I did clean on both and reset both, both working fine now, I now do a clean and reset for Every Install
10:23:53ravonPaul_The_Nerd: Sweet! :D
10:26:41Paul_The_NerdI take it there's been some improvement? ;-)
10:28:02ravonMajor :)
10:29:05Paul_The_NerdThere should be no reason why our firmware can't produce more bass than the stock firmware on most of the players, since you have access to both our EQ, and the Bass/Treble options in the normal sound menu (though I'm not sure if those are enabled for the X5 yet)
10:29:17Paul_The_NerdThough I can't for the life of me think why you'd want *that* much bass.
10:29:45B4gderthe "normal" ones aren't present on X5 since the X5 DAC doesn't have them iirc
10:29:48LinusNPaul_The_Nerd: remember that rockbox doesn't yet have a psychoacoustic bass
10:29:58B4gderyet? so they exist at least?
10:30:36B4gderI mean on HW
10:30:51LinusNit exists in my mind and some code fragments on my pc
10:31:06Paul_The_NerdLinusN: Yes, but on many of the players there's the HW bass, and that plus a low frequency boost on the equalizer will usually more than satisfy bassheads.
10:31:40ravonPaul_The_Nerd: What can I say, I enjoy a bass concussion every now and then :)
10:32:01peturmind your ears tho!
10:32:01ravonPaul_The_Nerd: Your suggestions made the day though.
10:32:25LinusNi think we might want to have some kind of auto-precut for the eq
10:32:33LinusNas an option
10:32:34leftrightravon, I'm surprised that much bass doesn't make you nauseas
10:33:24ravonleftright: I'm spoiled with the one-click-wonder settings in the iAudio stock firmware :(
10:34:13B4gderravon: I think you'll find people in here not generally agreeing that is a "wonder"
10:34:23Paul_The_NerdLinusN: At the very least, an automatic precut for the "Simple EQ Settings"
10:34:27ravonB4gder: It's a wonder to me :)
10:34:33leftrightif I have too much bass with my IEM's, I start to feel ill
10:34:37LinusNPaul_The_Nerd: yeps
10:35:13B4gderthe only eq I use is a perl operator ;-)
10:35:48Paul_The_NerdI've used the EQ to try to clean up some audiobooks so that they're more understandable.
10:36:08leftrightI use a 3bd of bass and a wee bit of treble
10:37:08 Part leftright
10:37:20 Join zblach [0] (
10:37:59zblachlemme just say that I'm really impressed with rockbox. I moved my ipod mini over a few days ago, and I don't regret it at all
10:38:08B4gderanyone for a coffee refill? I'm about to get one myself
10:38:13zblachB4gder: please
10:38:44zblacht must also seem like a slap in the face to apple, that their little player can handle all sorts of file formats without their using it
10:38:51zblachwhich I also find a bit strange
10:39:24zblachbut I have one concern
10:40:05zblachwhen running a media file, i sometimes get pauses at the start of the song. around 2 second, 4 second and 8 second mark. what could be causing that?
10:40:19Paul_The_NerdIs it just when you first play a file?
10:40:26LinusNzblach: the ipod not supporting other formats is a political/business decision by apple
10:40:42zblachLinusN: that would seem to make sense, but what about the free codecs?
10:40:45Paul_The_NerdThat's often just a buffering issue with the initial disk spinup, I think. My H120 does that sometimes too.
10:40:59zblachPaul_The_Nerd: and it ignoring some songs entirely?
10:41:11LinusNi.e apple doesn't *want* ipod to support other formats
10:41:14Paul_The_NerdThat's something different.
10:41:21Paul_The_NerdAre the songs being ignored AAC, or MP3?
10:41:45Paul_The_NerdAre they being ignored when you click on them directly, or skipped in a playlist, or what?
10:41:53LinusNzblach: large album covers in the tags?
10:44:40 Join _zblach [0] (
10:44:59 Part Paul_The_Nerd
10:45:11_zblachme again. accidentally plugged my ipod into my reset switch
10:46:57 Quit Ribs ("Leaving")
10:49:31 Quit midkay (Read error: 104 (Connection reset by peer))
10:50:04 Join midkay [0] (n=midkay@rockbox/developer/midkay)
10:50:21 Join Paul_The_Nerd [0] (
10:51:45 Join SereR0KR [0] (
10:53:03 Quit stoffel_ (Remote closed the connection)
10:58:11 Quit zblach (Read error: 110 (Connection timed out))
10:59:25nudelI'm assuming .SVG is a vector image format which adds further weight to the "they're original" claim, since I doubt PopCap have released the vector versions of their images (assuming they weren't drawn as bitmaps from the start).
10:59:35 Join Farpenoodle [0] (
10:59:44 Quit _zblach (Remote closed the connection)
10:59:58B4gdernudel: exactly
11:00:05nudelThey're blatant replicas of the Popcap graphics but I guess that isn't illegal.
11:02:10 Join zblach [0] (
11:03:57 Join RedBreva [0] (
11:05:58 Join Lynx [0] (
11:09:24 Quit Arrogant ("Leaving")
11:10:56RedBrevaIs it very quiet in here today, or do I have a connectivity problem?
11:11:23Bg3rit's quiet atm
11:11:40RedBrevaOK :D Just Checkin
11:12:13ashridahi just caught myself trying to drag the slider in the jpg Bjorn had in his email :)
11:12:28RedBrevaheheh, been there done that...
11:12:31zblachwasn't someone supposed to get me coffee?
11:12:35Paul_The_NerdImages should not have sliders.
11:12:48ashridahPaul_The_Nerd: it was a pic of gqview
11:13:56Paul_The_NerdI saw the pic.
11:14:05Paul_The_NerdI thought about sliding it too
11:15:21RedBrevaHaveing played with VMware for a week or so, I am very impressed, and am considering wiping my laptop HDD and installing Linux... Is there any (nooby proof) documentation on taking a clean install and getting it ready to be able to build Rockbox, Manuals and Sims
11:16:25ashridahRedBreva: there's documentation on what you need, but you'll need a standard C development environment for the linux distro (which, for example, ubuntu doesn't install by default if you choose it).
11:16:25infamisnooby proof?
11:16:33ashridahdocumentation on
11:16:36ashridaharghl even
11:16:44*ashridah smacks forehead
11:16:54Paul_The_NerdSpecifically the CrossCompiler wiki page, no?
11:17:36infamisthough I had to reinstall linux just to get rockbox working :)
11:17:48infamisfor sim compiles that is
11:18:13ashridahinfamis: then you did it wrong. all i did was install the SDL development headers :)
11:18:16ashridahdepends on the distro tho
11:18:19ashridahsome are easier than others
11:18:29ashridahand technically, if you're using gentoo, you reinstall the entire damned thing every week anyway
11:18:36RedBrevaI was looking at SuSe 10...
11:18:48ashridahRedBreva: whatever floats your boat.
11:18:51infamisredhat was what I had before...went with debian
11:18:57 Join Moos [0] (
11:19:16*ashridah can't wait to get his main workstation living again
11:19:47ashridahthen it's back to debian unstable for me
11:19:48infamisredhat didn't have the x11 includes or something & nothing I tried worked on trying to get the includes individually
11:19:52RedBrevaDon't really understand the diffs between distro's - as long as it's running KDE when I'm finished I am happy
11:20:05ashridahinfamis: yick.
11:20:32infamisnot like I use linux on daily basis so it was whatever
11:20:38infamisno bigge
11:22:05infamisanyways it's almost 4:30am and it's time to go home to go home from work.
11:22:12 Quit infamis ("I'm out like michael jackson in the 90s")
11:23:16 Quit Lynx_ (Read error: 110 (Connection timed out))
11:23:16 Nick Lynx is now known as Lynx_ (
11:23:38***Saving seen data "./dancer.seen"
11:24:10 Join lamed [0] (
11:24:55RedBrevalooked at some of the docs.. seems to be lots of setup guides for Windows, but the linux guides seem to assume quite a lot of general knowledge already...
11:25:35LinusNgenerally, windows users need more hand-holding than linux users
11:26:01B4gderthe docs are added on demand/need basis
11:26:06RedBrevaWell, if I install SuSe (or whatever) I assume I can (continue to) ask stupid questions here to get it working?
11:26:07lamedI was just wanting to ask why i can't get rockbox configure to know that i'm having sdl already installed on ubunto linux so here you go... (linus, i'm new to linux)
11:26:32B4gderlamed: configure checks for it and if it doesn't find it, something is wrong
11:26:57lamedwhat should i do? i've reinstalled it and there's not sdl-configure file nowhere
11:27:45RedBrevaLinusN: Not when they are Windows users trying to set up Linux ;-)
11:27:59ashridahlamed: Ubuntu doesn't include a development environment on the cd
11:28:05ashridahyou need to install those bits using the net
11:28:28LinusNlamed: you need to install the development sdl package
11:28:53ashridahso, something like apt-get install libsdl1.2-dev might do it (if it's similar in name to ubuntu's
11:29:14lamedmmm... thanks. I only wish i could connect to the internet to do that in the first place (i'm on winxp now)
11:29:19ashridahyou'll also need to install the compiler and C development package
11:29:30lamedalready have those
11:30:39lamedI think this information should be reminded in the how to build the simulator wikipage. has anyone succeded in running the sim on windows?
11:30:42 Quit sucka ("( :: NoNameScript 4.01 :: )")
11:31:52Paul_The_Nerdlamed: The SDL sim worked fine built on Cygwin last time I checked. I'll admit it's been a few weeks though.
11:31:56RedBrevaI can build the sim under Cygwin (Very Slowley) and run it under windows XP
11:32:38RedBrevaBut I like the fact that I can build 10 sims and 10 maunals as quickly under VMWare than I can build 1 sim under Cygwin!
11:32:59RedBrevamanunals = manuals
11:33:38lamedI'm getting the same no sdl message. bah... i probably have to install the same dev-package right?
11:34:09lamedand there's no apt-get on cygwin :/
11:34:18RedBreva - step 3
11:34:25B4gderin cygwin you should install the sdl package mentioned in the wiki page
11:34:42Paul_The_Nerdlamed: With cygwin you just run the cygwin installer, and find the sdl devel package in the list.
11:35:14 Join ze__ [0] (
11:35:17Paul_The_NerdB4gder: He found that quote on
11:35:54 Quit ze (Read error: 104 (Connection reset by peer))
11:35:57 Nick ze__ is now known as ze (
11:36:16Paul_The_NerdI was curious about it too.
11:36:38Paul_The_NerdBut I stumbled across it when reading another post before coming back to ask.
11:37:23lamedyes paul, that what's i'm doing, thanks.
11:38:19B4gderI added a question about it in the wiki page
11:39:02B4gdernow I see its listed under "known bugs"
11:39:05B4gderseems weird
11:41:39 Join muesli__ [0] (n=muesli_t@
11:43:31Paul_The_NerdI think the idea was that a full buffer fill on the first song picked is a bad idea, but *just* one song means a delay on the first skip, but 1.1 songs, (or 1 and some small percentage) means the first skip leaves you with some play time will the drive spins up, but doesn't have the overhead of caching the beginnings of several songs, or a full buffering on playback start.
11:45:54linuxstbWho's idea is that? I would prefer a full buffer fill on playback start...
11:45:58 Quit zblach (Remote closed the connection)
11:46:42Paul_The_Nerdlinuxstb: I'm not sure why there's only a partial fill on playback start, honestly. But right now if you look, it buffers exactly one song, then after that song, does a full buffer fill.
11:47:09linuxstbYes, I know it does that now.
11:47:44ashridahi don't recall it being an entire buffer, and iirc, it was mostly for the "i can't find the song i like" crowd, so it doesn't go nuts buffering if you're constantly switching songs
11:47:52ashridahs/entire buffer/entire song/
11:48:04ashridahit was some specific amount, wasn't it?
11:48:20Paul_The_NerdI think it's changed to "exactly 1 song" but I could be wrong.
11:48:43linuxstbI think filling the whole buffer could actually improve responsiveness for users who skip a lot - it keeps the disk spinning during the first few seconds, so if the user chooses to skip, then the disk is already running.
11:49:24Paul_The_NerdWell also, if the entire buffer is full, then a "next" early on is going to be an in-buffer skip anyway.
11:50:37linuxstbTrue. I forget that some people use lossy codecs...
11:51:51lamedthanks redbreva, I just had to add /usr/loca/bin to path.
11:54:08amiconnWell, while I also tend to think that a full buffer fill at start might be better, there's another point for a partial fill on startup than just for undecided people
11:55:27amiconnA partial fill spins down the hd earlier, and that might be desirable to protect the hd, as the unit likely receives some bumping and shaking after starting playback
11:56:58Paul_The_NerdHmm... It may be "1 song, or X size, whichever is smaller" because it definitely doesn't do a whole flac, but it always seems to have to spin up again with a next track on the first song.
11:57:24linuxstbI've got a feeling it's about 2MB of data.
11:57:28ravonHmmm, the Coldfire is big endian? That could be a reason why the dumb-plugin doesn't work.
11:57:36Paul_The_NerdI don't really have a view on whether 1 song or whole buffer is preferable though. Well, I prefer whole buffer, simply as it's probably better for my listening habits' battery life.
11:57:40lamedwhat #if sentence can I use to determine whether a button define is having 'BUTTON_REPEAT' in it? Meaning #define LIST_PGUP (BUTTON_ON | BUTTON_UP) = false, and #define LIST_PGUP (BUTTON_ON | BUTTON_REPEAT) = true?
11:57:41Paul_The_NerdBut that's just self-interest.
11:57:52 Quit ScootScat (Read error: 110 (Connection timed out))
11:57:54Paul_The_Nerdlinuxstb: Nah, it's more than 2mb. It was ~7 when I just buffered a flac.
11:58:01amiconnPaul_The_Nerd: It's a minimum of 1MB and a maximum of 8MB, aiming at one song
11:58:18Paul_The_NerdAh, see, an answer is known. :)
11:59:04linuxstbYes, just tried a 10.5MB vorbis file, and it buffered up to about 7.4MB. So that's probably 8MB, with the first 0.6MB being consumed during the disk fill.
12:00:26amiconnMaybe it's buffer_size/4
12:02:21lamedi guess no one understands what i'm asking..?
12:02:43linuxstbHere's the reason for that behaviour -
12:03:21linuxstblamed: Why do you ask "what #if" - do you mean "what #define" ?
12:03:41linuxstbOr you want to check if an existing definition uses BUTTON_REPEAT?
12:05:37lamedthe latter
12:06:12linuxstbMaybe something like "#if (LIST_PGUP & BUTTON_REPEAT)"
12:07:04 Quit ashridah ("Leaving")
12:07:04lamedi'm only trying to check button_repeat, without, say, list_pgup
12:07:36 Join ashridah [0] (
12:07:38linuxstbThen I am confused....
12:07:51linuxstbWhat are you trying to achieve?
12:08:07Paul_The_Nerdlamed: That line he showed is basically "Does LIST_PGUP include BUTTON_REPEAT"
12:10:05lamedthe thing is, i have a couple of targets that already have button_repeat in the button table, and others use something like BUTTON_LEFT, and i want to out-define the case or BUTTON_LEFT | BUTTON_REPEAT | BUTTON_REPEAT, which causes a compiler error. i'll try what fideling with what you've mentioned
12:10:37lamedfideling = fiddling
12:11:35linuxstbI think you should define LIST_PGUP_REPEAT, or something similar.
12:11:54linuxstbYou seem to be creating a new action, which requires you to define key mappings for all targets.
12:13:36lamedi'm really not... I'm just trying to clear up the button orgy i've made.
12:15:58linuxstbOK, I think I can see what you're doing. You want to put "#if ((LIST_PGUP & BUTTON_REPEAT) == 0)" around the "case LIST_PGUP | BUTTON_REPEAT:" line?
12:18:38lamedmmm... yes, you're warming up on it!
12:20:47linuxstb:) The problem is that you end up with duplicate case values...
12:21:22linuxstbSo a clearer solution could be "#if (LIST_PGUP != (LIST_PGUP | BUTTON_REPEAT))"
12:21:34lamednot exactly, because the idea is taking that case LIST_PGUP | BUTTON_REPEAT line
12:23:09linuxstbI don't understand again.... Can you pastebin the guilty code?
12:24:22lamed#if (LIST_PGUP & BUTTON_REPEAT) means that if BUTTON_REPEAT is included in the button #define, it calculates to 1, otherwise a zero?
12:24:49linuxstbNot quite - it will calculate to the value of BUTTON_REPEAT.
12:25:03linuxstbWhich is non-zero.
12:25:17lamedmmm. so same result
12:25:32lamed10x for the in-depth.
12:26:27lamedthat will do the trick then. I will publish a one-button vertical screen scroll patch soon :)
12:27:11ravonWhat's "stkov"? Got a "*PANIC* Stkov dircache"
12:27:20linuxstbA stack overflow.
12:28:49Paul_The_NerdFor some reason I just love errors that start with *PANIC*
12:30:45 Join VoltageX [0] (
12:31:39lamedlinuxstb: if you'r kind enough to help me on another one. I have two integers: w,h. I have to check, if (w-h > 0) then w-h, else w=0. in another words, w-h should always be a positive number. question is, is there a mathematical trick i can do to make this check without making an if sentence?
12:32:10Bg3rabs ?
12:32:13lamedthen =(w-h) else =0
12:32:40lamedis it allowed on rockbox?
12:32:49Bg3rshould be
12:33:04Bg3ru can make it as macro
12:33:30Bg3r#define abs(x) ((x)>0?(x):-(x))
12:33:52lamedyes, there are abs all over the code :)
12:34:10lamedthanks, i'm such a newbe
12:35:01Bg3reveryone has been a newbie...
12:35:34lamedbut not everyone couldn't have spelled newbie.
12:36:21VoltageXPaul_The_Nerd: hi, have you seen HCl?
12:37:18Paul_The_NerdVoltageX: Very briefly was he here.
12:38:12 Join tempi [0] (n=tempi@unaffiliated/tempi)
12:39:51lamedis everything ok with cvs? i can't cvs diff properly.
12:40:00 Part tempi ("Leaving")
12:40:30VoltageXok, stuck on 24kbit internet at a friend's place now, so I won't be fiddling with rockbox
12:40:31preglowthink i'll just forget spdif monitoring for now
12:44:13Paul_The_NerdAnd after people saw you in the logs and got all hopeful
12:47:45 Join XavierGr [0] (
12:49:44 Part Arcainus ("Leaving")
12:51:16lamedwhat's the alt key for ifps?
12:51:36Paul_The_NerdThe ifps don't really have a remotely final keymapping yet...
12:51:56 Join TeaSea [0] (
12:52:08amiconnlamed: abs() wouldn't be exactly what you want
12:52:13Paul_The_NerdAlso, whether there's a key that can be used to alternate the function of other keys somewhat depends on how the keys work from a hardware perspective
12:52:24lamedamiconn: why not?
12:52:35amiconnIiuc you do't want the value to be non-zero if w < h
12:52:45amiconn...but you want it to be zero
12:52:53amiconn...instead of negative
12:53:29lamedyes. I should have thought about it when i checked what abs does...
12:53:37amiconnRockbox has MIN() and MAX() macros. MAX(w - h, 0) should do the trick
12:54:06lamedisn't there a "positive" macro?
12:55:40lamedmax is fine, but i figured there's a macro to do exactly what i'm trying, isn't it?
12:55:58 Quit VoltageX ("I was using TinyIRC! Visit for more information.")
12:56:48amiconnlamed: MAX() does exactly what you want, and I see no point in an extra macro for 'positive'. MAX() is more flexible
12:57:02 Quit XavierGr (Read error: 104 (Connection reset by peer))
12:57:12preglowamiconn: you know of any specific stuff we can do to cut down binary size for rockbox?
12:58:43amiconnI'll look into that. Perhaps we should use -Os on archos, but that triggers some gcc bugs
12:59:10amiconn(since -Os is essentially -O2 without the size-increasing opts)
12:59:57 Join quobl [0] (n=quobl@tor/session/external/x-920614f44873f573)
13:02:46 Join XavierGr [0] (
13:09:31 Quit Paul_The_Nerd ("Leaving.")
13:11:44MoosSlasheri: red
13:12:44B4gder870 points
13:13:04Moosgreat score : )
13:13:05amiconn_all_ builds red except bootloaders...
13:15:14MoosB4gder: do we know the high score since you introduced this score thing?
13:15:41preglowmine was about 14000
13:15:42preglowBEAT THAT
13:15:43XavierGrbut I don't remember who was it and how much
13:15:53Moospreglow: haha
13:16:05B4gderI think preglow still has the "leader shirt"
13:17:16B4gderone day we'll have to buy him a beer for that feat
13:17:22preglowyes you will
13:17:32preglowi don't even remember what i did wrong
13:17:35Mooshehe :)
13:17:45preglowit was somehow related to the sim trying to use some inline asm, i think
13:19:15Slasheriargh, gorgot to commit the language file :)
13:19:37Slasherigrr, this keyboard
13:19:37preglowas a matter of fact, i just might buy myself a beer for that feat
13:19:45lamedamiconn: lcd-h100.c includes system.h, system.h has #define max, but the compiler still complains about undefined refrence ?
13:20:41linuxstbAre you using "max" or "MAX" ? It should be upper-case.
13:21:23 Join wehn [0] (
13:23:02peturspeaking of "leader shirt".. B4gder, what's up with the RB shirts?
13:23:27B4gderits been slightly delayed, but I'm gonna proceed with them soon now
13:23:41***Saving seen data "./dancer.seen"
13:23:46B4gderpaypal decided they'd give me some trouble
13:24:14B4gderEU money lawndring laws they blame
13:24:31Slasheriah, i had also forgot to ask about those shirts.. writing the mail soon today
13:25:27 Quit SereR0KR ("XChat Aqua")
13:31:53 Quit mtnbkr_ (Remote closed the connection)
13:31:55 Quit Farpenoodle (Read error: 110 (Connection timed out))
13:32:34 Join mtnbkr_ [0] (i=mtnbkr@tor/session/direct/x-7f20c339af1496fe)
13:36:14 Quit TeaSea (Read error: 101 (Network is unreachable))
13:41:58 Join TeaSea [0] (
13:43:55LinusNthe t-shirt thing is purely a money laundry scam :-)
13:44:19B4gderit'll make me rich!
13:45:03B4gderI haven't yet figured out how though
13:45:17*petur thinks it 'll also be a t-shirt laundry thing
13:46:44ashridah* B4gder scrawls "Rockbox Dev" in temporary ink on all his shirts and mails them out with a return, self-addressed stamped postpak
13:50:50 Part XavierGr
13:53:57 Join Jungti1234 [0] (n=jungti12@
13:59:32ravonHCl: Any luck on the dumb-plugin?
14:00:13ashridah* HCl runs
14:00:21ravonHCl: I've been fiddling around with it a bit. It does decode the modules correctly if you run in debug mode and dump to disk. However, playing the data will generate 0.5s of sound and then give a stack overflow.
14:00:22 Join midkay_ [0] (
14:00:23preglowravon: he's got no charger
14:00:28ravonlol, still? :)
14:01:32preglowpetur: you know the coldfire audio interface well?
14:01:37crwlis anyone ever going to try to implement DUMB as a codec?
14:01:40 Quit midkay (
14:01:58preglowcrwl: that would be what they're trying, i imagine
14:02:05crwlpreglow, oh
14:02:06peturpreglow: not very...
14:02:17LinusNpreglow: what do you want to know?
14:02:24crwli suppose that must be painful
14:02:28preglowi want to retransmit the optical input to the uda without having to use software to do so
14:02:39preglowthe coldfire manual says it's possible, but i can't find the details
14:02:53crwlif you have a big module file with big samples and not much memory available to the codec... anyway, that'll be interesting
14:03:06preglowi do tend to go blind after looking at a datasheet too long, though...
14:03:19pregloweven though the coldfire manual is nice enough
14:04:00 Quit RedBreva ("CGI:IRC")
14:04:57 Quit midkay_ ("Leaving")
14:05:49preglowi think i found out
14:06:07LinusNpreglow: set the TXSOURCE to ebuin1
14:06:18preglowthe txsource select field looks promising
14:06:50linuxstbravon: Are you saying that it works in the simulator, but running it on a real device gives you a stack overflow? Or that you only sometimes get a stack overflow on a real device?
14:07:30ravonlinuxstb: The decoding to disk works on the iAudio. I get the stack overflow when the data is PCM-played.
14:07:40preglowLinusN: btw, should i delete apps/pcm_recording.c? it seems to be the old debug recording screen
14:07:47LinusNdo so
14:08:46linuxstbravon: That's the "*PANIC* Stkov dircache" you mentioned earlier?
14:09:01ravonlinuxstb: Yup. And a USB-panic too
14:09:02 Join XavierGr [0] (
14:09:16linuxstbSounds like your code is overwriting memory it shouldn't be.
14:09:29ravonlinuxstb: I'm not sure it renders the sound fast enough. Gonna check how long this 3 second sample takes, but I'm pretty sure it took >8 seconds to render.
14:10:33ravon"Ticks spent: 1419"
14:10:48linuxstbThat's not too bad - assuming you're not using IRAM and haven't made any attempt at optimisation.
14:12:41preglowthere aren't many ways to use iram
14:13:15 Quit wehn ()
14:13:34linuxstbI don't know dumb at all.
14:14:07preglowLinusN: hmmm, seems i'll have a problem, i can't select spdif as a clock source
14:18:17preglowbeh, and i can't set my spdif output to 44.1khz to test if it works with the correct sampling rate either
14:18:24preglowi think i'm doing something else wrong to boot
14:19:16 Part XavierGr
14:23:15 Join RedBreva [0] (
14:33:32 Join mikearthur [0] (
14:33:41preglowthe fact that i don't get any audio anyway is a bit suspicious, of course
14:33:53preglowbut i just have to go out and catch some sun now
14:34:33Aveanyone know what kind of method apple is using for "playback time left estimation" on their side?
14:38:47lamedredundant in tree.h: TREE_POWER_BTN
14:39:43ravon2800 ticks when I switched to 8bit. Uh.
14:41:46B4gderAve: when do they say that about what?
14:47:30Aveapple surely has some time estimate on playback time on the ipod
14:47:42Aveand that it is "approximate"
14:47:49 Join midkay [0] (n=midkay@rockbox/developer/midkay)
14:47:58B4gderwhat playback time? you mean the total time it can play music until the battery runs out?
14:48:33Avebased on the sources, rockbox is using linear interpolation of battery voltage mapped to charge level, and knowing the battery capacity rockbox is using constant current figure for playback time estimation
14:49:07AveI think the current ifdef's cause ipod to use archos values actually
14:51:13B4gderarchoses use the nimh or 2200 defines
14:51:18B4gderthe ipods use the 1300 one
14:53:01Avees yes
14:53:08Avethat is true, but the currect DRAW figure
14:53:24Ave#define CURRENT_NORMAL 145
14:53:31Avethat is what is used for estimation
14:54:18Avebased on ajb 192kbps cbr mp3 playback with no backlight on (this is speculation)
14:54:37B4gderah, true
14:54:49Aveanyways I'm trying to find better figures for my ipod nano
14:55:10AveI ran battery benchmark once but the results did not get written to disk before battery ran out
14:55:26 Nick kyomi|off is now known as Kyomi (
14:55:54KyomiAnyone know why they went with tinySID instead of something like reSID?
14:56:09B4gderKyomi: feel free to do that and we'll pick the best one! ;-)
14:56:16KyomiSome of my PSIDs sound off and it only plays like barely half of the HVSC
14:56:49KyomiI know it's like... in the early stages but... I cant play most of the ones I wanna play ;_;
14:57:16B4gderI view it like this:
14:57:21B4gder1) we had no SID player
14:57:24 Join SereR0KR [0] (
14:57:25B4gder2) now we have a patch for it
14:57:40B4gdersure, you can probably improve it heaps
14:57:51KyomiNow we need a MIDI player :P
14:57:51B4gderbut talking is not gonna make a diff
14:58:29KyomiOn the Doom I/II thing... is there something that could be coded for strafe left/right in the set keys part?
14:58:39 Quit ashridah ("Leaving")
14:58:49KyomiIt'd make everything a WHOLE lot easier
14:58:52linuxstbKyomi: Are you running the latest version of the SID patch? There's been two or three updates since the initial version.
14:59:09KyomiI've got 1.1
14:59:19KyomiWhatever is in the experimental build
14:59:29linuxstbCheck the patch tracker - I think there is an improved version.
14:59:29KyomiSince yesterday
14:59:30ravonI think 1.2 is the latest.
14:59:45KyomiI dont compile my own...
14:59:49KyomiIt scares me ^^;
14:59:56B4gdercompiling is Good For You
15:00:16LinusNKyomi: resid is not easy to port to rockbox because:
15:00:21linuxstbThen you're only option is to hassle the person making the experimental build to upgrade the patch.
15:00:25LinusN1) It is written in C++
15:00:31KyomiI think my H320 battery is dying
15:00:32LinusN2) It is CPU hungry
15:01:39KyomiReason being.. I go in and out of Doom like 3 times to reset my settings to what I want then play a couple SIDs and it turns itself off
15:01:43B4gderwhy would you write an emulator in C++?
15:01:51KyomiAnd this is all in an hour
15:02:13linuxstbI think the excuse is to "model the hardware"...
15:02:38linuxstbpacbox was originally C++
15:02:42 Join goffa [0] (n=goffa@
15:03:09 Quit goffa (Client Quit)
15:03:12linuxstbBut I couldn't see any advantages to it.
15:03:17 Join Febs [0] (
15:04:21 Join goffa [0] (n=goffa@
15:04:27goffathat sucked last night
15:04:41goffapower failure started at like 8:30 pm
15:04:50goffaended at 5:20 am
15:05:14 Join XavierGr [0] (
15:05:23goffamy x5 had a pretty low battery
15:05:31 Quit XavierGr (Read error: 104 (Connection reset by peer))
15:05:36goffabut was good until 3 am or so :) (dozed)
15:06:15 Join XavierGr [0] (
15:06:22 Quit SereR0KR ("XChat Aqua")
15:07:13FebsWho is "Neon John"?
15:07:26goffanot bad
15:07:36goffadoesn't have a coming next from what i see
15:07:49B4gderFebs: long time rockbox mailing list member
15:07:50preglowa guy who hangs on the mailing list
15:08:16B4gderJungti1234: nice indeed
15:08:23ZagorJungti1234: very nice
15:09:25FebsI see that he has taken it upon himself to write to the guy claiming copyright infringement.
15:09:53B4gderyes, he mentioned that in his mail
15:09:53ZagorFebs: only personally, not on behalf of the project
15:09:55FebsRead his latest message to the developers' list.
15:10:01preglowi didn't bother to read all of it :>
15:10:03FebsIt's a bad idea nonetheless.
15:10:23Zagorhe's not the first
15:11:00FebsI love the way he criticizes "internet lawyers." What a hypocrite.
15:11:13Zagorhehe, he's quite a character :-)
15:12:24preglowthis isn't a bloody big deal
15:12:35preglowcan't people just be nice to each other!!
15:13:12KyomiI <3 that name
15:13:16KyomiMiika :3
15:14:14preglowSlasheri: you've got yourself an admirer now
15:14:21preglowthis is what being a rockbox coder gets you!
15:14:25preglowit's a glorious life
15:14:35B4gderfame and glory!
15:14:48B4gderfans outside your house screaming
15:15:21B4gderchanting *highscore* *highscore*
15:15:22*goffa lives too far away to attend screaming parties outside of coders houses
15:15:38goffabut i probably would :)
15:15:53preglowi need to go out and speak to my followers from time to time
15:16:00preglowto calm them down and prevent them tearing down the walls
15:16:59goffaand the women...
15:17:12B4gderwhat's that?
15:17:26 Join possie [0] (
15:17:46lostlogicB4gder: Rockbox should buffer the first song, and some seconds of the next song by default in the first buffer run. (really? I was under the impression this was very much debated and that there's no consensus about it being a good idea or not / DanielStenberg) (Currently it buffers only the first song, at least 1 meg, up-to the size of the filebuf / 4, this I don't think is the best option, first song plus X of next song, or full fill would b
15:17:46goffayou mean you aren't fighting them off with a stick? too many to handle
15:18:13preglowi hate that thing
15:18:17preglowi want it to do all the buffering at once
15:18:40B4gderyes, what is the point in stopping the buffering?
15:19:10LinusNthe idea was that rockbox was unresponsive during buffering in the early swcodec days
15:19:20lostlogicI didn't write it in the first place, I just made it smarter this time around
15:19:35possieif i install rockbox on my ipod G4, will i still be able to sync it with iTunes?
15:19:49lostlogicI'm _more than_ happy to turn it back to full buffer run at first, although the alst time I did that, one or two people on the forums bitched
15:19:59preglowas they always do
15:20:06B4gderlostlogic: why did they whine?
15:20:13B4gderI mean, what was the effect they didn't like?
15:20:30lostlogicbecause it wastes energy buffering at first when they might be fools and skip out of that song quite soon
15:20:37ravonJungti1234: I want those skins!
15:20:49preglowwe should probably implemt the ipod buffering scheme for people like that
15:21:01Jungti1234ravon: hehe
15:21:10preglowthough i suspect it'll be tricky
15:21:19preglowperhaps it'll be easier when we get metadata on buffer
15:21:54lostlogicpreglow: acutally no.
15:22:04B4gderlostlogic: but buffering the next song is also a waste if you don't skip, right?
15:22:06lostlogicpreglow: if we're going to do that we want to keep the current structure of two separate ring buffers
15:22:10lostlogicpreglow: and make the tracks one bigger
15:22:38lostlogicB4gder: uhhh ya mean if you never get to listenign to it
15:23:08B4gderno, I mean that if I have a 10MB song and you first load 7.5MB of it, Where would you load the next song?
15:23:15lostlogicpreglow: I'd imagine we buffer X of each of the next Y songs into the smaller ring buffer, with a jump pointer into the main filebuffer
15:23:21B4gderwould you leave a 2.5MB gap?
15:23:33B4gderI would guess you don't
15:23:38lostlogicB4gder: no, if the first song is that big then none of the next is loaded
15:23:42***Saving seen data "./dancer.seen"
15:23:53Mooswasn't this first buffering track a feature?
15:24:12lostlogicMoos: buffering only first track was a feature that was not liked by all, so we're deciding whether to keep it
15:24:12B4gdernow I see what your item in that page means
15:24:40t0mashm? I heard something about buffering the start of 3 tracks to make skipping faster?
15:24:52t0masthe current track, the next and the previous?
15:24:56Mooslostlogic: ok, but what do we loose if we just buff the 1st track then all removing?
15:24:59B4gderlostlogic: sorry, I thought you were referring to another kind of "feature" where you'd _always_ load a piece next song
15:24:59lostlogict0mas: people want it, I don't want to do it.
15:25:06lostlogict0mas: I dunno, something like that
15:25:08t0masah ok
15:25:08B4gderpiece of
15:25:33lostlogicB4gder: ah, k, well anyway, do I get to kill the initial short fill 'feature' or should I leave it be?
15:25:44B4gderimho we should kill it
15:25:50lostlogicI think that it's unneeded complexity, but I tend to turn my player on and ignoer it
15:26:00preglowand this time ignore the whiners
15:26:05preglowtell them to speak to me
15:26:05lostlogicyour wishes are my command, gentlemen.
15:26:07preglowso i can ignore them
15:26:10Moosbut what you loose man?
15:26:17MoosI canot understand
15:26:27lostlogicwe lose some code complexity
15:26:34B4gderwe gain simplicity
15:26:41Moosah ok
15:26:41B4gderand some battery ;-)
15:26:46Mooshehe : )
15:26:47lostlogicand possibly a few seconds of battery life due to doing a longer buffer fill which may turn out to be partly wasted at first
15:26:59lostlogicbut we possibly gain some battery for 'normal' people
15:27:02Moosok understood
15:27:04lostlogicwho load music that they want to listen to
15:27:32crashdseconded, who are these people who have playlists filled with tunes they dont like?
15:27:33lostlogicanywho, I'll do that next time I have voice working, voice is currently getting a KISS makeover too, so that could be a day or two.
15:27:40lostlogiccrashd: a wonderful question.
15:27:55crashdi just dont get it ;\
15:27:57MoosB4gder: btw, while you are here, what's about the X5 LCD and your lasr changes?
15:28:12B4gdera bug
15:28:20B4gderfix is pending
15:28:22Mooshihi :)
15:28:25B4gderneed to verify at home first
15:28:45Moos+some asm optimisation from amiconn too
15:29:01Mooshopefully at least : )
15:29:03B4gderyes, that'd be great
15:29:43B4gdera 10000 lap loop benefits from every little cycle we can remove
15:30:23*Moos go back working
15:32:53 Join warthawg [0] (
15:34:47 Part LinusN
15:36:49lamedhelp?! why am i getting ../tools/configure: permission denied on vmware when i'm logged in as root?
15:39:24crwlchmod +x ../tools/configure
15:39:47*goffa hopes his battery lasts until 11am on 1 hour of charge time
15:39:58goffaits 7:40 now
15:40:22lamed10x crwl
15:41:16 Quit lamed ("CGI:IRC (EOF)")
15:52:50 Quit Jungti1234 ()
15:57:15 Join bluebrother [0] (
16:00:29 Quit Siku ("reboot")
16:04:03 Join nomi [0] (
16:06:50 Quit nomi (Client Quit)
16:09:34 Join Siku [0] (n=Siku@
16:12:11lostlogicblah, voice does stuff that says "I'm working" but only plays blank data afaics :(
16:14:30goffajust tell it to shut up and get to work :)
16:14:38 Join _Lucretia_ [0] (
16:15:48 Quit XavierGr (Read error: 110 (Connection timed out))
16:17:05amiconnAve: Imho it doesn't make sense to adjust current draw figures before serious optimisation of the codecs
16:18:33 Join hardeep [0] (i=hardeeps@SDF.LONESTAR.ORG)
16:22:14 Join solexx [0] (
16:23:57preglowB4gder: can't we just remove the SDL option now?
16:24:03preglowdo the other sims even work anymore?
16:24:06B4gderyes we can
16:24:28B4gderI just need to coordinate the removal with a massive edit of automatic scripts
16:24:58preglowforgot that
16:27:33 Nick mtnbkr_ is now known as mtnbkr (i=mtnbkr@tor/session/direct/x-7f20c339af1496fe)
16:34:28 Quit solexx_ (Read error: 110 (Connection timed out))
16:36:28preglowwhat is the audio clock on the irivers?
16:37:02 Join XavierGr [0] (
16:38:17bluebrotherI think it would be nice having the configure script configureable −− like calling it with "-t<targetnumber> -b<buildtype>" etc.
16:38:37B4gderI agree
16:42:36 Join [TCK] [0] (
16:44:12 Quit _Lucretia_ ("Leaving")
16:46:28amiconnB4gder: If you did not yet implement a fix for the broken partial updates on X5 - don't bother (or just revert for now)
16:46:44B4gderok, sure
16:46:51 Join Daishi [0] (
16:47:03 Join webguest45 [0] (
16:48:16 Join Farpenoodle [0] (
16:50:37 Quit Febs ()
16:52:47preglowbluebrother: how should we refer to other settings? i've seen both \emph and `` ''
16:53:25bluebrotherpreglow: I already noticed that too. I was thinking if it would be better to have a macro for formatting all settings.
16:53:31preglowbluebrother: i agree
16:53:57preglowi think i rather prefer \emph formatting
16:54:02bluebrotherThat way we could think about using small caps, quote marks, \emph and change it easily.
16:54:17preglowdoes the standard latex font have proper small caps?
16:54:20preglowor simulated ones?
16:54:33preglowfontS, i mean
16:54:43bluebrotherThere are also some occurences like "if blah is YES" which I also don't like.
16:54:52preglowme neither
16:55:02bluebrotherI'm not sure on this.
16:55:22preglowbut i'd like to use small caps more
16:55:34preglowperhaps button names should be in small caps
16:55:44preglowmakes sense typographically
16:56:26bluebrotherI'd also like small caps for stuff like that better. But I'd like it even better having a new macro ;-)
16:56:26 Quit webguest45 ("CGI:IRC (EOF)")
16:56:36bluebrotherI think I'll do one later this day.
16:57:17bluebrotherWe just need to make sure we don't flood ourselves with macro
16:57:30preglowwell, i don't think we need to
16:58:03preglowbut for essential stuff like buttons and settings, we do need to
16:58:24bluebrotherI'd like to have the button macros a bit more generic.
16:58:30preglowas in how?
16:58:48bluebrotherLike a \Stop which expands on every target to the button that usually stops.
16:59:15bluebrotherafaics all targets have something like up, down, left, right and play, stop
16:59:16RedBrevaI'd vote for that one!
16:59:30bluebrotherespecially stop is annoying.
16:59:46B4gderthe x5 has no dedicated stop button
16:59:55bluebrother\opt{h1xx,h300}{\ButtonOff}\opt{archos}{\ButtonStop} etc.
17:00:05bluebrotherhow is it working on the x5?
17:00:08lostlogicamiconn: does anything come to mind that would cause the talk code to be sending blank voice clips (zeroed memory area) to the voice system? I'm getting some truly weird behavior on my local version :(
17:00:09preglowi think we might need to use opt for this a ton of places anyway
17:00:19preglownot all screens have the same possibilites on all platforms
17:00:36bluebrotherI was thinking to replace this with a \Stop that expands to the line above.
17:00:57preglowwell, if you find instances were it would be nice, just do it
17:01:05bluebrotherunfortunately I haven't had time looking in other targets than iriver more deeply.
17:01:24 Quit ravon ("Client exiting")
17:01:31bluebrotherhow is stop handled on x5? something like "always long press rec"?
17:01:40hardeeplostlogic: around?
17:02:00B4gderbluebrother: I don't even remember how to stop playback ;-)
17:02:09amiconnlostlogic: The talk module is supposed to send _one_ blank clip (not zeroed memory, but a real digital silence clip of ~0.3 seconds) at the end of a clip sequence
17:02:27bluebrotherI guess I need to play around with the sime a lot more ...
17:02:50amiconnThis is to avoid swallowed syllables, because the MAS doesn't play the very last frames of a stream
17:03:57goffaouch.. just read this on fark
17:04:02goffaglad i don't live in norway
17:04:11goffaNorwegian oil companies and goverment makes money hand over fist as oilprices soar above 72 USD/barrel. In other news, gasprices in Norway expected to soar to 8.80 USD/gallon this summer
17:04:22RedBrevaDo H3xx and H1xx needIRIVER_H100_PAD and IRIVER_H300_PAD... The buttons functions are the same, and it is a quite a verbose option to type, can't we just have IRIVER_Hxxx_PAD or similar, and be player specific (H1xx or h300) if needed?
17:05:05bluebrotherafaik h1xx and h3xx are identical except the naming −− h3xx has the "Navi" key while the h1xx it's called "select"
17:05:27bluebrother... which could be unified a bit.
17:05:41amiconnIn some cases it's desirable to assign buttons different on H1x0 and H300
17:06:10amiconnThe available buttons are the same, but their physical placement is different
17:06:35RedBrevaNo probs \opt{h1xx}{some button} \opt{h300}{different button} if that is needed!
17:06:47bluebrotheramiconn: I agree with that, but in most cases it's not necessary. So it would be easier to have a shorthand for the case it's identical
17:07:08bluebrotherI was thinking of something like \opt{iriver}{\ButtonStop}
17:07:24crwlgoffa, hey, that sounds expensive even compared to our current ~6.4 USD/gallon prices (if I did my maths correctly)
17:07:54*goffa thinks $3 is way too high
17:07:59B4gderM-x c-indent-line-or-region
17:07:59lostlogichardeep: at work, so yes around, but also slow.
17:08:03lostlogicamiconn: hmm, interesting.
17:08:03goffaits about $2.70 now
17:08:16RedBrevaI live that, but I would also live a common 'PAD' name... I don't really like mixing XXXX_PAD and dap names in one \opt - it looks untidy
17:08:52hardeeplostlogic: i'm not sure if you saw my comment in the irc logs but i found a small problem in the playback code...
17:09:04goffaacording to, the highest gas in the state is 2.75
17:09:04hardeeplostlogic: audio_play() doesn't set the "playing" status immediately
17:09:18lostlogichardeep: that's a problem?
17:09:20goffathat's 50 mi from here
17:09:23hardeepthis is different then archos behaviour and breaks some functionality in the playlist viewer
17:09:27hardeepand possibly other places
17:09:38lostlogichardeep: gotcha, so on archos, it sets playing true before the event is processed on-thread?
17:09:48lostlogichardeep: ok
17:10:00crwlgoffa, well, I'm not against high gas taxations and whatever so I don't really care
17:10:20lostlogichardeep: anything else you find, please let me know, or fix it yourself −− the playback code should be less fragile now, and so safe to edit ;)
17:10:44hardeeplostlogic: okay =) didn't want to mess with it while you were still making changes
17:11:33 Join Kohlrabi [0] (
17:11:54lostlogichardeep: I'll be working in the voice areas until I get voice re: working
17:12:31 Quit B4gder ("time to say moo")
17:18:17preglowbluebrother: btw, you think we should use the \unit command when we need to use units? Like \unit[500]{Hz}
17:18:37preglowit'll properly format it for us
17:18:39lostlogicWhy in the name of the FSM is voice never calling its advance_buffer callback!/
17:19:01preglowlostlogic: it has developed a taste for stronger liquors
17:19:05bluebrotherpreglow: haven't had the idea ... but it sounds good.
17:19:42lostlogicpreglow: :( seroiusly, I have a ci_voice that I feed to the voice codec and tell it "damnit, call voice_request_buffer and voice_advance_buffer" but it seems to be calling codec_advance_buffer instead :(
17:20:01preglowbluebrother: at least we'll have proper formatting
17:20:07preglowbluebrother: and it works in math mode as well, if we'll ever need that
17:20:36lostlogic:( was doing exactly what I told it to. I hate how programs do that.
17:22:53 Join _FireFly_ [0] (
17:23:16 Quit RedBreva ("CGI:IRC")
17:23:39 Join RedBreva [0] (
17:23:44***Saving seen data "./dancer.seen"
17:25:06 Join _Lucretia_ [0] (
17:25:16lostlogicamiconn: how does it insert that 0.3 seconds of silence? an additional clip, or appended through the get_more callback?
17:25:20preglowanyone know of vim settings that will automatically break lines while i edit?
17:25:38 Join solexx_ [0] (
17:25:40lostlogic(voice now plays again, but slightly more broken than before, simpler code though)
17:25:49lostlogicpreglow: textwidth
17:25:58lostlogicie set textwidth=72
17:26:37 Quit ProgramZeta (Read error: 110 (Connection timed out))
17:27:32bluebrotherpreglow: :set tw=76 (or any other value you want to use)
17:29:59preglowdoesn't work exactly like i want it to
17:30:32preglowit doesn't reformat when i delete text, and so on
17:31:33lostlogicpreglow: I think there are some plugins for autocmds to reformat but it's not supported by vim internally afaik
17:32:08 Join solexx__ [0] (
17:36:36 Join MusiFreq_ [0] (
17:37:08amiconnlostlogic: It's an additional clip that gets appended whenever the talk queue runs empty
17:37:27amiconnThis silence clip is also used during spelling, for space characters
17:37:51amiconnTake care not to break vcice for hwcodec...
17:37:56 Quit RedBreva ("CGI:IRC (EOF)")
17:39:41 Quit _FireFly_ ("Client exiting")
17:40:00 Quit solexx (Read error: 110 (Connection timed out))
17:40:46 Quit Lynx_ (Read error: 104 (Connection reset by peer))
17:40:49preglowwill anyone cry if monitoring is not enabled for spdif recording?
17:40:56lostlogicamiconn: I won't touch talk.c, I will just make swcodec work with it.
17:41:03preglowit seems i actually have to record the incoming sound, then potentially resample it
17:41:06preglowthen playing it backm
17:41:40lostlogicamiconn: waht does hwcodec do if you select a menu item, and then select another before the voice is done playing the first?
17:41:44 Join Lynx_ [0] (
17:41:55preglowlostlogic: preempt it, i think
17:41:59preglowcut it straight
17:42:02preglowas it should do
17:42:06lostlogicyes, I figured that, but _how_
17:42:41preglowby just starting to send some other mpeg data instaed of the current mpeg clip?
17:42:48preglowor are you talking lowlevel shiznit?
17:43:11lostlogicpreglow: I'm talking what does talk.c say to the hwcodec and or swcodec to ask it to stop
17:44:06 Quit solexx_ (Read error: 110 (Connection timed out))
17:44:39lostlogicpreglow: what color is your skin, and what country are you in? :-P
17:45:23 Quit Zagor ("Client exiting")
17:45:51preglowstraight out oslo gangsta
17:45:57preglowwhite as milk
17:46:29preglowi even got blondish hair to boot
17:46:41petur14.33.53 # <preglow> but i just have to go out and catch some sun now
17:46:49peturnot enough it seems
17:47:25preglowok, i may have exaggerating
17:47:28preglowbeen too
17:48:17preglowsome days in the mountain has even got me a wee nice tan
17:48:53preglowjust listened to doggystyle a couple of hours back, it seems to be festering in my brain
17:50:26 Nick mtnbkr is now known as mtnbkr_ (i=mtnbkr@tor/session/direct/x-7f20c339af1496fe)
17:50:59 Nick mtnbkr_ is now known as mtnbkr (i=mtnbkr@tor/session/direct/x-7f20c339af1496fe)
17:51:09 Quit MusiFreq (Read error: 110 (Connection timed out))
17:51:26 Quit [TCK] (Read error: 104 (Connection reset by peer))
18:01:32 Quit hardeep ("BitchX: so real, you'll wet yourself!")
18:04:07sharpeokay, this can't be correct, -8806939% percent emulation...
18:04:45peturso much you overflowed it ;)
18:05:02sharpei'm just that good at programmming an emulator.
18:05:47 Quit possie ("Leaving")
18:08:01sharpewell, some improvement, now it's at -984165%
18:08:24 Join dpro [0] (
18:15:33 Quit petur ("home << me+bike")
18:26:51 Join mikolas [0] (
18:32:26sharpe154% realtime, 14fps...
18:32:45 Quit RJ (Nick collision from services.)
18:32:52 Join RJ [0] (
18:33:26 Quit RJ (Nick collision from services.)
18:33:34 Join _rj [0] (
18:34:50 Part solexx__
18:35:42goffawoot on the speed :)
18:42:38 Join lamed [0] (
18:43:55lamedhi everyone, can someone compile & check an ondio simulator for me with a patch?\
18:44:51lamedfor some reason i'm able to build an vmware simulators but they don't show up like they should; they just get the vmware screen all vague.
18:45:00 Join ravon [0] (
18:46:43lamed... i've made the horizontal scroll work on repeated button press and I have to check if it's ok on an ondio
18:49:28lamed... or could someone offer me a hand dealing with the vmware bug?
18:50:41*lamed wonders why he keeps talking to himself
18:51:33 Join Evilnipple [0] (
18:52:05*lamed knocks his brand new tft screen and cable modem. they're not saying anything.
18:52:24dprolamed: hehe sorry ... no answer in my caser simply means "no I can't"
18:52:52lamed: /
18:53:42*lamed hates the new sdl sim
18:53:58 Join new5guser [0] (
18:56:10 Join stephano [0] (
18:59:28 Join wefds [0] (n=muesli_t@
19:02:10 Join pixelma [0] (n=M_@
19:02:52sharpeeh, you get used to talking to yourself.
19:03:07new5guserHi! I've been installing the dailies on my new ipod 5g for that week or so, and I just gotta say Thanks. You'all are fantastic!
19:03:40sharpeanother happy rockbox user :D
19:05:23 Join sandoaway [0] (n=lolsteam@
19:06:36 Part stephano ("Leaving")
19:06:40new5guserIs there a tool to update the tagcache off-line? ie. after I copy music and the player is still in disk-mode
19:07:20sharpebut that is an interesting idea.
19:07:47new5guserif it might be usefull i'll try hacking something together tonight
19:07:58sharpeheh, go for it
19:08:34new5gusermy concept is that it could also do the itunes db for music that can be played in the original firmware
19:08:41 Quit lamed ("CGI:IRC (EOF)")
19:08:49new5guseralla foo_pod
19:09:24sharpewell, if you need it, the fileformat for the tagcache is in /apps/tagcache.c
19:09:48sharpealong with functions to build it...
19:09:54 Quit muesli__ (Read error: 110 (Connection timed out))
19:09:57 Join muesli__ [0] (n=muesli_t@
19:10:14 Join cismo_ [0] (
19:12:37sharpethis is odd...
19:13:16Slasherinew5guser: if you want to implement an utility to build the tagcache offline, you could use the tagcache engine code directly for that
19:13:33Slasheriit shouldn't depend too much about firmware or anything like that
19:13:59 Quit cismo (Read error: 110 (Connection timed out))
19:14:42new5guserSlasheri: sharpe had suggested that as well. It sounds like the best way to go. I'm looking throught that code now
19:15:02Slasherisounds good
19:16:35 Join webguest29 [0] (
19:17:30sharpewell, i know how to slow it down to 10% realtime emulation speed...
19:17:47Slasheriyou will just need to call the following functions in this sequence to build the db: build_tagcache(); allocate_tempbuf(); commit(); free_tempbuf();
19:18:10Slasheriand do some minor #ifdeffing to the code to make it work stand alone
19:19:11new5gusergreat! now it doesn't seem like such a daunting task
19:19:23Slasherihehe :)
19:19:53sharpeallocate_tempbuf() in the standalone tagcache updater, could just use malloc...
19:20:17Slasheriyep it could, or just statically allocate a buffer of few megabytes
19:20:40sharpebut malloc is cooler to say that static.
19:20:48webguest29I'm thinking of buying a logic analyzer, do you have any recommendations?
19:21:03Slasheriyeah, then just malloc the few megabytes :D
19:21:17 Quit sando (Read error: 113 (No route to host))
19:21:17 Quit wefds (Read error: 110 (Connection timed out))
19:22:23sharpeokay, i know how to slow it down to 0 % too...
19:22:50Slasheriand the threading stuff should be ifdeffed completely out (as dircache & ramcache stuff too, but that will be automatically out)
19:23:01new5guserI'm still at the office, so I have to be productive for a while. I'll jump back on this channel tonight (EST) if I make any progress. Thanks sharpe and Slasheri
19:23:18sharpeyou're welcome
19:23:45***Saving seen data "./dancer.seen"
19:23:48 Part new5guser
19:24:07 Quit xmixahlx ("blah blah blah")
19:24:43 Join xmixahlx [0] (n=xmixahlx@
19:27:00 Join lamed [0] (
19:27:18lamedI'd love if someone takes a look :
19:27:24sharpei don't see how i'm only getting 14fps, when it should only be updating info on the 25th frame...
19:27:56sharpeyou used the wrong 'affected' :)
19:28:30 Join raf82 [0] (
19:28:36lamedwhy ho why can't you edit on FS?
19:29:36lamedregarding on that line, #, can _you_ build an run an ondio sim?
19:30:14lamed# = sharpe with a spelling mistake
19:30:23 Join wefds [0] (n=muesli_t@
19:30:49sharpegive me a minute.
19:31:00sharpeand i may try
19:31:03webguest29Thinking about the USBee AX ( which has 8 digital channels (maybe too few?) and one analog channel for oscilloscope function.. it also has a digital signal generator.. hm
19:31:16 Quit muesli__ (Read error: 110 (Connection timed out))
19:34:24 Join obo [0] (
19:36:25 Join mtnbkr_ [0] (i=mtnbkr@tor/session/direct/x-2aa9f9bbe80f1def)
19:36:40sharpeokay, so, you want me to build an ondio sim?
19:37:54 Join Ribs [0] (
19:39:20 Quit mikolas (".")
19:39:28 Join mikolas [0] (
19:41:14 Join PaulJ [0] (
19:42:23 Quit lostnihilist (Read error: 110 (Connection timed out))
19:43:38 Join muesli__ [0] (n=muesli_t@
19:43:47 Join hardeep [0] (i=hardeeps@otaku.freeshell.ORG)
19:44:13lamedsharpe: yeah, it needs to be tested, whether left, right, and menu+left/right works on it.
19:44:31lamed(on file view/playlist view or menu)
19:44:33 Quit lamed ("CGI:IRC")
19:45:18 Join Philip_0729 [0] (
19:46:56Philip_0729how would i go about compiling a bootloader?
19:48:00 Join Farpnut [0] (
19:48:21sharpewhen you run ./tools/configure to create the makefile, choose B for Bootloader
19:48:45Philip_0729i did that but then i run make
19:48:45sharpeif you have a build enviornment set up...
19:49:08hardeepSlasheri: minor (strange) bug with your latest changes −− the "All tracks" option correctly appears at the top of the list in all cases except under Genre where artists with numbers appear first (e.g. Genre->Rap->50 Cent)
19:49:41Philip_0729then make zip?? < that gives be a zip with a .rockbox folder and rockbox.iriver file.... not bootloader patcheer
19:50:31Slasherihardeep: ah, the sorting displaces it.. hmm, needs to be fixed somehow
19:51:18hardeepSlasheri: also, the bug that amiconn mentioned is still there −− chunking isn't done when starting a playlist so only the first max_files_in_dir tracks are inserted
19:51:54Slasherihmm, are you sure that is a tagcache related problem?
19:51:56hardeepSlasheri: I was just surprised that it didn't happen anywhere else re: sorting
19:51:56bluebrotherPhilip_0729: why do you want to build the bootloader?
19:52:13Slasheribecause it should use the chunked browsing in that situation too..
19:52:35Slasherihardeep: not all main lists are sorted
19:52:46Philip_0729to make it have a black background and white text... fit better with theme
19:53:04sharpelamed: you need to know if left, right, and menu+left/right works in the file view?
19:53:34bluebrotherbut you are aware of the fact that you can brick your player if the bootloader has errors?
19:53:45 Quit MusiFreq_ (Remote closed the connection)
19:53:47 Join MusiFreq [0] (
19:53:59Philip_0729yes i do not intend to do anything if i get any errors
19:54:10hardeepSlasheri: regarding second issues, it looks like it's tagcache related
19:54:23hardeepthere's no limit on number of tracks that can be inserted
19:54:38Slasherihardeep: hmm, i will make more checks regarding that issue then
19:54:42hardeeper, actually, you're limited to max_playlist_size which is much larger
19:54:45 Quit wefds (Read error: 110 (Connection timed out))
19:56:03 Quit Farpenoodle (Read error: 110 (Connection timed out))
19:57:05sharpelamed: they seem to work...
19:59:00Slasherihardeep: just verified and it's indeed a tagcache related problem, weird
20:00:20 Quit quobl ("Leaving")
20:00:32 Part raf82
20:00:45Slasheriah, hehe.. it overwrites the search struct..
20:00:49 Join raf82 [0] (
20:02:53 Quit bluefoxicy ("Ex-Chat")
20:03:50 Quit muesli__ (Read error: 110 (Connection timed out))
20:04:14 Join PuGz [0] (
20:04:30PuGzhey guys. i am in rockbox on my ipod 5g video. i entered the solitaire game.... how do i exit?
20:04:46 Join muesli__ [0] (n=muesli_t@
20:04:48sharpeshould be select_menu
20:04:52 Join archos-fmrecorde [0] (
20:05:24PuGzsharpe: i swear i tried that and it didnt work - but it did just then!
20:05:36PuGzi feel stupid
20:05:55PuGzi just installed ipod linux and the bootloader etc on my 5g ipod.
20:06:06sharpeheh... i've just rockbox on mine
20:06:13PuGzonce you have done that, installing rockbox is as simple as extracting the zip file to the root directory - done!
20:06:19PuGzsharpe: nice
20:07:11archos-fmrecordeHey all, I was about to embark on adding timed radio recording support into rockbox. Target is archos FM recorder, but I'll try to keep it general enough for other models to be supported too. Anyone know why this feature hasn't been added before, or have any comments/concerns/advice?
20:08:08sharpeinteresting idea. :)
20:08:28PuGzsharpe: ipodlinux has troubles loading mpd on my ipod atm... so i cant playback music! apparently this is a common thing on the 5g versions atm. so its good to play around with, but still need rockbox for music!
20:08:37PuGzsharpe: do you know if the itrip can be used with rockbox?
20:08:59PuGzsharpe: cause i tried using it and it didnt work. but i just found the car adapter setting for rockbox then... that might make it work?
20:08:59sharpePuGz: one sec
20:09:15 Join petur [0] (n=petur@rockbox/developer/petur)
20:09:22sharpewell, the car adapter is for the charger i believe.
20:12:28sharpeipod peripherals aren't supported in rockbox...
20:12:58PuGzbah - where's is say that?
20:12:58 Quit archos-fmrecorde ("CGI:IRC (EOF)")
20:13:05sharpeyou could however, get a normal fm transmitter
20:13:10PuGzyeah... i know
20:13:15PuGzi was trying to save the $$
20:13:28PuGzcause i already got a 5G itrip
20:13:33PuGzand it cost me $70
20:13:42 Join ep0ch [0] (
20:13:49PuGzi didnt know anything about ipodlinux or rockbox at the time
20:13:53sharpeeh, they're usually around 20$, can probably get one for 10-15...
20:14:34sharperegular fm transmitters
20:14:35PuGzi am in Australia. and they all cost about $80 in stores here
20:14:41PuGzi got mine on ebay for $70
20:14:45PuGzoh - regular ones
20:14:47 Join wefds [0] (n=muesli_t@
20:17:38webguest29does anyone know what make/model of LA that Linus have?
20:17:42PuGzcan get one off ebay for $7.95 and $8.95 postage. takes 2xAAA batteries or can use car charger. 4 preset channels to choose from. 30 feet transmit distance. led light indicators
20:17:58PuGzwebguest29: Linus?
20:17:58*t0mas is now playing: American Heartbreak - Things Are Looking Up
20:18:03t0masgreat song
20:22:30PuGzgotta goto bed
20:22:32PuGzcya guys
20:22:56 Quit muesli__ (Read error: 110 (Connection timed out))
20:24:28 Join muesli__ [0] (n=muesli_t@
20:30:27 Join muesli- [0] (n=muesli_t@
20:33:34 Part PaulJ
20:34:15 Nick Kyomi is now known as Kyomi\off (
20:34:22 Nick Kyomi\off is now known as Kyomi|off (
20:36:20 Quit wefds (Read error: 110 (Connection timed out))
20:38:59 Join webguest59 [0] (
20:39:53 Part webguest59
20:42:47 Quit muesli__ (Read error: 110 (Connection timed out))
20:46:50 Quit warthawg ("my work here is done")
20:47:12 Join ArrowHead [0] (
20:47:19 Part Philip_0729
20:47:31ArrowHeadI was wondering if there are any plans on porting RockBox to Creative players?
20:48:05phaedrus961Bagder: when you get a chance, mini 1g is ready to be added to daily builds
20:48:34 Join muesli__ [0] (n=muesli_t@
20:49:07amiconnlostlogic: talk.c tries to preemt the current clip when a new menu item is selected while the old clip(s) are still playing
20:49:52ArrowHeadI'll take that as a "no"... ?
20:49:56amiconnAs the MAS would produce a dropout while resyncing to the mp3 stream, we can't just cut the old clip at an arbitrary position and feed the new one
20:50:28 Join Rondom [0] (
20:50:55amiconnSo, talk.c stops DMA, scans for the next frame header, cuts the size of the currently playing clip down so it ends at the next frame boundary, then restarts dma and lets the frame end
20:51:34amiconnThis frame header scan is of course mp3 specific, and cannot work as-is on swcodec anyway, since it scans for the *bitswapped* frameheader
20:52:05 Part PuGz
20:52:05amiconnThis is why I asked whether it's possible to reset libamd and force it to resync
20:52:35 Quit mtnbkr (Nick collision from services.)
20:53:04 Nick mtnbkr_ is now known as mtnbkr (i=mtnbkr@tor/session/direct/x-2aa9f9bbe80f1def)
20:53:14sharpewell, i can get ~ 180% emulation...
20:54:41amiconnhardeep: This bug (not adding everything but just the first max-files-in-dirbuffer tracks) also existed with the old tagdb
20:56:37 Join Lear [0] (
20:58:19 Part ArrowHead
20:59:07 Join Farpenoodle [0] (
20:59:08 Quit muesli- (Read error: 110 (Connection timed out))
20:59:11 Join tvelocity [0] (
21:02:15amiconnlostlogic: I'm almost sure talk.c needs fixing too.
21:02:56obo_rj: around?
21:03:08lostlogicamiconn: hmm, indeed... there seems to be an extra layer of indirection in buffer fetches.
21:03:52amiconnIt's not the extra indirection, but I suspect at least some slight bugs in talk.c which just didn't show up on hwcodec
21:04:08lostlogicamiconn: mad requests data from playback via (new function in my local version) voice_request_buffer_callback(), which if it doesn't have any buffer requests new data from talk.c via voice_get_more()
21:04:12amiconnThere's a fix on the tracker which fixes a segfault with voice in the sim
21:04:55amiconnThe intermediate layers on swcodec must not buffer too much voice data b/c that makes the voice UI lag
21:05:21lostlogicof course
21:05:54amiconnI didn't check that fix yet, i.e. whether it breaks voice on hwcodec and/or swcodec target
21:06:07amiconnI hope to be able to do that later tonight
21:06:12amiconn(not @ home atm)
21:06:48amiconnMaybe it fixes the problem with voice running mad sometimes...
21:06:53lostlogicI'm going to ignore it until I'm done with my current rework −− most of the playback.c voice code has been refactored, I just hafta figure out what is wrong with the state that voice is leaving playback in.
21:07:52lostlogicbut anywho, the point is that the getmore as used for the MAS DMA whould be adapted so that the voice codec can call it directly instead of the playback.c voice_request_buffer stuff dummying up interfaces for both sides
21:07:56lostlogicit makes a mess.
21:08:37 Join lamed [0] (
21:10:04lostlogicbasically I need the request_buffer, advance_buffer and advance_buffer_loc functions properly implemented for communication between a codec and talk as opposed to codec and filebuf
21:10:43 Join SereR0KR [0] (
21:11:10 Quit muesli__ (Read error: 110 (Connection timed out))
21:11:31amiconnlostlogic: Hmm. I think you won't be able to ignore talk.c. Probably I should handle "the other side of things" and fiddle with it
21:11:43lostlogicamiconn: yeah, I'm beginning to see that
21:11:56lameddoes menu+ right vertically scrolls filenames (and menus) that are out of view?
21:12:07 Join carini [0] (
21:12:14 Join JBGood [0] (
21:12:20lamedthat is wonderful.
21:12:23lostlogicamiconn: <−− this is my refactor so far −− it breaks voice a bit worse, because of a bad end state, but the voice_request_buffer_callback and advance_buffer_callbacks are what need reimplementation with knowledge and help from talk.c
21:12:58lamedamiconn, lostlogic,: remember?
21:13:03lostlogicamiconn: but OTOH, I don't start the voice thread at all if there's no voice file, which is something I think you'd suggested −− not starting threads we don't need
21:13:07lamedthanks sharpe
21:13:11sharpeyou're welcome
21:13:44amiconnlostlogic: My suggestion was to start/stop threads as needed, saving ram especially on archos
21:14:19amiconnToday, playback and recording is handled by the same thread on archos, but probably not anymore when engines get unified
21:14:26lostlogicamiconn: right, so the voice thread will not be started if it won't be used, not exactly on demand yet, but hey, babysteps
21:14:39amiconnBut "mutexing" the threads we can reuse the stack
21:16:07lamedamiconn, lostlogic?
21:16:23amiconnlostlogic: The voice thread should probably be started (and the ram for the voicefile reserved) as soon as it finds _any_ .voice file
21:16:42lostlogicamiconn: yes, that's what it does
21:16:50amiconnThe user might have the wrong .lng loaded, but switch later to a .lng which has an associated voice file
21:16:50 Quit linuxstb (Read error: 110 (Connection timed out))
21:16:53lostlogicamiconn: the thread is started and left waiting for mp3 data
21:17:02lostlogiclamed: no idea
21:17:14 Join linuxstb [0] (
21:17:21amiconnJust keep in mind that we don't want to tie voice on swcodec to mp3
21:17:43*amiconn looks at preglow and whispers 'speex' ;)
21:17:51lamedlostlogic: vertical screen scrolling, I'm almost positive you where the submitter..!
21:18:10lostlogicamiconn: to change codecs for voice would involve implementing "track change" functionality for the voice codec which woudln't be rocket science, but is not done.
21:18:19lostlogiclamed: doesn't sound like me
21:18:54lamedlostlogic: whoops, horizontal screen scrolling :DDD and yeah, it's you
21:19:05lamed2006 22 Jan 04:24lostlogic
21:19:09lamedHorizontal screen scrolling part 3 (by Shachar Liberman)
21:19:11amiconnlostlogic: No, we don't need on-the-fly changes. The voice codec should be fixed to one format (e.g. by a #define)
21:19:29lostlogicoh, sure, I committed it, sure, doesn't mean I have any stake in it :-P
21:19:45amiconn...just not tied to mp3 so we can change voice to use speex by just changing that #define (and having the proper codec)
21:19:55lamedhmm, ok, could you submit the fourth part for me then?
21:20:01lostlogicamiconn: ok, no prob at all, whatever codec will just fireup and wait for data
21:20:07lostlogiclamed: not now, I'm rather busy.
21:20:15lostlogiclamed: is this a bug fix or an enhancement?
21:20:21lostlogiclamed: is it something that needs to go in for 3.0?
21:21:17lamedlostlogic: 1,sure, no rush. 2. both acctually, everything is documented in the tracker. 3.yes
21:23:36lamedI'll be off now, good night everyone and thanks for helping out, lostlogic sorry if i was a nag :) gday
21:23:47***Saving seen data "./dancer.seen"
21:25:32 Quit Farpnut (Read error: 110 (Connection timed out))
21:25:34 Quit lamed ("CGI:IRC (EOF)")
21:32:50 Part raf82
21:33:32 Quit Farpenoodle (Read error: 110 (Connection timed out))
21:42:20 Quit t0mas (Read error: 110 (Connection timed out))
21:59:08*Ribs finally removed all the Apple shit from his iPod
21:59:14RibsIt's just better with RockBox :>
22:01:44 Join Rondom_ [0] (
22:07:25 Quit Lear ("Chatzilla 0.9.72 [Firefox]")
22:08:29 Quit tvelocity ("Ex-Chat")
22:09:21twisted`oh hell... where is Paul_the_Nerd
22:10:19 Join damaki_ [0] (
22:10:44sharpegeneral concensus says he isn't here.
22:11:04peturuser list too ;)
22:11:54webguest29amiconn: do you know what kind of LA Linus have?
22:15:47Mooswebguest29: maybe Bagder know the logic analyzer used by Linus, since they are working on the same company
22:17:09 Join Nico_P [0] (
22:19:45 Join Arrogant [0] (
22:22:25 Quit Rondom (Nick collision from services.)
22:22:28 Nick Rondom_ is now known as Rondom (
22:23:36 Join t0mas [0] (
22:23:40 Join aegray_ [0] (
22:24:24 Quit SereR0KR ("XChat Aqua")
22:25:00 Quit stamppot (Remote closed the connection)
22:25:28 Quit damaki (Read error: 110 (Connection timed out))
22:29:32 Join stamppot [0] (
22:30:06 Quit aegray (Connection timed out)
22:33:47 Quit t0mas (Remote closed the connection)
22:37:44Bagderno I don't
22:37:53Bagderits Linus' own LA
22:38:41webguest29Bagder: aha, ok.. do you know what brand / model it is?
22:38:52 Join midgey34 [0] (
22:38:59BagderI'm sure he has told me and I've seen it but I can't remember
22:43:48 Join t0mas [0] (
22:44:15 Quit Arrogant ("Leaving")
22:44:19t0mascomputers work really bad without power
22:45:05sharpeyeah, that's how it seems to be most of the time
22:45:07yobesoomO RLY?
22:45:56 Join lostnihilist [0] (
22:47:53twisted`lol I noticed cus I did Paul[tab] and nothin happened... so...
22:48:18HClravon: any luck with libdumb?
22:49:52twisted`anyone here familiar with tagnavi.config ?
22:50:03 Quit sharpe (Read error: 104 (Connection reset by peer))
22:53:12twisted`cause I don't really get how it works
22:53:18twisted`"Artists" artist : album : song
22:53:33twisted`I want to be able to do Artists -> All Songs
22:55:55 Join sharpe [0] (
22:56:21 Join Kyomi [0] (
23:00:44*twisted` pwnd
23:01:22 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
23:01:46twisted`sharpe: u know how tagnavi.config works?
23:02:18twisted`cus it's fuckin undocumented >_<
23:03:30HClwelcome to opensource development ;x
23:03:31 Quit Nico_P (Remote closed the connection)
23:03:44twisted`most of the stuff has been documented so far :P
23:03:49twisted`better then the closed source crap ;)
23:03:54HCltrue :)
23:04:04twisted`let's hope this way OK
23:04:56twisted`didn't work
23:05:36twisted`I just want a menu where I can say: Artists -> All Songs, Albums -> Songs
23:06:04HClsounds like tagdatabase...
23:06:04 Join macox [0] (
23:06:13macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:13macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:13DBUGEnqueued KICK macox
23:06:13macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:19macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:20***Alert Mode level 1
23:06:20macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:23***Alert Mode level 2
23:06:23macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:30ender`oh great. now it's here
23:06:31***Alert Mode level 3
23:06:31macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:31***Alert Mode level 4
23:06:31macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:34Mode"#RockBox +o Bagder " by ChanServ (ChanServ@services.)
23:06:35macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:36sharpethat's nice.
23:06:37HClwhy do people think we don't know the ignore command?
23:06:39macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:43***Alert Mode level 5
23:06:43macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:44Mode"#RockBox +o logbot- " by Bagder (n=daniel@rockbox/developer/bagder)
23:06:45DBUGQ-Sent KICK HCl to server
23:06:46Kick(#rockbox HCl :No flooding!) by logbot-!
23:06:47***Alert Mode level 6
23:06:47DBUGsent MODE #rockbox +b *!*n=cisco@*
23:06:47macox#fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai #fazlamesai
23:06:47DBUGQ-Sent KICK scorche to server
23:06:47Mode"#RockBox +b *!*n=cisco@* " by logbot- (
23:06:48Kick(#rockbox scorche :No flooding!) by logbot-!
23:06:48 Join scorche [0] (n=scorche@
23:06:49DBUGQ-Sent KICK twisted` to server
23:06:50Kick(#rockbox twisted` :No flooding!) by logbot-!
23:06:51 Part macox
23:06:54 Join HCl [0] (i=hcl@rockbox/developer/HCl)
23:06:56HCland what was that for
23:07:07BagderI don't know why it did that
23:07:15HClmust be holding a grudge against me.
23:07:15sharpeoh he's mad now.
23:07:34 Join twisted` [0] (
23:07:37twisted`why the fuck was I kicked
23:07:46sharpewe love you
23:07:56twisted`23:06:50 -!- twisted` was kicked from #rockbox by logbot- [No flooding!]
23:08:05Bagderit seems logbot is quite lagged
23:08:07 Join RotAtoR [0] (n=e@rockbox/developer/rotator)
23:08:08HClseems logbot has been holding grudges
23:08:09twisted`I mean, you need to check logbot-'s settings
23:08:11Bagderor was at least
23:08:19HCland went "finally, ops! mahahhahaha"
23:08:26Mode"#RockBox -o logbot- " by Bagder (n=daniel@rockbox/developer/bagder)
23:08:27sharpelogbot is now senient...
23:08:37 Join SereR0KR [0] (
23:11:45 Nick MusiFreq is now known as axion_ (
23:12:00sharpewell... i get different readings for the emulation, but not it's like, 104% at ~10fps
23:12:08sharpenow, not not.
23:13:03 Join dj-fu [0] (
23:14:13 Part ep0ch ("Kopete 0.11.1 :")
23:14:47 Quit Sinbios ("If the definition of a klutz is someone who doesn't have eyes on their ass, then yes, I suppose I am a klutz.")
23:16:48***Alert Mode OFF
23:16:54lostlogicpreglow: wow, just actually _used_ crossfeed on a song that had some serious suck on earphones, amazing.
23:23:50***Saving seen data "./dancer.seen"
23:26:58 Join larcher [0] (
23:28:13 Join p0wr-toc [0] (
23:30:41amiconnwebguest29: Linus' LA is an Acute LA1000
23:31:35kkurbjundoes anyone know what's wrong with this statement: "lsr.l #16,%%d1 \n". I keep getting Error: operands mismatch −− statement `lsr.l #16,%d1' ignored
23:31:48kkurbjunfor coldfire asm
23:31:52amiconnof course
23:31:56scottderlostlogic: it can do magic on songs with hard stereo seperation
23:32:20lostlogicscottder: yep
23:32:43amiconnkkurbjun: Literal shift amounts have a range of 1..8
23:32:49lostlogickkurbjun: swap
23:33:16kkurbjunoh, so I'd have to load it to a register first then?
23:33:18lostlogicmove.w #0,%d1; swap %d1;
23:33:26scottderlostlogic: now crossfeed needs a bit of adjustability...that'd be awesome :)
23:33:38lostlogicscottder: it is quite adjustable...
23:33:39 Quit TeaSea (Read error: 101 (Network is unreachable))
23:33:53scottderIt is?
23:33:59kkurbjunlostlogic, ahh, that'd be faster then loading to a register and then doing that huh
23:34:02lostlogicdirect gain and 3 other options
23:34:09lostlogickkurbjun: yah
23:34:13amiconnFor higher shift amount, either (1) use multiple shift instructions, or (2) load a register with the shift amount and use shift-by-reg-value or (3) use some other clever construct
23:34:23scottderAhh hadn't
23:34:30amiconnlostlogic: That would shift left, not right
23:34:55lostlogicamiconn: damned byte order.
23:34:57amiconnkkurbjun: For >>16, there's a clever way:
23:35:11amiconnclr.w %%d1; swap %%d1
23:35:31lostlogicamiconn: how's that different than what I said?
23:35:37lostlogicother than it's faster
23:35:46amiconnoh, misread
23:36:00amiconnStill, move.w uses an extension word
23:36:07lostlogicyeah, you're right, I forgot about clr
23:36:15amiconnUnfortunately there's no moveq.w, only moveq.l
23:36:15kkurbjunso that should work then? what's the order on big endian again?
23:36:25 Join TCK [0] (
23:36:29lostlogicam starting to get used to arm where you have to ldr reg, #0
23:37:03amiconnThe arm load instructions are, ahem, weird...
23:37:15lostlogicamiconn: powerful :)
23:37:24amiconnNah, not necessarily
23:37:35 Join muesli__ [0] (n=muesli_t@
23:37:37amiconnYou can't just load an arbitrary literal
23:37:50kkurbjunwhat would be faster for a right 10 bit shift, loading to a reg and shifting of one shift 8 and one shift 2?
23:38:03kkurbjunof =or
23:38:04amiconnIt's equally fast
23:38:17amiconn...but with the 2 shifts, you save a register
23:38:38kkurbjunok, I'll do that, thanks for the tips
23:38:51amiconnBoth ways need 2 insns, and all these insns are single cycle
23:39:08 Join Jumpfroggy [0] (
23:39:16preglowlostlogic: yeah, it works nicely
23:39:24kkurbjunamiconn, is there a reference on how many cycles each instruction takes?
23:40:37amiconnThe important ones are CFPRM.pdf for the general coldfire instruction set, and MCF5249UM.pdf for 5249 hardware specifics
23:41:06amiconnThe latter has an instruction timing table, in the 'coldfire core' chapter
23:41:21kkurbjunahh, ok, I hadn't seen it in the general instruction set
23:41:37lostlogicamiconn: can you spontaneously generate similar documents for PP chips? :-D
23:42:47preglowjust rub him in the right way and they'll come
23:43:20mikolaspreglow, does the second core of pp appear as a coprocessor?
23:44:08preglownot as far as i know
23:44:29mikolasi'm just trying to get familiar with arm assembler
23:44:42mikolasseems quite simple compared to coldfire
23:45:04preglowyeah, i think so too
23:45:11mikolasbut the conditional stuff makes your brain hurt
23:45:31 Quit ze (Read error: 110 (Connection timed out))
23:45:40mikolaspretty different approach than with traditional cisc stuff :)
23:45:41preglowusing it properly is a bit hard
23:45:53preglowwell, conditional instructions can't really be said to be risc
23:45:58preglowi know of no other archs that do it
23:46:12preglowthe arm people took a quite different route to cisc
23:47:01mikolashaving programmed loads of stuff with 6502 and m68k, arm really is a bit different
23:47:40preglow6502, bleh
23:47:56preglownot really comparable, heh
23:48:01Bagdernow that is one easy asm
23:48:08mikolasthe best there is
23:48:15preglowthat's pretty easy, yes
23:48:19preglowthe lack of regs drives me crazy
23:48:26lostlogicI first learned 68020, so coldfire is just like what I knew only with a MAC pretty much
23:48:48preglowand less nifty features :/
23:49:12Kyomipaulheu needs to get his lil butt compiling
23:49:26KyomiThere hasn't been a new one since the 17th
23:51:20 Join wefds [0] (n=muesli_t@
23:53:27 Join TeaSea [0] (
23:54:09 Quit xmixahlx ("blah blah blah")
23:54:14SereR0KRhm rockdoom says missing base wad
23:54:29SereR0KRbut I have rockdoom.wad in /games/doom
23:58:14preglowhave you got both wads?
23:58:55SereR0KRboth wads?

Previous day | Next day