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 2010-08-27

00:01:37 Quit Jerom (Read error: Connection reset by peer)
00:06:02 Join _s1gma [0] (~d.d.derp@
00:06:36 Quit domonoky (Read error: Connection reset by peer)
00:08:43*pixelma always forgets the objdump syntax :/
00:10:14 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
00:10:40Bagderme too
00:11:18Bagderarm-elf-objdump -D −−target binary -marm [file]
00:13:44 Quit bertrik (Quit: :tiuQ)
00:13:55kaiscenecould someone explain what the glyph memory allocation does in regards to . is it a lookup cache? setting it to <20 causes display corruption in the WPS, == 1 causes simulator to lock up on skin load, >20 decreases skin memory usage hugely with no visible side effects.
00:15:41 Join b0hoon [0] (~quassel@
00:15:54Torneyou would generally use more than 20 characters of a font
00:16:23Torneit just sets how much memory the font gets
00:16:26pixelmakaiscene: are you sure the display corruption comes from setting something there? As it has been noticed before that was implemented
00:16:34Torneif you set it to 100 it gets enough memory for 100 different characters
00:16:47pixelmaTorne: 20 could be enough if you are using one font only for digits
00:17:02kaiscenepixelma: i'm experimenting with the new %Fl() parameter to change "glyph memory allocation", so it's due to changing that value
00:17:46 Quit pamaury (Ping timeout: 265 seconds)
00:17:54 Quit Hillshum (Quit: Page closed)
00:18:13pixelmaI see display corruption to different amounts in different themes which is also a bit different every time I reload the theme
00:18:28pixelmawhich player do you have by the way?
00:18:55kaisceneis this "glyph memory allocation" used to display only the current refresh/frame ?
00:19:15kaisceneso it gets reloaded with whatever text needs to be displayed per frame/refresh?
00:19:18pixelmahmm, seems to show often on monochrome targets
00:20:24pixelmakaiscene: no, it's how many characters of a font get stored in the "glyph cache" for quicker access to them. This speeds up WPS loading and drawing
00:20:30Tornekaiscene: whenever it needs to draw a character, if it's not in the cache it will ahve to read it from the font file on disk, kicking out the least recently used one
00:20:55kaisceneso this *is* a only a cache?
00:21:21Tornesort of.
00:21:24b0hoonTorne: hi. quick report - ipod 5.5 32MB: gives me 32MB :)
00:21:30Torneb0hoon: awesome
00:21:48Tornekaiscene: if you make it too small then whenever the text on the wps changes it will have to access the hard disk
00:22:06kaiscenei see. thank you both for explaining
00:22:13Tornekaiscene: e.g. if you set it to 100, and then a track starts which has a name in a foreign language which uses accents or other uncommon characters
00:22:31Tornethen it'll probably not have those uncommon characters saved in the cache of 100 glyphs
00:22:46Torneit will have to kick something else out to make room, which might then cause the *next* track name to need to read from disk again, and so on :)
00:23:00Tornei'm not surprised that setting it to very small numbers makes bad stuff happen
00:23:16pixelmaTorne: really?
00:23:39 Quit froggyman (Quit: CGI:IRC)
00:24:20Tornepixelma: not sure if it accounts for the overhead right..
00:24:25Tornei'm looking at it atm
00:24:54 Join krazykit [0] (~kkit@
00:26:57kaiscenewould you estimate a vast majority of song titles aren't usually more than 64 unique characters? so if I was using a font solely to display the song title, a glyph cache size of 64 would be satisfactory for most cases
00:27:30Tornekaiscene: er, ideally you want it to be enough to have *all the characters that will ever be used*
00:27:40Tornei.e. the total umber of unique characters in the user's mp3 collection ;)
00:27:51kaisceneyea. but i'm really hurting to decrease skin ram usage
00:28:01Torneyou don't want it to be just barely enough for *one song*, because that means it will have to access the disk every time a new song starts..
00:28:18kaiscenei see
00:28:29Tornefor a general purpose font you almost certainly want it to be at *least* 256
00:28:41Torneonly fonts used for one special purpose should really be using a tiny number
00:28:51Tornepixelma: it looks like it just multiplies the number of glyphs by the size of a font cache entry
00:29:06Tornepixelma: this doesn't allow for the fact that the buffer also has to contain the font header, the cache header and the offset table...
00:29:19 Quit ender` (Quit: Liar, n: one who tells an unpleasant truth.)
00:29:20Tornepixelma: so if you set it to a really small number it might not even be big enough to iitialise the font cache correctly
00:29:31Tornepixelma: i could be wrong..
00:29:35soapTorne, why have Rockbox detect iPod memory size and not Rockbox utility?
00:29:42pixelmaTorne: one of the arguments for it were "set it to small" for fonts you only for numbers, time, date, frequency and such
00:29:45Tornesoap: didn't we already do that?
00:29:46Tornepixelma: yes
00:29:52Tornepixelma: but i don't think that actually *works*, is my point
00:30:09Torneif you set it to 10, then the resulting buffer is probably only big enough for less characters than that
00:30:09soapTorne, THAT I don't know. I haven't paid enough attention to RBUtil it appears.
00:30:18Torneif you set it to 1 then it might be totally broken and cause crashes.
00:30:33Tornepixelma: because it doesn't allow for the per-font overhead, only the per-glyph overhead
00:30:44Torneunless it does in some way that i'm not seeing
00:31:27Tornesoap: anyway, whether rbutil does or not.. it doesn't seem like a bad sanity check
00:31:41Tornesoap: the code to detect is 6 instructions, though obviously you need a tiny bit to actually act on the outcome too ;)
00:31:54Tornesoap: then it works for manual installs/etc as well...
00:31:54pixelmasomeone should tell fredb if he comes here (or file a bug report). I still think that the screen corruption was already "in" before
00:32:11kaisceneyou think accessing the flash memory on the Clip+ to load glyphs after every song change would decrease battery life substainially?
00:32:26Tornekaiscene: unlikely
00:33:18Torneanyway if you are really desperate to save the ram i'd use the biggest number you can without running out of skin buffer ;)
00:33:25Tornebut try for 100-200 at least.. 64 is miniscule
00:33:42Tornejust a-z, A-Z and 0-9 is alrady 62 characters, remember
00:33:48Torneadd . : - ( ) etc
00:34:07Torneso even people with absolutely no unusual characters are going to frequently use 80+ chars
00:34:13kaisceneit would be nice to dynamically size the cache up to the parameter specified in %Fl
00:34:34Tornehow would that work?
00:34:57Torneif there's only one skin font then it could use the whole rest of the skin buffer for it, i guess
00:35:05Tornebut if there's more than one, how can it know what the best use of the memory is?
00:35:07kaisceneif the glyph to display isn't in the cache and the cache size is not > param in %Fl, increase cache size by 1 and insert glyph
00:35:32Tornethat would require a substantially more complicated cache impleentation
00:35:41Tornethe current one only works if the cache memory is all in one block
00:35:44kaisceneya :). that's why i said "it would be nice"
00:35:58Torneand it would alost certainly use more memory..
00:36:04Torneso you'd probably find you were worse off :)
00:36:08kaisceneoh lol
00:36:25Torneknowing it's all in one block is a space optimisation as well as making it simpler
00:36:30 Join Misanthropos_ [0] (
00:36:30Torneyou don't need to keep track of certain things
00:37:55 Quit einhirn (Read error: Connection reset by peer)
00:45:15 Quit b0hoon (Quit: - Chat comfortably. Anywhere.)
00:52:54***Saving seen data "./dancer.seen"
00:54:53 Quit bluebroth3r (Ping timeout: 265 seconds)
00:56:00 Quit GeekShadow (Quit: The cake is a lie !)
00:56:25 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
01:05:47 Quit Jennifur (Quit: I was raided by the FBI and all I got to keep was this lousy quit message!)
01:06:37 Join kugel [0] (~kugel@rockbox/developer/kugel)
01:07:53kugelpixelma: for *.o and *.elf objdump -D <file> is enough (-d if you only want the code, not data as well). for .bin/.codec/.rock you need −−binary -m arm as well
01:09:32pixelmakugel: thanks :) though I already found something in some old logs here and also where the crash after leaving early USB exactly happens
01:09:35kugelit's most useful to use objdump with the .elf files. they are completely linked, i.e. the fault addresses match with whatever the data abort (etc.) gives, while they still have useful debugging stuff in them (most importantly function, global variable names)
01:11:44pixelmain "lib/libskin_parser.a(skin_parser.o)", more exactly "9035954: 8b fb bf 903594e <_skin_parse+0x152>"
01:12:35 Join froggyman [0] (~me@unaffiliated/froggyman)
01:13:57*kugel doesn't know what bf should mean
01:14:48kugelI could help if you paste +- lines of the area when built for an arm target :)
01:14:56kugel+- 100*
01:15:35pixelmait's sh (on Ondio build)
01:16:50 Join fdinel [0] (
01:17:13pixelmaI could pastebin the whole function if you want
01:17:28 Quit jgarvey (Quit: Leaving)
01:17:49 Quit t0rc (Quit: Rawr)
01:18:13pixelma_skin_parse that is
01:18:52pixelma250 lines
01:20:07CIA-81New commit by kugel (r27900): Introduce a small api for loading code (codecs,plugins) from disk/memory. ...
01:20:14 Quit _s1gma (Remote host closed the connection)
01:20:15kugelsh doesn't tell me much :( but we could try
01:20:41kugeldoes the error only happen on sh?
01:22:08CIA-81r27900 build result: 1489 errors, 0 warnings (kugel committed)
01:22:57pixelmaI only know of the Archoses having early USB (connect USB from an off state and get the USB screen - for a while the statusbar is missing there and you get a crash if you disconnect and it would boot the "rest" if Rockbox)
01:24:16pixelmaor I only know of these having it for sure, I thought some other targets have early USB too but don't remember which and am really not so sure anymore)
01:26:30kugelI assume it says IllInst (or something like that) or 0x9035954?
01:28:14 Part toffe82
01:29:34pixelma"I09: CPUAdrEr at 09035954"
01:31:03kugelstrange. I doubt I can really help you with that (not tonight anyway), I guess it needs an sh expert
01:31:40kugelmaybe the same expert can tell me why that commit causes errors on sh
01:32:30pixelma might give a hint
01:36:17scorchekugel: (picking on you since you seem to be the only one here at the moment) Have you looked into submitting your GSoC code yet?
01:44:46kugelscorche: no, submitting code isn't "ready" already is it?
01:45:00scorcheit is afaict
01:45:03kugelIIRC they open the svn repos next monday
01:45:16scorchei guess so, but it seems set up to me
01:45:35scorchei guess they wanted to give the admins time to set up the project before telling students they can upload
01:45:47kugelanyway, the code sumbitting deadline is 10th sept or so, so I don't worry about that at the moment
01:46:16kugelbut I'll do it (at least) a few days before, I think they changed something this year
01:46:19scorchemight as well get it out of the way as soon as you can, but i do see that they havent told the students it is ready yet
01:46:49scorcheyes....there is this now:
01:48:08kugelI also need to have a quick talk with carol, I think they assume I have a set of patches which were committed by the mentor, but this isn't really the case with rockbox
01:48:26scorchei dont think they assume that
01:48:49kugelit sounded to me like that but I may be wrong
01:49:19scorcheit sounded to me like they expect the student of having some way of generating diffs/code written ;)
01:52:59kugelthat for sure :)
01:54:04kugelbut I think the noise introced by moving files is bigger than the actual code I've written, but that hopefully doesn't matter
01:57:22 Join JdGordon| [0] (
01:57:22 Quit JdGordon| (Changing host)
01:57:22 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
01:57:37 Part JdGordon|
01:59:10 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
01:59:38JdGordon|just as an FYI I'll be taking a break for a while, disabling flyspray emails and prob not spending muh time in here...
01:59:45 Part JdGordon|
02:02:10kugeland who fixes the skin engine bugs in that time? I don't think we have any other guy familiar with that area since the rewrite
02:02:48*kugel doesn't fucking understand the build errors
02:03:15 Join anewuser [0] (anewuser@unaffiliated/anewuser)
02:05:17DerPapstsim doesn't compile in cygwin anymore too
02:14:20 Quit Misanthropos_ (Quit: Ex-Chat)
02:15:44 Join Rob2222 [0] (~Miranda@
02:16:14 Quit amiconn (Disconnected by services)
02:16:16 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
02:16:29CIA-81New commit by kugel (r27901): Revert "Introduce a small api for loading code (codecs,plugins) from disk/memory." ...
02:16:36 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
02:16:56kugelhopefully someone can help me with that
02:17:08 Quit pixelma (Disconnected by services)
02:17:10 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
02:17:12 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
02:17:24 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
02:17:58 Quit Rob2223 (Ping timeout: 240 seconds)
02:18:28CIA-81r27901 build result: All green
02:18:39 Quit ps-auxw (Ping timeout: 276 seconds)
02:19:00 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
02:29:52CIA-81New commit by kugel (r27902): 2nd try: Introduce a small api for loading code (codecs,plugins) from disk/memory. ...
02:31:35CIA-81r27902 build result: All green
02:44:21kugeljhMikeS: ping
02:48:15 Join Strife1989 [0] (
02:51:19 Quit Strife89 (Ping timeout: 272 seconds)
02:52:56***Saving seen data "./dancer.seen"
02:56:00 Quit DerPapst (Quit: Leaving.)
02:59:40kugelI think I need to rename sleep()
03:13:54 Join krabador [0] (
03:32:33 Quit kugel (Remote host closed the connection)
03:38:14kaisceneclip can't get into the playlist menu since;revision=27897 . as soon as the down button is pressed it goes immediately to the pitchscreen - no opportunity for a long-press of the down button to show the playlist
03:59:27 Part froggyman
04:00:27 Join fyrestorm [0] (
04:10:48 Join steve|m1 [0] (
04:10:48 Quit steve|m (Disconnected by services)
04:11:00 Nick steve|m1 is now known as steve|m (
04:13:54 Quit Strife1989 (Quit: Quick reboot.)
04:16:09 Quit anewuser (Quit: FAMICOMPO 7 - August 15 Through August 31 JST -
04:26:58 Join [Saint] [0] (S_a_i_n_t@
04:27:04 Quit S_a_i_n_t (Ping timeout: 240 seconds)
04:27:39 Quit pixelma (Disconnected by services)
04:27:41 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:27:43 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:28:41 Quit amiconn (Disconnected by services)
04:28:43 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:29:03 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:30:41 Quit TheSeven (Ping timeout: 272 seconds)
04:31:33 Join Strife89 [0] (
04:35:18 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:36:13 Quit [Saint] (Ping timeout: 260 seconds)
04:37:54 Join S_a_i_n_t [0] (S_a_i_n_t@
04:39:29 Quit Barahir_ (Ping timeout: 260 seconds)
04:41:07 Join Barahir [0] (
04:43:37 Join froggyman [0] (~me@unaffiliated/froggyman)
04:52:20 Part froggyman
04:52:57***Saving seen data "./dancer.seen"
05:01:39 Join bunnyboi [0] (
05:12:04 Quit MethoS- (Remote host closed the connection)
05:22:08 Quit krabador (Quit: Sto andando via)
05:26:36 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
05:40:32 Join froggyman [0] (~4845cf9e@unaffiliated/froggyman)
05:42:02 Quit Horscht (Quit: Verlassend)
05:45:31 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:46:37 Quit t0rc (Quit: Buh-bye.)
05:50:04 Quit panni_ (Ping timeout: 240 seconds)
05:54:35 Quit z35 (Ping timeout: 265 seconds)
06:01:52 Join z35 [0] (
06:06:22 Quit Strife89 (Quit: Sleep.)
06:13:15 Quit bunnyboi (Read error: Connection reset by peer)
06:13:16 Join Jennifur [0] (
06:17:11 Quit z35 (Ping timeout: 265 seconds)
06:21:49 Join z35 [0] (
06:30:36 Join h3po [0] (
06:34:42 Quit z35 (Ping timeout: 265 seconds)
06:35:58 Join z35 [0] (
06:48:09 Join hebz0rl [0] (
06:48:16 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
06:49:41 Quit storm` (Ping timeout: 260 seconds)
06:52:59***Saving seen data "./dancer.seen"
06:53:48 Join storm` [0] (
06:54:48 Quit froggyman (Quit: CGI:IRC (EOF))
06:55:29 Quit z35 (Ping timeout: 265 seconds)
06:58:52 Quit Aslai (Ping timeout: 265 seconds)
07:09:05 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
07:09:23 Part JdGordon
07:12:16 Join z35 [0] (
07:14:40 Join Buschel [0] (
07:15:33 Quit h3po (Quit: Leaving.)
07:16:26 Join h3po [0] (
07:17:14 Quit z35 (Ping timeout: 265 seconds)
07:22:18 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
07:22:39 Join z35 [0] (
07:27:23 Quit z35 (Ping timeout: 265 seconds)
07:28:50 Quit h3po (Ping timeout: 276 seconds)
07:29:55 Join stoffel [0] (
07:31:13 Join z35 [0] (
07:34:39 Quit t0rc (Quit: pewpew)
07:36:05 Quit avacore^ (Ping timeout: 265 seconds)
07:39:57 Quit z35 (Ping timeout: 265 seconds)
07:41:21 Quit BHSPitMonkey (Quit: Ex-Chat)
07:45:55 Join z35 [0] (
07:46:31 Quit hebz0rl (Quit: Ex-Chat)
07:55:29 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
07:55:54 Quit z35 (Ping timeout: 265 seconds)
08:00:18 Join z35 [0] (
08:01:49 Quit Jennifur (Quit: IRC is just multiplayer notepad)
08:27:09 Join DerPapst [0] (
08:30:47 Join ender` [0] (
08:35:11 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:41:15 Join bieber [0] (
08:51:42 Join LinusN [0] (~linus@rockbox/developer/LinusN)
08:53:01***Saving seen data "./dancer.seen"
08:55:26 Quit bmbl (Ping timeout: 258 seconds)
09:08:19 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
09:13:49 Join einhirn [0] (
09:25:21 Quit bertrik (Read error: Connection timed out)
09:36:51 Join petur [0] (~petur@rockbox/developer/petur)
09:37:10 Join Jaykay [0] (
09:48:55 Quit DerPapst (Quit: Leaving.)
09:53:28 Quit Rob2222 (Quit: Rob2222)
09:53:55 Join Rob2222 [0] (
09:54:23 Join T44 [0] (
09:57:42 Quit Topy44 (Ping timeout: 265 seconds)
10:11:35 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:14:35 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:29:11 Join Jaykay_ [0] (
10:29:30 Quit Jaykay (Disconnected by services)
10:29:38 Nick Jaykay_ is now known as Jaykay (
10:33:59 Join DerPapst [0] (
10:44:12 Join AndyI [0] (~pasha_int@
10:45:03 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:45:42 Quit einhirn (Ping timeout: 255 seconds)
10:48:31 Quit seani (Ping timeout: 258 seconds)
10:50:45amiconnpixelma: The corruption I've seen on Clip+ affected the whole top line, not just top left corner
10:51:10amiconnRegarding early usb - all targets have that
10:51:22amiconnIt's just difficult to enter if there is also bootloader usb
10:53:02***Saving seen data "./dancer.seen"
10:53:40 Join esperegu [0] (~quassel@
11:01:29 Join seani [0] (~seani@
11:10:20pixelmaamiconn: screen corruption varies for me - most of the time it's just the top left corner and in case of cabbiev2 on the Ondio then the progressbar is a bit misplaced too, I know that on the Clip's cabbiev2 the top bar also is in the top row. Screen corruption can be way worse, see the imagebin link
11:10:46pixelmaand sometimes there is no visible screen corruption
11:11:54pixelma...I meant the progressbar is in the top row on the Clip's cabbiev2
11:20:57 Join wodz [0] (
11:21:43wodzI finally pinpointed what causes problems with USB transfers to MPIO
11:23:07wodzIt's buggy genesys usb-ata bridge + MPIO overriding usb descriptors for device so genesys specific workarounds arent picked by kernel driver
11:29:21wodzhmm the problem with transfers to my mini1G remained :/ So this are two separate issues, shit
11:31:55 Quit stoffel (Ping timeout: 240 seconds)
11:35:16 Quit Jaykay (Ping timeout: 265 seconds)
11:39:41 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:40:00 Join freddyb [0] (
11:48:03 Quit n1s (Ping timeout: 264 seconds)
12:06:42 Quit sasquatch (Ping timeout: 255 seconds)
12:07:52 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:23:51 Join sasquatch [0] (
12:24:34 Quit wodz (Quit: Leaving)
12:31:34 Join n1s [0] (~n1s@rockbox/developer/n1s)
12:32:04 Quit AndyI (Read error: Connection reset by peer)
12:33:13CIA-81New commit by kugel (r27903): Make getcwd match the posix variant, make get_current_file() behave similar to it and add a few sanity checks.
12:35:18 Join swilde [0] (
12:35:23 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
12:35:31CIA-81r27903 build result: All green
12:39:38 Quit dfkt (Ping timeout: 265 seconds)
12:48:15 Quit T44 (Read error: Connection reset by peer)
12:52:24kugelgah this bloody annoying mingw32 thing
12:53:03***Saving seen data "./dancer.seen"
12:54:31freddybRegarding the glyph memory allocation earlier between Torne and pixelma: the font structure is already allocated in font_table[] in skin_fonts.c. Loading a font with room for one glyph in a VP that only shows one glyph is ok.
12:55:03Tornenot the font structure.. the things font.c loads into the buffer
12:55:05Tornelike the offset table
12:59:03kugelcan I slap mingw guys for making the size parameter an int on their library?
12:59:08Tornepf->buffer_size is glyphs * get_glyph_size(), that's the buffer pointed to by pf->buffer and pf->buffer_start when it calls font_load
12:59:30 Join stoffel [0] (
13:03:29TorneOkay, i'm not *quite* right but i still think it can break if it's very small
13:03:42Torneit doesn't use any extra space after it's loaded the font
13:04:12Tornebut if you did glyphs == 1 and get_glyph_size() was small because the font wasn't very big, then the resulting buffer might be less than FONT_HEADER_SIZE
13:04:27Tornethe loading process assumes the buffer is at least that big, maybe more? not entirely sure
13:04:32Torneand your code doesn't guarantee that
13:06:33 Quit n1s (Quit: Lämnar)
13:09:56 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:12:12Tornei'm not entirely sure :)
13:12:24Tornebut it doens't look right, still
13:20:21freddybstruct font has a struct font_cache included, not just a pointer. get_glyph_size() pads for the cache overhead
13:20:45Torneyes, i realised that just now
13:20:55Tornebut it still uses the buffer itself as temporary space to load the header during font loading
13:21:06Torneand doesn't actually check that it's big enough
13:21:18Torneso if the buffer is smaller than FONT_HEADER_SIZE i think ti will just overwrite whatever is after that in the skin buffer?
13:21:27Tornesee the call to read() in internal_font_load
13:25:53freddybget_glyph_size() has its own struct font (not a pointer)
13:26:16Tornenothing to do with that
13:40:56pixelmafreddyb: welcome and thanks for reading logs :)
13:42:41mc2739kugel: w32 cross-compiled sims are broken again by r27902
13:43:04kugelmc2739: yes I know :)
13:43:18freddybI see now. I'm kinda thinking about converting the header code to just read directly from the font file...
13:43:34freddybpixelma: Thanks.
13:43:46 Quit TheSeven (Ping timeout: 276 seconds)
13:52:28kugelfreddyb: hey, welcome from me too \0/
13:53:34*gevaerts joins in the welcome celebrations
13:54:12freddybThanks, everyone. When do they pass out the paychecks?
13:54:28Tornepayment is in beer at devcon
13:54:41freddybGood to know.
13:54:42*n1s joins the welcomming too :)
13:55:04*pamaury too
13:55:38freddybThanks again.
13:56:25 Join MethoS- [0] (~clemens@
13:58:44n1sam i missing something or is FS #11573 a patch for a theme on the theme site?
13:58:59Torneyes, it is
13:59:01n1s(is this the way we want fixes for those)?
13:59:06Torneprobably not
13:59:52n1si know next to nothing about the themes so i'll leave it for someone else then
14:00:40Torneyeah that was what i did
14:02:04 Quit dfkt_ (Read error: Connection reset by peer)
14:36:48 Join panni_ [0] (
14:38:28CIA-81New commit by kugel (r27904): Use system headers a bit more: use host's fcntl.h for O_RDONLY etc. ...
14:39:53 Join teru [0] (
14:40:38CIA-81r27904 build result: 808 errors, 0 warnings (kugel committed)
14:41:46 Join balintx [0] (
14:43:40 Join _s1gma [0] (~d.d.derp@
14:44:05n1swhoops, seems like i broke the wiki access control
14:44:34n1scan a wiki admin revert the last edit to WikiUsersGroup ?
14:45:06n1spinging Bagder Zagor
14:48:00pixelmaLinusN too, maybe :)
14:48:32CIA-81New commit by kugel (r27905): Fix checkwps.
14:48:35Zagorooh, pretty :)
14:50:45Zagorn1s: fixed
14:51:00CIA-81r27905 build result: All green
14:51:18 Join dfkt [0] (dfkt@unaffiliated/dfkt)
14:51:29n1sZagor: thanks!
14:53:05***Saving seen data "./dancer.seen"
14:56:00balintxis it possible rockbox meshed up my fuzev2? i tried to copy a directory from an external SD card to the internal memory, and it said Copy( or paste?) was not successful. When I tried to open this directory, it opened the root of the internal memory, like a symbolic link. I tried to delete from USB, but rockbox was not even recognising the usb, just starts charging! I have booted OF, then plugged usb, and windows says Unrecognized device
14:56:00balintx both on MTP or MSC mode. The Rockbox build is r27430.
14:57:31kugelcan I rename sleep()?
14:58:02kugeleither that, or making it "unsigned sleep(unsgined)", it's erroring in the android port
14:59:29balintxnow it is "ANIC*" "kov main"
14:59:43balintxmay *PANIC*" but it not show properly
15:00:37balintxmaybe a formatting solves my problem..?
15:04:15 Join evilnick [0] (0c140464@rockbox/staff/evilnick)
15:07:20kugelgevaerts: any opinion?
15:10:37 Quit kaiscene ()
15:11:04balintxformatted the internal memory using original firmware, the problem is same, now I can't use rockbox tough
15:13:42 Quit krazykit (Quit: bbl)
15:14:11balintxidea? :/
15:14:44gevaertskugel: changing the prototype seems best I think
15:15:24kugelI don't feel strong but our sleep() also behaves quite different
15:15:58 Quit DerPapst (Quit: Leaving.)
15:16:22CIA-81New commit by teru (r27906): Fix bug adding a playlist to another playlist doesn't finish if it has UTF-8 BOM introduced by r24718.
15:17:38balintxinteresting. it loaded an sd card to database, and after 10 minutes the usb was recognized
15:18:02 Join jgarvey [0] (
15:18:47gevaertskugel: the #define sleep(x) sim_sleep(x) doesn't help?
15:19:19kugelnot exactly, because it also substitutes sleep() in unistd.h :)
15:19:20CIA-81r27906 build result: All green
15:19:51gevaertsah, I see how that can be an issue :)
15:23:33CIA-81New commit by fredwbauer (r27907): Added myself
15:25:23CIA-81r27907 build result: All green
15:26:31freddybKugel: Got one right...
15:27:25 Quit _s1gma (Remote host closed the connection)
15:29:24 Join anewuser [0] (anewuser@unaffiliated/anewuser)
15:29:46freddybAs a follow up to the skin fonts change I'm kinda thinking we should only write the glyphcache file for FONT_UI. That way a small skin font won't chop it too short.
15:36:54Tornethere's a downside to that, though, which is that if your UI is in english but you have lots of songs in another language it won't cache those chars if the WPS displays it in a different font
15:37:08Torneit's probably not that significant, though :)
15:38:17*kugel compiles the android app without uisimulator/common/io.c
15:39:28 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
15:41:35kugelbut it's broken :\
15:43:41 Quit Dreamxtreme (Quit: +++ OK ATH OK)
15:46:51 Quit dfkt (Read error: Connection reset by peer)
15:48:18freddybIt would be simple to use multiple glyphcache files. I'm going to experiment with one for each glyph slot.
15:50:23CIA-81New commit by teru (r27908): tweak lrcplayer.
15:52:41CIA-81r27908 build result: All green
15:55:35 Join Dreamxtreme [0] (~Dreamxtre@
16:02:04 Part t0rc ("Adieu")
16:04:15 Join dfkt [0] (dfkt@unaffiliated/dfkt)
16:05:13kugelit's interesting that remove_dir() (onplay.c) isn't horribly broken
16:05:35kugelconsidering what sim_readdir does (always return a static dirent struct)
16:05:56 Join DerPapst [0] (
16:06:02 Part LinusN
16:10:42kugeloh that directory handling is a can of worms actually
16:25:41 Quit petur (Quit: connection reset by beer)
16:35:54 Join Horscht [0] (~Horscht@xbmc/user/horscht)
16:37:52 Quit teru (Quit: Quit)
16:50:54 Join froggyman [0] (~4845cf9e@unaffiliated/froggyman)
16:53:09***Saving seen data "./dancer.seen"
16:54:02 Join Judas_PhD [0] (
16:54:39 Join toffe82 [0] (~chatzilla@
17:01:31 Quit Judas_PhD (Quit: This is a quitting message)
17:01:58 Join Judas_PhD [0] (
17:04:43 Quit dfkt (Read error: Connection reset by peer)
17:05:53 Quit Judas_PhD (Client Quit)
17:06:26 Join Judas_PhD [0] (
17:06:39 Quit Judas_PhD (Client Quit)
17:07:00 Join Judas_PhD [0] (
17:07:03 Part Zagor
17:07:54 Join kaiscene [0] (
17:08:24 Quit anewuser (Quit: FAMICOMPO 7 - August 15 Through August 31 JST -
17:08:28 Quit Judas_PhD (Client Quit)
17:08:51 Join Judas_PhD [0] (
17:09:11 Quit Judas_PhD (Client Quit)
17:09:36 Join Judas_PhD [0] (
17:10:32CIA-81New commit by fredwbauer (r27909): limit font_cache preloading to cache capacity
17:13:28CIA-81r27909 build result: All green
17:19:52 Quit froggyman (Quit: CGI:IRC)
17:43:57kugeldecent speed up from using native dir functions directly
17:44:01 Join Lear [0] (chatzilla@rockbox/developer/lear)
17:44:05 Join dfkt [0] (dfkt@unaffiliated/dfkt)
17:49:29 Quit S_a_i_n_t (Ping timeout: 276 seconds)
17:53:31 Quit Lear (Quit: ChatZilla 0.9.86 [Firefox 4.0b5pre/20100826030920])
18:01:34 Join Kitr88 [0] (
18:02:21 Join fdinel [0] (
18:02:47 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:03:20 Quit Kitar|st (Ping timeout: 240 seconds)
18:05:51 Quit Kitr88 (Ping timeout: 240 seconds)
18:11:24 Join Kitar|st [0] (
18:14:12 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
18:14:29 Quit bmbl (Client Quit)
18:15:09 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
18:31:20 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
18:31:32 Join Jerom [0] (~jerome@
18:35:52 Join t0rc_ [0] (
18:37:28 Join Strife89 [0] (
18:37:42 Quit t0rc (Ping timeout: 252 seconds)
18:38:47 Join [sko] [0] (~sko]
18:40:28 Quit pamaury (Remote host closed the connection)
18:41:51 Quit t0rc_ (Ping timeout: 245 seconds)
18:43:28 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:48:56CIA-81New commit by fredwbauer (r27910): close file when cached font load fails
18:51:49 Join Highlander [0] (
18:52:41CIA-81r27910 build result: All green
18:53:11***Saving seen data "./dancer.seen"
18:53:52 Quit Jerom (Quit: Leaving.)
19:01:30 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
19:03:01 Quit bertrik (Read error: Connection timed out)
19:03:22 Part timccccc
19:03:26 Join MichaelAdams [0] (
19:03:34 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:04:05 Join Jerom [0] (~jerome@
19:24:23 Quit sasquatch (Ping timeout: 276 seconds)
19:26:58 Join sasquatch [0] (
19:29:01 Quit sasquatch (Client Quit)
19:29:19 Quit Jerom (Quit: Leaving.)
19:36:42 Join sasquatch [0] (
19:40:27kugelthere's no way to get from dirent pointer to the complete path?
19:40:42kugeldirent->d_name only has the name of the entry, not its parents
19:43:14 Join Aslai [0] (
19:44:08 Join Jaykay [0] (
19:44:30 Quit Strife89 (Quit: Strife89 boots gevaerts! :P)
19:45:57 Quit toffe82 (Read error: Connection reset by peer)
19:46:05 Join Strife89 [0] (
19:52:55 Join toffe82 [0] (~chatzilla@
20:20:36 Quit stoffel (Ping timeout: 245 seconds)
20:25:27CIA-81New commit by fredwbauer (r27911): change get_glyph_size() to font_glyphs_to_bufsize(). fixes a bug when font glyph buffer < font header
20:27:02 Quit l403 (Ping timeout: 265 seconds)
20:28:20CIA-81r27911 build result: All green
20:29:24 Quit toffe82 (Remote host closed the connection)
20:39:16 Nick BeFalou_ is now known as BeFalou (~mamutoi@unaffiliated/befalou)
20:41:30BeFalouis there any runtime improvement since 3.6 with fuzev1?
20:42:12BeFalouI see that v2 runs for more 20 hours which is great
20:43:36 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
20:47:50 Join toffe82 [0] (~chatzilla@
20:49:55balintxno idea
20:53:12***Saving seen data "./dancer.seen"
21:02:22 Quit Aslai (Ping timeout: 252 seconds)
21:06:21bertrikI don't remember anything was done to improve runtime for fuze v1 since 3.6
21:08:22 Quit Zarggg_ (Read error: Connection reset by peer)
21:18:28 Join stoffel [0] (
21:37:25 Join McMusic [0] (
21:46:35kugelio.c less android working :)
21:47:06kugeldeath to libuisimulator!
21:49:51freddybGood job! Did you figure out the path thing or did you have to chase up on level at a time?
21:50:21 Quit seani (Ping timeout: 240 seconds)
21:51:35 Join powell14ski_ [0] (
21:55:58 Quit kugel (Ping timeout: 265 seconds)
22:01:38 Quit McMusic (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
22:02:40 Join seani [0] (~seani@
22:13:56 Quit pamaury (Ping timeout: 245 seconds)
22:14:20 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:14:41kugelfreddyb: what do you mean?
22:17:59 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
22:18:54 Join Strife1989 [0] (
22:20:21 Quit Strife89 (Ping timeout: 240 seconds)
22:24:04kugelah well
22:24:09 Quit evilnick (Quit: Page closed)
22:24:12kugelI figured out that it's not possible
22:24:42kugelnow I just save the dir path in the DIR* pointer and use it when getting to it
22:25:25 Join bunnyboi [0] (
22:33:03 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
22:37:19 Quit kugel (Remote host closed the connection)
22:38:37 Quit dfkt (Read error: Connection reset by peer)
22:47:17 Quit freddyb (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723085406])
22:53:16***Saving seen data "./dancer.seen"
23:03:38 Quit Highlander (Quit: Quitte)
23:08:36 Nick Strife1989 is now known as Strife89 (
23:09:11 Quit stoffel (Remote host closed the connection)
23:10:24 Join DerPapst1 [0] (
23:10:40 Quit DerPapst (Ping timeout: 240 seconds)
23:19:47 Quit Llorean (Quit: Leaving.)
23:22:43 Quit [sko] (Quit: Leaving.)
23:24:29 Quit Jaykay (Ping timeout: 240 seconds)
23:30:24 Join Misanthropos_ [0] (
23:40:01 Quit MichaelAdams (Quit: CGI:IRC (EOF))
23:44:46 Join Jennifur [0] (
23:45:43 Quit bunnyboi (Ping timeout: 252 seconds)
23:55:07 Quit jgarvey (Quit: Leaving)

Previous day | Next day