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

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

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

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

#rockbox log for 2014-02-15

00:01:12 Quit bulldog ()
00:05:22***Saving seen data "./dancer.seen"
00:12:28 Quit wodz (Quit: Leaving)
00:30:53 Nick SuperBrainAK is now known as DormantBrain (~andy@2001:470:8:a61::5f92:59a1)
00:33:56 Quit pamaury (Ping timeout: 260 seconds)
00:35:38 Quit ender` (Quit: We're not lost. We're locationally challenged.-- John M. Ford)
00:38:28myndziis there a one-color rockbox logo?
01:13:43gevaertsmyndzi: something like;a=blob;f=apps/bitmaps/native/rockboxlogo.160x53x1.bmp;h=7e34b23c6b21b22826b8f001c01c11de6ca3b3b8;hb=HEAD maybe?
01:17:03 Quit bertrik (Remote host closed the connection)
01:17:59gevaertsAlthought given that the regular logo is really only three colours, taking just the black from;a=blob;f=docs/logo/rockbox-logo.svg;h=a33189cefb84a5e9b34c33ab9fd7def9e87934e6;hb=HEAD *might* be reasonable
01:19:31 Quit [Saint] (Remote host closed the connection)
01:20:42 Join [Saint] [0] (~saint@rockbox/staff/saint)
01:21:28 Join nonMOOSE [0] (nonmoose@2001:470:f091:ec1b:3e11:cc7d:63a7:97b5)
01:22:21nonMOOSEso hey, I just loaded rockbox on my gf's 2gb ipod nano- and am totally confused (as it doesn't seem to work the way my old 30gb ipod classic did w/ rockbox installed)...
01:23:26nonMOOSEI wanna take the songs she's already got on the device, and de-itunes them (convert to regular filenames/folders, etc.)- and then set it up so I can just copy mp3's etc... onto the device like it's a jump drive
01:23:36nonMOOSEwell the latter is the most important piece...
01:24:06nonMOOSEwe can always re-rip and/or re-get all the songs she's got in her itunes library right now
01:24:34nonMOOSE(my goal is also to completely avoid using itunes- or any apple software if I can(
01:25:43nonMOOSEso yeah anyways... It's prolly been 4 or 5 years since I had that 30gb ipod w/ rockbox installed....
01:26:32nonMOOSEI somehow managed to lose the free ipod someone gave me :(
01:27:10nonMOOSEwell at least I've still never given apple a single cent of my money ;0
01:39:21 Quit tertu3 (Ping timeout: 246 seconds)
01:47:31 Quit ZincAlloy (Quit: Leaving.)
02:05:24***Saving seen data "./dancer.seen"
02:11:54 Quit lebellium (Quit: ChatZilla [Firefox 28.0/20140210161136])
02:19:38soapnonMOOSE, what is your question?
02:19:58soapHow to remove the iTunes-loaded songs and readd them properly named?
02:20:19soapOr "why does Rockbox not look like I recall it looking 4 or 5 years ago"?
02:27:00 Quit Cultist (Read error: Connection reset by peer)
02:28:11 Join Cultist [0] (
02:33:30myndzigevaerts: turns out i need it too small, i wound up tracing by hand in photoshop and adjusting for fire, plus removing the line details
02:33:56myndzialmost done except the treble clef intersects itself and i don't want it to cut out :(
02:45:41 Quit Guest80912 (Read error: Connection reset by peer)
02:47:50 Join Guest80912 [0] (
02:51:25 Join markun [0] (~markun@rockbox/developer/markun)
03:21:57 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
03:28:40myndzi ?
03:29:05myndzifirst definitely i think
03:50:18 Nick DormantBrain is now known as SuperBrainAK (~andy@2001:470:8:a61::5f92:59a1)
03:56:30 Join tapiralec [0] (
04:05:27***Saving seen data "./dancer.seen"
04:28:00 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:28:00 Quit amiconn (Disconnected by services)
04:28:03 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:28:25 Quit pixelma (Ping timeout: 246 seconds)
04:28:59 Join pixelma [0] (quassel@rockbox/staff/pixelma)
04:51:01 Join prof_wolfff [0] (
05:09:45 Quit Zarggg (Ping timeout: 260 seconds)
05:13:08 Quit TheSeven (Disconnected by services)
05:13:21 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
05:35:58 Quit [Saint] (Remote host closed the connection)
05:39:00 Join [Saint] [0] (~saint@rockbox/staff/saint)
06:01:29 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
06:05:31***Saving seen data "./dancer.seen"
06:06:14 Join tertu3 [0] (~tertu@
06:14:35 Quit kuldeepdhaka (Ping timeout: 252 seconds)
06:19:22 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
06:29:33 Quit tertu3 (Ping timeout: 260 seconds)
06:29:54 Quit kuldeepdhaka (Max SendQ exceeded)
06:31:16 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
06:38:34 Quit kuldeepdhaka (Max SendQ exceeded)
06:40:14 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
06:42:41 Quit kuldeepdhaka (Client Quit)
06:46:38 Quit tapiralec (Quit: Ex-Chat)
07:53:46 Join Gallomimia [0] (~gallomimi@
08:05:33***Saving seen data "./dancer.seen"
09:09:14 Join bertrik [0] (
09:09:14 Quit bertrik (Changing host)
09:09:14 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
09:10:02 Join ender` [0] (
09:21:56 Nick SuperBrainAK is now known as DormantBrain (~andy@2001:470:8:a61::5f92:59a1)
09:38:07 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
09:47:31 Join shai [0] (
10:05:37***Saving seen data "./dancer.seen"
10:05:48 Quit [Saint] (Remote host closed the connection)
10:10:52 Quit Guest24881 (Quit: Bye bye)
10:17:15 Quit Gallomimia (Quit: I am likely going to change locations)
10:25:49 Quit shai (Quit: Leaving)
10:40:06 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:52:28 Join [Saint] [0] (~saint@rockbox/staff/saint)
11:00:36 Join Scall [0] (~chat@unaffiliated/scall)
11:06:37 Quit Slasheri (Ping timeout: 265 seconds)
11:06:44 Join Slasheri [0] (
11:06:44 Quit Slasheri (Changing host)
11:06:44 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
11:22:06 Quit [Saint] (Remote host closed the connection)
11:23:56 Join [Saint] [0] (~saint@rockbox/staff/saint)
11:26:02 Quit dfkt (Read error: Connection reset by peer)
11:26:14 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:29:01 Join lebellium [0] (
11:39:27 Join wodz [0] (
12:05:40***Saving seen data "./dancer.seen"
12:29:03 Quit [Saint] (Remote host closed the connection)
12:30:20 Join [Saint] [0] (~saint@rockbox/staff/saint)
12:41:46 Quit akaWolf (Quit: my exit)
14:01:39 Quit Unhelpful (Remote host closed the connection)
14:05:19 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
14:05:41***Saving seen data "./dancer.seen"
14:12:24 Quit Unhelpful (Remote host closed the connection)
14:16:14 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
14:35:51 Join lorenzo92 [0] (
14:46:02lebelliumwouldn't it be possible to send an email to the authors who made a now broken theme? I feel sad seeing so many themes broken due to the new theme engine restrictions :(
15:02:38lebelliumthe theme engine is now more restrictive about width and height
15:02:46lebelliumthat's why so many themes are broken
15:03:45lebelliumcheck here for more info:,29851.0.html
15:18:49 Join ZincAlloy [0] (
15:25:35nonMOOSEsoap: well, back when I was using it before- the interface/theme seemed pretty similar, however the MUSIC, VIDEO, ETC. directories were clearly the 'main' focus- until a track started playing- at which point it did the old 'current playlist' deal, and then brought you right back to the music folder(s) and told you to queue up more songs....
15:27:50nonMOOSEdunno, this revision (at least on the ipod nano) basically hid the itunes tracks completely, and didn't create a MUSIC or whatever folder automatically- which 'kind of' leaves the n00b/dinosaur user going "Wait a minute, I thought this was a fucking mp3 player..? NOT ANYMORE!" heheh
15:28:50nonMOOSEheh- only a 2gb player is kinda pushin' the limits on minimum amount of storage
15:29:35nonMOOSEseeing as the first mp3 player ever (diamond mm 64mb) was well, 64mb's- and came out in ~1997/8 if I recall
15:30:11copperI have no idea what you mean
15:31:11copperbut then I have no idea what Rockbox looked like 5 years ago
15:31:33copperalso, what theme are you talking about? Cabbiev2?
15:33:41copperI think you want to use the database
15:34:08coppersince iTunes renames music files when transfering them, to something that's unusable
15:34:28copperthe database will read the metadata and present your tracks neatly organized by artist, genre, etc…
15:35:29copperalternatively, you can delete all music files present on the iPod and copy the original files in your iTunes music directory on your PC, which should have a usably file and directory structure
15:35:50ZincAlloyor do both :)
15:52:28 Join ulmutul [0] (
16:05:43***Saving seen data "./dancer.seen"
16:22:24*pamaury is experimenting with the speakers on imx233
16:24:21bertrikpamaury: I'm thinking the clip sport is either an AMSv2 (like the clip zip) or like a fuze+. Is there any way to tell? Like the presence of a bootloader mode that can be triggered, or a specific USB configuration?
16:25:47pamaurybertrik: recovery mode can tell you yeah
16:26:25pamauryimx recovery mode has usb 066f:3780 and is typically reached by holding a button when plugging usb
16:27:35pamauryor by plugging usb, holding the power button and pressing reset
16:28:13bertrikhm, now all we need is someone with a clip sport to try stuff like that! :)
16:28:34pamauryyeah, I asked on ABI but got no answer
16:28:53bertrikI wonder if the 128x128 resolution is easy to port, there was quite a bit of work for 96x96
16:30:49pamaurydo we have 128x128 targets ?
16:31:03ZincAlloyyes, iriver h10
16:33:00bertrikgogearhdd1630, lyreproto1, mpiohd200 are 128x128 and creativezenv is close with 132x132
16:36:27 Join petur [0] (~petur@rockbox/developer/petur)
16:36:39 Quit petur (Client Quit)
16:51:36 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
17:04:39soapnonMOOSE, Even 4-5 years ago you had to create the tagcache database before you would have been able to see your iTunes loaded music.
17:05:29soapIf I'm hearing you correctly you're complaining that on the 2GB Nano you don't automagicly have your music / videos waiting for you upon boot but you did have that on your older player some years back.
17:06:31soapIf I'm hearing you correctly I'm telling you that you misremember, as there was never a period in time where iTunes-added music was so easily available and never a time in which itunes videos are playable.
17:17:52 Quit lorenzo92 (Quit: ChatZilla [Firefox 27.0/20140130123732])
17:19:01soapbut copper has outlined where you need to go to deal with the situation I believe
17:31:02wodzpamaury: Any news with yp-cp3? I have some free time this evening so I could help
17:34:28wodzanyone have an idea why adding struct declaration to firmware/export/config.h breaks codecs.make ?
17:35:52pamaurywodz: maybe because it parses it and expect it to contain only #define stuff
17:37:38wodzyeah, it must be something like this
17:39:13 Quit lebellium (Quit: ChatZilla [Firefox 28.0/20140210161136])
17:40:27pamauryI don't see anything suspicous but that must the explaination somehow
17:43:48 Join lebellium [0] (
17:55:13wodzpamaury: I started experimenting with elfloader again. Now I would like to resolve external symbols (just like kernel modules works in linux). The problem I see is the considerable space increase as the symbols are referenced by name so you have to hold string + value for every exported symbol
17:56:52wodzI was thinking it should be wise to use hash instead of string. Do you have an idea how this could be done wrt to build system? How to build such array of structs {hash, value} and link it into the binary not affecting addresses obviously?
17:58:16pamaurywhat is the advantage of using symbols instead of a vtable ?
17:58:43pamauryappart from the obvious simplification from a programming/linking point of view ?
17:58:59wodzwhat do you mean by vtable?
17:59:41 Join lebellium_ [0] (
17:59:41pamauryjust like for codecs
18:00:35wodzyou cannot easily export from such modules
18:00:45 Quit lebellium (Ping timeout: 246 seconds)
18:00:50 Nick lebellium_ is now known as lebellium (
18:01:41wodzWhat I am still thinking about is to split some parts of *core* into loadable modules.
18:01:41pamauryoh you want to expert symbols too, indeed
18:02:14wodzlike for example tagcache, radio etc
18:02:27pamauryhow much space do symbol take and what is win of using a hash ? I mean real life figures
18:04:18wodzsymbol takes strlen(symbol name) + sizeof(void *) + sizeof(char *)
18:04:37wodzwith hashes it could take sizeof(int) + sizeof(void *)
18:05:14pamaury"readelf -s rockbox.elf | grep FUNC | wc -l" gives 2900 symbols but I would need a better estimation
18:05:30wodzand searching would be much faster. With names you need to perform binary search with strcmp() each time, with hashes you simpy compare
18:05:45***Saving seen data "./dancer.seen"
18:06:41wodzNot only FUNC, but I am thinking about explicit symbols exporting (like in linux) which should considerably reduce the number
18:07:18 Quit bertrik (Remote host closed the connection)
18:08:39ulmutulHello everybody! I'm working on a better scrollstrip implementation (my target: Vibe500)
18:09:09ulmutulIt should of course have adjustable speed settings
18:10:38ulmutulMy question is: in settings_list.c I can use CHOICE_SETTING(...), but I don't like the idea of strings named "fast, super fast, insanely fast, damn slow"
18:10:39pamauryreadelf -s rockbox.elf | awk '{ if($4=="FUNC" || $4=="OBJECT") print($8) }' | awk 'BEGIN { size=0; } { size+=length($1); } END { print(size); }'
18:10:39pamaurygives ~82000 so yeah quite a gain spacewise
18:11:38pamauryhowever it won't be that easy to have hashes instead of names
18:11:45ulmutulso another possibility would be to use TABLE_SETTING(F_ALLOW_ARBITRARY_VALS,...), but it's poorly documented.
18:12:39ulmutulI forgot to mention that the scrollstrip should also have an "off" setting, I think of something like "off, -2,-1,0,1,2"
18:12:41wodzulmutul: you should ask JdGordon probably
18:13:01wodzulmutul: why would you like to disable scrollstrip?
18:14:10ulmutulIt's better for some games, they just suck with scrollstrip. With strip setting "off" the upper/lower half of the strip is handled like ordinary up/down buttons when touched.
18:15:37pamaurywodz: the problem I see with symbols is for the main binary because these have to be included in the final non-elf binary
18:16:21ulmutulHm, I think I need some kind of "formatter" that's used for TABLE_SETTING.
18:16:25 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
18:16:30wodzpamaury: with strings thats is no issue. macro trickery + ld trickery do the job
18:16:59pamauryI mean for hashes
18:17:04wodzpamaury: with hashes, yes thats problematic and hence my ask for advice
18:17:22 Join Keanu73 [0] (~Keanu73@unaffiliated/keanu73)
18:17:22pamauryfor modules you can postprocess the file but for non-hash...
18:17:28pamauryfor non-module
18:17:35Keanu73EXCUSE ME
18:18:03wodzKeanu73: no you can't. This IS our logbot
18:18:08pamauryKeanu73: no caps please, this our logbot so no
18:18:20Keanu73get another nick instead PLEASE
18:18:24Keanu73like "ChanLogBot
18:18:34pamauryKeanu73: do you want to get kicked ??
18:18:46pamauryso please so it, like now
18:18:49pamaury*stop it
18:18:59Keanu73is logbot used with pircbot?
18:19:05Keanu73this one?
18:19:41Keanu73or with IRC syntax
18:19:54ulmutulBTW, anyone here with another scrollstrip target? I still have to do some code cleanup and then get into this git/gerrit stuff, but with some parameter tweaking the code should run with pretty much any strip-target.
18:20:28wodzKeanu73: Whats your problem actually? The nick is taken, get another one, simple
18:20:53wodzulmutul: I have hd300 somewhere in my drawer
18:20:55Keanu73wodz: is the bot made with pircbot at
18:21:16Keanu73because there's a similar logbot to this one:
18:22:03pamaurywodz: there is an "obvious" solution though which is to include strings and then post process but this way you waste space
18:22:05 Part Keanu73 ("I need to leave, some people are being mean to me or somewhat. Bye!")
18:23:42wodzI think I first make it work with strings and then I'll start to scratch my head how to do this with hashes
18:26:07pamauryyou could do it this way: post-process, add to the final binary (at the end) and then at the first beginning of loading, move it around so it ends up in so dynamically allocated buffer
18:29:39wodzits funny someone popped up with vibe500 while I am testing stuff on exactly this player :P
18:33:01pamaurywodz: I have another idea
18:33:21pamaurywodz: with macros you can add some empty space in the binary for each symbol and the string in another one
18:33:40pamaurythen you postprocess the binary to fill the empty section with the hash table from the string table
18:33:54pamauryeasy :p
18:34:01wodzcould work
18:35:19pamauryactually you would need to prefill the "empty" section with something not-empty of course, like pairs {addr, ptr to string} and then post process to replace by {addr, hash}
18:36:29wodzand strip the section with strings
18:36:50ulmutulwodz: I like this player, because it has all standard connectors.
18:37:09pamaurywodz: yeah exactly
18:37:22wodzulmutul: I meant its rather rare
18:37:48wodzpamaury: sounds like a plan. I need to make it work for strings first
18:38:33pamauryyeah definitely
18:44:59ulmutulsuccess with my TABLE_SETTING(..). I had to make my own scrollstrip_formatter().
18:45:52 Quit kuldeepdhaka (Ping timeout: 252 seconds)
18:50:36wodzhell, string version is working
18:53:00wodzpamaury: ^
18:54:21pamaurywodz: incredible :) nice work
18:55:02wodzI just resolved 2 external symbols. lets switch cube fully :-)
18:59:29 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:00:05 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
19:09:12 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
19:19:43 Join Rower [0] (
19:43:21 Part ulmutul
19:52:29 Nick DormantBrain is now known as SuperBrainAK (~andy@2001:470:8:a61::5f92:59a1)
20:05:48***Saving seen data "./dancer.seen"
20:49:28 Join rela [0] (~x@pdpc/supporter/active/rela)
20:56:22 Join Gallomimia [0] (~gallomimi@
21:15:13 Quit rela (Read error: Connection reset by peer)
21:22:34 Quit Rower (Ping timeout: 260 seconds)
21:41:11 Join Rower [0] (
21:57:36pamauryany ARM expert ?
21:57:49 Join tyllmoritz [0] (~robin@
22:05:49***Saving seen data "./dancer.seen"
22:08:31lebelliumkugel doesn't seem to be here but maybe someone else has an idea: FS #12951
22:08:32fs-bluebot 3Menu icons one pixel too down now (bugs, unconfirmed)
22:29:22 Nick SuperBrainAK is now known as DormantBrain (~andy@2001:470:8:a61::5f92:59a1)
22:35:35wodzpamaury: I am not an expert but You can try to ask the question actually :-)
22:37:26pamaurythis code doesn't seem to work: pamaury/9025537">
22:37:51pamauryI don't understand why
22:38:01pamaury(the IRQ and SVC mode are setup to use different stacks)
22:38:47wodzpamaury: first thing is that EABI dictates 8bytes alignment of the stack and you push 9 registers
22:40:20pamauryhum good point but since no code is run in IRQ mode except those two little pieces of code, is that relevant ?
22:40:39wodzdunno, just saying
22:41:37wodzyou force SVC 4bytes stack alignment as well
22:41:55wodzah, scrap that
22:42:54pamauryfor SVC I ensure 8-byte alignement
22:43:56wodzdo you reach _irq_handler or it fails earlier
22:45:09pamaurymost of the time the code seem to work but sometimes I get random crashes and out of the few times I can get an address it's a data abort usually on entry of _irq_handler (ie in the very first push {})
22:47:11wodzpamaury: the usual return sequence from irq is subs pc, lr, 4, I vaguely remember that you are supposed to update status on return
22:48:25 Quit tyllmoritz (Ping timeout: 252 seconds)
22:49:21pamauryI think that's what the "^" is supposed to do in ldm
22:49:26pamauryat least that's my understanding
22:50:01 Quit bluebrother (Disconnected by services)
22:50:06 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
22:52:54 Quit fs-bluebot (Ping timeout: 252 seconds)
23:11:05 Join fs-bluebot [0] (
23:24:00 Quit Rower (Quit: Hmmm...)
23:31:06 Quit y4n (Quit: We're fucking 3LN!)
23:36:01 Quit kuldeepdhaka (Quit: taataaaaa)
23:58:16 Quit bertrik (Ping timeout: 252 seconds)

Previous day | Next day