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

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

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

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

#rockbox log for 2007-03-26

00:00:01Soaplinuxstb: I've read that, but the only programs I can find that do it, fake it.
00:00:12XavierGrand the target can display 10fps without problem
00:00:15Lloreanandrewg867: There's a plugin you have to run to configure the random mode.
00:00:22Lloreanandrewg867: It shouldn't be necessary for non-random though
00:00:36 Join web-taz [0] (
00:00:44andrewg867Llorean: it still doesn't work on non random unless I have to reboot the player
00:00:53web-tazsry i got disconnected
00:01:00 Join JavaMan22 [0] (
00:01:06Soap23.976, 24, 25, 29.97, and 30 are the only ones I've ever made which aren't created with duplicate frames.
00:01:10Lloreanandrewg867: Reboot after choosing "Yes" and going back, or reboot after just enabling non-random without it ever having been "Yes" that boot?
00:01:24 Quit ender` (" MCSE = Moron Claiming System Expertise or Must Call Someone Else?")
00:02:14web-tazfonts: are they installed like themes/WPS or must i recompile the build?
00:02:21web-tazand where do i get them?
00:02:53pearldiverlinuxstb actually i just checked this 25fps one on PC and it gets out of sync too
00:03:01linuxstbSoap: So the programs you use would encode a 15fps file as 30fps (duplicating frames) - resulting in a file with a 30fps framerate?
00:03:01andrewg867Llorean: I haven't even tried using the random option before, I was just using yes and after I updated one day a couple weeks ago it didn't change the directory so I went in and saw the random option there but just switched to no then back to yes but it still doesn't work
00:03:32Lloreanandrewg867: The random option is a bit older than a couple weeks.
00:03:54linuxstbSoap: libmpeg2 supports the unoffiical framerates - so in theory you should be able to create genuine 10/12/15fps files.
00:03:56Soapwith the duplicate frame being a simple P frame saying "what <−−he said" as far as I can tell.
00:04:02Lloreanweb-taz: The manual explains the need to download the fonts, and where to get them
00:04:25andrewg867Llorean: I don't remember when it stopped working but it was a while ago
00:04:34web-tazmhh manual... i hate studying englisch texts ;)
00:04:46linuxstbSoap: libmpeg gives the user application a "frame duration" for each frame. So for a 15fps file, the duration would be double that of a 30fps file etc etc.
00:04:46andrewg867I should just manually go back in and revert it back to only having no and yes
00:04:57Lloreanandrewg867: Well, you're the first person to complain about it, which suggests to me at least it's not affecting other people, so it'd be nice to know when it began happening at least.
00:04:59JavaMan22how do i edit a theme
00:05:17JavaMan22i have this ubuntu theme and i want to edit ut
00:05:25Lloreanandrewg867: Avoiding it won't really help the bug get fixed, if there is a bug in it.
00:05:31 Join mirak [0] (
00:05:36linuxstbSoap: But as you said, it's unofficial (and commented as such in libmpeg2), so maybe very few encoders do it.
00:05:54Soapgive me a second, and I'll have a 12fps example up for you.
00:05:56miraklinuxstb: are you dave chapman ?
00:05:56andrewg867Llorean: I'll look at when it first appeared (the random option) and try reverting that file to see if that fixes it
00:06:03Soaphe's Batman.
00:06:12andrewg867Llorean: if it still doesn't work then something else must have broken it
00:06:26miraklinuxstb: hum yes you are sorry it's written ...
00:06:27scorcheJavaMan22: with a text editor such as vim, emacs, notepad2, etc
00:07:12JavaMan22ooo linuxstb is chapman
00:07:15Blastai'll try and work around some more with this stupid thing.
00:07:34JavaMan22Dave Chapman apps/plugins/mpegplayer/mpegplayer.c [diff]
00:07:34JavaMan22Initial attempt at rebuffering
00:07:39Blastathanks for developing rockbox :) made me not feel bad about purchasing that ipod after i found out how abysmal it's sound is ><
00:07:46JavaMan22because linuxstb said he would work on it yesterday :D
00:08:07 Join gratt [0] (i=d8e860d4@gateway/web/cgi-irc/
00:08:20miraklinuxstb: about idct for coldfire, I will not have much time to put in it (especially about coding) but I can provide verbal help :)
00:08:21 Join chris_hotwire [0] (
00:08:23 Quit Blasta ()
00:08:29 Quit My_Sic (Read error: 131 (Connection reset by peer))
00:08:40 Nick chris_hotwire is now known as hotwire__ (
00:10:12 Quit web-taz ("Nettalk6 -")
00:10:35linuxstbmirak: I can't do anything either - I don't have any experience of Coldfire asm, and I don't have a colour Coldfire target.
00:10:48linuxstbmirak: But it seems a shame to leave your patch rotting there...
00:10:51miraklinuxstb: one thing I guess was to put the functions in it's own .S file, I don't know how to do that. Other than that the only thing to change other than the idct.c file is to use a idct block of 128 bytes instead of 64.
00:11:33Lloreanandrewg867: If reverting that fixes it, that doesn't mean that Random is what broke it, though, it just means that reverting random fixes it. Try a full SVN build from shortly after random was introduced and see if it works, that'll tell you more or less if it ever worked. I know random, and folder advance, still worked for some people for a while after introduction, so Random alone didn't globally break it
00:11:34 Quit midgey ()
00:13:14miraklinuxstb: well if it was me I would have uploaded it to cvs and let some people integrate it better, but I understand that how it is the code quality requirements are not reached. (I mean the separate asm code). I don't know how to do that, and it seems that the atempt of RaeNye failed
00:13:45Moosamiconn too IIRC
00:14:42Moosand there are not a much of asm for coldfire around :)
00:14:45 Quit voltagex ()
00:14:55hotwire__Hey folks, I just tried out Elephant's Dream on one of the most recent builds with the latest mpegplayer updates on my Gigabeat F. Just wanted to report all was smooth for me, and generally phenomenal.
00:14:58miraklinuxstb: I don't think there is something to modify in the asm part, it was working great.
00:15:01Soap(if you care)
00:15:32 Join goffa [0] (n=goffa@
00:15:58grattHi , anybody that can help with sansa E200 playlist conversion patch I have written m3u2pla for Rockbox
00:16:09 Quit zylche ("-")
00:16:46 Join zylche [0] (
00:17:08 Quit web-taz2 (Read error: 110 (Connection timed out))
00:17:44 Quit printfXh4 (Read error: 148 (No route to host))
00:18:21 Quit Joely (Remote closed the connection)
00:22:26 Join gratty [0] (
00:23:29linuxstbmirak: I'm hoping someone who knows Coldfire asm will look at your patch and commit it - I just wanted to check if you were planning/able to do any cleaning up to get it ready
00:24:16 Quit joshin ("rebooting")
00:24:18linuxstbSoap: mplayer reports that file as 24.0fps
00:24:18 Quit gratt ("CGI:IRC (EOF)")
00:24:36Soapyea, 12fps with duplicated frames.
00:24:55linuxstbSo that's not the type of 12fps file that libmpeg2 supports.
00:25:10Soapso libmpeg2 will create honest 12fps files?
00:25:22linuxstbIt decodes them. yes.
00:25:25 Quit gratty (Remote closed the connection)
00:25:40Soapbut we don't know where to make them?
00:25:50linuxstbI'm sure mencoder would do it.
00:25:51 Join GRaTT [0] (
00:26:08linuxstbLet me try...
00:26:21JavaMan22linuxstb how come the iboy from ipl runs faster
00:26:28JavaMan22couldnt we just take their code
00:26:43 Quit lee-qid ("aufwiederbyebientotsayonara")
00:28:00linuxstbJavaMan22: Yes, you could.
00:28:43linuxstbSoap: Yes, mencoder seems to do it. The resulting file is reported as 12fps by mplayer.
00:28:57 Join midgey [0] (
00:29:33 Quit Thundercloud (Read error: 104 (Connection reset by peer))
00:29:41JavaMan22then why doesnt anyone do that
00:30:11linuxstbAsk anyone. All I can say is that I've not done it because I'm not interested in rockboy.
00:30:37LloreanJavaMan22: Why haven't you done it?
00:31:04BagderI've not done it because I'm not interested in rockboy.
00:31:09 Join Thundercloud [0] (
00:31:51dj-fuHas anyone done any work with visualisations? or a visualisation style plugin?
00:32:26dj-fuI didn't find anything on the cvs or forums
00:33:01markunJavaMan22: I also haven't done it because I'm not interested (enough) in rockboy
00:33:18SoapI can't run winamp visualisations on my 500 Mhz x86. I think you would be hard pressed to run them on any rockbox target outside possibly the Gigabeat.
00:33:30linuxstbmarkun: And I would guess it doesn't need optimising on the Gigabeat?
00:33:36miraklinuxstb: is it needed to know coldfire asm to do the separation, or is it just c and gcc skills ?
00:33:42dj-fuSoap, I see
00:33:50markunlinuxstb: longer battery life is always welcome :)
00:33:52dj-fumy 400mhz arm pocketpc handles them
00:34:15JavaMan22well certainly someone was interested to make rockboy to begin with
00:34:17linuxstbmirak: I think you need to know at least a little coldfire.
00:34:42scorcheSoap: think of the overhead of that OS versus rockbox though too, but point is taken
00:34:42 Quit daniel2023 ("CGI:IRC (EOF)")
00:34:44*markun doesn't care much for visulalisations (maybe visuals in general) either
00:34:50 Quit ompaul ("lights out")
00:35:03dj-funor do I, but, I figured someone might have done something
00:35:05LloreanJavaMan22: And now nobody's working on it. That's the way it is. If you want something done, you do it yourself, or file a feature request and wait patiently for someone to be interested.
00:35:08dj-fua fade-style peak analyzer, etc
00:35:55Soapscorche: touche - but Windows = overhead? bah! ;) More seriously = think of the floating point advantage x86 has.
00:36:14LloreanJavaMan22: Rockbox is NOT a game system. It is an audio player. This room is almost entirely full of people who focus on Audio features or, additionally, "media" such as pictures and video, with that being already a fairly low priority.
00:37:13scorchelinuxstb: PM
00:37:17 Quit midgey ()
00:39:18JavaMan22are you developers retired people?
00:39:30LloreanWe're all sorts of people
00:39:40scorchei dont think any are retired..
00:39:44*bluebrother is quite tired of all that "implement xyz" while xyz having nothing to do with music playback
00:39:47miraklinuxstb: I am aware I let the things a bit unfinished, but I think 98% of the job is done
00:39:53JavaMan22how come you guys have so much time
00:39:55Bagderwe're old and grumpy, but not retired ;-)
00:40:09scorcheBagder: speak for yourself!
00:40:12bluebrotherBagder, yeah :-)
00:40:21*Bagder pokes on scorche with a stick
00:40:35miraklinuxstb: I admit I hoped someone would do the 2% left
00:40:41bluebrotherit's great being old and grumpy ;-)
00:40:52*scorche takes the stick and leaves Bagder with no walking stick =P
00:40:55JavaMan22im 10
00:41:07BagderJavaMan22: in fact, many people have this much time but they spend it on other stuff
00:41:27mirakLlorean: I think rockbox is not just and audio player, it's relly an os isn't it ?
00:41:42 Join Joe_ [0] (
00:41:43LloreanMany people dedicate a ridiculous amount of time to trying to watch certain TV programmes each week.
00:41:45bluebrotherno :)
00:42:06JavaMan22i like rockbox because theres always updates :)
00:42:07Lloreanmirak: Its primary purpose is as an audio player. It's a jukebox firmware, not an OS, though it has to do many OS-like functions just to meet this purpose.
00:42:13 Quit Arathis ("Bye, bye")
00:42:30bluebrotherthat tv-thing is quite amazing. I barely even want to watch tv
00:43:15JavaMan22i dont watch a lot
00:43:35bluebrotherit also isn't an OS because it doesn't allow you to run arbitrary code. Only the plugins which are quite fixed to the build
00:43:58mirakLlorean: but it evolved, like windows evolved to be used for something else than text writing
00:44:28Lloreanmirak: It may have evolved what it can do, but it didn't really evolve where the core focus is.
00:44:29Bagderwell, rockbox is in fact an OS afaik the definition
00:44:32GRaTTHi, need character set info for m3u2pla playlist converter for sansaE200 patch is made but only works for latin UTF
00:44:46*Bagder runs away
00:45:07GRaTTit is real simple code
00:45:21Lloreanmirak: As Bagder said, it may technically be an OS, but that doesn't really change where most of the development focus is, overall.
00:46:02markunGRaTT: was it a question?
00:46:14JavaMan22ok well i hope that google team works on rockboy
00:46:49GRaTTthe OF need playlist with null bytes I use '\0' but it did not work for someone with other than latin UTF
00:47:09*bluebrother goes voting against doom again
00:47:12LloreanJavaMan22: I hate to say it, but we get to pick which applicants get in, which tends to mean we'll mostly pick the ones who are interested in features we consider core or important.
00:47:34bluebrotherwhat is "latin utf"?
00:47:36LloreanJavaMan22: Seriously, if you think iBoy is so much better, just use iBoy. And please quit making those faces, they just waste a line.
00:47:54markunbluebrother: latin1 would be a safe guess I think
00:48:18bluebrothersounds like a multibyte vs. singlebyte problem then.
00:48:19JavaMan22iboy has a bunch of color problems
00:48:24JavaMan22and not all roms work on it
00:48:24GRaTTbluebrother: specific character set. I know nothing about char sets.
00:48:53bluebrotherGRaTT, there is no "latin utf" −− utf is multibyte and includes like "everything"
00:48:55JavaMan22im playing metal gear solid and when i go up this elevator the game crashes on iboy
00:49:02bluebrotherin europe the usual charset is latin1
00:49:16JavaMan22i tryed to open the save file with rockboy but it didnt work
00:49:19bluebrotheror to be more precise was latin1, and got superseded by latin9
00:49:24jjorelatin1-something, usually, isn't it?
00:49:26GRaTTmarkun: I have modified it to use the defined NULL but unsure if this will fix it.
00:49:33bluebrotherwhich is basically the same but has the euro-symbol added
00:49:54markunGRaTT: I have no clue what m3u2pla is :)
00:50:01bluebrotherGRaTT, if you are writing chars the right character is '\0'
00:50:14mirakLlorean: respect the users ;)
00:50:33bluebrotherwhile NULL is used for pointers (and is not necessarily the value 0)
00:50:36GRaTTmarkun: it is a plugin I have made to convert the rockbox playlist to a version the Sansa E200 can read
00:50:49mirakLlorean: he come here from a user point of view so ...
00:51:13bluebrotherhave you tried comparing the problematic playlist in a sansa-version to a m3u2pla-version?
00:51:14mirakgood night all !
00:51:16markunGRaTT: maybe you should name it m3u82pla to make it more obscure and accurate :)
00:51:19Lloreanmirak: He's been told several times in the past that in Rockbox gaming isn't a priority.
00:51:23markunmirak: night
00:51:37markunmirak: don't forget to wake up at the right time tomorrow :)
00:51:39bluebrotherwe have iriverify, why not sansaify?
00:51:42Lloreanmirak: He's moved from "user without knowledge" to "someone who comes in and repeats the same questions even though he's been told an answer, just because he didn't like the answer"
00:51:51mirakmarkun: I prefer not think about it ...
00:52:06mirakmarkun: ah ! yes, I changed my clocks already
00:52:10markunok :)
00:52:22JavaMan22i have other questions also
00:52:25JavaMan22not just rockboy
00:52:34*bluebrother doesn't want to hear it
00:52:40mirakmarkun: the state is stealing us one hour of time I want to be paidback :)
00:52:42GRaTTmarkun: I converted it from a perl version I made the name was really unimportant
00:52:55 Join My_Sic [0] (
00:52:58bluebrothermirak, you'll get it back in half a year ;-)
00:53:05 Quit roolku ()
00:53:24markunGRaTT: I wasn't really serious
00:53:30mirakbluebrother: at this time I will think it as a happy present from the state :D
00:53:50GRaTTmarkun: did not really take you to be
00:54:01 Quit mirak ("Ex-Chat")
00:54:04bluebrotherI think that stupid DST thing should get dropped completely.
00:54:52GRaTTbluebrother: I changed to the NULL, it still worked for me and the little I know about char encoding I was hoping it might work.
00:55:43GRaTTbluebrother: might work for all char sets
00:56:58bluebrotherin most cases NULL and '\0' have the same value of 0.
00:57:24bluebrotherafaik this is true for all rockbox targets, so it shouldn't make a difference
00:58:42 Quit Joe_ (Client Quit)
00:58:45GRaTTIf the char set on the computer that is being used for compiling is different, would this change the vlue of '\0'
00:59:30bluebrotheranyway, this damn DST change stole me an hour, so it's quite late here. I'm out for sleep
00:59:40 Quit bluebrother ("zzzzzzz")
00:59:47 Quit obo ("bye")
01:00:00 Quit hotwire__ ("Chatzilla 0.9.77 [Firefox]")
01:04:24andrewg867there you go :)
01:05:05Lloreanandrewg867: As long as the file doesn't clip, it shouldn't distort all the way up to 0dB, I believe, unless something is wrong.
01:05:26LloreanThough I personally listen at about -40, on my other Rockbox players. :)
01:05:34andrewg867it will start to distort since I increased the DAC -> Mixer gain
01:06:03andrewg867I find -9dB is nice and loud for me, just for those times when you don't want to hear anything else
01:06:11Llorean0dB should be line level.
01:06:14***Saving seen data "./dancer.seen"
01:06:14LloreanWhich shouldn't distorty.
01:06:34LloreanIf you've got it so that "0dB" is actually distorting, it suggests to me there's a problem of some sort.
01:06:38andrewg867maybe someone could figure out the ideal gain level then :)
01:07:01LloreanWell, at 0, no gain should be added to or subtracted from the level of the file, if I understand.
01:07:01andrewg867its just the DAC -> Mixer gain that needs tweaking now
01:07:19andrewg867the gains I am talking about are internal chip gains
01:07:44LloreanYes, but those chips would be modifying the level before output, no?
01:08:17andrewg867nope, it is all software controlled from what I see
01:08:47 Join safetydan [0] (i=cbca159f@rockbox/developer/safetydan)
01:10:25Lloreanandrewg867: It looks like you're not even sure what you're setting the gain at?
01:10:31safetydanIs anyone crazy enough to try and get the mpeg player going on grayscale targets?
01:10:36*safetydan feels left out now with his h120
01:10:49safetydanJust pass the Y component through to the grayscale lib, easy! :)
01:11:03scorchesafetydan: well, the archos has had video for a long time =P
01:11:18safetydanscorche, true
01:11:23Lloreansafetydan: I think one feeling is that RVF should be adapted for all grayscale targets.
01:11:30linuxstbsafetydan: Feel free...
01:11:31Lloreansafetydan: Though I think mpegplayer should work too, in my opinion.
01:11:52andrewg867Llorean: you got that right ;) 0x16 should be about half between -45.43dB and 1.07dB
01:11:57 Join midgey [0] (
01:12:19andrewg867I really suck at converting between binary, hexidecimal and dB values
01:13:17safetydanlinuxstb, I'm not even sure if it's possible... doesn't the grayscale lib take CPU time? Is there enough left over after audio/video decoding for rendering?
01:13:52safetydanLlorean, on a side note, stereo Speex support is probably a long way off. The whole stereo decoding path is floating point.
01:14:24Lloreansafetydan: Honestly, I'm fine with that. :)
01:14:37Soaphow much "smearing" would using the greyscale library create?
01:14:47andrewg867Llorean: it should be -23.25dB gain on the DAC -> Mixer if my calculations are correct
01:15:22Lloreanandrewg867: I'm misunderstanding something then (not uncommon since I'm not familiar with how this hardware works) but wasnt it at -16.5 (more gain) before?
01:15:55safetydanLlorean, good, good. Now I just have to remember to commit the minor change that makes stereo files decode as mono rather than the noise they do now.
01:16:12linuxstbsafetydan: Yes, CPU usage will be tight.
01:16:39linuxstbsafetydan: Which is why I think rvf is preferred - better quality will be possible than with mpeg.
01:16:48andrewg867Llorean: hrm maybe their calculations are more correct than mine, but that new value (0x16) should still be higher
01:16:53Lloreanlinuxstb: How's the compression ratio on RVF?
01:16:57 Join Slowking_Man [0] (n=slow@wikipedia/Slowking-Man)
01:17:03linuxstbLlorean: There isn't any.
01:17:22andrewg867Llorean: it might be -20dB also, I haven't the slighest clue on how to calculate it
01:17:37linuxstbIIUC, it's mono bitmaps at the refresh rate of the LCD - around 70Hz I think.
01:17:49Lloreanlinuxstb: Okay, clarification, how's the ratio of RVF size to a color, compressed video of equivalent size?
01:18:14andrewg867Llorean: oops, I was looking at the wrong page in the datasheet
01:18:30Lloreanandrewg867: I'm just concerned about the distortion at 0dB. I don't get it with my other Rockbox players, and use that since most of them don't have a Line Out for me to use.
01:19:03linuxstbLlorean: Well, 160x128x70 gives 1.4Mbit/s... I'm not sure about the 70Hz part of that equation.
01:19:05andrewg867Llorean: let me try the calculations again, the maximum is 6dB which the 5bit volume would be 11111, and lowest of -40.5dB 00000
01:19:13 Quit Nico_P (Remote closed the connection)
01:19:17*scorche wishes the archos units had a proper line out
01:19:25Lloreanlinuxstb: Hm. So, larger filesizes, but likely better playback.
01:20:50amiconnscorche: The player does...
01:21:11Lloreanandrewg867: -7.5, I think
01:21:15*linuxstb spots someone who can correct my guesses about rvf playback
01:21:28andrewg867Llorean: also when I recorded earler I used a volume of -15dB on rockbox
01:21:41pixelmaLlorean: amiconn made a "elephant's dream.rvf" - you could encode it at same resolution as mpeg and compare
01:22:07pixelmaIIRC it is ~45MB
01:22:10Lloreanpixelma: What resolution is it?
01:22:22andrewg867Llorean: if you could try out the patch and try lowering the gain until you feel it is an inprovment
01:22:28 Nick pearldiver is now known as perldiver (
01:22:29pixelmaLlorean: 112x64 ...
01:22:30 Join marc [0] (
01:22:51safetydanLlorean, for some more details if you want to encode your own
01:23:15Lloreanpixelma: Alright. And the elephant's dream at 224x128 (four times the resolution) is 42m
01:23:39andrewg867Llorean: also what I though would be nice is a menu to select the gain level yourself almost like the sandisk "Normal" or "High" volume menu
01:23:42Lloreansafetydan: Thanks, it's mostly curiosity.
01:23:46amiconnlinuxstb: The 70Hz would be correct for H1x0 and M5
01:24:06Lloreansafetydan: Just trying to weigh in pros/cons of RVF on iRiver H100 in my mind.
01:24:31amiconnFor greyscale iPods this would be 88Hz
01:25:03linuxstbOuch, about 1.8Mbit/s for the video stream...
01:25:15amiconnLlorean: The big pro is that it would be definitely realtime, and less grainy than using the grayscale lib and mpeg video decoding
01:25:15Lloreanandrewg867: I disagree. The whole point of the dB scale on the output is to give users an objective view of the audio levels being produced, so I think it's an issue of getting things set up right. Otherwise I think if we had "Normal" and "High" we'd have to have something strange with how the volume is displayed.
01:25:26Lloreanamiconn: And I do agree, that's a very, very big pro.
01:25:36Lloreanamiconn: Actually being able to watch it is one of the most important features of video.
01:25:43amiconnWith mp3 audio it wouldn't even need to boost...
01:25:50amiconn(on H1x0 / M5)
01:26:28 Join aliask [0] (
01:26:34amiconnlinuxstb: This bitrate is roughly equal to pcm wav. Not excessively high imho
01:27:11*amiconn needs to fix some more plugins for the new charcell code :/
01:28:27andrewg867Llorean: remember that it also depends on the song that you are playing to determine your desired listening/line out recording volume. I found that with just the SVN build it was way too low with my headphones for any song
01:28:47linuxstbYes, not excessively high, but MPEG files at 160x128 don't need to be much more than 200kbps-300kbps, so they will be very large in comparison.
01:29:20Lloreanandrewg867: "Line out" volume shouldn't be used with anything but something that has a line-in. It's not the same concept as a listening level.
01:30:09andrewg867why not have a preset menu option to change the values around so you have a true 'Line out' where you can't change the volume
01:30:25Aloneaok, I am confused, the video file I had on there before that played fine except I didnt have the rebuffering of course, now says the file in incompatible.
01:30:30Lloreanandrewg867: Because 0dB should land perfectly on Line Out level, and -1dB should be exactly 1db below that level, and so on.
01:30:36andrewg867so it might be a menu called "Line out" and in there have "Yes" "No" and custom
01:30:47 Quit marc| (Read error: 110 (Connection timed out))
01:30:48linuxstbAlonea: That message probably means the plugin is incompatible.
01:30:49LloreanThat's the whole point in having an objective-scale volume level.
01:30:50amiconnARtificial option bloat?
01:31:05andrewg867it would technically be -1.25dB I think with the way the headphone amp adjusts gain
01:31:31linuxstbAlonea: i.e. your main "rockbox.xxxx" file and the plugins are out of sync.
01:31:32Alonealinuxstb: I am trying to watch an mpeg and look at the new rebuffering. I am using the same mpeg I was testing out about a month or so ago.
01:31:52LloreanSo is -4dB displayed -5 in reality, and then no change to -5, or is there compensation somehow? (rounding up the display so you get -1, -3, -4, -5)
01:32:02andrewg867each binary increase, ex: from 00000 to 00001 is 1.5dB gain change
01:32:08LloreanOr then -1, -3, -5, -6?
01:32:09Alonealinuxstb: ok, probably messed up somewhere there when I put the new rockbox on.
01:32:12LloreanEr -2, -3, -5, -6
01:32:21andrewg867Llorean: I'm not sure how it is currently calculated
01:32:48LloreanBecause the volume display needs to be objective. It should tell the user actual information about what they're getting out, rather than "You're one 'notch' below the highest level" etc.
01:32:52 Quit yossa ("Chatzilla 0.9.77 [Firefox]")
01:33:04linuxstbLlorean: I think the volume is stored in 1dB units, and then converted to the DAC's units when being set. So it would mean that some increments don't change the volume.
01:33:31andrewg867tenthdb2master() and tenthdb2mixer() in the as3514.c is where it is calculated I think
01:34:07Lloreanlinuxstb: I'm fine with that then. My main concern was if the further you got from 0, the less the displayed volume matched the internal on.
01:34:13Lloreanlinuxstb: And I'm still concerned about distortion at 0dB.
01:34:17 Part JavaMan22
01:35:05linuxstbLlorean: I think it happens elsewhere in Rockbox - e.g. the bass/treble controls on the Nano/Photo are in 1.5dB increments IIRC.
01:35:36Aloneaoh, ok, when it makes the zip, it no longer put the rockbox.gigabeat outside the .rockbox folder...
01:35:47Lloreanlinuxstb: If going from -2 to -3 doesn't change the gain (or -1 to -2) so that -3, -6, -9 all always match up exactly, I'm pretty much fine with it being that way
01:36:38linuxstbAlonea: No. You should delete the version in your gigabeat's root to prevent confusion.
01:37:26Alonealinuxstb: ok, so does the rockbox.gigabeat now go in the inside or outside of .rockbox?
01:37:56Alonealinuxstb: ok. I haven't updated the firmware in ages so i didn't know about the change
01:38:45markunAlonea: if you don't have the latest bootloader you might run into some problems..
01:39:05Aloneamarkun: when was the latest bootloader?
01:39:19LloreanAlonea: Anything after mid-januaryish should work.
01:39:26 Join twenty6ix [0] (
01:39:31LloreanIf I recall
01:39:32AloneaLlorean: nope. its not working. Rockbox error -2
01:39:40LloreanAlonea: Well, time to update.
01:39:55twenty6ixhi there, i have a question about scrolling with the scrollwheel in Rockbox. i've noticed it's not quite as fast as in the Apple iPod default software, anyone know if there's a way to speed it up?
01:40:03linuxstbI thought it was the 12 March release which was the first to check in the .rockbox directory.
01:40:08AloneaLlorean: ok, how do I get the usb to work.
01:40:27AloneaLlorean: wait, nm, there it goes
01:40:28 Join lokasenna [0] (
01:40:29linuxstbtwenty6ix: No. It's being worked on though, so future Rockbox versions will be better.
01:40:32 Part Llorean
01:40:36twenty6ixok, thanks linuxstb
01:40:40midgeylinuxstb: i know for sure that the march 12th release works
01:40:45twenty6ixi just installed Rockbox today and i'm loving it :)
01:40:50twenty6ixamazingly customizable
01:41:14twenty6ixi was worried before i wouldn't be able to watch mp4 movies, and i find mpg way too large once converted, but still being able to boot into the Apple OS is handy :)
01:41:38Alonealinuxstb: ok, where abouts on the rockbox site do I get the bootloader?
01:41:38twenty6ixanyone know of a Texas Hold'Em game for Rockbox?
01:41:44linuxstbIf your mpg file is too large, you're encoding at too high a bitrate.
01:41:57perldiver20min mpeg video file is 100mb
01:42:10midgeytwenty6ix: no such game exists
01:42:11twenty6ixwell i also checked the average framerates
01:42:27twenty6ix10fps for a 320x240 video
01:42:28perldiver(with mpegplayer wiki page vlc bat file settings)
01:42:40twenty6ixguess what resolution most of my iPod videos are
01:42:42lokasennarockbox on my e200 series sansa occasionally flips out, the screen looks like a random mess, is that something thats been documented and is being worked on? or is my sansa just flippy?
01:42:49linuxstbtwenty6ix: Yes, mpegplayer isn't really usable on the ipod video.
01:43:11twenty6ixyeah, that's why still being able to access the Apple OS is handy
01:43:24linuxstbAlonea: Follow the install instructions in the manual.
01:43:30twenty6ixall Rockbox needs is a good mp4 player and it'll be able to take over :)
01:43:53twenty6ixAlonea: i was nervous myself when i did it cause i thought i might mess it up, but it was actually really easy
01:43:55twenty6ixyou'll be fine
01:44:12linuxstbThat's unlikely for the ipod video - the CPU is too slow, and the hardware video decoder is undocumented.
01:44:29 Quit efyx (Read error: 54 (Connection reset by peer))
01:44:46twenty6ixi suppose it's possible someone could rewrite the decoder
01:44:49twenty6ixthat'd be a lot of work
01:44:58twenty6ixa lot of testing
01:45:06linuxstbExactly, and no-one cares enough to do it.
01:45:28twenty6ixmay as well just use the Apple OS for videos
01:46:36twenty6ixthanks for the help guys
01:46:42Alonealinuxstb: ok, sliiight problem. i thought it was connecting to my computer, but its not...I remember on the bootscreen it would tell you commands that you could do if things went wrong
01:47:33twenty6ixoh, also... is there no native FLAC tag support on Rockbox?
01:47:58linuxstbYes, Rockbox supports vorbis comments in FLAC files.
01:48:03SoapFLAC tags have been supersceded by Vorbis comments have they nopt?
01:48:09twenty6ixhmm, maybe these FLACs don't have tags
01:48:11twenty6ixi'll have to checj
01:48:30linuxstbSome (broken) encoders like to put id3v2 tags on everything...
01:48:44Soappeople used to put APE tags on flac, some even did ID3 on flac. There are a lot of funky things people did to them in the past.
01:48:52twenty6ixi'm surprised Rockbox didn't just read the filename and output that
01:48:59twenty6ixi was getting blank artist and title
01:49:14linuxstbIt does - if the WPS is coded correctly.
01:49:19Soapfoobar 0.8 would put APE on them by default, unless you turned that off, regardless of what you had used previously.
01:49:20twenty6ixwell Winamp still lets you tag FLAC files i think
01:50:07perldiverand newest winamp beta has FLAC extended tag support
01:50:17twenty6ixthey aren't tagged, i was right
01:50:20twenty6ixi'll have to tag em
01:51:16 Quit linuxstb ("CGI:IRC")
01:51:19twenty6ixthat was weird, Winamp just crashed when i tried to edit a FLAC tag
01:51:44 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
01:55:20Alonealinuxstb: ok, you rock. video works. Now, are there any limits file size wise?
01:55:45safetydanAlonea, probably just the usual 2GB FAT limit
01:56:10 Quit |Rincewind| ("Cya")
01:56:16Aloneasafetydan: ok, cool.
01:57:25 Quit twenty6ix ("( :: NoNameScript 4.02 :: )")
02:01:51 Quit aliask ("Chatzilla 0.9.77 [Firefox]")
02:02:14linuxstbAlonea: I tested a 1-hour file this afternoon (about 230MB). But as far as I know there shouldn't be any filesize limits beyond the 2GB FAT limit safetydan mentioned.
02:02:38linuxstbThe important counters are 64-bit...
02:02:53 Join joshin [0] (n=josh@unaffiliated/joshin)
02:02:54 Quit jhulst (Read error: 110 (Connection timed out))
02:03:20 Quit anathema (Read error: 110 (Connection timed out))
02:04:25XavierGrsafetydan: I played a bit with mpegplayer for H100, it shouldn't be too hard to port it, though I think that it wouldn't be good on performance
02:04:34Alonealinuxstb: alrighty. thats great! you are loved.
02:05:20safetydanXavierGr, it'd be nice to have it for completeness sakes though. I don't think it'd be particularly usable.
02:05:32 Quit Moos ("<night all>")
02:05:37XavierGrif I understand correctly you just need to initialize grayscale and then pass the gray_bitmap_part function the correct data
02:05:45 Quit entheh ("^~")
02:06:28XavierGrthe sims use the yuv_bitmap_part to draw video (and jpeg) so grayscale will have to use gray_bitmap_part instead
02:07:29XavierGrI am not sure if it works in the end though
02:12:24 Quit GRaTT ("using sirc version 2.211+KSIRC/1.3.11")
02:13:10 Join mako [0] (
02:14:00 Join merwin [0] (n=merwin@
02:15:22 Part mako
02:16:15 Quit My_Sic (Read error: 131 (Connection reset by peer))
02:16:55 Part pixelma
02:18:19 Join lhj [0] (i=c4ce6217@gateway/web/cgi-irc/
02:18:47 Quit lhj (Client Quit)
02:28:15 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
02:34:26 Join webguest96 [0] (i=415fbaa9@gateway/web/cgi-irc/
02:35:54 Quit z35 (Read error: 104 (Connection reset by peer))
02:36:12webguest96when is rockbox 3.1 coming out?
02:36:20Soapafter 3.0
02:36:30webguest96when is that coming out?
02:36:37linuxstbBefore 3.1
02:36:49linuxstbShort answer - there is no release planned.
02:36:50webguest96dates times secneds?
02:37:46 Quit webguest96 (Client Quit)
02:37:55SoapI don't mean to be an ass, it's just the question gets asked many times by "hit and run" IRC users. They pop in - ask the same question that's been asked thousands of times before, then never show up again.
02:38:04Soapand he's gone.
02:38:26linuxstbsafetydan, XavierGr: I haven't tried it on a real target yet, but I'm watching Elephants Dream in greyscale in the h140 sim...
02:38:55safetydanlinuxstb, cool!
02:39:40midgeymaybe we should spend some time cleaning up the ReleaseTodo page?
02:40:35linuxstbHere's a patch -
02:40:45JdGordonhey midgey, thanks for fixing up xobox... I forgot to check colors on the non-colour target :p
02:41:08midgeyno problem, its an issue i came across while converting to the old API
02:41:44midgeywhich macro should I be using for int settings? the old api was rb->set_int()
02:41:45JdGordonhopefully that example is good enough to get the other plugins going... although, I think there is a better way still
02:42:00JdGordonyeah, still use that
02:42:21JdGordonthe MENUITEM_SETTING macro wont work in plugins
02:42:37midgey:) i tried that once
02:43:40midgeycurrently i have the menu choices using MENITEM_STINGLIST and the int menus use set_int
02:43:51midgeybut once you get into an int setting, you can't exit
02:44:38midgeyam i missing the way to jump back into the parent menu?
02:44:53safetydanlinuxstb, that looks like a fairly minor changeset. Any reason not to commit it?
02:44:59dj-fuwrong channel
02:45:24JdGordonmidgey: is your do_menu() in a while() loop?
02:45:31linuxstbsafetydan: It needs testing, and I'm sure I've missed a few things.
02:45:39midgeyerr... oops
02:46:03linuxstbsafetydan: But yes, I can't see any reason not to commit it.
02:46:08linuxstb(once it's finished)
02:46:14JdGordonalso, you should pass an int* to do_menu() so it keeps track of the current selection
02:46:33midgeyJdGordon: is menu_exit unneeded now?
02:47:02JdGordondo_menu does menu_init, show, run and exit all in one :p
02:47:09midgeysweet no more accidental leaks
02:47:39midgeyso do_menu(&menu, &selection) is the call I'm looking for?
02:47:40linuxstbsafetydan: Hmm... It doesn't seem to work on a real target. I'm going to sleep now, but feel free to play with it. You can get a 160x128 sample video from the PluginMpegplayer wiki page.
02:48:08JdGordonmidgey: yep, bingo
02:48:10safetydanlinuxstb, I'll take a look when I get home to a compiler
02:48:28midgeyall right, i try to finish up the conversion tomorrow or sometime soon
02:48:35linuxstbsafetydan: Sound is working, but I just get a white screen.
02:49:02midgeyi currently have to write a voice-controlled Copter game in assembler for a final project
02:49:44JdGordonsounds like fun :)
02:49:59midgeyerr... something like that :)
02:50:26 Quit lokasenna ("Leaving")
02:53:44 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
02:55:57linuxstbsafetydan: Working now, but _very_ slowly.
02:56:28linuxstbI'm sure I must be using the grayscale lib wrongly...
02:56:31XavierGrI just got working mpegplayer in H100
02:56:38XavierGrah I am not alone in this
02:56:59amiconnThe grayscale lib isn't very fast when updating the whole screen
02:57:10XavierGrdamn linuxstb got me first
02:57:13amiconnUsing .rvf is *waaay* faster
02:57:28XavierGrwell the changes are not so many to make it work for H100
02:57:43XavierGrjust proper grayscale lib initialization and that's it
02:58:03XavierGrah and a few gray_show(true/false) here and there (menu)
02:59:05linuxstbYep... Here's my (working, but slowly) patch -
02:59:06XavierGrlinuxstb: did you use gray_gray_bitmap_part or gray_ub_gray_bitmap_part? because I tried the first only to discover it was the second one
02:59:17linuxstbXavierGr: Exactly what I did... :)
02:59:28amiconnYou need to use what matches your init
02:59:40linuxstbWhich would be the most suitable for mpegplayer?
03:00:00amiconnThe grayscale lib supports 2 modes: buffered and unbuffered
03:00:18XavierGrlinuxstb: hehe exactly what I did too but you beat me to it
03:00:33amiconnbuffered causes less flicker and should be faster if only a part of the pixels changes its values
03:00:51amiconnunbuffered causes more flicker, is faster when the whole screen changes, and uses less ram
03:01:04XavierGrnow I will try a fullscreen video it seems viewable :)
03:01:24XavierGrthough rvf has the nice hack with the contrast
03:01:36 Join bonbonthejon [0] (n=jon@
03:03:09linuxstbamiconn: How would you compare the CPU requirements of the grayscale lib with the yuv2rgb conversion - i.e. would you expect the H140 (or M5) to display faster or slower than the X5?
03:03:16XavierGrit sure will be interesting to port the rvf plugin, I will have a look on it but I am sure it will not be easy for me, mpegplayer was pretty straight forward
03:03:27amiconnlinuxstb: Definitely slower
03:04:03XavierGrI will try to measure performance though I doubt that the fps will display on top of the grayscale lib
03:04:08amiconnThe grayscale lib needs to do lots of bitshuffling in lcd_*gray_bitmap*
03:05:20amiconnOn H1x0 it manages to do ~8fps iirc with just fullscreen lcd_ub_gray_bitmap() calls boosted (it's boosted anyway due to the timer usage)
03:05:56amiconnArm seems to like bitshuffling much more: ~26fps on ipod mini G2 under the same conditions
03:06:17amiconn(of course the mini has an advantage here)
03:06:18***Saving seen data "./dancer.seen"
03:07:56XavierGrit is dead slow....
03:08:04XavierGreven with frame skipping is left back
03:08:30amiconnWell, mpeg decoding on grayscale targets could be made a bit faster by skipping luma decoding
03:08:36amiconnErm, chroma
03:08:55 Quit Alonea ("Chatzilla 0.9.77 [Firefox]")
03:10:59XavierGrwell definitely slower than H300 (with the same file of course)
03:11:40safetydanIs the H120 LCD 2-bit natively?
03:11:55XavierGrbut I don't see why it will be bad to include it on SVN, we have rockboy for archoses too, so just for the consistency it would be good to have it on grayscale targets too
03:12:27XavierGrsafetydan: iirc H100 can draw natively 4 shades of gray
03:12:42XavierGrand more with the greyscale lib
03:12:53safetydanI wonder what 4 color videos might look like? Would certainly be faster.
03:13:20linuxstbXavierGr: Is there any difference between your patch and mine? Did I forget anything?
03:13:31LloreanIt's worth a chance, I mean with the blurring of the LCD anyway, I'm not sure there's a huge advantage to graylib in the end
03:13:49XavierGrlinuxstb: let me check again but I don't think you did. And you took the time to #ifdef it properly :)
03:14:18*linuxstb wonders how well mpegplayer works on the Archos...
03:14:39linuxstbs/works/would work/
03:15:26XavierGrlinuxstb: I used the audio buffer for the grayscale init
03:16:06XavierGralso you don't update after the gray_ub function
03:16:31XavierGr gray_update_rect(output_x,output_y,output_width,output_height);
03:17:17linuxstbIs the update needed in unbuffered mode? It works without it...
03:17:49XavierGrI am not sure I just copied actions from the sim part
03:18:08XavierGrdo you uninitialize the greyscale lib at the end of the plugin?
03:18:19linuxstbAh, no...
03:19:18XavierGralso the fps display info won't be able to be shown while the video is playing
03:19:33linuxstbAh, it seems gray_release() just calls gray_show(false), which I was doing.
03:19:45linuxstbBut I've changed it.
03:20:13amiconnlinuxstb: It's currently equivalent, but it's not guaranteed that this stays
03:20:39linuxstbOf course - I was just wondering why it worked without it.
03:21:17XavierGrI say commit, it is fun to have it nonetheless :D
03:21:21amiconnIn unbuffered mode you must not call gray_update()
03:22:23linuxstbamiconn: Can you have a quick look at the patch?
03:23:15linuxstbI guess mpegplayer should now be built for all SWCODEC targets?
03:23:40linuxstb(excluding IFP, which doesn't get anything...)
03:25:12XavierGrwith all this video talk I can't wait for my Gigabeat to arrive...
03:25:39amiconnThe iFP could easily use the grayscale lib (but I guess there is no sufficient ram for mpeg decoding)
03:26:43linuxstbI think the CPU would struggle as well - it's just a single core ARM.
03:27:04linuxstbconfig-ifpxxx.h says the frequency is 30MHz...
03:27:54linuxstbNo wonder he's working so hard on optimisations (assuming that freq is correct).
03:29:56Lloreanlinuxstb: It's 60mhz, in reality, I believe
03:30:15amiconn30MHz is CPUFREQ_NORMAL
03:30:36amiconnIt's CPUFREQ_DEFAULT
03:31:02linuxstbBut yes, RAM would be the problem.
03:31:43*amiconn would like to see the grayscale lib adapted to iFP
03:31:58LloreanIsn't the iFP's screen the same as the H100 remote?
03:32:06amiconnSome plugins which use it don't need much ram. e.g. mandelbrot or cube
03:33:00*linuxstb commits mpegplayer for grayscale lib
03:33:44linuxstbamiconn: I assume you can't draw text in the grayscale lib?
03:34:11XavierGrlinuxstb: nice!
03:34:24XavierGrlinuxstb: not on top of it iirc
03:34:43XavierGrbut I might be wrong on that
03:34:50safetydanlinuxstb, very nice. Didn't think anyone would actually do it :)
03:35:16linuxstbsafetydan: It turned out to be trivial - I never expected it to work immediately...
03:35:21XavierGrsafetydan: yeah it was a shame not to have considering how easy it was to do it
03:35:53 Join toffe82 [0] (
03:37:02linuxstbBut as amiconn said, the next step would be to hack libmpeg2 so it doesn't decode the luma.
03:41:29JdGordonmidgey: looks good, except you shold look for the default: case where do_menu() returns < 0
03:41:42JdGordonwhen the user wants to quit....
03:41:49*linuxstb spots a sim warning
03:42:25JdGordonactually... you should really only quit plugin menus when you ge GO_TO_PREVIOUS or GO_TO_ROOT
03:43:33*linuxstb fixes warning
03:44:14safetydanbah, lousy work proxy blocks downloading things ending in .mpg
03:44:34midgeyJdGordon: does that mean everything's all good or do i need to account for those cases?
03:44:42amiconnlinuxstb: Now you can't, but you can temporarily clip the grayscale overlay in order to show text outside of it
03:44:59amiconngrayscale.c should show how this is done
03:45:38JdGordonmidgey: you _should) check for those cases... unless you dont want to allow the user to quit with the back or menu buttons...
03:45:43JdGordonbut that looks good...
03:45:43amiconnJust remember that you *must not* call any non grayscale lib function that writes directly to the lcd controller while the overlay is running
03:46:10amiconnThis is because the grayscale lib does write to the lcd controller in an isr
03:46:29linuxstbHmm... I should probably disable the "show fps" option for now then...
03:46:47linuxstbAlthough it's off by default.
03:47:15 Join jhulst [0] (
03:47:53amiconnThis is what gray_deferred_lcd_update() is for
03:48:35amiconnYou can call this if you have drawn something outside the grayscale overlay area, and would normally need to do lcd_update()
03:48:58amiconnIt will make the grayscale lib call lcd_update() the next time the isr is running
03:49:05Lloreanlinuxstb: MPEGplayer definitely does lose sync over time.
03:49:21linuxstbLlorean: It doesn't in my 1-hour test file...
03:49:40*linuxstb likes to blame encoders for sync issues
03:50:13perldiverlinuxstb how did you make it?
03:50:19perldiverwhat fps, settings, etc
03:50:29linuxstbThat's the problem, I can't remember....
03:50:44perldiverit gets _slightly_ out of sync
03:50:45linuxstbI made it around Christmas when I first started work on audio.
03:50:50perldiverits visible, not crucial
03:50:50Lloreanlinuxstb: I'll see if it gets worse by the end of the file, and then test that file in a PC-side player (the only copy of the file is on the player)
03:51:02LloreanI'm watching a movie on it right now, and it's lost about a quarter or half second of sync. Not much, but just enough to be clearly noticeable.
03:51:21perldiverwhat Llorean said, same experience here
03:51:44LloreanAnd I didn't notice it until shortly after I felt the disk spinning, making my brain jump to rebuffering conclusions, but I've no evidence of that except an assumption based on very little information. :)
03:51:45linuxstbperldiver: But didn't you say the same happened on your PC?
03:52:43linuxstbLlorean: If it plays in sync on a PC, can you upload the file somewhere, and I'll have a look.
03:52:45perldiverlinuxstb only one encode, initial 29.79 behaves nicely on PC
03:53:05linuxstbWas the source video 29.97?
03:53:19Lloreanlinuxstb: If it does, I will. I'll assume for now that it won't, and that it was the encoder, until I get proof. :)
03:53:35*amiconn thinks we should get rid of yuv_bitmap_part() in mpegplayer for the sim
03:53:46perldiverit happens about 10 minutes in a 20 min file
03:53:48amiconnThe sim needs to implement lcd_yuv_blit()
03:54:04perldiverin in*
03:54:25linuxstbDoes it suddenly lose sync, or is it gradual?
03:55:08perldiverlike Llorean said, it gets half a second out of sync
03:55:08linuxstbBut you can't notice any audio drop-outs?
03:55:17Lloreanlinuxstb: The audio is slightly ahead of the video, for me.
03:55:20perldiverso even if its gradually it's very hard to notice
03:55:39perldiverlinuxstb no, the picture and sound are smooth all the way through
03:56:21Lloreanlinuxstb: I do think it might be gradual though, I _think_ it's getting worse with time. It's hard to tell.
03:56:46*Llorean goes to try it on the PC
03:57:49Lloreanlinuxstb: The file appears to be perfectly in sync on a PC player when I seek to the same part of it.
03:58:06linuxstbWhat about if you play it from the start?
03:58:14LloreanI'm testing that now
03:58:26LloreanThe sync issues weren't noticed until about 15 minutes in, so I'll let you know in 15 minutes
03:58:47linuxstbI'll probably be asleep, but I'll check the logs in the morning.
04:01:30JdGordonif you enable cpu frequency... do you need to rebuild codecs also?
04:02:24linuxstbNo, you shouldn't need to. But you should rebuild plugins.
04:02:27perldiveralso, to connect usb while video is playing is a bad idea, locks the player
04:02:50 Join Joe_ [0] (
04:02:55*linuxstb notes too many bug reports for mpegplayer coming in...
04:07:03*linuxstb sleeps
04:07:21 Quit Joe_ (Client Quit)
04:07:29 Quit Thundercloud (Read error: 104 (Connection reset by peer))
04:09:00 Quit secleinteer ()
04:09:19 Join secleinteer [0] (n=secleint@
04:09:47 Join BHSPitLappy [0] (n=steve-o@
04:11:44 Join xomphos [0] (i=454c8c75@gateway/web/cgi-irc/
04:13:04xomphoshi, i just updated my version of rockbox by copying over the new files and I clicked on a jpg and it said incompatible version or something and wouldnt let me view it-does any1 know wats wrong?
04:13:25Lloreanxomphos: Which new files, exactly, did you copy over?
04:14:37xomphossry but i g2g, 4 some reason my client wont show any text
04:14:45xomphosthanks 2 those who tried 2 help if any tho
04:14:47 Quit xomphos (Client Quit)
04:17:22Lloreanlinuxstb: Yes, the file seems to stay in sync when played on my PC. It's quite large though, I'll try encoding something smaller and see if it experiences the same problem.
04:29:04 Quit borisyeltsin (Remote closed the connection)
04:33:54 Join borisyeltsin [0] (
04:34:45 Quit billytwowilly (Nick collision from services.)
04:34:54 Nick borisyeltsin is now known as billytwowilly (
05:00:55 Join mcs117 [0] (
05:01:06mcs117hey guys?
05:01:21mcs117how to fix a sansa e260?
05:01:32mcs117it always restart
05:03:44mcs117no one here?
05:03:46Lloreanmcs117: People aren't always here, or may not always know how to answer. Sometimes patience is a virtue.
05:04:04LloreanWhat exactly did you do to your sansa?
05:04:17mcs117exactly nuttin
05:04:22mcs117it just crashed
05:04:27mcs117so i restarted it
05:04:34LloreanWell, this isn't a general support channel. This is #Rockbox.
05:04:48mcs117im thinking of switching to rockbox
05:04:55mcs117the firmware
05:05:04LloreanWell, when you fix your Sansa and are ready to install Rockbox, feel free to come back.
05:05:15LloreanBut we can't provide support for the original firmware. We didn't write it, and aren't experts with it.
05:05:30LloreanYour best bet is to simply contact Sandisk's official support channel
05:05:32mcs117how do people revert it back to the orig firmware
05:06:19***Saving seen data "./dancer.seen"
05:06:38LloreanBy uninstalling Rockbox. We just remove the Rockbox data, we don't overwrite the original firmware, so the process of removing Rockbox just slides it back into place.
05:10:05 Quit jhulst ("Konversation terminated!")
05:10:25Lloreanmcs117: Why don't you just contact official support?
05:11:17mcs117i will if the updater doesn't fix it
05:12:10 Join hardeep_ [0] (
05:12:27 Quit hardeep_ (Client Quit)
05:16:53 Join jhulst [0] (
05:20:56 Quit hardeep (Read error: 110 (Connection timed out))
05:22:19 Quit mcs117 ()
05:24:22 Quit jhulst (Remote closed the connection)
05:26:42 Join jhulst [0] (
05:50:33Lloreanlinuxstb: Alright, I tested on both my Nano and my Gigabeat (I was curious if maybe a multicore target didn't suffer the problem) and have a 70mbish file that goes out of sync by the end on my Nano (and a file encoded from the same source at higher resolution does so as well on my Gigabeat). They play back in-sync on the PC side.
06:16:05 Join jhulst_ [0] (
06:16:52 Join ender` [0] (n=ender@
06:16:55 Quit jhulst (Read error: 104 (Connection reset by peer))
06:17:10 Nick jhulst_ is now known as jhulst (
06:17:58 Join lucas42 [0] (i=proxyuse@
06:22:03 Join aims_ [0] (
06:24:39 Part safetydan
06:24:41 Quit aims_ (Client Quit)
06:26:35 Quit Slickire ("CGI:IRC (EOF)")
06:35:24 Quit bonbonthejon (Read error: 131 (Connection reset by peer))
06:39:36 Join jhulst_ [0] (
06:40:21 Quit jhulst (Read error: 104 (Connection reset by peer))
06:42:47 Join ryan_r [0] (n=ryan@
06:42:59 Quit jhulst_ (Client Quit)
06:51:25 Join Vyrus001 [0] (
06:52:17 Quit joshin ("Gotta stop the kids from rioting")
06:52:21 Join joshin [0] (n=joshin@VDSL-130-13-11-10.PHNX.QWEST.NET)
06:52:46 Nick joshin is now known as joshin_ (n=joshin@unaffiliated/joshin)
06:53:47 Join printfXh4 [0] (
07:06:20***Saving seen data "./dancer.seen"
07:10:00 Quit scorche (" Try HydraIRC -> <-")
07:10:31 Join saratoga [0] (i=98039a42@gateway/web/cgi-irc/
07:10:48saratogawell, i may yet regret it, but I put in for the SOC WMA decoder project
07:12:20 Quit saratoga (Client Quit)
07:13:44 Quit dj-fu (Read error: 145 (Connection timed out))
07:29:27 Join dilinger [0] (
07:29:27JdGordonanyone know where the size of the pcm buffer is defined?
07:41:15 Join bonbonthejon [0] (n=jon@
07:48:00 Quit midgey ()
07:49:42 Quit ryan_r (Read error: 110 (Connection timed out))
07:52:11 Join ryan_r [0] (n=ryan@
07:59:33 Join lee-qid [0] (
08:08:10 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
08:08:13 Join miepchen^schlaf [0] (
08:10:51 Part toffe82
08:17:46 Join Rob222241 [0] (
08:26:36LloreanAnyone have a rough idea how much not plugging in headphones would affect battery life / power draw?
08:27:22midkayit would?
08:27:34Lloreanmidkay: No resistance on the circuit, so in theory it'd have some effect, right?
08:27:43LloreanOr maybe even no circuit, rather.
08:27:47midkayi wouldn't have thought so...
08:28:11LloreanI mean some power is needed to move the diaphragm in the headphones etc.
08:28:27midkayi didn't think so..
08:28:38 Join kaaloo [0] (
08:28:49LloreanAnyway, I'm at 3 hours on my "how long can the gigabeat run video" test.
08:28:58LloreanI'm hoping to hit the 3:30 mark that Apple advertises for the iPod Video.
08:29:06midkaynice :)
08:29:19LloreanThe video's at 600kbps with the audio at 128kbps, so it's a bit higher than it needs to be.
08:30:25JdGordonthe gb can keep the vid going in sync?
08:30:31LloreanJdGordon: Nope
08:30:35LloreanIt's playing back the audio too fast.
08:30:39JdGordonah ok
08:30:44Llorean(And I'm 100% sure it's not 'the video is too slow' now)
08:31:14LloreanI didn't note down the time I started my test, so I ran the same video on my PC, and got it aligned almost perfectly, the audio's now far early on the gigabeat compared to the PC, but both have the video perfectly in sync still
08:32:33LloreanIt's possible this is, once more, in the same family of "audio plays slightly faster than it should" that I discovered with speex.
08:32:41LloreanThe one I thought was sample rate related, originally.
08:32:48 Join himitsu [0] (n=himitsu@
08:35:16 Join vadim [0] (
08:35:54 Quit Rob2222 (Read error: 110 (Connection timed out))
08:39:43 Join GodEater_ [0] (
08:41:04 Join B4gder [0] (
08:41:42 Quit BHSPitLappy (Read error: 145 (Connection timed out))
08:43:59 Quit gtkspert (Remote closed the connection)
08:44:15 Join gtkspert [0] (
08:46:38 Quit midkay ("Leaving")
08:46:52 Join midkay [0] (n=midkay@rockbox/developer/midkay)
08:51:37 Join bluebrother [0] (i=kljD0rjX@rockbox/staff/bluebrother)
08:52:01 Join BHSPitLappy [0] (
08:52:20Llorean3:30! And 15% remaining supposedly
08:52:29 Quit Hadaka (Read error: 110 (Connection timed out))
09:04:45 Quit merwin (Read error: 54 (Connection reset by peer))
09:06:22***Saving seen data "./dancer.seen"
09:06:38XavierGrLlorean: not pluggin headphones will result in a slithgly better battery time iirc
09:06:48XavierGrbut not much
09:07:16LloreanXavierGr: I didn't expect it to be much but I was curious if it was less than 5%, really
09:09:01XavierGrI am not sure how much, but I remember someone doing amperage measurements with and without headphones and there was a measureable difference
09:10:15 Join noNickFound [0] (
09:12:02 Quit lucas42 (Read error: 104 (Connection reset by peer))
09:13:32noNickFoundanybody here was played the original xonix game?
09:13:49 Nick noNickFound is now known as std|denis (
09:14:13 Quit ment ("eof")
09:17:53LloreanOkay, just to confirm, the Gigabeat did _not_ have Video in its retail firmware, right?
09:22:26 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:22:41Llorean4 hours, and still going
09:22:55vadimLlorean, righ, no video
09:23:09Lloreanvadim: Strange. I really want to know what they intended with the thing then.
09:23:22LloreanRockbox already gets better battery life watching Videos than the 5G 30gb iPod.
09:23:29LloreanVideos on the Gigabeat, that is.
09:25:03markunLlorean: sweet results man!
09:25:13markunhi vadim btw!
09:25:15Lloreanmarkun: It's still going
09:25:23LloreanI'm at 4:03 and counting. :)
09:25:27 Quit lee-qid (Read error: 110 (Connection timed out))
09:25:31vadimhi markun, :)
09:26:05Lloreanmarkun: And I bet I could've gotten even more time if I lowered the bitrate a bit, I think 600 may be slightly overkill
09:26:39markunvadim: seems like a long time ago when rockbox didn't run on the Gigabeat :)
09:27:00markunLlorean: and lowered the backlight brightness maybe
09:27:22markunand some more decoder optimisations perhaps
09:27:33Lloreanmarkun: Will decoder optimizations help that much without CPU scaling?
09:27:58markunLlorean: yes, the CPU draws more when it's more busy
09:28:08B4gder... and we're on 13 gsoc applications
09:28:08vadimyes:) I'm still going to remove that 100 bytes offset where kernel loads, because i never had time to remove it and there is no real need in it except I was trying to be conservative :)
09:28:12markunat least that was what it looked like
09:28:15Lloreanmarkun: I know it'll help, but I just was curious in what sort of range.
09:28:27LloreanLike, if it's twice as busy does it draw twice as much, or only maybe 5% more?
09:28:49LloreanI'm just curious how much it might affect CPU draw, is all. :)
09:28:55LloreanB4gder: We seem popular.
09:28:57markunLlorean: maybe toffe could do some more measurements to find that out
09:29:05Lloreanmarkun: I'd be very interested to know.
09:29:13 Join aliask [0] (
09:29:15B4gderLlorean: indeed
09:29:24B4gderalthough I read that some get "hundreds"
09:29:39B4gdera bit overwhelming it must be
09:29:52 Join SlAtz [0] (
09:29:52LloreanB4gder: I imagine Firefox, OpenOffice, etc. It must be an effort to organize/rank them
09:29:52 Quit Slowking_Man (Connection timed out)
09:30:19 Join scorche [0] (i=scorche@rockbox/administrator/scorche)
09:30:37*amiconn ponders commiting the first step of player lcd code rework now
09:30:51LinusNgo! go! go!
09:30:53XavierGramiconn: everything that you commit is good :D
09:31:08amiconnNah, jumpscroll still isn't working...
09:31:16markunamiconn: only jump-scrolling is not working?
09:31:40amiconnBut the diff is already over 200K
09:32:02amiconn(a big part is of course the character list)
09:32:11markunthe unicode patch was also quite a big beats at one time :)
09:32:34 Join inversions [0] (
09:32:48amiconnAnd this change will cause a red delta for player
09:33:50 Quit ze ("leaving")
09:34:26markunamiconn: did you add support for cyrillic?
09:34:41amiconnNot yet, but this is a preparation for it
09:35:18amiconnAdding cyrillic is now 'just' a means of extending the character tables and defining the soft characters where we can't map to the hardware directly
09:35:59 Quit idnar (Nick collision from services.)
09:36:01 Join idnar_ [0] (n=mithrand@unaffiliated/idnar)
09:36:55SlAtzCan someone help me a little bit? =) I got Rockbox on my Sansa e200 yesterday,, but i can't figure out how to import my mp3 to rockbox :S
09:37:42SlAtzit's already loaded with mp3, pic's and videos, but i can't find the music while i've loaded rockbox
09:38:19LloreanSlAtz: Your music may be in a hidden folder. Either choose to set the File View in Rockbox to "all", unhide the music folder, or use the database. Both the file view and database options are described in the manual.
09:41:31preglowSlasheri: cool
09:44:13 Join scubacoles [0] (n=scubacol@
09:45:07markunLlorean: don't forget to tell us when the video test ends
09:45:30 Join Slickire [0] (i=5ac0e8c6@gateway/web/cgi-irc/
09:45:32Lloreanmarkun: I intend to
09:45:36Lloreanmarkun: It just hasn't ended yet. :)
09:46:17aliaskWhat's the video test?
09:46:36aliaskBattery life?
09:46:53Llorean320x176, 600kbps video, 128kbps audio.
09:47:03aliaskWhy 176?
09:47:09markunwhy not?
09:47:16Llorean'tis a movie
09:47:31aliaskAhh. Cool. How many hours so far?
09:47:37Llorean4 hours, 25 minuts
09:47:40 Quit himitsu (Remote closed the connection)
09:47:46markunon a F40?
09:48:24SlAtzLlorean: Thanx ;D
09:49:46SlAtzgotta get to school now, c ya'll
09:50:06 Quit SlAtz (Read error: 131 (Connection reset by peer))
09:50:10markunI believe the Gigabeat S is supposed to have 2.5 hours video playback :)
09:51:22Lloreanmarkun: Well, we have 4.5 at least, under my conditions. :)
09:52:14aliaskI think it was attributed to the switch from OLED -> LCD
09:52:41 Join Shaid` [0] (
09:52:52markunaliask: which OLEDs?
09:53:24markunThe S was supposed to have a OLED screen?
09:53:52aliaskI thought the F/X had an oled screen?
09:54:08aliaskStupid wikipedia.
09:54:11markunthe P does
09:54:17markunlittle flash player
09:55:23markunYes, I see that wikipedia claims a OLED display for the X, but I'm pretty sure that's false
09:56:27markunalso, it states a 2.5" screen for the X which I believe is 'only' 2.4"
09:56:29scubacolesI can confirm the X is just a boring old LCD
09:56:40scubacolescorrect again Markun
09:57:38markunluckily it's a wiki
09:58:05markundo you need to create a wikipedia account for each language?
09:58:22amiconnBagder: The website's diff link script still has the problem that you can't diff renamed files because [rename] isn't a link.
09:59:27markunpreglow: does this arcticle make any sense to you?
09:59:57 Quit inversions (Read error: 110 (Connection timed out))
10:00:36B4gderamiconn: right, but there was some problem related to that which is why it is like this now
10:00:47B4gderI don't recall the details right now
10:01:52preglowmarkun: not in the least
10:02:13preglowjust looks like gibberish
10:02:31preglowbut then again, i'm not an expert in compression :>
10:03:15markunpreglow: someone in #vorbis also thought it was nonsense
10:03:50markunI wonder if it should be the same as
10:04:32preglowi've at least heard about that
10:04:42preglowi remember .vqf files from years ago
10:04:45preglownever took off
10:04:59markunpreglow: the problem is that googling the subject gives a lot of hits which I think are just copies of the wikipedia contents..
10:05:15markunYes, I played with VQF a little
10:05:48preglowwell, it just looks like gibberish to me
10:05:59preglowjhMikeS: seems like you snuck some more tabs in when doing the last asm commit
10:06:09LinusN"destroyed hyperspace", isn't that what happened to the Millenium Falcon? :-)
10:06:36LinusNoh, that was the "hyperdrive" :-P
10:06:59markunLinusN: I remember the word mainly from the HHGTTG :)
10:07:08 Join DarkTurtle [0] (
10:07:44markunfestival TTS kept pronouncing that word completely wrong :)
10:08:01LinusNaha! so it is actually 42 = 5.22 / 4m !!!
10:08:03DarkTurtleoh heyy markun :P
10:08:43markunDarkTurtle: did you see that linuxstb added video rebuffering?
10:08:55DarkTurtleoh no i havnt been in the channel
10:09:12DarkTurtlebut ive made a theme :)
10:09:13LinusNthis must be the first time i see an "oops" commit from amiconn
10:09:30DarkTurtleuhm.. it says that i need 2 go here 2 ask for write permissions for the wiki
10:09:39DarkTurtlecan i have that so i can add my theme?
10:09:44LinusNDarkTurtle: what is your wikiname?
10:09:55LinusNhang on
10:09:59DarkTurtlekk thx :)
10:10:40LinusNDarkTurtle: done, happy wiki editing!
10:10:46DarkTurtlelol thx. :)
10:10:48 Quit BHSPitLappy (Remote closed the connection)
10:11:00*Llorean wonders if his video's planning to stop before it beast the 64mb iPod model.
10:11:11markunLlorean: how long is the movie?
10:11:17amiconnLinusN: Huh?
10:11:28Lloreanmarkun: The movie is 1:45, I'm on my third viewing of it.
10:11:29amiconnI did quite a lot of oops commits over time...
10:11:39LinusNyou usually get things right
10:12:00 Quit Shaid (Read error: 110 (Connection timed out))
10:12:01 Nick Shaid` is now known as Shaid (
10:12:03markunamiconn: maybe it's just because you get so anoyed when other people didn't test their patch before commit :)
10:12:05amiconnThis .h was a last-minute removal, and I didn't re-check build afterwards
10:12:16LloreanAh, it died.
10:12:21Llorean4 hours, 50 minutes.
10:12:33markundamn, that has to be improved :)
10:12:41Slasheriroolku: (if you can see this from logs), i hope the recent change does not affect the playcount/score of the last track when stopped with stop button. Otherwise scores are getting wrong if you just stop the playback in the middle of track and then resume it later. It should happen only at end of playlist
10:12:49LloreanAnd my battery wasn't _quite_ 100% full (maybe 97%ish)
10:12:57LloreanBut still, a good time for a first test. :)
10:13:14amiconnLlorean: That clearly beats rvf playback on archos recorders (with stock batteries)
10:13:15markunLlorean: power usage should drop by 2mA for each lower backlight brightness level
10:14:07Lloreanmarkun: I'm sure the time can be improved further.
10:14:22LloreanThough nearly 5 hours is quite good for video playback. Isn't that "normal" for PMPs?
10:14:51amiconnThis diff: (not viewable from the [rename] non-link) was the boring part...
10:18:11amiconnThe tricky part is in the new (renamed) lcd-charcell.c
10:20:58amiconnBtw, it looks like quite some space-saving would be possible by statically initializing the screen_access structures
10:22:42amiconn(should then even be const in order to be put in ROM for rombox)
10:24:11markunLlorean: yes, I think it's quite normal
10:27:52 Part LinusN
10:36:16 Join My_Sic [0] (
10:38:37 Join Llorea1 [0] (
10:40:00 Quit Llorean (Nick collision from services.)
10:40:01 Nick Llorea1 is now known as Llorean (
10:42:26 Quit scubacoles ("Leaving")
10:44:49 Join merwin [0] (n=merwin@
10:51:01linuxstbLlorean: Nice video runtime test, but I think a fairer comparisons with original firmwares would be a fullscreen video - it's much easier for mpegplayer to decode a 320x176 video than 320x240 (see the relative FPSs on the wiki page).
10:52:09 Join webguest13 [0] (i=c19fe442@gateway/web/cgi-irc/
10:52:48Lloreanlinuxstb: Yeah, I noticed that after I was about 4 hours in.
10:53:05Lloreanlinuxstb: I plan to do that next, though hopefully with a video where the audio stays in sync. :-P
10:53:34 Quit webguest13 (Client Quit)
10:53:36LloreanThe video is playing back at an equal speed to that on my PC, but the audio is playing slightly faster than my PC plays the audio.
10:55:19amiconnLlorean: You could try a synchronous playback test with an audio file (PC and gigabeat). Maybe the output sample rate on the gigabeat is slightly off?
10:55:50Lloreanamiconn: I think the output sample rate on all SWCodec is slightly off
10:56:22Lloreanamiconn: I was having a similar problem with speex, and discovered it also applied to AAC, with synchronous playback tests between my PC and my H120 and Gigabeat
10:56:27 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
10:56:31LloreanBut I had a very hard time reproducing it consistently
10:56:35 Join Rondom [0] (
10:57:04linuxstbThe clock of the DAC shouldn't affect a/v sync though - video is synced to the audio, so if that clock is wrong, both streams should be equally wrong...
10:58:18Lloreanlinuxstb: What if you have limit FPS on, and the file is 25fps, and the audio is slightly faster than realtime? What would it do to try to catch the video up to the audio if the video needed to go faster than 25?
10:58:55 Join Thundercloud [0] (
11:01:07 Join himitsu [0] (n=himitsu@
11:02:34Lloreanlinuxstb: I can provide a 22 minute file that is definitely out of sync by the end, it's about 70mb if there's a good way to get it to you (or someplace to upload to)
11:04:25B4gderI can provide the upload/download service
11:06:24***Saving seen data "./dancer.seen"
11:06:56LloreanB4gder: Mind if I go ahead and upload on the assumption that linuxstb will want to see it when he's got time?
11:09:40JdGordonplayback menu is now using the new menu api... :) are there any changes it needs? or new functions for it?
11:10:07JdGordonplugins cant use icons from recorder/icons.h :'(
11:12:43linuxstbLlorean: Yes, I want to see it, thanks.
11:13:05linuxstbLlorean: Can I just check you have both Limit FPS and Skip Frames enabled?
11:13:09Lloreanlinuxstb: Yes
11:13:47 Join BigBambi [0] (
11:13:55Lloreanlinuxstb: Both are enabled, and the audio is very clearly out of sync by the end of the file.
11:15:25amiconnJdGordon: The metronome actions for player are totally messed up. It's unusable and you cannot quit
11:15:52amiconnOne more case of the actions concept not fitting plugins at all
11:16:23amiconn(found this while checking for plugins which needed adjustment for the new player lcd code)
11:16:59JdGordonhold stop doesnt quit?
11:17:28amiconnHold stop shuts down the player by force less than a second
11:17:34JdGordonI havnt lost all hope for actions in plugins just yet...
11:17:42JdGordonoh.. ok
11:18:20JdGordonsholdnt it just use BUTTON_OFF then?
11:18:22amiconnMetronome has mapped volume up/down to Play/Stop, which is just weird, and pause/unpause is impossible
11:18:31amiconnThere is no BUTTON_OFF
11:18:39JdGordonSTOP even
11:18:42amiconnBUTTON_STOP is the poweroff button
11:18:56 Join My_Sic [0] (
11:19:10amiconnVolume should be mapped to Menu+Left/Menu+Right, like in wps
11:19:19amiconnAnd Play should be pause/unpause
11:19:27amiconnOn is tap, that's ok
11:20:44*JdGordon 's pluginlib_action.h file is full of WIP's....
11:23:15 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
11:26:19*amiconn spotted an almost 3.5 years old quirk in the bitmap vkeyboard code, added by LinusN back then...
11:26:32amiconnThe vkeyboard buttonbar strings aren't localised...
11:26:50LinusNha - i took you 3.5 years to find it - i won!
11:26:56LinusN"it" even
11:28:29amiconnI'm not sure whether this (Nov 2003) came even before localisation, and the one adding the latter just overlooked it
11:37:49 Quit DarkTurtle ("KVIrc 3.2.0 'Realia'")
11:42:13LloreanB4gder: File done sending.
11:51:12 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
11:52:46Lloreanlinuxstb_: You may have just missed the link to the file.
11:57:35linuxstb_I think I did... I'll check the logs.
11:57:43 Quit linuxstb (Nick collision from services.)
11:57:48 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
11:58:26linuxstbDownloading now...
12:01:45std|denisis it possible to boot from iRiver iHP-1xx? as from the external hdd
12:03:02Lloreanstd|denis: As far as your computer's concerned, it is just an external HDD.
12:04:39std|denisi'm just got a dead XP today [missing ntldr], so i'm wonder whether it possible to use my player as recovery tool, that is always with me :)
12:05:31 Join Moos [0] (
12:06:27 Join Shaid` [0] (
12:07:07Lloreanstd|denis: Remember, the h100 takes no power from USB, so the battery will have to be charged enough
12:08:02std|denisLlorean: of course. but its ok if there's no other way - have no CD drive, have no floppydisks
12:08:15LloreanYeah, just wanted to make sure you were aware of that
12:08:22std|denisjust as last chance to repair
12:08:24LloreanMost of our players can be powered or charged via USB.
12:08:41std|denisaha, i know that pity thing
12:09:23linuxstbLlorean: What was the source for your video? I noticed its encoded at 24fps, so maybe the problem is that it should in fact be 23.976
12:09:25*JdGordon slaps the build server with a large trout...
12:09:34JdGordonwhat the heck are you complaining about?? :'(
12:10:10linuxstbLlorean: Also, out of curiousity, which PC video player are you testing with?
12:10:27 Quit Rondom ("Ex-Chat")
12:10:49std|denismp3-codec seem to be not a bad-input proof
12:11:29std|denisi have an audio file, that hang my player, and uisim too
12:11:36Lloreanlinuxstb: The original is 23.98 (according to media player classic), I'm guessing that's rounded. The transcode batch has fps=24 in it, I assumed it's converting. The PC side video player I watched it in was Media Player Classic using the ffdshow filters.
12:12:35JdGordonis 64bit?
12:13:13amiconn#define SYS_USB_CONNECTED ((SYS_EVENT | ((long)1 << 27)))
12:13:18amiconn(in kernel.h)
12:13:29JdGordonah, ok
12:13:38amiconnComapre this with: int do_menu(const struct menu_item_ex *menu, int *start_selected);
12:14:16JdGordonah woops... used the wrong define then
12:14:38linuxstbLlorean: Could you try removing the fps=24 from the encode options? I'll still investigate your 24fps encode and see if I can make it playable though.
12:14:52 Join cadu [0] (
12:15:23amiconnHmm, and SYS_EVENT is 0x8000000, which is negative as a long32, but very positive as long64
12:15:37Lloreanlinuxstb: The file that went out of sync on the gigabeat was a 25fps original file without any changes applied to the framerate.
12:15:41amiconnThe latter always yields false when compared with an int
12:16:00amiconn...since an it that large *is* negative, and stays negative when extended to a long64
12:16:05linuxstbI think the issue is that mpegplayer bases its av sync purely on the "frame duration" of each video frame, rather than the PTS (presentation timestamp) stored in the program stream container. So if the frame duration is inaccurate, then you'll get problems.
12:16:40JdGordonso, the easy solution s disable sim builds on 64bit hosts.. :D
12:17:23amiconnThe easy solution is to code correctly :P
12:17:53Lloreanlinuxstb: Does that explain the 25fps file on my gigabeat though?
12:18:19 Quit himitsu (Remote closed the connection)
12:18:56amiconnlinuxstb: Can you make it use the PTS?
12:19:00linuxstbLlorean: What was the original source for the 25fps file? i.e. do you know its history?
12:19:16linuxstbamiconn: Yes, I think I should. It's just more complicated.
12:21:08 Quit BigBambi (Read error: 60 (Operation timed out))
12:22:45SoapLlorean: is this the file off a quarter second (or so) after 15 minutes?
12:23:17LloreanSoap: Give or take, yes.
12:24:29Lloreanlinuxstb: I don't know the history of the 25fps file well enough. And now that I think of it, my friend is rather fond of point and click for encoding DVDs.
12:25:03*linuxstb wonders if Rockbox's playlist features could be extended to use viewer plugins - jpeg, mpegplayer etc
12:25:09Soap (I don't know if the math is even right) 23.976/24=.999 15*60*.001=0.9 = almost one second off. Like I said, not sure of the math.
12:25:29Lloreanlinuxstb: I've thought of that before, but there'd have to be something in viewer.config that marks them as valid or not for playlisting
12:25:34SoapDVD movie or TV show?
12:25:46 Join BigBambi [0] (
12:25:49*Llorean imagines listening to music for 35 minutes and then hitting a .txt because he just inserted a folder.
12:25:59 Quit Shaid (Read error: 110 (Connection timed out))
12:26:00 Nick Shaid` is now known as Shaid (
12:26:10LloreanSoap: The 25fps one is a movie.
12:26:18amiconnlinuxstb: Hehe, that sure won't be gapless then, but it would 'magically' sort-of integrate wav playback on archos
12:26:21linuxstbLlorean: Yes, a directory with mixed content would be confusing... The user probably doesn't want music stopped and a text viewer popping up...
12:27:15LloreanIt'd also magically sorta make MIDI able to work
12:27:24linuxstbLlorean: Is the loss of sync in the 25fps file worse than your 23.976fps file?
12:27:36Lloreanlinuxstb: It appears to be about the same.
12:28:15Soapohh, two different mpgs
12:28:17LloreanI'm putting a 23.976 version of the television episode on my Nano now.
12:29:33linuxstbOK, that's odd... I'm convinced my 25fps test file was still in sync after an hour. This was encoded from a rock-solid 25fps source MPEG file.
12:30:19 Quit crashd ("Changing server")
12:30:23linuxstbI think the first test would be to check the PTS values of the first and last frames in the file, and see if they match the expected duration of the number of frames in the file...
12:31:03 Join himitsu [0] (n=himitsu@
12:31:10JdGordondoes anyone know where the pcm buffer size is defined?
12:31:37linuxstbamiconn: Viewers could be told to "play next N files in playlist" - meaning gapless could work...
12:34:03 Join lorijho [0] (n=lorijho@
12:34:19lorijhoHi guys
12:35:09 Join decayedcell_ [0] (
12:35:45*JdGordon doesnt like the idea of non-music tracks in the playlist...
12:37:19lorijhoAny of you using the voice functions in RockBox, e.g. the spoken directory names ?
12:37:53 Part LinusN
12:40:00 Quit decayedcell_ (Remote closed the connection)
12:40:19 Join decayedcell_ [0] (
12:41:00 Join Rondom [0] (
12:41:18JdGordonLlorean: if you boot your sansa with usb connected, does the OF connect to usb automatically?
12:41:39 Join crashd [0] (
12:42:47LloreanJdGordon: Yeah
12:43:05linuxstbLlorean: According to the first and last PTS values, your file is 1319.958 seconds long, but counting the frames (at 24fps) gives 1320.375 seconds. Counting frames at 23.976fps is even worse - 1321.697s
12:43:37JdGordonok, wierd... mine wont connect unless I disconnect usb first!
12:43:45linuxstb(the file has 31689 video frames)
12:44:39 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
12:45:17LloreanJdGordon: If I plug in a USB cable, that's all I have to do, it'll boot up and connect to USB.
12:46:06Lloreanlinuxstb: Wouldn't that mean, though, that the video is playing too slowly compared to the sound?
12:46:54linuxstbYes. But I thought that's what you said? i.e. that audio was before the video?
12:47:26LloreanYes, but when I played the 25fps file side by side with the PC, it stayed perfectly in sync with the video on the PC.
12:47:39LloreanWould the PC player be slowing down the audio to compensate then?
12:48:46linuxstbPossibly, yes. I guess it would depend on what the PC player uses for the master clcok.
12:49:11LloreanAlso, the file at 23.976 still goes out of sync
12:49:57JdGordonwoiuldnt the PC program be holding the frames on the screen longer instead of slowing the audio down?
12:50:05JdGordonthat makes more sense to me...
12:50:19LloreanJdGordon: If it were holding the frames on the screen longer, then it would've gone out of sync visually with the file playing on the player.
12:50:38 Join cadu_ [0] (
12:50:43JdGordonah ok
12:50:58LloreanThat's the thing, both Rockbox and my PC play the video at identical speeds.
12:51:06LloreanThey seem to play the audio at different speeds.
12:51:34cadu_hello guys! one question: i have tagged all my songs accordingly and correctly, would like to know what apps can be used 'efficiently' for synchronizing big libraries onto a rockboxed player....
12:51:52 Quit cadu (Read error: 110 (Connection timed out))
12:52:02cadu_i mean, like reproducing the library on a folder structure on the player and being able to identify removed files, renamed, retagged and just change these on the device accordingly
12:52:18decayedcell_using Tag&Rename
12:52:24decayedcell_you can make folders
12:52:31decayedcell_and rename the music according the tags
12:52:51decayedcell_e.g. artist/album/song
12:52:57cadu_my songs are already on a big library separated by artist/album/song
12:52:58decayedcell_that'll get it into a folder structure
12:53:09cadu_and they're all tagged and well renamed
12:53:17decayedcell_what player do you have? you could just copy it all to the player if its not an ipo
12:53:39 Join obo [0] (i=522e52e0@gateway/web/cgi-irc/
12:53:48Lloreandecayedcell_: And why exactly can't you copy it all to an iPod?
12:53:56linuxstbdecayedcell_: You can copy it all to the player if it's an ipod as well... You just can't use the original firmware (but that applies to other targets, such as the gigabeat as well)
12:53:56obocadu_: unison or rsync might be worth a look
12:54:07cadu_just wanted a way to do it automagically, like add a new album to this library and plug my rockbox, hit update/synchronize and it would update the new songs, delete the songs i deleted, retag things needed, like if i changed the name of a song from 'The Sirn' to 'The Siren'
12:54:33cadu_btw i did all the tagging job in Foobar2000 then made a nice script to rename/move all albums and songs on a nice tree structure
12:54:36decayedcell_yeah I forgot about that detail
12:54:45cadu_i have a toshiba F with rockbox
12:55:08cadu_obo, i was thinking about that, those mantain like a small database of files, so they just updated the updated files
12:55:23cadu_obo, heheh , that was a question :P
12:55:52obounison creates a database, rsync doesn't
12:56:15JdGordonLlorean: can you do me a wuit favour please? how many threads are running on your sansa?
12:56:35JdGordonand the bottom how many are on the COP?
12:56:41 Part decayedcell_
12:57:28cadu_obo, so if at host computer i add a folder (meaning new album, for example) to another folder and run unison, it'll only update the new folder ?
12:57:43JdGordonactually... the h300 and the sansa should have the same threads going right? except the extra "main" thread on the COP?
12:58:00LloreanJdGordon: 12, 2 of them are on (1)
12:58:23cadu_obo, unison and rsync seem nice alternatives, but ONE question remains in my head : my player has 20gb and i have like 24gb on my library, how would it cope with that?
12:59:55obocadu_: yes, it'll copy the new folder over. You can adjust the config file to exclude various paths/folders to cut things down a bit
13:00:43cadu_obo, it would nice to script over unison or something to make it go copying files from folders randomly until it fits 20gb
13:01:00cadu_like populating the player, you would have the biggest selection, not everything would be complete...of course :P
13:01:39cadu_obo, my library is composed of 4 folders : rock / electronica / brasil and _unorganized
13:02:27cadu_the new files/album folders are directly thrown in _unorganized, and my script in foobar shows them with a red color + missing fields, meaning : must organize it
13:02:38 Join james_75 [0] (i=james_ri@
13:02:39 Quit james_75 (K-lined)
13:02:49cadu_after i organize/tag it the warning goes off and i just have to move it to one of the three organized folders
13:03:02*lorijho came up with a small hack that creates files (spoken directory names) using (more) natural voices, I'm sure it's already been done before, anyway have a look at it...
13:04:13obocadu_: it's designed to sync one path to another - I'm not sure it'll do exactly what you're after...
13:04:54cadu_obo, i'm reading the documentation now, thank you :) by the way, do you know some of those 'synchro' programs like MediaMonkey ?
13:05:08cadu_but mediamonkey kinda sucks on the 'generic player' export option :(
13:05:18cadu_and inserts funny tags on your files too
13:05:19 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
13:06:26***Saving seen data "./dancer.seen"
13:07:41 Part LinusN
13:07:54amiconnping pong...
13:08:44*JdGordon 's new os stacks screen is soo fubared :D
13:09:32 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
13:14:08JdGordonis there any reason the debug os screen isnt in the sim?
13:14:11cadu_obo, what do you use ? :P
13:14:59linuxstbLlorean: Playing your file in mplayer, there's a value (called "ct" - for correction) which mplayer uses to maintain AV sync. This value starts at around 0.125 seconds, and is slowly increasing. Meaning there is some drift that mplayer takes account of and corrects....
13:15:54Lloreanlinuxstb: Ah, so the file itself is out of sync, but mplayer corrects for it and mpegplayer doesn't?
13:17:14LloreanIs ct an artifact of bad encoding, or is it something normal?
13:17:45linuxstbAre they different things? Maybe bad encoding is normal....
13:18:05linuxstbBut either way it seems very common so mpegplayer will need to deal with it.
13:18:17LloreanBad as in "non-standard" or something similar. :-P
13:18:30LloreanHow big of a fix is it?
13:19:30 Quit Kitt0s ("( :: NoNameScript 4.02 :: )")
13:19:40linuxstbIt probably isn't that much work. Currently mpegplayer ignores the PTS values (which is bad), at the very least it should take account of the offset between audio and video at the start of the stream. Your issue seems to be that this difference drifts throughout the file.
13:19:50 Join Kitt0s [0] (
13:20:18linuxstbSo PTS handling needs to be implemented anyway, and once that's done, calculating the PTS drift and adjusting the av sync should be relatively trivial.
13:21:40preglowlinuxstb: i'd say it's common anyway, yes
13:23:07linuxstbAnd just to annoy everyone, PTS values are 33 bits....
13:23:11 Quit Kitt0s (Client Quit)
13:23:34amiconnlinuxstb: Can't you just sync on the pts values, ignoring the fps value?
13:24:07amiconnI mean, instead of waiting a fixed precalculated frame time before showing the next frame, wait until the next pts is due
13:24:24linuxstbYes, that sounds sensible.
13:25:30*amiconn doesn't know whether the pts is optional
13:25:53linuxstbI was just wondering that...
13:26:53JdGordonok, who knows why a list item wont scroll when its selected?
13:27:03daurnimatorhi all
13:27:11LinusNJdGordon: huh?
13:27:16JdGordonevening daurnimator
13:27:32LinusNJdGordon: i guess the string length calculation is wrong?
13:27:35JdGordonLinusN: ive got a gui list set up, and the item which needs to scroll wont.. even if its the selected item
13:27:38Lloreanlinuxstb: A FAQ I'm reading says neither PTS nor DTS are mandatory
13:27:45daurnimatorJdGordon: sup
13:28:09LinusNJdGordon: perhaps the string length calculation is done with the default font by mistake?
13:28:27linuxstbLlorean: OK... I need to dig out my copy of the standard anyway.
13:28:53daurnimatorJdGordon: what suburb you in?
13:28:54JdGordonshouldnt be... not doing anything in the screen which would tell it to use the SYS font instead of UI...
13:28:58daurnimatorwas in glen eira
13:29:00JdGordoncaulfield north
13:29:35daurnimatornear that hospital?
13:29:57JdGordonWOW... scrolling on the sansa causes some odd artefacts :)
13:30:07daurnimatorhow is the sansa?
13:30:15JdGordonbut it does scroll in the tree and not my list :'(
13:30:18JdGordonits niice :)
13:30:45daurnimatorwhich model do you have?
13:31:13 Join roolku [0] (
13:31:29 Nick BjoernErik is now known as Bjoern-Erik (
13:32:41 Join ment [0] (
13:33:36roolkuSlasheri: I don't quite agree. If I stop half-way through a track (to switch the DAP off or select another track in the file browser) stats should be recorded for that track.
13:35:15Slasheriroolku: but if that was a good track and you stop playback, then the score will be lower
13:35:20roolkuSlasheri: your case where the track is resumed later is just one possibility and should be handled in the resume (because only then is it clear that a resume actually occurs)
13:35:28preglowSlasheri: got a patch? :)
13:35:49Slasherieventually you need to stop playback anyway and you dont want to wait until track is nearly finished to do that
13:36:02Slasheripreglow: i found some new issues, it's gonna take a while :)
13:37:33roolkuSlasheri: I suppose it depends on what you think is more important. For me it is the fact that I played the track at all
13:38:05 Join kavorka^ [0] (
13:38:31JdGordonamiconn: should I use the gui list for the player as well as the bitmap lcds?
13:38:38LloreanWouldn't the easiest thing to do be to score the song as a value (seconds played)/(song length)?
13:38:40Slasheriroolku: indeed, it's good probably good that the playcount increases. But i am worried about the negative effects to the autoscore
13:38:50LloreanIf you play the song through two hole times, it'll be a value of 2
13:38:58LloreanIf you play half the song, stop, resume, finish, it'll be 1
13:39:05Slasheriif it was a bad track, and you stop playback because of that, then it wouldn't be an issue
13:39:35Slasheribut if you would stop the playback anyway (or pause it and auto shutdown happens), score will be generated wrong
13:40:40roolkuSlasheri: I am not a big fan of the auto score thing I must admit
13:40:49kavorka^have a iaudio X5 60gb, have correctly extracted the ./rockbox folder in the root directory and placed the x5_fw.bin file in the firmware directory as the manual says. however when i power up with the charger nothing it goes to the native o/s. what am i doing wrong?
13:41:32 Quit obo ("CGI:IRC")
13:41:40LinusNkavorka^: it didn't update the firmware?
13:41:59kavorka^LinusN, nope...just went into 'charging' mode
13:42:07kavorka^like it normally does
13:42:18Slasheriroolku :) i use it quite often and that's why i think only full play of track (after end of playlist, or skip to other track, or end of track) should be accounted to the db
13:42:19LinusNis it perhaps an x5v?
13:42:35kavorka^x5v is without the fm radio yes?
13:42:36LinusNor did you download the x5v bootloader by mistake?
13:43:00amiconnJdGordon: I would think so. The current thread list on player isn't that helpful because it just displays the thread numbers and stack usage
13:43:03LinusNyes, x5v is without fm
13:43:04Slasheriroolku: btw, the manual rating support would be great as well :)
13:43:07roolkuSlasheri: I understand your problem, but I think the assumtion that the longer I listen to the song the better it is seems problematic
13:43:22Slasherithat's true..
13:43:44*roolku has manual rating
13:43:52kavorka^LinusN, defintely have the x5 60gb model with fm radio...and have d/l the x5_fw.bin not x5v_fw.bin
13:44:10kavorka^its something stupid im doing...i can feel it
13:44:12Slasheriroolku: hehe, just prepare and commit that too
13:44:14LinusNkavorka^: then i don't know what would be wrong
13:44:26roolkuSlasheri: I am just not sure the rest of the patch (lyrics * album art) is wanted in svn
13:44:29kavorka^LinusN, thanks...ill try again
13:44:42LinusNkavorka^: and remember to safely remove the usb
13:44:42roolkuSlasheri: so I would need to seperate it out...
13:44:45Slasherime neither, those can be problematic
13:44:52Slasheribut the rating support should be fine
13:45:46roolkuSlasheri: okay, I'll do it when I find the time
13:47:41JdGordonLinusN: I fgured it out... I forgot to stop it redrawing the list on ACTION_NONE :p
13:47:51Slasheriroolku: great
13:47:54roolkuwhat range should I use? currently I have 0=unrated and then 1-5 possible 'stars'
13:47:57LinusNJdGordon: aha...
13:48:15JdGordonamiconn: this screen inst in the sim, can you test a patch before I commit for the player?
13:48:16Slasheriroolku: i think something like that is fine, but no more than 10 steps
13:48:17roolkubut I think the old rating was on a scale from 0-9 (or 0-10 even)
13:49:02Moosroolku: 0-10 have my vote, more discriminatif rating
13:49:42 Join jborn__ [0] (
13:50:04 Join Rob2222 [0] (
13:50:58 Join _jhMikeS_ [0] (
13:52:24 Quit cadu_ (
13:52:24 Quit crashd (
13:52:24 Quit std|denis (
13:52:24 Quit ender` (
13:52:24 Quit secleinteer (
13:52:24 Quit amiconn (
13:52:24 Quit eggy (
13:52:24 Quit dan_a (
13:52:24 Quit RaRe` (
13:52:24 Quit FOAD (
13:52:24 Quit kclaf (
13:52:24 Quit NuclearDog (
13:52:24 Quit bospaadje (
13:52:24 Quit relaxed (
13:52:24 Quit ampleyfly (
13:52:24 Quit lostnihilist (
13:52:24 Quit Overand (
13:52:24 Quit tchan (
13:52:24 Quit slarti (
13:52:24 Quit linuxstb (
13:52:24 Quit Stalwart (
13:52:24 Quit desowin (
13:52:24 Quit Soap (
13:52:24 Quit daurnimator (
13:52:24 Quit smolyn (
13:52:24 Quit Lynx_ (
13:52:24 Quit bun-bun (
13:52:24 Quit netmasta10bt (
13:52:24 Quit hcs (
13:52:24 Quit BrianHV (
13:52:24 Quit lostlogic (
13:52:24 Quit roolku (
13:52:24 Quit idnar_ (
13:52:24 Quit gtkspert (
13:52:24 Quit Rob222241 (
13:52:24 Quit miepchen^schlaf (
13:52:24 Quit marc (
13:52:24 Quit Ribs_ (
13:52:24 Quit Everybody|away (
13:52:24 Quit jhMikeS (
13:52:24 Quit VeePee (
13:52:24 Quit Rick (
13:52:24 Quit z0de (
13:52:24 Quit Bagder (
13:52:24 Quit sslashes (
13:52:24 Quit Nimdae (
13:52:24 Quit Kohlrabi (
13:52:24 Quit toer (
13:52:24 Quit jborn_ (
13:52:24 Quit luckz (
13:52:24 Quit Vakio (
13:52:24 Quit nixphoeni (
13:52:24 Quit ment (
13:52:24 Quit lorijho (
13:52:24 Quit merwin (
13:52:24 Quit bluebrother (
13:52:24 Quit joshin_ (
13:52:24 Quit perldiver (
13:52:25 Quit DataGhost (
13:52:25 Quit Nibbier (
13:52:25 Quit atsea-142 (
13:52:25 Quit atsea- (
13:52:25 Quit andrewg867 (
13:52:25 Quit dukeman (
13:52:25 Quit pill_ (
13:52:25 Quit bagawk (
13:52:25 Quit sneakums (
13:52:25 Quit pabs (
13:52:25 Quit melter (
13:52:25 Quit crwl (
13:52:25 Quit Weiss (
13:52:28 Join Naked [0] (i=naked@
13:52:29 Nick Naked is now known as Hadaka (i=naked@
13:52:35NJoinmiepchen^schlaf [0] (
13:52:58NJoinRick [0] (
13:53:09NJoinpill_ [0] (
13:53:38NJoinVakio [0] (
13:53:56NJoingtkspert [0] (
13:54:05NJoinandrewg867 [0] (
13:54:08NJoinNimdae [0] (
13:54:26NJoinlorijho [0] (n=lorijho@
13:55:46JdGordonok, it seems we cant have the thread names updating correctly and scroll... so im rearranding the text so the usage is after the status and before the name
13:55:53JdGordonso we dont crae about no scrolling
13:56:53NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
13:56:53NJoindesowin [0] (n=desowin@unaffiliated/desowin)
13:56:53NJoinStalwart [0] (n=stalwart@
13:56:53NJoinSoap [0] (n=Soap@rockbox/staff/soap)
13:56:53NJoindaurnimator [0] (n=fake@unaffiliated/daurnimator)
13:56:53NJoinsmolyn [0] (
13:56:53NJoinBrianHV [0] (
13:56:53NJoinLynx_ [0] (
13:56:53 Join bun-bun [0] (n=bun@unaffiliated/bun-bun)
13:56:53NJoinnetmasta10bt [0] (
13:56:53NJoinhcs [0] (n=agashlin@rockbox/contributor/hcs)
13:56:53NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
13:56:54***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
13:57:22NJoinEverybody|away [0] (
13:57:36 Join marc| [0] (
14:00:34XavierGrSweet! RVF for H100 :D (no sound though...)
14:00:43NJoinluckz [0] (
14:01:13NJoincrwl [0] (
14:01:13NJoinsslashes [0] (i=sslashes@
14:01:13NJoinment [0] (
14:01:13NJoinroolku [0] (
14:01:13NJoincadu_ [0] (
14:01:13NJoincrashd [0] (
14:01:13NJoinstd|denis [0] (
14:01:13NJoinender` [0] (n=ender@
14:01:13NJoinsecleinteer [0] (n=secleint@
14:01:13 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
14:01:13NJoineggy [0] (n=eggy@pdpc/supporter/sustaining/yorkcc.eggy)
14:01:13NJoinrelaxed [0] (n=relaxed@unaffiliated/relaxed)
14:01:13NJoinFOAD [0] (
14:01:13NJoinampleyfly [0] (
14:01:13NJoinRaRe` [0] (
14:01:13NJoinlostnihilist [0] (
14:01:13NJoinbospaadje [0] (
14:01:13NJoindan_a [0] (n=dan_a@
14:01:13NJoinkclaf [0] (
14:01:13NJoinslarti [0] (i=tom@gentoo/developer/slarti)
14:01:13NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
14:01:13NJoinNuclearDog [0] (
14:01:13NJoinOverand [0] (i=overand@pdpc/supporter/active/Overand)
14:01:46 Join bluebrother [0] (
14:01:47***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
14:02:05NJoinjoshin_ [0] (n=joshin@VDSL-130-13-11-10.PHNX.QWEST.NET)
14:02:54XavierGrwhy do I get a filebuffer string error when I startup rockbox on H100? :\
14:03:02XavierGrI don't know what that error means...
14:03:09JdGordonmore than 16 items in viewer.config...
14:03:27JdGordono the combined length of the llines in that file are longer than some buffer.. iirc
14:03:41XavierGrit must be with recent additions of mpegplayer and the rvf I am playing now
14:04:15XavierGrso to fix it I have to increase the line buffer of the viewer.config?
14:04:28 Quit Hadaka (
14:04:49 Join barrywardell [0] (
14:04:53bluebrotherXavierGr: yes. The addition of mpegplayer triggered that message
14:04:55JdGordonanyone with a player got a min?
14:05:19XavierGr:\ now my viewer.config file has 27 lines
14:05:30bluebrothereasy solution: remove a line
14:05:35JdGordonthat would explain it then :D
14:05:45bluebrothergood solution: adjust the buffer size ;-)
14:06:09 Join ze [0] (
14:06:11XavierGr18 different plugins in there
14:06:21XavierGrso yeah it exceeds 16
14:06:26NJoinNibbier [0] (
14:06:34 Quit ze (Client Quit)
14:06:36 Join ze [0] (
14:06:45 Join Bagder [0] (
14:06:46***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
14:06:49zehmm er oh
14:07:08bluebrotherI think it was STRING_BUFFER_SIZE in filetypes.c
14:07:18 Quit ze (Client Quit)
14:07:18roolkuI might do 0-10 then, as I believe the current WPS use that.
14:07:21 Join ze [0] (
14:07:27 Quit ampleyfly (Remote closed the connection)
14:08:17NJoinHadaka [0] (i=naked@
14:08:46std|denisis it enough to add "-DDEBUG" into GCCOPTS to get a debug symbols in uisim build?
14:09:10mentany ideas how to make ipod mini remote connector without buying the original remote controller?
14:09:43bluebrotherment: no, but if you find a way I would be interested ;-)
14:09:56NJoinKohlrabi [0] (
14:09:57JdGordonstd|denis: the uisim has the debug symbols... you can run it through gdb...
14:10:26bluebrotherstd|denis: there is also a (D)ebug option in the advanced configuration
14:10:46bluebrotherbut IIRC this only affects the build for the target
14:11:43XavierGramiconn: do you know if it is possible for the RockVideo Converter to make rvf files with custom width height size? (other than archos resolution)
14:11:46std|denisJdGordon: yeah, it would be great, but i have lack of knowledge about gdb :) so i want to use something more familiar - VC build-in debugger or OllyDbg
14:11:50 Quit smolyn ("KVIrc 3.2.5 Anomalies")
14:12:13bluebrotherstd|denis: you can use ddd as frontend to gdb. Or insight
14:12:26std|denisbluebrother: is it some kinda gui?
14:12:31B4gder... or emacs! ;-)
14:12:39bluebrotherit is.
14:12:40mentbluebrother: I have some ideas how to make it but I think it wont be much reliable
14:12:45NJoinRibs_ [0] (n=ribs@
14:13:11bluebrotheror get the gdb cheat sheet and learn about gdb −− it's worth it :)
14:13:46 Join ampleyfly [0] (n=ampleyfl@
14:13:46 Join atsea-151 [0] (i=ariel@gateway/tor/x-870c96895aa201a8)
14:13:46 Join atsea-142140 [0] (i=ariel@gateway/tor/x-abe95c9645fbb72e)
14:13:46NJoinVeePee [0] (
14:13:46NJoindukeman [0] (
14:13:46NJoinmelter [0] (
14:13:46NJoinnixphoeni [0] (
14:13:46NJoinWeiss [0] (
14:13:50NJoinDataGhost [0] (
14:14:02 Join pearldiver [0] (
14:14:06NJoinsneakums [0] (
14:14:17std|denisbluebrother: just have not enough time to learn it now. i just wanna search a bug with id3 handling
14:14:34std|denisat least i suppose its in id3..
14:14:52NJointoer [0] (
14:15:06JdGordonbarrywardell: any idea how to make the e200 only send SCROLL events once for each click like the OF?
14:15:17 Join idnar [0] (i=mithrand@unaffiliated/idnar)
14:15:46std|denis"ID3 v2.3 specification" and "ID3 v2.3 Programming Guidelines" are dead links?
14:16:22sbehfind /mnt/mp3player | grep PP5022
14:16:27sbehwhere do i get it?
14:18:09std|denisaliask: thanx, i don't need them ;) i just found that they are returns 404
14:18:15linuxstbXavierGr: I'm pretty sure that there's an old attempt to port the rvf player to swcodec on the patch tracker if you're investigating that.
14:18:57NJoinbagawk [0] (n=lee@unaffiliated/bagawk)
14:19:00JdGordonoh bugger... red delta :'(
14:20:09XavierGrlinuxstb: there is? I am going to search for it
14:20:50 Quit atsea-142140 (Remote closed the connection)
14:20:50 Quit atsea-151 (Remote closed the connection)
14:22:32XavierGrcan't find it, searched for rvf, video etc but I couldn't find it :\
14:22:57 Join atsea-142140 [0] (i=ariel@gateway/tor/x-d984f9b19b532c98)
14:23:31 Nick Rick is now known as Gibbed (
14:23:32 Nick Gibbed is now known as Rick (
14:24:07 Quit Moos ("<reboot>")
14:25:04XavierGrI will try to create a custom resolution rvf using the linux command lines
14:25:53 Join Moos [0] (
14:25:56 Join atsea-151 [0] (i=ariel@gateway/tor/x-c753131a1549ef23)
14:26:15sbehhow much of ram does `gmake sansapatcher` need?
14:30:01 Part kavorka^ ("Client exiting")
14:30:20CtcpVersion from freenode-connect!freenode@freenode/bot/connect
14:30:20***Server message 477: 'logbot_ #rockbox :[freenode-info] if you need to send private messages, please register:'
14:33:28barrywardellJdGordon: I think we need to get an interrupt based driver working
14:35:40 Quit Rondom (Read error: 145 (Connection timed out))
14:39:25 Join kavorka^ [0] (
14:41:59barrywardellsbeh: why are you trying to build sansapatcher with that file?
14:44:47barrywardellsbeh: sansapatcher was meant to be built with a PP5022.mi4 containing the Rockbox bootloader
14:46:03 Quit kavorka^ ("Client exiting")
14:46:27sbehbarrywardell: i dont have this file, and this wiki-articel ( sais that i can get it there
14:47:12barrywardellso you already have an old version of the Rockbox bootloader installed?
14:48:06barrywardellthe instructions you linked are only for people who previously had an old bootloader installed
14:48:24barrywardellfollow the link to the Installation chapter for installation instructions
14:48:40sbehbarrywardell: i said that i got the from there by searching through the wiki
14:49:04sbehbarrywardell: step 2.2.3. is where i am
14:49:11sbehbarrywardell: Installing the boot loader
14:49:38linuxstbsbeh: The manual has links to sansapatcher binaries for download.
14:49:55sbehlinuxstb: i don`t have one of these operating systems
14:49:58barrywardellyes, but it doesn't say anything about needing a PP5022.mi4 file. Just use sanspatcher as linuxstb says
14:50:09barrywardellwhat operating system do you have?
14:50:24sbehbarrywardell: in makefile PP5022.mi4 is defined as required
14:50:41linuxstbsbeh: You can get that from
14:50:57linuxstbYou're right that that isn't documented anywhere...
14:51:08sbehthank you :D
14:51:16barrywardellI guess we need some "manual install" instructions
14:51:40linuxstbYes, I was thinking both sansapatcher and ipodpatcher could do with a "compiling from source" section.
14:52:00linuxstb(for trouble-makers like sbeh) :)
14:52:02barrywardellsbeh: out of curiousity, what operating system are you using?
14:52:25*linuxstb should soon have a PPC Linux setup
14:52:30sbehtry this
14:53:52sbehhm, how do i clean my local subversion-checkout?
14:53:54*GodEater_ will probably have one soon too - just as soon as it's installed on his PS3
14:54:31B4gderPS3 has "Cell" CPU(s), doesn't it?
14:54:40sbehit does.
14:54:44 Quit aliask (Remote closed the connection)
14:54:46linuxstbYes, but I thought it was PPC?
14:54:48B4gderso not strictly PPC
14:55:01linuxstbi.e. the PPC Linux distros apparently run fine on it.
14:55:08 Join Webgiest [0] (i=4a87b2ea@gateway/web/cgi-irc/
14:55:08sbehthe first cpu acts like ppc
14:55:19GodEater_all the PS3 distros are PPC based
14:55:45GodEater_I might even avoicd using gentoo on it if you guys will lay off ribbing me about it ;)
14:56:04sbehyou don`t have the full cell-power when using this linux-mode of ps3
14:56:13linuxstbGodEater: So have you chosen a distro yet? I've also just bought a PS3 and (as soon as I buy a USB keyboard) will be installing Linux.
14:56:21B4gder"A set of 128 registers, each 128 bits wide"
14:56:29GodEater_linuxstb, not yet no
14:56:36GodEater_linuxstb, have you ? :
14:56:46sbehlinuxstb: btw this file does not help
14:56:56GodEater_I've also got to raid my parts bin and see if I have a spare usb keyboard
14:57:02B4gderthe ps3 is nicely prepared to run linux though it seems
14:57:10GodEater_B4gder, seems to be yes :)
14:57:16sbehi dont think so
14:57:29linuxstbNo. I normally install Debian everywhere, but am also bored of fighting with Linux installations, so am thinking of finding something more tailored to the PS3.
14:57:31sbehif you run linux on ps3, the hardware gets some limits
14:57:40B4gderof course
14:57:48sbehyou need to hack to get full power of ps3 and run linux on it
14:58:13GodEater_linuxstb, so YellowDog ?
14:58:16B4gderit still is prepared to run linux fine
14:58:26sbehyes, but with lees power
14:58:26B4gdereven if the linux perhaps need improvements to run perfectly
14:58:29GodEater_when compared to say the Xbox260 ;)
14:58:32GodEater_or 360 even
14:58:49sbehB4gder: NO, ps3 hardware is limitated by hardware when you run in linux-mode!
14:58:57B4gder"linux-mode" ?
14:59:07B4gderafaik, it is "otheros"
14:59:19B4gderso that is limited by design?
14:59:34sbehno, it is activ limited by onboard-firmware
14:59:44LinusNthat sucks
14:59:44B4gderthat's what I mean by design
15:00:42linuxstbGodEater_: I expect I'll probably just go with Debian...
15:01:02GodEater_linuxstb, defeatist ;)
15:01:17sbehB4gder & LinusN:
15:01:19linuxstbI can't live without "apt-get"...
15:01:35*B4gder shrugs
15:01:38sbehjust use mplayers
15:01:48bluebrotherit's mpeg4
15:02:01*bluebrother already watched a couple of 23c3 videos
15:02:20sbehlinuxstb: where do you think is the problem at?
15:02:23sbehmy compiler?
15:03:13GodEater_oh is that the Xbox360 vid ?
15:03:22sbehits the 'console hacking 2006'-vid
15:03:37sbehxbox360 is a gameing-console, so its in there too ;)
15:03:54GodEater_I guess I'll just watch it and see
15:03:58*B4gder prefers good old web pages to videos
15:04:09GodEater_me too
15:04:24sbehB4gder: this vid just tells you about ps3 what i did
15:04:27GodEater_I'm also fed up with fighting with mplayer and ffmepg all wekend
15:04:37B4gdersbeh: good, then I don't miss anything ;-)
15:04:38sbehB4gder: otheros-mode is running with limited hardware-ressources
15:04:42GodEater_and apparently learning how to type really badly
15:05:10sbehhm, mplayer works fine here
15:06:25GodEater_sbeh, I was trying to use it to transcode some divx movies to x264
15:06:27***Saving seen data "./dancer.seen"
15:06:43linuxstbsbeh: Can the PS3 handle HD video (e.g. 1920x1080i MPEG-2) in Linux?
15:06:45sbehB4gder: do you realy think that they will allow you with your linux on there cell-processor to outperform there serverhardware? :)
15:06:58B4gderhow to access the SPEs using linux
15:07:02B4gderseems easy enough
15:07:43linuxstbsbeh: I can't help you with your compiling problem - does your gcc install work fine when compiling other things?
15:08:15sbehB4gder: just watch the video ;)
15:08:46sbehlinuxstb: yes, building world a the moment (1h task on a p-m 1.6ghz)
15:08:53 Join perldiver [0] (
15:09:01sbehlinuxstb: > gcc −−version
15:09:01sbehgcc (GCC) 3.4.6 [FreeBSD] 20060305
15:09:08B4gdersbeh: if you can't tell me in a line then I rather find a factual web page ;-)
15:09:29sbehB4gder: my english is realy bad as you see
15:09:59B4gderand that video is the only resource explaining this?
15:10:02 Quit cadu_ (Read error: 104 (Connection reset by peer))
15:10:05B4gderif so, then I start to doubt it
15:10:22sbehB4gder: i surely can't know _all_ sources
15:10:44B4gder... yeah, I asked you for _all_ sources...
15:12:27GodEater_it's not a bad presentation actually Bagder
15:12:44B4gderso what is the limitation then, put shortly?
15:12:52GodEater_I'll let you know when I get to it ;)
15:12:58B4gderI'm at work, I really really don't watch videos here
15:13:20GodEater_ok - I'll give an executive summary assuming I understand it once I've watched it
15:13:33sbehlinuxstb: wtf
15:13:39sbehlinuxstb: i am so stupid!
15:13:58 Quit pearldiver (Read error: 110 (Connection timed out))
15:13:59sbehlinuxstb: did not do a `gmake clean` after fetching your url
15:14:14sbehlinuxstb: so i still had the old bootimg.c ;)
15:14:16 Join jhulst [0] (n=jhulst@
15:14:35sbehlinuxstb: works fine now
15:14:48linuxstbsbeh: Good. Did you see my question about playing HD video on the PS3?
15:16:17sbehlinuxstb: yes i did, but i can't tell you, sorry :/
15:16:31linuxstbOK.. I guess I just need to test myself.
15:17:32amiconnXavierGr: I never used the DirectShow rvf converter if that's what you mean
15:17:42sbeh#elif defined(__FreeBSD__) || defined(__NetBSD__) || defined(__OpenBSD__) \$
15:17:43sbeh || defined(__bsdi__) || defined(__DragonFly__)$
15:17:43sbeh sprintf(sansa->diskname,"/dev/da%d",i);$
15:19:26sbehbtw the bootimg.c was 39M of code, harhar :D
15:20:08amiconnXavierGr: For proper rvf support on H1x0, M5 and greyscale ipods we have to teach the converter tool(s) how to create custom resolutions, framerates, and pixel packings in case of the ipods
15:20:11linuxstbsbeh: Are you sure you have the correct PP5022.mi4? It should only be about 365KB
15:20:31sbehlinuxstb: it was
15:20:39linuxstbAnd bootimg.c is 39MB?
15:20:58amiconnCustom framerate is already possible with the cmdline tools, but I think there should be a "target" option that selects sensible parameter defaults for all these
15:21:00sbehlinuxstb: was with the old PP5022.mi4 from
15:21:22sbehlinuxstb: this was the reason why cc1 did consume >500m of ram ^^
15:21:38linuxstbAh, OK...
15:22:07sbehwhat is UMS-mode?
15:22:20sbehi only have MSC and MTP
15:22:31linuxstbUMS = USB Mass Storage = MSC
15:22:48XavierGramiconn: I think I can create custom resolution if I use the saperate tools, I am testing them right now
15:23:12GodEater_it seems the limitation with the PS3 is really that it's processor is geared for single precision vector work - and isn't very good for "standard" applications
15:23:20GodEater_it's not really that it's locked down
15:23:29GodEater_it's just not very good at "other stuff"
15:23:34 Join pondlife [0] (
15:23:36GodEater_you get about half the power apparently
15:23:52XavierGramiconn: if I manage to make this thing work with sound I will try to see if I can hack RockVideo too (if there is source code for it)
15:23:55GodEater_when trying to do integer work / double precision work
15:24:17B4gderGodEater_: that's not really a "firmware limitation" as I see it though, but I guess we can drop that now ;-)
15:24:20amiconnXavierGr: RockVideo isn't open source, unfortunately
15:24:27amiconnIt's [IDC]Dragon's work
15:24:29GodEater_B4gder, nor as I consider it either
15:26:11 Nick jborn__ is now known as JoeBorn (
15:26:21XavierGramiconn: ah too bad! maybe a script could make things easier
15:26:54amiconnAsk him when he's around
15:34:36sbehyeah, thats overkill
15:34:44sbehyou`re realy leet person :D
15:35:52 Join Arathis [0] (
15:37:24sbehhm, laggy
15:39:58 Quit std|denis ("bye")
15:41:11 Part LinusN
15:44:09GodEater_linuxstb, looks like you will be able to play back your hidef video
15:48:36sbehhm, in show audio track can i see that the buffer is fully filled, but sound still laggs, is sounds to me that an value becomes to how an then the 'current frame' gets dropped
15:48:53sbehframe like 100msec music
15:49:30sbehhm, in show audio track i can see that the buffer is fully filled, but sound still laggs, it sounds to me that an value becomes to high an then the 'current frame' gets dropped
15:49:58B4gderyeps, there are bugs left in the sansa audio playback
15:50:55sbehthis drops only occur on loud parts in the current song :)
15:52:07JdGordonthere is a patch which seems to fix the low audio problem on the sansa... but the ui is still impossible to use if audio is playing
15:52:19 Part kaaloo
15:52:59barrywardellthe laggyness is apparently due to the lcd driver constantly flushing the cache
15:53:35B4gderbtw, we need a USB stack mentor volunteer
15:53:35barrywardelland I think the drops are the headphone over-current protection kicking in
15:53:44 Join mattzz [0] (i=c19fe442@gateway/web/cgi-irc/
15:54:22 Quit Thundercloud (Remote closed the connection)
15:54:32sbehbarrywardell: hm
15:54:44sbehbarrywardell: so its sounds like to me
15:55:08sbehbarrywardell: patch to get from?
15:55:14 Quit amiconn (" zzzap")
15:55:34 Join vcardenas [0] (i=c8767629@gateway/web/cgi-irc/
15:55:36JdGordonbarrywardell: I have 6911 and 6908 and cpu frequecny scalling enabled and audio works much better, but yeah, its impossible to use the ui
15:55:43*linuxstb wonders why MPEG encoders would write PTS values that don't match the expected duration of the raw streams...
15:56:08*GodEater_ laughs at the console hacking video "Debian didn't work so I had to use Fedora..."
15:56:16 Join lee-qid [0] (
15:56:34B4gderpoor bastard
15:56:38sbehGodEater_: maybe he wantet debian not to work?
15:56:47GodEater_no - he said he'd have installed it if it did
15:56:52sbehGodEater_: this ccc-guys are crazy sometimes oO
15:57:31barrywardellsbeh: that link has the patch
15:57:38sbehbarrywardell: thanks
15:58:01GodEater_hmm - you don't get full low level access - the linux OS sits on top of a hypervisor
15:58:21sbehGodEater_: as i said ;)
15:58:26GodEater_you don't get complete access to the disk, and no access to the 3d acceleration
15:58:40barrywardellJdGordon: 6908 seems to be a workaround for the cache flushing I was talking about
15:58:44GodEater_well that's not quite what you said
15:58:44sbehand now u can think of what reasons, hehe
15:59:03B4gderI honestly don't think there's a conspiracy behind it
15:59:04sbehGodEater_: oh, its some time away since i watched the vid :(
15:59:21GodEater_B4gder, nor me - it makes perfect sense to me (if I were Sony)
15:59:26B4gderafter all, they made this box to easily run a 3rd party OS
15:59:53GodEater_B4gder, I'll live with it - it does enough that I don't care. It's a MUCH nicer solution than on the Xbox360
15:59:56sbehno, it was the only way to go to prevent to many people from hacking it
16:00:17B4gderwhy does it prevent anyone?
16:00:37B4gderI mean, why can't it be hacked as well?
16:00:48markunB4gder: it can, but you don't gain so much
16:00:59markunso nobody is putting any energy in it
16:01:07sbehit can be hacked, but there are not so much people out who want to
16:01:08markunnot such a bad move from Sony
16:01:12sbehbecause it can already run linux
16:01:17GodEater_it's pretty intelligent really
16:01:17sbehso where is the fun to hack?
16:01:18barrywardellJdGordon: so 6908 only improves audio playback, not the UI?
16:01:25B4gderbut still, if that really was "downgrading" it so much as sbeh claims then I think people will
16:01:26markunsbeh: so basically what I was saying :)
16:01:43sbehmarkun: sry
16:02:13*B4gder gets grumpy when forced in a corner with no mmu
16:02:24JdGordonbarrywardell: I have both compiled.. so I cant actually say how much each affects it... the volume one works great tho
16:02:39mattzzbarrywardell: 6908 improves audio and and fixed the white-screen-bug
16:04:07GodEater_the chances of the PS3 getting hacked further I think are probably higher than with the XBox360 - since they allow you to run your own stuff on the PS3. The exploit in the XBox360's hypervisor would have been much harder to introduce since it would have a required an exploit in someone *else's* code.
16:04:46Webgiestmattzz: but did you notice the screen artifacts , little horizonal lines that last about a second when navigating through the menu and directories?
16:05:10mattzzWebgiest: I noticed artifacts in scrolling text
16:05:19Webgiestthat too
16:05:33Webgieststill very nice!
16:05:34sbehGodEater_: i heard of an update the burns some pieces on the xbox360 to prevent it from downgrading oO
16:05:49XavierGrtoo bad that rvfmux will make any resolution video to 112*64 (not to mention that it seems that rvf without sound can't play properly)
16:05:50 Quit billytwowilly (Remote closed the connection)
16:05:54sbehGodEater_: they fight with heavy weapons :D
16:06:14barrywardellthe thing about 6911 is that it disables the over-current protection. I think that is a bad thing
16:06:24 Join billytwowilly [0] (
16:07:17barrywardellare there LCD artifacts with 6908? that's what the cache flush is supposed to avoid
16:07:45GodEater_sbeh: that's kind of my point :)
16:08:05 Quit B4gder ("Time to say moo")
16:08:16JdGordonI have lcd artefacts.. dunno if its 6908 or any of my hacks though.. ill check
16:08:18mattzzbarrywardell: I only observed some artifacts in horizontal scrolling text.
16:08:56 Join crop [0] (i=c27f0812@gateway/web/cgi-irc/
16:09:13XavierGrat least rvfmux has source and seems quite easy to add custom resolution support
16:09:21 Quit SirFunk (Connection timed out)
16:09:27cropJdGordon: any news / new ideas about fixing
16:09:47 Join Hoffmann_ [0] (
16:09:56 Join SirFunk [0] (
16:09:56JdGordoncrop: im trying various things... havnt got it nice enough to commit yet though
16:11:53bluebrotherJdGordon: I looked into root_menu.c this weekend ... restructuring the code and removing the adgument in the struct saves a couple of bytes
16:12:29barrywardellJdGordon: I don't notice any extra US sluggishness when playing music
16:12:41bluebrotherbtw, how does fast switching using button shortcuts work? The appropriate function returns a GO_TO_something?
16:12:48JdGordonreally? with cpu scaling disabled?
16:13:01JdGordonpretty fast...
16:13:04barrywardellwith SVN
16:13:38*JdGordon just rabbed brand new svn
16:13:46cropJdGordon: I also have an idea. I think we should clean up the code in tree.c, menu.c, and root_emnu.c a little bit. As a first step, I's suggest to pull out the definitions of GO_TO_xxx constants to a separate header file (e.g. screen_navigation.h) since they are used everywhere. And then write comment to all browsing funcs (what they return). And then things will become clearer. I tried to to this but my laptop cr
16:14:47JdGordonbluebrother: there is no ui code between a return from a function and loading the next... so it should happen very quickly... misread
16:14:59pondlifeJdGordon: You're up late today?
16:15:07JdGordonyeah.. just return GO_TO_SOMETHING
16:15:17JdGordonpondlife: DST wierdness.. we moved 2 hours closer :p
16:15:18croppondlife: tomorrow :-)
16:15:38JdGordonits only 12.15am here...
16:15:47pondlifeJdGordon: So are you in one of those states that DOES do DST..?
16:16:03JdGordonyea :)
16:16:08*pondlife doesn't want to start any DST debates again
16:17:29 Join midgey [0] (
16:17:42JdGordonbarrywardell: you have no ui slowness with audio playing? the scroll wheel is heaps more jerky than usual here...
16:17:47JdGordon...with svn
16:18:02 Join perl|work [0] (
16:18:06JdGordonpossibly only while it buffers though...
16:18:11barrywardelli've just done a svn up, I'll let you know in a second...
16:18:20bluebrotherbbl ...
16:20:20 Quit crop ("CGI:IRC")
16:20:20XavierGrI wonder how one compiles rvf_mux.cpp, on windows it says it is missing header files on linux I am getting an error of "lookup" with no type
16:20:45 Part mattzz
16:20:53 Quit Hoffmann (Read error: 110 (Connection timed out))
16:21:30Webgiestwith a fresh svn and 6908 there are no artifacts while scrolling but the same choppy navigation while play music over here
16:21:41Webgiestsound good though
16:22:12Webgiestno more vinyl!
16:22:31JdGordonhmm... wtf? im getting nothing but static!
16:22:36barrywardellhmmm. now I get the choppy navigation
16:22:42barrywardellwith SVN
16:23:03Webgiestwell the choopiness has smoothed out some now as it plays longer
16:23:42barrywardellis it only while the file is being loaded from disk?
16:24:03JdGordoni tihnk so.. its too hard to navgate to the audio debug screen to see :p
16:24:26JdGordon... and i forgot that there is a virtual led which would do just as well :p
16:25:09JdGordonis the cpu actually at 11mhz? or is it only saying it is?
16:25:18barrywardellit's only saying it is
16:25:45JdGordonso is it at 30 or 74?
16:25:48barrywardelli don't know what it's actually at
16:25:55JdGordonok :)
16:25:57barrywardellprobably 75MHz
16:27:12 Quit Hoffmann_ (Connection timed out)
16:27:18barrywardelltest_disk gives reasonable results
16:28:06Webgieston a side note elephants dream is out of sync again
16:28:55SlasheriJdGordon: argh, you just edited that debug_menu :P
16:29:15Slasherinow i have hard time to convert it with the updated scheduler :)
16:29:35 Join Hoffmann_ [0] (
16:30:06Slasherinever mind :)
16:30:18JdGordonyou fiddled with the thread debug screen?
16:30:36Slasheria little bit, because scheduler structure changed a bit
16:30:53Slasherifor example there is no longer thread pool per core but a single pool of all threads
16:31:34JdGordon:'( bugger... that caused a big PITA for the screen :p
16:31:41JdGordonyou want me to revert the commit?
16:31:54Slasherinope, i will just figure out how to fix it :)
16:32:15JdGordonok :)
16:34:19SlasheriJdGordon: hmm, just wondering what that thread_ids is supposed to do
16:34:26Slasheri(why it's needed)
16:34:43Slasherisome kind of sorting?
16:34:56JdGordonthats used to put all threads in a single array, which sounds like it isnt needed anymore...
16:35:08Slasheriah, then i will remove that :)
16:35:11Slasheriit's not needed
16:37:35SlasheriJdGordon: hehe, you should add an ability to kill threads to that screen =)
16:37:55JdGordonI thought about an option to change priority... not to kill them though
16:38:03JdGordonthat sounds like a bad thing....
16:38:09JdGordonsome idiot might try to kill main :)
16:38:11SlasheriJdGordon: hmm, priority change and killing.. that would be nice ;)
16:38:21Slasheriwell, it's debug screen anyway :P
16:38:43perl|work"elephants dream is out of sync again"?
16:40:25JdGordonbarrywardell: well, the scroll wheel sluggishnees isnt completly because of the disk access... I made the wheel light act as a hdd light and its only slightly worse when the disk is on.. and only for a short time...
16:40:52JdGordonuntill we get proper buttons should I try giving the buttons their own thread maybe?
16:41:45barrywardellJdGordon: Ideally, we need to figure out how to enable GPIO interrupts
16:41:45JdGordonanyone here get seizures from flickering lights? I have a patc for you :p
16:42:33Webgiestperl|work: with 6908 it is haven't tried a clean build though
16:42:34 Quit VeePee ("leaving")
16:42:51JdGordonof course... but untill then?
16:42:57sbehJdGordon: ahah, nice, i need that patch :) (wheel as hdd-led)
16:43:51barrywardellJdGordon: no harm in trying to see if it works for now
16:44:31sbehJdGordon: or just tell me where i`ve to look at
16:44:39sbehJdGordon: thanks :D
16:44:49 Join pabs [0] (
16:44:52 Quit Slickire ("CGI:IRC")
16:45:08perl|workWebgiest ah
16:45:37 Join Nico_P [0] (
16:46:05 Join toffe82 [0] (
16:47:20Webgiestdo the ipods not have an up and down button? just scroll forward/back?
16:48:26JdGordontick tasks are done in the main thread right?
16:48:50SlasheriJdGordon: no, in ISR
16:49:15Slasheriso those are guaranteed to happen in time
16:50:01JdGordonso putting the button checking in a thread would be pointless then?
16:50:21Slasherihmm, yes. what do you had in mind?
16:50:53 Quit JoeBorn ("rebootus maximus")
16:51:02markunlinuxstb: I found a way to 'fix' my mencoder frame skipping problem: -mc 0
16:51:22JdGordontrying to make the sansa more responsive... looks like we cant do anything about it untill interupts are figured out...
16:51:30barrywardellJdGordon: do you think maybe it's just the lcd not updating, rather than the buttons being unresponsive?
16:51:35markunIt just disables the a/v sync correction (which is incorrect)
16:52:05JdGordonbarrywardell: no idea... any idea how to test it?
16:52:08linuxstbmarkun: I knew a/v sync would cause all the problems....
16:52:31markunlinuxstb: me too, but I still don't understand why it thinks the a/v is out of sync
16:53:03markunmkv with 24fps xvid + 48000Hz Ogg Vorbis
16:53:06 Join My_Sic [0] (
16:53:32markunlinuxstb: anyway, now it works! :)
16:53:36barrywardellJdGordon: the wheel led comes on straight away, but scrolling is sluggish. So the buttons are being read frequently, just the UI isn't updating
16:53:42 Quit Nico_P (Remote closed the connection)
16:53:43JdGordonbarrywardell: although, if it was the not updating, shouldnt it just jump to the selection, instead of actually scrolling down eventually?
16:54:08markunlinuxstb: didn't you at one time say that the libmpeg2 generated m2v used much less CPU than the one generated with ffmpeg?
16:54:23linuxstbmpeg2enc does that, not libmpeg2 (libmpeg2 is just a decoder)
16:54:24 Join Nico_P [0] (
16:54:29barrywardellJdGordon: yeah, I was wondering that too. not sure what's up there???
16:54:41markunlinuxstb: ah, so is the difference big?
16:54:54 Join JoeBorn [0] (
16:54:59 Join webguest35 [0] (i=55b36c3e@gateway/web/cgi-irc/
16:55:07linuxstbI forget the details, but I think it was significant. i.e. 2 or 3 fps more.
16:55:14barrywardellJdGordon: possibly the cpu is overloaded with playback, disk access and lcd updates?
16:55:46JdGordonI just changes the PRIORITY_BUFFERING prio to 8 and it is _much_ better now
16:56:33barrywardellok, so I guess it is a cpu problem then?
16:56:47JdGordonlooks like it
16:57:10JdGordonit has no problems keeping up with it at that prio...
16:57:54 Quit jhulst (Read error: 110 (Connection timed out))
16:59:05webguest35i always hear a low scratching if i'm playing songs on my sansa
16:59:21webguest35sometimes also if i don't hear a song
17:01:02acathlawebguest35, yep, same problem
17:01:45webguest35i presume this is due to the early developement status?
17:01:48acathlaa reboot and everything is fine
17:01:59webguest35i'll try it
17:02:00acathlai don't know
17:02:30webguest35unfortunately it persists :-(
17:06:28***Saving seen data "./dancer.seen"
17:08:20 Quit bonbonthejon (Read error: 131 (Connection reset by peer))
17:08:34 Quit JdGordon ("zzzzZZZZZzzz....")
17:10:53 Join jmsbwtr_ [0] (
17:11:19Webgiestwhen playing music and running the oscilloscope plugin when pressing |<< or >>| to change the speed there is no lag in the response time, is it doing something special?
17:12:24pondlifeSlasheri: Quick question..? I don't have a device at the moment, so can't check. Does the CPU thread/stacks display include any boost info?
17:12:33pondlifeOr did I dream it?
17:17:09Slasheripondlife: doesn't seem to
17:17:21Slasheribut that should be easy to add
17:17:41pondlifeGo on, you know it makes sense ;)
17:17:48Slasherihehe :)
17:18:07pondlifeOr maybe I can make that screen work in the sim, and do it myself.
17:18:19pondlifeI guess it ought to be available in the sim really.
17:18:33Slasheriplease don't touch that screen until i commit the scheduler stuff (soon)
17:18:39pondlifeNope, I won't.
17:19:13*pondlife touches the threads screen, then hides
17:20:38pondlifeI like the "Test code!!!" comment at the top of dbg_os, by the way...
17:21:52sbehCFLAGS:= ${CFLAGS:C/-m(arch|cpu)=[a-zA-Z0-9]+//}
17:22:04sbehgives me something like that: -Os -pipe -m
17:22:07sbehwhats wrong in here?
17:22:40 Quit webguest35 ("CGI:IRC (EOF)")
17:22:50preglow_jhMikeS_: there?
17:22:57preglow_jhMikeS_: ah, you just fixed it :P
17:23:37_jhMikeS_ preglow: yeah
17:23:44 Nick _jhMikeS_ is now known as jhMikeS (
17:25:52jhMikeSnot sure what's up with that since the settings are the same for me for .S and .c,.h, etc.
17:28:27preglowdon't worry about it, do it all the time myself :P
17:33:18 Join web-taz [0] (
17:34:13jhMikeSI assumed I had the problem solved 'till I scanned the logs and saw your message :\
17:39:33*jhMikeS sees that he's going to have to go on a global huntdown of repeated use of the same array index with array notation, changing: a[n], a[n], a[n] to t *p = &a[n], p->, p->, p-> ... a cheap way to save thousands of bytes and make code faster
17:39:57 Join Insectoid [0] (n=boinkboi@
17:40:37 Join Thundercloud [0] (
17:44:04SlasherijhMikeS: hmm, why didn't you use add_to_list() to put the "wake up tasks" to end of running list?
17:44:08Slasherithat puts them to last
17:44:34jhMikeSSlasheri: Cause I just wanted to do it one op
17:44:58Slasherihmm, one op?
17:45:03 Join TheSphinX^ [0] (
17:45:13 Quit ryan_r (Read error: 110 (Connection timed out))
17:45:23jhMikeSTake the entire list...put it at the end. Not loop and add them one by one
17:45:37jhMikeSor maybe I missed something...possibly
17:46:07Slasheriah yes, you want to put entire list at once there
17:47:16InsectoidI want to do a bit of vanity code for fun. What file is one of the first executed so that I could stick in like an ascending piezo scale when my Ipod comes on?
17:48:22jhMikeSI actually wanted to detect running in IRQ context...perhaps this is easy with other CPUs but preserving that flag in set_irq_level on coldfire would just bloat the code to no end. you can see how much space was saved by just removing two bytes from that function.
17:48:29Webgieston what is the shift function under File browser and menus? i've added the e200 but im not sure what the shift is.
17:48:33 Join krz [0] (i=nastyboi@
17:48:43markunInsectoid: apps/main.c ?
17:49:16InsectoidMarkun That would make sense wouldn't it?
17:49:50 Quit Ribs_ (Remote closed the connection)
17:50:06SlasherijhMikeS: i am going to commit the scheduler cop changes soon (without playback on cop yet)
17:51:19 Join Ribs [0] (n=ribs@
17:51:28 Quit Guile (Read error: 104 (Connection reset by peer))
17:53:49 Join Guile [0] (n=Guile@
17:54:55jhMikeSgood deal...I've been trying to stay out your way as much as possible there. I do want to implement some of the newer features more cleanly but am hold out for that.
17:56:36 Part Llorean
17:57:47preglowjhMikeS: if it wasn't for gcc, the index stuff would gain even more
17:57:55 Part jjore
17:58:10preglowjhMikeS: i have cases where gcc doesn't code *pointer++ as (aX)+
17:58:17 Quit Arathis ("Bye, bye")
18:00:31jhMikeSpreglow: it is just awful isn't it? especially on cf where array access makes for more extension words and those add up hugely.
18:01:41 Join Arathis [0] (
18:01:55jhMikeSI wonder if there's some compiler flag that should be added?
18:02:36 Quit Everybody|away ("( :: NoNameScript 4.02 :: )")
18:04:35 Join Everybody [0] (
18:05:15 Quit printfXh4 (Read error: 145 (Connection timed out))
18:05:47sbehhm, i think i need to switch to linux to build it
18:05:58 Join printfXh4 [0] (
18:06:08sbehtoo buggy for other oses
18:06:19jhMikeSpreglow: It would probably optimize away common subexpressions that also have side effects though if so...hehe
18:06:32sbehbtw line 433: OUTPUT_ARCH(1 )
18:10:19 Quit krz (Read error: 110 (Connection timed out))
18:10:53 Quit Nico_P (Remote closed the connection)
18:11:51 Join Thundercloud_ [0] (
18:20:28barrywardellyou need gcc 4.0.3 to compile it. I think you have 2.95.3?
18:21:32sbehbarrywardell: line 25
18:21:54barrywardellyeah, that's where I got it from :)
18:22:03sbeh> gcc -v |& tail -1
18:22:04sbehgcc version 3.4.6 [FreeBSD] 20060305
18:23:18markunlinuxstb: I could use crossfeed for mpegplayer :)
18:24:50sbehbarrywardell: its tools/configure's decision, not mine :P
18:25:16 Join Llorean [0] (
18:26:35 Quit Thundercloud (Read error: 110 (Connection timed out))
18:26:48barrywardelltry arm-elf-gcc -v
18:26:49 Quit dilinger (Read error: 110 (Connection timed out))
18:27:15sbeh> arm-elf-gcc -v |& tail -1
18:27:15sbehgcc version 2.95.3 20010315 (release)
18:27:39barrywardelldid you compile that yourself?
18:27:54sbehportinall devel/arm-elf-gcc
18:27:56sbehdid that for me :P
18:28:07barrywardellyou could use in the tools directory to compile a correct version
18:30:39sbehshould i make a self-build-guide for other freebsd users out there?
18:31:17 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
18:31:27Lloreansbeh: What guide were you following before?
18:31:42sbehLlorean: docs/README
18:31:55perl|workyay rating support :P
18:32:45Lloreansbeh: Ah, that's an exceedingly undependable file, oddly enough
18:35:26Lloreansbeh: That's basically a two year old file with some tweaks to "modernize" it, but Rockbox requires specific compiler versions not mentioned in it (but mentioned in most other locations regarding compiling)
18:37:34*jhMikeS going to do a quick experiment using priority_yield in the dsp loop
18:38:14 Join My_Sic [0] (
18:38:35SlasherijhMikeS: btw, i will commit the changes in a few minutes :)
18:39:51Slasherijust doing final testing now
18:41:54 Quit Insectoid (Read error: 104 (Connection reset by peer))
18:44:20jhMikeSI'm quite sure I'll have to run a mixer thread in that manner where load and timing must be rather constant
18:46:31 Join krz [0] (
18:54:51 Join inversions [0] (
18:55:38SlasherijhMikeS: committed, hopefully it didn't break anything :)
18:58:54jhMikeSwe'll find out shortly ;)
19:01:33sbehbarrywardell: seems to work fine
19:01:37 Join jhulst [0] (n=jhulst@
19:01:46sbehdoes somebody commit my update to :)
19:04:35Lloreansbeh: How is changing which tool it uses really an update?
19:05:23LloreanAlso, there is a proper process for patches. :-P
19:05:24sbehLlorean: because of freebsd-users only have fetch from base
19:06:31***Saving seen data "./dancer.seen"
19:06:44jhMikeSguess it shouldn't suprise me that priority_yield has no effect on already hiprio thread since it's really meant as a way for low priority threads to be done with resources in a timely manner if they yield while holding those.
19:06:53Lloreansbeh: That would've been more useful knowledge than you simply posting a link to a code snipped without comment or explanation. Anyway, post it to flyspray, I imagine someone will be interested, but I personally don't think I should be tinkering with the scripts.
19:08:56 Quit Guile ("KVIrc 3.2.0 'Realia'")
19:09:35jhMikeSSlasheri: RED!
19:09:58 Part Llorean
19:10:10 Join PaulJam [0] (
19:11:53PaulJamlittle question: is there still a way to show the database autoscore in the WPS now that the %rr tag shows the manual rating?
19:13:02 Quit atsea-142140 (Read error: 104 (Connection reset by peer))
19:13:02 Quit atsea-151 (Read error: 104 (Connection reset by peer))
19:13:40barrywardellsbeh: I don't know enough about that script to be comfortable with committing changes to it.
19:14:05barrywardellbut great that you got it working in the end
19:15:24 Join jac0b [0] (
19:16:26 Join dilinger [0] (
19:17:56jhMikeSSlasheri: So far so good on H120
19:19:12 Quit krz (Read error: 110 (Connection timed out))
19:19:38sbehsystem.c:813: error: 'boostctrl_mtx' undeclared
19:19:54 Join webguest66 [0] (i=5933599e@gateway/web/cgi-irc/
19:20:13sbehforgotten to add a file?
19:20:31Slasheriifdef problem, fixing soon
19:21:04*jhMikeS is happy to see green on the single core targets
19:24:31*pondlife wishes that the sim thread control was a bit more similar to the targets
19:25:44 Quit webguest66 ("CGI:IRC (EOF)")
19:26:46jhMikeSpondlife: I think it could be on windows anyway since it has fibers. does linux?
19:26:56jhMikeSor we don't run windows now...:P
19:27:03pondlifeHmm, depends on SDL I guess
19:27:16preglowpondlife: in what way, really?
19:27:37jhMikeSAt least with fibers we could run the same scheduler
19:27:48bluebrotherroolku: shouldn't the new rating tag be only added if HAVE_TAGCACHE is defined?
19:28:14pondlifepreglow: The thread handling on the sim doesn't allow tracking in the same way (I tried enabling the debug screen). Also it's not a very good simulation of the real thing.
19:28:28pondlifeThat latter part being my main point.
19:29:14pondlifeAnd AFAIK SDL does not support fibers
19:31:25jhMikeSreally, couldn't we just run a single os thread...and to simulate cores, two os threads?
19:32:19pondlifeI would think so.
19:32:56pondlifeIt might not be worth the effort, but if it makes the sim a truer simulation and might actually simplify the code a bit....
19:33:07 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
19:36:05jac0bon the sansa am I supposed to leave the rockbox.e200 in the .rockbox folder?
19:36:27jhMikeSSo long as the host permits a stack switch for the thread
19:38:04 Quit Ribs ("Ex-Chat")
19:38:31barrywardelljac0b: yes
19:40:54jac0bokay cause I took it out but it functions correctly
19:41:13barrywardellit will work either in .rockbox or in the root. but it defaults to .rockbox
19:41:53 Join Ribs [0] (n=ribs@
19:44:01jac0bOH ALRIGHT THANKS
19:44:09jac0bopps sorry about the caps
19:45:05 Join petur [0] (n=petur@rockbox/developer/petur)
19:46:33roolkubluebrother: not sure what you mean by tag? - the menu is only active if runtime gathering is on which in turn depends on tagcache
19:48:19 Join nls [0] (
19:49:09roolkubluebrother: or do you mean the declaration in id3.h? I suppose this could be ifdefed, but it concerns a few other as well
19:49:13toffe82rbutils doesn't update or delete the rockbox.gigabeat which is on the root if it exist there, so you have 2 differents files (one in the root and one in rockbox)
19:50:41roolkuPaulJam: not at the moment, if there is a need for it, I could make an %rs (or similar) tag
19:52:01 Quit ender` (" I know you think you understand what you thought I said, but I am not sure that what you heard is not what I meant.")
19:52:40MoosSlasheri: still red
19:52:46PaulJamroolku: thanks for the info.
19:52:47 Join Nico_P [0] (
19:53:39SlasheriMoos: i know, hadn't time to fix it until now
19:53:42scorcheroolku: well, it is more an issue of that ifdef encircling anything to do with tagcache...why have the code in there and waste space, when tagcache is not being built?
19:53:52 Quit PaulJam (".")
19:53:53roolkubluebrother: or do you mean wps tag? - yes, I suppose I should make all runtime stuff dependent on tagcache
19:53:59MoosSlasheri: no worries :)
19:54:18roolkuscorche: I agree - just trying to work out what bluebrother meant
19:55:04roolkuscorche: as I did nothing different to any of the other 3 runtime values
19:55:06 Join ender` [0] (n=ender@
19:57:10barrywardelltoffe82: but it looks inside .rockbox first anyway, so that's not a problem is it?
19:58:31toffe82I don't know, but for a couple of days, when I switch on my gigabeat I have a message saying loading and I just find out this, I will try later if it is a problem
19:59:45barrywardellI think it should be ok, but no harm in checking just to be sure
20:04:50 Join bob_brain [0] (
20:05:03 Join elinenbe_work [0] (i=d1c4c007@gateway/web/cgi-irc/
20:05:19 Quit jhulst (Read error: 110 (Connection timed out))
20:05:55elinenbe_workSlasheri: is the COP as fast as the main processor on the ipod?
20:06:08Slasherielinenbe_work: yes, it should be
20:06:47elinenbe_workSlasheri: that's mostly awesome... considering rockbox runs pretty damn well on the single processor, it should rock on both.
20:06:55Slasheriroolku: btw, i really would like to have that autoscore fixed to not react on stop button (ie. only when playlist really ends)
20:07:10Slasheribecause it's starting to ruin my scores :)
20:07:43elinenbe_workSlasheri: can that help at all in running video or something like rockboy? Also, does the ipod linux project utilize the cop?
20:07:51Slasherielinenbe_work: indeed, it was very speedy when i played audio on cop
20:07:56 Join jhulst [0] (n=jhulst@
20:08:02Slasherielinenbe_work: video already runs on cop
20:08:24Slasheriand i have no idea about ipod linux, probably no
20:08:54roolkuSlasheri: do you have a suggestion how? As I really would like to record that I played the track, even if I interrupted it
20:08:56elinenbe_workSlasheri: then how come the video is so slow here:
20:09:11elinenbe_workDo we know anything about the video processor?
20:09:31roolkuSlasheri: only solution I see is to have additional run time data
20:09:34elinenbe_workSlasheri: also, does that affect battery life at all?
20:09:42 Quit pondlife ("disconnected has pondlife")
20:09:49linuxstbelinenbe_work: I assume you're talking about the ipod video?
20:10:02elinenbe_workwell, ipod v5 and v5.5
20:10:14linuxstbThen read the last sentence of the first paragraph on that page...
20:10:23Slasheriroolku: hmm, i wouldn't if considering the following: you have just finished playing your "last track" and stop playback and shutdown player after the next track starts to listen to later again
20:11:02Slasherinow it also considers that latest track to be played and decreases the overall score for sure
20:11:27Slasherielinenbe_work: atm, shouldn't affect
20:11:39elinenbe_worklinuxstb: thanks! We needs to find out some info on them chips!
20:11:46 Join donski [0] (i=5ac0e8c6@gateway/web/cgi-irc/
20:11:49roolkuSlasheri: yes, but this is only one scenario - what about if you don't resume the track
20:12:00bob_brainHow would I encode video files for rockbox on linux?
20:12:15roolkuSlasheri: ideally we need to detect the resume and not increment playcount if resuming
20:12:25Slasheriroolku: then i don't want it to be played either
20:12:57Slasherii only consider track as "played" when it truly has gone to its end, either by normally ending or skipping
20:13:05Slasheristopping isn't a normal track ending
20:13:25elinenbe_workwell, the progress is great. Good work guys!
20:13:37roolkuSlasheri: I am fairly certain this is not true. I can skip and it will still increase playcount
20:14:12Slasheriroolku: and it should.. skipping is "normal ending"
20:14:19roolkuSlasheri: I had to do this routinely to 'fix' the missed tracks from my 'not yet listened' tracks
20:14:39Slasheriand if you skipped it, you probably didn't like it atm so we can decrease score too
20:14:49roolkuSlasheri: what is the difference between skipping and stopping to select a new track?
20:14:50Slasheribut if you stop a track, there might be other reason for it
20:15:17roolkuSlasheri: well, there might
20:15:18 Join krz [0] (
20:15:27roolkuSlasheri: not neccessarily is
20:16:42perl|workthis page got pretty heavy and getting hard to scroll through even
20:17:01 Quit bob_brain (Remote closed the connection)
20:17:35 Join thebigearl [0] (
20:17:35roolkuSlasheri: I can see the problem that resumed tracks are recorded twice, but this has to be solved on the resume phase as it is not known if a track will be resumed
20:17:44 Quit donski ("CGI:IRC")
20:17:47 Quit inversions ()
20:18:57 Quit jac0b ("Chatzilla 0.9.77 [Firefox]")
20:19:37Slasheriroolku: probably a better autoscore system needs to be implemented
20:19:54 Quit jhulst (Remote closed the connection)
20:20:33Slasheribut i have no idea yet how that should be done
20:21:08elinenbe_workSlasheri: what game are you talking about?
20:21:27Slasherielinenbe_work: game?
20:21:41 Join donski [0] (i=5ac0e8c6@gateway/web/cgi-irc/
20:21:49Slasherielinenbe_work: the song statistics
20:23:02 Join jhulst [0] (n=jhulst@
20:23:46sbehProblems in last 30mins: Normal Shutdown does not work, Sound still laggy (drops msec when song becomes fast/loud), FreeBSD disables the usb-port because the bootloader does non-compliant usb-communication, LCD slowles (~5-8min) becomes complete white due to white horizontal lines
20:24:33 Join Domonoky [0] (
20:25:59 Join atsea-142140 [0] (i=ariel@gateway/tor/x-350c3a19c25d4dbe)
20:26:09sbehdevd does now auto-update my sansa :D
20:27:01bluebrotherroolku: I meant it like scorche said −− the added stuff in onplay.c jumped in my eyes first. If you create a build without tagcache it doesn't make sense to have that present
20:27:11 Quit web-taz ("Nettalk6 -")
20:28:22bluebrotheralso, the tag itself shouldn't be mentioned inline in the manual but only in the wps tags table in the appendix
20:29:08barrywardellsbeh: they're all known problems, except for shutdown - that should work
20:29:26 Join inversions [0] (
20:29:37roolkubluebrother: I'll encapsulate all the runtime stuff in have_tagcache later tonight (a little busy right now)
20:29:54bluebrotherbtw, will there be a warning if a user tries to set a rating when the database isn't initialized?
20:30:06barrywardellsbeh: what exactly happens when you try to shutdown?
20:30:14 Join atsea-151 [0] (i=ariel@gateway/tor/x-a74072f2b9485b51)
20:30:17roolkubluebrother: feel free to fix, I didn't want to meddle with the manual - just trying to help
20:30:35donskiHi, can I please get wikiaccess? my wiki name is DonalFarrell
20:30:39sbehbarrywardell: it hangs on 'Shutting down ...' one moment please, secounds ago it worked
20:30:43bluebrotherroolku: I really appreciate you added something to the manual. Will change that.
20:30:44roolkubluebrother: I am not sure if the macros are okay either, must be the first time I have done tex :)
20:31:16bluebrotherthe diff looks fine
20:31:26barrywardellsbeh: It shuts down for me here...
20:33:09 Join harlequin [0] (i=444e19c8@gateway/web/cgi-irc/
20:34:28sbehbarrywardell: ok, it does definitely fail when 'Commiting database [x/y]' message occured and you want it to shutdown
20:34:56 Join matsl [0] (
20:35:22 Part juxtap
20:35:33Slasherisbeh: you can't shutdown when the db is committing
20:35:40sbehSlasheri: after...
20:35:48Slasheriah, probably a new bug then
20:35:57bluebrotherdonski: done. Should work now
20:36:23donskibluebrother: Thanks !
20:36:27harlequinI have a question.....sooo I downloaded rockbox for my toshiba f10 gigabeat and sofar it has been working I'm in the process of downloading a new build for album art when I just found out you could watch videos on the do you do this? Like how could I get a video encoded and played on the gigabeat?
20:37:39perl|workalbum art is not a supported feature at the moment so you'd need a custom build with the latest svn updates and album art patch
20:37:43perl|workto enjoy both
20:38:49bluebrotherharlequin: if you encounter any problems with an unofficial build please make sure to check with an official build first −− unofficial builds are completely unsupported.
20:39:08harlequinooh ok
20:39:23harlequinwill they be added as official builds later on
20:39:28bluebrotherother than that, there are a couple of unsupported builds in the forums. You might find one for you there
20:39:35bluebrothernot soon.
20:40:04bluebrotheran official implementation of Album Art will at least require Metadata on Buffer, which itself won't come soon
20:40:15bluebrotherso don't hold your breath for it
20:40:55 Join anathema [0] (
20:41:17bluebrotherstill, it is likely to happen. Just don't ask when, we don't estimate times ;-)
20:42:35harlequin:) ok
20:42:42harlequinthanks for the help
20:44:32sbehit's time for kgdb over usb, isn't it? :P
20:45:20 Quit krz (Read error: 110 (Connection timed out))
20:45:30 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:46:47donskibluebrother: Hi, sorry it didnt seem to work. I am trying to put my name on the Summer of Code page, but it gives me access denied
20:49:59 Join Redbreva [0] (
20:51:10*linuxstb pings a non-present Llorean
20:51:25 Quit Moos ("Glory to Rockbox")
20:51:25BigBambidonski: I think SoC page needs more rights than 'standard' people have
20:51:35linuxstbNo it doesn't.
20:51:38barrywardellsbeh: If we could get a working usb driver first...
20:52:00BigBambiIt was discussed in here a few days ago to make it like that, and Bagder said good idea
20:52:14BigBambiSo I assumed it happened given, donski couldn't edit it
20:52:20BigBambiBut I am probably wrong
20:52:22donskiBigBambi: ah rite, ok
20:52:29sbehbarrywardell: :)
20:52:43BigBambidonski: but I'm probably wrong
20:52:56BigBambidonski: can you edit any other page?
20:53:08donskiBigBambi: no , everything gives me an access denied
20:53:19BigBambiOK, I was wrong and something else is up :)
20:53:31linuxstbdonski: Are you registered in the wiki?
20:53:38donskiI have just registered tho, so maybe the system just needs a refresh
20:54:13linuxstbWhat's your wiki name?
20:54:42BigBambidonski: try now
20:54:50BigBambiI just fixed it (I think)
20:55:19donskiWayy! thanks, all good now!
20:56:48 Quit inversions (Read error: 110 (Connection timed out))
20:57:26 Quit BigBambi ("Leaving")
21:00:03linuxstbAnyone want to test what may or may not be an avsync improvement to mpegplayer?
21:02:29 Quit donski ("CGI:IRC (EOF)")
21:02:33TheSphinX^linuxstb one moment =)
21:02:38 Join robin0800 [0] (
21:05:26TheSphinX^linuxstb tzzz tzz unused variables, what options should be prefeared to test?
21:05:43 Quit vcardenas ("CGI:IRC (EOF)")
21:05:46linuxstbThe usual Skip Frames and Limit FPS
21:06:34***Saving seen data "./dancer.seen"
21:07:57 Quit Thundercloud_ (Remote closed the connection)
21:08:17linuxstbTheSphinX^: Do you have any files which played out of sync before? And are they files your device can decode in realtime?
21:09:56TheSphinX^i tested your last patch, and now im testing the same movies and a new, I'll try every config combination
21:10:41 Join bigearl [0] (
21:10:45 Join _Veseliq_ [0] (
21:10:49LearYikes, svnversion is really slow on Cygwin. :/
21:11:07bluebrotherd'oh! Seems I did a typo when editing TWikiUsersGroup. Oh my.
21:11:17linuxstbTheSphinX^: Don't bother disabling those options - I'll probably remove them now anyway.
21:11:40TheSphinX^linuxstb: for example it's the first time my currently running movie is sync while limited fps and no skipping
21:11:40bluebrotherLear: create a file docs/VERSION and put any version string you want to get displayed there
21:11:52TheSphinX^and it's decoding in realtime ^^
21:12:18LearYes, I know, just creating an "update" script that does that (after doing an svn update).
21:12:51bluebrotherthat's a nice idea −− haven't thought of that myself ;-)
21:13:01bluebrotherbut on native linux svnversion isn't too slow
21:13:11 Quit secleinteer ()
21:14:02 Nick Everybody is now known as Everybody|away (
21:14:08LearJust tested Linux myself. Seems like a filesystem issue mainly, so maybe it would be better if I didn't build on FAT32...
21:16:07LearAh, much better without that delay of 10-70 seconds. :)
21:16:47 Quit ment ("eof")
21:17:09 Join secleinteer [0] (
21:18:56linuxstbTheSphinX^: So that patch is working well for you?
21:20:00 Join z0de [0] (
21:21:49TheSphinX^linuxstb after looking 2 movies witl 4 differrent settings per movie, yes very well
21:22:13TheSphinX^so i looked every movi 4 times =)
21:22:19 Join web-taz [0] (
21:22:29 Join marc [0] (
21:24:32perl|workso im guessing the optimal fps would be 24 or 25
21:25:24perl|work29.97 looks too problematic
21:25:47linuxstb29.97 should work for widescreen videos on the gigabeat (i.e. 320x176)
21:26:45 Quit thebigearl (Read error: 110 (Connection timed out))
21:27:20Webgiestlinuxstb: even with all the patches in my build it syncs up just fine at 23.9 fps
21:28:44linuxstbWebgiest: That's with my new patch?
21:28:49 Quit jhulst (Remote closed the connection)
21:29:15Webgiestyes and many more today im afraid and crashes when ever I exit O_o
21:29:41 Quit lee-qid ("aufwiederbyebientotsayonara")
21:30:38 Quit web-taz ("Nettalk6 -")
21:31:06 Quit marc| (Read error: 110 (Connection timed out))
21:31:46Webgiestlinuxstb:data abort at 00041AC0 but I also applied the no vinyl patch so it could be that too I'll give it go with a fresh check out
21:32:15Webgiesteverytime i hit the quit button
21:33:03 Join jhulst [0] (n=jhulst@
21:36:38 Quit Ribs (Remote closed the connection)
21:39:43 Join Ribs [0] (n=ribs@
21:39:56 Join amiconn [0] (
21:40:21TheSphinX^hmm on my x5 no error and i've many patches, but the most affecting other plugins.
21:40:21TheSphinX^I wish my exams were done, than I will have more time for the gif-viewer and doom.
21:41:48 Nick Everybody|away is now known as Everybody (
21:42:16WebgiestTheSphinX: animated GIFs?
21:43:09TheSphinX^currently i'm trying to get the old outdated gifviewer to wor, and than i will tray to animate them
21:43:32TheSphinX^step by step
21:47:12scorcheBagder: there?
21:48:29amiconnlinuxstb: 64 bit issues in DEBUGF() call...
21:48:31 Join mattzz [0] (
21:48:38 Quit _Veseliq_ (Remote closed the connection)
21:49:52Bagderscorche: yeps
21:49:54 Join mirak [0] (
21:53:44*Bagder detects gsoc application #14
21:54:23preglowBagder: when do projects get handed out?
21:54:32Bagderapril 11
21:54:42bluebrotherwho reads Llorean?
21:54:47Bagderits deadline for applications today
21:54:59Bagderbluebrother: llorean, christi, me, linus and zagor
21:55:21preglowBagder: the llvm one is weird
21:55:31elinenbe_workBagder: let's get viewports and metadata for ID3 image tags working ;-)
21:55:42jhulstare we going to have to wait until April 11 to know what was accepted?
21:55:43elinenbe_workBagder: have you gotten any/many applications yet?
21:55:52Bagderelinenbe_work: 14 so far
21:56:01 Quit himitsu (Read error: 111 (Connection refused))
21:56:07preglowBagder: i don't the (ownership requested) stuff. what's needed for that to appear?
21:56:12elinenbe_workBagder: That's awesome! More than I thought... that's great!
21:56:18preglowBagder: don't GET
21:56:30Bagderpreglow: they need a volunteering mentor
21:56:31linuxstbpreglow: I think that means someone has offered to mentor it.
21:57:26 Quit zylche (Connection timed out)
21:58:03 Join vcardenas [0] (i=c8767629@gateway/web/cgi-irc/
21:58:52jhMikeSpreglow: any chance we can just use 24bit audio and leave me a working byte? I plan on running two codecs and one voice channel throught the same eq without attenuation.
21:59:09 Join krz [0] (
21:59:17*amiconn wonders how much cpu power would be needed for this softmodem idea
22:00:07bluebrotherBagder: did someone contact you regarding this german podcast? I was wondering if I should write them as it looks like a couple of them barely speak english
22:00:39BagderI don't think anyone has ever mailed that press email
22:01:15bluebrothermaybe I'll just tell them about the address in case they haven't found it yet
22:01:36 Join andy^rb [0] (
22:01:42jhMikeSpreglow: I'm probaby gonna work up a prelimiary low latency system without voice and crossfade at first just to evaluate any looming problems
22:03:12 Join juxtap [0] (
22:06:55 Quit harlequin ("CGI:IRC (EOF)")
22:10:36 Join web-taz [0] (
22:12:56 Quit matsl (Remote closed the connection)
22:16:03 Join lokkju_wrk [0] (n=lokkju@unaffiliated/lokkju)
22:19:28 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
22:20:45 Join Domonoky_ [0] (
22:21:19preglowjhMikeS: you mean 24 bit out of the codecs, or?
22:21:59lokkju_wrkare there any current forums/channels for disscussing modification to the actual apple ipod firmware? I know there is basic stuff, like iPodWizard, but what about more advanced editing, such as menu choiuces, and arbitrary command codes?
22:22:00 Join Insectoid [0] (n=boinkboi@
22:22:51 Quit barrywardell ("Konversation terminated!")
22:23:24InsectoidIs anyone having trouble with Ipod 5g and latest svn? After scrolling a bit through the menus it just locks up −− the menu is still displayed but there's no speech and no movement. It could be some of the patching I've tried doing, I'm doing an svn revert now but I'm checking to see if anyone else's having the problem first.
22:23:29 Quit linuxstb (Remote closed the connection)
22:23:45 Join linuxstb [0] (i=5343d4aa@rockbox/developer/linuxstb)
22:24:07jhMikeSpreglow: yes
22:24:12preglowjhMikeS: why?
22:24:43jhMikeSI want no worries about clipping or limiting until the last pass
22:24:50preglowjhMikeS: oh, you mean mix pre eq?
22:24:53preglowhmm, well
22:25:43jhMikeSthen we have one eq/tone control set with one history for all affected channels
22:25:50preglowsounds tempting, but that means we'll have to use high-precision shifting for emac if we want to have true 24 bit output
22:26:08 Quit Hoffmann_ (Connection timed out)
22:26:19preglowbut the gains are very tempting indeed
22:26:23preglowno more worrying about shifter overflow
22:26:29preglowand one history for everything
22:26:37 Join Hoffmann_ [0] (
22:26:49preglowi guess you should just go ahead
22:27:05jhMikeSthat must not happen as it is...I suppose I'll calc the exact bit limit...perhaps 25 or 26 is ok
22:27:09preglowsome codecs will need modifying, the only one i know works at 24 bits currently is tremor
22:27:45jhMikeSHow much modification to use interleaved stereo as the standard?
22:28:12preglowdepends on the codec
22:28:32preglowwhy'd you want that, btw? avoid having two pointers?
22:29:08jhMikeSno, the way data will be packeted at the start, it's easier to link a channel out of memory blocks if the audio is interleaved
22:29:17 Quit Domonoky (Read error: 60 (Operation timed out))
22:29:42jhMikeSbut I think I can do a trick and just have half the block be left and half right
22:31:32 Join kraizee [0] (
22:33:11preglowi'm starting to think using interleaved audio is better anyway
22:33:19preglowbut it doesn't exactly make burst reading/writing any easier
22:33:56 Join lnxmomo [0] (
22:34:02jhMikeSwhy not? you'd burst LRLR LRLR instead of LLLL RRRR
22:34:24lnxmomoi am in rockbox atm in my ipod and i want to get back into the apple firmware, how do i do that
22:34:38lnxmomoi have tried menu+select then rewind
22:34:41jhMikeSanother problem is then you need different routines for mono and stereo...mega bloat and I'm not sure it would gain anything
22:35:33nlslnxmomo: shut it down by holding the play button the press the play button to start it again and quickly toggle hold
22:35:42jhMikeSeh, just forget all that interleaved stuff :P
22:36:20lnxmomothanx nls
22:37:08preglowjhMikeS: can do! :)
22:37:11jhMikeSAbout the biggest thing to do is have all dsp functions take a single parm so they can be plugged in and a list of calls blindly iterated through. I know you lose some arm advantage there.
22:38:43jhMikeSmaybe not though...most of the data you need in these calls seem to be in memory anyway and not directly in the regparms
22:39:43jhMikeSno,'s my beastie and I'll take care of that horribly boring aspect :)
22:39:45 Quit andy^rb ()
22:43:09 Join solexx [0] (
22:44:29 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
22:45:52solexxis there a known breakage in current Iriver H120 builds?
22:46:17solexxI updated my weeks-old rockbox this morning and have been greeted with "codec failure" since then
22:46:45bluebrothersolexx: you didn't update correctly
22:46:50linuxstbrockbox.iriver has moved into the .rockbox folder - you probably have an old bootloader that's only looking in the root directory.
22:47:00linuxstb(or what bluebrother said...)
22:47:14bluebrotheryou need to replace the *complete* build, not only rockbox.<whatever>
22:47:24solexxargh, thanks. This is probably a FAQ now... Sorry.
22:47:32*nls is surprised that so many h1xx users have bootloaders that are sooo old
22:47:44 Join dj-fu [0] (n=ajc@unaffiliated/dj-fu)
22:47:51bluebrotherwhat bootloader for h100 didn't look into .rockbox?
22:48:03TheSphinX^the newest
22:48:03nlsbluebrother: v5 and older i think
22:48:06bluebrotherI started with v5 like 2 years ago and that already did.
22:48:18bluebrotherIIRC at least ;-)
22:48:41bluebrotherwe should ship the FAQ as text file with the builds ...
22:49:05nlsIriverBoot says it was added to v6
22:49:45linuxstbsvn says 19 Nov 2005
22:49:49solexxWhoah! Removing /rockbox.iriver did it.
22:50:20linuxstbAh, maybe one version looked in the root first, then .rockbox...
22:50:25solexxI'll need some time to get accustomed to the new menu.
22:50:32 Join Shaid` [0] (
22:50:33nlswiki says 10 Jan 2006 for v6
22:50:43 Quit krz (Read error: 110 (Connection timed out))
22:50:57bluebrotherthen my memory is wrong :)
22:51:03linuxstbAh yes, the 19 Nov 2005 commit just switched the order...
22:53:30 Join Thundercloud [0] (
22:54:06 Quit Insectoid (Read error: 110 (Connection timed out))
22:55:18 Quit Nico_P (Remote closed the connection)
22:56:04 Join bob_brain [0] (
22:58:05 Quit anathema (Nick collision from services.)
22:58:21 Join inversions [0] (
22:58:41 Join zylche [0] (
23:00:34 Quit lnxmomo (Read error: 110 (Connection timed out))
23:01:32 Quit TheSphinX^ ("Miranda IM! Smaller, Faster, Easier.")
23:01:43 Join Guile [0] (n=Guile@
23:03:53 Join Soul-Slayer [0] (n=jonno@
23:04:06 Join Insectoid [0] (n=boinkboi@
23:04:23InsectoidWhen running make in my build directory, what could make it not generate sysfont.h?
23:05:00Soul-SlayerHey everyone, lil problem. On my H10 5GB, using the latest SVN and a newly built bootloader, my H10 keeps freezing up after a few seconds. I've tried with older builds and there's no problem, it's just the latest one.
23:05:44bob_brainSo you have a new bootloader and build?
23:05:58bluebrotherthat's the problem
23:06:02bluebrotherupdate the bootloader
23:06:14Soul-SlayerI just did. Just compiled it a moment ago.
23:06:18InsectoidDoes the freeze happen after you do something? Or can you just let it sit at the start screen and it freezes?
23:06:18bob_brainhe has a new one lol
23:06:36***Saving seen data "./dancer.seen"
23:06:36bob_brainwait do you have a mtp or ums ver?
23:07:16Soul-SlayerIt's after trying to do anything that accesses the disk it seems... It freezes up when I try and navigate through the Files menu, and it doesn't even begin to initialise a database, just freezes on 'Found 0 files'
23:07:18bluebrotherwow, the delta-table is quite red :(
23:07:21Soul-SlayerI have a UMS.
23:08:14 Quit Shaid (Read error: 110 (Connection timed out))
23:08:15 Nick Shaid` is now known as Shaid (
23:08:22nlsSoul-Slayer: and you are sure that it works with an older build, not just that it used to work?
23:08:22bob_brainI don't know if this could be the problem but the new bootloaders look for the rockbox.h10 inside the .rockbox folder
23:08:41bob_braindelete the one from the root
23:08:41Soul-SlayerJust to give you some idea, I went (From root folder)...
23:08:41Soul-SlayerMedia -> Music -> (artist name)... And it refuses to navigate any further than this, my H10 has become completely unresponsive.
23:08:54bluebrotherthe current builds have rockbox.h10 in the .rockbox folder
23:09:01Soul-SlayerYes, I've deleted the rockbox.h10 in the root dir, and I've tried two other older builds and they work fine.
23:09:16InsectoidThis smells a lot like what's happening with my Ipod.
23:09:36bob_braingive me a second I will give it a try on my H10 though it is mtp version
23:09:40Soul-SlayerAt first I thought there was something up with my hard disk (When I listen to it, I can hear it endlessly spinning but no tell tale clicking that it is actually doing something)
23:09:40nlsSoul-Slayer: how old were the older builds
23:09:49Soul-SlayerBut both the OF and all older build work as expected.
23:09:52bluebrotherwhich bootloader did you install? The one from
23:09:54InsectoidSoul-Slayer: Do you have music queued up that you can resume? Does it let you play? Or does it freeze before it gets to that point.
23:10:07Soul-SlayerI can't get as far as playback.
23:10:14bluebrotherthat still sounds pretty much like a COP issue with the bootloader
23:10:16Soul-SlayerAnd the bootloader I compiled from the latest SVN source.
23:10:16 Quit My_Sic (Read error: 104 (Connection reset by peer))
23:10:20InsectoidNot even resuming playback from the main menu?
23:10:44Soul-SlayerI'll copy in playlist control file and see, hold.
23:11:48InsectoidMy silly makefile isn't generating sysfont.h. And things are breaking when everything looks for it.
23:12:38InsectoidIs there any way to force it to/generate it manually?
23:12:41Soul-SlayerIt definitely seems to be a disk issue... Now it's freezing up on the 'Scanning disk...' part, and I can hear the hard disk spinning but not clicking at all.
23:12:43nlsSoul-Slayer: why not install an official bootloader binary, just to rule out your own bootloader being the problem?
23:12:46linuxstbInsectoid: sysfont.h is (I think) the first file generated, so it's normally a symptom of something wrong in your build environment.
23:12:53Soul-SlayerOkay, will do.
23:13:08BagderInsectoid: run 'make V=1' to get more info
23:13:45InsectoidLinuxstb: It compiled fine about ten minutes ago, but then I rechecked svn trunk out, made a new build directory, ran ../tools/configure, and make now doesn't make the sysfont.h
23:14:28Bagderand gsoc #15...
23:14:38linuxstbInsectoid: Did you do a "make clean"?
23:14:49Insectoiddoing so now.
23:15:10InsectoidDidn't think to do that before because it was a totally fresh directory.
23:15:21Soul-SlayerUsing an official bootloader now. Still won't get beyond 'Scanning disk...'
23:15:25bob_brainI can confirm that a bootloader from the current svn seems to not load rockbox on my H10
23:15:29Soul-SlayerI heard the disk clicking a couple of times, now it's doing nothing at all
23:15:36 Join My_Sic [0] (
23:15:42Soul-SlayerWell, it's spinning, just no clicking
23:15:45nlsSoul-Slayer: and an older build works?
23:15:56InsectoidSame. I get codepages is up to date, then
23:15:58Insectoidmake -C /root/rockbox/firmware OBJDIR=/root/rockbox/build/firmware
23:16:04Soul-SlayerI'll check the revision of the older build I have
23:16:10InsectoidSo it's like it skipped the sysfont.h
23:16:17nlsSoul-Slayer: I guess you have to wait for barrywardell then :-/
23:16:32linuxstbInsectoid: Are you logged in as the root user?
23:18:08 Join barrywardell [0] (
23:18:14bob_brainA build that I compiled yesurday works fine on my H10
23:18:18Soul-SlayerRevision 12887 is the latest old build I have, and it works perfectly. That's ~50 commits ago
23:18:23bob_brainonly current svn doesn't
23:18:26 Join bepe86 [0] (n=bepe@
23:18:45bob_brainsomething between yesturday and today
23:19:32 Join bonbonthejon [0] (n=jon@
23:19:36Soul-SlayerCould it be r12926? 'Make scheduler functions thread safe core wise'
23:20:06Arathisbob_brain: do you compile the bootloader every day? oO
23:20:17bob_brainnot everyday
23:20:21bob_brainbut some
23:20:48bonbonthejonits built with the firmware
23:21:01nlsSoul-Slayer: my guess would be either that or 12915...
23:21:27bob_brainhow do you get an svn from a certian revision?
23:21:46Soul-SlayerI'll get a source from the 25th and compile it and see if it works
23:21:58bob_brainI know it works from the 25th
23:22:04bob_brainI just tried one
23:22:12Soul-SlayerWhich player do you have?
23:22:18bob_brainwould have been afternoon
23:22:21bob_brainH10 5gb
23:22:29bob_brainsame problems as you
23:22:32Soul-SlayerAre you getting the same problem I am with the latest?
23:22:34Arathisbob_brain: i just use the svn builds of rockbox, but i don't install a new bootloader ever time
23:22:49bob_brainArathis: you don't need to
23:23:00bluebrotherbob_brain: svn co -r <revision>
23:23:05bob_brainArathis: only when there are updates to it
23:23:12Arathisbob_brain: that's why i asked ;)
23:23:16bob_brainbluebrother: thx
23:23:26bob_brainArathis: asked what?
23:24:01Arathisbob_brain: if you compile a new bootloader every time you udpdate rockbox. it appeared to me that you thought you'd have to
23:24:11Soul-SlayerBob, try the revision 12915, then we can work out whether it was that or 12926 doing it
23:24:22bob_brainArathis: ah no no
23:24:36Soul-SlayerI'd imagine that people would have noticed earlier had it have been the later though, and wouldn't have continued trying to improve?
23:24:36barrywardellproblems with the H10 builds?
23:24:45bob_brainthe man is here
23:25:07Soul-SlayerBelieved to be revision 12915 or 12926, but we are getting lockups when trying to access the disk in general
23:25:13bob_brainThe current build getts stuck at boot
23:25:23bob_brainwe have rulled out the bootloader
23:25:25barrywardellhave you tried older builds?
23:25:34Soul-Slayer25th of March works fine.
23:25:39bob_brainI tried one from yesturday works fine
23:25:40Soul-SlayerAnd before that.
23:25:52bob_brainyesturday afternoon
23:25:56barrywardelli'm just rebuilding now...
23:26:20bob_brainapears to be somthing with the disk spin because it gets stuck at "scanning"
23:26:54Soul-SlayerAnd when I get beyond there, I can only navigate around 3 layers of folders before it freezes up.
23:27:06Soul-SlayerAnd initialising the database freezes on 0 files.
23:27:29 Quit petur ("sssssssssss---------PLOP!")
23:28:16 Join entheh [0] (
23:28:43bob_brainWhy can't I use 007quick as a nick?
23:28:50 Part perl|work
23:29:08barrywardellcan you work backwards with 'svn update -r <revision>' and wory out which revision commit the problem?
23:29:46barrywardelli'll do the same here
23:30:37Soul-SlayerJust compiling 12915
23:30:56bob_brainI'm building the last one
23:31:36 Nick bob_brain is now known as H10_007quick (
23:31:54H10_007quickMan doesn't like my nickname
23:32:28barrywardell12921 here
23:33:00 Quit mattzz ("Leaving")
23:33:41scorcheBagder: application was short and sweet, because i figure that you already know me and what my proposal is, at least by the summary
23:34:14 Join funky [0] (n=repulse@unaffiliated/funky)
23:34:21Bagderwe got it
23:34:56Soul-SlayerAlright, just about to boot into 12915
23:35:21H10_007quickbarywardell: did you see my new script to compile?
23:35:45Soul-Slayer12915 is working fine.
23:35:53H10_007quickbarrywardell: alows one to pick up to 9 builds and it compiles them
23:36:21Soul-SlayerIn which case my next suspicion is 12926... I'll try it
23:36:29barrywardellI haven't seen it. I'll check it out
23:36:45H10_007quickbarrywardell: at the top
23:37:08InsectoidOn a brand new, just checked-out 12932, it doesn't generate sysfont.h. This makes me kind of sad.
23:37:27Soul-SlayerHave you tried earlier builds?
23:37:51Soul-SlayerOk, can't even compile 12926... 12928 then ^^.
23:37:51H10_007quickbarrywardell: Oh and disregard the message at the top if you haven't already looked at it, I will delete that soon
23:38:06linuxstbInsectoid: Can you type "make V=1 &> logfile" and upload that logfile somewhere?
23:38:14bluebrotherH10_007quick: wow, quite lenthy and complicated ... you can enter the player name instead of the number to configure
23:38:28H10_007quickcould make it that way
23:39:06H10_007quickbluebrother: My first script though and one would have to type out the name exactly so wouldn't be that great
23:40:09H10_007quickbluebrother: better to stick to numbers I think
23:40:19bluebrotherI usually use the names.
23:40:29bluebrotherand you can always map them ...
23:40:59H10_007quickbluebrother: what do you mean by "map" them?
23:42:03bluebrotherI use this for preparing my build dirs:
23:42:21scorcheBagder: also, my existing mentor app will just sit there i suppose...i had to register under a new account, as they couldnt do anything with my existing mentor app, and because i apped to be a mentor, i couldnt app as a student
23:42:58H10_007quickbarrywardell: 12931 does not work trying ...30
23:43:00bluebrotherthe way you do it now you can still get the user enter a number and then case $number in 0) name=foo; esac
23:43:15 Join webguest40 [0] (i=8e34510c@gateway/web/cgi-irc/
23:43:45Soul-Slayer007, stop
23:43:51Soul-Slayer12928 doesn't work, as I expected
23:43:59barrywardell12921 doesn't work
23:44:16H10_007quickbluebrother: I don't see anything different from what I did
23:44:21Soul-SlayerSo it's > than 12915 and >12921
23:44:26bluebrotherit's shorter :)
23:44:46Soul-SlayerOr possibly 12921 itself
23:45:00bluebrotheryou repeat the same stuff for each possible build
23:46:05Soul-SlayerAnyone wanna take a guess as to which revision could be making the problem 12915> and <12921 :P?
23:46:46Soul-SlayerBecause they all look fairly innocent...
23:46:46H10_007quickshould I try 12920?
23:47:28InsectoidLinuxstb: It's on as make log
23:47:40Soul-SlayerCan't hurt... Barry, any ideas?
23:48:14barrywardelli'm guessing 12916...
23:48:29H10_007quickyou going to do it?
23:48:34Soul-SlayerI'll test that if you like.
23:49:12barrywardelli'm compiling it now...
23:49:17Soul-SlayerAh, okay.
23:49:32 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
23:49:37Soul-SlayerWell, bearing in mind 12915 was fine, if it goes wrong at 12916 then we have our culprit.
23:49:58Soul-SlayerBut it seems strangely unlinked to the H10?
23:50:18 Quit ender` (" Sometimes I lie awake at night, and I ask, "Where have I gone wrong?" Then a voice says to me, "This is going to take more t)
23:50:20Lloreanlinuxstb: Longer than 4 hours at 320x240 25fps video at 600kbps/128kbps. I don't know how much longer because I got distracted, but 4 hours at least.
23:50:25 Quit atsea-151 (Remote closed the connection)
23:50:25 Quit atsea-142140 (Remote closed the connection)
23:52:07barrywardell12916 doesn't work either
23:52:20Soul-SlayerSo it was something there which broke it then..
23:52:38Soul-SlayerBut I honestly can't see how any of the commits link up with the H10?
23:53:21linuxstbInsectoid; I don't understand why it isn't working for you. You can create sysfont manually with the command "./tools/convbdf -h -o sysfont.h ../fonts/rockbox_default.bdf"
23:53:27 Join atsea-142140 [0] (i=ariel@gateway/tor/x-4d6d984c11b8f1aa)
23:53:40linuxstb^that should be ../tools/convbdf
23:54:01linuxstbLlorean: And how long was your 320x176 test?
23:54:05InsectoidThank you. Hey... I could have a larger font by default then?
23:54:59linuxstbTry it and let us know... But there are probably lots of places where the size is assumed.
23:55:24Lloreanlinuxstb: 4:50
23:55:38 Nick kraizee is now known as krz (
23:56:05linuxstbLlorean: OK... Sounds about right.
23:56:29linuxstbBTW, your test video wasn't the best to try and debug a/v sync with - there's very little talking...
23:56:32Lloreanlinuxstb: I think it was closer to 4:10 or 4:15 for the 320x240, I just wasn't watching it closely enough to be sure.
23:57:08 Quit robin0800 (" HydraIRC -> <- Go on, try it!")
23:57:24linuxstbHave you had chance to test the latest a/v sync improvements?
23:57:29Lloreanlinuxstb: Yeah, it wasn't originally intended to be a test video, but the loss of sync was pretty apparent to me in the last bit of dialogue between Smart and 99 at the end.
23:57:33LloreanI'm just about to test them
23:58:16linuxstbThe opening sequence (a voice on a tape-recorder) isn't ideal either...
23:58:50LloreanIt was originally just encoded because I wanted to see how suitable TV was for the Nano's screen

Previous day | Next day