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 2010-02-17

00:00:03gevaertsbut that's not what I care about right now. Getting the things to work for real is more important
00:00:07*stripwax knows that the wps doesn't use 0 CPU while the screen is off, but still...
00:00:20 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
00:00:20gevaertsstripwax: they're called from the wps redraw code
00:00:30stripwaxcool
00:00:38kugelstripwax: depends, the skin engine "sleeps" if the backlight is off, but that only applies for targets with backlight and non-transflexive screens
00:00:39gevaertsand they're loaded in the wps buffer
00:01:05stripwaxgevaerts - so the refresh rate is determined by the wps update rate? or could the pebble instruct the wps how often to refresh?
00:01:44stripwaxkugel - I'm pretty sure the skin engine doesn't sleep on ipod video if the lcd controller is powered off when backlight is off (ipod video is translective but lcd can be turned off too)
00:01:58Buschelsudoman: I never saw any 3G bench before. but yours looks strange. maybe a new battery is a good idea −− they are not too expensive. as soon as you have one please report the runtime again, of course :o)
00:01:59gevaertsstripwax: it probably should. I'm still not entirely sure that I can get the actual loading to work properly for all CPU architectures, so I don't care too much right now :)
00:02:07stripwaxotherwise I'd expect ipod video runtimes with/without wps to be essentially the same
00:02:21kugelI believe the ipodvideo updates if the backlight is off but stops updating if the lcd sleep kicks in
00:02:27JdGordon_stripwax: you can easily disable the wps if you want
00:02:43stripwaxgevaerts - yep - I agree, proving the concept right now probably more important!
00:03:01JdGordon_if pebbles works we can maybe move the playlist viewer to it to show a complicated example
00:03:03stripwaxJdGordon - true. but, I kinda feel I shouldn't have to / it shouldn't make a difference
00:03:31stripwaxkugel - hm, maybe it was changed since the most recent runtime changes on ipodruntime
00:03:50 Quit m3dlg (Quit: RAGE QUIT)
00:03:56JdGordon_without being able to call local functions I wonder how painful pebbles are going to be to make though
00:04:01kugelbasically, the skin engine calls lcd_active() to see whether it should update or not
00:04:02 Join CptJesus [0] (~60ea966d@gateway/web/freenode/x-kkvcgisqvmjkffzx)
00:04:45CptJesusHi everyone
00:04:48sudomanbuschel: the apple firmware shows a series of icons for a "charging battery" and an icon for a "charged battery". does this mean that there is software charge control/monitoring in the 3G or is it just a timed progression? i'm curious because i don't know whether to charge with the apple or rockbox firmware
00:05:27kugelJdGordon_: so you're not going to add %ac etc to the playlist viewer?
00:05:31TheSeventhere is software monitoring at least
00:06:17kugelI wonder why it doesn't simply contain sublines so that the normal skin engine code would handle this all?
00:06:19TheSevennot sure about software control on those old devices
00:06:23 Quit ender` (Quit: The word 'politics' is derived from the word 'poly', meaning 'many', and the word 'ticks', meaning 'blood sucking parasites'. -- Larry Hardiman)
00:06:43Buschelsudoman: I am charging via apple firmwae as this is a lot faster on my 5.5G. I assume this is the same on the 3G
00:06:49CptJesusI've been trying to install rockbox on my 30 gb iPod video...and I've been completely unable to install the bootloader. ipodpatcher.exe has a write failed error. Its being run as an administrator. If anyone could help me, that would be really awesome...I can't for the life of me figure out the issue
00:06:55sudomaninteresting, because there doesn't seem to be any software monitoring installed with the rockbox firmware
00:07:11JdGordon_kugel: not any time soon. Ideally get_line() should be directly usable from the playlist viewer, to cut down on *everything* a very stripped down parser was reimplemented for it which only stores the token id and not a full token struct for each item
00:07:12sudoman*on a 3G
00:07:18 Join Rob2223 [0] (~Miranda@p4FDCAC79.dip.t-dialin.net)
00:07:25JdGordon_no conditionals are allowed, no sublines, no difficult tags
00:07:51JdGordon_I'd like the playlist viewer to become much more integrated so its not a simple skin reimplementation
00:08:36kugelyes that would be nice
00:08:45pamauryTheSeven: did you managed to test the nano2g thing ?
00:08:59kugelbut IMO at the very least it should support align tags
00:09:00TheSeveni'm in the process of doing so
00:09:05evilnick_BCptJesus: If there's an error message then it'd be helpful to post that
00:09:14TheSevenneeded to clean up my working copy a little bit to make it work again ;-)
00:09:19CptJesusAlright, 1 sec, I'll retreive it again
00:09:37JdGordon_kugel: then by all means, go for it. a naive implementation could do like %s where its the whole line
00:09:37Buschelsudoman: I just submitted a patch to take care of your runtime tests. As nobody else has got some better measurements I took yours :o) So please report, if a new battery changes the behaviour.
00:09:40BuschelGood night
00:09:41JdGordon_which wouldnt be too complicated
00:10:03sudomanbuschel: thanks a bunch ;)
00:10:03TheSevendamn, that build machine is slow as hell
00:10:08 Quit Buschel ()
00:10:41TheSeveninstalling...
00:10:54 Quit Rob2222 (Ping timeout: 256 seconds)
00:11:28CptJesusevilnick: moving images to create room for new firmware −−> Error reading from disk: [ERR] Write failed −−> :invalid argument −−> [ERR] Image movement failed −−> [ERR] −−install failed
00:11:47TheSevenrolo'ing...
00:12:33*TheSeven yells at rolo
00:12:38evilnick_BCptJesus: Can you access the drive through explorer (or equivalent)?
00:12:52CptJesusYes, its accessible. And I am using Windows
00:13:02TheSevenpamaury: seems to be fine, but i don't have high nesting levels
00:13:15evilnick_BHmmm, actually, you should be able to use rbutil, not ipodpatcher.exe
00:13:19pamauryAt least it doesn't crash like before ?
00:13:37TheSevenyep, didn't get a crash yet
00:13:44TheSevenand i think it used to crash on boot
00:13:56CptJesusevilnick_b: I tried that. It simply fails to install the bootloader. Do you need a log?
00:14:01TheSevendircache thread is at 4% stack usage
00:14:10pamauryLogically, if the fault was on dircache, it can't crash anymore with depth level < 16
00:14:29pamaurywow, that's strange !
00:15:04pamauryTheSeven: can you disable dircache and renable it (and reboot) ? To see if stack usage maintain
00:15:39evilnick_BCptJesus: It does say that you should be logged in as an Administrator, which *may* be different to Running as Administrator. Can you try that out?
00:16:13evilnick_BCptJesus: And is it a Windows iPod?
00:16:13CptJesusevilnick_b: I am logged in as administrator =(
00:16:17CptJesusYes
00:16:23CptJesusFormatted with FAT32
00:16:31CptJesusBy iTunes
00:16:45sudomanwhat if CptJesus reformats?
00:17:07gevaertsreformatting won't help if it's the bootloader install that fails
00:17:18TheSevenpamaury: it stays at 4%, whether it's enabled or not
00:17:23CptJesus[RbUtil] Bootloader Post-Installation, error state: true
00:17:30CIA-8New commit by 03pamaury (r24711): Remove obsolete comment.
00:17:35CptJesusThat's all that the log gives me
00:17:58TheSevenif it's enabled, i get a splash saying "scanning <something>..." very very briefly (a quarter of a second maybe?)
00:18:07TheSevenis that suspiciously fast=
00:18:07 Quit parafin (Read error: Operation timed out)
00:18:08TheSeven?
00:18:15pamauryTheSeven: that's too low to be real, 4% of DEF+0x200=0x600 is 60 bytes ! I can't believe it scan the whole disk with 60bytes of stack
00:18:23pamauryIs it still enabled ?
00:18:31pamaury(check the dircache option in debug menu)
00:18:55TheSeveni disabled it, rebooted, checked stack: 4%, enabled it, rebooted, got the scanning splash, checked stack: 4%
00:19:30 Quit bertrik (Quit: De groeten)
00:19:34gevaertsCptJesus: rbutil has a system information item in a menu somewhere
00:19:37TheSevenCache initialized: yes
00:19:37pamauryDid you check the dircache entry in the debug menu, the one will statistics like the size, number of entries, ...
00:19:42pamauryHum
00:19:43*gevaerts can't remember where exactly
00:19:55CptJesusI have it, what do you need from it?
00:20:21pamauryThat's stupid but this 4% seems suspicious to me :)
00:20:37TheSevensize: ~100KB, limit: ~6MB, reserve: 0/65536, scanning took 0 seconds, 1389 entries
00:20:57gevaertsCptJesus: can you paste the entire thing on http://pastebin.com and give us the resulting link?
00:21:15CptJesusDo you want the System Trace or the System Info?
00:21:16pamauryWhat is the memory type of the nano2g ?
00:21:21TheSevenhow much stack usage is reported on your devices?
00:21:26TheSevenpamaury: 32MB of SDRAM
00:21:31gevaertsSystem Info I think
00:21:36pamauryTheSeven: sorry, storage ;)
00:21:44TheSeven8GB flash
00:21:51 Join m3dlg [0] (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
00:21:55pamauryOn mine, something like 50% at least with 0 depth level
00:22:02CptJesusgevaerts: http://pastebin.com/d1683087e
00:22:08 Quit Barahir (Ping timeout: 260 seconds)
00:22:11 Join parafin [0] (parafin@paraf.in)
00:22:12CptJesusThats system info. I can give you the trace if you want it
00:22:13pamaury(mostly taken by the fat core+ata code I imagine)
00:22:27gevaertshm, it doesn't see any USB devices
00:22:31gevaertsThat's strange
00:22:36evilnick_BAttached USB devices - nothing??!
00:22:43pamauryTheSeven: hum, perhaps that's because it's flash then but still...
00:22:50gevaertsCptJesus: is itunes running?
00:23:01CptJesusYet both rbutil and ipodpatcher see the iPod. And no, it is not
00:23:04TheSevenpamaury: the "ata code" (ftl in that case) won't use much stack, I reduced that to the absolute minimum possible (no sector buffers at all)
00:23:30TheSevenbut it will use more than 60 bytes for sure
00:23:49 Join Barahir [0] (~jonathan@gssn-5f756b48.pool.mediaWays.net)
00:23:52gevaerts6.1 is windows 7, right?
00:23:57CptJesusCorrect
00:23:58pamauryYes, but even if you fit the ftl code in say, 10 bytes, my code uses at 40 bytes, the fat uses way more bytes
00:24:05pamaury*at least
00:24:35pamauryNo that's impossible, something is wrong but if you say it's working...
00:25:05evilnick_BCptJesus: Does Autodetect work, i.e. does it find the correct drive?
00:25:10CptJesusevilnick_B: Yes
00:25:18*gevaerts doesn't know what's going on
00:25:44CptJesusgevaerts: That makes 2 of us =(
00:25:54gevaertsI suspect some windows 7 specific thing, but I could easily be wrong
00:25:59evilnick_BCptJesus: Maybe you could do the system trace too
00:26:08evilnick_Bs/do/put on pastebin/
00:26:22CptJesushttp://pastebin.com/m516e3f87
00:27:04pamauryTheSeven: I know I'm boring but would you mind doing a little modification and recompile dircache.c to retest ? ;-)
00:28:03evilnick_BCptJesus: I suspect the latter part of that log could be useful ;)
00:28:40pamauryoops, I just noticed that my dircache code doesn't propage dircache generation errors. Perhaps that's related but then I don't understand how it works
00:29:03CptJesusevilnick_B: The part where it spams [Settings] GET U: "mountpoint" "G:/" ?
00:29:22evilnick_BCptJesus: Yeah, that's almost definitely not a good sign
00:29:51TheSevenpamaury: damn, i smashed my ftl
00:29:55 Quit MagusG (Read error: Connection reset by peer)
00:29:56CptJesusevilnick_B: Am I allowed to blame witchcraft at this point?
00:30:13pamauryTheSeven: what does that mean ?
00:30:40TheSeveni might need to restore the whole thing, but I'll try to debug some things first
00:30:53gevaertsJdGordon_: the way things are set up now, you can load (well, not right now, that bit needs work...) the same pebble multiple times. We could do separate load and use tokens to fix that, or we could cache the names. If my current work succeeds and I get globals working, reusin the same loaded pebble for multiple instances could cause trouble though
00:31:25pamauryTheSeven: so you can try and modify something ?
00:31:30gevaertsHaving an extra pebble_api shouldn't be too hard
00:31:39TheSevencurrently it won't mount
00:31:39JdGordon_resuing the same blog would be far more useful than allowing globals inside the pebble
00:31:44evilnick_BCptJesus: That or Win7 :) - I believe that the devs who work on rbutil are away from IRC now, but they'll presumably check the logs.
00:31:58JdGordon_not allowing local funcitons is still the bigest issue imo
00:32:02*JdGordon_ back in 30
00:32:11gevaertsyes, maybe that will be fixable though
00:32:56CptJesusevilnick_B: Definitely witchcraft. That seems like the most logical assumption. However, for reference, I can manually put the rockbox files on the ipod through explorer. The bootloader fails no matter what method I use
00:33:04pamauryJdGordon_: does gcc supports nested functions in C ?
00:33:07pamaury*doesn't
00:33:24evilnick_BCptJesus: I
00:33:43CptJesusI guess I'll come back to this tomorrow or something...I'm too frustrated to continue working on this right now
00:34:03gevaertsCptJesus: the rbutil developers aren't here right now. I'd suggest either trying with a pre-7 windows or coming back tomorrow, a few hours earlier (the devs you need are in Europe, and it's past midnight here)
00:34:22evilnick_B'm baffled, to be honest.
00:34:36evilnick_BCptJesus: Could you use the WinXP VM inside Win7?
00:34:36CptJesusThanks for the help though =)
00:34:53CptJesusI'll give it a whirl and be back later, I have to take my sister to Gymnastics
00:35:01CptJesusI appreciate the effort though
00:35:26*TheSeven seriously needs to work on nandfsck
00:35:39CptJesusTake it easy
00:35:46 Quit CptJesus (Quit: Page closed)
00:38:36 Quit evilnick_B (Quit: Page closed)
00:39:07 Quit TheUser (Ping timeout: 258 seconds)
00:41:28JdGordon_pamaury: no idea, but that sounds more scary than goto's :)
00:41:33 Quit jgarvey (Quit: Leaving)
00:42:14pamauryTo compile, yes. To code, no
00:42:18pamaury:)
00:44:08kugelnested functions? are you crazy?
00:44:24pamaurywhy ?
00:44:49pamauryThere are plenty of languages with nested functions
00:45:11JdGordon_EUREKA! we have a mechanism where pebbles can load other pebbles, then you can functions :)
00:45:48pamauryomg, that's horrible ;)
00:46:45pamauryHow do you pass parameters ? With a ugly void* ptr ?
00:48:36gevaertspamaury: what do those nested functions look like? I doubt if they are very different from normal functions in the binary
00:49:03stripwaxsaratoga - on coldfire (h120), mdctexpt seems only fractionally faster than trunk (which is probably to be expected as mdctexp has almost no coldfire asm). so I think good to merge!
00:50:03pamauryIt highly depends on how there are compiled but yes generally there are compiled to normal functions (though the prolog and epilog are different and the access to variables of others stack frames are much more complicate). It's even more challenging if you want to do it without global variables !
00:50:07stripwaxhm, actually mdctexp slightly slower than trunk for high bitrate vorbis.
00:50:47BiontJdGordon_: Now I have 21 viewports for the playlist viewer :) http://img24.imageshack.us/img24/5094/12499034.jpg
00:51:03JdGordon_ah crap!
00:51:53pamaurygevaerts: but the concept of funtions is not well defined at binary level so you can have one "binary function" that indeeds contains the code of several C functions but that's weird :)
00:52:11gevaertspamaury: *you* make it work then! :)
00:52:37*pamaury hides from gevaerts ;)
00:52:43TheSevenhttp://pastie.org/828126
00:52:51TheSevenhmm, clean nano2g svn build
00:54:17JdGordon_Biont: why arnt you using multifont also! :<
00:54:33JdGordon_kugel: look at that screenshot... you dont need the alignment tags!
00:55:19***Saving seen data "./dancer.seen"
00:56:13 Quit pamaury (Quit: abort();)
00:57:26BiontJdGordon_: I'm having trouble getting my han ds on a recent simulator. The one at the extras page doesn't have multifont yet and I failed miserably at compiling my own one several times. Multifont is pretty much the only reason this is not at the theme page yet, because I want to further improve the playlist viewer and song info
00:57:36JdGordon_kugel: by the way, if you want to do RaaaaAAAaa you should do it for the mini2440 instead of n900 :) seen as you have one already
00:57:59JdGordon_Biont: windows sim I guess?
00:58:11BiontJdGordon_ Yes :/
00:58:24kugelwho talked about n900 btw?
00:58:32JdGordon_didn't you?
00:58:44kugelno
00:59:03kugelthe sim already runs on it, so it's boring :)
01:00
01:00:03JdGordon_silly duffer... remove the SIMULATOR define and then say that!
01:05:04Biontbtw: I do think alignment tags would be nice in the playlist viewer. this is one of the things I forgot to mention in the forums. It's not a big problem, but you can't align the playtime properly at the moment. when a song is longer than 9:59 minutes, it will "stand out"
01:05:38BiontI think you all have faced worse things before, though
01:07:39gevaertsOK, ARM+colour LCD people can now also look at the clock. FS #11027 has my current code
01:08:47gevaertsmonochrome and greyscale people need to change it a bit I guess
01:11:18gevaertsnon-arm might work. I have no idea
01:18:22TheSevendoes anyone have an idea where that weird build error could be coming from?
01:18:24JdGordon_Biont: I'm bored so I'll setup stuff so I can build a windows sim for you if you want
01:19:38BiontJdGordon_: Really? That would be great, thank you
01:21:51BiontJdGordon_: Maybe I can get it at the theme site before going to bed then :)
01:23:40kugelgevaerts: considered -fPIC?
01:24:57gevaertskugel: I use -fPIC, but on it's own that's not enough. I *think* that the fixup I do now is correct for all architectures, but I don't really know why it doesn't work with multiple functions
01:25:33kugelit shouldn't be a problem if they're static is it?
01:25:43gevaertsunless I did something wrong while testing of course
01:25:52gevaertsmy test was with a static function
01:26:00gevaertsthe thing locked up :)
01:26:05Biontdoes multifont also work for the .sbs?
01:26:06kugelstrange
01:26:16kugelBiont: of course :)
01:26:25JdGordon_Biont: jdgordon.info/rockbox/d2w32sim.zip">http://usa.jdgordon.info/rockbox/d2w32sim.zip
01:26:27Biontgreat
01:27:02JdGordon_rasher: apparently I can build w32 sims now, do you want me to do a full set?
01:27:27kugelgevaerts: the code in your current test pebble looks fun :)
01:27:34TheSevengevaerts: can you upload a binary of a failing multi-function pebble somewhere?
01:27:39TheSeveni'd like to look at the disassembly
01:27:49 Quit checker (Quit: CGI:IRC)
01:28:15*gevaerts tests again first to make sure
01:28:16TheSeveni don't see any reason why gcc should ever generate non-position-independent code on arm, as long as everything is in one file
01:28:33kugelgevaerts: what target do you test on?
01:28:37gevaertskugel: e200
01:29:10kugelcpucache_invalidate() needs to be called on COP-less archs as well I think
01:29:41gevaertsin that case, the standard plugin loader is also wrong
01:29:42kugelah, it's there, at the end
01:29:47BiontJdGordon_: thank you very much. am I supposed to overwrite the .rockbox folder in my simulator with this one? (sorry for being stupid, but it says "codec failure" if I do)
01:29:54JdGordon_yes
01:30:01JdGordon_extract the whole folder over the top
01:31:56BiontJdGordon_: Well, then there's the codec failure
01:32:02gevaertsTheSeven: gevaerts/pebble/">http://www.evonet.be/~gevaerts/pebble/ has everything
01:33:04gevaertsTheSeven: the loader fixes the first four bytes to adjust for the load address. See apps/pebble.c in the patch on FS #11027 for details
01:33:37 Join ecio [0] (~Moo@244-108.202-68.tampabay.res.rr.com)
01:33:57stripwaxJdGordon - won't Biont need the updated sim executable too?
01:34:08JdGordon_its not in the zip?
01:34:27Biontnope, just a .rockbox folder
01:34:27stripwaxnope. rockboxui.exe
01:34:39stripwaxhence the codec failure...
01:34:43JdGordon_lame
01:34:45JdGordon_ok 1 min
01:34:52Biont:)
01:34:55stripwaxnot really. sim builds don't use zip.
01:35:12JdGordon_Biont: jdgordon.info/rockbox/rockboxui.exe">http://usa.jdgordon.info/rockbox/rockboxui.exe
01:35:16TheSevengevaerts: where's the entrypoint of that thing? at offset 4?
01:35:19JdGordon_someone needs to fix make zip then :)
01:35:25gevaertsTheSeven: yes
01:35:26stripwaxeh.
01:35:32gevaertsat least, I hope so :)
01:35:34stripwaxmaybe.
01:36:18*stripwax wonders - what's the chance of mdctexp merge with trunk actually building first time?
01:36:31kugelJdGordon_: make zip never worked for sims, but Raapp would solve it
01:36:34 Join CaptainKewl [0] (jds@207-237-117-89.c3-0.80w-ubr2.nyr-80w.ny.cable.rcn.com)
01:36:51kugelstripwax: did you merge already?
01:36:54stripwaxman, cygwin builds are sooooo slow
01:36:57gevaertsstripwax: that depends on how many test builds you do :)
01:37:07stripwaxkugel - locally, yes. not committed it yet, as I want to see what builds first
01:37:16stripwaxbut I might not get to it if I fall asleep first ....
01:37:31JdGordon_stripwax: if you are impatient with cygwin you can ssh into my box and go nuts
01:37:49TheSevengevaerts: heh, it's a very trivial issue
01:37:53stripwaxJdGordon - nah, but thanks anyway
01:38:01TheSevenhelper is the function at 0x4, not pebble_do :-D
01:38:11BiontJdGordon_: Thank you :)
01:38:17gevaertshm, that would explain it :)
01:38:50JdGordon_so multi funciton pebbles are doable?
01:39:04*gevaerts tries
01:39:04TheSeveni don't know if there is a better way, but i would just use a trivial ASM stub consisting of a branch to pebble_do and link it with that
01:39:07kugelthere should probably be a pepple header which just expands to a function calling the pebble's main? or look how plugins do it :)
01:39:22 Quit fyrestorm (Read error: Connection reset by peer)
01:39:33gevaertsI'm trying a .lds fix first
01:39:57*stripwax whoops - seems to work!
01:40:00stripwaxLook out for the red...
01:40:25TheSeventhe rest seems to be safe
01:40:46gevaertsif this really works I think I'll have a go at doing this sort of thing for normal plugins though
01:41:25JdGordon_I wonder if we could use the pebbles for things like metadata parsers so only the required ones are in ram?
01:41:35kugelso, I imagine pebbles being shipped with themes which makes up for a boring licensing issue discussion
01:41:37JdGordon_and other things which arnt used often
01:41:56JdGordon_kugel: no, why would they? they'd be part of the source tree
01:42:03JdGordon_still gpl requirements
01:42:10gevaertskugel: they still depend on the exact plugin API version
01:42:22kugelJdGordon_: why should they be part of the source tree?
01:42:48JdGordon_same reason as plugins
01:43:07kugeland what's the reason for plugins?
01:43:09rasherkugel: probably same situation as with rocks. In theory they could be separate, but in practice all the usual ones will get included
01:43:17stripwaxoh wait - some ugly conflicts cos some changes to atrac3 are in trunk and some other changes are in branch. ugh. won't commit just yet ...
01:43:18rashers/usual/useful/
01:43:56 Part toffe82
01:44:04kugelthere are a billion themes, each could potentially have a few pebbles, I don't think we're going to ship all of them
01:44:37kugeland any themer could say "I put it into my theme.zip and don't care about it"
01:45:01JdGordon_we should reject any themes that include pebbles
01:45:04gevaertsyes, but they would only work for one target, and not for very long
01:45:12kugelJdGordon_: scrolling lines reveal the menu backdrop here
01:45:14JdGordon_especially when they are going to be target dependdant
01:47:03kugelI don't like the idea of rejecting themes that have pebbles
01:47:12kugelwhat's the point of the feature then?
01:47:42kugeland they won't necessarily be target dependant
01:48:13LloreanUnless you make them in lua.
01:48:22LloreanOtherwise they're at least target-restricted.
01:49:03kugeldepending on how they're done they might be arch-dependant only
01:49:16gevaertsTheSeven: I don't see a quick fix. I think I'll move back to a header like the normal plugins have
01:49:25 Join MagusG [0] (magusg@c-98-192-69-229.hsd1.ga.comcast.net)
01:49:29Lloreankugel: That's target dependency though
01:49:55CIA-8New commit by 03stripwax (r24712): Merge from branches/mdctexp - faster ifft+imdct in codec lib
01:49:57*stripwax >- fingers in ears <-
01:50:06TheSevengevaerts: a quick hack would probably be swapping those functions and using a forward declaration
01:50:47LloreanIs there any way to make pebbles lua rather than compiled?
01:51:16kugelsure if you integrate lua into the core
01:51:51kugelI think the overhead would be quite huge though, especially with multiple pebbles
01:51:59gevaertsTheSeven: maybe, but right now I'm more tempted to try the same trivial .got fixup on normal plugins. If that works, we gain a lot more
01:52:08gevaertsit's also late :)
01:52:09LloreanWell I mean the only options are target-independent pebbles, reject themes with pebbles that aren't included in SVN, or themes are broken down by target rather than screen.
01:52:13stripwaxJust noticed OndioSP build is red &minus;&minus; always &minus;&minus; anyone care?
01:52:27JdGordon_noone thats still awake :)
01:52:28pixelmastripwax: could have been a bit more info I think. Has it been tested on coldfire a bit (as I always saw people talking about results on ARM when talking about it here)?
01:52:39LloreanI don't see why you can't include pebbles in SVN and reject themes with nonstandard ones, much like tags
01:52:48pixelmastripwax: broken rombox
01:52:48JdGordon_Llorean: LUA pebbles would also be very useful
01:52:59JdGordon_ls
01:53:08LloreanJdGordon_: They'd certainly be more designer-friendly since you wouldn't need a dev environment to create / test them
01:53:12stripwaxpixelma - yes. plus I also tested on coldfire a short while back. seems fractionally slowee than trunk, but only at high bitrates, from what I can tell (limited testing)
01:53:39gevaertsLlorean: in theory the lua interpreter itself could be a pebble :)
01:53:40LloreanI really don't expect a proliferation of compiled pebbles on the same scale as the number of created themes - most theme authors won't want to set up the dev environment, etc.
01:53:45Lloreangevaerts: That would be the ideal.
01:53:51Llorean:-P
01:53:52JdGordon_Llorean: yes, but that shouldnt mean that they should be done instead of c pebbles (I dont think you were suggesting that though)
01:54:33stripwaxpixelma - it probably needs a coldfire expert/enthusiast to see if anything needs changing for coldfire targets, but nobody so far has volunteered to make any changes in branch
01:54:40pixelmastripwax: doesn't sound very assuring
01:54:47 Quit robin0800 (Remote host closed the connection)
01:54:52kugelJdGordon_: not sure if you overlooked it, but scrolling lines is broken
01:54:54stripwaxpixelma - it is faster for wma 192kbps and below, and vorbis 192kbps and below
01:54:59 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
01:55:08 Quit krabador (Read error: Connection reset by peer)
01:55:20JdGordon_kugel: I thuoght I fixed all that, should be a quick fix
01:56:59CIA-8New commit by 03stripwax (r24713): Fix yellow (ff_fft_permute_c an unused function in codeclib)
01:58:56CIA-8New commit by 03stripwax (r24714): Fix yellow (bad function declaration)
01:59:43*gevaerts has a running two-function pebble :)
02:00
02:00:02 Quit efyx_ (Remote host closed the connection)
02:00:26kugelso, someone needs to mail that windows 3.1 guy so that he can reimplement it as a pebble
02:00:47CIA-8New commit by 03stripwax (r24715): FIx typo in coldfire asm which caused yellow
02:00:52gevaertsI swapped the functions, after which I got two .got entries, so I had to fix the .got fixup
02:01:21kugel2 gots?
02:01:34stripwaxOk, I think I've fixed those other guys' code and hopefully no warnings now :)
02:01:45gevaertskugel: two entries, one .got table
02:01:52stripwaxpixelma - please add your findings to the FasterMDCT page
02:01:56saratogawoot
02:02:01JdGordon_gevaerts: get rid of pebble globals, having them is going to cause issues when they are loaded more than once
02:02:06JdGordon_and FUCKING AYE :)
02:02:17gevaertsJdGordon_: globals and functions are the same :)
02:02:30gevaertswell, not entirely
02:02:50JdGordon_so we just make sure pebble authors know not to use them?
02:02:59JdGordon_or static variables of any type?
02:03:21*gevaerts thinks that people are moving too fast. We're still experimenting to see what's possible!
02:03:33kugelah yes, I forgot that!
02:03:36JdGordon_nonesense!
02:03:49kugelI was sure it's ready to commit already :)
02:04:15 Nick Strife1989 is now known as Strife89 (~michael@adsl-220-102-6.mcn.bellsouth.net)
02:09:30Unhelpfulwe are *going* to have some relocations to deal with most likely, might as well handle simple relocs in some way
02:09:39Unhelpfulmultiple (and background) plugins, go!
02:09:47CIA-8New commit by 03stripwax (r24716): split_radix_permutation also unused (since now a hardcoded table) - but left intact so we can/could regenerate the table
02:10:22stripwax^^ oddly enough, only [some of] the Sim builds complained about the last one. weird.
02:10:51JdGordon_Unhelpful: we are already better than iphone OS :)
02:12:46 Join killan [0] (~nnscript@c-38fd70d5.06-397-67626721.cust.bredbandsbolaget.se)
02:13:02 Quit TheSphinX^ (Quit: XChat@Linux)
02:14:16stripwaxsaratoga - green! :)
02:14:39UnhelpfulJdGordon_: not *yet*...
02:18:15 Quit `Assassin (Read error: Connection reset by peer)
02:18:20*stripwax sleeps .....
02:18:27 Quit stripwax (Quit: http://miranda-im.org)
02:19:46 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.87)
02:20:33Stephen__is there a patch to increase the charge from the h340 ? instead of 100mah get 500mah ?
02:20:33 Join freddyb [0] (~fred@pool-68-238-8-141.chi.dsl-w.verizon.net)
02:21:30sudomanStephen_: i think you need a new battery for that
02:21:53 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
02:22:44Stephen__isn;t it the same as the ipods it doesn't pull the full charge from the battery irregardless of the batterys newness
02:23:00 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
02:24:33 Quit JdGordon_ (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
02:25:19 Quit GeekShad__ (Ping timeout: 260 seconds)
02:28:57kugelJdGordon: can you have a look if the playback.c change of fs#11004 is fine?
02:31:31 Quit kugel (Remote host closed the connection)
02:32:24 Quit MethoS- (Remote host closed the connection)
02:35:39 Quit Kitr88 (Ping timeout: 252 seconds)
02:36:22 Join Kitr88 [0] (Kitr88@BSN-182-120-56.dial-up.dsl.siol.net)
02:39:59 Quit freddyb (Quit: Konversation terminated!)
02:40:34 Quit Biont (Quit: CGI:IRC)
02:45:05 Quit Stephen__ (Quit: Leaving)
02:46:51 Quit minus (Ping timeout: 272 seconds)
02:46:57sudomanStephen_: i don't know these things. i have an upgraded battery that's a little old. honestly, i did expect to get more out of it, even when it was new.
02:48:11 Quit S_a_i_n_t (Quit: There are 10 types of people, those who understand binary, and those who don't.)
02:49:14 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
02:55:23***Saving seen data "./dancer.seen"
03:00
03:10:32 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.87)
03:11:28S_a_i_n_tdo you need to have all the toolchains installed to make a simulator build in CygWin?
03:11:28 Join CptJesus [0] (~60ea966d@gateway/web/freenode/x-wjonuzasbehzpotg)
03:12:00CptJesusgevaerts: Are you available?
03:13:05saratogaS_a_i_n_t: no just a native compiler
03:13:05JdGordonS_a_i_n_t: you dont need any to build the sim
03:13:15 Join crazed [0] (~cr4z3d@unaffiliated/cr4z3d)
03:13:31S_a_i_n_tI had to reinstall Cygwin, and now I get some message regarding SDL and it won't make sim builds anymore...
03:13:37crazedwhat's a recommended way of tagging mp3s i pulled off an ipod?
03:13:47crazedi have the whole iPod_Control/Music directory
03:13:48 Join BlakeJohnson861 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
03:14:05LloreanS_a_i_n_t: Do you have SDL installed?
03:14:48saratogahopefully hes checked that
03:15:03S_a_i_n_tI *should* do, I have all the 'required packages' from the wiki, and didn't touch the base packages.
03:15:15 Quit BlakeJohnson86 (Ping timeout: 260 seconds)
03:16:06S_a_i_n_tthe only thing that should be different between this CygWin istall and the last, is I didn;t bother building toolchains I won;t use.
03:18:10 Quit Strife89 (Quit: Bed.)
03:18:14sudomancrazed: most songs should be tagged. do you need to change the file names of the songs?
03:18:29crazedwell unless picard is lying i don't think any of these songs are tagged
03:19:16S_a_i_n_tthey should all be tagged, just in a weird truncated folder system that the ipod puts them in
03:19:33S_a_i_n_twith names like DDFGH etc.
03:19:38crazedyeah
03:19:42sudomani think there might be some trick you can use in itunes to get it to write the stored information in the database to the id3 tags
03:19:44crazedthat's what it looks like but there's no tags
03:20:09sudomanwhat os do you have?
03:20:14S_a_i_n_tisn't there a util on the RB wiki for this...?
03:21:36 Quit n17ikh (Read error: Connection reset by peer)
03:23:03saratogaif your itunes didn't write tags then your only hope is to find an app that can read the ipod database and rename the files based on what it says
03:23:04Tomisno, itunes definitely edits the mp3/m4a tags
03:23:04Tomisfields for which there is no tag standard, are stored only in the database, such a ratings
03:23:08saratogai have no idea if such a thing exists
03:25:31TheSevenwhich ones were the ubuntu packages needed to build the manual?
03:26:46TheSevenis that documented somewhere?
03:27:20TheSevennvm, found it
03:29:38 Join n17ikh [0] (~n17ikh@host-69-59-126-212.nctv.com)
03:29:51crazedyeah there's definitely no id3 tags on here
03:29:53crazedfuug
03:31:49sudomanTomis: i downloaded some talks online and put them into itunes. they had correct file names but itunes did not update the id3 tags. i had to use TriTag on os x to give track name tags to those songs.
03:32:19Tomisso the files didn't have id3 from the get-go
03:32:45Tomisbut it pulled in the file's name as the display name, without editing the id3
03:33:44 Quit CptJesus (Quit: Page closed)
03:34:11sudomanyeah, basically. i'm not sure if this would work, but someone suggested getting the info for the song in itunes and then clicking OK. but that would be tedious for many songs. it might be feasible for doing album names.
03:34:56Tomisi suspect it would only update it if you actually made a change
03:37:31S_a_i_n_ti think it depends on the "allow itunes to manage my music" setting in itunes
03:37:53S_a_i_n_tif set to no, it doesn;t touch your tags at all.
03:38:05S_a_i_n_t.......damn semecolon!......
03:38:23S_a_i_n_t*semi
03:40:36 Quit Adnyxo (Ping timeout: 264 seconds)
03:47:43 Part froggyman
03:49:19 Quit sudoman (Quit: sudoman)
03:50:41 Quit S_a_i_n_t (Quit: There are 10 types of people, those who understand binary, and those who don't.)
03:55:12 Quit DerPapst (Quit: Leaving.)
04:00
04:01:52 Quit saratoga (Quit: CGI:IRC (EOF))
04:13:55 Quit TheSeven (Disconnected by services)
04:14:07 Join The_Seven [0] (~theseven@rockbox/developer/TheSeven)
04:14:17 Nick The_Seven is now known as TheSeven (~theseven@rockbox/developer/TheSeven)
04:21:34 Join medial [0] (~neo@85.26.92.11)
04:22:37 Join saratoga [0] (~9803c57f@rockbox/developer/saratoga)
04:23:01 Join logiclost [0] (~lostlogic@erudite.lostlogicx.com)
04:23:04 Join blithe_ [0] (~blithe@72.14.176.144)
04:23:47 Quit blithe (Write error: Broken pipe)
04:24:43 Join Tuplis_ [0] (~jani@adsl-77-109-221-158.kymp.net)
04:28:30 Join Barahir_ [0] (~jonathan@gssn-5f75462d.pool.mediaWays.net)
04:28:35 Quit Tuplis (Ping timeout: 265 seconds)
04:28:35 Quit lostlogic (Ping timeout: 265 seconds)
04:31:09 Quit Sajber^ (Read error: Connection reset by peer)
04:31:45 Quit Barahir (Ping timeout: 248 seconds)
04:38:23saratogathe ffmpeg AAC_HE decoder is really interesting, it uses an MDCT to implement the synthesis QMF
04:43:21 Quit GeekShadow (Quit: The cake is a lie !)
04:55:26***Saving seen data "./dancer.seen"
05:00
05:28:57 Join phanboy4 [0] (~benji@c-24-98-43-198.hsd1.ga.comcast.net)
05:32:03 Join Hillshum [0] (~hillshum@75-165-233-186.slkc.qwest.net)
05:33:21 Quit Horscht (Quit: Verlassend)
05:34:20 Quit ps-auxw (Ping timeout: 260 seconds)
05:36:19 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
05:43:28 Join nimak [0] (~nima@adsl-75-45-228-161.dsl.sfldmi.sbcglobal.net)
05:45:35 Quit nima (Ping timeout: 245 seconds)
05:53:23 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
05:58:09 Quit ecio (Ping timeout: 248 seconds)
05:58:22 Join Anaxite [0] (~anaxite@pdpc/supporter/professional/anaxite)
05:59:23 Join phanboy_iv [0] (~benji@c-24-98-43-198.hsd1.ga.comcast.net)
06:00
06:01:38AnaxiteI figure it's a bit late but I've got a question for anyone working on Rockbox; I've got a Sansa Clip v1 I'm willing to donate to the project. Is there anyone interested in having it?
06:04:01 Quit Rob2223 (Read error: Connection reset by peer)
06:04:01 Quit phanboy4 (Read error: Connection reset by peer)
06:04:01 Join Rob2222 [0] (~Miranda@p4FDCAC79.dip.t-dialin.net)
06:04:02 Quit lyngaas (Read error: Connection reset by peer)
06:04:02 Quit danopia (Ping timeout: 256 seconds)
06:05:11 Join danopia [0] (~danopia@fullcirclemagazine/developer/danopia)
06:06:34saratogaAnaxite: post in the New Ports thread and see if anyone takes you up on the offer
06:06:50AnaxiteUnder the general Sansa thread?
06:08:00saratogayeah thats your best bet I think
06:08:07saratogano one is going to be awake now
06:09:00AnaxiteOk, that's fine. Alternately, I'm on IRC during the daytime as well if it's better not to crowd the forums with requests/questions.
06:09:44AnaxiteThanks.
06:09:52saratogawould't hurt to ask online later
06:10:06saratogaoff hand i'm not sure who would be interested, theres probably some people with just a fuze or a v2 player
06:10:16HillshumAnaxite: Many of the devs are European, and it's a bit late there
06:11:22 Join lyngaas [0] (~staale@19.81-167-149.customer.lyse.net)
06:11:33AnaxiteI half-expected it, but thanks for the confirmation. :)
06:12:27HillshumAnaxite: How much capacity does this Clip have?
06:12:37AnaxiteIt's a 2GB Clip.
06:14:05saratogaFWIW i've been offering a 2GB ClipV2 to anyone who wants it for a while
06:14:19saratogabut the V1 is more interesting since rockbox already mostly runs on it
06:14:36Hillshumsaratoga: Who does 'anyone' extend to?
06:14:59saratogapeople i think are worth the effort to mail it to
06:15:21saratogadev badge on the forums helps with that
06:15:30AnaxiteYeah, I decided to give rockbox on the clip a shot... apart from the battery life and the USB driver, it's working quite nicely. I was quite impressed!
06:16:18HillshumAnaxite: How does the battery compare on yours? I've never really run the Original Firmware on mine
06:17:58AnaxiteIt has a bit less than half the battery life of the OF when idling.
06:18:28HillshumHmm.
06:18:28AnaxiteI know that without rockbox, it has reached around the 16 hour mark, possibly a few more (didn't count much past that)
06:18:38Hillshumidling?
06:18:50AnaxiteSorry; I meant when not playing music.
06:24:23 Quit panni_ (Read error: Connection reset by peer)
06:42:31 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
06:55:30***Saving seen data "./dancer.seen"
07:00
07:00:06 Quit liar (Quit: Verlassend)
07:02:31 Quit saratoga (Quit: Page closed)
07:04:03 Join flyback- [0] (~teac@c-98-219-129-239.hsd1.pa.comcast.net)
07:07:09 Quit flyback (Ping timeout: 272 seconds)
07:09:02 Nick Tuplis_ is now known as Tuplis (~jani@adsl-77-109-221-158.kymp.net)
07:13:15 Join TheUser [0] (~theuser12@80.93.115.26)
07:23:09 Join stoffel [0] (~quassel@p57B4DFAA.dip.t-dialin.net)
07:30:44 Quit Hillshum (Ping timeout: 265 seconds)
07:35:20 Quit Bagder (Ping timeout: 252 seconds)
07:38:36 Join kramer3d_ [0] (~kramer@unaffiliated/kramer3d)
07:39:25 Quit kramer3d (Disconnected by services)
07:39:30 Nick kramer3d_ is now known as kramer3d (~kramer@unaffiliated/kramer3d)
07:42:12 Quit TheUser (Ping timeout: 260 seconds)
07:42:28 Join TheUser [0] (~theuser12@80.93.115.26)
07:49:31 Quit TheUser (Ping timeout: 246 seconds)
07:49:56 Join TheUser [0] (~theuser12@80.93.115.26)
07:58:30 Quit TheUser (Ping timeout: 245 seconds)
08:00
08:17:09CIA-8New commit by 03jdgordon (r24717): change the %St tag so it only shows the filename of file settings, not the full path (i.e cabbiev2 instead of /.rockbox/wps/cabbiev2.wps)
08:23:37 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
08:26:05 Quit stoffel (Remote host closed the connection)
08:31:46 Join ender` [0] (krneki@foo.eternallybored.org)
08:39:03 Quit togetic (Quit: WeeChat 0.3.0)
08:39:28 Join togetic [0] (~togetic@unaffiliated/ibuffy)
08:40:27 Quit medial (Quit: My damn controlling terminal disappeared!)
08:46:17 Quit CaptainKewl (Ping timeout: 240 seconds)
08:47:20 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
08:49:52 Quit TheSphinX^ (Client Quit)
08:50:23 Quit tchan (Ping timeout: 272 seconds)
08:51:52 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
08:52:37 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
08:52:42 Quit TheSphinX^ (Client Quit)
08:54:39 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
08:55:32***Saving seen data "./dancer.seen"
08:55:59 Quit shai (Read error: Connection reset by peer)
08:56:02 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
08:56:19 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
08:57:09 Join EricDraven [0] (~de7e5f3c@giant.haxx.se)
08:57:20 Quit shai (Read error: Connection reset by peer)
08:58:02 Quit EricDraven (Client Quit)
08:58:05 Join petur [0] (~petur@rockbox/developer/petur)
08:59:58 Quit TheSphinX^ (Quit: XChat@Linux)
09:00
09:00:04 Join flydutch [0] (~flydutch@host154-132-dynamic.15-87-r.retail.telecomitalia.it)
09:00:14 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
09:03:37 Join stoffel [0] (~quassel@p57B4DFAA.dip.t-dialin.net)
09:03:42 Quit BHSPitMonkey (Quit: Ex-Chat)
09:12:28 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
09:12:55 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
09:13:25AlexPJdGordon: I agree with kugel - the playlist viewer is not usable for prev/next preset nae without align tags
09:13:35AlexPs/nae/name/
09:14:25 Quit phanboy_iv (Ping timeout: 248 seconds)
09:15:12AlexPI would like left side of screen &minus;&minus;>Prev preset Next preset<&minus;&minus; right side of screen. Each of those is in its own viewport, but the next preset needs to right aligned within its viewport
09:16:46amiconnWhat would prev/next preset name be good for?
09:17:51 Join Bagder [0] (~daniel@rockbox/developer/bagder)
09:18:06amiconnIt's certainly different from next track info in the wps. The next track *will* start playing after some time without user intervention, whereas the fm station never changes automatically
09:19:48 Quit Anaxite (Quit: KVIrc Insomnia 4.0.0, revision: 3725, sources date: 20091205, built on: 2009-12-21 09:01:14 UTC http://www.kvirc.net/)
09:21:14AlexPamiconn: I like to see what I'm going to change onto next
09:21:18 Quit TheSphinX^ (Quit: XChat@Linux)
09:21:22 Quit shai (Read error: Connection reset by peer)
09:21:45 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
09:22:05AlexPAnd regardless of whether it is something you would use, if it is intended to be possible then it should be possible to align it properly
09:22:57amiconnWe don't have previous track info in the wps either. Wouldn't you need that when skipping back? ;)
09:23:23*amiconn uses the preset list when he needs to select a specific preset
09:23:24AlexPIdeally, but that is much harder to do I believe
09:23:36AlexPIt is also less likely
09:23:47 Join DerPapst [0] (~DerPapst@p4FE8EEF5.dip.t-dialin.net)
09:23:47 Join T44 [0] (~Topy44@f049051198.adsl.alicedsl.de)
09:25:10 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
09:26:13 Join einhirn [0] (~Miranda@bsod.rz.tu-clausthal.de)
09:26:27 Quit TheSphinX^ (Client Quit)
09:27:31 Quit Topy44 (Ping timeout: 256 seconds)
09:27:31 Quit TopyMobile__ (Ping timeout: 246 seconds)
09:32:42 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
09:37:32 Quit karashata (Quit: The fluffy dragon has left completely!)
09:42:49 Join TopyMobile__ [0] (~topy@f049051198.adsl.alicedsl.de)
09:43:36 Quit kramer3d (Ping timeout: 264 seconds)
09:44:03 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
09:55:22 Quit TheSphinX^ (Quit: XChat)
09:55:42 Join TheSphinX^ [0] (~cold@p54A5D6BD.dip.t-dialin.net)
09:56:58 Quit TheSphinX^ (Client Quit)
10:00
10:00:46 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
10:16:43 Quit Zagor (Quit: SSD install)
10:24:53 Quit kramer3d (Quit: Leaving)
10:29:09 Join shai_ [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
10:31:41 Join Zagor [0] (~bjst@46.35.227.87.static.tab.siw.siwnet.net)
10:31:41 Quit Zagor (Changing host)
10:31:41 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
10:31:58 Quit shai (Ping timeout: 256 seconds)
10:46:20 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.223)
10:46:47S_a_i_n_tTheSeven: Ping
10:49:10S_a_i_n_tAnyone to field a Nano2g/DirCache question..?
10:55:33***Saving seen data "./dancer.seen"
11:00
11:17:06S_a_i_n_tHmmm....ok, new way of looking at this...Does anyone know why "DEFAULT_STACK_SIZE" in 'firmware/common/dircache.c' was changed so drastically? IIRC it *was* at 0x900, now its down to 0x200?!? The reason this surprises me is that to stop the Nano2g from *PANIC*Stkov DirCache-ing *see #FS10679. I need to set it to 0x1900 (at least, but 0x2000 seems to run better) to stop this...so dropping it even further seemed a bit odd to me.
11:18:13S_a_i_n_tI don't want to risk trying a build at 0x200 because I don't feel like trashing all my data on the 2g's for the millionth time, but the sudden drop in size makes me wonder what I may break setting it to 0x1900~2000 again...
11:18:44 Join shai__ [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
11:21:50 Quit shai_ (Ping timeout: 256 seconds)
11:24:55 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:25:49 Join Sajber^ [0] (~Sajber@h-143-173.A213.priv.bahnhof.se)
11:28:30 Join zumbi_ [0] (~zumbi@77.230.237.25)
11:28:32 Join yawny [0] (user36@pr0.us)
11:28:57 Join nima [0] (~nima@adsl-75-45-228-161.dsl.sfldmi.sbcglobal.net)
11:28:58 Join Frampis_ [0] (famas@noppakerho.com)
11:29:11 Join koniiiik_ [0] (johnny64@phoenix.wheel.sk)
11:29:16 Join ranmacha1 [0] (ranma@yumi.tdiedrich.de)
11:31:20 Join maraz_ [0] (maraz@kapsi.fi)
11:33:52 Quit nimak (*.net *.split)
11:33:52 Quit BlakeJohnson861 (*.net *.split)
11:33:52 Quit parafin (*.net *.split)
11:33:52 Quit martian67 (*.net *.split)
11:33:52 Quit moos (*.net *.split)
11:33:52 Quit koniiiik (*.net *.split)
11:33:52 Quit cjcopi (*.net *.split)
11:33:52 Quit Zarggg (*.net *.split)
11:33:52 Quit ranmachan (*.net *.split)
11:33:52 Quit zumbi (*.net *.split)
11:33:53 Quit Frampis (*.net *.split)
11:33:53 Quit maraz (*.net *.split)
11:33:53 Quit elcan (*.net *.split)
11:35:37 Quit stoffel (Ping timeout: 246 seconds)
11:36:58 Join efyx_ [0] (~efyx@lap34-1-82-225-185-146.fbx.proxad.net)
11:38:01 Join parafin [0] (parafin@paraf.in)
11:40:48 Quit parafin (Client Quit)
11:40:51 Join parafin [0] (parafin@paraf.in)
11:41:34 Quit parafin (Client Quit)
11:41:37 Join parafin [0] (parafin@paraf.in)
11:49:28 Join ecio [0] (~ecio@244-108.202-68.tampabay.res.rr.com)
11:59:35*S_a_i_n_t *finally* (after missing it several times...) spots r24708 (fix for #FS10679), and decides to run with it...thanks pamaury ;) , I assure you I'll let you know if *PANIC* Stkov DirCache still happens on the Nano2G
12:00
12:00:37 Join Zarggg [0] (~zarggg@65-78-69-194.c3-0.eas-ubr6.atw-eas.pa.cable.rcn.com)
12:04:01 Quit m3dlg (Ping timeout: 248 seconds)
12:11:40 Join watto [0] (~watto@193.203.81.165)
12:27:01 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
12:39:33 Join stoffel [0] (~quassel@p57B4DFAA.dip.t-dialin.net)
12:42:24*TheSeven wonders if he should file a bug agains rockboxdev.sh
12:42:30 Join OnlyByGrace [0] (~5fee1b37@gateway/web/freenode/x-afecxfflngxoanli)
12:43:02 Join MethoS- [0] (~clemens@134.102.106.250)
12:44:06S_a_i_n_twhat's the problem with it?
12:44:18*S_a_i_n_t is happy to see someone else alive in here :D
12:44:37 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
12:44:48TheSevenit fails building an EABI compiler on a fresh ubuntu install
12:45:18Bagdercan't we just blame ubuntu? B)
12:45:28S_a_i_n_t....easier :D
12:46:47S_a_i_n_tTheSeven: can you verify that dircache is fixed now (for Nano2G) as of 24708?
12:46:51 Quit OnlyByGrace (Ping timeout: 252 seconds)
12:46:57TheSevenS_a_i_n_t: I think it is.
12:47:03S_a_i_n_tyou're the only other person I know (with a 2g) to ask...
12:47:10S_a_i_n_tawesome :D
12:49:15TheSevenoh, and it does fail with a "cannot create directory: file exists" message if i run it for the second time
12:49:27TheSevencan't it just kill that?
12:51:12S_a_i_n_tyou'd think it would be smart enough to not bother creating the dir if it already exists...
12:51:44BagderI made it not do it by itself once upon the time just to avoid mistakes
12:51:54Bagderbut I'm not sure it helps anyone
12:52:16TheSevenrm, deleting it's own temp directory won't hurt, huh?
12:52:25TheSevenhm*
12:52:27TheSevenwtf
12:52:54S_a_i_n_t*shouldn't* hurt...
12:52:57BagderI think my theory was that I'd leave the rm -rf'ing to the user to avoid the very very slim chance the dir isn't actually "ours"
12:53:25S_a_i_n_tI managed to get away with deleting all rockboxdev'sh's tmp stuff without a hitch.
12:53:28TheSevenit should do that with rbdev-build then, but not the subdirectories
12:53:31S_a_i_n_tin CygWin though.
12:54:00TheSevenit happily overwrote the stuff in rbdev-build, but then complained about rbdev-build-e
12:54:22S_a_i_n_todd.
12:55:09Bagderno strong opinions either way from me
12:55:37***Saving seen data "./dancer.seen"
12:55:49 Quit Sajber^ (Read error: Connection reset by peer)
12:55:50TheSevenhm, which packages do i need to have installed for an eabi toolchain build to complete?
12:56:00S_a_i_n_tI'm of the opinion that a tmp should be cleared once the ap is done with it...otherwise it's not "tmp" by definition
12:56:20TheSevenS_a_i_n_t: that wouldn't be helpful when it crashes
12:56:26BagderS_a_i_n_t: there again, I took the safe route. Like when you want to debug problems etc
12:57:16Bagderand I'd still call it "temp", just that "temp" is extended to expire manually ;-)
12:57:28S_a_i_n_t:D
12:57:42TheSevenit's complaining about GMP 4.1+ and MPFR 2.3.2+, but I can't find those in the ubuntu repositories
12:57:43AlexPor automatically on reboot
12:59:45TheSevenaha. who would have thought that the package for gmp 4.2 is called "libgmp3".
13:00
13:00:08Bagderand libmpfr-dev perhaps?
13:00:45Bagderwell the number in the package names are usually soname numbers, not actual version number
13:00:57TheSevenyes, libmpfr-dev seems to depend on everything i need
13:07:05 Join minus [0] (~minus@p5DC70FBA.dip0.t-ipconnect.de)
13:14:47 Join perfectdrug__ [0] (~marko@p5B0ED07B.dip.t-dialin.net)
13:16:38 Nick shai__ is now known as shai (~Shai@l192-117-110-233.cable.actcom.net.il)
13:17:08 Nick shai is now known as Guest63500 (~Shai@l192-117-110-233.cable.actcom.net.il)
13:17:47 Quit Guest63500 (Quit: Leaving)
13:18:07 Join shai_ [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
13:18:55 Quit perfectdrug_ (Ping timeout: 272 seconds)
13:20:25 Join kugel [0] (kugel@rockbox/developer/kugel)
13:29:52TheSevenhow do i tell configure to use eabi?
13:34:21kugelchose e when it's asking for the crosscompiler
13:38:11TheSevenit isn't asking...
13:46:13kugelTheSeven: ah sorry, I confused it. pass &minus;&minus;eabi
13:48:30 Part minus ("Verlassend")
13:49:53 Join joeytwiddle [0] (~joey@cpc4-aztw4-0-0-cust207.aztw.cable.virginmedia.com)
13:59:45 Join anewuser [0] (anewuser@unaffiliated/anewuser)
14:00
14:04:03 Join froggyman [0] (~sopgenort@pool-72-69-210-48.chi01.dsl-w.verizon.net)
14:06:21 Join teru [0] (~teru@KD059133108225.ppp.dion.ne.jp)
14:06:56 Part froggyman
14:11:52 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:22:32 Quit kugel (Ping timeout: 265 seconds)
14:25:26 Join FOAD__ [0] (~dok@dinah.blub.net)
14:27:51 Quit FOAD (Read error: Operation timed out)
14:27:51 Nick FOAD__ is now known as FOAD (~dok@dinah.blub.net)
14:29:28 Join kugel [0] (kugel@rockbox/developer/kugel)
14:38:52 Join kugel_ [0] (kugel@141.45.206.62)
14:39:09 Join froggyman [0] (~187b533e@giant.haxx.se)
14:41:02 Join kugel__ [0] (kugel@141.45.206.62)
14:41:32 Quit kugel (Ping timeout: 276 seconds)
14:43:05 Quit kugel_ (Ping timeout: 240 seconds)
14:45:39 Join Farthen [0] (~chatzilla@e179233161.adsl.alicedsl.de)
14:46:46 Join Adnyxo [0] (~aaron@adsl-065-013-002-216.sip.asm.bellsouth.net)
14:48:24 Join kugel [0] (kugel@rockbox/developer/kugel)
14:50:25 Quit kugel__ (Ping timeout: 248 seconds)
14:54:58 Join archivator [0] (~archivato@stu0279.keble.ox.ac.uk)
14:55:41***Saving seen data "./dancer.seen"
14:56:37 Join funman [0] (~fun@rockbox/developer/funman)
14:56:48 Quit kugel (Ping timeout: 264 seconds)
14:58:56 Quit anewuser (Quit: http://xrl.us/WinterChipV =ooo ϢINTER ϾHIP 5iVE is OOON!!)
14:59:55archivatorre: gsoc 2010 - I was thinking of doing TTS in core, integrating flite to be precise. However, I am not entirely sure that concatenative synthesis is the right tool for the job. Something like espeak might indeed be more suitable for in-core tts. Any thoughts on that or rockbox TTS in general? What are the use cases?
15:00
15:02:41funmanblind people i guess
15:03:43archivatorwhere does the current method fail that we need TTS? Just song artists/titles?
15:04:16funmanand folders, basically everything that isn't known when compiling
15:04:39funmani never used voice so there might be other use cases i don't know about
15:05:35gevaertspossibly some plugin use, like the text viewer
15:06:08archivatorgevaerts: I guess you would want a naturally-sounding voice for that right?
15:06:28gevaertsYou'd want the best possible voice in all cases I'd say :)
15:06:48archivator"best" is subjective.
15:07:13AlexPBut whatever that best is, you always want it
15:07:14archivatorFormant synthesis has its uses - I see no reason why something as short as menu labels should be naturally sounding.
15:07:49archivatorHowever, if you have to listen to more than 10 words, it starts being annoying (for me personally at least).
15:08:22gevaertsI think this is one of those cases where the candidate student is expected to make a documented proposal :)
15:08:41archivatorfair enough, I'm just doing my research first.
15:08:42AlexPAll assuming we get accepted :)
15:09:03gevaertsyes, that's not done yet
15:09:51archivatorso, we don't know what's going to be synthesized (open domain), we want it to be naturally sounding (implying some sort of unit selection) and we want it to be fast. Now pick 2 of the 3. :)
15:10:34gevaertsYou forgot one
15:10:39AlexPThe first and third would be most important to me
15:10:45gevaertsWe want it to work for lots of languages
15:10:50AlexPthat too
15:11:06AlexPAlthough the first implementation should probably be English
15:11:17S_a_i_n_tdoes "natural sounding voice" and "tts" go together?
15:11:25AlexPOthers could be added later (as long as it was made possible)
15:11:31*S_a_i_n_t thinks of "MS Sam" and decides not :P
15:11:43archivatorS_a_i_n_t: yes. You'd be surprised how far TTS has gone. It's not all 2000-era MS Sam-like :)
15:12:03*gevaerts stopped caring after Macintalk
15:12:39archivatorI think TTS should first be built with an English voice, keeping the current scheme for all other cases.
15:13:02AlexPIt'd be in a branch at first anyway
15:13:05archivatorI'm not even sure if flite can do loadable voices (I think everything is compiled into the binary by default)
15:13:07S_a_i_n_tI've been looking for (with limited success) a decent, natural sounding SAPI voice for ages....any suggestions?
15:13:27archivatorAlexP: right. It's just that I don't see it replacing the current scheme any time soon.
15:13:31AlexPBut yes, I imagine the GSOC would be for English, but it would be necessary that it is possible to extend to other languages
15:14:27CIA-8New commit by 03teru (r24718): playlist catalog: improve UTF-8 BOM handling. correct length of array passed to create_playlist_list.
15:14:32gevaertsI'd not say that it's necessary
15:14:34archivatorS_a_i_n_t: sorry, I don't do SAPI. In Linuxlandia, try festival with a clustergen or multisyn voice. Even better, try building your own limited-domain voice (see festvox). You'd be surprised how accurate those can be.
15:14:41 Join ecio_ [0] (~ecio@32.135.232.200)
15:15:12gevaertsIt's highly desirable, but if it appears that only one TTS system would actually work on rockbox, and it only does English, we still want it
15:15:38S_a_i_n_tarchivator: thanks :D
15:15:40AlexPyeah, OK
15:15:55AlexPbut it'd have to be an impossibility
15:16:36gevaertsThere will be tradeoffs whichever system is chosen
15:16:37archivatorgevaerts: it's not a matter of "it only does English" - at the most basic level no TTS system cares about the underlying language. It's more about refactoring the system in a way that allows runtime loading of voices (I hope that flite supports that though I have my doubts).
15:17:21gevaertsarchivator: if you can do it with runtime loading of chunks of code, that's fine
15:17:40 Quit ecio (Ping timeout: 260 seconds)
15:17:40 Nick ecio_ is now known as ecio (~ecio@32.135.232.200)
15:18:17 Quit bmbl (Quit: Bye!)
15:19:25archivatorYeah, you're right, that would solve it. I'm gonna dive into flite now to see just how much floating point math there is...
15:19:47*FlynDice has lighted display on clip+ \o/
15:20:05ThomasAHFlynDice: yeah!
15:23:47funmanFlynDice: cool!
15:24:04funmani'm curious to see your code (can't wait the commit :P)
15:25:30 Join Sajber^ [0] (~Sajber@h-143-173.A213.priv.bahnhof.se)
15:27:40FlynDicefunman: Here's the diff http://pastie.org/828922
15:28:09FlynDicefunman: have you flashed yours yet?
15:29:01 Join Mokli [0] (~vali-moel@p5B0EB6FE.dip0.t-ipconnect.de)
15:29:09Moklihi
15:29:23Moklii have a problem with my nano 4th gen
15:29:26funmannope, still didn't look at mt firmware
15:29:35funmanFlynDice: is the output normal (SSP code) ?
15:29:36FarthenMokli: nano 4g is not currently supported
15:29:41Moklii will use ipod linux on it
15:30:05Moklibut the installer for windows says me: "no ipod found"
15:30:09FlynDicefunman: no, garbage output but the screen is lit
15:30:09Mokliwhat i can do?
15:30:25funmanMokli: nothing
15:30:30Farthenit is also not supported by ipodlinux and this is no ipodlinux support channel
15:31:01Mokliyes^^ but the ipod linux support channel is very bad...
15:31:27Mokli19 user are in the chat of ipod linux...
15:31:29Bagderipod linux doesn't work on the nano 4th gen either
15:32:01Moklino, on the website i read it!
15:32:10Farthenipodlinux is dead anyway. it is old and no one uses it
15:32:22BagderMokli: ipod 4th gen perhaps, not nano
15:32:40Moklihttp://ipodlinux.org/wiki/Project_Status
15:32:52Mokliohh
15:32:56Moklisorry
15:33:08Moklii think youre right^^
15:33:14*Bagder takes the point from Mokli and shows the audience
15:33:33Moklii have look for 4th gen, but no for 4th gen nano^^ sorry
15:33:49Moklithank you very much!
15:34:33funmanFlynDice: OF use DMA for SSP transfers, not sure how they mix that with I2SOUT and SD
15:34:37AlexPAlso, regardless of the lack of support in the ipodlinux channel, this isn't an ipodlinux support channel
15:34:55funman(and the fact that there is only 2 channels in as3525, perhaps they have more channels in the as3525v2 ?)
15:35:18Moklihttp://l4n.clustur.com/index.php/Main_Page
15:35:34Mokliwhats that?
15:35:46Farthen#linux4nano
15:35:54S_a_i_n_tthey have some code running on the 4g....but that;s besides the point.
15:35:56AlexPMokli: Nothing to do with Rockbox
15:36:02Mokliyes..but how i can use it?
15:36:18AlexPyou can't
15:36:27Bagderas a doorstop!
15:36:32archivator:( flite is full of floating point math. There is a fixed-point version of the main bits but quite a lot of floating point math surrounding them. I'll have to ask on the mailing list how feasible it is to convert the remaining code into fixed point..
15:37:09S_a_i_n_tMokli: Try in linux4nano....AND HAVE SOME PATIENCE!
15:37:09AlexPMokli: Neither Rockbox nor ipodlinux nor anything else run on the 4th gen nano. There has been some progress with finding an exploit, but nothing uses it yet. It will be a long long time if ever that anything runs
15:37:10CIA-8New commit by 03funman (r24719): mkamsboot: update dualboot.c / dualboot.h in place ...
15:37:21S_a_i_n_thelp WILL come, if you wait.
15:37:22FlynDiceback later
15:37:29Mokliokay, i'am sorry..
15:37:36 Part Mokli
15:37:44 Join Mokli [0] (~vali-moel@p5B0EB6FE.dip0.t-ipconnect.de)
15:41:49stoffelspeaking about ipod: what hdd-based player is recommended today? i'm thinking about retiring my h340 (2nd dead battery)
15:42:34AlexPThere aren't any that are still made
15:42:43S_a_i_n_twhy is "a HDD-based player" "speaking about iPod"?
15:43:11stoffelS_a_i_n_t: only because i've been looking at the ipod 5th generation today ;)
15:43:14AlexPThe gigabeats are good if you don't mind no recording or radio (F/X)
15:43:52AlexPbut it really depends what you want etc.
15:44:13 Join moos [0] (moos@rockbox/staff/moos)
15:45:07stoffelwell, playback only would be ok. also i don't need a radio. but a ~60-80GB hdd would be nice
15:45:52gevaertsgigabeat F or X is a nice player then. They come out of the box up to 60GB
15:45:53AlexPIf you don't want to change the disk, then a gigageat F60/X60/S60 or an ipod 60/80 GB (make sure to get the correct gen) are the options
15:46:08AlexPS60 is slightly less supported
15:46:56funmanclip+ bootloader fits in 0x50000 bytes of ram
15:47:56stoffelalternatively, can anyone recommend a shop for quality batteries for the h340?
15:48:08AlexPcameron sino are a good brand
15:48:11funmanFlynDice: the only thing which looks wrong in dualboot.S is the fact that GPIOB_DIR is reset to 0x0, can you try resetting it to the original value (after a sanity check in case this could cause a brick)
15:48:16AlexPYou can find those in various places
15:48:22stoffeland thx for the recommendations for other players
15:48:57 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:49:31 Quit teru (Quit: Quit)
15:53:17 Part Mokli
15:54:26 Quit froggyman (Quit: we're hard against impotence)
16:00
16:07:12 Join robin0800 [0] (~quassel@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
16:13:09 Join CGL [0] (~CGL@190.207.235.219)
16:15:14TheSeveni would like to commit boosting for nano2g. any objections?
16:16:26 Nick perfectdrug__ is now known as perfectdrug (~marko@p5B0ED07B.dip.t-dialin.net)
16:20:59 Join TheUser [0] (~theuser12@80.93.115.26)
16:22:03 Quit MethoS- (Read error: Connection reset by peer)
16:25:21 Quit S_a_i_n_t (Ping timeout: 248 seconds)
16:25:23 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.55)
16:28:49 Join MethoS- [0] (~clemens@134.102.106.250)
16:31:01 Join markun [0] (~markun@rockbox/developer/markun)
16:32:01 Quit CGL (Quit: Saliendo)
16:32:19 Join Schmogel [0] (~Miranda@p3EE219AB.dip0.t-ipconnect.de)
16:33:27 Join CGL [0] (~CGL@190.207.235.219)
16:33:58 Quit petur (Quit: *plop*)
16:35:22 Quit TheUser (Read error: Connection reset by peer)
16:35:51TheSevenis it a known issue that doom fails to build with eabi?
16:39:10 Join TheUser [0] (~theuser12@80.93.115.26)
16:40:01ZagorTheSeven: I didn't know, at least
16:45:34 Join m3dlg [0] (~m3dlg@212.183.140.22)
16:45:50TheSevenZagor: it doesn't fit into PLUGIN_RAM by 1288 bytes
16:46:03Zagoraha
16:46:17 Join _TheUser_ [0] (~theuser12@80.93.115.26)
16:46:40Zagorwhat makes the size grow with eabi?
16:47:45 Quit TheUser (Ping timeout: 248 seconds)
16:47:59TheSevenno idea, but the main binary is also only ~2K smaller, and there is at least one point where it has a higher stack usage
16:49:21CIA-8New commit by 03theseven (r24720): Increase iPod Nano2G NAND stack size (fixes FS #11030)
16:52:13 Join toffe82 [0] (~chatzilla@12.169.218.14)
16:54:50CIA-8New commit by 03theseven (r24721): Enable iPod Nano 2G boosting (without undervolting for now)
16:55:31S_a_i_n_tTheSeven: Someone's been busy :P
16:55:45***Saving seen data "./dancer.seen"
16:55:46 Quit m3dlg (Quit: RAGE QUIT)
16:56:32CIA-8New commit by 03theseven (r24722): Somewhat calibrated battery parameters for Nano2G - still not very accurate though, especially while charging.
16:56:41 Quit CGL (Ping timeout: 240 seconds)
16:57:04S_a_i_n_tbusier stikk...
16:57:15S_a_i_n_t*still rather
16:57:52 Quit _TheUser_ (Ping timeout: 256 seconds)
16:57:55CIA-8New commit by 03theseven (r24723): Add more information to the battery debug screen for Nano2G
16:58:49TheSevenOh damn...
16:59:44 Join TheUser [0] (~theuser12@80.93.115.26)
17:00
17:00:32 Quit TheUser (Client Quit)
17:00:42 Join CGL [0] (~CGL@190.207.235.219)
17:01:09 Join TheUser [0] (~theuser12@80.93.115.26)
17:06:57 Quit Farthen (Ping timeout: 248 seconds)
17:12:24CIA-8New commit by 03theseven (r24724): Revert accidental changes from r24723
17:16:26 Join jgarvey [0] (~jgarvey@cpe-071-070-229-181.nc.res.rr.com)
17:26:32 Join Tomis2 [0] (~Tomis@70.134.96.35)
17:29:13 Quit Tomis (Ping timeout: 276 seconds)
17:29:14 Nick Tomis2 is now known as Tomis (~Tomis@70.134.96.35)
17:29:46 Join matsl [0] (~matsl@dhcp115.contactor.se)
17:30:17 Quit matsl (Client Quit)
17:31:21CIA-8New commit by 03theseven (r24725): Reduce Nano2G main thread stack size to the default value again. The huge sector buffers of the FTL have been moved to the bss some time ago, so this ...
17:32:26*TheSeven eats CIA-8
17:32:27*CIA-8 tastes crunchy
17:33:55archivatorIs there any reason why there isn't integrated static analysis of the rockbox codebase (splint and the like)? Theoretically, it could catch a few slip ups here and there..
17:34:41 Join pamaury [0] (~pamaury@ALyon-551-1-70-117.w92-137.abo.wanadoo.fr)
17:35:46 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
17:40:24JdGordonAlexP: OK, fine, I'll add very basic alignment support to the playlist viewer.
17:43:53TheSevenhuh, ramsize doesn't include the main thread's stack?
17:47:00 Join fyrestorm [0] (~nnscript@cpe-24-90-81-175.nyc.res.rr.com)
17:48:19 Join phanboy_iv [0] (~benji@c-24-98-43-198.hsd1.ga.comcast.net)
17:52:34 Join anewuser [0] (anewuser@unaffiliated/anewuser)
17:53:58 Quit Bagder (Quit: It is time to say moo)
17:54:17AlexPJdGordon: Ta very much :)
17:58:50gevaertsarchivator: the usual reason I guess
17:59:27archivatorgevaerts: no one's gotten around to doing it?
18:00
18:00:04gevaertsyes
18:04:07 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
18:05:14funmancool! i have an answer from the person who mentioned using the as3525/as353x linux code on the chinese forum
18:05:54 Join Luca_S [0] (~5d3fc54b@giant.haxx.se)
18:05:59 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:07:21 Join Bullet` [0] (~Bullet@AMontsouris-159-1-85-243.w92-140.abo.wanadoo.fr)
18:07:37 Join Guest [0] (~7375c95d@giant.haxx.se)
18:08:01 Join freddyb [0] (~fred@pool-68-238-8-141.chi.dsl-w.verizon.net)
18:08:11 Quit Luca_S (Client Quit)
18:10:17GuestHi, I had come here a few days back regarding USB mode for Ipod with V3.5. I am facing same issue again. When I connect ipod video 30 gig to pc, it shows rockbox's usb screen. For connecting as ipod's USB mode/screen, I need to press Menu while connecting. Is that expected??
18:10:24 Quit komputes (Ping timeout: 248 seconds)
18:10:38freddybI rewrote FS #11019. Now it just adds a flag to specify the background image file.
18:12:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:14:37 Join grndslm [0] (~grndslm@174-126-14-4.cpe.cableone.net)
18:16:11 Quit robin0800 (Remote host closed the connection)
18:16:27funmanI hope I can get the "Board Support Package" for as352x/as353x
18:16:30 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:17:28FlynDicefunman: I have recognizable output for clip+ now
18:19:00funmannice
18:19:03FlynDiceI'm changing dualboot.S now but I think the whole GPIOB manipulation is unneeded...
18:19:10GuestHi, I had come here a few days back regarding USB mode for Ipod with V3.5. I am facing same issue again. When I connect ipod video 30 gig to pc, it shows rockbox's usb screen. For connecting as ipod's USB mode/screen, I need to press Menu while connecting. Is that expected??
18:19:22funmanFlynDice: in your tests it was only needed for reading power button ?
18:19:37funmanif that's the case then i guess you could just remove it completely
18:20:09FlynDiceYes, I also posted some commented disassembly on the forum to support that
18:21:05funmanis it in one of the pasties?
18:21:12FlynDiceyes
18:22:00gevaertsGuest: I don't understand what you mean. If you connect 3.5 normally, it should reboot to the Apple Emergency Disk Mode, if you hold menu, it should remain in rockbox, without going to mass storage mode. Are you seeing something different?
18:22:54FlynDicefunman: My preference would be to remove the GPIOB manipulation altogether and just check C3 and A1
18:22:54 Quit archivator (Remote host closed the connection)
18:23:14funmani'm going to try http://pastie.org/829247 , do you see anything wrong with it ?
18:23:26FlynDicelooking
18:23:46Guestgaverts: yes. If i connect ipod to pc, when it is switched off, it frst boots to rockbox, then shows usb cable and "multimedia mode" for a couple of seconds. Then it moves to a screen showing "do not disconnect" but that is not a apple screen
18:24:17GuestIf i press menu while connecting to pc, then actual apple's "do not disconnect" screen comes.
18:24:24funmanFlynDice: you had tested without setting GPIOB iirc (I don't see it in the git log, but i remember pestering you so you add this check)
18:25:19FlynDicefunman: That looks like clipv2 you changed for. Yes, I started out with no GPIOB changes and it worked fine.
18:25:21gevaertsGuest: both of those are apple screens, and yes, that's expected
18:25:59funmanFlynDice: i changed the RAM_SIZE of Clip+ from 1MB to 0x50000 because the compressed OF+bootlaoder fit in there
18:26:15Guestohh, so you mean the screen with small font "do not disconnect" and small circle with a line is apple's screen, not rockbox's?
18:26:20funman(and decompressed)
18:26:48gevaertsGuest: it is, yes. That's the Apple in-ROM USB mode
18:26:55FlynDiceI think setting B0 is required to read the pwr button. I could not read the pwr button without B0 set
18:27:12funmanclipv2 firmware block is 383K while clip+ is 131K (much closer to clipv1)
18:28:12FlynDiceIn the diff you asked about I see line 10 +#if defined(SANSA_CLIPV2) not defined(SANSA_CLIPPLUS)
18:28:36funmanthis line only affects RAM_SIZE (see 3 following lines)
18:29:18 Join mt [0] (~mtee@rockbox/developer/mt)
18:29:48FlynDiceOh, sorry, problem reading a diff instead of applied patch ;-)
18:29:49Guestgaverts: okay. Searching " Apple in-ROM USB mode" on google to confirm.
18:30:34funmani think we should also check if decompressed OF/rockbox + uclpack function fit in the memory
18:30:43AlexPGuest: You don't need to, he both knows what he is on about and is telling the truth
18:31:06 Join JdGordon1 [0] (~jonno@68-27-107-67.pools.spcsdns.net)
18:32:04GuestAlexP: yes sir, I know no one lies here. I was just checking. Sorry if that hurt you/gaverts
18:32:16AlexPIt is gevaerts
18:32:22JdGordon1AlexP: alignment will be crap though. it will be the whole line, not sublines
18:32:23AlexPAnd no, of course it didn't hurt
18:32:27gevaertsI don't know if it hurt this gaverts person. I myself don't mind :)
18:32:36AlexPI was just trying to save you the trouble
18:32:48 Join karashata [0] (~karashata@74-220-162-11.wightman.ca)
18:33:06AlexPJdGordon1: What does subline mean in this context?
18:33:20JdGordon1%alleft%arright wont work
18:33:26AlexPah, OK
18:33:28JdGordon1its one or the other (or %c)
18:33:28AlexPThat's fine
18:33:36GuestAlexP and gevaerts. thanks. Agreed
18:34:13AlexPFor me it is just for using to do next or prev preset
18:34:29AlexPSo it'll only have one line and one piece of info
18:34:55FlynDicefunman: Ok, I'm with you now. The button reads look fine to me. I can't comment on the ram size change though....
18:35:37 Join Riku [0] (~Lss@cm48.omega219.maxonline.com.sg)
18:35:47funmanthanks, i'll double check this change and wait for your lcd code so i can expect a big surprise (or a brick)
18:36:05 Nick Riku is now known as Guest25299 (~Lss@cm48.omega219.maxonline.com.sg)
18:36:54 Quit TheSeven (Read error: Connection reset by peer)
18:37:00 Quit Guest94132 (Read error: Connection reset by peer)
18:37:20 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
18:37:45 Join archivator [0] (~archivato@stu0279.keble.ox.ac.uk)
18:38:32FlynDicefunman: If you want to to check the code you want to use it's relatively simple
18:38:53FlynDiceI've got a sort of a test setup that wrks pretty well
18:39:01 Quit Guest (Quit: CGI:IRC (Ping timeout))
18:39:05 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:39:24funmanif the button check works on my clip+ (because it didn't on mt's clip+)
18:39:27FlynDiceI guess test is what I really meant, not check
18:39:38 Join leavittx [0] (~leavittx@cl-534.mbx-01.si.sixxs.net)
18:40:26funmanmkamsboot happily accepts an OF too big to fit decompressed in RAM
18:42:07JdGordon1AlexP: http://pastebin.com/m76595ca2
18:42:51AlexPoooh, lovely, ta :)
18:46:16 Quit JdGordon1 (Ping timeout: 240 seconds)
18:46:23 Join saratoga [0] (~9803c6dd@gateway/web/freenode/x-kkicxqfnpyugyniu)
18:49:00kugelJdGordon: it dosn't have the special handling get_line() has but it looks good enough, thanks :)
18:49:13FlynDicefunman: with this diff: http://pastie.org/829296 I get a blue ROCKBOX that scrolls off the bottom of the screen and some numbers I have'nt investigated yet.
18:49:45 Join JdGordon_ [0] (~Miranda@rockbox/developer/JdGordon)
18:49:46 Quit freddyb (Quit: Konversation terminated!)
18:50:41funmanperhaps the screen isn't aligned on 0x0
18:51:20JdGordon_kugel: what special handing?
18:51:27JdGordon_you mean splitting the line?
18:51:44kugelif %ac and %al are on one line and the strings overlap, for example
18:51:51 Join bertrik [0] (~bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
18:51:55JdGordon_yes, see above, I said I wouldnt do that
18:52:13JdGordon_although it might not be so hard after all, *if* we ignore overlapping
18:53:00kugelhow hard would it be to call get_line()?
18:54:31 Quit stoffel (Ping timeout: 246 seconds)
18:54:37JdGordon_a complete rework
18:54:46JdGordon_the playlist viewer doesnt store full tokens
18:54:48leavittxHi everybody! I'm using latest svn rockbox build on my Cowon D2 and when I run jpeg plugin, it says "Incompatible version". Wtf? :)
18:54:50JdGordon_only token types
18:54:58JdGordon_I want to change the whole system to work properly
18:55:08leavittxSorry for interrupting :-\
18:55:23JdGordon_leavittx: sounds like you didnt unzip the whole .rockbox folder
18:55:46***Saving seen data "./dancer.seen"
18:55:48leavittxJdGordon_: no, I did it
18:56:07leavittxJdGordon_: I did it many times before actually =)
18:56:39JdGordon_well, thats about the onl reaosn you'd get that error, so I dunno what's wrong
18:56:54AlexPleavittx: Could you check the file times to make sure?
18:57:11leavittxAlexP: what files?
18:57:43leavittxAnd there is no way to exit f*cking boomshine!!!
18:58:34AlexPbetween the main Rockbox binary (.rockbox/rockbox.somethingorother) and the jpeg viewer (.rockbox/rocks/viewers/jpeg.rock)
18:59:14Torneleavittx: there is a way to exit every plugin
18:59:46 Quit flydutch (Quit: /* empty */)
18:59:50leavittxTorne: hard reset isn't a good way
19:00
19:00:04Torneno, I mean there is a *proper* way to exit every plugin
19:00:08Tornesome of them are just not very obvious
19:00:27Tornewe don't include plugins you can't get out of
19:00:29kugelTorne: I don't think that's true for all targets
19:00:46JdGordon_worst case is using a hammer or pin (depends on target) to reset :)
19:00:56leavittxAlexP: dates are equal =(
19:01:05Tornekugel: really?
19:01:19AlexPHmmm, boomshine doesn't appear to be in the manual
19:01:24Torneit isn't, no
19:01:28kugelyes, I'm sure there are a dozen of plugins on the yh* which you can't exit properly
19:01:28pixelmaI have my doubts about boomshine (the only lua plugin)
19:01:30*AlexP scolds the person that added it without a manual entry
19:01:49gevaertsDidn't the jpeg viewer and the bmp viewer get merged into one image viewer?
19:01:51Tornedoes the lua plugin not have a global exit?
19:01:55UnhelpfulTorne: untrue for some plugins using PLA - if you map over wrong thing when mixing contexts the action you assign to exit may be unavailable.
19:02:02TorneUnhelpful: that's by accident, though :)
19:02:07kugelgevaerts: no, not exactly
19:02:08pixelmait's not even playable (visible (enough)) on all players it exist
19:02:11Tornei.e. a bug
19:02:19kugelsame ui code but they're separate plugins still
19:02:45Unhelpful[09:27:36] <Torne> we don't include plugins you can't get out of <- insert "intentionally" somewhere, then? ;)
19:03:07Torneexit button for boomshine is ACTION_KBD_ABORT from CONTEXT_KEYBOARD
19:03:08FlynDicefunman: I tested the button code in your diff and It works just fine. Once again, the ram size change I can't comment on ;-)
19:03:10pixelmagevaerts: and there is no bmp viewer yet (if I didn't miss the commit) but a patch - and a png viewer for colour tarrgets
19:03:11Torneaccording to the code
19:03:12 Join kugel_ [0] (~kugel@e178122131.adsl.alicedsl.de)
19:03:21Tornenot sure what button that is on D2
19:03:22Torneif any :)
19:03:53funmancool, i want to test with skyeye before trying on target but i hav esome trouble
19:04:04Torneand it tests that *first* so even if th ebutton mapping overlapped with something else it would work
19:04:38AlexPJdGordon_: What did you generate that patch with? patch claims the input is garbage :)
19:04:48AlexPI'll apply it manually :)
19:05:00TorneYeah that is indeed not a button on the D2, ah well
19:05:34JdGordon_svn diff :(
19:05:44Torne...how does the keyboard even work on d2 then. it doesn't have a confirm button either ;)
19:06:34funmani'm executing an elf file which does mov r0, #0; bx r0
19:06:41funmanb *0; stepi; stepi causes a break
19:06:56 Quit kugel (Ping timeout: 248 seconds)
19:06:57funmanb *0; continue just locks and ^C aborts the program so I can't see where it is executing
19:07:03JdGordon_AlexP: hows your c skills? if that works it just needs to _RTL cases fixed up and can be commited, if you want to do it :)
19:07:20AlexPummm :)
19:07:36Torneok i can't see how the keyboard can possibly work on d2
19:07:39Torneahwell :)
19:07:41JdGordon_thats just a matter of adding some is_lang_rtl()'s
19:07:46Torneperhaps it doesn't
19:08:32Tornethe author of boomshine has assumed that the virtual keyboard mappings are valid :)
19:13:49 Join Adub- [0] (~Rotting@xplr-ts-t11-208-114-159-122.barrettxplore.com)
19:15:10AlexPJdGordon_: Should this work alongside the fms patch?
19:15:28JdGordon_you'll need to merge it manually
19:15:35JdGordon_I was working from a clean tree
19:15:39AlexPyeah, I did - but it fails to compile
19:15:40JdGordon_shouldnt be too hard to merge
19:15:48AlexPOne mo, I'll pastebin the errors
19:17:21*gevaerts sees three differently sized and coloured analogue clocks on his wps :)
19:18:05AlexPJdGordon_: http://pastebin.com/m66cfd4a1
19:18:21JdGordon_gevaerts: apart from being loadable anywhere, is there any actual difference between your pebbles and rocks?
19:18:22CIA-8New commit by 03b0hoon (r24726): Packard Bell Vibe 500: Full RTC support.
19:18:42gevaertsJdGordon_: the difference is slowly shrinking
19:18:48funmanmkamsboot makefile rules are broken: successive runs of make randomly rebuild some targets
19:19:45gevaertsJdGordon_: is there a place where I can call cleanup routines for wps things?
19:20:13leavittxAlexP, JdGordon_: problem with jpeg plugin fixed with one minute brain activity and "make clean". Make is so f*cking make. Or it isn't it's bug? ;)
19:20:31JdGordon_no. right now it just blows everything away and important things get re-init when the new cgf is loaded
19:20:52gevaertsok. Good enough for me right now :)
19:21:05JdGordon_AlexP: line 327 needs to be change to viewer->vp->width
19:21:16AlexPOK :)
19:21:38JdGordon_can you pastebin that whole block? im not sure about the rest
19:21:46 Join steckel [0] (~5ce483cd@gateway/web/freenode/x-vrhjogrudxaxisdv)
19:21:53JdGordon_sounds like you need a break; at the end of the second switch
19:22:03AlexPWhole block of the source? Coming up
19:22:20JdGordon_300-350
19:22:43 Join Bagder [0] (~daniel@rockbox/developer/bagder)
19:22:53funmanclip+ patched, OF boots!
19:23:07AlexPJdGordon_: http://pastebin.com/m60fb928d
19:24:00JdGordon_http://pastebin.com/m33b69e05
19:24:10JdGordon_you can probably just copy that straight over the top
19:24:18AlexPcheers
19:24:37JdGordon_oh, the getstringsize line is missing a ,
19:24:45JdGordon_bufm should be buf,
19:26:36 Join kugel__ [0] (~kugel@e178109168.adsl.alicedsl.de)
19:27:53CIA-8New commit by 03funman (r24727): Clip+ remove setting GPIOB PIN0, it seems needed only for power ...
19:28:21AlexPJdGordon_: I noticed the missing , but was just about to pastebin you the errors of it moaning about bufm being undeclared :)
19:28:29 Quit CGL (Ping timeout: 265 seconds)
19:28:41AlexPI didn't twig that m is next to , :)
19:28:54AlexPcool, compiling now
19:29:05domonokyhm, the new Vibe 500 target isnt in the automated builds. Should it get added ?
19:29:51*FlynDice see's funman's success and cheers!
19:29:54 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:30:24 Quit kugel_ (Ping timeout: 264 seconds)
19:30:27funmanboth home & left act as expected
19:32:31 Quit steckel (Quit: Page closed)
19:34:35 Join tomers [0] (~chatzilla@bzq-84-109-85-100.red.bezeqint.net)
19:36:05ThomasAHfunman: pressing home+left together will boot OF?
19:36:13funmanany of them
19:36:24 Quit TheUser (Ping timeout: 264 seconds)
19:36:30 Quit phanboy_iv (Read error: Connection reset by peer)
19:36:38AlexPJdGordon_: Am I putting the %ar within the %Vp ? i.e. %Vp|1|%ar%s%Tn||
19:37:03ThomasAHfunman: and what else is already visible? Just an illumniated screen or even more?
19:37:43funmanlcd display is correct but off the screen
19:37:45pamauryTheSeven: I have an explaination of the suspicious 4% of yesterday !
19:38:02funmantoo much on the left and too much on the top, you can't see the left part and the bottom is garbled
19:38:04TheSevenwhich would be?
19:38:19ThomasAHfunman: correct as in Rockbox screen?
19:38:27TheSevenbtw, it's 77% now
19:38:35funmanThomasAH: Rockbox screen?
19:39:14ThomasAHfunman: menu, clock or something else which is not a black screen :)
19:39:54pamauryIs there a good reason why dircache does the generation of the cache in the main thread instead of in his own thread when doing a non-transparent build (that is, on boot) ? Can't the kernel do threading at the that time ?
19:40:21 Join martian67 [0] (~martian67@about/linux/regular/martian67)
19:40:31funmanhmm something bad happened: the OF just started (refreshing DB) and when i plugged USB the screen went off
19:40:32JdGordon_AlexP: yes, except get rid of %s
19:40:38funmanoh it's back
19:40:46AlexPOK, I'll try without %s
19:40:58AlexPIs it not possible to have scrolling + alignment?
19:41:11funmanThomasAH: it's only the bootloader (no menu), but it's rockbox code and it displays fine
19:42:21JdGordon_AlexP: http://pastebin.com/m3b680988 should work with %s
19:42:21CIA-8New commit by 03b0hoon (r24728): Packard Bell Vibe 500: add RTC alarm.
19:42:22funmanah i got it to display fine
19:42:31AlexPJdGordon_: cool :)
19:42:36JdGordon_remember I wrote that change in 10min on the bus :p
19:42:46 Quit danopia (Ping timeout: 256 seconds)
19:42:48AlexPI'm just trying without for now
19:43:03AlexPAnd it isn't showing the correct names with the alignment tags
19:43:07AlexPWithout it is fine
19:43:13ThomasAHfunman: ah, the boot loader has some visible parts. Never saw rockbox booting (only running on a co-worker's machine)
19:43:16JdGordon_what's it showing?
19:43:16CIA-8New commit by 03funman (r24729): Clip+: LCD display (with FlynDice help)
19:44:03funmanFlynDice: i corrected lcd_write_data(), and reduced the changes to GPIO*_DIR, can you check if it worsk for you ?
19:44:47AlexPJdGordon_: So I have %Vp|1|%ar%Tn|| so it shoukd be showing the next preset. What it is showing is the previous preset and the next preset concatanated
19:44:53FlynDicegimmee a few mins, paying bills right now.. cool!
19:45:06funmanwhat, paying bills is cool? i'll send you mine :)
19:45:07AlexPSo e.g. Radio1Radio3 instead of Radio3
19:45:17*ThomasAH checks out svn now to make the life of his clip+ a bit more dangerous :)
19:45:55funmanThomasAH: you're aware that 2 clip+ have been bricked already and we're not sure what happened exactly ?
19:46:36funmanright now it's 2 working and 2 bricked, so you could help making the statistics on one side or the other, but keep in mind it's experimental code
19:46:59AlexPgevaerts: Where is the funky screenshot? :)
19:47:14 Quit Zagor (Quit: Clint excited)
19:47:24ThomasAHfunman: yes, and if one of your clips is bricked earlier than mine, just ask me ... and if mine bricks earlier, you'll have a better statistic :)
19:48:20JdGordon_AlexP: wierd, and without %ar it shows correctly?
19:48:28AlexPYep
19:48:45gevaertsAlexP: especially for you, gevaerts/pebble_clock.png">http://www.evonet.be/~gevaerts/pebble_clock.png
19:49:07JdGordon_thats a pretty silly screenshot! they all show the same thing
19:49:10JdGordon_BORIGN!
19:49:11AlexPgevaerts: cool! :)
19:49:26JdGordon_make the clocks fly around the display!
19:50:49AlexPJdGordon_: I've tried the latest fms patch two - a couple of little things. The menu backdrop shows through on scrolling lines, and when you exit the fms to the menu the menu backdrop isn't properly drawn again.
19:50:57AlexPWould you like them on flyspray?
19:51:06JdGordon_the latest latest?
19:51:15JdGordon_scrolling and backdrop both worked fine for me
19:51:30AlexPThe one at 07:11 today
19:51:32JdGordon_unless I'm using a backdrop which doesnt make it obvious enough
19:51:35JdGordon_arg
19:51:59AlexPI'll take a couple of screenshots, one mo
19:52:42 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
19:53:03JdGordon_AlexP: ah, add a tempbuf[0] = '\0'; under the line "alignment - viewer->....."
19:53:04funmanperfectdrug: if you want to do some drawing, we need a Clip+ image for the sim
19:53:35perfectdrugfunman: sure is it any different?
19:53:42funmanhm the Clip simulator use a real picture
19:54:03perfectdrugpretty much all simulators use real pictures
19:54:17funmansvgs are used in the manual right?
19:54:21perfectdrugyes
19:54:23funmani could take a picture of my clip+ then
19:54:52perfectdrugplease do or make a scan
19:55:10perfectdrugwhat is the difference?
19:55:21 Join Strife89 [0] (~michael@168.16.237.214)
19:55:22funmanwith clipv1/v2 ?
19:55:24 Join CGL [0] (~CGL@190.207.235.219)
19:55:49bertrikfunman, FlynDice nice work!
19:55:55 Join Anaxite [0] (~aka@pdpc/supporter/professional/anaxite)
19:56:16perfectdrugfunman: yes
19:56:17funmaneh, i just joined the party, leave the credits to FlynDice :)
19:56:36AlexPJdGordon_: Better but not quite - now the last few letters are cut off so instead of Radio3 I see Radi :)
19:57:08AlexPhmmm, but not always
19:57:12AlexPsometimes it is OK
19:57:37*FlynDice just sands the edges smooth...
19:57:52JdGordon_well that doesnt make much sense
19:59:07perfectdrugif I remember correctly the clip was my first SVG I made and it looks a little bit rough now:D
19:59:21funmanperfectdrug: the buttons are different (clip+ has volume on the left side, power on top, no hold button), obviously the µSD slot (on the right side, bottom), it has a more square design, a square buttons pad, a square home button, a not removable clip
19:59:29AlexPAnd on short names (three letters) they seem to be left aligned
20:00
20:00:22perfectdrugfunman: ok sounds fun a scan would be useful or a nice photo from the right angle
20:00:52 Quit Adnyxo (Ping timeout: 265 seconds)
20:01:24 Quit ecio (Quit: Colloquy for iPhone - http://colloquy.mobi)
20:02:15AlexPJdGordon_: http://aeparker.com/files/dump1.png and http://aeparker.com/files/dump2.png
20:02:44AlexPer, they aren't meant to be identical - one mo :)
20:04:02 Join ecio [0] (~Moo@adsl-065-081-069-051.sip.mco.bellsouth.net)
20:04:15AlexPOK, dump1 updated
20:04:36AlexPJdGordon_: dump1 shows a scrolling line in the fms, and dump2 shows having just exited the fms to the menu
20:05:05JdGordon_oh, the backdrop in the fms is correct? its when exiting that shows the error?
20:05:38AlexPThe backdrop in the fms is correct except for the small section at the top (that is also not correct in the WPS, but is fine in svn)
20:05:58AlexPbut the main issue is exiting the fms, yes
20:06:43 Join stoffel [0] (~quassel@p57B4DFAA.dip.t-dialin.net)
20:07:07AlexPOn dump1 you can also see the issue with the right aligning - the preset that says "Boome" is actually "Boomerang"
20:07:14funmanhttp://img214.imageshack.us/img214/8272/clipu.jpg < photo looks bad, i'll try the scanner
20:07:28JdGordon_AlexP: how wide is that viewport?
20:08:10JdGordon_haha, typo...
20:08:11AlexPJdGordon_: 110 - %Vl|o|130|219|110|16|1|FFFFFF|000000|
20:08:14FlynDicefunman: Screen works for me just fine
20:08:20JdGordon_x = width = length
20:08:25JdGordon_guess which is wrong :p
20:08:42AlexPheh :)
20:09:08funmannice
20:09:13perfectdrugfunman: maybe lit up the display so I can see the boundaries
20:09:21*JdGordon_ doesnt like coding on his netbook.. damn keyboard is just too small
20:09:32AlexPyeah, I bet
20:09:45FlynDiceany idea what the dma&minus;&minus;> 2000 6D06 is from?
20:10:19topikwhy not ask google images, funman?
20:10:23funmanFlynDice: sd-as3525v2.c debug code (from init_storage())
20:10:31AlexPJdGordon_: So - length?
20:10:46JdGordon_good guess :)
20:11:33AlexPJdGordon_: OK, that works nicely :)
20:11:46AlexPI'll just try to add the %s bit in
20:12:16funmanhttp://img526.imageshack.us/img526/2432/outo.jpg < scanner (the blue "sansa" isn't on the screen)
20:12:34funmantopik: we'd need an image with appropriate licence
20:12:52pixelmaperfectdrug: saw a few pictures on webshop pages of the Clip+, maybe not highest resolution but from the right angle and good enough
20:13:06funmanbtw we don't have an image for clipv2, do we want to have the exact correct picture of a v2 (with the different font in the sansa logo) or copy the clipv1 one ?
20:13:44perfectdrugfunman: this is perfect!:)
20:14:03funmancan you see the screen borders ?
20:14:20 Join Adnyxo [0] (~aaron@adsl-065-013-002-216.sip.asm.bellsouth.net)
20:15:07topiknice pic that one
20:15:29topikgiant killer clip
20:15:34 Quit stoffel (Remote host closed the connection)
20:15:47funman:p
20:17:03perfectdrugfunman: not really no but it is not that important for now, is the border around higher than the inner part, does it stand out a little bit?
20:17:05funmanhttp://img59.imageshack.us/img59/2903/out0003.jpg < the screen is more contrasted (i put a light under the scanner)
20:17:23FlynDicefunman: are you actually getting the ROCKBOX logo to display without scrolling off?
20:18:26JdGordon_AlexP: assuming %s works, can you post the diff and those screenshots on the tracker please?
20:18:30funmanyes, i added a while(1) ; after backlight_init() in bootloader/sansa_as3525.c
20:19:02perfectdrugfunman: this is even better thanks, the round corners are the hardest part, let me try
20:19:13 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
20:19:13 Quit pixelma (Disconnected by services)
20:19:30 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
20:19:39 Quit amiconn (Disconnected by services)
20:19:40 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
20:20:04 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
20:21:36 Join Farthen [0] (~chatzilla@e179233161.adsl.alicedsl.de)
20:24:24AlexPJdGordon_: scrolling works, and sure :)
20:24:37JdGordon_sweet :)
20:24:46JdGordon_*almost* ready
20:24:56AlexPyep :)
20:25:06funmanFlynDice: which capacity has your clip+ ?
20:25:39FlynDice2GB... got the cheapest in case I bricked it ;-)
20:25:55funman^^
20:26:12funmani've 4GB, i'd like to see if the CSD differ
20:27:05FlynDicenp just another 30 sec and I'll have that for you....
20:28:42*FlynDice needs to go, back later
20:31:09CIA-8New commit by 03pamaury (r24730): Fix error handling in dircache generation.
20:32:21TheSevenpamaury: so what do you think did cause the 4% we observed yesterday
20:32:45AnaxiteOk; it's daytime now so I'll ask again. I've got a Clip v1, 2GB, that I don't need and am willing to donate for rockbox. Is there any interest?
20:33:38 Quit shaggy-h (Read error: Connection reset by peer)
20:33:41pamauryTheSeven: that's absolutely stupid ! When the system boot, it ask dircache to init. If dircache cannot load its state from cache, it does a non-transparent rebuild ... on the main thread ==> it uses the main thread stack, so dircache thread stats at 4%
20:33:53pamaury*cannot load its state from storage
20:34:25funmanuisdl.h doesn't appear in simulator's make.dep
20:35:34JdGordon_Anaxite: thanks, but I tihnk just about anyone interested in a clip has already thanks to many woots
20:36:18AnaxiteFigures, ok :)
20:36:29gevaertsUnless funman has broken another one :)
20:36:40 Join shaggy-h [0] (~kiwi@78-86-164-31.zone2.bethere.co.uk)
20:36:45funman^^ i still have the clip soap sent me
20:36:57TheSevenah crap. nano2g rolo seems to be broken once again
20:37:01AlexPAnaxite: That's a kind offer though, thank you
20:37:03TheSevendidn't I fix that days ago?
20:37:16gevaertsJdGordon_: if I want my shiny new %P tag to have a single string argument, how do I do that?
20:37:45AnaxiteWell, is there a hardware wishlist? I'd like to support the project somehow, but didn't see such a list anywhere (might have missed it?).
20:37:53pamauryTheSeven: so if you want to check the real dircache stack usage for sure, you have to do something like disable dircache, renable, reboot, reboot, check
20:37:56funmanAnaxite: best thing you can do is enjoy rockbox and tell us if you have problems
20:38:14JdGordon_gevaerts: add a parsers for %P (top of skin_parser.c), then inside that you need to handle it however you want
20:38:23JdGordon_usually in the form %P|you string|
20:38:25funmanAnaxite: i don't think there are developers who can't work because of missing hardware
20:38:36AnaxiteOk.
20:38:54AlexPI have a hardware wishlist but it is very seperate to Rockbox :)
20:39:33kugel__JdGordon_: the backdrop problem only happens if there's an sbs I think
20:39:58JdGordon_probably, my test theme doesnt have one
20:40:02bertrikfunman, jobec did another runtime benchmark with his sansa clip v1 with the radio explicitly powered down and recovered about 1.5 hours runtime :P
20:40:16bluebroth3rdoes anyone have the mbr for a nano2g 8gb around?
20:40:18JdGordon_and my main DAP doesnt have fm which means testing is only on sim :/
20:40:33funmanbertrik: nice! is it always powered on at startup?
20:41:34bertrikI don't know exactly. I think there is a generic bug in the radio code, it tries to power down the radio on init, but the function does nothing because it thinks it's already powered down
20:43:20 Join Stephen__ [0] (~S@86-40-164-98-dynamic.b-ras2.srl.dublin.eircom.net)
20:43:21TheSevenbluebroth3r: yep
20:43:43TheSevenstill need it?
20:44:09bertrikwe could modify the si4700 init to make it start up in full power-down mode
20:45:17bluebroth3rTheSeven: yep. Can you send it to me?
20:45:54Schmogelive got a clip+ 8gb. tell me if i can help in any way ;)
20:46:05funmanSchmogel: are you willing to risk bricking it ?
20:46:12Schmogelmh not really :D
20:46:55CIA-8New commit by 03funman (r24731): Sansa Clip+: simulator ...
20:47:09funmanyou can try running mkamsboot from current svn, it works on mine and FlynDice's clip+ but a previous version bricked mt's clip+
20:47:11Schmogelthe chance of bricking it is very high, isnt it?
20:47:21funman1/2 atm
20:47:33Schmogelmhh ok.
20:47:42funmanalthough the exact current code worked 100% of the 2 times it was tried
20:50:20TheSeventhere isn't even a hidden jtag on them?
20:50:50funmanprobably, i haven't opened mine
20:51:05saratoganot worth jtaging a 20 dollar player
20:51:08saratogajust buy another :)
20:51:20funman20 dollars for a clip+? my 4gb costed 65€
20:51:36funmanABI only has a clipv1 disassembly it seems
20:52:14funmani tried to solder wires on a dead v1 but it was difficult and i didn't know how to detect the correct pinout
20:53:20TheSevendoes anyone feel like potentially bricking his nano2g btw?
20:53:39pixelmaJdGordon_: with your recent %St changes, I could display the fmr file with only the file name?
20:53:43TheSeveni need more testers for norloader, it's supposed to be working on all devices now
20:53:55TheSevenand one can test it with a relatively low bricking risk
20:53:56JdGordon_pixelma: yes, that was the point :)
20:55:19JdGordon_I had that on my todo list, coulnt remember if it was you or AlexP who requested it, so just did it and hoped you'd notice :)
20:55:39AlexPNot me :)
20:55:47***Saving seen data "./dancer.seen"
20:55:55AlexPJdGordon_: I've attached the comment, screenshot and patch
20:56:00JdGordon_ta
20:56:56AlexPJdGordon_: The patch is against a tree with the fm patch, so the line numbers will be out
20:57:50JdGordon_arr :) I meant the full diff, but thats fine too
20:57:51funmanFlynDice:
20:57:57funmanhttp://pastie.org/829514 (for csd)
20:57:58JdGordon_I'll unmerge it and commit it seperately
20:58:31funmanSchmogel: if you decide to try mkamsboot, could you use this patch too? we'd have 2GB 4GB and 8GB
20:59:15AlexPJdGordon_: Full diff with the fmpatch?
20:59:26AlexPI can do that too - I thought you'd prefer this one :)
20:59:27JdGordon_yeah, dont worry though
20:59:32JdGordon_this one is better
20:59:44AlexPOK, it'll take 30 seconds if you want the other :
20:59:44AlexP)
20:59:46Schmogelok. but first i have to understand how to work with mkamsboot
21:00
21:01:27funmanyou can look at README, "boot file" is bootloader-clipplus.sansa which you have to build from svn
21:01:38*pixelma is somehow reminded of the problems with %?St and bass and treble info
21:02:20funmanSchmogel: i can send you the bootloader, do you have a svn checkout and gcc to build mkamsboot from svn ? (released versions won't work for clip+)
21:03:36Schmogelno, i don't think so :D you know, im new to this ;)
21:04:12funmanwhich OS do you run ?
21:04:30Schmogelwin 7 and ubuntu 9.10
21:04:37 Part watto
21:04:48funmanamd64 ?
21:05:06Schmogelyes, windows
21:05:16funmanand ubuntu ?
21:05:37Schmogelx86
21:05:45JdGordon_pixelma: whats that problem? unless its in the tracker its unlikely anyone else remembers
21:06:17 Quit kugel__ (Quit: exit(0);)
21:08:38funmanSchmogel: give me your email (in private) and i'll send you a bootloader and mkamsboot binary for ubuntu 32 bits
21:08:56Schmogelok, thx
21:10:42pixelmaJdGordon_: I couldn't make it work on my M5 (I wanted to make a WPS for use with my home stereo, which would notify me when bass/treble were not 0). Tried that with conditionals (it's a lot values but I think I did it correctly) and seem to remember that I found a comment in the code saying something like "assume all possible values are 0 or greater than 0" which is not true for bass/treble
21:10:55pixelmaforgot to put it on the tracker :\
21:11:23ThomasAHfunman: ok, I have bootloader-clipplus.sansa and mkamsboot from current svn and OF as clppa.bin in ~/rbinstall ... what now?
21:12:32funman./mkamsboot clppa-original.bin bootloader-clipplus.sansa clppa-patched.bin
21:12:42funmancopy clppa-patched.bin to Clip+, rename as clppa.bin and eject
21:15:16ThomasAHcompleted ...
21:15:54ThomasAHdmg -> 2000 6d06
21:16:33funmancool
21:16:39funman(should be dma)
21:17:12funmanThomasAH: can you apply http://pastie.org/829514 and rebuild bootloader ? what capacity is your clip+ ?
21:17:16ThomasAHoh, yes ... now Home key?
21:17:26funmanhome or left
21:17:34funmanyou need to keep pressing power to power it off first
21:18:10ThomasAHfunman: 8GB
21:18:39ThomasAHfunman: when should I press home or left? together with power for power-on?
21:19:05ThomasAHah, works
21:20:05funmanhm the CSD reports 3584MB but OF says 3744MB
21:20:21*ThomasAH tries to apply patch single-handed ... baby on shoulder
21:20:42funman^^
21:21:01funmanteaching the baby how to patch ? :P
21:21:20TheSeven3584MiB = 3758MB
21:21:28 Quit anewuser (Ping timeout: 240 seconds)
21:21:29TheSeven14MB gone?
21:21:52 Quit Frampis_ (Ping timeout: 240 seconds)
21:22:22funmanthere are 30MB reserved for the firmware
21:22:55funmanbut what i see is 3584MB (not MiB), i'll double check my calculation
21:22:56 Join anewuser [0] (anewuser@unaffiliated/anewuser)
21:23:04 Quit Guest38414 (Ping timeout: 256 seconds)
21:23:23 Join Biont [0] (~4dbd813d@giant.haxx.se)
21:23:25*TheSeven just noticed that the page footer of the nano2g manual just says "iPod Nano"!
21:24:32funmanTheSeven: hm but the OF could well remove 14MB for its database
21:25:13ThomasAHfunman: flashing ...
21:25:22funman3933732864 bytes according to dmesg or 3751.5 MB
21:27:40ThomasAHfunman: 0x0a4040af 03b377f80 0x5b590000 0x400e0032 (not double checked!)
21:28:21funmanThomasAH: looks correct at first sight
21:28:25BiontI converted a new font to use in my theme. It's a free font with lots of line spacing added (by me) to improve touchscreen usability. Now I'm confused about how to name it properly. The setting I used in the converter was 6, but the font itself is more like a 17 in size while all the line spacing would add up to a font size somewhere about 25+. What size would be correct?
21:28:58JdGordon_pixel height
21:28:59ThomasAHfunman: now double checked
21:29:16Biontalright, thank you
21:29:36funmanThomasAH: can you tell me how much sectors are reported in dmesg ? [31046.481823] sd 13:0:0:0: [sdf] 7683072 512-byte logical blocks: (3.93 GB/3.66 GiB)
21:29:41pixelmaBiont: including spacing (so probably the 25+ but nore accurate)
21:30:00Bionthm....
21:30:25ThomasAHfunman: [54379.183860] sd 5:0:0:0: [sdb] 15462400 512-byte logical blocks: (7.91 GB/7.37 GiB)
21:30:31pixelmaso that people can easily chose a different one with the same height (line height)
21:30:53funmanhum i think i just have an overflow ..
21:30:54BiontI was thinking it might be a good idea to name it "17_t_XYZ.fnt", the "t" indicating the linespacing (t->touchscreen)
21:31:11BiontBut I wanted to ask before doing so :)
21:31:23JdGordon_Biont: no, well, the number shuold be the line height which would be 25 or so
21:31:40JdGordon_it's expected that any size 17 font fits in the same theme
21:31:57Biontalright, I understand
21:32:51BiontThank you, I'm almost ready to commit my new theme now. JdGordon, thanks again for compiling the simulator for me :) I haven't been lazy
21:33:44 Quit fyrestorm (Quit: lamers envy me like they envy bill g -- main boot xp, just the way it should be!)
21:34:20ThomasAHfunman: you or the numbers? (information or integer overflow?)
21:34:43BiontI assume the line selector height equals the total font height, so it would be 24 then
21:34:50funmanThomasAH: my code, when intepreting the numbers
21:35:01ThomasAHah :)
21:35:05funmanyour Clip+ is presumably 14548480MB
21:35:30*ThomasAH has an 8GB card inserted
21:35:42 Join Tomis2 [0] (~Tomis@70.134.64.213)
21:35:46ThomasAHshould I eject and repeat?
21:35:46funmancard is untouched atm
21:36:20ThomasAHoh, MB, not KB ... 14548480MB would be far too much :)
21:36:49funmanin KB as well
21:36:49AlexPI want a 14 TB clip+!
21:37:05ThomasAHAlexP: clip++ :)
21:37:49 Join froggyman [0] (~sopgenort@pool-72-69-210-48.chi01.dsl-w.verizon.net)
21:38:56 Quit Tomis (Ping timeout: 248 seconds)
21:38:57 Nick Tomis2 is now known as Tomis (~Tomis@70.134.64.213)
21:40:41 Quit Farthen (Read error: Connection reset by peer)
21:40:43 Join feisar [0] (~jljhook@irkki.fi)
21:41:12 Nick feisar is now known as Guest82304 (~jljhook@irkki.fi)
21:41:22funmanSchmogel: do you experience the bug of https://bugs.launchpad.net/ubuntu/+source/libgphoto2/+bug/355998 ?
21:41:23 Join Frampis [0] (~frampis@noppakerho.com)
21:43:48 Part Anaxite ("No matter how dark the night, somehow the Sun rises once again")
21:44:44funmanC_SIZE is 5.3 bigger on 8GB than on 4GB (and incredibly too high on both)
21:44:45Schmogelnope. it just didnt do anything in mtp-mode. but in msc ubuntu mounted the clip+
21:44:50 Join Zagor [0] (~bjst@46.35.227.87.static.tab.siw.siwnet.net)
21:44:50 Quit Zagor (Changing host)
21:44:50 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
21:45:49funmanbetween 4GB & 8GB only C_SIZE and CRC change
21:46:24CIA-8New commit by 03Buschel (r24732): Allow batteries with higher capacity for iPod Mini1G, Mini2G and Video (FS #8973).
21:48:17ThomasAHfunman: any known issues when connecting the turned-off clip+ to USB port? Or should I avoid that for know?
21:48:27ThomasAH-k
21:48:34funmanThomasAH: if you connect it to USB it will just power on and run rockbox
21:50:06funmanI don't get the CRC7 definition in chapter 4.5 of SD spec: G(x) and M(x) are defined in function of x and of the "bit string" to check, but I don't know what 'x' is
21:50:58 Quit moos (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
21:51:41 Quit Frampis (Quit: Heippa.)
21:51:55 Quit Guest25299 (Read error: Connection reset by peer)
21:52:31 Join Lss__ [0] (~Lss@cm48.omega219.maxonline.com.sg)
21:52:44 Join Frampis [0] (famas@noppakerho.com)
21:54:58 Join stooo [0] (~sto@g226199185.adsl.alicedsl.de)
22:00
22:01:35CIA-8New commit by 03tomers (r24733): Touchscreen: Yet another WPS keymap tweak ...
22:02:04funmanLinux has a crc7 implementation (for bytes). http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=lib/crc7.c;h=f1c3a144cec123d73bee9897922bca0fa4dace48;hb=HEAD#l13 gives x^7+x^3+1 = 0x0 for index 0, but 0^7+0^3+1 == 1 for me ..
22:13:31tomerscan some suggest a console tool in Ubuntu that dumps id3/4?
22:13:57funmaneyeD3 (dunno about id4)
22:14:19funmanif you mean id3v2, it'll dump those as well
22:14:40tomersfunman: thanks! installing (it's in the repository)
22:20:19 Join stripwax [0] (~Miranda@87-194-34-169.bethere.co.uk)
22:21:29 Join panni_ [0] (hannes@ip-95-222-52-93.unitymediagroup.de)
22:22:36tomersI get a crash when going to id3 screen while playing a certain file
22:22:55tomersI am now opening a flyspray item for this. What severity should it be?
22:23:11 Quit Strife89 (Ping timeout: 264 seconds)
22:23:52tomersAlso, can I see id3 properties of this file from the file browser (not from WPS)?
22:24:07tomersI couldn't find anything that matches in Open With
22:24:48funmantomers: properties
22:25:11tomersproperties works - no crash
22:27:58 Join Luca_S [0] (~57064cb7@giant.haxx.se)
22:28:40 Join Farthen [0] (~chatzilla@e179233161.adsl.alicedsl.de)
22:28:49 Join fyrestorm [0] (~nnscript@cpe-24-90-81-175.nyc.res.rr.com)
22:34:57tomersFS #11033 - Rockbox crash when displaying ID3 info
22:35:51 Quit stooo (Ping timeout: 260 seconds)
22:36:04funmanafaict the CRC in CSD are wrong
22:37:26 Join GeekShad0w [0] (~Antoine@91.189.204-77.rev.gaoland.net)
22:38:42 Quit pamaury (Quit: abort();)
22:39:54 Quit GeekShadow (Ping timeout: 246 seconds)
22:40:01stripwaxsaratoga - for max N=8192 size mdct (=real), we use an FFT of size n/4 (complex) = 2048 complex pairs - so I can't help thinking that 1) max revtab should be 1<<11 not 1<<12 and/or 2) we're (I'm) setting revtab_shift incorrectly since if the table is indeed 1<<12 then we should be shifting by 13-nbits not 14-nbits.
22:40:23stripwaxBut if the code is actually wrong I don't understand why/how it doesn't affect the audio
22:42:28funman2 wrongs make one right ?
22:43:06JdGordon_tomers: using multifont at all?
22:45:07 Quit Bullet` (Ping timeout: 276 seconds)
22:45:39tomersJdGordon: not sure what you mean, but it's clean build with default settings
22:46:02 Join stooo1 [0] (~sto@g226199185.adsl.alicedsl.de)
22:46:42JdGordon_the location of the crash is worrying
22:47:09tomersI know
22:47:17 Part stooo1
22:47:51JdGordon_is fcache->_lru == NULL?
22:50:18tomersno
22:50:40JdGordon_index ?
22:51:34JdGordon_I'll investigate this evening if I get a chance. That code wasnt touched recently
22:51:37JdGordon_which font?
22:51:47tomershttp://pastie.org/829752
22:51:57funmannow i'm at 5375MB for 4GB and 28415MB for 8GB
22:52:07tomersdefault font. i think helvetica 14...
22:52:08 Quit Luca_S (Quit: CGI:IRC)
22:52:57JdGordon__size = 7536757, _capacity = 7536672, _index = 0x700075} looks wrong
22:54:52 Quit Zagor (Quit: Clint excited)
22:55:06saratogastripwax: it does sound like its a factor of 2 too big
22:55:12saratogawould you like me to post a smaller table?
22:55:48***Saving seen data "./dancer.seen"
22:56:37stripwaxi'm still trying to understand it actually.. something seems off somewhere but not in a way that affects decoding at all (very odd). I actually hope/assume I'm just forgetting how it fits together (currently reading the original ffmpeg sources to compare)
22:57:35 Join Hillshum [0] (~hillshum@75-165-233-186.slkc.qwest.net)
22:57:42 Join einhirn [0] (~Miranda@p548588AC.dip0.t-ipconnect.de)
22:58:05 Join axon [0] (~masmos@141.30.30.31)
23:00
23:00:26axonhello out there - for more than two hours i read some files in your wiki - an i remarked that my english is not as good as i thought - finaly i have one question: Can i install rockbox (or any other linux distribution) on a SilverCrest DM-67?
23:00:41 Join Bullet` [0] (~Bullet@AMontsouris-159-1-33-45.w90-46.abo.wanadoo.fr)
23:00:54saratogano
23:01:03Hillshumaxon: Not Rockbox. Rockbox isn't Linux either
23:01:05tomersaxon: No. If it's not listed in the front page, it's not supported
23:01:46axonsaragtoga , Hillshum, tomers: why not?
23:02:04stripwaxaxon - please read the FAQ
23:02:11axonread it
23:02:22axonalready
23:02:40Bagderaxon: every device is unique, every port is especially crafted to work on that particualr device
23:03:07Bagdernobody has done the work on "SilverCrest DM-67" (which I've never even heard of in the past)
23:03:12saratogasame reason you can't install Windows on your toaster, software only runs on computers its made for
23:03:41 Quit tomers (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20100106054534])
23:03:46axonso - is there any possibilty to adapt the software?
23:03:53Bagdersure
23:03:55stripwaxaxon - you can try, if you have the skills.
23:03:56Stephen__Bagder, http://www.mysilvercrest.de/en/artikel.php?a=141
23:04:02saratogasee the NewPort wiki page
23:04:12BagderStephen__: yes, I found that too
23:04:21saratogastripwax: I looked at the new ffmpeg AAC-HE decoder
23:04:42saratogaits got a function nearly identical to libfaad in overall purpose, but based around an MDCT instead of a strange filterbank in libfaad
23:04:49saratogaits also extremely slow looking in libfaad . . .
23:05:06funmanin fact the CSD are correct (you should see 15462400 512-byte sectors in dmesg on 8GB)
23:05:06saratoganearly identical to a function in libfaad, not the whole codec
23:05:37stripwaxsaratoga - gotcha. I don't really know anything about faad though, maybe one day ..
23:06:32stripwaxaxon - also read this (which is linked, by the way, from the Third question in the FAQ) http://www.rockbox.org/wiki/TargetStatus "If you truly want to see Rockbox happening on a new platform, you should start collecting as much info as possible about it with great detail. Post the info in a wiki page or the forum thread related to the device and start figuring out what similarities the new player has with existing supported p
23:06:32stripwaxlayers. Hopefully you and your friends will get enough info about it to make someone with clues possible to write code for it and eventually start porting Rockbox."
23:06:33saratogastripwax: AAC-HE is basically a subband decomposition thats run before encoding to AAC, so its got a filterbank that in principle could be phrased as an MDCT
23:07:44 Quit grndslm (Read error: Connection reset by peer)
23:07:49saratogaAAC-HE is actually somewhat similar to ATRAC3 in that regard
23:10:09 Join Xerion_ [0] (~xerion@82-170-197-160.ip.telfort.nl)
23:12:31Unhelpfulsaratoga: how likely it can be simply dropped in?
23:12:50saratogaUnhelpful: the function is pretty short, so hopefully pretty likely :)
23:12:58 Quit Xerion (Ping timeout: 272 seconds)
23:12:59 Nick Xerion_ is now known as Xerion (~xerion@82-170-197-160.ip.telfort.nl)
23:13:22axonstripwax - in fact i don't no really what you mean with skilled - so i'm getting the impression that i'm not - surely i can program in c and c++ a little bit, even in java or php but never did something like that - so in fact i have to know which hardware is used in this drive ?! how can i find this out? is there any description ?
23:13:30axonany manual or how to?
23:13:31saratogaunhelpful: http://pastebin.com/m34952ad8
23:13:51saratogaaxon: did you read the wiki page I suggested?
23:14:08stripwaxaxon - read the NewPort page, on the wiki, as saratoga suggested
23:14:51saratogaUnhelpful: oh huh i mixed up analysis and synthesis, so i'm not sure where it goes in libfaad
23:14:52stripwaxaxon - but basically, you need to identify all of the hardware inside your device, how it communicates with each other, and how to program the relevant device drivers.
23:15:10saratogabut even so the analysis and synthesis filterbanks look quite simple in ffmpeg, so maybe they both can be moved over
23:15:23saratogaand we could replace libfaad piece by piece until it doesn't suck
23:18:32saratogaoh ffmpeg's sbr_qmf_synthesis probably goes with libfaad's sbr_qmf_synthesis_64
23:19:35 Quit ecio (Ping timeout: 264 seconds)
23:19:41Unhelpfuland then we just have the tricky job of figuring out whether we've hurt quality or not... much easier when you can check w/ md5sum :/
23:19:48 Join CptJesus [0] (~60ea966d@gateway/web/freenode/x-fcwqpxudmbawbjfa)
23:19:58CptJesusHi everyone
23:20:18CptJesusAre any rbutil devs online currently?
23:20:20stripwaxsaratoga - (re ffmpeg/mdct). maybe i'm just going crazy. so N=8192 => mdct nbits=13 . fft initialised (back when it had init) to nbits-2 => fft_nbits=11 (for size fftn=2048). So bitrev_shift = 14-13 = 1. And the revtab is 1<<12. so we shift by 1 bit, and interpret the revtab as being 1<<11 = 2048. so maybe it does just work
23:20:53 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
23:21:17*stripwax should know since he wrote that bit anyway
23:21:23*stripwax doesn't really remember though
23:21:35 Join bluebrother [0] (~dom@f053153127.adsl.alicedsl.de)
23:21:36 Quit bluebrother (Changing host)
23:21:36 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
23:21:44saratogaso we're still 2x too big
23:21:56stripwaxAnyway, long story short, revtab probably a factor of two bigger than it needs to be and (at the same time) we should change bitrev shift to be 13-nbits rather than 14-nbits
23:21:57stripwaxyeo
23:21:59stripwax^yep
23:22:01CptJesusI was in here yesterday inquiring about an issue, and I was told to try and talk to a rbutil dev today
23:22:08stripwaxUnless we want to support *FFT* of size (complex) 4096?
23:22:34saratogai don't see any reason to
23:22:42domonokyCptJesus: yep, just ask
23:23:00stripwaxthinking maybe the fft plugin (which doesn't use codeclib's fft anyway but ... could, maybe?)
23:23:04saratogaif we need bigger ffts for plugins, we can just change the fft header file to let you provide your own bitrev table
23:23:11stripwaxsaratoga - ok cool, good idea.
23:23:50saratogaif 2048 only need a 512 table, and all codecs essentially only use that or smaller 99.999% of the time, we should probably think about putting it in IRAM
23:24:09CptJesusdomonoky: Alright, I've been trying to install Rockbox on my iPod video 30gb, and for some reason, it is completely unable to install the bootloader. I asked a few people yesterday, and none of them could narrow down the problem. Neither iPodpatcher.exe or rbutil is able to install the bootloader. I posted the system trace from my rbutil yesterday if you need it
23:24:11saratogaits a shame theres no easy way to do that neatly . . .
23:24:56 Join grndslm [0] (~grndslm@174-126-14-4.cpe.cableone.net)
23:24:59domonokyCptJesus: can you pastebin the system-trace, after you tried the bootloader install ?
23:25:19CptJesusdomonoky: Yep, will do right now
23:25:25 Quit bluebroth3r (Ping timeout: 276 seconds)
23:26:31 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
23:27:33 Quit linuxstb (Read error: Connection reset by peer)
23:27:35CptJesusdomonoky: http://pastebin.com/m12f76224
23:27:53 Join Casainho [0] (~chatzilla@87-196-154-59.net.novis.pt)
23:28:04 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
23:28:12stripwaxsaratoga - we would have to have two copies, one sized for 512 in iram, and one sized for 2048 in non-iram [except for targets that can fit the 2048 in iram? Is 4KB too much to fit?]
23:28:35 Quit anewuser (Quit: http://xrl.us/WinterChipV =ooo ϢINTER ϾHIP 5iVE is OOON!!)
23:28:56stripwaxactually I was just about to try building for coldfire with revtab in ICONST_DATA to see what would happen [dunno if h120 has enough space though in iram]
23:29:36domonokyCptJesus: hm, unfortunatly no output from ipodpatcher code (rbutil just uses ipodpatchers code for ipod installs)
23:29:47domonokyCptJesus: but i am just reading the logs..
23:29:55Unhelpfulstripwax: why not have it in or out of iram with two different names, and pick one w/ a define before including the header?
23:30:21stripwaxUnhelpful - we actually want the 'most commonly used' on in iram , unless the whole table fits in iram.
23:30:25 Join ecio [0] (~ecio@32.135.232.200)
23:30:29stripwaxthe 'most commonly used' one is a smaller table.
23:30:44stripwaxyou can transform the 'whole table' into its subset easily enough
23:30:57Unhelpfulstripwax: right, but i'm saying you could ask for the larger one in iram via an include...
23:30:57CptJesusdomonoky: iPodpatcher.exe gives me a write failed error. I detailed it yesterday, but I can do it again if it would help
23:31:08domonokyi am just reading it :-)
23:31:15stripwaxsaratoga - yeah, looks like the sign of output is inverted compared to oggdec (what I was fearing yesterday)
23:31:45saratogawell thats easy to fix then
23:32:32domonokyCptJesus: but if ipodpatcher also fails, i think there is nothing rbutil could do. I suspect a issue with windows 7, do you have access to a Pc with a different OS ?
23:33:02 Join Meddler [0] (~5d60801a@giant.haxx.se)
23:33:14stripwaxsaratoga - yep (but dunno where exactly - is the FFT broken or the MDCT pre/post/symmetry bit)
23:33:17Unhelpfulerm, via a define-before-include. possibly. :)
23:33:18stripwaxI'll assume the latter
23:33:32CptJesusdomonoky: I have access to a computer with Windows Vista
23:33:40CptJesusdomonoky: Should I give that shot?
23:33:47BiontHm, I'm trying to upload my theme, but it says it didn't pass checkwps. apparently, there's something wrong with my sbs, but everything works perfectly in the simulator. At least I couldn't find anything wrong with it. the debug console also doesn't complain. help?
23:33:57domonokyCptJesus: would be worth to test.
23:34:00stripwaxUnhelpful - and include it twice? on a target with only a small amount of iram, we want the 'small' table in iram, and the 'whole' table in dram (just in case it is needed, even though it's not commonly needed)
23:34:07CptJesusdomonoky: Be back shortly
23:34:14Biont"Checking .rockbox/wps/Breeze.sbs... WPS parsing failure " <&minus;&minus;-this is what the theme site says
23:34:31stripwaxit uses more memory but at least you're guaranteed to have the 99.99% case data in iram. and you only use the non-iram table in the remaining 0.01% of cases
23:34:35Unhelpfulstripwax: right, but i'm saying that if a plugin or something wanted only the "big" table....
23:35:33stripwaxUnhelpful - true. a bunch of codecs will only need the 'small' table. probably only vorbis needs the big table (and even then the 'small' table is sufficient in 99% of cases and probably 100% of common portable audio cases)
23:35:41domonokyBiont: this meant it didnt pass checkwps on any of the possible targets. *i should really try to improve the upload page, the checkwps output should be shown too *
23:36:42 Quit Meddler (Client Quit)
23:37:11gevaertsBiont: do you have the theme somewhere? If you don't have a recent checkwps, someone else could run it for you
23:37:26Biontdomonoky: So what can I do?
23:37:31saratogaUnhelpful: why is bs_log2_tab in memory for WMA?
23:37:53Biontgevaerts: I can upload it somewhere else, yes
23:38:33 Quit CGL (Quit: Saliendo)
23:38:47Unhelpfulsaratoga: because it's using log2 somewhere... i actually found clz to be faster with a function in dram and table-follows-function than using a much smaller inline code + iram table. the same would apply to log2, it would just need a different table. ;)
23:38:56CptJesusdomonoky: Exact same thing. No difference
23:39:10saratogais there someway to get the DRAM only version?
23:39:22saratogalog2 performance is unimportant
23:39:26CptJesusdomonoky: Am I missing something here? Am I supposed to do something to the iPod before installation that isn't specifically in the log?
23:39:28Biontgevaerts: I hope this not some stupid mistake of me :/
23:39:43domonokyCptJesus: no, it should just work.
23:39:47 Quit Lss__ (Read error: Connection reset by peer)
23:39:55Biontalright, here's the theme http://www.mediafire.com/?nxmzmzzlfyj
23:40:04Unhelpfulsaratoga: and faster with dram anyway... i was going to have a second copy of the support-arm.S clz, with the table modified for log2, and use that to handle log2 on armv4
23:40:18Unhelpfuli've just not gotten it done yet
23:40:48CptJesusdomonoky: Well in that case, I can't imagine what the problem is. I can try and get access to a Windows XP machine if you think that might help
23:41:36Biontbefore someone wastes his time: could this have to do with multifont? do I need to supply all fonts used by the theme? I only used standard fonts except my custom font which is included in the theme
23:42:09domonokyCptJesus: i dont know if it helps. you could try a itunes restore, to exclude any hd problems.
23:43:32domonokyBiont: you would get a different message if the fonts or other files are the problem.
23:43:51CptJesusdomonoky: How would I go about that?
23:44:21 Quit axon (Quit: Ex-Chat)
23:44:34domonokyCptJesus: use itunes to restore to factory state, and then try again. *dont know if it helps, but could be worth a try*
23:44:44CptJesusdomonoky: kk, I'll give it a try
23:44:49FlynDicefunman: 2GB clip+ 928040DF FEFBCFFF 5F5A83AE 00260032
23:45:37Biontdomonoky: Yes I noticed that when uploading the previous version of the theme (It's a very comfortable feature btw, awesome work). I just thought something like that might be the problem, since the only difference from my side is the usage of multifont
23:45:54funmanFlynDice: dmesg says 3800064 512-byte sectors ?
23:46:31domonokyBiont: perhaps checkwps has problems with multi font tags ?
23:46:59 Quit Bullet` (Quit: Bye les gens =))
23:47:00funman2GB uses version 0 CSD (where maximal codec capacity is exactly 2GB)
23:47:14CptJesusdomonoky: That did it!
23:47:15domonokyBiont: if you can compile, you could try to compile the correct checkwps yourself, and see what it says.
23:47:35stripwaxsaratoga - oh wait. I think the sign is missing in the ifft actually. hm. can you see if liba52 has the same sign inversion problem?
23:47:39CptJesusdomonoky: Not sure what the issue was, but its installing perfectly now. Thanks a ton for the help
23:48:12domonokygood :-)
23:48:16gevaertsBiont: it's definitely related to the fonts. Commenting the %Fl tags out makes it parse
23:48:25Biontdomonoky: Nope, sorry. I'm even too studid to compile a simulator myself.
23:48:37funmanbtw clipv2 & clip+ respect the "always 1" flag, unlike fuzev1/clipv1
23:48:45Biontand I'm too studid to spell stupid
23:48:54Biontgevaerts: hm...okay
23:49:04Biontgevaerts: Guess I'm too early^^
23:49:28gevaertsand checkwps doesn't give motr information
23:50:14*domonoky pokes JdGordon with this issue..
23:51:39stripwaxsaratoga - yeah, ffmpeg fft flips the sign of the sin table for Inverse. obviously we use a single trig table (with positive-sign sin) so it's kinda geared for the forward fft (oops). I must have missed that when I put in the static trig for fft. but would be good if you can confirm liba52 is also sign flipped
23:52:07 Quit evilnick_B (Quit: Page closed)
23:52:27Stephen__i've noticed problems with the ipodvideo sim and multi-font, it crashes upon restart. works on target tho. could be the them as deleating config.cfg works to bring it back but it's only on a theme with multifont this is caused
23:53:21 Part domonoky
23:54:11Biontoh yeah, my D2 sim also behaves strangely after bringing in multifont
23:54:47BiontIt wasn't annoying enough to bring it up here, but if we're at it anyway...
23:56:35Stephen__is it when a multifont theme is selected then you quit the sim adn try to start up again it'll crash ?
23:57:45gevaertsok, I think I found why checkwps fails
23:58:03BiontIt happens sometimes when I reload the theme.cfg from within the simulator. It will fail to load most .bmps, but it crashed only once so far (and for other reasons, I think)
23:58:26 Quit CptJesus (Quit: Page closed)

Previous day | Next day