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 2010-06-07

00:00:41***Saving seen data "./dancer.seen"
00:03:23 Quit bertrik (Ping timeout: 260 seconds)
00:08:14 Quit M3DLG (Ping timeout: 264 seconds)
00:14:36 Quit notlistening (Quit: Leaving)
00:14:55simonrvnso what tool should i use to pop open my fuze? i don't want to break the plastic clips.
00:20:16 Nick fxb is now known as fxb__ (
00:31:40 Join jxself [0] (~jxself@fsf/member/jxself)
00:44:46jxselfOh wait. v 3.6 makes it possible to use Rockbox on Sansa e250v2? Awesome, and thanks. :)
00:49:21 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
00:50:02 Join Jennifur [0] (
00:50:09 Quit wincent (Ping timeout: 265 seconds)
00:53:54 Join Antibuddha [0] (
00:53:56Antibuddhahow hard would it be to make a rockbox software hard drive based music server from scratch parts?
00:54:30Antibuddhai mean like the bare necessities... a hard drive, battery, spdif output, and lcd screen, with simple controls like an mp3 player
00:54:55Antibuddhathere is great need for software for diy music transports
00:55:18Antibuddhafreaking audiophile companies want to charge $1000-3000 for a hard drive and software
01:00:03 Quit ender` (Quit: I have a theory that it's impossible to prove anything, but I can't prove it.)
01:01:17 Join hebz0rl [0] (
01:01:52DhraakellianAntibuddha: ?
01:02:11Antibuddhacool tyvm i will read
01:02:25Dhraakellian'though I tihnk their goals may be a little bit different
01:02:39Antibuddhai have suggested to rockbox to make their own open source player for years
01:02:46Antibuddhaits good to see they are trying to do it
01:05:57DhraakellianI seem to remember something about getting a company to manufacture it rather than it being quite so DIY, but I haven't been paying close attention
01:07:30Antibuddhathere's this one small business in china, audio-gd, with great quality transports and other electronics, but they don't have software or any products like a hard drive based audio transport
01:07:51Antibuddhai bet if they had access to the software they could make a great hard drive transport, and their prices are very low
01:11:00 Nick _[CGL]_ is now known as [CGL]_ido (~CGL@
01:12:26 Quit efyx (Remote host closed the connection)
01:15:02Antibuddhathe company is going to release a battery powered dac amp soon
01:15:19Antibuddhawould be nice if they put it on hold and tried making an mp3 player
01:18:57Antibuddhaim going to ask people at cisc memory player if they are interested in an arm9 rockbox based music transport. they use computers with special bit perfect and ram buffering software to increase sq, but i figure arm9 controller must be superior
01:19:23 Join Mobius-1 [0] (
01:19:51 Part jxself ("Support freedom:")
01:20:31Mobius-1good day rockbox users.
01:22:02Mobius-1i love the lua scripting for rockbox. kinda wish the documentation wasn't lacking so much :\
01:46:44 Quit MethoS- (Read error: Connection reset by peer)
02:00:43***Saving seen data "./dancer.seen"
02:09:36 Join FlynDice [0] (
02:11:46 Quit robin0800 (Remote host closed the connection)
02:20:07 Join FlynDice_ [0] (
02:21:01 Quit FlynDice (Ping timeout: 265 seconds)
02:34:26 Quit DerPapst (Read error: Connection reset by peer)
02:35:02 Join DerPapst [0] (
02:39:43 Quit Jennifur (Ping timeout: 248 seconds)
02:41:10 Join Jennifur [0] (
02:45:35 Quit Mobius-1 (Remote host closed the connection)
02:46:52 Join TopyMobile [0] (
02:48:41 Join FlynDice [0] (
02:48:47 Quit FlynDice_ (Ping timeout: 248 seconds)
02:57:17 Quit tom` (Ping timeout: 265 seconds)
02:58:10 Join nave7693 [0] (
02:58:36 Part nave7693
03:02:19 Quit TopyMobile (Ping timeout: 240 seconds)
03:04:19 Quit [CGL]_ido (Ping timeout: 240 seconds)
03:08:28CIA-6New commit by bieber (r26639): Theme Editor: Implemented document load functionality
03:15:10 Quit t0rc (Quit: Leaving)
03:19:38 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
03:24:09 Join anewuser [0] (anewuser@unaffiliated/anewuser)
03:24:38 Quit Antibuddha (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
03:27:18S_a_i_n_t_JdGordon: Any particular reason you asked about pacbox?
03:27:24S_a_i_n_t_I use it fairly often.
03:31:25S_a_i_n_t_Torne: USB charging for iPod, congrats ;)
03:32:40hamish_Is there a SUB driver being worked on for the sansa fuze?
03:33:36S_a_i_n_t_(immesnse sarcasm) no, they figure they'll
03:33:44S_a_i_n_t_leave it as it is ;)
03:34:14hamish_Is there someway I can help?
03:35:16saratogadig through the current driver and see what its missing
03:38:17JdGordonS_a_i_n_t_: there is a patch i was looking at for it
03:38:18simonrvnit's on the wiki. seems simple enough for a coder.
03:38:47S_a_i_n_t_JdGordon: Does it do anything particularly interesting?
03:39:09JdGordonim not sure
03:39:33saratogaor i guess dig through the fuze USB driver and see how it works
03:39:41saratogafuze binary's driver i mean
03:43:23saratogahamish_: you might also talk to pamaury the next time hes on
03:54:31JdGordonok, skin breaking is going in pretty soon...
03:54:43JdGordonthis is pretty muc the last chance to change any tags without causing much annoyance
04:00:17 Join grndslm [0] (
04:00:47***Saving seen data "./dancer.seen"
04:04:12hamish_saratoga, I
04:04:26hamish_'m only a beginner coder... :P
04:09:54 Join Zigtown [0] (
04:10:39 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
04:18:25JdGordonsaratoga: are you going to commit thw pictureeflow+wps patch?
04:18:34JdGordon7 sub patches now :/
04:26:52 Join DerPapst1 [0] (
04:29:15 Quit DerPapst (Ping timeout: 240 seconds)
04:29:33 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
04:38:47 Quit simonrvn (Read error: Connection reset by peer)
04:40:04 Quit kramer3d (Quit: Leaving)
04:42:31 Quit hamish_ (Remote host closed the connection)
04:44:38 Quit JdGordon (Changing host)
04:44:38 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
04:49:54 Quit pixelma (Disconnected by services)
04:49:54 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:50:01 Quit amiconn (Disconnected by services)
04:50:03 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:50:12 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:50:16 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:51:51saratogaJdGordon: no you probably should since i haven't looked at the pictureflow code
04:52:29saratogawell theres always that email i sent to the list
04:57:40 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
04:58:37 Join Barahir_ [0] (
04:59:57 Quit grndslm (Remote host closed the connection)
05:01:49 Quit TheSeven (Ping timeout: 264 seconds)
05:02:03 Quit Barahir (Ping timeout: 248 seconds)
05:02:57 Join grndslm [0] (
05:05:55 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:16:29 Quit Zigtown (Remote host closed the connection)
05:17:07 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
05:19:02 Quit hebz0rl (Quit: Ex-Chat)
05:22:22 Join tom` [0] (
05:23:03JdGordonsaratoga: I dont really see the need for someone who actually worked on pictureflow to review the patches
05:23:16JdGordonif they work then a general review should be good enough
05:23:26JdGordonespecially if csavery wants to contnuie working on it
05:25:19JdGordonS_a_i_n_t_: oy
05:25:43CIA-6New commit by bieber (r26640): Theme Editor: Added a preferences dialog and allowed modification of the syntax highlighting and editor colors
05:27:29 Quit kenguest (Read error: Connection reset by peer)
05:27:36 Join kenguest [0] (
05:27:45JdGordonbieber: woo! looking good
05:28:31JdGordondo you know how you're going to the renderer yet?
05:29:32bieberTentatively, I still need to read up more on Qt's graphics API. My plan is to try and give each element of the tree its own graphics object, and then have the parent nodes composite all their children and so on until it gets to the top of the tree
05:29:55bieberThen apparently Qt will let me do things like handling mouse clicks on the individual objects, so I don't have to deal with collision detection and all that
05:30:33 Join CGL [0] (~CGL@
05:30:48JdGordonso the code wont be usable at all for rockbox?
05:31:08bieberNo, I'm planning on straight C++
05:31:19JdGordonI've got a patch ready which changes the draw order which shuold make things line up more
05:32:46bieberbtw, will the .cfg files for the themes get mad if I insert a field that Rockbox doesn't use?
05:32:51JdGordonthinking ahead, (and maybe to get help anyway) can you do a tutorial (or at least document) on how to add new tags once you know how it will all work?
05:33:05JdGordonas long as it starts with a # it will be fine
05:33:39LloreanWhat would you put in themes that Rockbox wouldn't use?
05:34:24bieberJust diagnostic info for the editor
05:34:41bieberI want to be able to maintain projects using just the .cfg files, but the editor needs to know what target the theme is for
05:34:42*JdGordon is slightyl worried that eventually someone will add a tag and forget or not know how to add it to the editor
05:35:06bieberJdGordon: No problem. Parsing, of course, is easy. Hopefully rendering will also be easy when I get that built
05:35:25bieberI also need to update my project Wiki page, so I'm thinking tonight I'll just sit down and write documentation for a while
05:35:36JdGordonna, docmentation is boring
05:35:45LloreanWhy does the editor need to know what target the theme is for?
05:35:54JdGordonok, here we go!
05:36:00LloreanShouldn't you have the option to display the theme as it'd be seen on various targets, since themes can often be used on many targets?
05:36:34JdGordonscreen size+depth isnt the only thing that determines if a theme is usable on a target
05:36:56LloreanJdGordon: Yes, so the "view on target" list should restrict itself based on what features the theme seems to require.
05:37:36JdGordonthat does make sense
05:37:38bieberYou can view on different targets, but it would be nice to at least have it remember the primary target you're working with, so you don't have to pick from a list every time you load the project
05:37:52bieberI'm sure I could include a list of targets in the .cfg file as well
05:38:02Lloreanbieber: Can't the list be computed?
05:38:17LloreanYou look at what the theme requires, and eliminate targets that can't deliver it (below minimum resolution, don't have RTC, etc...)
05:38:32bieberThat's doable
05:39:12LloreanI'm just not too fond of inserting lines into the theme.cfg that don't serve a purpose for Rockbox since people downloading the .cfg may not even know of the theme editor.
05:39:39LloreanIf you download a theme for your D2, you don't want to see "#target: ip5g" or whatever in the .cfg and be confused
05:40:24LloreanI also think people using the theme editor will *generally* be using it for one target (whichever one they own / design for) so it's probably pretty safe to just keep track of what the last used "display on" option was and go with that, prompting if the current theme won't work on that device for some reason
05:41:30LloreanJdGordon: besides resolution, what features restrict themes? I thought use of the RTC tag had a fallback if the RTC wasn't present, because of the H100 RTC situation. Are there other tags that represent possibly non-present hardware?
05:41:37LloreanOr is it just color depth and resolution?
05:41:52JdGordonfm tags, rec tags
05:42:20LloreanAh, right.
05:42:22LloreanOther screens.
05:42:55JdGordonI would have no problem with the config having the line "# theme build by the theme editor for screen size 176x220x16" and from that work out the targets it will work on based on the tags
05:43:49bieberI might also try for a project file that just compresses all the associated files together
05:44:00bieberThen you could generate a theme distribution from that
05:44:11LloreanJdGordon: Would it make sense to make that into a genuine value somewhere (like in the .wps) so that Rockbox rejects WPS files that would draw viewports or whatnot off screen?
05:44:14CIA-6New commit by jdgordon (r26641): Changeover to the new skin format. SkinBreakingChange in the wiki has the runown of the changes. ...
05:44:17bieberI'll mull it over tonight, have to go for now
05:44:31JdGordonLlorean: no, I dont think so
05:44:49LloreanWhy not?
05:46:08JdGordonbecause doing that isnt catastrofic
05:46:51 Quit Topy44 (Ping timeout: 248 seconds)
05:48:19 Join Topy44 [0] (
05:49:25LloreanLoading a theme designed for a screen size larger than your screen can cause some noticeable issues. Why not include "screen size" as an actual property of the theme?
05:50:19JdGordonnothing dangerous though.. it wont stop you from using rockbox and being able to rundo it
05:50:32JdGordonIt might not even work at all... I cant remember how strcit the viewport parsing is
05:53:53 Quit tom` (Ping timeout: 240 seconds)
05:55:41CIA-6New commit by jdgordon (r26642): fix errors
06:00:40CIA-6New commit by jdgordon (r26643): this should be the last of the warnings
06:00:51***Saving seen data "./dancer.seen"
06:07:08 Join tom` [0] (
06:10:47JdGordon%Cl updater with the updater might not work with svn... it seems the change in the parser for it didnt get into that diff... I'll fix that after lunch
06:11:49 Join hamish_ [0] (~hamish@
06:11:51CIA-6New commit by jdgordon (r26644): bah
06:13:36 Quit hamish_ (Client Quit)
06:16:23 Join kaylinsigswort [0] (
06:22:52 Join simonrvn [0] (
06:31:06CIA-6New commit by jdgordon (r26645): undo accidental commit from the big one
06:40:53CIA-6New commit by jdgordon (r26646): make the parser work with the changed %Cl format. ...
06:46:12 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
06:55:34 Join Buschel [0] (
06:58:05Buschelsaratoga: you there? did you already see the libmad measurements w/o COP?
06:58:40saratogaBuschel: was just typing my response
06:58:50saratogaThe original ARM optimizations were for the ifp port, which was an old device with 64KB of IRAM and 1 MB of RAM, so it was probably meant to get the DCT entirely into IRAM.
06:59:38saratogaas long as you're looking at these things, can I interest you in the ffmpeg dct32 routine:;a=blob_plain;f=libavcodec/mpegaudiodec.c;hb=HEAD
07:00:02saratogai was just counting its add/muls now
07:01:55 Quit mc2739 (Quit: leaving)
07:02:28saratogaactually going over that code it looks like it does a lot more muls
07:03:04 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
07:05:05saratogathe ffmpeg mp3 decoder is supposed to be a bit faster then stock libmad on x86, but everywhere i look it seems like it choses a worse algorithm
07:05:25saratogai wonder if theres just some really stupid thing libmad does that makes up for all these little advantages it has
07:05:46Buschelsaratoga: it looks like the dct32 which I used in mpc before. you can recognize this in the COS-tab (there is a 10.19 value visible). this implementation needs some pre- and postshifting to avoid overflow in interal operations (see BF for COS0_15).
07:06:08Buschelsaratoga: I changed this version to the current one, which is kind of derived from libmad's
07:07:36saratogai think i looked at the imdct libmad uses and decided it was optimal from a mul/add stand point
07:07:43saratogaany other ideas what can be improved in libmad?
07:08:24saratogafurther improvements won't help PP so much, but they're quite welcome on the arm9 targets
07:12:01Buschelsaratoga: maybe it is possible to re-arrange dct32 and make use of ldm/stm. nevertheless this will not improve speed that much −− I remember dct32 to need about 5 MHz on mpc, libmad should be similar with -O1
07:12:25saratogathat'll help a little on PP, but won't make much difference for arm9
07:12:39saratogait ldm is no faster there, though it saves a bit of memory
07:13:03saratogalet me test your patch on arm9e
07:14:10Buschelsaratoga: I also remember libmad to need MHz's for synth and dct32 that were nearly identical to mpc (which is more or less maximum optimied :). I do not think we should research this area −− except for code beautification. The filterbank implementation is a mess...
07:15:15Buschelsaratoga: I cannot make statements regarding the imdct implementation. Is it reasonably fast compared to aac?
07:18:16saratogaits hard to compare since its very small
07:18:26saratoga36 point verses 2048 in aac
07:18:36saratogabut the algorithm it uses is known to be very nearly optimal
07:18:55saratogaaccording to my googling of IEEE transactions papers from the 1990s
07:19:18Buscheldo we have measurements of the CPU-needs for libmad's imdtc?
07:20:05saratogawhen i tested 2 years ago it needed 27% of CPU time on PP IIRC
07:20:13saratogabut that number is from memory so i could be off
07:20:31Buschelso, about 10-12 MHz
07:21:00BuschelWe should re-measure this.
07:21:21saratogachecking on arm9 would be good too, since its radically different then pp
07:22:43saratoga34.94MHz on Clipv2 with your patch
07:22:58Buschelso, from recalling the profiling looks like => synth-dewindow=12MHz, dct32=5MHz, imdct=12MHz => 29MHz. There are still 10 MHz left.
07:23:22Buschelwell, profiling it again would be better than recalling ;o)
07:25:24saratogabitstream parsing, frequency inversion, etc
07:25:43saratogaall that weird mpeg crap
07:26:08saratogai'll try commenting out synth_full and running test_codec on that
07:26:25saratoga(on clipv2)
07:26:36Buschelhow fast is libmad on clipv2 using svn?
07:27:00saratogai'm not sure, funman keeps changing the iram clock
07:27:38saratogathat seems odd
07:28:04Buschelsomething went wrong with applying the patch?
07:28:47saratogaoh i used regular arm-elf-gcc
07:30:20saratogahow do i configure it to use eabi?
07:32:56 Join webguest14 [0] (
07:33:45 Join esperegu [0] (~quassel@
07:33:51JdGordongevaerts: your build client is broken :)
07:35:19Buschelsaratoga: configure −−eabi
07:35:33Buschelsaratoga: have to go now (first to dentist, then to work)
07:35:41Buschelwish me luck for the dentist...
07:35:53saratogaok i need to sleep anyway
07:36:10Buschelsee you later :)
07:36:11 Quit Buschel ()
07:36:28saratoga35.61MHz if I just comment out the ASM code
07:36:47saratogawill try eabi now
07:38:26 Quit webguest14 (Quit: CGI:IRC (Ping timeout))
07:40:18 Join LinusN [0] (linus@rockbox/developer/LinusN)
07:40:45saratoga34.11 MHz
07:40:59saratogaso its 900kHz faster on ARM9E too
07:44:21saratoga16.87MHz if I disable synth_full completely on EABI
08:00:53***Saving seen data "./dancer.seen"
08:08:17 Quit parafin (Quit: So long and thanks for all the fish)
08:15:28 Quit BHSPitMonkey (Quit: Ex-Chat)
08:30:02 Quit TheSeven (Ping timeout: 264 seconds)
08:34:58 Join flydutch [0] (
08:39:08 Join kugel [0] (~kugel@rockbox/developer/kugel)
08:40:43 Join ender` [0] (
08:42:51 Quit kaylinsigswort (Remote host closed the connection)
08:43:00 Quit detaos (Ping timeout: 258 seconds)
08:43:01 Join Zagor [0] (
08:43:02 Quit Zagor (Changing host)
08:43:02 Join Zagor [0] (bjst@rockbox/developer/Zagor)
08:45:36 Quit DerPapst1 (Quit: Leaving.)
08:49:39 Join detaos [0] (
08:52:56CIA-6New commit by jdgordon (r26647): temporariliy use S for the timeout param because the parser doesnt handle decimals yet
08:53:10literal< ocin> it looks the same like rockbox_default
08:53:33literalocin: that's because the iAmp .wps file has errors, so it fails to load, and you get rockbox_default instead
08:55:25JdGordonarg, there is a problem with my skin updater... it always escapes ;'s which mean sublines wont work
09:00:10bieberJdGordon: Do you want decimals in the parser?
09:00:14CIA-6New commit by jdgordon (r26648): the updater uncondiotnally changed ; to %; which meant alternating lines stopped working.. hopefully this fixes them
09:00:29JdGordonbieber: limited decimal support
09:00:42JdGordonI think its better to keep that as a S anyway
09:00:48bieberOh, okay
09:01:01JdGordonthe value needs to be 0.2 * SOME_SPECIAL_DEFINE
09:07:42 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:17:08 Join stoffel [0] (
09:22:21 Join DerPapst [0] (
09:26:01JdGordonthe sooner we drop the legacy parser in the core the better :/
09:31:18 Quit stoffel (Remote host closed the connection)
09:32:30CIA-6New commit by jdgordon (r26649): make %t work
09:36:15 Join petur [0] (~d408b802@rockbox/developer/petur)
09:38:11 Quit kugel (Ping timeout: 272 seconds)
09:41:24 Join B4gder [0] (~daniel@rockbox/developer/bagder)
09:46:27 Join Rob2222 [0] (
09:49:59 Quit Rob2223 (Ping timeout: 265 seconds)
09:51:09 Quit JdGordon (Quit: Leaving.)
09:51:12 Quit JdGordon| (Remote host closed the connection)
09:54:30 Join JdGordon [0] (~7bf38c1f@gateway/web/freenode/x-rodsthxfxgrdkugo)
09:57:58CIA-6New commit by bieber (r26650): Theme Editor: Added font selection
10:00:51bieberJdGordon: I'm getting some errors opening up random files from the wps directory in the theme editor, is that to be expected?
10:00:55***Saving seen data "./dancer.seen"
10:01:54simonrvnError on line 7: Argument list expected for instance ^_^
10:02:53 Join _jhMikeS_ [0] (
10:02:53 Quit jhMikeS (Read error: Connection reset by peer)
10:02:53 Quit _jhMikeS_ (Changing host)
10:02:53 Join _jhMikeS_ [0] (~jethead71@rockbox/developer/jhMikeS)
10:04:10bieberlol, yes, like that one
10:04:18JdGordonbieber: simonrvn: which wps?
10:04:23JdGordonand with latest svn?
10:04:49bieberYeah, latest SVN
10:04:59bieberiCatcher is giving me an error
10:05:03JdGordonmake install again
10:05:13simonrvnone based on cabbie-enhanced-info. i wouldn't worry about that ;). i think it has to do with the file browser, rather than a wps per se.
10:05:54CIA-6New commit by bieber (r26651): Theme Editor: Made text editor scroll horizontally
10:06:04bieberAck, you're right, I do need to get line numbering in there
10:06:16simonrvnmy bad, any of them. cabbiev2, mine,
10:06:52simonrvnsame, latest
10:07:10JdGordonwhich target?
10:08:07JdGordonok hang on... booting my desktop back into linux
10:08:13bieberRight now I'm looking at iCatcher.128x64x1.220x176x16
10:08:29bieberI think they're all giving me some errors though...on an insanely long line, no less
10:08:52JdGordonyour editor is failing on icatcher?
10:08:53simonrvnit doesn't seem to affect anything
10:09:19 Join JdGordon1 [0] (
10:09:39*simonrvn needs to sleep
10:09:41bieberJdGordon: Yeah
10:09:59JdGordon1arg, I gotta go... back in an hour or so
10:10:06bieberYou can shift-select all of them in the open dialog and just flick through the tab bar to see which ones are failing
10:10:13 Join efyx [0] (
10:10:29JdGordon1the update probably didnt work perfectly :/
10:10:40bieberIt could be parser bugs, too
10:10:46 Quit phanboy4 (Ping timeout: 265 seconds)
10:10:59bieberI'll take a closer look at some of them tomorrow
10:11:02 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
10:13:07 Quit JdGordon (Ping timeout: 252 seconds)
10:13:41bieberJdGordon: I just found that boxes.220x176x16.wps is using some tags (%Vf and %Vb) that aren't in the tag table
10:15:07JdGordon|damn i thought i commiyed tjem
10:15:20JdGordon|thrg are I
10:15:44 Join Highlander [0] (
10:19:18 Nick fxb__ is now known as fxb (
10:20:37 Quit Jennifur (Ping timeout: 272 seconds)
10:22:06 Join Jennifur [0] (
10:23:38 Part LinusN
10:30:06 Join LinusN [0] (linus@rockbox/developer/LinusN)
10:42:54 Join MethoS- [0] (~clemens@
10:43:50 Join deleete [0] (
10:44:00 Quit deleete (Client Quit)
10:44:29 Join webguest40 [0] (
10:52:49 Quit FlynDice (Remote host closed the connection)
10:55:25GodEaterUSB Multimedia key mode doesn't appear to play nice with Windows 7 / foobar2000
10:55:40GodEatervolume control works, but playback controls don't
10:58:37 Join notlistening [0] (
11:04:01 Quit Jennifur (Ping timeout: 276 seconds)
11:08:49 Join einhirn [0] (
11:11:07*S_a_i_n_t_ wonders if JdGordon1, JdGordon| is still around.
11:11:30S_a_i_n_t_Primarily, wondering what he was "oy" 'd at about ;)
11:14:21S_a_i_n_t_Arg! S_a_i_n_t_ sees that if he wants this build to work, that he should get around to updating his theme files to the new syntax.
11:17:21 Join chrissavery [0] (
11:23:17chrissaveryHello - I'm interested in any feedback on FS #11270 - PictureFlow - WPS Integration Patch. This patch may be ready for svn. It is a set of 7 patches that provide some setting options for using PictureFlow in a more seamless way. It also has several bug fixes and smaller enhancements. Comments welcome...
11:24:19 Part Zagor
11:26:46S_a_i_n_t_JdGordon1: Having some problems with
11:26:56S_a_i_n_t_it never actually outputs anything. :/
11:34:16 Quit webguest40 (Quit: CGI:IRC (EOF))
11:35:08 Quit Kitr88 (Ping timeout: 248 seconds)
11:35:27 Join Kitr88 [0] (
11:39:30Lloreanchrissavery: I haven't read *all* the comments yet, but I don't like the idea of it going to Pictureflow when playback ends/stops.
11:40:02LloreanEven if it's just when Pictureflow is the hotkey, it means that if someone starts playback via filetree but has pictureflow as the hotkey, they end up in pictureflow after and that's not really ideal.
11:42:11chrissaveryis there a way to detect if wps was started via file/database/pictureflow. The reason it does this is because otherwise to get back to pictureflow you have to go to root>plugins>demos>pictureflow. So this assumes you want tighter integration. When it goes to pictureflow on stop I think at that point a single << key gets you into whatever was last menu.
11:42:35Lloreanchrissavery: I think that's preferable to dropping people in pictureflow when they may not even have run it this boot though...
11:43:03LloreanIf you can fix the integration so that it remembers where the WPS was launched from later, that's great.
11:43:25LloreanBut this is kinda an ugly hack that shouldn't really be an "until we get it better, let's do this" kinda thing - the "until we get it better" should be preserving the safer current behaviour there
11:44:09 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:45:40LloreanUsers have no reason to expect changing the hotkey function changes what happens when they leave the WPS - if the WPS is going to return somewhere else it should at the very least be a very explicit option of its own (but, as I said, a proper solution is probably best).
11:45:57 Join swilde [0] (
11:47:44chrissaveryActually I don't consider it undesirable. The reason I created the wps integration was because I wanted to use it as my main interface - default album select mode. So having it go there is exactly what I wanted for a more visual and friendly interface. I would have to explore if there is a return state upon exit of WPS that I can use to decide where to return to. But personally I would prefer it as is since PictureFlow is my pref
11:48:29LloreanYes, you would
11:48:42LloreanBut my point is that it should do everything it can to not cause unexpected behaviour for people who *don't* want this.
11:49:00Llorean"Pictureflow on hotkey" is a separate feature from "return to pictureflow from WPS" and should be treated as such
11:49:07 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:49:27LloreanIf you feel it needs to go back to pictureflow from the WPS, it should be done the "right" way. But the hack should be left out.
11:50:04chrissaverywhat "hack" are you talkign about?
11:50:31LloreanThe one of returning to pictureflow based on the hotkey setting.
11:50:47 Quit S_a_i_n_t_ (Ping timeout: 264 seconds)
11:52:49 Join S_a_i_n_t [0] (S_a_i_n_t@
11:54:30Lloreanchrissavery: If I've misunderstood how it works, I'm sorry.
11:54:50LloreanI may have overlooked something in some of the comments, a lot of them seemed to be about icons and other things.
11:55:41 Join liar [0] (
11:56:32chrissaveryWhat is the current behaviour upon playlist ending? I thought it was going to whatever the hotkey is set as. At least, when I tested I found that it always went to the screen that my hotkey was set for. Perhaps it is currently just returning to whatever started WPS? If so then I would need to add a GOTO_PICTUREFLOW root menu value so that when WPS is started with that value it will return via root to pictureflow rather than just s
11:56:32 Join Zagor [0] (bjst@rockbox/developer/Zagor)
11:56:50Lloreanchrissavery: Yes, it returns to wherever the WPS was launched from.
11:57:47 Quit MethoS- (Remote host closed the connection)
11:59:13 Join robin0800 [0] (
11:59:44chrissaveryI'll look into this. I added a GOTO_PICTUREFLOW for enabling pictureflow start screen so that may make it easy to see if a return mode could be used rather than a direct load.
12:00:18 Join pondlife [0] (~Steve@rockbox/developer/pondlife)
12:00:57***Saving seen data "./dancer.seen"
12:06:57 Quit GeekShadow (Ping timeout: 240 seconds)
12:14:57 Join markun [0] (~markun@rockbox/developer/markun)
12:17:04S_a_i_n_tbieber: you here/
12:18:34S_a_i_n_tNot sure if you have anything to do with it or not...but, I found a parsing error in skinupdater.exe, not sure if that's solely JdGordon's domain or not.
12:19:25S_a_i_n_tIt seems it doesn't like it when the .sbs ends with a comment, or in my case *many* commented lines.
12:19:57 Quit robin0800 (Remote host closed the connection)
12:20:29 Join watto [0] (~watto@
12:21:15S_a_i_n_tAaaanyway, here's what it did to my comment box at the end of the .sbs
12:21:30S_a_i_n_tit seems like it actually choked on the UI viewport.
12:23:50S_a_i_n_t*there*'s what id did the the UI viewport rather, easy to fix, but strange.
12:28:52 Part LinusN
12:36:14*JdGordon1 returns
12:36:32 Join gibbon_ [0] (
12:36:47gibbon_hello rockbox world
12:37:25JdGordon1S_a_i_n_t: haha wtf? is back up and hopefully ready to build again... regarding bjorns mail, i also added the eabi compiler
12:39:17gibbon_what advantage to we hope to get from this?
12:41:02GodEaterUMS is bloody choppy on Windows 7 too :(
12:42:47 Join webguest95 [0] (
12:45:19CIA-6New commit by jdgordon (r26652): add the Vf and Vb colour tags to the theme editor
12:47:24CIA-6New commit by jdgordon (r26653): fix FS #11368 - %V was eating the whole line so the %Vf() and %Vb() tags were being ignored
12:48:02gevaertsJdGordon1: I've stopped it now. It can't be broken though, I followed the instructions very precidely!
12:49:31GodEaterprecidely ? :)
12:49:33JdGordon1chrissavery: IIRC it is possible to know where the music was started from
12:49:52Zagorgibbon_: binsize improvement, mainly
12:51:28JdGordon1S_a_i_n_t: I am entirely to blame for the updater... pstebin the erroring sbs
12:53:23gibbon_Zagor: ok, thanks
12:54:05chrissaveryjdgordon: I'm just changing transfer code between wps-pictureflow now. When I started coding it was all new and I didn't know that control passed through root_menu.c. I have a version now that behaves better and uses normal wps returning. It's more standard now but not the way I'd prefer so I'm considering an setting to force stop -> pictureflow. Std works fine if you stick to just wps-pictureflow flips but if you decide to go to
12:55:18chrissaveryjust exploring now.
12:56:46JdGordon1I havnt looked at that code in ages, but I think you just need to set the preivous_browser variable when entering pictureflow, when the wps stops music it returns GOTO_PREVIOUS_BROWSER so magi happens
12:56:50JdGordon1I *think*
12:56:52Lloreanchrissavery: There isn't a setting to pick where "stop" returns you to normally. If one's added it should be a separate patch that gives you all the options, rather than pictureflow specific
12:56:53 Nick JdGordon1 is now known as JdGordon (
12:57:06Lloreanchrissavery: Generally you should try to change as little as possible, and split along feature lines for different patches.
12:59:03JdGordonchrissavery: I'm free now to look at the minimal changes if some of those patches are ready
13:00:05JdGordonbah, I see the early patch is the one which touiches the core
13:00:10JdGordonthat makes things difficult
13:00:29chrissaveryI'm thinking when you set Hotkey it could be "PictureFlow" or "PictureFlow Forced" - so that the user can decide what flavour they like. But I'm still looking for a return value that will provide default desired behaviour. When it returns from wps with last_screen it only works when directly entered last screen from pictureflow, so it doesn't become a kind of default screen. I see a old_previous_screen that maybe will work.
13:01:25*JdGordon is looking at your patch
13:01:56chrissaveryJdGordon: You may want to look at #2-#7 as they are all pretty small. #1 is the big wps integration and I'm thinking a small patch #8 to come to fix how returning works. If I chg #1 then it will possibly break #2-#7 and I don't want the headache of syncing all them again...
13:02:12Lloreanchrissavery: Where you return to absolutely shouldn't be part of the 'hotkey' function
13:02:18JdGordonin your change to root_menu.c all you need to do is before starting the plugin, do previous_browser = next_screen
13:02:20LloreanThat's not what "hotkey" is for.
13:02:58JdGordonI really dont like the plugin_load() line beign there 4 times in the patch...
13:03:16chrissaverywhich patch#?
13:04:11CIA-6New commit by uchida (r26654): text viewer: can selected scroll by column/scroll by screen for the horizontal screen. ...
13:04:49JdGordonyeah, looks like you just need to set previous_browser = GO_TO_PICTUREFLOW: before breaking that case and it will work
13:05:02JdGordonassuming next_screen after exiting is GO_TO_WPS
13:06:17chrissaveryya, that's what I'm looking at right now. It could be just two places - one for root_menu.c: load_plugin_screen and then once for hotkey code, though even that could be removed if I can code in a wps return value GOTO_PICTUREFLOW that lets only root_menu do the loading.
13:06:21JdGordonthe change to wps.c is not needed
13:06:40JdGordonthat would be ideal
13:07:36 Join ucchan [0] (
13:08:00 Join teru [0] (
13:08:42JdGordonI also dont like "while (!tagcache_is_usable()) sleep(HZ);"
13:10:09JdGordonsplashing "db not ready" or "loading" would be FAR better
13:10:26JdGordonand checking for action_user_abort() to cancel the loop
13:10:44JdGordonalso you have to make sure the db is even initialised
13:13:52ucchanrasher: login ?
13:13:57CIA-6New commit by jdgordon (r26655): off by one. FS #11369
13:14:01chrissaveryin my changes now it has an ifdef so that GOTO_PICTUREFLOW case is only if have tagcache
13:14:39JdGordonok, but the database isnt automatically intiialised
13:15:04JdGordon*yes* if you use pf you will hve it inited but that still isnt a safe check
13:15:11JdGordonthat is an infinite loop otherwise
13:15:13chrissaverythat tagcache_is_usbale line is because during start screen it has to wait for tag_cache init otherwise pictureflow returns dead. maybe that waiting should occur in pictureflow not in root_menu
13:15:58JdGordonno, it is ok in root_menu... it just needs to be cancelable
13:16:12chrissaveryI was going to add a max wait time there but haven't done it yet. I'd prefer if it didn't marr the startup with a message...
13:16:42chrissaveryI should look at user_abort method - I didn't knwo there was such a thing.
13:17:45chrissaveryalso now changing code in onplay.c so it sets return value instead of direct plugin_load.
13:18:10JdGordonis perrikwp here?
13:21:13*JdGordon grumbles about lack of willing testers
13:25:16GodEaterugh, UMS is SO bad on Windows 7 I can't even use it :(
13:29:05chrissaveryJdGordan: I'll have an updated #1 soon that fixes a number of these issues. It won't chg pictureflow.c at all so won't break #2-#7 fortunately.
13:29:27JdGordonI've been distracted by skin bugs, but Ill be here for a while
13:29:59JdGordonchrissavery: what timezone you in?
13:30:27chrissaveryI'm in Bangkok, Thailand so it's UTC+7 or 6:31pm here now.
13:31:25JdGordonok cool
13:33:55chrissaveryI keep weird hours - will be around for many hours now.
13:39:48JdGordon+7 works well for me, I'm +10
13:45:04CIA-6New commit by jdgordon (r26656): Fix FS #11370 - BEWARE when useing viewport colours. ...
13:45:23JdGordonok, back to pf
13:53:32 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
13:57:40 Quit B4gder (Quit: It is time to say moo)
14:00:59***Saving seen data "./dancer.seen"
14:03:34CIA-6New commit by teru (r26657): text viewer: some fixes related to font. ...
14:15:40 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:29:19 Quit kugel (Ping timeout: 265 seconds)
14:35:08 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:37:34 Quit ucchan (Quit: Leaving...)
14:41:00CIA-6New commit by jdgordon (r26658): this probably fixes FS #11373
14:45:40 Join perrikwp [0] (
14:50:52 Join marc2003 [0] (
14:51:39perrikwpjdgordon: r26658 did fix FS #11373
14:51:59JdGordonok, show me screenshots and the wps
14:52:46perrikwpok in just a couple of minutes
14:57:39kugelJdGordon: "probably"? but you do an actual test, don't you?
14:58:03*kugel thinks commit messages shouldn't be chat messages
14:58:18JdGordonwell seen as I didnt get a great description of the problem it wasnt exactly easy to know if that bug was the one or not
14:59:19kugelJdGordon: maybe you could be a bit more descriptive on some commits, e.g. r26644?
14:59:51kugelI would appreciate it
15:00:02perrikwpdo you want the wps code and a screenshot of the fixed wps or a before/after screenshot? Also, sorry if I wasn't descriptive enough in the bug report.
15:00:44JdGordonscreenshot of the broken wps
15:00:53perrikwpoh ok
15:01:08JdGordonwait, fixed? was this a problem with the updater or a real bug?
15:01:18CIA-6New commit by teru (r26659): text viewer: gigabeat f: add buttons to do one line up/down, and one column left/right.
15:01:20perrikwpa real bug
15:01:51kugelbtw, shouldn't the updater be used very soon before people start uploading manually fixed themes?
15:03:58perrikwpI do think the updater has a problem also with alternating sublines. It inserted a '%' between each subline, I manually fixed that. Unless that was on purpose.
15:04:37JdGordonyes, I know about that problem... not really sure how to fix it
15:09:32 Join Jaykay [0] (
15:10:08perrikwpI do have another wps I have tried updating, but have gotten stuck. It partially works, as in only a couple of viewports show up. Maybe you could help me debug it?
15:10:40CIA-6New commit by teru (r26660): fix red.
15:12:03JdGordonI'm more interested in these phantom )'s
15:12:31marc2003jdgordon: cabbiev2 is broken with r26658:
15:12:43marc2003that's r26652 on the left
15:13:09marc2003i'm not sure which one inbetween broke it. :p
15:13:18JdGordonwhich is correct?
15:14:47perrikwpcabbiev2 works fine here with r26658 on a Gigabeat S
15:16:05marc2003r26652 on the left is the correct one
15:17:29perrikwpAlthough the problem your having looks exactly like the problem I'm having with trying to update my other theme gigaWidgets
15:17:57 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:18:10perrikwpsome of the lines/viewports don't show up but others do
15:19:12chrissaveryJdGordon: Just posted patch #1a to replace #1 on FS #11270. Should be much more agreeable for integration. Simplifies changes and restores normal flow.
15:20:44JdGordonok cool
15:22:15kugelchrissavery: the mpegplayer has some backlight settings too (but more advanced), maybe that can be reused?
15:26:16chrissaveryI'm not sure why pictureflow was set with disabled backlight originally. I added the option to have it as normal. I'd just as soon have removed the disable code but thought this was more flexible because someone must have wanted it disabled. I'll have a look at mpegplayer - not sure what other backlight options would be useful though.
15:27:46 Join n1s [0] (~n1s@rockbox/developer/n1s)
15:28:31 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:30:25 Join jgarvey [0] (
15:31:33 Join kaylinsigswort [0] (
15:32:23chrissaverykugel: ah, I see it has brightness control. if people want that I could look at adding it. I don't need it myself. My main thing was to restore normal dimming.
15:34:16kugelyea, it just an idea
15:34:42pondlifeI'd think that Pictureflow should use the standard settings - it's really just a replacement for the album list (conceptually)...
15:35:15 Join hebz0rl [0] (
15:35:44pondlifeUlitmately, it would be nice if it could be driven from any suitable Database selection... i.e. where art is available
15:43:03chrissaverypondlife: on my todo list is options for a few different sort orders - artist+album, album, genre+artist+album. I'll be doing that once this current bundle is done. It would be nice if you could select an order in the database to use as default for pictureflow. Right now my tag_cache skills are poor. I can see the basics but I'm still figuring out how to do ordering with multiple tags. If anyone knows good example code for that, p
15:43:32pondlifeheh, hence my use of "Ultimately" ;)
15:44:09pondlifeI was mainly thinking of not complicating it too much - just look at it as a flashy list (albeit one that scrolls sideways, not up/down)
15:45:58chrissaverysure. it's a nice way to choose an album to play. which is why I wanted it integrated. it was too hard to get at using the menu.
15:47:11chrissaveryI was going to add a root menu item for it but saw that menu changes are somewhat taboo.
15:52:55pondlifeDoes it (still) require the database? I've been away...
15:58:29pondlifeIf so, then maybe a sensible menu access would be via Database > Album > (Pictureflow)
15:58:48pondlifeA special entry akin to <All tracks> and <Random>
15:59:44pondlifeThen the database code could pre-index for you, perhaps? (Says /me with no knowledge of how complex this would be...!)
16:01:03***Saving seen data "./dancer.seen"
16:01:09 Join bzed_ [0] (
16:01:24 Quit bzed (Ping timeout: 276 seconds)
16:01:31 Nick bzed_ is now known as bzed (
16:02:05 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
16:04:07kugelpondlife: that's relatively easy actually, I did it once
16:04:47kugelbut people said it shouldn't be done unless it offers the same features as the other entries (e.g. the context menu)
16:05:10pondlifeWell, it ought to. But that's possible, I'd think.
16:05:24pondlifeAnd not a showstopper if it doesn't initially.
16:05:30kugelthe context menu isn't as easy
16:06:28pondlifeNo, but adding this doesn't prevent people using the "normal" lists..
16:07:43 Join funman [0] (~fun@rockbox/developer/funman)
16:12:05chrissaveryI could see a context menu item that jumps into pictureflow with the current list of items. If it saved the list then pictureflow could load it as default. That would be fairly easy, maybe.
16:16:02pondlifeContext menus normally refer to the single selected item - so be careful how you'd fit that into the UI... a simple option within the Database listings would probably be easier.
16:18:02 Quit bluebrother (Disconnected by services)
16:18:04 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
16:20:10 Quit JdGordon (Changing host)
16:20:10 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
16:26:50 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
16:28:10chrissaveryIt could save the list of the items under the current selection, eg. Artist selected, saves the list you would get if you choose that. It's just an idea. In pictureflow it could have a setting for Normal order or Saved order as default. This would give a lot of flexibilty without writing a lot of new db code.
16:32:07pondlifeI was thinking you'd select <Pictureflow>, then the existing db code would work out the list for you. Then rather than displaying a standard list, it would pass this to pictureflow somehow (a file I assume). So the change wouldn't impact the db core at all, just the UI end of things.
16:32:30*pondlife keeps thinking of Rockbox-as-a-web-app ;)
16:34:49 Join chrissavery1 [0] (
16:35:00 Quit grndslm (Quit: Leaving)
16:35:28 Quit chrissavery (Ping timeout: 265 seconds)
16:40:08 Part Zagor
16:41:11saratogaFS #11374 - Spanish Translation (translated with Yahoo! Babel Fish)
16:41:13saratogawhat the heck
16:42:05kugelbrandon dowdy again
16:42:37kugelI think he posted a patch before in which he used google translate for afrikaans.lang :)
16:43:37 Quit slck (Ping timeout: 240 seconds)
16:44:38saratogahopefully my closing message was clear enough
16:44:57 Quit kaylinsigswort (Remote host closed the connection)
16:45:24chrissavery1is there some reason a "chrissavery1" shows up as a user here now? forgive my irc newbie-ness.
16:45:39chrissavery1I was just chrissavery a few minutes ago.
16:46:02kugelchrissavery1: you timed out (lost internet connection?)
16:46:19saratogaif you drop from IRC sometimes it does't realize it for a while, so your old name is still in the network and you need a new one until it flushes
16:46:39saratogairc is strange
16:46:41kugelsome clients reconnect before the server notice that you timed out, and then you get a different nick (which depends on the client) because the first one is still there
16:46:59Bagderit realizes it, it just locks the name for a while to make sure the effect isn't abused
16:47:20Bagderirc is a shaky protocol
16:47:22chrissavery1hmm. ok. pidgin strangeness I guess. no problem. I thought maybe someone was spoofing me now.
16:47:37kugelchrissavery1: '/nick chrissavery' should work
16:47:52Bagderand you can/should register your nick, then you can always get it back
16:48:00 Nick chrissavery1 is now known as chrissavery (
16:48:13chrissaverythanks. ha ha.
17:03:49 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
17:06:01 Part chrissavery
17:10:37 Quit domonoky (Read error: Connection reset by peer)
17:11:46 Join slck [0] (Venci@Slackware.SlackPix.Com)
17:11:47 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:13:40 Quit marc2003 (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
17:17:36 Join chrissavery [0] (
17:19:52 Part chrissavery
17:19:57 Join chrissavery [0] (
17:22:05 Quit teru (Quit: Quit)
17:26:01 Quit tom` (Read error: Operation timed out)
17:28:18 Quit webguest95 (Quit: CGI:IRC (EOF))
17:30:31 Join phanboy4 [0] (
17:30:41funman_jhMikeS_: <- where are those macros ?
17:31:31CIA-6New commit by mc2739 (r26661): Slovak language update - FS #11350 by Peter Lecky
17:32:54CIA-6New commit by mc2739 (r26662): Update of Czech language - FS #11371 by Marek Salaba
17:37:15S_a_i_n_tHmmm, unlikely...but JdGordon: around?
17:37:42JdGordonnot doing rb atm though
17:38:35 Join tom` [0] (
17:39:53 Nick mikroflops_ is now known as mikroflops (
17:40:12 Quit slck (Ping timeout: 276 seconds)
17:40:27 Quit markun (Ping timeout: 258 seconds)
17:41:13*funman broke microphone recording on fuzev1/e200v2 in 3.6
17:41:43 Join kugel_ [0] (
17:41:52notlisteningdomonoky, hey I think i've tracked down all the issue, well most anyways
17:42:01 Quit kugel (Disconnected by services)
17:42:05 Nick kugel_ is now known as kugel (
17:42:07domonoky:-) i will look into it later..
17:42:11 Quit kugel (Changing host)
17:42:11 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:42:28notlisteningdomonoky, yeah no worries, I'll pastebin that .sbs just for the sake of you knowing it's here though ;) <−−Full .SBS <−−Part where it spat the dummy.
17:43:55S_a_i_n_tSeemingly on the UI viewport, it seems to have mixed up the FG/BG colours also. :-?
17:44:21 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
17:44:23mc2739funman: it's broke in svn, too
17:44:32JdGordonok, I'll look in the morning
17:45:10funmanmc2739: yep it's r26444 in svn - locking is not enough when recording microphone on AMSv1
17:46:45 Join kugel_ [0] (
17:46:53 Quit kugel (Disconnected by services)
17:46:58 Nick kugel_ is now known as kugel (
17:47:00 Quit kugel (Changing host)
17:47:00 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:57:17S_a_i_n_tchrissavery: Around?
17:58:17 Quit kugel (Ping timeout: 240 seconds)
17:58:19S_a_i_n_tPerhaps it's just weird wording, but...are those pictureflow patches supposed to be installed sequentially(Urrrgh sp?)?
17:58:33 Quit petur (Quit: work->home)
17:58:46S_a_i_n_tI've just been using the last one each time, wording on IRC earlier makes me doubt that's correct now.
17:59:22 Join grndslm [0] (
17:59:44S_a_i_n_tIIRC, I have #15 in my build now...I wasn't supposed to install 2~15 ON TOP of #1 was I?
17:59:45chrissaverythe original ones are composite so only the final one should be done. but the last block on FS is where I break it down into separate patches for the purpose of getting in svn.
18:00:09chrissaverybreaking them down allows for reverting one if it causes trouble without removing everything.
18:00:30S_a_i_n_tjust a few comments I parsed (apparently incorrectly) made me question if they were standalone or not.
18:01:05***Saving seen data "./dancer.seen"
18:01:10funmanchrissavery: breaking them down also allows finding the one bit which causes trouble if it happens ;)
18:02:04chrissaverythe naming is slightly different, pictureflow-wps-# are composite so last one is good. pictureflow-#-<reason> is the breakdown. last full patch is picture-wps-16.patch which is the one broken down in the last two posts.
18:02:49 Quit GeekShadow (Read error: Connection reset by peer)
18:06:12funmanhm only the pcm-as3525.c bit is relevant
18:06:47chrissaveryAt this point I would stick with the final post #1a -> #7 ones as it has changes in how the integration is done.
18:08:37S_a_i_n_tHmmm, I have no problems with #15 (I think it's 15, I needed to sync it to my own I have my own versioning of it now which is slightly confising)
18:09:11S_a_i_n_tIf you could figure out why "insert" doesn't work from pictureflow for the Nano2g, I'd love-you-long-time.
18:09:28funmancabbiev2 progress information is a bit too low on fuze
18:09:42funmanthe shuffle/repeat icons hide the bottom part
18:10:35S_a_i_n_toh, bugger.. chrissavery, that was aimed at yourself if you didn't guess already.
18:10:38chrissaverysaint: you mean when you try to add to the playlist?
18:10:46S_a_i_n_tThough, I know you have a lot on your plate.
18:11:12S_a_i_n_tFrom what I can see, there should be no difference...but, there is.
18:11:21 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:11:29S_a_i_n_tNano1g == fine, Nano2g != fine.
18:11:31chrissaverydoes it work for replacing and playing current selected album?
18:11:41S_a_i_n_tyes, that works.
18:11:54S_a_i_n_tI just can't append to the playist with the Nano2g
18:11:57chrissaveryhmm. no idea off hand. I don't own any apple products. Just one Fuze v2 here.
18:12:13S_a_i_n_tFair enough.
18:12:39chrissaveryThere may be a button mapping that prevents the expected key from doing that function?
18:12:47S_a_i_n_tI've lokked at how it's doing the appending (slightly) and from what I can *should* work :/
18:12:59chrissaveryIs it just single track or whole album or both?
18:13:08S_a_i_n_tchrissavery: That's the thing, they both use the same keymap :/
18:13:16S_a_i_n_tso, I'm greatly confused about it.
18:13:37S_a_i_n_t<chrissavery> Is it just single track or whole album or both? <−− both
18:15:33mc2739funman: that works on e200v2
18:17:22 Quit bieber (Ping timeout: 252 seconds)
18:17:35 Join bieber [0] (
18:17:35CIA-6New commit by funman (r26663): as3525v1: fix r26444 which broke mic recording ...
18:17:44 Join Strife89|PalmTX [0] (~cstrife89@
18:20:04*S_a_i_n_t cries about the horrible state of disrepair his patches/tree is in following devcon/skinbreak stuff :/
18:20:12chrissaverysaint: just verified again that appending works on my Fuze
18:20:50 Nick fxb is now known as fxb__ (
18:21:33 Join toffe82 [0] (~chatzilla@
18:21:33S_a_i_n_tchrissavery: Thanks, but no need really. I run the same revisions of My Nano1Gs and m Nano2Gs...1G == fine, 2G == No worky with append to playlist :/
18:22:29S_a_i_n_tOh, by the way chrissavery: Pidgin *may* support this, not sure...Usually, if you just type the first letter of a nick and press TAB you can cycle through the available nicks.
18:22:46S_a_i_n_t*most* IRCclients do this, /heaps/ easier ;)
18:23:03chrissaveryS_a_i_n_t: cool. works.
18:23:22CIA-6New commit by funman (r26664): Remove CORELOCK_SWAP (unused since r16981)
18:23:26chrissaveryand there I was typing peoples names manually. <doh>
18:26:53 Quit pamaury (Remote host closed the connection)
18:28:31 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:37:20 Quit Strife89|PalmTX (Quit: Job hunt.)
18:37:45AlexPfunman: Do you think r26663 is worth a 3.6.1 for the fuse? Broken recording seems something worth a bug fix release
18:39:42funmanAlexP: i'm not sure, myself i only use svn, and this only breaks recording from microphone (FM is fine), and on e200v2 too (all the stable AMSv1)
18:39:59 Quit esperegu (Read error: Connection reset by peer)
18:40:10AlexPfunman: I don't know either, maybe I'll send a mail
18:40:13funmani guess we can at least wait a few days in case another problem is found in 3.6
18:40:17n1sif it affects stable targets i thing a bug fix release is a good idea
18:40:24 Join esperegu [0] (~quassel@
18:43:37 Quit domonoky (Ping timeout: 264 seconds)
18:43:41 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:44:27 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:45:29kugelfunman, AlexP: I think it's worth it yes
18:47:59 Join jhMikeS [0] (
18:47:59 Quit jhMikeS (Changing host)
18:47:59 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
18:47:59 Quit _jhMikeS_ (Disconnected by services)
18:51:43AlexPkugel: Is it generally done for all targets, or just the ones with the problem do you know?
18:52:04kugelI think only the affected ones
18:52:28kugelIIRC we had 3.something.1 for archos only because early usb broke
18:52:44AlexPah, rombox or something too?
18:52:58funmanah no this one was for all targets
18:53:00 Join lilmackie [0] (
18:53:05n1sit makes sense to only release a bugfix for affected targets
18:53:28lilmackiehey i'm having an issue with my sansa e270 after updating its build
18:54:32funman3.0.1 commits were made in 3.0 branch, and v_3_5_1 was tagged from it
18:54:35S_a_i_n_tissues such as...?
18:54:36 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
18:54:41AlexPYeah, OK so just fuze and e200v2?
18:55:44lilmackiewell i decided to also put some new themes on as well and i have no idea if that's what screwed it up.. but my wps screen is now white font on a black background, showing the title, album, time etc, but not formatted outside of the font
18:56:06lilmackiei've changed themes, i've changed the skin for the wps but it's not changing the black and white thing it's got going on
18:56:25S_a_i_n_tdoes the default "CabbieV2" theme work?
18:56:26lilmackieweird thing is that when i go to explore the database or w/e THAT has the current theme
18:56:48AlexPsaratoga: r.e. FS #11374, see FS #11284 and FS #11285 :(
18:57:10lilmackieyes actually S_a_i_n_t
18:57:24lilmackiebut it appears none of the others do =/
18:57:33*S_a_i_n_t starts to wonder about the skinbreak stuff...
18:57:47S_a_i_n_tDid you get theme from the themesite, or use Rbutil to install them?
18:57:50lilmackieis it because i used the rockbox utility to install the themes?
18:57:56bertrikAlexP, funman, and I guess clipv1 too, right?
18:57:57lilmackiefirst time too
18:58:03n1slilmackie: the skin handling stuff was redone today so that broke any old themes
18:58:04lilmackiebefore i've just done it manually
18:58:12funmanbertrik: afaik it's not released
18:58:19n1sdunno if the themesite has been converted
18:58:19lilmackieoh joy
18:58:21S_a_i_n_tlilmackie: Get them from the themesite manually.
18:58:31lilmackieso i should remove all the previous stuff?
18:58:43S_a_i_n_tit *should* overwrite it.
18:58:50S_a_i_n_tbut, if you want...then sure.
18:59:12S_a_i_n_tI assume you installed a "current build", and not 3.6?
18:59:22funmanbertrik: i think we should have made a 3.6 build for clipv1 anyway, i heard some people have it crash but i have seen no real bug report
18:59:31lilmackiecurrent build
18:59:49lilmackieoverwrite or remove S_a_i_n_t ?
19:00:05S_a_i_n_tlilmackie: Sweet, you can either go to the themesite and get the themes manually, or use Rbutil and only use the 3.6 build.
19:00:15S_a_i_n_tthis shannanegans will be fixed for 3.7
19:00:25S_a_i_n_tthere was a HUGE change in the skin syntax.
19:00:41S_a_i_n_tI think you'll be one of the first in a long line of people messing this up.
19:00:42lilmackieis there anything wrong with the current build other than the skinbreak?
19:01:21S_a_i_n_tthe skinbreak isn't "wrong", just "different, incompatable"
19:01:31 Join stoffel [0] (
19:01:31domonokythe themesite isnt converted at moment i think. So there are no working themes for svn there at moment.
19:01:34lilmackieif i downgrade it to 3.6 the themes should work without being manually fixed?
19:01:44lilmackiehow do i downgrade it?
19:01:50S_a_i_n_tinstall 3.6
19:01:51n1sinstall 3.6...
19:01:51AlexPJust install 3.6
19:01:59domonokyjust use rbutil and select the release build :-)
19:02:03lilmackieyes but how, can i do rbutil
19:02:06S_a_i_n_ttripple info-mania!
19:02:10S_a_i_n_tyes, you can
19:02:13lilmackiehold on a sec. lemme try this
19:02:20bluebroth3rbeware that voice files for release builds are currently broken (in Rockbox Utility!)
19:02:28 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
19:02:40S_a_i_n_tI suspect that this will get *pretty* annoying soon...
19:02:45S_a_i_n_t(skinbreak help)
19:02:52n1sdomonoky: would it be easy to fix rbutil to download a release voice for releases and a current voice for current builds instead of always using current voice (since that won't work now)
19:03:01lilmackiei don't use the voice files
19:03:10bertrikfunman, I would not bother making a clip v1 3.6 release now, but it's a nice goal to get it released with 3.7
19:03:11S_a_i_n_tthen, don;t worry abut it
19:03:16domonokyn1s: i think bluebrother already fixed that in svn
19:03:19CIA-6New commit by funman (r26665): backport r26663 : fix microphone recording on as3525v1
19:03:24domonokywe just need a new rbutil release...
19:03:26bluebroth3rn1s: I've committed that yesterday :)
19:03:30funmani just committed the fix in 3.6 branch
19:03:31literalthe theme site isn't showing any themes for the Clip+ because they all fail for the current build. Shouldn't it show all the ones that work with 3.6 at least?
19:03:37n1sdomonoky, bluebroth3r aha, very nice :)
19:03:40bluebroth3rhandling of different theme files is also already done.!?
19:03:58CIA-6New commit by bluebrother (r26666): Fix typo.
19:04:06CIA-6New commit by bluebrother (r26667): Distinguish between release and current build when installing fonts. ...
19:04:07 Quit TheSeven (Ping timeout: 252 seconds)
19:04:09S_a_i_n_twho knows what's going on with this skinbreak mess?
19:04:10bluebroth3rnow the same fix for fonts (though not critical) ...
19:04:16S_a_i_n_tthis is going to be a nightmare!
19:04:16domonokyliteral: i think the themesite still misses the 3.6 checkwps binarys... and clip hadnt a 3.5 release..
19:04:36literaldomonoky: yeah, it only shows 3.5 and current
19:04:46domonokyS_a_i_n_t: it wont... we will fix it in the next hours/days..
19:04:46n1sbluebroth3r: is anything holding back a release?
19:04:49bluebroth3rS_a_i_n_t: then simply go into a cave for a couple of days and come back when things have been sorted out :)
19:05:04AlexPS_a_i_n_t: Saying that repeatedly isn't helping, it is just adding pointless noise to this channel
19:05:07S_a_i_n_thehehe, I use my own themes, I'm not bothered :)
19:05:21bluebroth3rn1s: I'd like to do at least some basic testing first. And I've got to check if there are issues open :)
19:06:07bluebroth3rdomonoky: can I commit to the t.r.o tree without affecting the live system?
19:06:31domonokyit will only go live, if i do svn up on the server..
19:07:01S_a_i_n_tlilmackie: *apparently* you'll have to install 3.6 with rbutil, and then download the themes with rbutil also.
19:07:09S_a_i_n_tthis *should* work.
19:07:16bluebroth3rok. I'll commit my changes to rbutilqt.php then shortly
19:07:23n1sor just wait a few days for everything to be sorted
19:07:49S_a_i_n_tI had assumed it'd be done so people couldn't get into this mess, but it must have been more work than was thought at first.
19:08:04lilmackiek now
19:08:07lilmackiei have rbutil up
19:08:12lilmackiehow do i downgrade
19:08:15lilmackiei know i sound stupid
19:08:22AlexPJust install the release build
19:08:29n1swell, if you want stuff to work, use the relaeses, that's what they're for :)
19:08:38literalalso, the theme site uses a new themeid whenever a theme is updated, so the old link doesn't work, though it still keeps the same "downloaded $x times" stats
19:08:52n1s<lilmackie> how do i downgrade it?
19:08:53n1s<S_a_i_n_t> install 3.6
19:08:53n1s<n1s> install 3.6...
19:08:53DBUGEnqueued KICK n1s
19:08:53n1s<AlexP> Just install 3.6
19:09:00lilmackien1s: from site?
19:09:10AlexPUsing rbutil
19:09:13n1sor with rbutil, whatever you prefer
19:09:17AlexPSelect install build, then select release
19:09:21S_a_i_n_tn1s: Why would *I* instal 3.6?
19:09:30AlexPS_a_i_n_t: He was quoting you
19:09:40S_a_i_n_taha, gotcha.
19:09:48lilmackieit sez stable version archived build, current build
19:09:55AlexPyes, stable
19:10:01AlexPThe release is the stable one...
19:10:11lilmackiethe 3.6?
19:10:21lilmackieok nvm i see it
19:10:25AlexPJust click on it, it tells you
19:10:27lilmackieonce i selected it it sez 3.6
19:10:32AlexPwe know
19:10:33CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:10:33*lilmackie is such a noob sometimets
19:10:48 Quit stoffel (Remote host closed the connection)
19:11:09lilmackiethanks for agreeing AlexP :P
19:11:22 Quit n1s (Quit: Lämnar)
19:11:23AlexPyou are welcome :)
19:12:46 Quit DerPapst (Quit: Leaving.)
19:12:54lilmackieseems to have fixed it
19:13:03lilmackieand now i have AA o_O
19:13:32lilmackiek another question while i'm here
19:14:12lilmackiei decided to update the build because i found that my player was lagging. on the wps it wouldn't show the next song properly (it was stuck on a previous title/artist) or it would take awhile to forward through songs
19:14:17lilmackieany idea what causes that?
19:14:27lilmackiecuz i tried to initalize it or update the database and that didn't fix it
19:14:42AlexPthe database doesn't affect that
19:14:49AlexPWhat player do you have?
19:15:27lilmackiesansa e200
19:15:35lilmackiemore specifically e270
19:16:02AlexPforward through as in skip, or fast forward?
19:16:26lilmackieas in next song
19:16:49lilmackieit would show the now playing song fine... but the next song would read as a song that past like 20 tracks ago
19:16:56AlexPThat is odd
19:17:06lilmackieit was super bizarre
19:17:11lilmackieand began to really piss me off
19:17:13bertriklilmackie, can you reproduce that, or was it a one-time thing?
19:17:39lilmackiewhat do you mean reproduce?
19:17:49lilmackiei can't exactly make it happen... it would do it on it's own
19:18:16S_a_i_n_tdoes it happen if you skip multiple tracks in quick succession?
19:18:45lilmackieno. i thought that's what it was first, like it wasn't reading fast enough or the cpu was slow but it would happen even if i listened to a song all the way thru
19:19:38bertrikI meant if you have a list of specific things to do to make that problem appear again on your player.
19:19:59lilmackieturn it on and listen to music... it would do it whenever it felt like
19:20:11lilmackiebut it became more and more of an apparent problem because it was happening more often
19:23:43bertrikOk, that sounds like a bug in rockbox. Fixing bugs like that is usually much easier when there is some way to make the problem appear every time, so that's why I was asking.
19:24:09*domonoky just updated the release checkwps binarys to 3.6 on the themesite and runs them now over all themes...
19:24:56lilmackielets see if the new stable release makes a difference
19:25:08lilmackiestupid web irc
19:25:20lilmackiecuts off my letters
19:25:25S_a_i_n_tdomonoky: Is there a way to tell if the theme parses correctly? I had a pretty bad parsing error on one of my themes here, and if I'm correct my themes on the site should do it to.
19:25:40S_a_i_n_tI guess, download it an d see...yeah?
19:26:17domonokyS_a_i_n_t: you mean you got a parsing error after converting it ?
19:26:37domonoky(the themes on the server arnt converted yet)
19:26:53funmankugel: have you seen any volume difference between fuzev1 and v2 ?
19:26:55S_a_i_n_tOh,, it *mauled* it trying to convert it to the new syntax.
19:27:03S_a_i_n_tjust the UI viewport line though.
19:27:04 Quit liar (Ping timeout: 265 seconds)
19:28:06 Join Akranis [0] (
19:29:16lilmackieokie dokiie
19:29:18lilmackieim out fellas
19:29:23lilmackiethanks for the help!
19:29:48 Quit lilmackie (Quit: CGI:IRC)
19:32:44 Join Xerion [0] (
19:32:48 Quit jgarvey (Ping timeout: 240 seconds)
19:33:50CIA-6New commit by Domonoky (r26668): add a link to the themesite copy which holds the non-converted themes.
19:34:03CIA-6New commit by bertrik (r26669): Catalan language update and cleanup, FS #11366 by Adrián Cereto Massagué
19:36:09funmansaratoga: <- clipv2 recording looks different (I used 0dB on rockbox and max 'high' volume on OF)
19:39:20 Join liar [0] (
19:39:45 Quit esperegu (Ping timeout: 245 seconds)
19:40:11funman <- difference between OF & rockbox is more visible with this graph (although i'm not sure what it represents)
19:40:29 Join Zagor_ [0] (
19:40:30 Quit Zagor_ (Changing host)
19:40:30 Join Zagor_ [0] (~bjst@rockbox/developer/Zagor)
19:41:32funmandB probably
19:42:05bertrikr0b-, during devcon I could borrow pamaury's clip+, which seems to have the same radio as yours
19:42:41funmanan audio sample at 0dB is 0xffff on 16 bits sample depth?
19:42:51bertrikit turns out it contains a radio chip with a different keymap than the si4700 compatible chip found in most ams sansas
19:43:25CIA-6New commit by bluebrother (r26670): Add workaround to redirect to copy of theme site holding old themes. ...
19:43:48bertrikThis radio chip is most likely some other model from silabs (who also makes the si4700) because it has the same i2c address and some radio registers seem to be at the same location
19:44:14bluebroth3rdomonoky: you should be able to put the updated script online and old versions of Rockbox Utility will get redirected to the old site copy.
19:44:15funmanbertrik: : this radio chip (0219) is 4702 firmware version 19 ?
19:44:19 Join esperegu [0] (~quassel@
19:44:54bertriklikely it will take a few days/weeks until we figured out the new register mapping and are able to do first tests
19:45:04domonokybluebroth3r: good. will do that when the checkwps run is finished :-)
19:45:18 Join sandking [0] (
19:45:24ranmafunman: I wonder why I never looked at the amd5536udc driver, the geode does indeed use the same IP block for USB OTG as AS3525 AFAICS
19:45:25bertrikfunman, yes, that's a known one, which should work with our existing driver
19:45:29sandkingany linux users here?
19:45:44bluebroth3rwithout a new Rockbox Utility we will then have the same situation as now, except that you can update the "non-old" version of the site
19:45:59bluebroth3rsandking: mostly Rockbox users, but a lot of the developers use linux
19:46:22bluebroth3rbieber: a few comments on the latest theme editor changes:
19:46:23funmanranma: perhaps because you've not read the long wiki page ;)
19:46:24sandkingbluebroth3r: well - i'm looking for rockbox users that use linux :]
19:46:32funmanranma: any progress running the OF on clip+ ?
19:46:43ranmaI had read it, but I completely forgot about it :)
19:47:02bluebroth3rthe tab bar on the left in the preferences dialog looks weird. Did you use that intentionally? Most applications have that bar on the top
19:47:28bluebroth3ralso, I'd prefer to have a non-proportional font as default. But maybe people here think otherwise?
19:47:36*bluebroth3r wonders who already tries the theme editor
19:47:37funmanranma: there is some minor variations in registers between as3525 datasheet and amd5536udc (according to the amd datasheet at least)
19:47:42ranmafunman: No, I haven't tried anything further so far
19:48:23S_a_i_n_tsandking: It's better to just ask your question.
19:48:40bluebroth3rbieber: and how hard would it be to hilight syntax errors in the editor? Maybe by using a red background or similar?
19:48:40ranmafunman: Is there? I've been looking over the datasheet and so far the only discrepancies seem to be stuff that the as3525 patch doesn't document.
19:49:39 Join DerPapst [0] (
19:49:43sandkingS_a_i_n_t: yeah, i just realized that it's more Linux specific so I'm asking it also on Ubuntu. I just formatted iRiver h120 and I realized that I cannot boot properly. device is seen in palimpsest but doesn't mount properly - i just wonder how to mount it so i could install rockbox
19:49:50chrissaveryfunman: do you know if it's possible on fuzev2 to just twiddle the usb data pins (not as usb at all)?
19:50:14bluebroth3rbieber: is there interest in having svn binaries of the theme editor available so people can play around? Or do you think it's too early for that?
19:50:17funmanchrissavery: 'twiddle' ?
19:51:22chrissaveryjust toggle the pin high/low without regard for any protocol. so that if I connected a simple circuit to it I could trigger something.
19:51:33 Join jgarvey [0] (
19:51:49funmanranma: (from april 2009 forum post)
19:52:45funmanchrissavery: you want to use them from software? i have no idea on how to do this
19:54:08chrissaveryperhaps those pins are mapped to some register that would allow changing them outside of usb stack?
19:54:56literalis it possible to specify that the "rest of the line" is allowed to scroll, without using viewports? I don't want to use a viewport in this case because the beginning of the line is of variable size and I want to utilize the space efficiently
19:55:04funmani think there is more chance that you can control a pin of the dock connector rather than directly usb data
19:55:34S_a_i_n_tliteral: If you don;t use a viewport, the whole line scrolls, or none of it.
19:55:37 Join petur [0] (~petur@rockbox/developer/petur)
19:55:38chrissaverythat would be fine if other GPIO pins were exposed on the connector.
19:55:42literalalright then
19:56:49 Quit Akranis (Quit: Lämnar)
19:57:00 Join Highlander_ [0] (
19:57:37 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:58:32funmanchrissavery: might have some info
19:58:43funmanchrissavery: you have something you can dock on the fuze?
20:00:09 Join Jerom [0] (~Jerom@
20:00:47sandkinghow can i descramble the bootloader?
20:00:57sandkingi mean - from where i can get the tool?
20:01:07***Saving seen data "./dancer.seen"
20:01:20funmanrockbox svn should have all tools in /tools
20:01:25 Quit Highlander (Ping timeout: 265 seconds)
20:02:13sandkingfunman: ah, i got only install utility
20:04:19ranmachrissavery: It should be possible to control the pullup on one or both of the datalines
20:06:01funmanRob2222: where are you about fuzev2 which can't upgrade to a patched rockbox firmware?
20:06:27sandkingsorry - i can't find proper download of svn - can anyone tell me where to look for it?
20:06:56 Quit bieber (Ping timeout: 240 seconds)
20:07:12chrissaveryfunman: just reading through that link you gave me. I have a project I'd like to build that would use any 1 or 2 data pins on the connector to control a transistor switch. I can say more but maybe should post in community as it's kind of off topic here.
20:07:13 Join bieber [0] (
20:08:05funmanchrissavery: ok, i think custom project is better on #rockbox-community ; i was asking that because i don't think we know how to use sansa dock accessories properly on sansa AMS
20:08:54chrissaverythanks for the link. reading.
20:08:59sandkingi thought that installing rockbox was easier i don't have any idea how this descrambling works. i looked into tools folder and can't find it :/
20:09:14mc2739funman: I think Rob2222 returned his Fuzev2 and got one that he could install rockbox. Last I heard, it locked up and he was waiting for the battery to drain
20:09:46funmanhm ok, too bad he hasn't the old one anymore
20:10:27mc2739funman: see FS #11367 regarding dock & line out
20:10:37domonokysandking: for normal install you dont need scramble or else, just use rockbox utillity.
20:10:41bieberbluebroth3r: That sounds good to me
20:10:46S_a_i_n_tIt would have been nice to figure out what kept RB off Rob222's unit :/
20:10:56S_a_i_n_t*sigh* even
20:11:44funmanmc2739: ah nice, so plugging dock triggers USB insertion in ascodec register?
20:11:51saratogafunman: if you want to check the volume level, computing the replaygain value is a very easy way to get a single number
20:12:16funmancan i do that from audacity ?
20:12:17sandkingdomonoky: i installed longbox a long time ago on my iriver and somehow i've done it. now i try to use this install utility but it asks me for bootloader. i downloaded the bootlader from here but rockbox utility doesn't recognize the bootloader
20:12:50saratogafunman: i don't know about audacity, but lots of software can compute the replaygain value for a wav
20:12:55domonokysandking: no, it asks you for a original firmware.. please read the message again :-)
20:12:57*bieber typed in the wrong window :/
20:12:58mc2739funman: yes, I think taht came from ranma's ascodec interrupt commit
20:13:11saratogain Windows I just use foobar2000 to check volume levels
20:13:22sandkingdomonoky: well, yeah, but it doesn't accept the original one :|
20:13:24ranmafunman: USB insertion is triggered if there is power on the USB pins (charger)
20:13:40domonokysandking: and what is the error message ?
20:14:20sandkingdomonoky: after checking md5 it displays "firmware file not recognized"
20:14:54ranmas@the USB pins@USB VCC@
20:15:07saratogasandking: this is for the bootloader or the OF?
20:15:10sandkingi got this file
20:15:33S_a_i_n_toriginal firmware
20:15:34bertrikon the fuze v1, dock detection is on dbop pin 9 IIRC, I guess it's on a GPIO / button input for other sansas too
20:16:06domonokysandking: thats the rockbox bootloader, not the iriver firmware..
20:16:08sandkingi don't think i understand the question
20:16:31ranmamc2739: You probably want to setup the GPIO to generate interrupts on level change for dock detection :)
20:16:53sandkingbut it tells me that bootloader installlation requires me to provide original hex file
20:16:53saratogawhy would we ask you to download the rockbox bootloader? rbutil obviously knows what the rockbox bootloader is and where to get it
20:16:54bertrikDo we really need an interrupt for that, can't we just poll?
20:17:16saratogasandking: what happens if you downloada the hex file instead of the bootloader? does that work
20:17:33sandkingok, i think i get what i'm doing wrong
20:17:40sandkingproblem is i didn't get the original
20:18:09mc2739I didn't think we could get an interupt from dbop_din
20:18:12bluebroth3rbieber: the binaries thing?
20:18:30ranmamc2739: Ah, right, from dbop_din probably not
20:18:38 Part pondlife
20:19:14mc2739unless it is also connected to a GPIO like several of the buttons
20:20:43ranmaShould be portb bit5
20:21:15bluebroth3rI'll mostly build windows binaries though −− Linux users should be able to build themselves, at least as of the current state.
20:21:55bluebroth3rplus, I'd have to start building manually so it won't be done too frequently. Will try to keep a useful update rate though :)
20:23:21bieberThank you :)
20:23:42bluebroth3ryou're welcome
20:24:31 Quit CGL (Remote host closed the connection)
20:27:14ranmabertrik: Hmm, of course since dbop is already polled for buttons probably makes most sense to add dock support there :)
20:27:39 Join Highlander__ [0] (
20:28:17 Quit ender` (Read error: Connection reset by peer)
20:28:22 Join ender` [0] (
20:28:27 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
20:28:33 Quit Highlander__ (Client Quit)
20:28:57bertrikranma, yeah, I was thinking that too about dock detection (and similarly headphone insertion detection)
20:30:25 Quit Highlander_ (Ping timeout: 265 seconds)
20:30:48domonokycan some please craft me a short sh command line or script to run the recursivly on a folder with many folders with theme.zips in it ? fails for me.
20:31:24S_a_i_n_tJdGordon was gonna look at it this morning.
20:31:28bluebroth3rdomonoky: i.e. find all .zip files and run on it?
20:31:36S_a_i_n_t(about 4~5 hours)
20:31:53bluebroth3rfind . -name "*.zip" -exec ./ {} \;
20:31:57literalfind -name "*.zip" -exec ...
20:32:00literalbeat me to it
20:32:00domonokyS_a_i_n_t: ok, so i will wait with the converting a bit..
20:32:02bluebroth3r{} will get replaced by the filename found
20:32:32S_a_i_n_tIt seems to deflate the zip fine, but then...nothing.
20:32:47S_a_i_n_ttry with a single theme if you want, if it wors for you...great.
20:32:52S_a_i_n_tdidn't for me though.
20:33:17S_a_i_n_tI needed to use skinupdater.exe
20:34:41S_a_i_n_tI could be my themes specifically though, I have no idea. Though, if it is, you'll run into it on the themesite at least once.
20:36:11mc2739ranma: you were correct - it does appear to be GPIO B5
20:36:37 Quit esperegu (Read error: Connection reset by peer)
20:38:10 Join esperegu [0] (~quassel@
20:41:07*domonoky tried it with a test zip on the themeserver, and it fails because it misses the 'zip' command. Maybe its the same cause for you ? :-)
20:41:38S_a_i_n_tit unpacks/deflates whatever, then...nope.
20:42:23S_a_i_n_tI even made sure to do "./ <> <workingfolder>
20:42:42domonokyand no error message at the end ?
20:44:02S_a_i_n_tIt fails if <workingfolder> is already present too...which I though was odd.
20:44:17S_a_i_n_tas I tried just pointing it to a tmp foldr, but, no dice.
20:45:29domonokythat worked for me.. it created a <workingfolder>/<zipname> folder..
20:46:30S_a_i_n_tdomonoky: Wow...guess where it fails is truly erratic then? It gave some cryptic message about the working dir not being allowed to exist.
20:46:40S_a_i_n_tas it attempted to create it, then found it.
20:46:44S_a_i_n_tthis is weird :/
20:46:46domonokythe .pl file also doesnt move the zip back... so more work on that needed, before i can convert the themesite.
20:47:00*bluebroth3r needs to make a few adjustments to
20:47:13domonokyif the dir it wants to create exists it surely will fail...
20:47:19bluebroth3rbut it didn't need much adjustments to make it work with the themeeditor :)
20:47:25bluebroth3rdomonoky: mkdir -p
20:47:27domonokymaybe it failed to extract the themename for you ?
20:48:00S_a_i_n_tit wasn't anything exotic...I renamed it "" as a test.
20:48:23bluebroth3rbieber: maybe you want to give the wiki page a little update?
20:48:33 Quit flydutch (Quit: /* empty */)
20:48:35S_a_i_n_tI did exactly this "./ theme
20:49:00bluebroth3rwhat is the last argument to
20:49:10domonokyah, i tried that now to... that fails to get the correct path..
20:49:14S_a_i_n_t(quotes there for reference, didn't actually use quotes)
20:49:30S_a_i_n_tbluebroth3r: Last arg == working folder.
20:49:35funmanmc2739: if we can detect dock insertion/extraction, we could only enable/disable lineout at this point?
20:49:45domonokyit works if i use: ./ <fullpath to zip> <full path to temp>
20:49:45bluebroth3rdoes it work with something else? I.e. not the basename of the theme?
20:50:00S_a_i_n_twell, not for me.
20:50:02bluebroth3rnice :o
20:50:05 Quit toffe82 (Read error: Connection reset by peer)
20:51:00domonokyfrom the code it looks like it expects at least a path for the zip file..
20:51:02bertrikfunman, mc2739 I wonder if there's an existing rockbox thread that we can somehow make responsible for monitoring dock detection and headphone insertion and taking action accordingly
20:51:48domonokyso that shouldnt be a problem for me... but i dont have zip on the themeserver :-/ scorche, ping ?
20:52:31S_a_i_n_tdomonoky: I pestered JdGordon about correct syntax a few times, as I kept forgetting, the example I used (even checked it in the IRC logs) was stated as the one I mentioned.
20:52:44S_a_i_n_t./ workingfolder
20:52:51domonokyS_a_i_n_t: only the code has the truth :-)
20:53:08S_a_i_n_tyou'd think he'd know it if he wrote it though?
20:53:20domonokyuse ./ <path to zip>/ workingfolder
20:53:24 Join wincent [0] (
20:53:41S_a_i_n_tAha...kinda irrelevant now. ;)
20:53:55S_a_i_n_tI used skinupdater.exe eventualy :P
20:54:34S_a_i_n_twhich ended up choking on my UI viewport line massively though, bu tit was easy to fix by hand.
20:54:36mc2739bertrik: since dock connection triggers the USB detect interupt, won't we need to hook into that?
20:55:26 Quit guymann (Ping timeout: 260 seconds)
20:56:02ranmamc2739: I think the problem is that you have to change ascodec registers, which is easiest done from a thread.
20:56:05domonokyS_a_i_n_t:do youz know if that choking on UI viewport already fix in svn ? (i cant run it on the server if it doesnt convert properly)
20:56:22S_a_i_n_tJdGordon hadn't looke dat it yet.
20:56:28ranmaAt least if you have to change more than one register
20:56:29funmanbertrik: headphone insertion detection is possible on AMS, but not extraction
20:56:38S_a_i_n_the was busy when I told him about it, but he knows of it at least
20:57:14 Join guymann [0] (
20:57:20S_a_i_n_tit seems to only happen if there's a comment after the UI viewport line, but that would involve opening up every theme to check it. :/
20:57:28domonokyoki.. so the converting has to wait.. but at least i can prepare everything.
20:57:31bertrikmc2739, I'm not familiar with docks. I can imagine that the USB interrupt is only triggered when the dock is powered, while a dock detect event also happens with an unpowered dock.
20:57:54domonokyS_a_i_n_t: it has to be automatic.. there are a few hundred themes on the server..
20:58:06S_a_i_n_tyeah, that was my point ;)
20:58:08ranmabertrik: AFAICS the anythingbutipod thread suggests that a dock has to be powered to be detected
20:58:12gevaertsZagor_: the new doesn't seem to produce a working toolchain for me
20:58:26Zagor_gevaerts: what doesn't work?
20:58:40sandkingstrange - i downloaded original firmware from iriver and both patcher and rockbox utility tells me that the firmware isn't recognized
20:58:48ranmaSince it has to pull up on pin and not all models provide power for that on the dock connector
20:59:30mc2739bertrik: my dock does not show to be connected when unpowered
20:59:41bertrikSomehow I can't view the abi thread.
21:00:18bertrikmc2739, what do you mean? the player does not detect the dock in the OF?
21:00:43Zagor_gevaerts: do you have both arm-elf compilers installed?
21:01:00sandkingi got older one and it seems ok now
21:01:38gevaertsZagor_: I have arm-elf and arm-elf-eabi, both do the same thing
21:02:11gevaertsm68k-elf too
21:02:24Zagor_gevaerts: right, but did you remove the old arm-elf binaries before running the new
21:02:30bertrikI built the arm-elf-eabi last night and it seemed to work fine for me (ubuntu 10.04), I did have to remove the old one first IIRC
21:02:32gevaertsnot before, no
21:02:57gevaertsThey're gone now, but they were installed (and in the path) when I built the new
21:02:59Zagor_gevaerts: that's my best guess. it works fine for me
21:03:25mc2739bertrik: my dock does not trigger trigger a dock detect when the dock is unpowered. It triggers USB and dock detect when the dock has power
21:04:05*gevaerts tries
21:04:24 Quit esperegu (Read error: Connection reset by peer)
21:04:37funmangevaerts: check out your PATH
21:04:38mc2739my understanding that the dock detect is a voltage from the dap that is looped back into the dap on another dock pin
21:04:53funman'as' is installed in /path/to/prefix/arm-elf-eabi/
21:05:02funman'arm-elf-eabi-as' is installed in /path/to/prefix/bin
21:05:15gevaertsprefix is /usr/local here
21:05:27saratogait could be that powered docks use an external voltage rather then looping a power pin from the dock connector
21:05:45saratogaprobably easier that way since you need fewer pins on the connector
21:05:53saratogaif you just short the power supply rail
21:06:04funmani meant, 'as' is in /usr/local/arm-elf-eabi/bin and arm-elf-eabi-as in /usr/local/bin
21:06:08mc2739yes, that sounds likely
21:06:46gevaertsfunman: I know, I just wanted to clarify. /usr/local/bin is in my PATH
21:06:49bertrikit's not like the sansa dock connector has too few pins :P
21:07:05 Join esperegu [0] (~quassel@
21:07:09funmangevaerts: and not /usr/local/arm-elf-eabi/bin ?
21:07:23gevaertsThat one has never been in my PATH
21:07:33funmanhm it's not the same problem I had
21:08:16Zagor_gevaerts: no but /usr/local/arm-elf/bin was
21:09:24Zagor_I'm not entirely sure what is causing the problem, but having the old installed while building the new appears to cause problems
21:09:38Zagor_s/installed/in PATH/
21:09:42*gevaerts finds it non-obvious that the presence of arm-elf-as in the path makes the gcc build produce a binary that tries to use as instead of arm-elf-as :)
21:09:51bertrikI think I found a gcc built by somewhere in my path
21:10:35Zagor_gevaerts: no, /usr/local/arm-elf/bin contains 'as', not 'arm-elf-as'
21:11:19*S_a_i_n_t had some small confisuin regarding the "packaged by Rockbox" CygWin toolchains and the "generated by" CygWin toolchains being in two totally different places for a small time :/
21:11:21*gevaerts looks what the old path had
21:11:53funmanmc2739: if needed i can check if leaving line output always enabled affect battery life
21:12:26Zagor_S_a_i_n_t: yeah, in hindsight perhaps "correcting" the install path caused more problems than gains
21:13:33gevaertsZagor_: ok, so what's going on is that the switch to /usr/local/ from /usr/local/<arch>/ makes the non-prefixed tools get installed in the old path?
21:13:51Zagor_gevaerts: yes
21:15:02 Join Zigtown [0] (
21:15:04mc2739funman: thanks, that would help
21:15:29 Quit mc2739 (Quit: leaving)
21:16:23gevaertsZagor_: maybe the build server should accept "arm-gcc403" (and others) as aliases for "arm"?
21:16:32 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
21:19:14literalsupreme irony: the UnicodeGuide page in the rockbox wiki has encoding problems ->
21:19:24Zagor_literal: hehe
21:20:28Zagor_the website is unfortunately still iso-8859-1
21:21:30Zagor_that page looks to be using odd characters a little unneccesarily though
21:25:01S_a_i_n_tdoes anyone know why building a voicefile isn't smart enough to do "nothing to be done for `all`"?
21:25:36Zagor_S_a_i_n_t: because nobody fixed it? :)
21:26:29sandkingwhere i can find startup log?
21:26:43S_a_i_n_tZagor_: Good enough for me ;)
21:26:50sandkingi installed rockbox and during bootup there's some error and it boots into original firmware
21:27:13S_a_i_n_tsandking: IFAIK, RB doesn't log errors on boot.
21:27:55sandkingis it impossible?
21:28:18S_a_i_n_tI guess not, just not a "default thing"
21:28:37Jerom"Data abord at 30086356 FSR ( domain 0, fault 8 ) address 0x5F6464C1" with latest svn revision on Fuzev2
21:28:38sandkingI can't figure out what's happening
21:28:52Zagor_sandking: much of the boot is done before the disk is up so we couldn't log even if we wanted
21:28:52gevaertssandking: the bootloader should have shown the error on the screen
21:28:58sandkingi got firmware on the device, loaded patched original firmware and it doesn't work
21:29:20sandkinggevaerts: yep, it does. but it's too quick to figure it out
21:29:44funmanJerom: doing what?
21:30:15Jeromloading rockbox.sansa from microSD1
21:30:16S_a_i_n_tDoes a bootchart build need to actually boot successfully to write the boot log?
21:30:27 Quit simonrvn (Ping timeout: 260 seconds)
21:30:28sandkingresult: -1?
21:30:30S_a_i_n_tif not, you could *maybe* see how far it's getting.
21:30:34sandkinglast line display -1
21:31:12sandkingLoading Firmware and then Result: -1
21:31:44funmanJerom: do you have the rockbox.elf in your build dir? you can see which function is at 30086356
21:32:59JeromHow can I do that ?
21:34:08mc2739Jerom: did you download or build your own Rockbox?
21:34:15Jerommc2739: build
21:34:37mc2739look in your build diectory for rockbox.elf
21:34:38 Join simonrvn [0] (
21:35:07JeromI have done a gdb rockbox.elf
21:35:20sandkingis there some info about startup messages?
21:35:43funmanJerom: can you upload it somewhere ? ( for example)
21:38:28domonokysandking: is the rockbox main firmware actually installed ?
21:38:29funmanthe address is in a string
21:38:49*domonoky thinks -1 was rockbox binary not found..
21:39:00sandkingdomonoky: well, i went through manual installation as well as utility installation with no luck
21:39:47domonokysandking: rockbox consists of two parts.. bootloader and the main binary (plus extra files).. from the message we know the bootloader is working :-)
21:40:19domonokynow check if there is a .rockbox dir on your player with a rockbox.<ext> file in it (and more)
21:40:38Jeromfunman: So I'm out of luck ?
21:41:41sandkingdomonoky: well, i got .rockbox on it, but don't know what's rockbox.<ext>
21:42:18 Join toffe82 [0] (~chatzilla@
21:42:40sandkingbut anyway - rockbox utility should take care of the process, right?
21:42:40funmanJerom: loading rockbox.sansa from microsd works fine for me
21:43:03JeromI've tried coying it to the main memory
21:43:11 Quit phanboy4 (Read error: Connection reset by peer)
21:43:12 Part watto
21:43:19JeromSame thing happens
21:43:45funmanhow old is the build you are rolo'ing from ?
21:43:55 Join phanboy4 [0] (
21:44:02domonokysandking: a file named rockbox with some extension (that differs between targets)
21:44:42sandkingdomonoky: well, yeah, i got rockbox.iriver
21:45:47sandkingand i got latest build for h120 from here so i don't know what i could do wrong here
21:47:01JeromShould I try loading the latest build without Rolo ?
21:47:31*domonoky just looked into the code. and -1 is really file not found
21:47:50funmanJerom: yeah
21:48:20bluebroth3rmaybe run chkdsk /f?
21:48:27bluebroth3rcould be filesystem issues
21:49:17JeromNow I can't boot rockbox :p
21:49:27funmanthe firmware is checksumed but it's a simple sum so it could be easily confused by filesystem issues
21:50:37Jeromsame error as before
21:51:45saratogado you have some odd file system on your disk?
21:51:58saratogaJerom: not you
21:52:02Jeromsorry ^^
21:56:57 Join mbr [0] (
21:59:54mbrfunman: Any ideas why some new fuze v2 are not installable?
22:00:21mbrI bought two for my children, both not installable
22:00:34JeromStrange, I copied it over the OF instead to the microSD and it worked
22:00:35funmanno idea
22:00:53funmanmbr: if you ship one of them to me i could look
22:01:10***Saving seen data "./dancer.seen"
22:01:14funmani have a fuzev2 here but it's not mine, if the owner agrees perhaps we could exchange
22:01:18chrissaverywasn't there a report that changing some bytes in the OF made it installable?
22:02:10mbrchrissavery: Hmm didn't read about it. do you have a pointer?
22:02:42CIA-6New commit by Domonoky (r26671): fix list of themes for specific targets in admin view.
22:02:55funmanchrissavery: i think the report just said that changing a few bytes still allowed patching from/to the 2 installable OF versions
22:03:15funmanwhen patching with rockbox we just change a few more bytes (but not so much)
22:03:34mbrfunman: need to check this. problem is I bought them via amazon and exchange might be a problem in case of warranty
22:04:13funmanthese fuzes probably need to be opened anyway
22:04:34mbrin the forum I reported that exchanging those bytes works, meaning flashing works
22:05:09chrissaverymy mistake. it was mbr that posted about that but I forgot the details of it. had to chk.
22:05:33funmanmbr: you should try swapping 2 contigous bytes so the checksum changes, and modify the checksum accordingly
22:06:02 Quit sandking (Quit: Leaving)
22:06:15mbrfunman: you mean it might be necessary to do the unbricking procedure
22:06:46mbrI'll try to do do the checksum stuff
22:07:00funmanmight yes, and also there could be data present in the hidden part of the flash which isn't in the OF file
22:07:37 Quit bieber (Ping timeout: 264 seconds)
22:07:51 Join bieber [0] (
22:08:17chrissaverymaybe fuze devs looked at how the rockbox patched and purposefully added code to prevent patched installs? how would you get around that?
22:08:47gevaertschrissavery: change the patching method of course
22:09:02S_a_i_n_tbut, if it was added to the FW, and not the device...vouldn;t it just be removed?
22:09:10chrissaverycould a trial be run that tries that?
22:09:10S_a_i_n_targh, "couldn't"
22:09:27funmani don't think sandisk mind us, preventing european people from working around the high volume regulation is more plausible to me
22:09:34AlexPchrissavery: What do you mean a trial?
22:09:45AlexPIt needs a dev to look at it
22:10:02S_a_i_n_tfunman: What's with that volume regulation? We don;t have that here...
22:10:05chrissaverymaybe diff OF versions and see if some suspicious init code is changed.
22:10:10AlexPS_a_i_n_t: Try google
22:10:24S_a_i_n_tso DAP companies don't deafen people? and wind up liable in court?
22:10:31Torneit would be in the flashing code, rather than the init code, probably..
22:11:21chrissaveryI meant for someone who knows how the patching works to create a different patch method that could be tested on a "retarded" fuze
22:11:41funmangoing to new versions and then back to old versions works for me
22:12:30funmanchrissavery: we don't know how the patching works, we use the OF code to do that; we only know (a part of) the file format that this patching code uses
22:14:30chrissaveryhas anyone tried setting the region code to "rest of world" and then doing an upgrade?
22:14:48 Join moparx [0] (~moparx@unaffiliated/moparx)
22:17:33bertrikah, so the idea is that the OF refuses an update to a different region?
22:19:22 Quit Strife89 (Quit: Leaving)
22:19:26 Quit bmbl (Quit: Bye!)
22:20:01chrissaverywell, if they are trying to prevent bypassing the volume limit then they may have only coded a restriction on the EU version. Though why it may limit NA I don't know. But I haven't heard of anyone proving it isn't region specific. They would have to force a reset and choose a different region. I haven't heard of that being tested though.
22:21:36chrissaveryAlso, does the Fuze have ROM code that inits the OF? I know there is some mention of it in the AS3525 docs but not sure if it is present. They must have coded something into the hardware since I can install the same OF and it allows me to install RB.
22:22:50chrissaveryIf we could identify different bytes in such a ROM we may be able to determine where it checks those bytes in the OF and bypass that check.
22:23:44funmanchrissavery: it could be in the hardware yes, or also in the reserved part of the flash which isn't filled by the OF file
22:23:54bertrikI do think it has a ROM in the processor for booting, which runs before the OF
22:26:17bertrikchrissavery, have you tried to flash it with a rockbox bootloader file with a name matching your current region?
22:26:39bieberbluebroth3r: I'm going to do that now
22:26:49saratogamy theory is that AMS did a new stepping of the as3525v2 chip thats more stringent in checking the firmware headers
22:26:59mbrchrissavery: Just installed a vanilla 2.3.33
22:27:49saratogaand we simply didn't notice some header byte before because it didn't matter in the previous stepping
22:27:49chrissaverymy Fuzev2 has no issues with RB installing, it's a version from last August. I'm just wondering about the future users who won't be able to use RB.
22:27:49funmansaratoga: i think there could also be a new checksum since they added one between AMSv1 and AMSv2
22:28:52saratogafunman: i thought swapping bytes was reported to have worked in the forums
22:28:52saratogathat would imply theres no new checksum
22:28:52mbrthen generated a fresh patched 2.3.33 and tried to install that in "Rest of World" mode
22:28:52mbrno success
22:28:52chrissaverybertrik: I don't know about region-specific firmwares. Is there something about that? I've always just named it fuzpa.bin.
22:28:52funmansaratoga: swapping doesn't alter the checksum
22:28:53 Quit perrikwp (Quit: CGI:IRC (Ping timeout))
22:28:59funmanchrissavery: does fuzpe.bin work?
22:28:59saratogaso they added a new checksum that works the same as the old?
22:29:03saratogaseems kind of silly
22:29:32funmanthey added a checksum in the header part, additionally to the firmware block part, and there is also the whole file checksum
22:29:43*mbr tried to flash fuzpa.bin
22:29:47CIA-6New commit by bieber (r26672): Theme Editor: Changed anged error reporting method in parser, added error messages to status bar in editor
22:30:21chrissaverymbr: ok. that shoots that idea down for good.
22:30:56bertrikI vaguely remember seeing some file on the OF filesystem of the amsv1 sansas that contained the region specific settings
22:31:26bertrikmaybe it helps to delete that file, just an idea
22:31:42funman is related
22:31:59bertrikwell, if it exists on the amsv2 sansas too
22:35:35 Quit Jaykay (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
22:39:05*funman will claim these unpatchable fuzev2 don't exist if he can't get his hands on one of them
22:39:11chrissaveryfunman: for me fuzpa.bin has always worked. I don't have a new Fuze though.
22:39:41amiconnWhat was the binsize impact of the skin format thingy?
22:40:05bertrikWe just keep the ams sansa thread locked, and not move it, right? In some places the wiki refers to the thread and I'm not sure what happens when we move the thread.
22:40:31AlexPbertrik: Probably best for now
22:40:53AlexPbertrik: If it moved it'd only be to removed posts, and then only badged (non-artist) people can see it
22:41:15funmanbertrik: the thread can be useful for finding old information
22:41:34funmani guess MrSomeone could delete all the support requests & answers from it
22:41:35mbrOK, I still have a few days to decide if I return me fuzes
22:41:40kugelbertrik: I think the links would still work
22:41:55bertrikOk, I do think it's a good idea to copy the information from the thread to the wiki instead of only referring to it
22:42:02AlexPkugel: I think so too, but users wouldn't be able to get to it
22:42:12kugelyep, keep it there.
22:42:57mbrneed to go
22:43:05S_a_i_n_tamiconn: My themes ended up less than 1kb smaller.
22:43:24AlexPS_a_i_n_t: He said binsize
22:43:28funmanbertrik: true
22:43:35kugelnobody cares about the theme size really :)
22:43:53*funman goes designing a 1920x1080 theme
22:44:06 Part mbr
22:44:23S_a_i_n_tmy rockbox.ipod is a couple of Kb larger.
22:44:41 Quit notlistening (Ping timeout: 252 seconds)
22:45:05gevaertsamiconn: IIRC around 1K
22:45:08S_a_i_n_tAlmost 2kb, compared to a build before skinbreak.
22:45:23gevaertsright, 1K average RAM/BIN
22:45:30*gevaerts isn't sure now
22:46:09kugelthe binsize doesn't really matter though, since the legacy parser is still in
22:46:18kugelWe should look at it once the new parser gets used
22:46:44gevaertskugel: why is that?
22:47:19 Nick fxb__ is now known as fxb (
22:48:01kugelgevaerts: why is what?
22:48:16gevaertskugel: why is it still in?
22:48:47kugelthe new parser isn't finished IIUC
22:48:53kugelit's the one the theme editor also uses
22:49:21*gevaerts isn't sure he understands
22:49:30*kugel thinks there should be a mail clearing it
22:49:41gevaertsSo the new format is parsed using a patched old parser?
22:50:32kugelbut that's all more or less a guess going by their talk and the commits
22:50:55kugelthere hasn't been an "official" announcement of what's happening with the rockbox side I think
22:52:25 Join panni_ [0] (
22:52:28 Join S_a_i_n_t_ [0] (S_a_i_n_t@
22:52:33 Quit S_a_i_n_t (Ping timeout: 276 seconds)
22:54:24 Quit dfkt (Read error: Connection reset by peer)
22:55:47kugelgevaerts: according to the parser is completed actually
22:56:50gevaertskugel: I'd read that as "handed over to JdGordon for integration"
22:57:33 Quit Zigtown (Remote host closed the connection)
22:59:06 Quit S_a_i_n_t_ ()
22:59:16 Join Buschel [0] (
23:00:49kisakfunman: about how many files do you use to test with on your Fuze v2?
23:01:37 Quit moparx (Ping timeout: 264 seconds)
23:02:47funmanto test what?
23:03:13kisakgeneral use and playback
23:03:32funmana few
23:03:35kisakmy Fuze v2 is having a hard time keeping up
23:04:39kisakwhen updating the database in the background, it causes longer delays between files
23:04:54funmansounds normal
23:04:57 Quit chrissavery (Ping timeout: 240 seconds)
23:05:18 Join S_a_i_n_t [0] (S_a_i_n_t@
23:05:46kisakwhen I made the idle shutdown longer I started seeing the player become non-responsive
23:05:59kisakbacklight off, no apparent response
23:06:07bertrikwe have a mechanism now to give some threads more priority to access storage, the database update thread has low priority
23:06:34kisakthis was in attempt to give the database time it needed to finish
23:07:49kisakalso, there's been 8-10 second delay in starting the first file
23:08:32kisakI can't find any setting to turn off to try to test with less stuff turned on
23:09:14CIA-6New commit by bieber (r26673): Theme Editor: Added a clear erros function to the parser
23:10:16kisakseems to me that something isn't scaling up well (to 2700+ files)
23:10:46 Join moparx [0] (~moparx@unaffiliated/moparx)
23:10:57 Quit GeekShadow (Read error: Connection reset by peer)
23:10:58Buschelsaratoga: can you please add your test results (mpa, eabi, dct32) to flyspray? I would like to keep all information together.
23:11:38gevaertskisak: do you have dircache enabled?
23:12:08kisakgevaerts: I assume it is because I see messages saying that it is full
23:13:12kisakgevaerts: there are messages when I enter directories saying that the dircache is full
23:13:32kisakthen they go away and it doesn't seem to stop me
23:13:43gevaertsHave you checked the filesystem recently?
23:13:59kisakI've had no reason to check it
23:14:29funmana good reason is that the unstable SD driver can corrupt the filesystem at anytime
23:14:43kisakwill a fsck be sufficient?
23:14:44CIA-6New commit by pixelma (r26674): Rework the button table in the graphical eq section. Clean up the code, correct falso info and add the missing mappings of the: Fuze, H10s, C200, M3 ...
23:15:19pixelma+ Packard Bell Vibe 500...
23:15:54funmankisak: should work i think
23:17:00CIA-6New commit by amiconn (r26675): Fix byte order for Mini G2, and fix comments.
23:17:19saratogaBuschel: done
23:18:27 Join audio-i [0] (
23:18:28 Join perrikwp [0] (
23:18:49saratogaBuschel: i wonder if we could optimize synth_full a little more by eliminating some of those 64 bit accumulations
23:19:50kisak I've never encountered this message before
23:20:04 Quit audio-i (Client Quit)
23:20:12 Join audio-i [0] (
23:20:21Buschelsaratoga: you mean using 32 bit mla?
23:20:27saratogaBuschel: yeah
23:20:34saratoga64 bit seems a bit excessive ?
23:20:37kugelBuschel: btw, the flac slowdown can be fixed by using -O2
23:20:50 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:21:00kugelthat makes both arm and arm-eabi faster, but arm-eabi benefits more and is even faster than arm with it
23:21:04 Quit hebz0rl (Quit: Ex-Chat)
23:21:20kugelnoticeably faster I might add
23:21:32Buschelkugel: good news :o) so, no reason to not submit this now ;)
23:21:54kugelthis = -O2 for flac?
23:21:55 Quit audio-i (Client Quit)
23:22:13 Join audio-i [0] (
23:22:16 Join elinenbe [0] (
23:23:01Buschelsaratoga: I messed around a lot with 32 bit mla/mul in the past with mpc synth and dct32. You'll need excessive pre/post scaling to avoid overflow and to keep reasonable precision. the difference in speed is not that much.
23:23:16gevaertskisak: that's a pretty common one. Pick randomly :)
23:23:30audio-idomonoky: is the theme site process finished? there are still lots of themes missing, and some of the shown themes do not reflect that they work with release 3.6
23:23:32kisakoh great...
23:23:50kisakI pick the one that claims the drive is sane
23:24:11Buschelsaratoga: for mpc I switched back to full precision a while ago. few gain in speed but hard to maintain.
23:24:13domonokyaudio-i: i didnt start the conversion yet, there seems to be a problem with the skinupdater which needs fixing first...
23:24:52domonokyi am also just rerunning checkwps, so all themes should be tested against 3.6 in a bit..
23:25:02audio-iah ok thanks
23:25:31 Part Jerom
23:25:34domonokyif there are themes missing, please tell me.. i just removed all old hidden themes, and maybe i have done something wrong.. :-)
23:25:38saratogai guess for a real improvement you'd need to use 32x32=32 multiply accumulate and that would involve a lot of careful thought
23:26:11Buschelsaratoga: if we might experiment with precision loss we could also just downscale coefficients by >>8 or even not use some of the dewindowing coefficients (no load, no mul, no add)
23:26:12audio-idomonoky: at this moment, yes there are a lot of themes missing :)
23:26:58domonokyplease look again in ~1/2 hour.. and i need exact infos about missing themes then.
23:27:13audio-iwill do
23:27:25saratogawhich are the dewindowing coefficients?
23:27:32Buschelkugel: sorry, just read your question. yes: "this = -O2 for arm"
23:28:15Buschelsaratoga: D[][] in synth.c
23:28:25saratogaaren't those the filter coefficients?
23:29:05Buschelsaratoga: yes. that's the same.
23:29:18Buschel(maybe resorted though)
23:29:22saratogahow could you not use them? wouldn't everything still be modulated?
23:29:54Buschelsaratoga: just take a look at mpc's filter coef declaration. it's more obvious there
23:30:24BuschelDi_opt[] in synth_filter.c
23:31:26saratogaspeaking of mpc, maybe putting the dct into the codeclib makes sense?
23:31:34Buschelsaratoga: you will see that the first and the last value in each line can be dropped without impacting the filtering too much.
23:32:17saratogado other decoders do this?
23:32:32Buschelsaratoga: no. mpc uses a data structure that is different to libmad's. the dct implementations are similar but not identical
23:32:59saratogai know various embedded mp3 decoders use tricks though
23:33:28Buschelsaratoga: I am sure performance tweaked codecs do so. but this implies loss of precision
23:33:49saratogado you think its a lot of precision? we have quite a bit to spare right now
23:34:39 Join Guest5624 [0] (
23:34:49Buschelsaratoga: we could measure this. I never did
23:35:12saratogado you think RMS error is a good enough measure, or is the error likely to be concentrated in frequency or time?
23:35:34 Join MethoS- [0] (~clemens@
23:35:34funmansaratoga: i should look at the replaygain track gain for each file, right?
23:35:40saratogafunman: yeah
23:35:48saratogait should scale linearly with volume
23:36:28 Quit Guest5624 (Client Quit)
23:36:31saratogaRMS works just as well, but its a little harder to calculate unless you like matlab
23:36:56 Join Guest5624 [0] (
23:37:43Buschelsaratoga: I would expect that the side band attenuation will be reduced. As a consequence more aliasing should be introduced. I would prefer to create .wav output and analyze it with a tool like CoolEdit or something similar.
23:38:16 Quit Guest5624 (Client Quit)
23:38:30Buschelsaratoga: The spectrum of the additional aliasing/noise is of interest.
23:38:41funmanattenuation is much lower for clipv2
23:38:58funmanit's the same for the OF on all clips but varies on rockbox
23:39:37funman16.16dB on clip+, 15.71dB on clipv1 => can this .45dB difference be easily heard
23:39:57kugelBuschel: there's no patch to submit though ;)
23:42:18 Join Guest5624 [0] (
23:42:21bertrikfunman, 0.5 dB is not a lot
23:42:46saratoga1dB is where you can start to notice without abx
23:43:07Buschelsaratoga: what I did a while ago was to compare 32bit to 64bit multiplies. The additional noise was highly correlated to the signal itself. It was interesting to compare this spectrum to the masking threshold. Masking threshold can be visualized when subtracting the original wav from the encoded/decoded wav. The additional noise (introduced through 32bit calculation or dropping coefs) can be visualized when subtracting encoded/decoded 64bit from en
23:43:59Buschelsaratoga: the result was that the introduced noise was covered by the masking threshold. so, no audible impact.
23:44:00bertrikthe as3514 codec has a 1.5 dB volume difference per step
23:44:42Buschelkugel: ?
23:45:26kugelBuschel: nobody has made a patch for the -O2 thing
23:46:28Buschelkugel: I know :o) I thought you would just submit this one.
23:48:48 Quit kugel (Remote host closed the connection)
23:52:12Buschelsee you tomorrow, gotta get some sleep now
23:54:13domonokyaudio-i: checkwps run finsished... if there are still themes missing, please give me some sort of list and i will look into it.
23:55:03 Quit t0rc (Remote host closed the connection)
23:55:35funmanwhen using DAC output and not SUM output on the headphones, the track has much higher volume
23:55:41domonokyaudio-i: name + lcd size to indentify the theme would be good.
23:55:48S_a_i_n_tIIRC there is 2 themes missing from the Nanos, I'm used to seeing 29 themes, but there's only 27 can I view the "old" themesite to confirm this?
23:56:12domonokyS_a_i_n_t: there is a link to the copy on the site
23:56:17funmanat rockbox volume 0, RG attenuation is 16.64dB with DAC, vs 13.09dB with SUM
23:56:55S_a_i_n_tdomonoky: Thanks, just saw it then...cleverly hidden in plain sight ;)
23:57:05 Quit Buschel (Ping timeout: 264 seconds)
23:58:28 Quit bertrik (Quit: De groeten)

Previous day | Next day