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

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

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

#rockbox log for 2007-11-03

00:00:01jhMikeSit seemed associated with setting it up. I rather arbitraryily named it "CACHE_PRIORITY" with no clue as to the purpose.
00:00:12 Part WalkGood
00:00:26jhMikeSIt's set differently for COP/CPU
00:02:01 Quit Domonoky_ (Read error: 104 (Connection reset by peer))
00:03:06 Quit hcs ("Leaving.")
00:04:29Bagdermediaproc.mi4 kind of implies...
00:04:35Bagdera separate processor
00:04:36resky_nanok: i don't know i cannot push my wma files from amarok to my ipod.
00:05:04resky_can i just copy from my shell?
00:05:52zivan56ok, got a topic up in the forum for the View. Now if only Sandisk is generous enough to give the project a player...
00:06:49Bagderand I posted this =>
00:06:52 Quit midkay (Nick collision from services.)
00:06:53 Nick midkay_ is now known as midkay (
00:07:58resky_perhaps i should ask the other way around: what is the best method to push my files from a linux box on my ipod which has rockbox?
00:08:59*amiconn thinks the best method is a plain file copy
00:09:01oblibresky_, it's just a USB hard drive. Copy it
00:09:14amiconnJust 'cp' or a file manager of your choice
00:09:18Bagderzivan56: it being Rhapsody, we can at least expect R nastiness
00:09:25resky_oblib: the original ipod software had problems with this
00:09:44nanokresky_: just copy them, like any normal files
00:09:48zivan56Bagder: I see. But wasn't the only problem on the rhapsody not knowing the key?
00:09:57resky_oblib: must i put them somewhere in the file structure?
00:09:58oblibresky_, yeah, that's one of the things that rockbox does better
00:10:16 Quit davina ("xchat on Ubuntu 7.04")
00:10:23Bagderzivan56: (oops) no, it was more knowing where/how to upgrade the original bootloader
00:10:33nanokresky_: so the question was not how to listen to them on rockbox, but how to get them to the ipod ;)
00:10:36oblibresky_, nope, your choice. If you want to have access to the files by genre, artist, title, whatever, read about using the database
00:10:38Bagdersince that one does signature verfication correctly
00:10:42 Join amin [0] (
00:10:46zivan56ah, I see. Didn't follow the development of that port much
00:10:49nanokresky_: put them wherever you want your music to be
00:10:56resky_neat stuff
00:11:10oblibresky_, my ipod has a /Music directory in root, and I copy it all into that directory
00:11:13Bagderzivan56: our "fake" DSA sig approach doesn't work on the R models, we have to binary patch the original BL and put it on the target
00:11:35Bagderzivan56: and the R can't upgrade its bootloader by itself, we have to upload a vanilla bootloader to it to do the upgrade!
00:11:44resky_can i just delete the ipodcontroll directory and such?
00:11:46zivan56Sounds messy. Hopefully Mr.H is interested in helping out :)
00:11:56Bagdermessy it is
00:12:10zivan56at least here we have a rom file and it appears to upgrade like the vanilla e200
00:12:21nanoktalking about sansa r?
00:12:23 Quit toffe82 (Read error: 110 (Connection timed out))
00:12:24zivan56then again, I don't actually have a View to verify anything
00:12:25oblibresky_, I usually leave it, because when I hook up the ipod to the computer, I let it boot back into Apple's firmware because it copies a lot faster
00:12:30*nanok is thinking r must stand for rubish
00:12:31Bagderzivan56: very true, the .rom is included...
00:12:36Bagdernanok: sansa view
00:12:44oblibresky_, if you blow away that directory, the Nano will reload the Apple directory and take longer
00:12:50amini hav rockbox installed but i cant bigger the fonts ... are there any updates with the themes available and graphics or a website that can make my rockbox cooler ?
00:12:50nanokaahm, they extended the rubish to the new ones :(
00:13:06Bagdersince we know it, it isn't so rubbish
00:13:11nanokamin: check, extras
00:13:51zivan56would the view be the fastest CPU supporting rockbox if it were to be ported? Seems quite powerful
00:13:59zivan56err player with fastest CPU
00:14:20Bagderzivan56: my guess is that the fast "CPU" is for video/display in the same way ipod videos do it
00:14:50Bagderthus possibly not easily accessible/usable as a stand-alone cpu
00:15:21Bagderbut this is just my speculation of course, I don't know
00:15:22nanok..thus close to worthless..
00:15:32nanoki would still like to see a sigmatel ported
00:15:36amiconn250MHz wouldn't be the fastest even if not taking the gigabeat s into account
00:15:41nanokthough i guess that;s almost offtopic here :)
00:16:18zivan56hmm, well according to the nvidia page, it uses this core:
00:16:58zivan56although it mentions different speeds, I guess just instruction compatible with it
00:19:13bertriklostlogic: still here? I just watched it happen. Playback stopped as usefl ran down close to 0.
00:19:48lostlogicbertrik: got the logf with buffering?
00:21:29resky_my ipod just went dead
00:21:48resky_i copied my stuff and restarted my ipod and now i cannot turn it back on
00:22:23resky_and its not just the battery. even my computer cannot revive it.
00:23:05oblibWhat kind of ipod?
00:23:17resky_ipod nano first generation
00:23:19bertriklostlogic: I'm getting the feeling it is easily reproducible, just takes a long time for the usefl buffer to completely drain
00:23:46oblibresky_, You push select and it does nothing?
00:24:05oblibresky_, did you install the bootloader?
00:24:19resky_yeah it worked, i got even new styles
00:24:38linuxstbHold MENU+SELECT for about 20 seconds
00:26:01 Quit petur ("*plop*")
00:26:56 Quit cooz ()
00:27:30 Nick Soap_ is now known as Soap (n=Soap@rockbox/staff/soap)
00:27:39 Join toffe82 [0] (
00:28:42 Part toffe82
00:29:48Ebertwhy can't i install any of these Safari beta's? "volume does not meet requirements"
00:30:34Ebertsorry wrong channel
00:37:23 Quit ompaul (Client Quit)
00:37:38bertriklostlogic: ah, easily reproducible now, about 1 in 3 times the usefl nears 0, playback stops
00:37:58resky_what can i do? the ipod does not work anymore
00:38:58pixelmaresky_: the reset will always work, keep trying. Sometimes it needs flicking the hold switch on and off and then holding menu+select for a while (up to 30 seconds, a minute?) and don't wiggle your fingers
00:39:50 Quit Daniel_S ("CGI:IRC")
00:42:26zivan56the Sansa View "mediaproc.rom/mi4" also appears to be PP based, although I can't get a key out of either of them or decrypt it with the current key. Perhaps a newer version of mi4 format
00:43:02 Quit Thundercloud (Remote closed the connection)
00:43:06 Join h121h [0] (n=h121@
00:43:07resky_does not work
00:43:31Bagderzivan56: it decrypts with the same key
00:43:43Bagderthe mi4 I mean
00:44:29zivan56does it? I can't get it to
00:44:40Bagderwell it says it can't verify it
00:44:44Bagderbut it works
00:45:01zivan56ah, but I couldn't find any strings in there, so I assumed it didn't succeed
00:45:42BagderI objdumped it decrypted, and it certainly looks like fine code
00:45:47Bagderand an exception table at 0x200 as usual
00:45:58linuxstbAnd I can see lots of strings in it...
00:45:59zivan56actually you are right, there are some things describing ascii and stuff
00:46:26*zivan56 was probably looking at the unencrypted rom for some reason
00:47:38 Quit jhMikeS (Read error: 104 (Connection reset by peer))
00:48:40 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
00:49:40Bagderit seems the view is planned to appear in Sweden on monday
00:49:50pixelmaLlorean: what's "OCD"?
00:50:23linuxstbDoes the View mean Sandisk are discontinuing the e200?
00:50:30BagderI believe they already did
00:50:44Lloreanpixelma: obsessive-compulsive disorder.
00:50:49BigBambipixelma: Obsessive Compulsive Disorder
00:51:04zivan56although they are still selling them here (e200s), I do believe they are not updating the firmware or anything anymore
00:51:18pixelmadidn't know the abbreviation...
00:51:29zivan56good to hear the view is coming to's not coming to Canada any time soon apparantly
00:52:13pixelmaLlorean: thank you... ;)
00:52:19Bagderfunnily I found like ten stores selling it, but they all just mentioned that it hasn't arrived yet
00:52:33Bagderwell, those that I checked at least, I didn't check them all...
00:53:05zivan56sounds like what happened with the US release, it just appeared on day
00:53:31nanokyeah, i had a hard time getting my e200 one week ago: some even told me they are waiting for the new model
00:53:49nanokfor fucks sake, i don;t want the new bloody model, i want this one..
00:53:53Dark_ApostropheHello. Normally I use the 7words theme in Rockbox, but I decided to try some other ones out, which caused a problem: When I switched back to 7words, the icontheme didn't switch back to tango
00:53:59LloreanI wonder if the View will be as easy as the e200/c200 or will show enhanced security like the e200R when time comes to upload something
00:54:01Dark_Apostrophecould anyone help me out?
00:54:37BagderLlorean: at least they provide the .rom file, which is a good sign
00:54:42zivan56I found only one store, that's online only, selling it here: $260 CDN for 16 GB vs $199 USD for 16gb in the US. But 1 USD = 0.93 CAD
00:54:52nanokDark_Apostrophe: if you switch back to the default team, and after to the one you want, is it okay then?
00:56:23LloreanBagder: Does the view support Rhapsody channels, or just the normal WMA encryption, do you know?
00:56:33Dark_Apostrophenanok: No, already tried that
00:56:37Bagderthe firmware mentions Rhapsody
00:56:49BagderI mean in strings
00:56:58zivan56the rhapsody channels appear only if you use the rhapsody software to enable them
00:57:24LloreanMaybe the extra hidden partition and signature turned out to be a hassle for Sandisk as well.
00:58:07pixelmaDark_Apostrophe: if you browse .cfg files, is there one that could be your icon set?
00:59:01zivan56the Sansa View tries to hide UMS/MSC mode, and you have to use a "hack" of sorts to enable it. Maybe they are trying to hide the partition behind MTP mode?
00:59:10*Bagder dives into bed
00:59:19 Quit amin ()
00:59:37Dark_Apostrophepixelma: I'll check, once I find my USB cable
00:59:52pixelmaI mean on your player
01:00:04Dark_ApostropheWell, .rockbox is hidden
01:00:16Soapzivan56: see PM for (useful?) (interesting?) off-topic.
01:00:27nanokBagder: are people still using those? i thought sleeping in your chair, head on the keyboard, is standard these days
01:00:37nanokconfortable too, if you have a decent keyboard
01:00:48*nanok thinks beds are a waste of space
01:01:08pixelmaDark_Apostrophe: you could choose "browse .cfg files" from the menu...
01:01:28Dark_ApostropheOk, gimme a few mins
01:02:59 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
01:03:37 Join sdoyon [0] (
01:06:46 Quit ender` (" I find television very educating. Every time somebody turns on the set, I go into the other room and read a book. -- Grouc")
01:07:50resky_ok i got it to load again ;)
01:08:15resky_how do i tell my database taht it should look in a specific directory, say /Music
01:09:30LloreanIt always looks in all directories
01:10:46resky_doesnt work. i went to settings and general settings and tehn database
01:11:03resky_there i klicked update now and initialize now
01:11:23resky_now my database is empty, but i have like 6 wma files in /Music
01:12:57SoapDo they have proper metadata?
01:13:08resky_yeah. everything.
01:13:28resky_title, artist etc.
01:13:56***Saving seen data "./dancer.seen"
01:14:09resky_rythmbox displays everything fine.
01:14:11 Quit MrRichard ()
01:15:06Dark_Apostrophepixelma: In config.cfg, there are these lines:
01:15:07Dark_Apostropheiconset: /.rockbox/icons/-.bmp
01:15:07Dark_Apostropheviewers iconset: /.rockbox/icons/viewers.bmp
01:15:24resky_is there some trick to get the metadata from wma files?
01:16:00 Quit h121h (Read error: 110 (Connection timed out))
01:16:11Dark_Apostrophepixelma: .rockbox/icons contains tango_small.bmp tango_small_viewers.bmp tango_small_viewers.icons viewers.bmp
01:16:36Dark_ApostropheSo do I change iconsen, viewers iconset, both or neither? And to which files?
01:16:56resky_the wma files do NOT have drm.
01:18:20pixelmaDark_Apostrophe: I think you have to chose the .icon file
01:18:36Dark_ApostropheBut the current one is .bmp
01:18:55resky_its kind of strange: the now playing displays everything like it should be, but the db cannot find any files.
01:19:05pixelmaDark_Apostrophe: err... wait
01:21:11 Quit oblib ("Leaving")
01:21:38pixelmaif I understand the wiki right you have to chose a .cfg file - and not look inside the icon folder. If I understand correctly, it should be there when browsing .cfg files from the menu... (but I don't use that feature myself)
01:21:46nanokresky_: did you reboot after database initialization? or was this not the first initialization?
01:22:31pixelmaDark_Apostrophe: or maybe in "browse themes" as an own "theme"
01:23:52resky_nanok: it was not the fist one, but i tried to initialize and update a few times and somehow it worked. very strange
01:23:52 Quit bertrik ("bye")
01:24:02pixelmaor at least that's implied by the example zip I'm currently looking into
01:25:11nanokresky_: so now it is okay?
01:25:41resky_yeah somehow. but dont ask me why it did not work the first 3 times or so.
01:25:59nanokresky_: it usually takes a while
01:26:06 Quit moos ("nighty...")
01:26:11nanokthe database is updated in the background
01:26:45 Nick fxb is now known as fxb__ (
01:26:55nanokmaybe it would be nice to have a status of the db, when going to the db view "updating" if it is currently not done
01:27:00nanokor something like that?
01:30:03resky_i really thought it did not work.
01:30:21resky_well now i know that this takes some time.
01:33:30 Quit zivan56 ()
01:36:09 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
01:38:07 Quit XavierGr ()
01:38:23 Quit zicho ("*.net *.split")
01:39:52 Join XavierGr [0] (
01:39:53 Quit obo ("bye")
01:46:11sdoyonIs there a need for VOICE_CURRENT_TIME, since there is a LANG_CURRENT_TIME with the same voice entry? Hmm the voice has an extra colon...?
01:47:21 Join japc [0] (
01:53:05 Join darkapostrophe [0] (
01:54:02 Join Thundercloud [0] (
01:54:46 Join MrRichard [0] (i=Tentacle@
01:55:02MrRichardI've tried everything and can't seem to work out what's wrong with my database :(
01:55:22 Quit FOAD (Read error: 104 (Connection reset by peer))
01:56:42 Join FOAD [0] (
02:01:34sdoyonLast change to info screen, r15381 by lenzone10... A few oddities in there. Was there any discussion of that? I can't seem to find anything in the IRC logs.
02:03:45pixelmathere was a short one the next morning. Everyone agreed that it was wrong of him, especially without asking about it first. Noone reverted it though and I don't know if he was contacted
02:05:27sdoyonI'll look it up. thx.
02:05:29pixelmalink to the talk
02:07:45pixelmaI'm off to bed now though, night.
02:08:26sdoyonOK thanks. Err on that commit, I think there may be an error or two in there, regardless of what the order should be...
02:09:30 Quit Dark_Apostrophe (Connection timed out)
02:09:30 Nick darkapostrophe is now known as Dark_Apostrophe (
02:09:55pixelmahmm, ok. Maybe I try to catch someone to talk about it tomorrow again
02:10:47sdoyonSigh... Sometimes I wish I were in another timezone :-).
02:11:51pixelmaI'd also like to know if it really improves things for blind or visually impaired but I'm too tired for it :)
02:12:09pixelmato talk about it now I meant
02:13:22 Quit resky_ ("Lost terminal")
02:13:47sdoyonWell, I'll just comment on it here, and you can look it up in the logs tomorrow, even if it feels like talking into thin air.
02:14:03pixelmawas about to suggest that :)
02:14:41 Part pixelma
02:14:57 Quit matsl ("Riece/3.1.2 XEmacs/21.5-b28 (fuki, linux)")
02:21:46sdoyonAbout the order: TIME used to be first and that was nice because blind users don't have it anywhere else. However blind users really need quicker access to this somehow, and sighted users don't need to see it in that screen, plus it won't update of course. Date is seldom consulted. VERSION is essential yet rarely checked, and the spelling is unpleasant. I imagine a sighted user might like it on top, and a blind user not. Battery lev
02:22:42hcssdoyon: you got cut off at "Battery lev"
02:23:23sdoyonOh? thx.
02:23:38sdoyonBattery level would be the second most useful after time.
02:23:56 Quit TotallyInfected (Remote closed the connection)
02:24:22sdoyonBufffer size is rarely consulted of course. lenzone10's idea might have been that using UP to wrap around, items at the bottom are closer than those in the middle.
02:24:37 Join TotallyInfected [0] (
02:24:53sdoyonHowever it feels rather strange to me when going through the items in order.
02:25:56sdoyonHe also tweaked something wrt disk size/free, haven't figured out exactly what at this point.
02:26:25 Join toffe82 [0] (
02:26:53sdoyonsnprintf(buffer, MAX_PATH, "%s %s %s", str(LANG_DISK_NAME_MMC), s1, str(LANG_NOT_PRESENT), s2); looks wrong to me.
02:27:44 Quit animeloe (Read error: 110 (Connection timed out))
02:27:58sdoyonHe changed LANG_CURRENT_TIME to VOICE_CURRENT_TIME, no idea why exactly. In fact these look like duplicates to me (except the colon that should do nothing) and one of them should probably be eliminated.
02:29:54sdoyonAnother caveat: the spoken info is current as of when we start talking, even if the screen does not update.
02:31:14sdoyonOh and "micro Secure Digital card:" (voice for LANG_DISK_NAME_MMC) is just as painful to listen to as it is to read. Needs to be shortened.
02:32:17Mouser_XHow about Micro SD card?
02:32:24 Quit Thundercloud (Remote closed the connection)
02:33:37sdoyonYeah. I guess I should just shut up and go change that :-).
02:36:09 Join lespea [0] (
02:43:05lespeadoes anybody here use foobar?
02:43:50Soaplespea: please keep the conversation on topic in this logged channel. There is a #foobar2000 channel right here on freenode, though.
02:44:20lespeaerr... that was my lead-in question to my real question haha
02:44:22 Join WalkGood [0] (i=WalkGood@unaffiliated/walkgood)
02:44:45lespeaI was wondering if anybody had like a script or something to transfer the ratings in foobar to the rating system that rockbox uses
02:46:35Soaplespea: didn't mean to be so presumptive.
02:46:55lespeaSoap: oh no problem, it was a good point :)
02:47:50Mouser_Xlespea: I'd have to say that it's unlikely. I'd like something like that for Winamp (what I use), but I've never heard of one existing (then again, it's not like I've looked extensively either).
02:47:50lespeathe ratings are stored in a (non standard) id3v2 tag 'rating' from 1 to 5
02:49:12lespeaI guess I could write a perl script or something but I'm quite lazy atm lol
02:49:22 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
02:52:26sdoyonpixelma: he apparently swapped the #else cases for DISK1 and DISK2 so that free comes before size. I think I agree with that.
02:52:37 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
02:53:10sdoyonpixelma: Oh and saying "not present" instead of nothing, that's definitely a plus. An unspoken line in a list is pretty confusing.
02:53:43 Quit linuxstb ("ChatZilla [Firefox]")
02:55:23sdoyonpixelma: Another thing that lenzone10 should NOT be blamed for: the white space damage would appear to be a result of the CRLFs that were previously in that file.
02:56:08 Quit kubrick (Read error: 104 (Connection reset by peer))
02:57:21 Part WalkGood
03:11:28 Join sarixe [0] (
03:11:58 Quit sarixe (Client Quit)
03:12:38 Quit kugel ("ChatZilla [Firefox]")
03:14:01***Saving seen data "./dancer.seen"
03:16:05 Quit sneakums ("blorp")
03:18:47marcosourcehey, I managed to run my rockboc'ed sandisk sansa e270 today from 10 until 7 today, and the battery timer cet saying there was about 4 hour left from start to end, and it still has about 2 hours left acording to the timer, it might need a revision or two
03:19:06lostlogicbagh, how does the bufopen logf get ahead of the Buffering track logf?
03:22:12 Join webguest52 [0] (i=58a10615@gateway/web/cgi-irc/
03:23:10webguest52little question: can i use a e200tool with a NOT bricked Sansa just to see if it's recognised ?
03:24:29 Quit animeloe (Read error: 110 (Connection timed out))
03:26:09 Join TMM [0] (
03:26:27 Join donsdw [0] (
03:27:26webguest52I connected my Sansa, ran "./e200tool off" but only obtains "Searching for device 6666:e200 ... 9 8 7 6 5 4 3 2 1 0 not found!"
03:27:50lostlogichcs: still around by chance?
03:27:57hcslostlogic: yep!
03:28:00 Join annulus_ [0] (
03:28:07 Part donsdw
03:28:22lostlogichcs: cool, you're still having freezes and badness on your sansa right?
03:28:33lostlogicI'm looking into it now, how long you likely to be around for?
03:29:07hcsProbably a few hours.
03:29:23lostlogicawesome, will ping if I come up with something
03:29:42hcsgreat, thanks
03:29:49 Join darkapostrophe [0] (
03:30:40 Quit Dark_Apostrophe (Nick collision from services.)
03:30:41 Nick darkapostrophe is now known as Dark_Apostrophe (
03:32:20 Join sneakums [0] (
03:36:49 Quit webguest52 ("CGI:IRC")
03:39:22lostlogicdoes an O(n) on stack size operation over num_handles sound like a bad idea?
03:39:26 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
03:40:32lostlogic2400 bytes, growing by 8(?) bytes per recursion is still 300 handles...
03:44:54lostlogic-58 on ipod video is not actually entirely muted... I can detect the music in my silent house with my sensitive noise blocking earphones
03:46:58hcslostlogic: I'm getting periodic freezes in MP3 playback, as well, but I'm using 15407 as it was the last with your hack for SPCs in
03:48:16 Join ddalton [0] (
03:48:16lostlogichcs: gotcha
03:48:23ddaltonhey sdoyon
03:48:30lostlogicI still can't figure out how the latest is different for SPCs than the hack
03:48:36lostlogicit should be the same path.
03:48:56ddaltonsdoyon: your beeping notification doesn't seem to work for me when playback is on now.
03:48:59ddaltonnot sure why
03:49:10sdoyonddalton: Hi
03:49:33sdoyonddalton: beeping notification?
03:49:34hcslostlogic: well, I'm putting the latest on now to see how it handles MP3s, did you see my update to the bug report?
03:49:45lostlogicno, still suck at reading fs
03:49:51lostlogicwhich fs#
03:49:53ddaltonsdoyon: sorry I mean beep-list...
03:50:16 Join midgey [0] (
03:50:59sdoyonddalton: Must be some of the recent playback changes I suppose...
03:51:06ddaltonsdoyon: Im not coding on rb so won't update. im using r15273 I think but let me check.
03:51:32ddaltonno. I didn't update. It just broke on the ame build. Maybe I did something.
03:51:49ddaltondon't worry when I start coding again ill try and fix it
03:51:51lostlogichcs: oh joy.
03:52:21hcsyeah, it's still sitting there not loading, but now blocking everything up
03:52:25sdoyonddalton: do you use .talk clips? Are they fully spoken, when NOT playing music at the same time?
03:52:27ddaltonsdoyon: can you join the other rb channel? So I can ask you a couple of questions that are OT here
03:53:09ddaltonsdoyon: yes just let me see something
03:53:49sdoyonddalton: Uh OK but quickly, I was in the middle of something. What is it called again?
03:54:04Mouser_X#rockbox-community ?
03:54:33Mouser_X(I assume that's the one he's talking about.)
03:55:29ddaltonsdoyon: I have seen it happen. But it works fine now. I never saw it with disk cash...
03:55:29lostlogichcs: you'll be overjoyed to know that I've just committed two patches that won't have any impact on your issues :-D
03:55:40ddaltonyes community
03:55:48 Quit MrRichard (Read error: 110 (Connection timed out))
03:55:59hcslostlogic: I can't handle this kind of excitement!
03:56:36lostlogicit's amazing the things that I run across while tryign to dig up some crazy race condition that only hits sansa
03:57:06hcsI have two, want the other?
03:57:49lostlogichcs: other race condition that only hits sansa?
03:58:24hcsheh, I meant two sansas
03:58:47lostlogicoh, YES! I'd love to be able to reproduce these things.
03:58:54lostlogicwhat region are you in?
03:59:05hcsNew Jersey
03:59:11lostlogicshit, 3k miles from here
03:59:34lostlogicalmost closer to ship to Nico in France :-d
03:59:51Mouser_X"I've come millions of miles-" (whispered) "thousands" "-thousands of miles!"
04:01:27lostlogicokokok, let's think about this logically. ata is active is always false on sansa, this is the most likely difference between it and other flash targets
04:01:34lostlogicso what does that change about the buffering behavior?
04:03:44 Quit jhulst (Connection timed out)
04:06:51lostlogichcs: try this 1-liner for me:
04:07:36 Join psycho_maniac [0] (
04:10:09hcslostlogic: no good
04:10:16lostlogicworse, same?
04:11:08 Quit animeloe (Read error: 110 (Connection timed out))
04:11:57lostlogic2800 miles newark to seattle, 3600 newark to paris (just in case anyone else was curious)
04:12:48hcsshipping to you'd be cheaper
04:13:13lostlogichcs: if I ahven't solved this by monday, we'll think about it again :)
04:13:27lostlogicI have a fair amount of free time this weekend to read and anylize this code
04:15:11psycho_maniacdo i need to upgrade my bootloader on my ipod video with the recent comments in svn?
04:15:22lostlogicpsycho_maniac: no need
04:15:50lostlogicpsycho_maniac: in general, as long as bootloader works, you can stick to it. Most of the time we are having to fix things that we put into the main build that would break a perfectly functional boot loader ;)
04:17:01psycho_maniaci thought i might of because of this comment: firmware/target/arm/ipod/video/lcd-video.c [diff]
04:17:02psycho_maniacBootloaders are single core... (aka fix red)
04:18:10lostlogicpsycho_maniac: right, he'd committed something that broke the bootloader, so had to fix it
04:18:31lostlogicpsycho_maniac: technically the final result is a bootloader with a faster LCD update than it had yesterday but... who cares?
04:19:06 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
04:19:26psycho_maniaci would if it made my ipod work better.
04:19:51lostlogichcs: can you by chance turn on logfqueues and logfqueues_sys_timeout in both buffering and playback and hit me with a logf from it breaking?
04:20:06lostlogicpsycho_maniac: it might make it boot 20ms faster, that would be the only change
04:20:19lostlogicpsycho_maniac: once the main firmware is loaded the bootloader used to load it doesn't matter
04:20:48hcslostlogic: ok, with the patch you just sent?
04:20:59psycho_maniacalright. i see
04:20:59lostlogichcs: doesn't matter one way or the other ;)
04:23:49 Quit midgey ()
04:23:51lostlogicI wonder if some junk is getting on the queue and that's preventing it from ever hitting the fill_buffer code.
04:24:21lostlogicI also still wonder how the heck the current SVN doesn't work but the hack did.
04:30:17hcsI wonder, does that potentially useless watermark setting thing I have in there mean anything?
04:30:41lostlogichcs: which watermark setting do you have set to what?
04:32:28hcslostlogic: in spc.c I do ci->configure(CODEC_SET_FILEBUF_WATERMARK, ci->filesize);
04:32:58hcsanyway, flyspry updated with the logf you requested
04:33:12lostlogicyes, that is what breaks it.
04:33:15lostlogicfor spc at least.
04:33:27 Join stevenm [0] (n=stevenm@
04:33:38lostlogicsmart man.
04:33:43lostlogicI can see why you did that
04:33:46lostlogicand I can see how it broke us
04:34:01lostlogicdon't change it, probably smarter for us to change the buffering to a <=
04:34:01hcswhat's it supposed to do? and does it do something different from that?
04:34:12stevenmWell, it's Friday night insomina-induced boredom, and you know what that means. How do people feel about the idea of seeking support for midiplay?
04:35:03lostlogichcs: that setting is how low the buffer will be allowed to run before the buffering thread will fill it up
04:35:26lostlogichcs: since spc never advances the buffer (I'm guessing here) it also never gets below the low watermark
04:35:59lostlogicthen since it doesn't have ata disk active it never even checks the high watermark which would cause it to recover from that bad state on non-sansa targets
04:36:37lostlogichcs: do other codecs set the watermark?
04:36:49hcsI think I fiddle with it in ADX
04:37:12lostlogichmm, if ever codec set the watermark to an appropriate value as it started up it would be good
04:37:19lostlogicunless they all do it, none of them should do it though :-\
04:37:24lostlogicwho wants to tackle _that_ one
04:37:26hcsyeah, also set it to the whole file size...
04:37:53hcsNSF doesn't touch it
04:38:15lostlogichas NSF been OK with the current SVN? I remember it was mediocre pre-hack and ok w/hack
04:38:15hcsand, accordingly, is working in the current rev
04:38:34hcsADX works too, though, but the files are of a respectable size
04:39:03lostlogicare they also not read as a unit?
04:39:05 Quit n17ikh|Lappy (Connection timed out)
04:39:57hcsuh, it attempts to read them as a unit
04:40:22hcsit's particularly screwy on account of seeking backwards
04:40:36lostlogichcs: if they are 'respectable' in size, that's definitely not legal... no file > 32k can be read as a unit safely
04:40:51lostlogicunless it's small enough to be treated in the atomic category by the buffer
04:40:57hcsit checks how much is successfully read, though
04:41:20lostlogichcs: the api doesn't allow for a non-complete return atm, it probably should.
04:41:35 Join donutman25 [0] (n=chatzill@
04:42:16hcsIt'd be nice, again, to see these laws. Is it written somewhere and I've just consistently missed it?
04:43:23lostlogicno, it is not −− the headers should be commented
04:44:34hcsI sort of get the feeling that it's grown organically, is there anyone who actually has this all in his head?
04:45:51hcsfun fact: with multiple ADXs in a directory, they don't work
04:46:33psycho_maniacwhat do you mean not work? skipping to the next one does not work? can you play them one at a time?
04:46:33hcsso it's just my own damn meddling breaking things now
04:46:48hcspsycho_maniac: precisely
04:46:56lostlogichcs: I'm guessing not, we jus tread the code as needed ;)
04:47:27lostlogichcs: not really −− that is still only part o the puzzle
04:47:32lostlogicthat solves the "why spc" part
04:47:34 Join midgey [0] (
04:47:36lostlogicbut not the "why sansa" part
04:47:42 Join n17ikh|Lappy [0] (
04:48:00lostlogicI mean sure the sansa doesn't have the disk active, but that stills houldn't break us and bertrik still has a bug
04:48:49lostlogichcs: last summer I had it all in my head :-P
04:49:47Mouser_XAlso, ADX = broken? D:
04:49:53*Mouser_X likes ADX.
04:50:03lostlogicMouser_X: should be fine with my latest commit
04:50:06hcsa) on sansa b) if you're trying to play several of 'em
04:50:08lostlogichcs: check it out fo rme ;)
04:50:13hcsoh, cool, sure
04:50:26 Quit scorche|w ("CGI:IRC (Session timeout)")
04:50:29lostlogicstill doesn't explain the 'bertrik' bug
04:50:39lostlogicwhy would larger files still run down the buffer? grr.
04:51:21hcsso did you find a reason why the watermark adjustment would specifically affect sansa, or just fix the watermark handling?
04:51:30lostlogicthe latter
04:52:03lostlogichcs: would it be feasible in these tiny-file codecs to call the advance buffer callback once the file is done playing back?
04:52:20lostlogichmm, actually no, don't.
04:52:22lostlogicI lied.
04:52:34lostlogicthey can be moved since I call them atomic.
04:54:12ddaltonsdoyon: re my talking battery levels patch should it voice the level on boot? or only after a certain time?
04:57:52 Quit psycho_maniac (" HydraIRC -> <- Chicks dig it")
04:58:14sdoyonddalton: leave it on boot I think.
04:58:25 Join ramon8 [0] (
04:58:27hcslostlogic: r15417 still not working, same behavior
04:58:29ddaltonsdoyon: ok then.
04:58:37lostlogichcs: wtfnowai
04:58:53lostlogichcs: I hate your bug.
04:59:08sdoyonA battery question: with CHARGING_SIMPLE, when the charger is connected, is it correct to assume the battery level reading is meaningless?
04:59:43hcsI'm going to pull out the watermark bit to see if that has any effect
05:01:01lostlogichcs: cool
05:02:09hcsseemingly no
05:03:22hcsI'll do some comparisons with the NSF codec
05:04:41 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:05:01 Join miepchen^schlaf [0] (
05:09:45stevenmWhee, midiplay can seek.. I think... in svn.
05:10:02stevenmbackwards too, I mean
05:10:28Mouser_XI should get some MIDIs then.
05:10:45stevenmI hope I didn't break any builds... added a new key, hopefully right.
05:11:04stevenmMouser_X, it doesn't work too well on portalplayer targets (ipod, etc) because it's only using only one CPU at the time
05:12:01*Mouser_X has a Gigabeat.
05:12:14stevenmit may work.. I only have h300
05:12:20hcslostlogic: I don't see anything different, after removing the watermark stuff, between SPC and NSF. Maybe it's the fact that SPC files are consistently just over 64K (twice 32K)?
05:12:39hcsi.e., I wonder what would happen if I padded some NSF files to 64K+1
05:13:12lostlogichcs interesting thought
05:13:14hcsbut then again, ADX also fails
05:13:26lostlogicwhat size are adx?
05:13:50hcslarge, the ones I have here are > 10 MB
05:14:03***Saving seen data "./dancer.seen"
05:14:07hcsbut I don't have many NSFs over 64k
05:14:09lostlogichate you
05:14:37hcsdon't hate the player, hate the playback code
05:14:57lostlogichate sansa
05:15:04sdoyonddalton: Did anyone else comment on your battery announcement patch?
05:16:03 Quit bb (Nick collision from services.)
05:16:08 Join bb_ [0] (n=bb@unaffiliated/bb)
05:16:53hcsaha, largish NSFs (144k) do not work
05:17:01ddaltonsdoyon: one guy tried it out I think. I know JdGordon doesn't like it though.
05:17:22ddaltonsdoyon: so do I just put that stuff you mailed me in .emacs and open emacs and start coding?
05:17:48lostlogichcs: problem is the requests for > 32k reads then most likely
05:18:06lostlogicso my atomic stuff isn't working right
05:18:10hcsI had some files between 32k and 64k that seemed ok, though
05:18:20lostlogicthen it becomes probabilistic
05:18:21 Quit Rondom (Nick collision from services.)
05:18:39 Join Rondom [0] (
05:20:36hcsheh, if you copy/paste a file multiple times into a folder in Nautilus, it calls the first X, the second X (copy), the third X (another copy)
05:21:12hcsanyway, tested and working with 56k NSFs
05:21:32stevenmthe only problem with seeking that I can see is that playback should be turned off while seeking is taking place. But I don't know how to do that.. the basic implementation is there, anyway.
05:21:47sdoyonddalton: I guess. Don't know if it'll all work with your version. Has to be .c extension. And for RB do control-c dot stroustrup.
05:26:33 Quit midgey (Read error: 104 (Connection reset by peer))
05:29:00ddaltonsdoyon: Im just writing up a reordered info screen patch. I think im going to have to resync it from about two commits ago to now.
05:29:08lostlogichcs: ok, I once again think I know what's going on, but it'll be somewhat in depth to fix it, especially for adx
05:29:14ddaltonsince your changes. But anyway ill see how I go.
05:29:45ddaltonwhat do you think about: -battery level -time -date -disk 1 -disk 2 -version -buffer
05:29:50ddaltonis that ok ordering?
05:29:52sdoyonddalton: Thought you weren't coding on RB for a while...
05:30:45ddaltonsdoyon: yes but im waiting for my linux box and got nothing else to do. Going away for the night in about half an hour. and this is fairly simple
05:30:46sdoyonddalton: If you could go read the logs of earlier tonight, I already typed all I had to say on that :-).
05:31:09ddaltonoh ok ill do that tm haven't got time now. sorry.
05:31:14stevenmoh well, time to go meander before finally succumbing to sleep. bye all
05:31:16 Quit stevenm ("Connection reset by beer")
05:31:35hcslostlogic: care t enlighten me?
05:32:38lostlogichcs: actually, to confirm my theory I need you to do another logf for me :-\ you can turn off the LOGFQUEUE*s if you want and use the code that I just dcommitted moments ago
05:32:41 Part toffe82
05:33:01lostlogicI need to know whether the SPCs and NSFs are loading getting marked TYPE_ATOMIC_AUDIO or not and that logf will tell me so
05:33:09ddaltonsdoyon: anyway what were you saying? do you disagree with what I said?
05:33:29hcslostlogic: sure
05:34:25sdoyonddalton: about what?
05:35:55ddaltoninfo screen...
05:36:26ddaltonsdoyon: also could you send me your updated patches? I just don't want to resync much and just the combined one is ok.
05:36:29ddaltonor both...
05:37:56sdoyonddalton: info screen: as I said, please go read the IRC logs of earlier tonight. Well tonight for me :-). Say 4hours.
05:39:00sdoyonddalton: re stack of patches: OK,next time I put it together. Lately I've been working closer to stock RB.
05:40:16lostlogicgah, this is going to hurt my head.
05:40:24lostlogicI might hafta pencil and paper this out.
05:41:39ddaltonsdoyon: ok thanks. How much resyncing do you think combined.diff will require since the last one you sent me?
05:41:49ddaltonr15273 I had it working on
05:42:45sdoyonddalton: No idea sorry. I did resync more recently though.
05:43:09sdoyonddalton: If you'll just let me work, I can finish committing your patch :-).
05:43:53hcslostlogic: erm, not seeing the Loading atomic message
05:44:02ddaltonok thanks
05:44:34lostlogichcs: that's what I was afeared of :(
05:44:41lostlogicso that's _part_ of the problem
05:44:49lostlogicat least that would solve SPC and NSF if it was showing up
05:44:50ddaltonBTW if you are using the patch I uploaded I think a test line slipped in there. might want to remove that.
05:45:25hcsdo you want the logf anyway?
05:45:42lostlogichcs if the atomic message isn't in there, I don't want to see the "goddamned" thing
05:45:46sdoyondalton: Indeed. I saw it but thanks for the warning. Must remember the plugin bump...
05:45:52lostlogic(I'm not as angry or frustrated as I might sound ;))
05:48:12Soaphe's more
05:48:19 Join perrikwp [0] (n=chatzill@
05:48:22*lostlogic pulls Soap's hair
05:48:29ddaltonsdoyon: oh yeah. whats the plugin bump? didn't quite understand that.
05:48:36ddaltondon't worry if your busy though
05:48:41hcsangry people don't quote "goddamned"
05:48:51lostlogichcs: that was my intent ;)
05:55:01lostlogichcs: so why the heck?
05:56:02hcslostlogic: if I answered that question with "because" would you be satisfied?
05:56:26lostlogichcs: satisfied enough to throw you out the window, yes.
05:56:43hcssatisfaction through defenestration
05:56:56lostlogic:-D best word evar
05:59:13*Llorean is fond of the word.
06:00:58lostlogiczomg I gets it
06:01:07lostlogicit bounces _itself_ out of the fucking fill_buffer
06:01:11lostlogicbecause fuck
06:01:13lostlogicwtf fuck
06:01:27lostlogicthe next metadata puts it over the threshhold
06:01:35lostlogicbut why doesn't that then enter the other buffer loop?
06:02:33ddaltonsdoyon: are you going to commit 7910?
06:02:54sdoyonddalton: Done.
06:03:10ddaltonsdoyon: thanks. Lets see what blind users think.
06:03:18ddaltonhalf them don't build from source...
06:04:38sdoyonddalton: You might want to contribute an update to the manual to explain the new setting.
06:06:19ddaltonsdoyon: oh yeah good idea. Ill have to do it when I get back.
06:06:26ddaltonmaybe I can do it when I am away
06:08:04ddaltonsdoyon: do I need to add it for every target?
06:08:11ddaltonor just modify one file
06:08:30sdoyonddalton: I don't know actually.
06:08:49sdoyonddalton: Uh wait no it's only in one place.! now I can't build meg-fx. :(
06:11:23lostlogicI didn't do it! yay.
06:12:18*jhMikeS sees the culprit nearby >:[
06:13:54ddaltonsdoyon: where is it?
06:14:21jhMikeSsdoyon: RED!
06:14:35sdoyonUh oh.
06:14:38hcsscramble the green deltas!
06:14:54*ddalton 's patch broke something?
06:15:03*jhMikeS threw in another commit too :)
06:15:35jhMikeSLooks like the build is stuck too...10min = too long
06:16:19Mouser_XUnless it's done on my craptop. Then that'd be a new record.
06:17:08Mouser_X(It's an hour+ on my laptop.)
06:17:11jhMikeSwell it was due now 12m ago
06:17:13sdoyon'LANG_BATTERY_TOPOFF_CHARGE' undeclared. That's one of my info screen commits. A feature problem...?
06:19:08jhMikeSsdoyon: Probably. Not sure what the common thing is for the iRivers, Gigabeat and FM is off hand. Needs fixing though.
06:19:11lostlogichcs: well I'm once again pretty sure I can fix "something"
06:19:22lostlogicbut it won't really fix nsf and spc unless we can make them buffer as atomic :(
06:19:43sdoyonTOPOFF, /* Can occur for CONFIG_CHARGING == CHARGING_CONTROL */
06:19:57*jhMikeS wonders why the build servers only need a kick when the swedes are noonwhere to be found :\
06:20:01sdoyonBut the lang entry exists only for recorder. Does this make sense?
06:20:19sdoyonGee that build page is complicated.
06:20:23jhMikeSlostlogic: buffer as atomic?
06:20:33lostlogicjhMikeS: atomic => read as a unit, may not wrap
06:21:01lostlogicjhMikeS: most audio is packet => will be read in < GUARDBUF size bits
06:21:08jhMikeSor change the spc codec a little bit to accept a wrap. It stores the data locally anyway.
06:21:20Mouser_XSPCs require 100% of the file to be available, or else it can't be played.
06:21:22lostlogicjhMikeS: it does? I thought it something something
06:21:29lostlogicMouser_X: right, hence my atomic
06:21:31hcsNSF doesn't
06:22:04 Quit Mouser_X ("CGI:IRC (EOF)")
06:22:15lostlogicjhMikeS: I don't think they need a kick btw, my server is still chugging on builds
06:22:16jhMikeSlostlogic: It just reads into a local buffer but it should be able to accept wrapped data and do two reads.
06:22:35lostlogicjhMikeS: if it reads into a local buffer it can, but I thought it just took a pointer to the buffer
06:22:42lostlogicwhich is how request_filebuf callback works
06:23:08hcsjhMikeS: that's true, it only needs the whole buffer so that it can instantly restart, but we could just use the normal mechanism for seeking backwards for that
06:23:24jhMikeSlostlogic: not that cheats
06:23:47hcsNSF needs is, though, as the data is not a fixed size
06:24:05 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:24:08jhMikeShow big can an NSF be anyway?
06:24:11hcsthough I suppose bank switching could be implemented as seeking
06:24:33hcsmaximum? over a megabyte
06:24:43hcsI'd have to look at the header
06:25:01 Quit Mouser_X (Client Quit)
06:26:09 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:26:17 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
06:26:54jhMikeShow'd the server kick? Is the kicker interface implemented?
06:27:15jhMikeSor is it automated now? somethings different.
06:27:21hcsjhMikeS: largest I've ever seen is about 300k
06:27:26sdoyonThink I have the fix. Trying to build an affected target locally.
06:28:39Mouser_XHCS: I was going to say. A 1 MB NSF would be huge! I've never seen one anywhere near that size.
06:28:59 Join qweru [0] (
06:30:37jhMikeSmeaneth that one should not account for? :p
06:31:01hcshm, at 4k per bank and 256 maximum banks, biggest would be exactly 1MB, actually
06:31:44jhMikeSso the codec malloc buffer can do half that
06:32:00hcsthere's a codec malloc buffer?!?!
06:32:08jhMikeS512KB of it
06:32:27hcswell I'd say the problem is solved, then
06:32:37hcsI'd be glad to fail files that don't fit in that
06:33:08jhMikeSthough it is sort of a target for being rid of it but can work for now I guess.
06:33:29lostlogicgrr, jhMikeS can you take a look at the the bits near the pattern AFMT_SPC and see if you know why that logf isn't happening (playback.c)?
06:33:32lostlogicI dont' get it
06:33:34hcsfor ADX, the whole-file-reading bit is just an attempt to make backwards seeking work better, but it's nothing inherent in the format that requires it. it can work on 18 bytes at a time
06:33:59lostlogichcs: then attempt reads < 32k at a time.
06:34:01hcsit could work on one byte at a time, really...
06:34:38jhMikeSlostlogic...which lines?
06:34:43hcsonly if you change the buffering to fail on a request > 32k to force me into it
06:34:44lostlogicso are we telling me that I don't need TYPE_AUDIO_ATOMIC any more?
06:35:05hcshow does SID work?
06:35:06lostlogicjhMikeS: 2431 and the 20 above and 10 below
06:35:08jhMikeSlostlogic: well, we will need flat buffering services
06:36:02*jhMikeS was hoping that switch would be eradicated
06:37:09lostlogicflat buffering?
06:37:31hcsah, SID's a maximum 64k, of course
06:37:36jhMikeS1) the offset <= 0, or track_id3->codectype != AFMT_SPC ??
06:38:00jhMikeSlostlogic: for MOD file which are large and must not wrap at all
06:38:18hcsSID reads into its own buffer w/ read_filebuf, if I used that would I have to worry about 32k?
06:38:43jhMikeSif it's less than the guardbuf size, then no
06:39:00lostlogicjhMikeS: shit, offset.
06:39:19hcsjhMikeS: so is there any read that'll work reliably regardless of the size?
06:39:45lostlogicjhMikeS: I knew you'd figure it out :)
06:39:51hcsor do I have to sit there and grab a chunk at a time myself?
06:39:59jhMikeSlostlogic: the offset will always be 0
06:40:36jhMikeShcs: if it returns less size than you need, keep reading
06:41:29hcsjhMikeS: and will that work? or will it just freeze up sometimes if I request > 32k like it does now?
06:41:56lostlogichcs: as long as each read < 32k you're fine
06:42:14lostlogichcs: the buffering api fills a minimum of 32k per read by going to the disk if needed
06:42:26lostlogicbut it can only guarantee one file chunk before it might get interrupted
06:42:46lostlogicjhMikeS: gah, then why it not works if offset == 0
06:43:00hcsfine. is this 32k a constant somewhere I can use?
06:43:16 Join animeloe_ [0] (n=animeloe@unaffiliated/animeloe)
06:43:34lostlogicthose two shouldn't really be allowed to not be equal.
06:44:02hcsI'll use CONF_FILECHUNK, makes more direct sense
06:44:09 Join hannesd_ [0] (
06:45:01 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
06:45:05jhMikeSlostlogic: I suppose the cheapest flat buffering would be is if the file would exceed the end of the guardbuf, buffer it at the start of the buffer.
06:45:46 Quit animeloe (Read error: 113 (No route to host))
06:45:49lostlogicjhMikeS: yeah, that's how TYPE_ATOMIC_AUDIO works
06:45:58lostlogicif I could get it to just set itself right on bufopen *grr*
06:46:29jhMikeSplayback.c still needs alot of straightening out.
06:46:52lostlogicjhMikeS: yeah, there is a lot of duplication I'm seeing still between buffering and playback :(
06:47:46jhMikeSplayback should be modified to comply with buffering if anything. don't hack up new code to make playback work. :P
06:48:03lostlogicjhMikeS: yes.
06:48:15lostlogicplayback should pass in the TYPE_ATOMIC_AUDIO
06:49:31jhMikeSI do have a plan to try to make a decent buffer rotation long as buffering can move all the handles around with it nicely.
06:50:42lostlogichcs: for spc, or any codec which will not advance the buffer throughout playback, please set the watermark to 2x the file size, otherwise the buffering api has no way to know when to load more of them until that file _finishes_ causing a pause in payback while it buffers.
06:51:09lostlogicI'm about to commit what I _think_ will fix spc for _most_ cases and not cause anything to be worse
06:51:19lostlogicbut I'm not 100% sure so I expect bug reports :)
06:51:32hcslostlogic: twice the size of the current file?
06:52:58sdoyonFixed BTW.
06:53:07 Join animeloe__ [0] (n=animeloe@unaffiliated/animeloe)
06:53:18 Quit tchan (SendQ exceeded)
06:53:26lostlogichcs: that works, the correct size is: sizeof(struct memory_handle) + sizeof(struct mp3_info) + filesize + 1 I think.
06:53:41jhMikeSlostlogic: not sure if that makes sense to me
06:53:53jhMikeSthe think about 2x that is
06:54:20lostlogicjhMikeS: if the codec never advances during playback, the buffering api doesn't know that the file is about to end and can't know to start buffering the next stuff, so it needs to start buffering when the previous file is closed
06:55:12hcshm? doesn't it buffer the next stuff anyway?
06:55:19jhMikeSlostlogic: it should fake the advancement since it has an ET associated with it. don't know why it shouldn't just do that.
06:56:01jhMikeSI would think it should buffer until the buffer fills or until data runs out.
06:56:31jhMikeSAnd when are bufferings going to happen across directories? :) Any directory awareness should be expunged from playback.
06:56:31 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
06:57:31hcsAll this is making me want to spend a few months writing my own dang buffer management.
06:58:01Mouser_XI thought that this was why Nico_P wrote the MoB stuff...
06:58:08Mouser_X(Or something like that.)
06:58:10jhMikeSplayback should just be a dumb engine that simply plays a list of files given to it.
06:58:25 Quit hannesd (Connection timed out)
06:58:26 Nick hannesd_ is now known as hannesd (
06:58:32lostlogicjhMikeS: it can't fake it −− the buffer only knows about bytes and if it frees some bytes, buffer could overwrite them but it needs thema ll in order to be playing
06:58:54hcslostlogic: which is something I encountered with ADX early on
06:59:20lostlogicso the threshhold for filling more needs to be large enough to trigger when the previous track is done
06:59:29lostlogicunless someone else has a better idea
06:59:43*lostlogic = pragmatic problem solver. "will it work?"
07:00:31 Quit animeloe_ (Read error: 110 (Connection timed out))
07:01:07jhMikeSlostlogic: you're saying only one SPC is buffered and not the entire list of them?
07:02:58lostlogicjhMikeS: no, imagine a list large enough to fill the buffer. it buffers them all, plays through them, when it gets to the last one on the buffer, that one's full size remains valid on the buffer until that file is doen, now the disk must spin and fill _after_ the song finishes befor ethe next track will play
07:03:09lostlogicthe way to solve this is to have the buffer fill when the 2nd to last song on buffer is done.
07:03:19lostlogicand for an atomic track I don't see another practical solution
07:06:10jhMikeSlostlogic: I have a similar dilemma with MPEG packets and stream lookahead but it's much easier to solve with flat windows which are implemented on the buffer but the stream parser doesn't do the lookhead yet to expand the stream window. It must pin down the data starting at the previous position until it has parsed out the next packet but at the same time must look ahead to trigger while parsing.
07:06:38 Quit sdoyon ("ircII EPIC4-2.4 -- Are we there yet?")
07:07:48lostlogicok, so the other option in t his case is to have the playback engine do a "test the waters" read on the next file when it's anti-skip seconds from the end of the playing one.
07:07:49jhMikeSThe buffer just contains the byte range that is in memory and the stream requests the byte range it wants. it is notified when the range exists.
07:08:41lostlogicI'm not against the idea, but it's a large departure from the current architecture for low buffer detection
07:08:42jhMikeSThere is no wrapped calculation anywhere in mpegplayer now.
07:08:49lostlogicputtin the onus on playback instead of on buffering
07:09:45jhMikeSI think you should convert buffering to do the same and life will be easier and code simpler.
07:10:33lostlogicjhMikeS: yes, I see your poitn, in the meantime, can I please just have the 3 line fix tha twill make it work!?!?!?!?!
07:11:38lostlogicjhMikeS: if I can get the immediately BROKEN stuff working, I'll think about more major rearchitecture after that
07:12:23*lostlogic snippy for real now, sorry. late, tired, don't need argument about the architecture when just trying to make the immediate bugs die.
07:13:50jhMikeS3 lines? it _has_ to 3? :0
07:14:05***Saving seen data "./dancer.seen"
07:14:18*lostlogic sighs :)
07:14:19lostlogicI'm tired.
07:14:26*jhMikeS should be snippy given something weird is happen in what he's doing and can't find a bug in the code
07:14:53jhMikeSthis is video stuff that I'm talking about
07:15:25jhMikeSbut anywho, can't the position crap just be for certain formats?
07:15:32lostlogicahhh, joy of joys (a piece of code that will never get my loving attention because... video!? on my ipod!? naw.)
07:15:40lostlogicwhich position crap?
07:15:46lostlogicthe one you suggest or the existing one?
07:16:20jhMikeSmove the offset checks into the cases? will that help?
07:16:21lostlogicsorry, I'm still confused
07:16:38lostlogicjhMikeS: oh gotcha
07:16:46hcswell, goodnight all, I expect to see a completely rewritten playback engine in the morning
07:16:48lostlogicyeah, I'll do that before I commit −−
07:16:52lostlogichcs: laugh
07:16:57lostlogicI'm sleeping soon
07:17:02lostlogichopefully I'll break it first
07:17:14hcsoh, was there to be a commit impending?
07:17:30lostlogichcs: yes, my test cases just passed.
07:17:53*jhMikeS hopes the SPC codec doesn't change radically or anything so the DC version can be resynced easily.
07:19:01hcsjhMikeS: all that will be changing is the loading
07:19:07hcsif I ever get to doing that
07:21:51lostlogiclet me know how badly it breaks, in the next 20 minutes before I pass out
07:22:00hcsI fetch
07:22:19lostlogichcs: thanks for your patience with the new buffering stuff and all of the bugs that Nico and I have variously fixed and introduced.
07:22:29*jhMikeS hands lostlogic some coffee...or something stronger is for that :P *joking*
07:22:45hcslostlogic: heh, thanks for doing all the work.
07:23:19lostlogicit's such a tangled web of code and someone who understands it should really put a lot of effort into untangling it so that any old person can fix bugs in it... sigh.
07:23:32lostlogicjhMikeS: :-P
07:23:48lostlogicit's funny, I get frustrated and/or snippy about things, but I _love_ this shit.
07:24:08*jhMikeS summons john conway...who iirc did alot of research on knotting...or was that someone else? :\
07:24:36lostlogicthere's a guy who has a whole website dedicated to tying and lacing shoes
07:24:38hcslostlogic: success!
07:24:43lostlogichcs: :)
07:25:07lostlogichcs: that should make everything but ADX work
07:25:16hcslostlogic: yeah, just tested on ADX
07:25:16lostlogicand your work on ADX to make it use incremental reads should solve that
07:25:30jhMikeSlostlogic: I think we need a theoretical mathematician for this one :)
07:25:57hcsjhMikeS: do you really want to know exactly how impossible it is?
07:26:03lostlogicjhMikeS: sigh, I'll never be that... like I mentioned, I code by feel ;)
07:26:43*jhMikeS codes by, well, has no idea and feels numb after it
07:26:57hcsjhMikeS: godshatter
07:27:02lostlogic*giggles like a stoned school girl*
07:27:26hcsso, ADX for tomorrow, NSF and SPC seem to be working nice now, adios
07:27:50*jhMikeS totally doesn't understand hcs :?
07:28:52lostlogicscorche: hey, you've seen me _live_ late at night.
07:28:56lostlogicthat sounds bad.
07:29:20scorcheyeah...i also saw you sleep all night while we were busy chatting around you
07:29:31lostlogicahahahahaha, alcohol... it was the alcohol
07:29:54scorcheyou were just the dead person who was exactly in the middle of our oval
07:30:02hcsjhMikeS: godshatter - a term from A Fire Upon the Deep, for programs left in the minds of lower lifeforms by "gods", completely incomprehensible but possessing some of their power
07:30:07lostlogicyeah, I remember waking up and mumbling a couple of times
07:30:25lostlogichcs: I need to reread that book...
07:30:47hcsI'm on A Deepness in the Sky now
07:30:59lostlogicah, I've read almost all of his work
07:31:03lostlogicgot to chat with him at MarCon one year
07:31:06 Join webguest48 [0] (i=4ca81f01@gateway/web/cgi-irc/
07:31:22jhMikeSthe gods must be crazy then
07:31:27lostlogicso true
07:31:31Mouser_XGreat movie.
07:31:44Mouser_X(The Gods Must be Crazy, that is.)
07:31:52lostlogichas anyone else updated to my latest commit other than me and hcs?
07:32:19lostlogicI'd like to get 1 or 2 more "it works, I mean ti plays and the disk both spins and stops in the expected manner" reports before I pass out
07:32:22jhMikeSno...I'm scared :)
07:32:26webguest48Hello everyone. i have rockbox working well and everything. i know that this isnt a ipodlinux thing but if anyone has that on their ipod can they please help me?
07:32:42lostlogicjhMikeS: you can't see them (because I'm weak) but I'm shooting hate rays out of my eyes at you
07:32:50scorchewebguest48: ipl has their own channel
07:33:03webguest48may i have the link?
07:33:16*jhMikeS will saving some heating costs this winter :)
07:33:21scorchewell, you are using our web client which is restricted to #rockbox
07:33:48webguest48so put in #ipodlinux ?
07:33:52Mouser_Xlostlogic: I'd update, but I should be going to bed.
07:34:07lostlogicMouser_X: update so _I_ can go to bed!!! :)
07:34:11jhMikeSok, updated and building...
07:34:12scorchewebguest48: join that channel using
07:34:17Mouser_X(Also, I don't want to mess up my current playlist...)
07:34:32webguest48thank you very much scorche
07:34:48scorchesorry... .net
07:34:49jhMikeSzippity do da...
07:35:47webguest48oh ok thank you
07:35:58jhMikeSextractity do da...
07:36:09 Quit J3TC- (".•«UPP»•.")
07:36:20webguest48oh ok thanks. i wondered why i needed authentication lol
07:37:13webguest48do i put in #ipodlinux? im sorry for all these questions
07:37:22jhMikeSlostlogic: well, it's playing. any stunts to perform?
07:37:48lostlogicjhMikeS: if possible, skip to the last buffered track (you'll know when the next is empty, because of another bug) and let that one play out
07:37:50scorchewebguest48: isnt there a spac for a channel?
07:38:02lostlogicif that works and the disk boths pins up and spins down for the buffer run then I'm happy for now
07:38:20jhMikeSnext unbuffered? that would be next folder for me
07:38:37lostlogicjhMikeS: ooh, a complicated user :)
07:39:01jhMikeSit played fine.
07:39:27lostlogicjhMikeS: cool.
07:39:28jhMikeSis playing anyway
07:39:29webguest48oh this gigantic list of channels came up
07:39:36lostlogicscorche: update. now.
07:40:02lostlogicyou heard me, update your player to latest svn and tell me what I've broken.
07:40:05jhMikeSlostlogic: are aware I was getting boxes in the WPS after the current track was done decoding?
07:40:46jhMikeSthe "default char" for every line. happened with SPC on every change. let's see in a moment.
07:41:28*scorche ignores lostlogic and continues trying to find what to start transferring to the new drive first
07:42:32*lostlogic kicks scorche in the teeth
07:42:34jhMikeSyep, still happens a few seconds before the end of the SPC until the new one starts
07:43:01lostlogicnot my problem for now :-D
07:43:02jhMikeSlooks like the metadata just disappears for a second or two
07:44:09lostlogicjhMikeS: yeah, I've seen some metadata wierdness since MoB, I'm hoping Nico gets to it because that's outside of my alleyway
07:45:20lostlogic(my ipod lives on a counter or in my pocket, what do I care if the metadata is wrong spuriously)
07:46:17jhMikeSmmmm...cause it looks flakey?
07:46:36lostlogicjhMikeS: :)
07:48:39lostlogicI'm not saying I'd never get to it, but the idea of having playback tell the buffer when it's going to want the next track soon has my interest at the moment... not sure if it'll always be practical but still interesting
07:49:12hcslostlogic: uh, MP3s don't work
07:49:19hcsfreeze like ADX
07:49:22lostlogichcs: like at all?
07:49:28lostlogichcs: at buffer end?
07:49:45lostlogichcs: you're dreaming, go back to sleep :-D
07:50:42hcssome are working, others are not, it seems
07:50:57lostlogicdoes it have to do with files or buffer position of files?
07:51:18jhMikeSMP3 at least starts fine here
07:51:47*jhMikeS doesn't use playlists and uses directory changing
07:51:49hcsfile size, maybe?
07:51:56lostlogicI've skipped + listened to several buffers-full of mp3 and ogg w/o trouble on my ipv
07:52:04lostlogichcs: maybe, but unlikely imo
07:52:20hcsI've got an 8MB MP3 that works as 2nd in a playlist but not first
07:52:37lostlogichcs: you're fired.
07:52:42lostlogicthat is the weirdest bug I've ever heard.
07:53:07Mouser_XYay! More fun stuff! :(
07:53:15*jhMikeS would tend to suspect playlists
07:53:42lostlogichcs: this mp3 that breaks, won't even start if it's first in playlist?
07:53:59hcslostlogic: it also works if I play only it, but if I play the whole directory it gets stuck
07:54:41lostlogichcs: beginning or middle of file?
07:55:06hcslostlogic: beginning, same sort of freeze as SPC had and ADX still has
07:55:21lostlogichcs: I can only conclude that it hates you
07:55:29jhMikeSdon't cross the street in the middle in the middle in the middle in the middle in the middle of the block :P
07:55:37lostlogicmidnight is my limit (I know it's 3 for you, but whatever)
07:55:47lostlogichcs: I'll revisit the mp3 with you logfing (if you're up for it) tomorrow.
07:56:10jhMikeSis logfing to point at whoever caused the bug?
07:56:25hcsI think the problem is that I'm trying to play Metallica.
07:56:40hcsLars is probably hating my Sansa to pieces.
07:56:51jhMikeSthe bits are distored then
07:56:59lostlogicjhMikeS: no, it's no fun to point fingers at two people (me and nico)
07:57:14hcsThe MP3 that should not be.
07:57:33jhMikeSone logfing per core to do it twice as fast :)
07:58:00lostlogic2:05 until I turn into a pumpkin
07:58:10hcsalright, to put an end to this I'm getting off irc so I can put myself to sleep
07:58:14 Quit hcs ("Leaving.")
07:58:40lostlogicsigh, I am not happy about that delta on my last commit. Will endeavor to shrink tomorrow
07:59:00jhMikeSnighty night
07:59:24 Quit Mouser_X ("CGI:IRC 0.5.9 (2006/06/06)")
08:03:45 Join ameeen [0] (
08:05:06ameeeni installed .rockbox and now the songs that i transferred thru itunes, i can access them but they are shown as "untagged" without the names, though they are playing fine, just to mention: I have themes,fonts enabled
08:05:55scorcheyou need to tag your files properly then
08:07:05ameeenscorche what's taging ..
08:08:00ameeeni cant find anything related to tagging here :
08:10:19 Join homielowe [0] (
08:20:57 Join Rob222241 [0] (
08:24:59lespeaameeen: in itunes, do all of the columns have entires?
08:25:05lespeaameeen: does that make sense even ?
08:25:28ameeenlespea let me check
08:25:49ameeenlespea no, only the tracks are present with Name
08:26:35lespeathat is your main problem, your music is (sorry to say) Extremely disorganized
08:26:55lespeatags are everything, filenames don't mean much
08:26:59ameeenlespea does that mean I need to edit each and every track ? to edit the artist and album columns ?
08:27:09lespeaameeen: pretty much yeah
08:27:18lespeaameeen: it sucks, but once it's done you're golden
08:27:50lespeaameeen: i have ~8000 songs all tagged correctly −− the effort is way worth it
08:28:12ameeenlespea can you please tell me which columns shall I edit (enter the information) for .rockbox
08:29:02ameeenartist and album would be good enough yeah ?
08:29:37 Quit karashata ("Leaving.")
08:29:52ameeenand is there any other software .. to edit a huge list of songs (i mean to tag them), in a lot less-time-consuming way ?
08:31:06micolsno lespea the effort is not worth it
08:31:15micolsi have ca. 22k songs
08:31:47lespeaameeen: what os are you using?
08:32:09micolslinux has some smart utils for this
08:32:11lespeamicols: to each their own... if you like chaos then w/e
08:32:22micolsi once wrote my own id3 tag editor in visual basic for winxp
08:32:32micolsit did edit the tags of ca. 10k songs finely
08:32:37lespeaameeen: i recommend either "the godfather" or mp3tag
08:32:52lespeamicols: visuall basic == suck
08:33:10lespeamicols: why re-invent the wheel?
08:33:12ameeenlespea which one is best ?
08:33:37lespeaameeen: i like the godfather, but mp3tag is more user-friendly
08:33:50scorchethere are many ways to tag...manually, automatically based on file/folder structure, automaticallt based on acoustic fingerprint, etc
08:33:54lespeaameeen: godfather is harder to find kinda
08:34:21lespeascorche: yeah "acoustic fingerprinting" sucks from my experience
08:34:52lespeascorche: indeed, thanks couldn't remember prog lol... not too impressive
08:35:00lespeascorche: but i'm a tag freak so....
08:35:10scorchewell, this is getting a bit offtopic anyway...
08:35:18lespeascorche: #1 default action is ... remove all tags and rebuild
08:35:19ameeenfor example my track names are like: 099 - Peter Bjorn & John ft. V. Bergsman – Young Folks - [Torrent-Tatty] (™ CA -RED i-Tunes @128) ... which method shall I go for, mp3tag ? lespea, scorche
08:35:37lespeaameeen: oh good god
08:35:43lespeaameeen: good luck ???
08:36:08scorcheameeen: kill 2 birds with one stone by organizing you music into artist/album/track, and then tag all songs based on that
08:36:10ameeenlespea not all the songs are like that,, few of them
08:36:11lespeaameeen: the godfather has very good parsing... but ... yeah... filenames are pretty bad
08:36:40ameeenmp3tag wudn't help ?
08:36:41lespeaameeen: scorche has a very good point.. if you want to manually organize like that.. the godfather will own
08:37:11ameeenso shal lI go for godfather instead of mp3tag ?
08:37:45lespeaameeen: i'd say yes
08:37:45*amiconn just found that the PP5002 also allows readings its exact cpu name
08:37:49lespeaameeen: have you found it?
08:37:59ameeen ????
08:38:10amiconnAnd the iPod 2nd gen firmware also knows about a PP5001...
08:38:39 Quit ompaul (Read error: 113 (No route to host))
08:39:30 Quit Rob2222 (Read error: 110 (Connection timed out))
08:39:34lespeaameeen: yeah that'll work... though i think the original site is dead???
08:39:49lespeaameeen: if that doesn't work... i have a zip somewheres i can send to you
08:40:17ameeenlespea but that's godfather 0.70 .. is this the same one, you got ?
08:40:48lespeaameeen: no that's newer... haven't upgraded in a while... to be honest with each new release more features were killed :(
08:40:56lespeaameeen: perhaps they got the best of him :(
08:41:29lespeaameeen: i lied... i have .70 lol
08:41:40ameeenlespea it's opened and Im scared from the scary interface as still I cant figure out what exactly to do
08:41:54lespeaameeen: oh god...
08:41:57lespeaameeen: lol
08:42:07lespeaameeen: i don't htink you want my drunk ass explaining that to you
08:42:19lespeaameeen: hmm... okay... huh... lol :(
08:42:33ameeenlespea would appreciate it though
08:42:36lespeaameeen: does none of it make sense?
08:42:44ameeenlespea most of it does
08:43:17lespeaameeen: i cant' exactly fire it up (running linux atm) and i haven't had to do anything crazy (like you're trying to do) so i haven't used it in a bit
08:43:43lespeaokay... well... i would folwo scorche's advice and manually organize it
08:43:53lespeathen you can go in the folders and mass-tag things
08:44:13lespeabecause honestly... it sounds like hte filenames are FUBAR
08:44:19ameeeni am in the folder and all the list of songs are on main window of the software
08:44:48lespeahow many songs do you have again?
08:45:10ameeen099 - Peter Bjorn & John ft. V. Bergsman – Young Folks - [Torrent-Tatty] (™ CA -RED i-Tunes @128)
08:45:49lespeais your music just an incoherent mess... like is every song by a diff artist really?
08:46:15ameeenthe folder I am present in, atm, all songs are like that
08:46:32lespeaand they have nothing in common?
08:47:32ameeen001, 002, 003
08:47:50 Join Shaid` [0] (
08:47:59ameeen[Torrent Tatty] is also common
08:48:22lespeaokay, you have a LOT of work to do... or none
08:48:23ameeenlike song name is starting with: 001 or 002 or 068 or 054
08:48:55lespeado you htink you'd benefit from knowing the artist/song name/album/etc or is everything so random it wouln't make much differnece?
08:50:02 Join bertrik [0] (
08:51:29ameeenlespea did u get the notice
08:52:08lespeaameeen: no
08:53:44 Join Buschel [0] (
08:53:44 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
08:55:59 Quit Ebert ()
09:03:03 Quit ameeen (Nick collision from services.)
09:03:10 Join amin [0] (
09:03:28lespea:o wtf
09:04:09 Nick amin is now known as ameeen (
09:05:07 Quit Shaid (Read error: 101 (Network is unreachable))
09:05:07 Nick Shaid` is now known as Shaid (
09:10:04 Quit webguest48 ("CGI:IRC (EOF)")
09:12:49 Join desowin [0] (
09:14:04 Quit TotallyInfected ()
09:14:09***Saving seen data "./dancer.seen"
09:21:25 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:29:38 Quit lespea (Read error: 104 (Connection reset by peer))
09:29:43 Quit ameeen ()
09:56:12amiconnMy 1st and 2nd gen are both PP5002E0. I'm curious about the 3rd gen...
10:02:27 Quit ramon8 (Read error: 104 (Connection reset by peer))
10:03:15 Join Siku [0] (
10:03:16 Nick fxb__ is now known as fxb (
10:03:21Buschelamiconn: good morning :o)
10:04:29 Quit desowin ("use linux")
10:12:29 Join barrywardell [0] (
10:21:55 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
10:27:32 Join davina [0] (
10:28:46 Join ameeen [0] (
10:32:49ameeenwhen I go in Database > Artist > artistname > (all tracks) > track.mp3 .. and then return to back screen it takes me directly to (all tracks) >artistname that I selected (the same artist name that I am playing) but does not happen when I go through Tracks > trackname.mp3 and when I push back it takes me to the top of the list of tracknames but not the trackname.mp3 that I selected.. how to solve this issue
10:33:26 Join Thundercloud [0] (
10:38:13crashd__ahoy thundercloud
10:38:16 Nick crashd__ is now known as crashd (
10:38:27bertrikameeen: I always end up at the top of the list
10:38:45Buschelamiconn: there is some further speed up of yuv-blit possible for ipod video
10:38:52bertrikI don't think that's a bug or an issue
10:40:44ameeenbertrik but what if I want to end up not on the top of the list but on the track that I just came from
10:41:02bertrikthat's a missing feature :P
10:41:34bertrikI don't know how it's supposed to be really
10:42:41ameeenwhere shall I extract
10:42:54ameeenin the .rockbox or .rockbox/<some folder>
10:46:31ameeeni am unable to charge the ipod (thru usb) while RockBox is playing music
10:46:43ameeenit takes me to the boot screen whenever i plug in
10:46:54linuxstbameeen: Hold the MENU button as you insert the USB cable
10:48:16ameeenlastebil yeah cool that worked. Thanks. But i cant see the icon of charging
10:48:21ameeenlinuxstb **
10:48:36 Join kubrick [0] (n=repulse@unaffiliated/funky)
10:52:20 Join spiorf [0] (
10:53:11 Quit hannesd ("Client suicide")
10:53:20 Join hannesd [0] (
10:54:49*amiconn wonders about Buschel
10:56:21ameeenomg I cant see my ipod as a MASS Usb device in My Computer :|
10:56:28ameeenit's plugged but not there
10:56:38*Buschel reaches 29.5fps yuv-blit at 80MHz full screen
10:56:48ameeenthough itunes can detect it :|
10:57:20ameeenamiconn can u pls help ..
10:57:53pixelmathere might be an option in itunes (think that's what I read but neither have itunes nor an Ipod...)
10:58:19linuxstbYes, you need to prevent itunes hiding the disk - click the "enable use as disk" option for the ipod.
10:58:24 Join JdGordon [0] (
10:58:45 Join merbanan [0] (n=banan@
10:58:45JdGordoneveing all
10:59:13ameeendone, thanks again
10:59:43ameeenshall i paste the rockdoom in F:\.rockbox
11:00:42bertrikameeen: rbutil can install the doom files almost automatically i think
11:02:39 Join Mantaar [0] (
11:04:23 Join Domonoky [0] (n=Domonoky@
11:07:24 Quit RaRe (Read error: 110 (Connection timed out))
11:12:28ameeenaddons=wads ?
11:13:59 Join ender` [0] (
11:14:10***Saving seen data "./dancer.seen"
11:15:03ameeenpixelma im there but im confused with the addons=wads ?
11:16:31linuxstbaddons are things you add onto wads.
11:19:02ameeenlinuxstb i downloaded smb_v102 and in that folder there is skymaybe.exe .. shall I copy the skymaybe.exe in f:\.rockbox\doom ?
11:21:25 Join lee-qid [0] (
11:21:39 Nick parafin|away is now known as parafin (
11:24:40JdGordon... and who said PP was dead?
11:24:52*JdGordon wants to get a sansa view now
11:25:53JdGordon$330 AUD :O
11:26:21JdGordon260 for the 8gb version
11:27:31sbeh<- sansa e200 4gb
11:30:02 Quit barrywardell ()
11:30:50JdGordonyeah, i have a 2gb e200 already... might get a view duty free when i go to america :p
11:30:52linuxstbameeen: I've never used addons, but I know Rockbox won't read a .exe file
11:34:16markun"rockbox: the gaming software that also plays music"
11:34:36*Buschel updated FS #8075, yuv-blit possible with 29.5fps @80MHz on ipod video (+8% vs. trunk)
11:34:53markunBuschel: great result!
11:35:20markunanything keeping it from being committed?
11:37:23Buschelask amiconn, he is the main worker in this area
11:39:05markunamiconn: 17 bits of course
11:41:48amiconnBuschel: Btw, you're right that the chroma buffer size is wrong in svn
11:42:54Buscheloops, typo. should of course be 565 :/
11:43:08amiconnHmm, extending the chroma buffer to 32bit means allocating quite a lot of stack... usually not a good idea on embedded systems. Almost 2KB here
11:44:01Buschelamiconn: yep, thats one thing where i needed your opinion. nevertheless it amkes using stm/ldm possible.
11:45:31 Join desowin [0] (
11:45:45 Join moos [0] (
11:45:54Buschelbtw, does anybode know what fps mpegplayer does reach at 320x240 without the yuv-conversion? so, pure decoding only?
11:47:21 Quit spiorf (Remote closed the connection)
11:48:38linuxstbI may be out of date, but I think it used to be around 21/22fps. But it would be easy to test.
11:49:27linuxstb(just comment out the calls to vo_draw_frame() in mpegplayer/mpegplayer.c)
11:50:06Buscheli wonder why the last optimization for lcd/yuv did not scale that much on high resolutions. afaic the mpeg-decoder runs on cop and mp3-decoding and lcd/yuv run on main cpu?`
11:50:23linuxstbNo, lcd/yuv is on the COP as well
11:50:33linuxstbaudio decoding and disk i/o are on the main CPU
11:51:24Buschelwould it make sense and is it possible to do lcd/yuv on main cpu?
11:51:56 Quit merbanan (Remote closed the connection)
11:51:57linuxstbThe problem is cache coherency - both cores have independent caches. So you would need to flush/invalidate caches.
11:52:16linuxstbI'm not sure what our current knowledge is about that - jhMikeS is probably the person to talk to.
11:52:42linuxstbBut I think it's definitely worth testing.
11:55:11linuxstbAnother problem is synchronising the yuv blits, and ensuring that the decoding of the next frame isn't overwriting the buffer containing the frame currently being rendered.
11:55:15amiconnI don't think that putting yuv on cpu would help, I rather think the opposite will happen
11:55:36amiconnThe cpu is already >50% loaded with just decoding mp3 audio
11:55:58linuxstbHow fast is just the yuv blit?
11:56:12*linuxstb checks wiki
11:56:32Buschelamiconn: which bitrate for mp3 is used?
11:56:32JdGordonI'm updating the yes/no screen for the mr500 so it has to big-ish buttons to press.. (a green tick and a red X), should I do it as a ifdef for certain screens? or for all bitmap to make it a bit nicer?
11:56:35linuxstbAh, only 27.,3fps...
11:56:51*linuxstb reverts his previous comments and agrees with amiconn
11:56:59amiconnBuschel: Whatever you want... can be vbr as well, but *must* be 44.1kHz
11:57:40linuxstbJdGordon: I tried the sim for your "music wardrobe" the other day, and the mouse didn't appear to do anything. Should it?
11:58:16 Join RaRe [0] (
11:58:17JdGordonamiconn: i finished the info list patch but imho its crap... the text really need to be done differently if its a 1 or two line list... so i tihnk we should just ifdef it for charcell instead of adding it
11:58:23linuxstbAlso, is there a reason you made Rockbox portrait, when (IIUC), the OF is landscape?
11:58:45JdGordonthe mouse isnt setup int hte sim yet, and thats the way its wired up apparentyl!
11:58:53JdGordonkkurbjun is working on making it landscape
11:59:15Buschelamiconn: but loworing the mp3-bitrate will have no effect on fps as all the yuv/mpeg-stuff is placed on another cpu...
11:59:29linuxstbJdGordon: I thought you did work on the mouse - is it not committed yet?
11:59:32amiconnJdGordon: Two-line is also useful on small bitmap screens, or when using large fonts
11:59:51JdGordonlinuxstb: not commited yet.. I dont really like the way i did in that patch
11:59:55amiconnBuschel: Yes, but putting yuv on the cpu instead of cop won't speed up things
12:00:05linuxstbJdGordon: OK ;) That would explain why the mouse didn't work...
12:00:11JdGordonamiconn: ok, but I think we should ifef those screens based on the display and not add the option
12:00:12Buschelwas just a suggestion...
12:00:34JdGordonwe have doom going.. so its about time to release isnt it :D
12:00:42JdGordons/we/karl :p
12:00:58amiconnJdGordon: I'd want two-line on my H300 when using it with my 'in-car' font
12:02:00JdGordonit'll bloat it a bit adding the code to do the correct text for the optino..
12:02:12JdGordonbut if your ok with a red delta then ill put it in...
12:02:57amiconnWhat option?
12:03:12JdGordonto display as a 1 or 2 line list
12:03:12 Quit atsea-39 (Remote closed the connection)
12:03:27JdGordonits not so easy to do it based ont he line count
12:03:33JdGordonwithout really hacking the drawing code
12:03:45JdGordonthe text formatter doesnt know what display its on
12:04:22amiconnI'd want the list to auto-switch.... but that might be difficult
12:04:30*JdGordon would actually like to get into the list code and redo it so it doesnt have the current multi screen wrapper code
12:04:57JdGordonthere is no reason for the list users to worry about the multiscreen api anymore
12:05:36JdGordonand calling the text callback twice every update for every line is a waste also
12:06:39JdGordon+ the interface is bloated imho
12:14:15ameeenwhich build of rockbox is this:
12:14:55linuxstbJdGordon: Do you mean you wouldn't use the screens api at all in the list code?
12:14:55ameeen2.3 or 2.5..
12:15:52linuxstbameeen: That's the latest "current build". I'm not sure what you mean by 2.3 or 2.5
12:16:03JdGordonlinuxstb: yeah, or as little as possible anyway.. currently there is the list for a display and every list function does FOR_ALL_SCREENS()... which isnt really needed
12:16:24JdGordonthe drawing is the only thing which cares about the different displays
12:16:43JdGordonamiconn: fs8080 is the patch if your interested
12:16:50ameeenlinuxstb so I must opt for Voice files for builds after 2007-08-06 ???
12:17:19JdGordonlinuxstb: there are very few struct members which change for the 2 displays
12:17:52linuxstbThen can't they just be factored out?
12:18:35 Join atsea-39 [0] (i=atsea-@gateway/tor/x-7b70e8ea490069de)
12:19:57linuxstbameeen: What page are you reading? Go to and click on the Daily builds and voices link for up to date voice files.
12:21:26ameeenlinuxstb did that .. now i have downloaded: "" and it asks me to save in .\rockbox\langs .. but in this folder there are files *.lng .. shall I save this "" in .rockbox\langs\ anyway ?
12:22:07linuxstbWhen you say "it asks me", what are you talking about?
12:23:02ameeenThis is the place to download voice files for making Rockbox' menus speak. Learn how to do this. These files should be placed on your player in the /.rockbox/langs folder.
12:23:28pixelmaalso the voice file you link to is already kind of old and most probably won't work correctly...
12:23:51ameeenpixelma then which voice file shall I download ..
12:24:16*linuxstb repeats - go to and click on the Daily builds and voices link for up to date voice files.
12:24:44pixelmaI guess you're on the VoiceFiles wiki page - it is stated there and linuxstb already answered that...
12:25:03pixelma"Starting at the 6th of August 2007, all voice files are target specific so you need to download a new voice file built for your particular target. We host daily built US English ones at - built with Festival on Linux."
12:25:18ameeenlinuxstb did that .. shall I save that voice file in .rockbox\langs ?
12:26:07ameeenyes pixelma Ive downloaded that file .. now where to save it ? in .rockbox\langs ?
12:26:31pixelmayes, but you need to rename it to english.voice
12:26:43Domonokyameeen: maybe try rbutil, it does all this things for you :-)
12:27:25ameeenDomonoky rbutil=?
12:28:41amiconnAh, 0x70000080 seems to be some kind of gpio, but unlike the already know gpios, it is 32 bits wide
12:28:56amiconnAnd it looks like hdd power is hooked up there on the ipods
12:30:07 Quit kclaf (Read error: 113 (No route to host))
12:30:30JdGordonhey pixelma, your a bit arty yeah?
12:31:06ameeenim confused .. I installed .rockbox from .. is this same as .. from this website:
12:31:12JdGordonif your bored, can you do a "yes" and a "no" image for the mrobe for me please? I tinhk 64x64 pixels each would be good
12:32:26pixelmaJdGordon: a bit, but I'm not bored - got a few things to do before leaving for the weekend
12:33:13JdGordonok :)
12:34:09pixelmaameeen: no, it's not the same. The first is the so-called "current build" which can change a few times a day, the other on is the "daily build" made once a day. The current build is usually preferred
12:34:28 Join MethoS- [0] (
12:34:40linuxstbameeen: The Installation chapter in the manual describes the different types of Rockbox builds
12:34:41pixelmabut under normal circumstances you can use the daily built voice file with it
12:35:28*pixelma thinks that the voice files should be provided in the correct directory structure and name (as everything else)
12:35:37pixelmathe daily ones of course
12:35:45 Quit MethoS- (Remote closed the connection)
12:36:21ameeenpixelma and shall I copy: { downloaded from } to f:\.rockbox\langs ?
12:37:36pixelmaalready answered that 10 minutes ago...
12:40:11pixelmaand by the way... it is also explained in the manual ;)
12:40:57amiconnThere's a universal gpio manipulation function in the pp502x ipod roms, which sets or resets a single bit in either 0x70000080, or the usual gpios, selected by a bit number (continuous range across all gpio ports), and an on/off boolean
12:42:20amiconnAnd that function also confirms that the known GPIOs can do atomic bit manipulation via their usual address + 0x800
12:43:12ameeenpixelma thanks. i found it in the manual
12:44:12 Quit Shaid ("I will see you, in the end. And I will laugh at your pain...")
12:45:40amiconnVery interesting stuff....
12:45:51*amiconn gets closer to controlling hdd power on the pp502x ipods :)
12:46:07Buschelamiconn: go, go, go!
12:46:56bertrikamiconn: I think I saw a patch on the tracker for a debug menu that allows easy setting/clearing of GPIO bits
12:47:59amiconnNot for the newly discovered gpio for sure...
12:48:21amiconnBtw, 0x70000080 seems to be involved in the G5 backlight brightness setup as well
12:48:22bertrikok, but maybe you can adapt it
12:48:38bertrikit's FS #8054
12:49:39amiconnIn fact I don't need that. I am analysing what the of does so that we can do the same in rockbox
12:50:09amiconnOf course that newly discovered gpio will go into the debug screen as well; I'm not sure how to name it though
12:50:21Dark_ApostropheI'm trying to compile RockBoxUtilityQt
12:50:23amiconnGPIO32 perhaps, because it's 32bits wide unlike the others
12:50:32Dark_ApostropheCan't seem to get qmake to do its part of it.
12:50:38Dark_ApostropheCan anyone help me out a bit?
12:55:55DomonokyDark_Apostrophe : on which system are you compiling ?
12:56:16Dark_ApostropheDebian Sid AMD64
12:56:45Domonokyoki.. and waht doesnt work ?
12:57:03Dark_ApostropheWell, I run qmake-qt4 and it just outputs the help page
12:57:15Dark_ApostropheI tried adding options like qmake-qt4 -makefile -unix
12:57:20Dark_ApostropheDidn't help
12:57:23Domonokyand you are in the rbutil dir ?
12:57:40Dark_ApostropheI've compiled Qt apps before - I use KDE ;)
12:57:45Domonokytry qmake
12:58:10Dark_ApostropheNo such file
12:58:31Domonokythen you arent in the right dir.. there should be a .pro file ... :-)
12:58:45Dark_ApostropheI downloaded rbutilqt-v1.0.2, tar -jxf-ed it and cd -ed into it
12:58:45JdGordonlinuxstb: ok, just checked in muse support
12:59:27Domonokythere is a tar ball of rbutil ? please try svn ..
12:59:40 Join MethoS- [0] (
13:00:11 Quit ompaul (Remote closed the connection)
13:01:31DomonokyDark_Apostrophe: you downloaded the tar from the rbutil wiki page ?
13:01:51Dark_Apostropheah, in the SVN, I see a .pro file
13:01:52Domonokythats a binary :-) no need to compile ..
13:02:08*Dark_Apostrophe slaps self
13:02:19Dark_ApostropheI blame it on the fact that I just woke up ;)
13:07:50 Join MethoS-- [0] (
13:07:54amiconnHmm, looks like hdd power (on mini G2) is on plain GPIO J bit 2 :)
13:09:54 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
13:10:07pondlifeJdGordon: You know about data0?
13:10:30JdGordonyeah, tookl me a bit to realise what the error was
13:10:33JdGordonfix commited
13:14:14***Saving seen data "./dancer.seen"
13:16:55pondlifeJdGordon: Am I correct to think that your wiki-attached patch is needed to test the touchpad sim? Or is that in SVN now?
13:17:14 Join spiorf [0] (
13:17:34 Quit Mantaar (Read error: 110 (Connection timed out))
13:18:37JdGordonpondlife: the lists all work in the sim in svn.. the other screens havnt been commited yet
13:18:51*JdGordon was too lazy to update that page properly
13:18:56pondlifeUrgh, my sim won't run
13:19:42pondlifeEndless recursion through battery_status_update(), battery_level () and send_battery_level_event().
13:19:54pondlifeInstant stack o/flow
13:19:55JdGordonah, so not my changes :)
13:20:29pondlifeNope, but a lovely backtrace - about 30000 items deep
13:21:17pondlifeMake that 85000 entries and counting
13:21:33pondlifeAh 97720
13:22:31 Part pixelma
13:23:31*pondlife makes clean and rebuilds
13:26:10 Quit MethoS-- (Network is unreachable)
13:28:09 Join stripwax [0] (
13:30:21 Quit MethoS- (Read error: 110 (Connection timed out))
13:31:41 Join stripwax_ [0] (
13:33:25amiconnHehe. Rockbox trying to read from the hdd while the hdd is powered off makes it freeze
13:36:51*stripwax_ looks at all the ipod5g lcd speedups!
13:37:18stripwax_101fps?! WOW
13:39:57bertrikargh, the sim is borken indeed, I was just working on removing some #ifndef SIMULATOR
13:40:05ameeenfrom where I can get the latest themes/icons for rockbox along with wallpapers if any
13:41:02stripwax_ameeen - here?
13:41:39 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
13:42:05 Nick jumpatrain is now known as tictoc (i=tabac@gateway/gpg-tor/key-0xB9002659)
13:42:55ameeenstripwax thanks
13:44:46amiconnstripwax: Nice, eh? :)
13:44:46 Quit bertrik ("bye")
13:45:25 Nick tictoc is now known as jumpatrain (i=tabac@gateway/gpg-tor/key-0xB9002659)
13:46:30 Quit Nico_P (Remote closed the connection)
13:46:51stripwax_amiconn - definitely! was the trick just to do the updates on a tick task? (I say 'just' but..)
13:46:58 Join DerPapst [0] (
13:47:12DerPapstamiconn: one of my 3Gs is a pp5002D
13:47:54amiconnstripwax: That, and using the same update rectangle all the time, plus the finding that we can write to the BCM's internal buffer while an update is running
13:48:05amiconnDerPapst: Interesting...
13:48:27DerPapsthaven't checked the other one yet. has to charge a bit before installing a current build.
13:48:36amiconnI would think that 'D' is older than 'E0' - but that doesn't match the generation order...
13:48:38stripwax_amiconn - great! I'll take a look at the old irc logs and see how it happened :)
13:49:02 Quit stripwax (Read error: 110 (Connection timed out))
13:51:17 Join perrikwp_ [0] (n=chatzill@
13:51:45 Nick fxb is now known as fxb__ (
13:52:45amiconnHmm. On G5, hdd power is adifferent pin than on mini G2 :\
13:53:03ameeenI have a Black 30G Video ipod.. And I have scratches all over (from front and back) can someone tell me a way to get them off
13:54:02BigBambiameeen: Stay on topic please
13:55:49stripwax_amiconn - awesome work, making me want to donate again!
13:56:32 Join tuffins [0] (
13:56:32 Nick stripwax_ is now known as stripwax (
14:07:02 Quit Domonoky (Remote closed the connection)
14:07:13ameeenis there a difference between reboots #1 (holding Play/Pause) #2 (Holding Select Button and Menu), in the ipod30Gb-video, I have an IpodLinuxLoader that has boot options for Apple, Rockbox, Disk mode, Sleep
14:09:20 Quit perrikwp (Read error: 113 (No route to host))
14:10:16lostlogicyawn, how's buffering?
14:10:37stripwaxameeen - well yes. holding play/pause shuts down rockbox. holding select+menu doesn't shutdown, it's a hard reboot, you lose any settings you've changed, etc
14:11:03stripwaxany questions on ipodlinuxloader probably belong in an ipodlinux channel though, not here
14:11:17ameeenstripwax thanks, appreciate it
14:13:22 Nick animeloe__ is now known as animeloe (n=animeloe@unaffiliated/animeloe)
14:14:54tuffinswould anyone be as kind as to tell me how to install these bloody games on my daughters ipod 5.5g
14:16:06stripwaxtuffins - huh?
14:16:29stripwaxis that a request for help installing Rockbox on an ipod, or something else?
14:16:48sd__tuffins: yes, rockbox has games
14:16:54sd__if you have rockbox, that is
14:20:25tuffinsno sorry me was on about ormal games but guess this wrong place sorry
14:20:33stripwaxtuffins - "normal games"?
14:21:08stripwaxif you want to install Apple's games on her ipod, you use iTunes (nothing to do with rockbox)
14:22:10tuffinsyeah the hacked one but me suffering
14:22:24tuffinsso what is rock box
14:22:38pondlifetuffins: will tell you all
14:25:53pondlifejhMikeS: ping
14:25:56stripwaxtuffins - rockbox completely replaces the Apple firmware on your (daughters) ipod. It changes everything (more than just "hacked".) If you don't know what that means, you probably don't want to do it.
14:26:30pondlifejhMikeS: queue_broadcast() doesn't seem to work on the sim, is that to be expected?
14:26:52pondlifelostlogic: You dream algorithms too?
14:27:12lostlogicpondlife: more like i dream data flow diagrams :-\
14:27:21 Quit ameeen ()
14:27:23pondlifehaha, thought it was just me
14:27:29stripwaxBut then I don't know what you mean by "hacked" or "normal games" so I can't help
14:27:47tuffinsyeah got good idea firmwared me psp and unbricked it
14:27:58tuffinsbutr just cant get ipod games to install
14:28:31sd__tuffins: ... and you call that _bloody_? :)
14:28:56stripwaxtuffins - what do you mean by "ipod games" and "install"? If you have already installed Rockbox, then you already have the games.. And if you haven't installed Rockbox, then you use iTunes to install the games.
14:32:53 Join zicho [0] (
14:34:04 Quit TMM (Read error: 110 (Connection timed out))
14:35:55 Quit lee-qid ("aufwiederbyebientotsayonara")
14:37:02 Join obo [0] (n=obo@rockbox/developer/obo)
14:37:20tuplanollahello. what kind of videos does rockbox play?
14:37:30tuplanollai know it's mpg, but i need more details
14:39:54 Join perrikwp__ [0] (n=chatzill@
14:39:57 Nick perrikwp__ is now known as perrikwp (n=chatzill@
14:40:47tuplanollaa, thanks
14:42:37 Quit japc (Read error: 104 (Connection reset by peer))
14:49:34 Join merbanan [0] (n=banan@
14:49:44lostlogicany sansa (broken) buffering (broken) testers (broken) ... wait... anyone experiencing 8077 want to test a patch?
14:49:52 Quit JdGordon (Remote closed the connection)
14:51:22lostlogicThe race on Sansa is between the codec asking for data and the buffering thread filling on its own, I think and if I'm right I have a fix that occurred while I was partially/entirely sleeping
14:52:21pondlifelostlogic: Aside from any races, the test for ata_disk_is_active() really means "is there a penalty for disk access", so should always allow buffering on flash targets, right?
14:52:45lostlogicpondlife: I see you fixed a recursion stack overflow... how deep? I put in a recursion up-to 256 levels deep in bufering yesterday and I hope it doesn't break as handles increases
14:52:46 Join Gnostic [0] (
14:52:58lostlogicpondlife: yeah, but that's an optimization _only_
14:53:05pondlifeI got about 94000 deep...
14:53:14lostlogicpondlife: good, I think I'm safe for now.
14:53:24lostlogicmy math said 300 was safe on that thread
14:53:30lostlogicso I set it to a power of 2 < 300
14:53:42pondlifeOn sim, this was. I doubt it'd get so far on target
14:53:53lostlogichehe yeah
14:54:05 Join J3TC- [0] (
14:54:27lostlogicit's size_t per level + variables pushed right?
14:54:40lostlogicI should read the asm.
14:54:49 Quit tuffins ()
14:57:40*pondlife continues to be annoyed by the yellow and wonders if those gcc 4.2.x clients could be not handed sim builds...
14:59:35 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
15:00:49 Quit Echelon (Read error: 113 (No route to host))
15:02:35 Quit perrikwp_ (Read error: 110 (Connection timed out))
15:02:48 Part pondlife ("Gone")
15:04:27 Quit Gnostic ()
15:06:32 Join stripwax_ [0] (
15:08:24 Quit DerPapst ("So Long And Thanks For All The Fish!")
15:09:25 Join DerPapst [0] (
15:14:15***Saving seen data "./dancer.seen"
15:14:55 Join bertrik [0] (
15:18:30 Quit nanok (Read error: 110 (Connection timed out))
15:21:11 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
15:23:49 Quit stripwax (Read error: 110 (Connection timed out))
15:24:17 Join stripwax [0] (
15:25:05Buschelanyone has a good fast method to mask x = x & 0xffff in arm-asm? is there any possibility to not need to use an additional register which contains 0xffff?
15:25:12 Quit zicho (Read error: 110 (Connection timed out))
15:29:36lostlogicBuschel: shift left 16, right 16
15:30:17 Join Frazz [0] (
15:30:31Buschelhmm, this needs 2 commands. so, no way to do it via 1 cycle?
15:30:44lostlogicnothing comes to mind
15:31:12lostlogicNico_P: I know I broke it so it doesn't fill cmpletely a lot of the time, I'm working on it.
15:32:09Mouser_XDang. It looks like there's still a lot of work to be done on the MoB stuff...
15:33:21sd__Buschel: maybe via mul & shr option
15:33:23linuxstbBuschel: Depending how you are using x subsequently, you may be able to combine the second shift with another instruction - e.g. add r1, r2, r3 lsr #16 (where r3 is x << #16)
15:33:26sd__but not sure it would be one cyce
15:33:57lostlogicMouser_X: naw, it's pretty good now, just tweaking
15:35:06 Join ameeen [0] (
15:36:20Mouser_XWell, the build I have (it's about a week old) is working, so that's good. And I still have all my previous builds (before MoB commit), so if need be, I'm good there as well.
15:36:38lostlogicneed testers on current build.
15:36:46lostlogiccan't fix otherwise.
15:38:09Mouser_XThat's true. I guess I *could* aid in testing (though, I don't have a build environment right now), but once I start a playlist, I usually like to see it through.
15:38:20*Mouser_X is going through his 1500+ OCR stuff.
15:38:33Mouser_XI'm in the K or L area.
15:40:14linuxstbYou should just be able to backup your .rockbox folder (e.g. rename it to .rockbox-safe), test a current build, then restore, with playlists and settings intact.
15:40:50Mouser_XD'oh! I thought of that last night, but had already forgotten by morning...
15:42:17 Join mf0102 [0] (n=michi@
15:43:12 Quit stripwax_ (Read error: 110 (Connection timed out))
15:45:19*Mouser_X has a Gigabeat. Thus, he'd only be able to verify that any Sansa fixes didn't break anything else...
15:46:12lostlogicMouser_X: the sansa only has made apparent bugs, the bugs existed on other platforms, were just easy to trigger on sansa
15:46:53lostlogicI just need people other than me to run the code and tell me how it changes between builds because my usage isn't going to catch all things
15:50:37 Join webguest51 [0] (i=53c37844@gateway/web/cgi-irc/
15:51:21lostlogicspeex has a bunch of logf warnings.
15:51:29lostlogic%d used when %ld should be
15:52:11lostlogicbertrik: <3 I'm working on it... getting closer apparently
15:52:28bertrikah, ok, thanks in advance
15:53:33lostlogicwebguest51: DRY (Don't Repeat Yourself)
15:53:42DerPapsts/o/i :P
15:54:53webguest51Is there a wiki page dedicated to the 5G hardware (registers and other)?
15:55:10lostlogicshould be
15:55:17Nico_Plostlogic: apparently we shouldn't use ata_disk_is_active
15:55:23Nico_Pand use the ata idle callback instead
15:55:40lostlogicNico_P: yeah, whatever, it's just an optimization, I'm more interested in fixing things regardless of the optimization
15:55:44webguest51sorry, the refresh is pretty slow on my computer...
15:56:22Nico_Plostlogic: yeah, but amiconn was pretty adamant ata_disk_is_active was broken on sansa
15:56:40Nico_Pit always return false I know, but maybe other things like that are borken
15:56:50lostlogicNico_P: sure, but it always returning false isn't particularly a concern to us, buffering should work fine
15:56:56DerPapstwebguest51: i'm not too sure if there is such a wiki page. most of the information is gathered from and the other information is in rockbox' source code in comments
15:59:45webguest51I am looking in the index but cannot find any ipod hardware related info
16:00:10lostlogicugh, now it freezes if I skip off the end of the buffer but works fine if it runs down.
16:01:20Mouser_Xlostlogic: If I were to test your builds (I can't compile patches right now), how would I do so? Also, how useful/helpful would that be?
16:01:26*Mouser_X is getting his Gigabeat.
16:02:15Bagdermi4code update
16:02:31webguest51ok, i have seen the ipolinux wiki
16:03:29webguest51but I can't find some info about the broadcom register (except like you said, in the .C)
16:03:42 Quit jhMikeS (Nick collision from services.)
16:03:48 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
16:04:04lostlogicMouser_X: for now I really just need you to run latest SVN and tell me if there's anything weird going on other than it buffering only one track at a time in some cases
16:04:48Nico_Plostlogic: I'll be helping ASAP but right now I have HW problems to solve on a computer
16:04:59lostlogicNico_P: no probs, I'm having fun playing :-P
16:05:13lostlogicNico_P: any comments about the stuff I committed last night are welcome though
16:05:28Nico_PI still need to read the diffs closely
16:06:39webguest51it would be nice to gather all the informations that has been found via reverse engineering in a rockbox wiki page (PP50xx and broadcom)
16:10:20 Quit J3TC- (".•«UPP»•.")
16:10:55 Join MajorC [0] (
16:11:01BigBambiwebguest51: Feel free (it is a wiki after all)
16:11:40 Quit webguest51 ("CGI:IRC (Ping timeout)")
16:11:42ameeensince Ive installed rockbox ive noticed that my battery is consumed more as compared to applefirmware .. do need any particular settings to save my battery ..
16:12:13 Join kclaf [0] (i=kclaf@
16:12:15BigBambiameeen: Try reading the ipod FAQ on he wiki
16:12:25BigBambiand search the forum
16:12:30BigBambiThis has been asked o death
16:12:54ameeenBigBambi the ipod FAQ or the rockbox FAQ ?
16:12:59 Join webguest51 [0] (i=53c37844@gateway/web/cgi-irc/
16:13:12BigBambi<BigBambi> ameeen: Try reading the ipod FAQ
16:15:08 Quit webguest51 (Client Quit)
16:16:02 Join bertrik_ [0] (
16:21:52 Quit Frazz (Read error: 104 (Connection reset by peer))
16:22:36 Join Echelon [0] (
16:29:24 Quit bertrik (Read error: 110 (Connection timed out))
16:29:49 Join toffe82 [0] (n=chatzill@
16:29:52 Nick bertrik_ is now known as bertrik (
16:35:42 Join J3TC- [0] (
16:36:07 Part toffe82
16:41:29puzzlesmy 5.5 just came \o/
16:43:26DerPapsthappy rockboxing ;)
16:43:32 Join toffe82 [0] (
16:44:55puzzleswell, they didn't include any sample clips, so i won't miss anything
16:48:35lostlogicok, so great thing about the ipod video screen is that it can logfdisplay two columsn wide on a readable font
16:48:40lostlogicZOMG LOGFUPTHEWAZOO
16:51:11 Quit jhMikeS (Nick collision from services.)
16:51:17 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
16:55:03lostlogic48 lines of logf is finally enough to see the sequence of events that i'm trying to track :-D
16:57:44 Join zicho [0] (
16:58:49 Join Gnostic [0] (
17:03:26 Quit Gnostic (Client Quit)
17:05:38 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
17:05:47Dark_ApostropheI just compiled RockBoxUtilityQt, and I have a question:
17:06:03Dark_ApostropheSo far, I haven't touched the GigabeatF20's bootloader
17:06:17Dark_ApostropheI've just updated rockbox, and left the original bootloader intact
17:06:50Dark_ApostropheIs it all with Rockbox Utility wise? As in, if I do it, is there a more likely than not chance of brickage?
17:07:49Mouser_XWorst case scenario (that I've noticed) with the Gigabeat, is that if you brick it (which is quite unlikely from what it looks like, based on what you said), it's still fixable.
17:08:38Mouser_XIt just requires opening up the DAP, and unplugging the HDD. Once unplugged, plug turn it on and plug it into a USB plug. There's a wiki about this.
17:09:00Mouser_X*uplugged, turn the Gigabeat on, and plug it into a USB port.
17:12:22 Join GnosticE [0] (
17:12:40krazykitright. the only way to truly brick the gigabeat is to mess with the Updater. as long as you don't do that, it's recoverable.
17:13:17krazykitand by "Updater" I mean Toshiba's updater, the hacked one somewhere on the boards.
17:13:39 Quit perrikwp (Read error: 110 (Connection timed out))
17:14:16***Saving seen data "./dancer.seen"
17:15:15toffe82krazykit: ther is no updater for the gigabeat F only for the S
17:16:19krazykittoffe82, that isn't true:
17:16:56krazykitperhaps i was unclear in my wording, though. not so much like "updater tool" as "update files that you put in a folder and it updates itself"
17:17:19toffe82krazykit: you are right , I didn't remember this one ;)
17:19:18 Part GnosticE
17:19:20 Join Arathis [0] (
17:22:33 Quit billytwowilly (Remote closed the connection)
17:24:07 Join Gourgas [0] (
17:24:50GourgasI was wondering if there is any way to make fm transmitters other than the Itrip work with rockbox on an ipod mini?
17:25:07GourgasThe ones that connect to the bottom port of the ipod
17:26:28toffe82look there
17:27:59Gourgasyeah its not there
17:28:00Gourgasoh well
17:28:23Gourgas"This is because they use Apple's proprietary Apple Accessory Protocol, which is only partially documented"
17:28:57Gourgasif this cheap chinese made transmitter can use that prot, I strongly doubt youguys cannot do it :)
17:32:03 Quit desowin (Read error: 110 (Connection timed out))
17:32:31 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
17:33:03 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
17:33:12 Quit linuxstb (Nick collision from services.)
17:33:19 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
17:37:19 Quit Gourgas ()
17:42:10 Join billytwowilly [0] (
17:43:06 Quit MajorC ()
17:43:20lostlogicbertrik: around?
17:43:36bertrikI need to go shopping now, store closes in 17 minutes
17:44:00bertrikback in 15 minutes
17:44:01lostlogicbertrik: enjoy, hit me when you get back, have a patch for you
17:46:26 Join MethoS- [0] (
17:47:35 Join MethoS-- [0] (
17:49:12 Join ackbahr [0] (
17:50:08 Join illissius` [0] (
17:51:21 Nick bb_ is now known as bb (n=bb@unaffiliated/bb)
17:54:11 Quit linuxstb ("Leaving")
17:57:20 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
17:59:54 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
18:02:53 Quit ameeen ()
18:05:22 Quit MethoS- (Connection timed out)
18:05:23puzzlesuh-oh, this doesn't look good.
18:05:28puzzles[ERR] Firmware partition doesn't contain Apple copyright, aborting.
18:06:16 Join stripwax [0] (
18:06:37 Quit illissius- (Read error: 110 (Connection timed out))
18:09:11 Quit mirak (Nick collision from services.)
18:10:00 Join mirak [0] (
18:13:45lostlogicTesters needed: If you have MoB related bugs please try:
18:14:47 Join dandin1 [0] (
18:14:58 Quit MethoS-- (Read error: 110 (Connection timed out))
18:15:12 Quit mirak (Nick collision from services.)
18:15:27linuxstbpuzzles: Does your ipod boot normally into the Apple firmware?
18:15:42puzzlesi wiped the firmware partition, it seems
18:16:01 Join mirak [0] (
18:16:01lostlogicpuzzles: no worries, just use itunes to install it ;)
18:16:07puzzlesi'm installing gayTunes on another box and i'm just going to restore it and format the drive
18:16:16puzzlesi was trying to avoid touching windows or itunes ;)
18:16:31 Quit mirak (Nick collision from services.)
18:17:13puzzlesoh, slick
18:17:20 Join mirak [0] (
18:20:14lostlogichcs: wake up, we went to sleep at the same time and I've been here for hours.
18:20:44 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
18:25:16 Join bluebrother|cafe [0] (i=5b036bdc@gateway/web/cgi-irc/
18:25:22 Quit spiorf (Read error: 110 (Connection timed out))
18:25:48 Join spiorf [0] (
18:26:38*bluebrother|cafe waves
18:26:59 Join karashata [0] (
18:27:06lostlogicbluebrother|cafe: coming to us live from coffee?
18:27:50 Quit kclaf (Remote closed the connection)
18:27:51bluebrother|cafenot exactly −− only some internet cafe (as I don
18:28:15bluebrother|cafet have an internet connection at home right now)
18:28:32bluebrother|cafebah, the web client sucks. And this stupid keyboard
18:28:35*lostlogic loans you some internets
18:28:37bluebrother|cafeyeah, I guess it
18:28:54bluebrother|cafe'll take 4 weeks or so until i'm back again
18:29:06lostlogicto get a phone line and dsl setup?
18:29:31bluebrother|cafeyep. And my final place −− I
18:29:39*lostlogic makes threatening noises about committing code that only I've tested :-D
18:29:40bluebrother|cafe'm at some temporary place this month
18:29:56lostlogicbluebrother|cafe: protip: enter != ' :-P
18:30:15 Quit Siku (Connection timed out)
18:30:35lostlogicbluebrother|cafe: I went through a month of temporary living for most of August, sux even if the place itself is nice.
18:30:41 Join Siku [0] (
18:30:42bluebrother|cafeyeah. I'm used to touch typing and this keyboard is messed up. Some laptop-like keyboard on a desktop maching
18:31:07bluebrother|cafecurrently the place is cold and rainy. But I'll see how nice work will get.
18:31:29bluebrother|cafejust started yesterday. At least the company and people seems to be great
18:31:31lostlogichaha, what region?
18:31:52bluebrother|cafe"northern" germany. Brunswick to be exact.
18:32:27bluebrother|cafebut coming from the south it's quite nothern ;-)
18:32:39lostlogicout here in Seattle we're expecting 8 months of mostly rain :-D
18:33:40bluebrother|cafethere is only one thing that's nice in this internet cafe: they provide laptop seats. Next time I need to bring my machine and try if the do packet filtering
18:33:58 Join Ebert [0] (
18:34:02bluebrother|cafeand hopefully I can update rbutil binaries −− this windows bug was rather nasty
18:35:34lostlogicahh yeah −− I prefer using my own computer on their network.
18:36:40 Join DrMoos [0] (
18:37:09 Quit moos (Read error: 104 (Connection reset by peer))
18:37:14bluebrother|cafeif they don't filter I could svn up ... *dream*
18:37:17*lostlogic makes more threatening noises about committing this code because it's burning a hole in my hard disk :-D
18:37:41lostlogicbluebrother|cafe: you could probably get one of us to setup a port-80 ssh server or proxy for you
18:38:11 Nick DrMoos is now known as moos (
18:38:13 Quit spiorf (Remote closed the connection)
18:39:03bluebrother|cafewell, I'll try another day. I had a bit of a walk to find this place here. There's enough stuff left to do ;-) (like finding a permanent home)
18:39:13 Join GnosticE [0] (
18:40:30 Quit ackbahr (Read error: 104 (Connection reset by peer))
18:42:01 Join kclaf [0] (i=kclaf@
18:52:46 Join stripwax [0] (
18:56:27*lostlogic gave up and committed, I hope I didn't break anyone too badly :)
18:57:38zichoWhen i make dynamic playlists, how do i add entire directories to them?
18:57:46 Join bertrik_ [0] (
18:58:01lostlogicbertrik_: ahahaha, seconds too late, I just committed, update and let me knwo what you get :):)
18:58:20bertrik_merge conflict probably :P
18:58:49lostlogicI'm pretty confident, I've abused this code as well as I can on my ipod and gotten not so much as a sniff of trouble
18:59:16bertrik_I've been hacking a bit today at simplifying the calculations done with track_ridx and track_widx
18:59:17 Join lee-qid [0] (
18:59:21linuxstbzicho: Hold down the select button on the directory name in the file browser, and select "insert".
18:59:55lostlogicbertrik_: ahh, that'd be very nice... playback.c can definitely be simplified a fair bit more now that buffering.c exists to do some of its thinking for it
19:00:52bertrik_would it be possible to create some kind of automatic test suite for it?
19:01:09lostlogicbertrik_: possible yes. am I going to do it? no.
19:01:09bertrik_or perhaps add assert statements to do sanity checking
19:01:19bertrik_no, I understand, a lot of work
19:01:41lostlogicactually buffering is pretty well sanity checked (thanks Nico!) playback not so much
19:03:34bertrik_I'm seeing warnings about 'long long int' in speex, ##!% ?
19:03:43lostlogicjust logf crap
19:03:59lostlogicSomeone Should Fix It (TM)
19:04:19*bertrik_ is still compiling
19:04:41lostlogicnote to all: %d != %ld
19:05:09 Quit stripwax (Read error: 104 (Connection reset by peer))
19:08:04lostlogicYES! I have successfully negated and reversed my binary size increase from last night.
19:08:06 Join ilgufo [0] (
19:08:07lostlogicthat's what I like to see.
19:10:17 Join stripwax [0] (
19:10:53bertrik_lostlogic: so far it works :)
19:11:08lostlogicbertrik_: sweet, let me know if tha tchanges :-P
19:11:25bertrik_it survived at least one usefl low situation, need to torture it some more
19:12:59 Quit stripwax (Read error: 104 (Connection reset by peer))
19:13:15bertrik_usefl is supposed to fill up almost completely after it reaches 0, right?
19:13:26lostlogicshould yeah
19:13:42 Quit bertrik (Read error: 110 (Connection timed out))
19:13:45 Join DrMoos [0] (
19:14:13*bluebrother|cafe leaves for today ... cu
19:14:18***Saving seen data "./dancer.seen"
19:14:26 Quit bluebrother|cafe ("no more coffee for today ;)")
19:14:31 Quit moos (Read error: 110 (Connection timed out))
19:14:33 Nick DrMoos is now known as moos (
19:15:02bertrik_argh, playback stopped again, but I could just be my own changes, and I realise logf is not enabled for buffering
19:15:19bertrik_I'll compile a fresh SVN version with logf for playback and buffering
19:15:31lostlogicbertrik_: thanks
19:15:41 Quit Buschel ()
19:17:10 Quit dandin1 ()
19:18:19 Join stripwax [0] (
19:21:43 Quit SirFunk (Read error: 110 (Connection timed out))
19:27:19 Quit mirak (Read error: 110 (Connection timed out))
19:28:41 Nick bertrik_ is now known as bertrik (
19:28:57 Join Frazz [0] (
19:29:00puzzlesyay for rockbox!
19:29:50 Quit GnosticE ()
19:30:39 Join linuxstb_ [0] (
19:31:04 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
19:31:56 Quit linuxstb (Nick collision from services.)
19:32:07 Nick linuxstb_ is now known as linuxstb (n=chatzill@rockbox/developer/linuxstb)
19:34:31hcslostlogic: I hear you were looking for me?
19:34:47lostlogichcs: yeah, just want you to test the latest SVN
19:35:00lostlogicI'm not sure if i hope you break it or not... but I _think_ you won't.
19:35:08 Join spiorf [0] (
19:35:38lostlogicADX might still break it
19:35:46lostlogicsince it's too big to treat as atomic
19:35:56hcsprobably, I won't worry about that
19:37:24hcsshould this care about the watermark stuff, which is still in the spc codec?
19:37:52bertriklostlogic: playback stopped again under the same scenario as initially described in FS #8077
19:38:13bertriki have a logf of playback and buffering
19:38:17lostlogichcs: it should absolutely not break regardless of watermark but it will probably end up pausing
19:38:25lostlogicbertrik: ugh, reopen and post :(
19:38:29*lostlogic punches sansas
19:38:50lostlogic(pausing while the flash is accessed after the last track on buffer finishes)
19:39:59bertrikhcs: ?
19:40:31hcser, lostlogic: gotcha
19:40:41 Join lespea [0] (
19:41:33lostlogicbertrik: also, does the buffering thread screen look the same when it breaks now? I'm specifically wondering about the real suddenly dropping to zero part
19:42:04 Quit stripwax (Read error: 104 (Connection reset by peer))
19:42:24bertrikI didn't pay attention to that just now, next time I'll have a better look
19:42:29lostlogicbertrik: thanks
19:42:44 Quit lespea (Client Quit)
19:42:46 Quit daurn (Read error: 113 (No route to host))
19:43:05 Join lespea [0] (
19:43:06bertrikcan you reopen the bug report?
19:43:25bertrik(so I can add the latest logf of SVN 15436)
19:44:01lostlogicsorry didn't realize you couldn't reopen
19:44:20qwmwhat time is it now in GMT/UTC?
19:44:46BigBambiqwm 18:44
19:46:20bertriklostlogic: pcm dropped to 0, alloc and real are still full, usefl=7716
19:46:36hcslostlogic: SPC working, those troublesome MP3s from last night working
19:46:41lostlogicbertrik: ok, good to know
19:46:50lostlogichcs: ok, I've narrowed the field *sigh*
19:46:53lostlogicthanks guys
19:47:10bertrikthank you for looking into it
19:47:41 Join stripwax [0] (
19:48:28Nico_Plostlogic: I'm afraid I'll only be able to look at your commits tomorrow
19:48:55lostlogicNico_P: no worry, I think I'm zeroing in on the bastard
19:49:31Nico_Pmaybe I'll need a few more details than what's in the commit messages
19:49:54lostlogicNico_P: I should be around tomorrow morning to chat about it
19:49:58lostlogic(morning PDT)
19:50:06midkaylostlogic: just jumping in, which bastard are you talking about?
19:50:24lostlogicmidkay: hehe :) 8077, 8073, 8074
19:50:42Nico_PI have to go now... see you ;)
19:50:43lostlogicand whichever one was hcs' spc bug
19:50:47lostlogicsee ya
19:50:51 Quit Nico_P (Remote closed the connection)
19:50:58lostlogicof those I think only 8077 is still alive
19:51:11midkaylostlogic: ah, right, i've had this myself :)
19:51:32lostlogicmidkay: anything to add to the bug report that might give more clues?
19:52:01hcsmine's 7331, I'll close it
19:52:01 Part kubrick
19:52:34lostlogic7331 is gameboy :-P
19:52:47midkaylostlogic: hmm, it may not be the same thing and i haven't really investigated, but in the past week or two, in general i have a lot of sudden playback stops..
19:53:17hcslostlogic: damnit, i have to wake up before trying to do this crap
19:53:30lostlogicmidkay: gotcha, logfs or buffering_thread reports from when that happens would be useful, if you catch some
19:53:34lostlogichcs :)
19:54:13lostlogichcs: I closed the SPC one because I was _that_ confident that it'd be dead... 'course I also closed 8077 which isn't dead :-D
19:54:18lostlogicbertrik: got that logf?
19:54:57bertrikyes, I just added it
19:56:26lostlogicbertrik: stopping and resuming playback still works when it pauses like this?
19:56:33lostlogic(or does it crash?)
19:56:59bertrikstopping it causes a freeze, pausing/starting has no effect
19:57:50lostlogicbertrik: ugh, would you mind terribly enabling logfqueues on playback and buffering and getting me another logf? I keep forgetting to ask for those up front.
20:04:03lostlogicwait wait, I have a theory! Yes, I have a theory! The logfqueues will help confirm or deny it
20:04:07lostlogiccomplicated to solve
20:04:10lostlogicstupid flow.
20:04:20 Quit Rondom ("Ex-Chat")
20:04:30 Join desowin [0] (
20:10:42bertriklostlogic: attached the logf with LOGQUEUES enabled to the bug tracker
20:12:21 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
20:14:44 Quit ilgufo ("So Long, and Thanks For All the Fish -")
20:15:14 Quit Llorean ("Leaving.")
20:15:56 Join SirFunk [0] (
20:16:56 Join ilgufo [0] (
20:19:55DerPapstamiconn: my other 3G is a PP5002D too.
20:21:13 Join japc [0] (
20:35:48 Join dandin1 [0] (
20:36:51 Quit bertrik ("bye")
20:36:54 Quit Arathis ("Bye, bye")
20:41:08 Quit Mouser_X ("CGI:IRC (Ping timeout)")
20:43:04 Join lazka [0] (
20:43:41 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
20:46:15 Join StillOcean [0] (
20:46:28hcslostlogic: it may interest you to know that ADXs also work
20:47:24 Part StillOcean
20:52:37 Join w1ll14m [0] (i=d55d29c4@gateway/web/cgi-irc/
20:55:21w1ll14mscorche: are you there ?
20:55:26 Join fm2 [0] (n=chatzill@
20:55:39fm2from playback.c: "bool has_codec; /* Codec length in bytes */" Does the comment still apply?
20:56:02 Join dandin1x [0] (
20:56:28linuxstbfm2: I think it's safe to say it doesn't...
20:56:41 Quit dandin1 (Nick collision from services.)
20:56:45 Nick dandin1x is now known as dandin1 (
20:57:12fm2linuxstb: he-he, a very diplomatic answer! :-)
20:58:25linuxstbWe can blame lostlogic for that line - r15435 ;)
20:58:33fm2Is the field still needed at all? What are the valid values for hid? Only non-negative?
20:59:36fm2Then codec_hid could just be set to -1. And a macro could be defined such as HAS_CODEC (ci) (ci->codec_hid >= 0)
21:00:06hcslostlogic: retraction, ADX doesn't seem to switch between tracks successfully, but we don't care much anyway
21:00:22fm2BTW: will the MOB solve the problem with track change at the end of the track?
21:00:55linuxstbfm2: It seems to be. I wouldn't like a macro, but making 0 a valid handle id would be consisted with file descriptors in Rockbox, so IMO would be clearer.
21:02:10fm2linuxstb: same thoughts here. Ok, checks can be made with ... >= 0
21:02:44fm2linuxstb: any info about the track change question?
21:04:16fm2So the bug ist still there with mob?
21:04:30w1ll14mhmm rockbox boots now on my ipod nano 1g.....
21:04:38w1ll14mbut playback was impossible
21:04:56w1ll14mit seems my memory is corrupted?!
21:05:10w1ll14mwell, i'll re format it and test it again
21:05:42 Quit fm2 ("ChatZilla [Firefox]")
21:11:09 Quit Mouser_X (Nick collision from services.)
21:11:33 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
21:14:19***Saving seen data "./dancer.seen"
21:14:49 Join stewball [0] (n=WTFOMGBB@
21:15:53 Join MethoS- [0] (
21:16:23 Join eigma [0] (n=cat@
21:16:29eigmakkurbjun: ping
21:17:32hcslostlogic: retraction 2, MP3 froze up in the middle of a track
21:20:07 Join MethoS-- [0] (
21:24:37 Quit MethoS-- (Remote closed the connection)
21:26:05 Quit amiconn (Nick collision from services.)
21:26:14 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:27:19kkurbjuneigma, hey
21:27:37 Quit stewball ()
21:27:41eigmaoh, I just wanted to ask you where the low-level LCD drawing code was.. found it in the mean time
21:27:57 Join stewball [0] (n=WTFOMGBB@
21:28:08kkurbjunI noticed that you were getting an error when building saying that plugin ram was full
21:28:21kkurbjunyou should just have to re-run ../tools/configure
21:28:28kkurbjunand then rebuilt
21:28:39eigmaone of your changes hadn't applied entirely to my tree.. I reverted some files and it worked
21:28:59kkurbjunoh gothca
21:30:34 Join Robin0800 [0] (
21:32:19 Quit Robin0800 (Client Quit)
21:32:35 Join Robin0800 [0] (
21:33:48 Join webguest89 [0] (i=543f2003@gateway/web/cgi-irc/
21:34:13 Quit Robin0800 (Client Quit)
21:34:44 Join Robin0800 [0] (
21:36:00 Quit MethoS- (Read error: 110 (Connection timed out))
21:36:35webguest89Hi, I just got myself an e260 and it runs rockbox (remembered it from the archos jukebox times). Now it randomly stops playing. If i press play again it restarts the song. is there any more information i could collect for a bug report?
21:37:37lostlogichcs: yeah, that's most likely 8077 manifesting on your world
21:37:40 Quit Robin0800 (Client Quit)
21:37:56lostlogicoops, comment wrong
21:37:57 Join Robin0800 [0] (
21:38:24lostlogicunfortunately, it would have taken a bit more fiddling to make the has_codec flag go away entirely
21:38:44webguest89And another thing that probably doesn't belong here: i can't "crossdev -s1 −−target mk68-elf" on my gentoo box. the gcc build fails
21:39:27 Quit Robin0800 (Client Quit)
21:39:27linuxstbtarget should be m68k-elf
21:39:43 Join Robin0800 [0] (
21:39:54webguest89sorry typo on my part, but i did copy-paste it from the instructions page
21:40:05webguest89(the one i pasted to the command line)
21:40:26linuxstbBut we recommend you use the tools/ script in the Rockbox source to build the cross-compilers.
21:41:06webguest89won't that mess up the gentoo package system?
21:41:13 Quit eigma ()
21:41:20 Quit Robin0800 (Client Quit)
21:41:36 Join Robin0800 [0] (
21:41:57linuxstbIt's independent of your distro's packaging system. You can even install the compilers somewhere within your home directory if you wish.
21:43:00webguest89okay i'll try that thanks
21:44:36 Quit midkay ("Leaving")
21:46:45 Join Robin_0800 [0] (
21:47:05GodEaterspeaking as a gentoo user myself - I recommend the route - crossdev is a pain in the ass
21:47:58w1ll14mis someone here busy with nano 1g problems ?
21:48:30 Join [1]Robin0800 [0] (
21:48:48 Quit lee-qid ("aufwiederbyebientotsayonara")
21:52:29GodEaterw1ll14m, can you be more specific ?
21:54:18w1ll14mGodEater, well, with the new builds nano's have different problems
21:54:24 Join webguest56 [0] (i=47fece89@gateway/web/cgi-irc/
21:54:45w1ll14mGodEater, there is a FS:
21:54:46 Join webguest44 [0] (i=47fece89@gateway/web/cgi-irc/
21:55:04webguest44Does Rockbox support DIVX or AVI videos?
21:55:20GodEaterw1ll14m, I still believe no developer owns a problem nano - making it difficult to debug
21:55:22GodEaterwebguest44, no
21:55:29 Join lee-qid [0] (
21:55:29GodEaterwebguest44, only mpeg2
21:56:09w1ll14mGodEater, could it be that it has problems with the way frequencies are defined ? i have my ipod now running at 30/80 Mhz
21:56:13webguest44GodEater: ok, thanks
21:56:37GodEaterw1ll14m, I believe the consensus is that it's do with the ATA controller
21:56:41w1ll14mGodEater, for som more specific details you can read the last comments in the FS task
21:57:04 Quit webguest44 (Client Quit)
21:57:19w1ll14mGodEater, that was my thought to, as my memory was corrupted with the previous build
21:57:50w1ll14mand before i had a even worse problem 'no partition found'
21:57:55GodEaterw1ll14m: I don't think it's to do with the frequency it's being run at though. However - I don't own a nano at all, and haven't been looking into this.
21:58:08w1ll14mGodeater, ok
21:58:41w1ll14mis there any i can do to test ?
21:59:05w1ll14mi would borrow my ipod to a dev if the ipod was mine ;)
21:59:25 Quit webguest56 ("CGI:IRC (Ping timeout)")
22:00:09 Quit SirFunk (Remote closed the connection)
22:04:24lostlogiclinuxstb: 0 => valid hid on the way ;)
22:05:06 Quit Robin_0800 (Read error: 110 (Connection timed out))
22:06:37lostlogicI wonder if I should treat the case when there are long lived handles (long enough for the code to cycle through all valid hids)
22:06:37 Quit Robin0800 (Read error: 110 (Connection timed out))
22:06:37 Nick [1]Robin0800 is now known as Robin0800 (
22:13:10 Quit Robin0800 (Read error: 104 (Connection reset by peer))
22:13:24linuxstblostlogic: Are we going to live that long?
22:18:34lostlogiclinuxstb: a 30 meg buffer fill uses 16, so that's 2048 buffer fills on a 16 bit signed counter... which is what I limitted it to because I wasn't positive if there were any int => 16 bit swcodec targets
22:18:39lostlogicare there?
22:18:48lostlogicif there aren't any 16 bit int swcodec targets then I can make it a lot lot bigger and not worry about it
22:19:28lostlogicI wish fm2 was still here since I killed off has_codec too
22:19:31linuxstbNo, there are no current targets with 16-bit ints. That doesn't mean an insane person won't try such a port in the future though..
22:19:55linuxstbBut this will be the least of their problems...
22:20:06lostlogic*nod* *bumps the mask*
22:20:48lostlogicok, now we won't live that long.
22:22:11lostlogichcs: can you reliably cause the mp3 stoppage that you mentioned?
22:22:54hcsnot sure, I'll try the same sequence again, but it takes around 20 minutes
22:23:12hcser, 20 MB, not sure how many minutes
22:23:21lostlogichcs: you can wait until the buffer fills completely then skip to the last track on buffer and let it play out to speed that up
22:23:32lostlogic(you'll know hwn you are onlast track because the wps will show nothing in next track info)
22:23:49lostlogicI think I know how to fix it.
22:23:52 Join cooz [0] (
22:24:00 Quit Bagder ("*plopp*")
22:24:06lostlogiccrap, except my commit just now is all borked.
22:24:40lostlogicI didn't copy the right zip to the ipod when I thought I was testing it :(
22:26:05hcsI'm going out shortly, I'll see if I can reproduce it consistently and check on things when I return
22:26:11 Quit hcs ("Leaving.")
22:26:27 Quit ilgufo (Read error: 110 (Connection timed out))
22:30:15lostlogicsomeone broke their build server.
22:30:50lostlogicGodEater: ping!
22:32:25 Join Fraser [0] (
22:32:47 Quit Frazz (Read error: 104 (Connection reset by peer))
22:35:08amiconnGodEater: Please fix your build server, or block ssh from the master...
22:35:31lostlogicit's kinda good that it's showing red now because that build breaks playback any way :-D
22:36:55amiconnCertainly not for hwcodec
22:37:21lostlogicamiconn: hah, trunuf
22:37:56amiconnah, hmm
22:38:08amiconnxtrm abbrv
22:39:24 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
22:40:18Mouser_XHow do I add comments, or comment something out, in a WPS?
22:40:55*Mouser_X is trying to remove patch dependencies for a WPS. It's not failing in checkwps, but it's crashing the sim.
22:43:01 Quit stewball ()
22:44:34lostlogiclinuxstb: I'm going to hafta revert :( It's rather complicated (it turns out) to make zero a legal hid
22:45:52 Join kugel [0] (i=kugel@unaffiliated/kugel)
22:50:13amiconnlostlogic: Why do handle ids need to be 32 bit? Can't they be reused?
22:51:06 Quit animeloe ("Leaving")
22:51:13lostlogicamiconn: they can but we would hafta search a linked list to know which ones are available
22:51:51amiconnWhy search? aren't they always contiguous?
22:52:19lostlogicamiconn: no.
22:52:40 Quit Siku ()
22:52:40lostlogiccodecs are freed out of the middle of the buffer atm
22:54:21 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
22:55:11 Join Robin0800 [0] (
22:59:09*lostlogic abuses build servers... while godeater's is down
23:00:21 Quit Fraser ("Leaving")
23:01:41BagderI'll disable his server for now
23:05:15amiconnWould anyone complain if I remove the clock setup related values from the PP ports debug screen?
23:05:48amiconnI'll put DEV_EN2 there instead, plus the newfound GPO32
23:05:55linuxstbWeren't you the person who added them?
23:06:21Mouser_XIs there anything wrong with this line (in regards to WPS stuff)? "Genre: %?ig<%ig|−−−−>"
23:06:54Mouser_XIf I have that commented out, the sim doesn't crash. If it's not though, the sim is crashing everytime, for me (Sansa e2xx)
23:07:24Mouser_XIt also crashes on "Genre: %ig"
23:09:12 Quit sd__ (Read error: 113 (No route to host))
23:09:35 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
23:10:06 Join mirak [0] (
23:11:34saratogahow can i check if the hold switch is on for the Sansa?
23:14:22***Saving seen data "./dancer.seen"
23:14:33linuxstbbutton_hold() I think
23:15:32SoapMouser_X: can you pastebin the entire WPS quickly?
23:15:38saratogawhich h file is it defined in?
23:15:58saratogaah i see it in the target tree
23:16:48 Quit merbanan (Remote closed the connection)
23:17:31saratogalinuxstb: how do I import a .h file from the target tree?
23:17:56saratogais the compiler smart enough to know which target it should import the h from or do I need to put the full path
23:18:51amiconnNeeding the full path would be silly
23:19:25amiconnThe Makefile sets the include dirs so that the correct .h files for the target are used
23:20:01amiconnTarget tree is there for reducing #ifdef clutter, not for increasing it
23:20:50Mouser_XSoap: (This is the one I've modified to remove patch dependencies. If line(s?) 49 and/or 50 are commented out, the sim doesn't crash.)
23:20:51amiconnThat's why the .h files for a certain module should be called the same for each target (if they exist at all). E.g. system-target.h, *not* system-e200.h
23:21:09 Quit desowin ("use linux")
23:21:14 Quit jhMikeS (Nick collision from services.)
23:21:20 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:21:23Mouser_X(I've only had either both, or neither, commented out, so I'm not 100% certain which it is.)
23:21:50amiconn.c files, however, should be named after their target. This only requires #ifdefing in firmware/SOURCES, where it's bearable
23:21:50 Join random_desu_is_s [0] (
23:22:00saratogaamiconn: thanks, good to know
23:22:55 Quit lespea (Remote closed the connection)
23:23:53Soapodd. In general, Mouser_X, wasn't the new WPS parser supposed to fail gracefully, even in the Sim?
23:25:40Mouser_XNo idea. I do know that in the sim, it either hangs, and I have to close it, or Windows brings up the "This made a boo-boo, send report?" box, and then the sim gets closed when I tell that window no.
23:25:40 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:26:22 Quit davina ("xchat on Ubuntu 7.04")
23:26:33 Join sd [0] (n=sd@
23:26:45*amiconn always disables error reporting in windows in order to avoid those annoying requests
23:26:56 Join random_desu_is_s [0] (
23:27:19Mouser_XSoap: This is a build from rasher's site. I didn't compile it myself. I don't think it's from the current build.
23:27:40Robin0800what does id =15 in this rockbox file rockbox-info.txt on the ipod video should this be 22?
23:28:09karashataas far as I know the sim should do the same thing as the actual builds when confronted with invalid tags and bad WPS files, and that's load the default WPS
23:28:21Mouser_XThe crashing is probably due to some MoB error, since what it's crashing on is Metadata stuff.
23:28:24amiconnid is the target id, which is unique and has nothing to do with the numbering in the configure menu
23:28:54Mouser_Xkarashata: The WPS is good. I've removed all of the parser problems.
23:29:03karashatathat's good
23:29:13w1ll14mamiconn, do you now someone with the nano problem we talked about yesterday who is willing to test a patch ?\
23:29:13 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:29:17Mouser_X(Well, good, in that checkwps said it successfully parsed it.)
23:29:18karashatathen it likely is a problem with the MoB stuff
23:29:33amiconnw1ll14m: No, but what patch?
23:29:45Soapw1ll14m: if you are talking about the clock-revealed Nano problems...I have such a device.
23:30:31Robin0800amiconn, is there a list of target id's somwhere?
23:30:45amiconnRobin0800: In configure...
23:31:53amiconnscramble should also know about them (at least those which are passed through scramble)
23:31:54 Join random_desu_is_s [0] (
23:32:17Robin0800amiconn,but why are thy not the same as the configure numbers?
23:32:23w1ll14mi've tested a patch (created my self) wich allows me to run rockbox
23:32:47amiconnThe menu numbers in configure are allowed to change. The id stays unique for each target
23:33:14w1ll14mSoap, maybe you could test something for me ?
23:33:14 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:33:54 Join random_desu_is_s [0] (
23:34:25Soapdoes it address the problem, w1ll14m, or does it put lipstick on a pig? ;)
23:34:32*amiconn found hd poweroff on the G5 :D
23:35:13amiconnIt's on the newfound GPO32, bit 30, low active like on the other PP502x targets (where it's on different ports though)
23:35:14BigBambiamiconn: woot!
23:35:16w1ll14mwell, it changes the way PLL_CONTROL is calculated.... let me give you an example
23:35:41amiconnw1ll14m: The way PLL_CONTROL is calculated is correct like it is now
23:36:54w1ll14mamiconn, i agree but it fixed my nano problem
23:37:25*linuxstb looks at a disassembly of the View firmware, and sees familiar PP502x registers being used.
23:37:25 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:37:27w1ll14musing a normal build gives me the error 'no partition found blah blah'
23:37:27 Join TMM [0] (
23:37:37amiconnhi linuxstb :)
23:38:24w1ll14mand now it runs just fine at 30/80 MHz
23:38:55amiconnlinuxstb: I've found hd poweroff on mini G2 and the G5, but unlike I thought first, they're not identical
23:39:10 Join random_desu_is_s [0] (
23:39:22amiconnSo we need to find/verify hd poweroff on G4, Color, and mini G1
23:39:38linuxstbNice. How do you verify it? Just turn it off and see if disk accesses break?
23:39:56amiconnI'll have a quick look on the Color rom now. Could you perhaps perform a quick test afterwards?
23:40:17saratogahmm if i call power_off() in the bootloader, it appears to work but the sansa wheel LEDs never power back off
23:40:19 Join colin_ [0] (
23:40:23 Quit TMM (SendQ exceeded)
23:40:23amiconnJust disable it while the disk is spinning and you'll get the distinctive click (emergency park)
23:40:25saratogado i have to manually turn those off?
23:40:37amiconnNot that distinctive on mini, but still noticeable
23:41:11 Join TMM [0] (
23:41:11 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:41:19 Quit jhMikeS (Nick collision from services.)
23:41:19amiconnsaratoga: Iirc power_off needs some special magic on the sansa. I wouldn't rely on it working properly in the bootloader
23:41:19saratogaamiconn: congrats, should make a few people happy
23:41:25 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:41:58 Join random_desu_is_s [0] (
23:42:59amiconnlinuxstb: What I did: Set hdd poweroff timeout to something high (e.g. 30 seconds), dircache enabled. Then powerdown, boot, and while dircache is still scanning enter the debug menu and 'View HW info' (where I placed the poweroff port write). -> click
23:44:00amiconnI still have to check whether actually using it works, or whether we have to do some additional init when powering back up (but I doubt that, as on H10 it's also just a port bit flip)
23:44:01 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:47:02 Join random_desu_is_s [0] (
23:48:22 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:48:56amiconnlinuxstb: I thought I had an ipod color rom dump, but I don't...
23:49:07linuxstbI can fix that...
23:49:53amiconnkkurbjun: Are you there?
23:50:08 Quit donutman25 ("ChatZilla [Firefox]")
23:50:29*amiconn expects the mini G1 hd poweroff pin to be the same as mini G2
23:50:56 Join random_desu_is_s [0] (
23:51:56 Join bertrik [0] (
23:55:26 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:57:00amiconnlinuxstb: Looks like the color has hd poweroff onn GPIO J bit 2 like the mini
23:57:01 Join random_desu_is_s [0] (
23:57:24*amiconn is now almost sure that G4 greyscale and mini G1 are the same
23:58:09linuxstbOK, I'll test. Is 0 or 1 off?
23:58:10 Quit HellDragon (Client Quit)
23:58:12saratogawill be interesting to see how big of a difference this makes for battery life
23:58:19amiconnLow active, i.e. 1 is off
23:58:33amiconnYou can use my newly introduced atomic gpio manipulation macros
23:58:55amiconnGPIO_SET_BITWISE(GPIOJ_OUTPUT_VAL, 0x04);
23:58:56 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))

Previous day | Next day