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

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

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

#rockbox log for 2010-07-15

00:00:09 Join Topy [0] (~topy@2a01:608:ffff:2627:2e0:4dff:fe83:a45e)
00:00:36 Quit fml2 (Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100628124739])
00:05:58 Quit captainkewlll (Quit: Page closed)
00:10:19 Quit evilnick_B (Quit: Page closed)
00:11:20 Quit krazykit (Quit: Connection reset by beer)
00:11:45 Join FlynDice [0] (~AndChat@12.130.106.87)
00:12:03 Quit p3tur (Quit: here today, gone tomorrow)
00:14:19 Quit petur (Quit: Zzzzz)
00:16:49 Part FlynDice ("Leaving")
00:19:21 Join Scofield [0] (www-data@giant.haxx.se)
00:19:32Scofieldhi
00:19:55 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
00:29:46 Quit Scofield (Quit: CGI:IRC)
00:31:38 Quit advcomp2019_ (Read error: Connection reset by peer)
00:32:15 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
00:32:54 Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
00:35:53 Part watto
00:40:16 Quit robin0800 (Ping timeout: 276 seconds)
00:44:52 Join krazykit [0] (~kkit@206.183.185.8)
00:47:03 Quit kugel (Remote host closed the connection)
00:58:21 Quit JdGordon (Ping timeout: 264 seconds)
01:00
01:02:49 Quit bertrik (Quit: :tiuQ)
01:04:19bieberJdGordon1: It's definitely crashing the theme editor, just not sure quite why yet
01:07:26bieberIf you do a text search in the project for atof it will find you the places where I was using decimal values, and you can fix those by replacing the atof line with just param[index].data.number
01:07:53bieberHowever, if you open up, say retroTape's WPS file there's a line that will no longer parse due to changes in the parser
01:08:39bieberMy guess is that it's miscounting the sublines, but I'll take a closer look at it tonight
01:08:46 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
01:09:31 Join kugel [0] (~AndChat@g231106195.adsl.alicedsl.de)
01:10:52bieberMake that param[index].data.number / 10.0
01:11:35 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
01:13:35bieberThe code that scans a decimal value looks a little weird, if you haven't been around when I get back I should be able to fix it
01:14:16 Join yelped [0] (www-data@giant.haxx.se)
01:14:48yelpedAnyone want to take a lokk, pleases? http://forums.rockbox.org/index.php?topic=25267.0
01:15:02yelped*look.
01:16:19gevaertssaratoga: don't forget the midterm evaliation!
01:16:33gevaertskugel: same for you :)
01:18:08 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
01:21:23 Quit advcomp2019 (Ping timeout: 265 seconds)
01:25:46 Quit Topy (Ping timeout: 276 seconds)
01:31:10 Join Topy [0] (~topy@2a01:608:ffff:2627:2e0:4dff:fe83:a45e)
01:34:31 Quit kugel (Read error: Connection reset by peer)
01:35:25mc2739yelped: check your forum thread
01:38:41yelpedYou should check it, too! :)
01:45:14yelpedGreat job! Thanks!
01:46:48 Quit yelped (Quit: CGI:IRC (EOF))
01:47:37 Quit bunnyboi (Ping timeout: 240 seconds)
01:48:39 Join JdGordon [0] (3a601b9e@gateway/web/freenode/ip.58.96.27.158)
01:48:49 Quit n1s (Quit: Lämnar)
01:50:35 Join bunnyboi [0] (~androgyne@cpe-72-224-19-1.nycap.res.rr.com)
01:57:13 Join JdGord [0] (3a601b9e@gateway/web/freenode/ip.58.96.27.158)
01:58:29 Join binaryhermit [0] (~binaryher@adsl-99-141-196-197.dsl.emhril.sbcglobal.net)
01:59:24 Quit GeekShadow (Quit: The cake is a lie !)
01:59:57***Saving seen data "./dancer.seen"
02:00
02:15:07 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.111)
02:15:22 Quit [Saint] (Ping timeout: 240 seconds)
02:15:40 Join fredreichbier [0] (~fred@unaffiliated/fredreichbier)
02:16:34 Quit DerPapst (Quit: Leaving.)
02:18:06fredreichbierhey there. i'm using rockbox on a sansa clip+. it's awesome! :D however, i don't know how to activate "hold". the sansa clip+ doesn't have a hold button anymore, with the original firmware, i need to press the "home" button for some seconds. can anyone give me a hint? :)
02:18:26JdGordread the manual
02:18:37JdGordI think its hold+play or something like that
02:18:54 Quit mt (Read error: Connection reset by peer)
02:19:29 Join [Saint] [0] (S_a_i_n_t@203.184.0.124)
02:19:52mc2739fredreichbier: try home and select
02:19:56 Quit S_a_i_n_t (Ping timeout: 245 seconds)
02:22:06fredreichbierah
02:23:02fredreichbierJdGord, mc2739: thanks for your help, home+select works! the manual uses an old clip scheme picture (which had a hold switch), so i got a bit confused =)
02:23:48JdGordyou didnt click the clip or clipv2 manual instead of clip+ did you?
02:24:56fredreichbierJdGord: i used this one: http://download.rockbox.org/daily/manual/rockbox-sansaclipplus/rockbox-buildch3.html#x5-240003
02:27:30JdGordah ok.. looks like a bug for pixelma then :)
02:28:23mc2739it does use the clip images and I do not find any mention of the soft-hold
02:31:22fredreichbieryeah, the soft-hold would be useful. the image issue isn't too urgent i guess, though
02:31:51fredreichbier(all in all, it's a really cool experience to run doom on a two-color 1cm x 2cm display. :D)
02:35:56mc2739FS #11480 created for clip+ manual problem
02:37:15JdGordfredreichbier: cool isnt the word I'd use... :)
02:41:51fredreichbierhehe
02:46:34fredreichbieri'm just asking myself why i didn't afford the extra 20€ to get a fuze instead ;)
02:53:58fredreichbierhowever, i'm going to bed now. thanks for the help and keep up the great work!
03:00
03:05:42 Quit efyx (Remote host closed the connection)
03:11:20 Join davharster [0] (~hartmand@ool-44c117b9.dyn.optonline.net)
03:20:38 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
03:22:22 Quit [Saint] (Read error: Connection reset by peer)
03:32:04 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.173)
03:34:51 Join robin0800 [0] (~quassel@149.254.61.169)
03:43:06 Quit davharster (Quit: Leaving.)
03:43:16 Quit robin0800 (Remote host closed the connection)
03:47:07 Part flyback ("Leaving")
03:54:45 Quit S_a_i_n_t (Ping timeout: 264 seconds)
03:55:20 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.2.34)
03:59:58***Saving seen data "./dancer.seen"
04:00
04:07:23 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:07:23 Quit pixelma (Disconnected by services)
04:07:25 Quit amiconn (Disconnected by services)
04:07:27 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:07:38 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:07:47 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:11:04 Join itmandrea [0] (~irc@ip68-98-28-195.ph.ph.cox.net)
04:13:49 Quit scorche (Disconnected by services)
04:13:59 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
04:17:18 Quit S_a_i_n_t (Ping timeout: 240 seconds)
04:18:22 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.207)
04:22:59 Quit evilnick (Ping timeout: 248 seconds)
04:23:07 Join evilnick- [0] (~evilnick@2001:0:4137:9e76:877:19f1:b5a6:f655)
04:23:24 Nick evilnick- is now known as evilnick (~evilnick@2001:0:4137:9e76:877:19f1:b5a6:f655)
04:25:57 Join T44 [0] (~Topy44@f054172050.adsl.alicedsl.de)
04:26:53 Quit TheSeven (Ping timeout: 240 seconds)
04:26:53 Quit Topy44 (Ping timeout: 265 seconds)
04:27:47 Quit S_a_i_n_t (Ping timeout: 260 seconds)
04:28:19 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.129)
04:29:25 Quit Zarggg (Quit: Zarggg)
04:30:10 Quit Topy (Ping timeout: 252 seconds)
04:31:44 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:32:10bieberJdGord: Did you get my earlier comments about your patch?
04:32:24JdGordI havnt had a chnace to look through the logs yet
04:32:26JdGordso no :)
04:33:00 Join Topy44 [0] (~topy@2a01:608:ffff:2627:2e0:4dff:fe83:a45e)
04:33:06bieberOkay
04:33:18JdGordI'll have a look now
04:33:34bieberThe short version is that it crashes the theme editor (which is an easy fix) and it messes up parsing somewhat (which shouldn't be a hard fix either)
04:35:16JdGordit shouldnt mess up the parser :) I copied that code out of the old parser
04:35:40JdGordI assume the ediotr crashes because it doesnt handle the DECIMAL case?
04:36:26bieberNo, that's just a missing feature
04:37:10bieberThe crash comes when I was interpreting decimal values, because I'm using atof on the text. Just do a text search for atof and you'll find the three lines that need fixing to make it stop crashing
04:38:24JdGordok, I'm at work so wont be able to touch it for 6 hours or so
04:39:01bieberOkay. I'm going to be coding in a couple hours, so I'll see if I can get it to work: it's probably something small like advancing the cursor too far after scanning a decimal value
04:39:17bieberWant me to commit the patch for you when it works?
04:40:23 Part itmandrea
04:43:19JdGordyes please
04:47:31 Join [Saint] [0] (S_a_i_n_t@203.184.1.54)
04:47:48 Quit JdGord (Quit: Page closed)
04:48:23 Quit S_a_i_n_t (Ping timeout: 260 seconds)
04:51:58 Quit t0rc (Quit: Leaving)
04:54:53 Quit Barahir (Ping timeout: 240 seconds)
04:56:51 Join Barahir [0] (~jonathan@frnk-590fd0e0.pool.mediaWays.net)
05:00
05:04:15 Quit bunnyboi (Ping timeout: 246 seconds)
05:04:25 Join Jennifur [0] (~androgyne@cpe-72-224-19-1.nycap.res.rr.com)
05:11:18 Quit shai (Quit: Leaving)
05:11:36 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
05:11:49 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
05:25:45 Join Zarggg [0] (~zarggg@65-78-69-194.c3-0.eas-ubr6.atw-eas.pa.cable.rcn.com)
05:25:56 Quit evilnick (Ping timeout: 248 seconds)
05:26:38 Join evilnick [0] (~evilnick@2001:0:4137:9e74:28a5:1241:b5a6:f655)
05:31:14 Quit Horscht (Quit: Verlassend)
05:34:00 Quit [Saint] (Ping timeout: 246 seconds)
05:38:45 Quit panni_ (Read error: Connection reset by peer)
05:45:03 Join togetic [0] (~togetic@unaffiliated/ibuffy)
05:49:49 Join mt2 [0] (~chatzilla@41.233.143.19)
05:53:45 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.143)
05:59:59***Saving seen data "./dancer.seen"
06:00
06:03:37 Join JdGord [0] (3a601b9e@gateway/web/freenode/ip.58.96.27.158)
06:07:49 Quit evilnick (Ping timeout: 276 seconds)
06:26:37 Quit n17ikh (Ping timeout: 258 seconds)
06:31:22 Join n17ikh [0] (~n17ikh@c-68-59-25-51.hsd1.sc.comcast.net)
06:34:23S_a_i_n_twhere is the "rockbox media player" expected to install it's driver from?
06:34:42S_a_i_n_tConnected to Windows XP and saw this message for the first time...ever.
06:35:20S_a_i_n_tAn auto install wizzard for "Rockbox Media Player", that fails because it has no idea what driver to use or where to look for it.
06:39:37 Join evilnick- [0] (~evilnick@ool-4a5909aa.dyn.optonline.net)
06:43:04saratogaIIRC thats just on players without proper USB support
06:43:25saratogaso theres no driver, its just a popup on windows for players without an actual USB rockbox driver
06:43:43saratogabecause rockbox didn't enumerate a proper USB driver
06:44:46S_a_i_n_tHmmm, it's a Nano2G, so makes sense.
06:45:01S_a_i_n_tItnteresting it popped up as "Rockbox Media Player" though.
06:45:07S_a_i_n_tFirst time I have seen that, ever.
06:45:16S_a_i_n_tEven my Nano1Gs don't do that.
06:47:54S_a_i_n_tThe 2Gs usually mount correctly, and have the correct volume lable, but get called "Unknown Device" by the root hub, or sometimes manage to get labled as an HID keyboard :-S
06:52:05 Join Orc [0] (~Orc@83.149.8.142)
06:53:49saratogaprobably some bug in the driver or windows that caused it to not mount as a USB devices correctly
06:55:55 Quit scorche (Disconnected by services)
06:56:05 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
06:56:31 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
07:00
07:13:27 Quit JdGord (Quit: Page closed)
07:18:22pixelmathe Clip+ manual mentions the soft hold in the WPS button table, it's called "keylock on/off". There is one weirdness though as the same button combo is said to be "play/pause"
07:19:28pixelmasaratoga: didn't you change some button mappings for the Clip+ semi-recently?
07:19:45 Join bieber_ [0] (~quassel@162-78.97-97.tampabay.res.rr.com)
07:19:53saratogapixelma: yes I changed something, but it was reverted
07:20:57pixelmaok, something more to check. I'll look what I can do with the picture, if it isn't a quick fix then I can only do so in the evening though
07:25:42 Quit Orc ()
07:30:39 Quit BHSPitMonkey (Remote host closed the connection)
07:38:12CIA-6New commit by mt (r27425): Initial cleanup for libwmapro.
07:38:53JdGordoncan anyone think of a reason why the parser would ever need to handle negative decimal params?
07:39:54CIA-6r27425 build result: All green
07:41:27saratogagevaerts: thank you
07:43:17 Join antil33t [0] (~Mudkips@124-197-51-80.callplus.net.nz)
07:44:23pixelmathe fix for the front picture would be easy but then many screenshots would not be used (because they have the "-sansaclip" in theirn ame) I wonder what the reason for this is. I'm aware that there are other screens with this resolution but we don't have screenshots of the remotes in the manual and I personally don't think we are going to... maybe for the m200
07:44:59 Quit Jennifur (Quit: need the ram)
07:46:38pixelmaalthough a black on white 128x64x1 is more standard
07:56:33 Quit mt2 (Ping timeout: 240 seconds)
07:59:58 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
08:00
08:00:03***Saving seen data "./dancer.seen"
08:01:48 Quit S_a_i_n_t (Ping timeout: 240 seconds)
08:03:00 Join mitk [0] (~mitk@195.117.162.130)
08:10:35 Join bunnyboi [0] (~androgyne@cpe-72-224-19-1.nycap.res.rr.com)
08:11:25 Nick solrize is now known as Male23935QX (~phr@adsl-67-119-12-184.dsl.snfc21.pacbell.net)
08:11:36 Nick Male23935QX is now known as solrize (~phr@adsl-67-119-12-184.dsl.snfc21.pacbell.net)
08:11:45 Quit solrize (Changing host)
08:11:45 Join solrize [0] (~phr@unaffiliated/solrize)
08:18:48 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.2)
08:18:54pixelmathe manual also only knows about one Clip pad
08:24:15CIA-6New commit by bieber (r27426): Theme Editor: Committed FS #11477 to add a DECIMAL parameter type in the parser and adapt the Theme Editor to accomodate the change by Johnathan ...
08:25:51CIA-6r27426 build result: All green
08:30:14 Quit jfc (Read error: Connection reset by peer)
08:30:27 Join jfc [0] (~john@dpc6682208002.direcpc.com)
08:37:49 Join mt2 [0] (~chatzilla@41.233.142.57)
08:45:09 Join JdGord [0] (3a601b9e@gateway/web/freenode/ip.58.96.27.158)
08:50:43 Quit M3DLG (Ping timeout: 245 seconds)
08:51:00 Quit sinthetek (Read error: Connection reset by peer)
08:51:08 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
08:54:10 Quit bieber_ (Remote host closed the connection)
09:00
09:00:28 Quit bunnyboi (Quit: IRC is just multiplayer notepad)
09:04:11S_a_i_n_tAnyone tested FS #11475 on a *not* iPod Colour?
09:04:55S_a_i_n_talternatively, what iPods share the same keyclick driver with the Colour?
09:10:55Tarvishey, what compression method does the themes site use?
09:10:59Tarvisit keeps rejecting my .zip
09:11:13S_a_i_n_tit uses zip.
09:11:21Tarvisneed more info than that
09:11:35S_a_i_n_tare you positive there's no hidden extension?
09:11:39TarvisI'm guessing it uses Deflate
09:11:44Tarvisyes
09:12:04 Join MxxCon [0] (donuts@ool-18b9baa8.dyn.optonline.net)
09:12:43Tarvisthere, that did it
09:13:21Tarvisalright, boxamp is back up
09:13:22Tarvishttp://themes.rockbox.org/index.php?themeid=914&target=iriverh300
09:13:47MxxConhello folks. i'm trying to install 'manufacturer mode' drivers for my sansa e200r, but i'm on win7 64bit and it refuses to install those drivers unless they are signed
09:15:48MxxConany idea what to do?
09:16:08S_a_i_n_tcan't you turn off that setting in the OS? (refusing unsigned drivers?)
09:16:35*S_a_i_n_t is pretty sure there's an option for this somewhere in XP at least...
09:16:57MxxConwin7 64bit requires signed and can't turn it off
09:17:25MxxCon32bit can install unsigned but not 64bit
09:17:36TarvisYou can
09:17:45TarvisYou have to do it during a boot screen though
09:17:55TarvisReboot the computer and mash F8 after you get past POST
09:18:19TarvisThen there's an option to Disable Driver Signature ENforcement
09:18:30Tarvisnote that you'll have to do this every time you use the drivers though
09:18:53TarvisTHere's other methods though
09:18:55Tarvishttp://www.raymond.cc/blog/archives/2009/08/24/loading-unsigned-drivers-in-windows-7-and-vista-64-bit-x64/
09:19:14 Join petur [0] (d408b802@rockbox/developer/petur)
09:19:30TarvisThe second method modifies the boot loader to automatically always use that option (ReadyDriver Plus_
09:19:33Tarvis)*
09:19:42MxxConi'll need to do this only once to run e200rpatcher.exe, right?
09:19:49Tarviswell
09:19:59Tarvisdepends on when the driver is loaded
09:20:01S_a_i_n_tin theory, yes
09:20:07Tarvisif it's loaded on startup you'll get constant nag
09:20:18Tarvisyou can remove the driver after the patcher, though
09:20:54MxxConoff to try that
09:20:56MxxConthanks
09:21:12 Quit MxxCon (Read error: Connection reset by peer)
09:29:38 Join n1s [0] (~n1s@rockbox/developer/n1s)
09:32:18JdGordbieber: just saw the commit, thanks alot.
09:32:32bieberNo problem
09:32:43JdGordnot having the parser throw an error on is a bit annoyng but I suppose makes sense
09:33:06bieberIt would probably make sense to throw in an error there if it doesn't encounter a digit initially
09:33:22JdGordor a .
09:33:46JdGordnow resyncing my patch is going to be anoying again :/
09:38:06bieberOh, sorry
09:39:23 Join MxxCon [0] (donuts@ool-18b9baa8.dyn.optonline.net)
09:39:30MxxConyay, that worked
09:39:41MxxConthank you for your help
09:39:59S_a_i_n_tglad it worked :D
09:40:33MxxConand i even managed to get it to work w/ newer firmware :D
09:41:30MxxConso now i have latest sansa firmware and rockbox. sexylent
09:42:11 Join Rob2223 [0] (~Miranda@p4FDCB5F9.dip.t-dialin.net)
09:42:59 Join swilde [0] (~wilde@aktaia.intevation.org)
09:44:02MxxConnow to find some media player that can keep track of tracks i played on rockbox..
09:45:19 Quit Rob2222 (Ping timeout: 240 seconds)
09:45:35 Quit Kitar|st (Ping timeout: 258 seconds)
09:51:02 Join Kitar|st [0] (~Kitar_st@BSN-182-56-95.dial-up.dsl.siol.net)
09:51:46 Part MxxCon
09:56:21 Quit bieber (Ping timeout: 248 seconds)
09:56:34 Join bieber [0] (~quassel@162-78.97-97.tampabay.res.rr.com)
09:58:22 Quit JdGord (Quit: Page closed)
10:00
10:00:07***Saving seen data "./dancer.seen"
10:04:50 Join wodz [0] (~wodz@skatol.ch.pw.edu.pl)
10:06:10n1shmmm, the version of init_vlc from current ffmpeg does a malloc that is easy to replace with a VLA on stack but the worst case in libcook seems to need about 4.8k of memory... which smells like a stkov waiting to happen
10:07:45 Join flydutch [0] (~flydutch@host21-164-dynamic.15-87-r.retail.telecomitalia.it)
10:10:44n1si suppose a 5k static buffer would be fine
10:16:15n1sotoh it does work with with the stack butter and it does seem like the worst case triggers on every cook file anyway :)
10:18:37*n1s wants atrac test files
10:19:01 Join JdGord [0] (~jdg@122.110.252.218)
10:29:12n1sstach usage of cook doesn't go over 64% so it seems fine
10:29:15n1sstack
10:40:33 Quit Tarvis (Read error: Connection reset by peer)
10:50:24 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:52:47 Join einhirn [0] (~Miranda@bsod.rz.tu-clausthal.de)
10:55:25 Quit JdGord (Quit: Bye)
10:56:09 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:56:18 Join JdGordon2 [0] (~jonno@122.110.252.218)
11:00
11:00:54 Quit mt2 (Ping timeout: 240 seconds)
11:10:25 Quit r0b- (Ping timeout: 248 seconds)
11:10:45pixelmagot (I think) a nice solution to the Clip/Clip+ image problem in the manual and it works -> additionally use the playertype option for images. Images will be used if present in the following order e.g. image-128x64x1-sansaclipplus > image-128x64x1-Clip > image-128x64x1, "Clip" would be true for all Clips, including the Clip+, "sansaclipplus" would only be this one. What do others think?
11:10:48 Quit Kitar|st (Read error: Connection reset by peer)
11:10:52 Join r0b- [0] (~nnscript@adsl-76-235-199-132.dsl.klmzmi.sbcglobal.net)
11:10:52 Join Kitar|st [0] (~Kitar_st@BSN-182-56-95.dial-up.dsl.siol.net)
11:12:20 Quit JdGordon2 (*.net *.split)
11:12:20 Quit S_a_i_n_t (*.net *.split)
11:12:20 Quit bluebrother (*.net *.split)
11:12:20 Quit dionoea (*.net *.split)
11:12:21 Quit mk429 (*.net *.split)
11:12:21 Quit ThomasAH (*.net *.split)
11:12:21 Quit Hadaka (*.net *.split)
11:12:21 Quit r0b- (*.net *.split)
11:12:21 Quit einhirn (*.net *.split)
11:12:21 Quit n1s (*.net *.split)
11:12:21 Quit jfc (*.net *.split)
11:12:21 Quit TheSeven (*.net *.split)
11:12:22 Quit Utchybann (*.net *.split)
11:12:22 Quit GodEater (*.net *.split)
11:12:22 Quit crwl (*.net *.split)
11:12:22 Quit detaos (*.net *.split)
11:12:22 Quit Guest5171 (*.net *.split)
11:12:22 Quit incant_ (*.net *.split)
11:12:22 Quit ps-auxw (*.net *.split)
11:12:22 Quit Kamyk (*.net *.split)
11:12:22 Quit Torne (*.net *.split)
11:12:44 Join mk429 [0] (mk429@spoon.netsoc.tcd.ie)
11:14:37 Join Hadaka [0] (~naked@naked.iki.fi)
11:15:31*bertrik is going to apply some language updates from FS
11:17:39CIA-6New commit by bertrik (r27427): Update Thai translation - FS #11474 by Phinitnun Chanasabaeng
11:18:43 Join r0b- [0] (~nnscript@adsl-76-235-199-132.dsl.klmzmi.sbcglobal.net)
11:18:43 Join JdGordon2 [0] (~jonno@122.110.252.218)
11:18:43 Join einhirn [0] (~Miranda@bsod.rz.tu-clausthal.de)
11:18:43 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:18:43 Join jfc [0] (~john@dpc6682208002.direcpc.com)
11:18:43 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.2)
11:18:43 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
11:18:43 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
11:18:43 Join Utchybann [0] (~Utchy@rps6752.ovh.net)
11:18:43 Join dionoea [0] (~dionoea@yop.chewa.net)
11:18:43 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
11:18:43 Join crwl [0] (~crwlll@dsl-jklbrasgw1-fe10fb00-173.dhcp.inet.fi)
11:18:43 Join detaos [0] (~quassel@ip72-218-104-242.hr.hr.cox.net)
11:18:43 Join Torne [0] (torne@rockbox/developer/Torne)
11:18:43 Join Kamyk [0] (kamyk@szluug.org)
11:18:43 Join ps-auxw [0] (~arneb@dyn37.ps-auxw.de)
11:18:43 Join incant_ [0] (~philippe@zone42.rdc.uottawa.ca)
11:18:43 Join Guest5171 [0] (jljhook@irkki.fi)
11:18:43 Join ThomasAH [0] (~thomas@aktaia.intevation.org)
11:19:41CIA-6r27427 build result: All green
11:20:33JdGordon2sweet! I figured out how to hopefully not break current skins without going back to the old draw order
11:20:34JdGordon2mc2739: ping?
11:21:19JdGordon2As long as noone relies on a empty line being cleared... which I shouldnt think they do
11:22:55 Join mt2 [0] (~chatzilla@41.239.51.181)
11:24:20 Join halmi [0] (~netbook@93-82-32-34.adsl.highway.telekom.at)
11:26:03CIA-6New commit by bertrik (r27428): Update of Czech language - FS #11479 by Marek Salaba
11:26:04wodzwhy is synaptics MEP driver use bit-bang I2C implementation? (synaptics-mep.c) ?
11:26:42bertrikI don't know, maybe it's simply not connected to an i2c controller.
11:26:44 Join simonrvn [0] (simon@211.36-ppp.3menatwork.com)
11:27:18bertrikWe do have a lot of separate bit-bang I2C implementations in rockbox by the way
11:27:37CIA-6r27428 build result: All green
11:31:29 Quit TheSeven (Ping timeout: 260 seconds)
11:32:38bertrikThere is a generic I2C driver in firmware/drivers/generic_i2c.c that does the I2C-level stuff and calss supplied function pointers to do the target specific GPIO toggling
11:33:32CIA-6New commit by bertrik (r27429): Slovak language update - FS #11481 by Peter Lecky
11:33:42wodzI know but mep driver do this 'in it own way'
11:34:09wodzthis plain stupid mep protocol layer should be separate to the physical layer
11:34:47bertrikwodz, do you have an idea which players use the mep protocol?
11:35:04CIA-6r27429 build result: All green
11:36:12 Join DerPapst [0] (~Alexander@p4FE8F70B.dip.t-dialin.net)
11:38:12 Quit halmi (Quit: halmi)
11:38:37wodziriver h10, packard bell vibe500, phillips HDD1630 HDD6330 SA9200, m:robe 100
11:44:26wodzhmm iriver h10 do not use mep protocol
11:46:04amiconnwodz: Did you see 14:24:24amiconnwodz: (for the logs): amiconn.dyndns.org/~jens/hd200_grey.diff">amiconn.dyndns.org/~jens/hd200_grey.diff">http://amiconn.dyndns.org/~jens/hd200_grey.diff :D (Sunday)?
11:46:40bertrikwodz, I think synaptics-mep.c is not implemented i2c
11:46:58wodzamiconn: nop, I'll test this when at home
11:47:19wodzbertrik: I think it is
11:47:27wodzit is mep over i2c
11:47:58wodzhttp://www.synaptics.com/sites/default/files/511-000039-01b.pdf
11:48:11bertrikbut it uses 3 wires + interrupt as far as I can see, while i2c uses only 2
11:49:24 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:52:45wodzbertrik: You are right - it uses MEP 3-wire interface
11:53:26wodzbut still it mixes interface layer with protocol layer
11:53:54wodzI mean synaptics-mep.c
11:54:46n1smt2: do you have encoders for atrac and wma pro? it would be nice to have some test files in the test_files set
11:58:29mt2n1s: I have one for wma pro I think .. but not sure if I have one for atrac .. I'll check but I'm not sure I could make the files today, most probably tomorrow.
11:58:32 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:58:54wodzamiconn: am I right that move.w and move.b have the same timings? If so it would be better IMO to change move.w to move.b when transferring data to LCD. This way it will be explicit that lcd bus is only 8bits wide.
12:00
12:00:08***Saving seen data "./dancer.seen"
12:05:06 Join efyx [0] (~efyx@lap34-1-82-225-185-146.fbx.proxad.net)
12:08:50 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:12:57 Join M3DLG [0] (~M3DLG@212.183.140.16)
12:14:50 Join watto [0] (~watto@193.203.81.165)
12:19:42JdGordon2mc2739: your theme is crashing because the font names dont have the .fnt (I think)
12:19:46JdGordon2bieber: woo! crashed the editor
12:22:06JdGordon2bieber: http://www.rockbox.org/tracker/task/11470?getfile=22284 open the sbs... line 44, change %t(0) at the end to %t(1)
12:23:05JdGordon2also it looks like that line isnt being parsed correctly
12:25:08JdGordon2gdb says it then parses correctly but crashes rendering it
12:27:26JdGordon2mc2739: hmm, no seems I'm wrong... .fnt is there
12:32:48JdGordon2ah! I see what is happening... it tries to reload the font every time... thats a bug in svn also, maybe I should fix it there
12:33:48wodzhow do You understand the sentence: "Error checking will take the form of a Packe Error Check (PEC) checksum (...). The PEC byte is an additive 8-bit unsigned checksum. (..) Overflows encountered during the summation are ignored."
12:33:51 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
12:34:34wodzis sum of all 1 in data?
12:35:43 Quit dfkt (Ping timeout: 245 seconds)
12:37:25 Quit JdGordon2 (Ping timeout: 248 seconds)
12:39:07 Join JdGordon2 [0] (~jonno@122.110.252.218)
12:40:51pamaurywodz: I understand this sentence that is there is an overflow, it should just behave as if the computation was modulo 256
12:41:44 Quit M3DLG (Ping timeout: 276 seconds)
12:42:11wodzthats clear but how checksum is computed - is it simply sum of all data modulo 256 or is it sum of 1 in all data (like parity check) or what
12:43:54pamauryI think it's the sum but that's a bit unclear: "additive 8-bit unsigned checksum". I would say the sum of data.
12:45:03 Quit binaryhermit (Ping timeout: 258 seconds)
12:45:15pamauryWhere did you find this ?
12:47:16wodzhttp://www.synaptics.com/sites/default/files/511-000039-01b.pdf
12:47:30wodzpage 11
12:52:30pamauryThe best way is probably to have some data to compare against. On the internet, PEC can be found in other devices and they talk about CRC-8 but the synaptic document doesn't mention a CRC polynomial
12:54:09wodzit may be as well sum + invert which is often used as simple checksum
12:56:57pamaurydo you have some data ? Is it for the MPDIO ?
12:57:11pamaury*MPIO
12:57:46wodzit used in mpio hd300 key reading routine as it uses synaptics capacitive touch module
12:57:54wodzs/it used/it is/g
12:58:34pamauryDo you need to generate such a checksum or just read and check ?
12:59:27wodzI need to generate
12:59:47pamauryThen I guess you first need to read data to check what the checksum is :)
13:00
13:00:23wodzor get solid understanding of dissasembly :-)
13:26:20JdGordon2mc2739: ok, I was wrong about why theme crashes, I'm mostly clueless as to why still :p
13:31:08JdGordon2bloody hell! fixed
13:38:45 Quit S_a_i_n_t (Ping timeout: 248 seconds)
13:39:26 Quit wodz (Quit: Leaving)
13:39:40JdGordon2Torne: any news on the font loading stuff?
13:39:49Tornei've not worked on it any more, i'm afraid
13:40:29 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.236)
13:50:40 Join punt2 [0] (5b314678@gateway/web/freenode/ip.91.49.70.120)
13:51:02 Quit S_a_i_n_t (Ping timeout: 246 seconds)
13:52:49 Join S_a_i_n_t [0] (~st.lasciv@203.184.0.207)
13:53:32 Quit ved (Quit: leaving)
13:54:43 Join ved [0] (ved@ddsbox.co.cc)
13:57:33 Join halmi [0] (~netbook@93-82-32-34.adsl.highway.telekom.at)
14:00
14:00:11***Saving seen data "./dancer.seen"
14:02:50 Quit S_a_i_n_t (Ping timeout: 265 seconds)
14:04:35 Join funman [0] (~fun@rockbox/developer/funman)
14:05:59punt2hi funman - you are just the person I was lookig for :)
14:06:22punt2just about to set-up a build environment to test your "patch" from her: http://forums.rockbox.org/index.php?topic=25259.15
14:07:06 Quit scorche (Disconnected by services)
14:07:16 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
14:08:08 Quit mitk (Quit: Leaving)
14:08:18funmanhi
14:08:30funmanpunt2: you are MichaelScofield?
14:08:35punt2no
14:09:12punt2but I rockboxed my clip+ yesterday, noticed the same behavior and found the thread via the board search
14:09:29funmanok
14:20:06mc2739JdGordon2: not sure if I mentioned this in FS, but the theme loads on fresh boot, it just crashes when switching to that theme from another theme
14:20:21JdGordon2I fixed that
14:20:29JdGordon2latest patch should be good
14:20:39JdGordon2but apparently it is out of sync because git is playing up
14:22:33mc2739ok, I'll test now
14:22:38 Quit halmi (Ping timeout: 240 seconds)
14:26:03n1smt2: no rush, it's just nice to have test files in one place for easy testing and benchmarking
14:26:50JdGordon2mc2739: hang on a minute, resyncing.... just making sure it compiles ok now
14:27:54JdGordon2mc2739: grab 6a
14:46:09 Quit DerPapst (Ping timeout: 245 seconds)
14:46:23punt2someone should add the portion about rockboxdev.sh from http://www.rockbox.org/irc/log-20100621 to http://www.rockbox.org/wiki/VMwareDevelopmentPlatform - would have saved me some googling :)
14:47:10mc2739JdGordon2: looks good now, just some minor viewport problems on the fm screen, but I haven't played with the line order yet
14:47:22JdGordon2great :)
14:48:56funmanpunt2: about PATH still having the old toolchain?
14:49:47 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.77)
14:50:02punt2yeah, about the arm-eabi toolchain
14:50:25punt2more specifically the "arm-elf-eabi-gcc: command not found"-error
14:50:57punt2I guess eventually there will be an updated vmware image
14:51:30 Quit dfkt_ (Read error: Connection reset by peer)
14:51:51funmanif you have writing skills just register on the wiki and tell us your username so we can give you write access :)
14:52:01JdGordon2bieber: converting the tag parse functions to the new parser is soo damn simple :D
14:52:12 Join komputes [0] (~komputes@ubuntu/member/komputes)
14:53:44punt2I might just do that when I am off from work.. and if I manage to recall my old board/wiki username
14:55:53gevaertspunt2: we want real names on the wiki. I hope you can recall that one :)
15:00
15:00:00punt2ooops... well - at least I won't forget my login a second time then, I guess :)
15:00:43punt2building the toolchain really takes forever on this slow machine :/
15:02:35 Quit antil33t (Read error: Connection reset by peer)
15:02:41 Join antil33t [0] (~Mudkips@124-197-51-80.callplus.net.nz)
15:04:40funmanthat might change when gcc gets to use C++
15:08:52punt2after a make clean rockbox is finally compiling now.. phew.. that was fun :)
15:09:56punt2how long does the compiling of rockbox take compared to building the toolchain? just a rough estimate would do, so I know how long of a coffee break I can take :D
15:10:15S_a_i_n_ta lot less ;)
15:11:25S_a_i_n_tthe first compile can take a few mins (2~3 to 15) depending on your machine speed, then susequent xompiles significantly less time.
15:11:36S_a_i_n_t*compiles
15:12:06funmanccache might help
15:12:18S_a_i_n_tor, hinder ;)
15:12:33S_a_i_n_t(for a damn slow machine)
15:13:09funmanpunt2: according to http://build.rockbox.org/dev.cgi , our network of fast machines makes every build in less than 2 minutes
15:13:28punt2should be a lot faster on my machine at home.. but as I am at work my painfully slow netbook must do
15:13:48S_a_i_n_tthose are *fast* machines though..., like, *really* fast.
15:13:58punt2ah, there we go.. did not even take 10 minutes :)
15:14:28JdGordon2not if you change config.h :/
15:14:32 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
15:14:47*JdGordon2 has big plans for a bunch of existing tags once this rework is done!
15:14:55JdGordon2how does conditionals inside the playlist viewier sound?
15:15:16S_a_i_n_t..odd.
15:15:16S_a_i_n_tIn a word.
15:15:27bertrikJdGordon, what can I do specifically to help testing the new skin parser? (I'm more of a firmware/ guy)
15:15:45S_a_i_n_tI can't imagine a use case off the top of my head, but, I seldom use the playlistviewer at all.
15:16:06JdGordon2possibly sublines also...
15:16:14funmanJdGordon2: btw, I would have committed the incomplete patch to SVN, to get full testing
15:16:16S_a_i_n_tsublines, yes...definitely.
15:16:32JdGordon2bertrik: just ruin the build, you shouldnt see any difference
15:16:34S_a_i_n_tconditionals...if you can do it, great, but don't stress about it ;D
15:16:41bertrikhaha :)
15:16:57*S_a_i_n_t thinks he meant "run"
15:16:58JdGordon2funman: oh will if I dont get enough help...
15:17:07S_a_i_n_t...but, that was an amusing typo
15:17:24JdGordon2S_a_i_n_t: I'm hoping to be able to use all the current (read: new) code to do it so it all just happens by magic
15:17:24JdGordon2and haha :)
15:18:07S_a_i_n_t"just ruin the build"...classic ;)
15:18:14S_a_i_n_tIsn;t that what you're doing now? :P
15:18:29JdGordon2I'm un-ruining it now
15:18:35S_a_i_n_t;)
15:20:25 Join dfkt [0] (dfkt@unaffiliated/dfkt)
15:23:11punt2funman: setting MAX_CONTRAST_SETTING to 250 did not change anything for me
15:23:23punt2I can't see a difference between a contrast of 0 and 250
15:23:50 Quit mt2 (Ping timeout: 240 seconds)
15:24:05funmanpunt2: let me have a look, i have other ideas to try (and now that you have a build environment it'll be fast to test)
15:24:18punt2okay :)
15:24:32 Join mt2 [0] (~chatzilla@41.239.51.181)
15:24:41punt2the only time when contrast changes seems to be when the display is fading in/out
15:26:33S_a_i_n_tdo OLED screens *have* a contrast?
15:26:42*S_a_i_n_t thought it was technically brightness.
15:26:42funmanhm weird i have no fading here
15:26:48punt2the display seems to be brighter than without the change, though.. but that could also just be my imagination
15:27:08funmanpunt2: did you try 250, 249 etc?
15:27:09punt2afaik they actually only have contrast, not brightness
15:27:21funmantry between 250 and 240
15:27:41funmanS_a_i_n_t: but according to torne, contrast on OLED acts like brightness on LCD (at least it's like that on the clip screen)
15:28:10S_a_i_n_tAha, that seems to stack up with what I though, thanks.
15:28:17S_a_i_n_t*thought
15:28:37Tornewell, on an LCD brightness is how bright the backlight shines, and contrast is the mapping between input values and how much the LCD pixel elements respond by opening/closing to light
15:28:37punt2should contrast change on-the-fly or only after actually confirming the selection of a value?
15:29:06funmanit changes when browsing the list
15:29:08Torneon an OLED there's no backlight; the contrast controls the responsiveness of individual pixels to their input values, but since the pixels directly emit light the visible effect looks awfully like brightness
15:29:33punt2I can't notice any change at all :/
15:29:36funmanpunt2: for example if you change the volume through the list menu, you'll hear the volume change at each step
15:29:42S_a_i_n_tHmmm, thanks Torne. Ever helpfull as usual ;)
15:29:59punt2I see.. thanks for the clarification.. this is my first rockboxable device in 3 years :)
15:30:02TorneS_a_i_n_t: the datasheets call it contrast but I think that's just from habit wrt. LCDs :)
15:30:03funmanpunt2: which bootloader do you have installed?
15:30:16TorneS_a_i_n_t: i've got a bunch of OLED-display devices and it sure seems like a brightness scale to me ;)
15:30:30punt2I used the rockboxutility yesterday, so I would guess the most recent one - but let me check
15:30:33 Join panni_ [0] (hannes@ip-95-222-52-93.unitymediagroup.de)
15:30:48S_a_i_n_tme also, that's why I always thought it was "brightness", not "contrast", live and learn huh?
15:31:07funmanpunt2: is the bootloader logo strictly aligned with current build (rockbox.sansa) logo? or is there a 2 pixels shifting?
15:32:08punt2btw. the screen appears to be brighter now, only changing contrast does not seem to work. and the display fades (going from lowest contrast to highest) in every time the backlight gets activated
15:32:14punt2let me check if I can see what you mean
15:32:43bertrikpunt2, are you running some kind of patched build now?
15:33:37punt2just a freshly checked-out trunk build, using the little adjustment proposed by funman @http://forums.rockbox.org/index.php?topic=25259.0
15:33:59*S_a_i_n_t didn't think the Clip(s) had backlight fading in/out...nice.
15:34:48punt2the device boots so fast I can't really look at the logo.. but it says boot 2.0
15:36:02punt2actually it is more like a flickering than a fading
15:36:27punt2when the backlight is reactivated it seems to be at the highest contrast setting, then it goes to the lowest and slowly "fades" up to the highest again
15:36:28funmanpunt2: the fading in/out is only present with the patch?
15:37:05punt2good question.. I think it was present before, but not as noticeable as the overall brightness was lower
15:37:15punt2but let me check
15:37:25funmankugel: should clips screen fade in/out ?
15:37:57kugelfunman: isn't the last step (from lowest brightness to off) too big?
15:38:10kugelI naturally think it would be nice :)
15:38:11bertrikfunman, didn't we have to enable the DCDC15 for some clips?
15:38:26funmanbertrik: for clipv1
15:39:04funmankugel: i mean is it enabled in the code? because i never noticed it on my clips
15:40:50bertrikMaybe this DCDC15 voltage now controls brightness/contrast on some newer clips
15:41:21funmani've found that lcd_set_contrast() does something different for the 2 Clip+ controllers
15:41:48kugelfunman: I doubt it is
15:41:53punt2mine should be the most recent revision of the clip+, if that helps
15:41:56funmanthe argument value differs: (3*X)+1 for the one i have, (10*X)+1 for the 2nd type
15:42:12punt2rolling back to the latest official build now to see if the patch really had an effect or if I am just imagining things
15:42:27 Quit S_a_i_n_t (Ping timeout: 276 seconds)
15:43:21 Join halmi [0] (~netbook@93-82-32-34.adsl.highway.telekom.at)
15:43:31 Quit preglow (Remote host closed the connection)
15:44:28bertrikJdGordon, with current svn and your patch I get: /home/bertrik/rockbox.clean/apps/gui/skin_engine/skin_parser.c:1944: error: ‘struct wps_data’ has no member named ‘backdrop’
15:44:29 Join preglow [0] (thomj@tvilling2.pvv.ntnu.no)
15:45:06 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.182)
15:45:59JdGordon2bertrik: bah, which patch number?
15:46:26bertrik6a, I'm compiling 7 now
15:46:28funmanbertrik: the lcd init procedure is different for the 2 controllers in the OF
15:47:25 Quit JdGordon (Disconnected by services)
15:47:32 Nick JdGordon2 is now known as JdGordon (~jonno@122.110.252.218)
15:48:53JdGordonbertrik: ah, mono target?
15:49:08JdGordonwrap that line with "#if LCD_DEPTH > 1 || defined(HAVE_REMOTE_LCD) && LCD_REMOTE_DEPTH > 1"
15:49:15bertrikyes, the sansa clip v1 to be specific
15:49:48JdGordonok, the next version will have that change, not tonight though
15:50:26JdGordonit might fail for the same reason elsewhere :/
15:51:50 Quit mt2 (Ping timeout: 240 seconds)
15:52:44JdGordonbertrik: umm, thanks but actually I tihnk this isnt ready for the clipv1 yet.. I havnt fixed the memory usage yet so it uses double the skin buffer of svn builds
15:52:52JdGordonwhich might be bad
15:53:02 Quit halmi (Ping timeout: 265 seconds)
15:53:26JdGordon(I was going to say 400KB but I actually think its about 12K for the clip)
15:53:50kugelhopefully the rewrite makes using a dynamic buffer more easy
15:54:11bertrikJdGordon, I don't know much about the memory layout on clipv1, I guess funman is the expert on that
15:54:56 Join hebz0rl [0] (~hebz0rl@dslb-088-067-209-182.pools.arcor-ip.net)
15:55:09JdGordonkugel: yeah, it should do. I'm hoping to parse the text into the plugin buffer and count how much is needed, then reparse into the real buffer and load images and stuff.. should be simple
15:55:58funmanpunt2: i'll try to make a patch for you to test later (hopefulyl this evening), will you stay around on irc ?
15:56:12kugelparsing twice doesn't seem ideal
15:56:39JdGordonand it gets better... I'm thinking about removing all the wps_data structs from the various skins and having a manager have a bunch of slots so skins are loaded on demand
15:56:44JdGordonthats your idea isnt it?
15:56:57JdGordon(text will be in ram the whole time)
15:57:27punt2funman: sorry for the delay. had to get some actual work done. with the current official dev build the screen *seems* to be just as bright/dark (noticeably darker than with the original fw) as with your modification, but still the same fading/flickering
15:57:48funmanso it's bright, but no difference between 0 and 250 ?
15:57:55punt2sure thing. but I might reconnect later with an actual irc client and a less random nickname :)
15:58:02JdGordonthe other option is loading into the plugin buffer and then relocating all the tree elements into the skin buffer if the whole thing would fit
15:58:43JdGordonthat is going to be a crazy about of memmove() though
15:58:45punt2well I am running the latest official build now, so I again only have 50 as highest contrast
15:58:53punt2but 50 or 250 with your patch did not seem to make a difference
15:59:36punt2and changing the contrast setting does not change the brightness/contrast in any of the two builds
16:00
16:00:13***Saving seen data "./dancer.seen"
16:00:29funmandid you go directly from 0 to 250 and vice versa?
16:01:19punt2yes, I tried it both directly and by actually scrolling through the whole list
16:01:31punt2never noticed any change in contrast
16:02:32punt2I just closed the blinds so I could get a better look at the brightness
16:02:50punt2with the official build going from 50 to 0 and vice-versa actually _does_ make a slight difference
16:03:28punt2but I have to look very hard to notice it.. nothing nearly as extreme as the difference during the fading in
16:03:29funmanhm perhaps 250 is too high
16:05:38punt2switching back to the patched build now to test some other values
16:06:21*bertrik wonders where the fading in effect comes from
16:07:10punt2yeah.. if there was supposed to be fading, it would have to be fixed
16:07:47JdGordonany ideas how slow doing say 1000 <1K memmove()'s would be? (on the slowest targets I guess...)
16:07:47punt2because right now it's like this: highest brightness for a split second then lowest brightnesst -> slowly fading up to highest brightness
16:08:17JdGordonactually make than <100B moves
16:08:18 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
16:09:37punt2okay.. a jump from 0 to 250 seems to make the same difference as from 0 to 50 in the unpatched build
16:09:45bertrikI'd guess that memmoves are very fast
16:11:00JdGordonok, cool
16:12:03 Join halmi [0] (~netbook@93-82-32-34.adsl.highway.telekom.at)
16:12:25punt2funman: anything else I could do to help you pinpoint the bug?
16:13:57funmanpunt2: try all the values between 0 and 250 and see if there is a clear difference at some point
16:14:19funmanshould be very visible
16:16:55 Quit halmi (Ping timeout: 264 seconds)
16:17:59punt2there is a slight difference between 250 and 0, but I have to really focus on the battery logo to actually notice it
16:18:24punt2just went through all the 250 values and could not see clear difference at any point
16:18:51punt2and the 250 to 0 is not nearly as different as the lowest to brighest during the fading in
16:20:01 Join halmi [0] (~netbook@93-82-32-34.adsl.highway.telekom.at)
16:20:52 Quit kugel (Remote host closed the connection)
16:21:08punt2a contrast of 0 in rockbox seems to equal the lowest contrast in the stock fw and 250 seems to equal the second lowest in the stock fw
16:24:04 Quit JdGordon (Ping timeout: 245 seconds)
16:28:14 Quit evilnick- (Ping timeout: 245 seconds)
16:40:45 Join MethoS- [0] (~clemens@134.102.106.250)
16:41:31punt2funman: going to be afk for a bit. please query me if there is anything you need me to do/test. will reply once I get back :)
16:41:35punt2later guys
16:41:41 Nick punt2 is now known as punt2^afk (5b314678@gateway/web/freenode/ip.91.49.70.120)
16:43:37rasherpunt2^afk: Please don't do that
16:47:11 Join jgarvey [0] (~jgarvey@cpe-065-190-066-089.nc.res.rr.com)
16:56:18funmani've received an "unpatchable" fuzev2
16:56:42 Join bz [0] (~bz@81.5.80.170)
16:58:22 Quit t0rc (Quit: Leaving)
17:00
17:01:58bzHi folks! Could you please clear one thing out: will iPod with rockbox be recognized and a usual USB stick so files could be simply copied without iTunes?
17:02:18funmanwith a current build, yes
17:02:51funmanhum, with 3.6 also in fact
17:03:55bzThat is great! Thanks!
17:12:01funmanpunt2^afk: can you try this diff? http://pastie.org/1045740
17:13:56 Quit Kohlrabi (*.net *.split)
17:13:56 Quit BeFalou (*.net *.split)
17:13:56 Quit mikroflops (*.net *.split)
17:14:00 Join Kohlrabi [0] (~Kohlrabi@frustum.nosebud.de)
17:14:15 Join BeFalou [0] (~mamutoi@unaffiliated/befalou)
17:14:24 Join mikroflops [0] (~yogurt@90-224-31-157-no112.tbcn.telia.com)
17:18:30 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:24:13*funman spots a 'emcfs_ams.c' string in fuzev2 OF
17:35:02 Quit MethoS- (Remote host closed the connection)
17:40:10 Quit pamaury (Remote host closed the connection)
17:40:17 Join Luca_S [0] (www-data@giant.haxx.se)
17:42:28 Quit anewuser (Quit: http://xrl.us/detroitVI Sixth edition of 25 samples + 48 hours + detroit party (july 31) (maybe streamed online!))
17:43:05saratogaBagder: could you upload these wmapro test files: http://duke.edu/~mgg6/rockbox/wmapro/
17:50:39 Join kugel_ [0] (~kugel@e178121077.adsl.alicedsl.de)
17:50:59 Quit kugel (Disconnected by services)
17:51:03 Nick kugel_ is now known as kugel (~kugel@e178121077.adsl.alicedsl.de)
17:51:07 Quit kugel (Changing host)
17:51:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:53:10 Join evilnick- [0] (~evilnick@ool-4a5909aa.dyn.optonline.net)
18:00
18:00:15***Saving seen data "./dancer.seen"
18:02:12 Join stoffel [0] (~quassel@62.237.32.210)
18:05:35 Quit petur (Quit: *plop*)
18:06:26 Quit Luca_S (Quit: CGI:IRC)
18:06:39 Join halmi_ [0] (~netbook@93-82-32-1.adsl.highway.telekom.at)
18:09:20 Quit halmi (Ping timeout: 240 seconds)
18:10:05 Join BlakeJohnson861 [0] (~bjohnson@c-24-118-162-123.hsd1.mn.comcast.net)
18:10:19 Quit BlakeJohnson86 (Quit: Leaving.)
18:10:48 Quit dfkt (Read error: Connection reset by peer)
18:19:15*n1s crosses fingers
18:19:21CIA-6New commit by nls (r27430): Sync codeclib bitstream code with upstream ffmpeg code. Build ffmpeg_bitstream.c as a part of the codec lib. Use this codeclib implementation in ...
18:19:40bertrikfunman, nice to hear you have one of those fuze v2s
18:20:16funmani suppose sandisk introduced new hardware not handled by OF older than 2.3.31
18:21:13CIA-6r27430 build result: 0 errors, 3 warnings (nls committed)
18:22:23 Nick punt2^afk is now known as punt2 (5b314678@gateway/web/freenode/ip.91.49.70.120)
18:22:30punt2funman: compiling as we speak
18:22:41kugeln1s: you could also use #ifndef ROCKBOX to denote stuff unused in rockbox
18:24:43n1skugel: yeah, but even if you would build this stuff standalone none of our codecs use the commented out stuff, which i think #ifndef ROCKBOX implies
18:28:40 Quit kugel (Ping timeout: 265 seconds)
18:29:53balintx_the previous build sansa fuze v2 was bad.
18:30:07balintx_rbutil could not unpack at about 8 %
18:30:25 Quit ThomasAH (Remote host closed the connection)
18:30:30balintx_total commander said (when I downloaded it manually) it is a bad Compressed file
18:30:50CIA-6New commit by nls (r27431): Hopefully fix 'comparison between signed and unsigned' warning
18:31:13n1sbalintx_: maybe try downloading the file again?
18:31:25balintx_27429 was the build
18:31:33balintx_the newer 27430 which is committed now works
18:31:37 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:31:41*n1s wonders why only the mips gcc warned for that comparison
18:31:47balintx_and yes, rbutil downloaded, and i downloaded too
18:32:14kugeln1s: different gcc version probably
18:32:31n1skugel: yeah its 4.1 but still
18:32:32CIA-6r27431 build result: 0 errors, 3 warnings (nls committed)
18:32:33kugelI find it a bit strange that we now use 4 different gcc versions for 4 architectures
18:32:37n1sdamnit!
18:33:06*gevaerts agrees with kugel
18:33:10gevaertsWe should have five!
18:33:18n1skugel: yes, it's a bit annoying
18:33:21balintx_:D
18:33:29 Quit tchan (Quit: WeeChat 0.3.3-dev)
18:34:00kugelgevaerts: we do soonish! distros on desktop pcs will begin shipping gcc 4.5
18:34:03n1si don't have the mips toolchain and would rather not try more blind fixes...
18:35:35funman4.5.0 didn't work on AMS
18:35:44 Join GArik_ [0] (~wesnoth@95-29-242-42.broadband.corbina.ru)
18:36:13n1sx.y.0 gcc is usually buggy
18:36:27kugelon non-x86*
18:36:33punt2funman: the patch seems to have done the trick
18:36:39punt2there is a huge difference between 25 and 0 now
18:36:44n1shey, i had mips toolchain on the netbook
18:37:02punt2and the single steps actually change contrast now, too
18:37:21punt2the fading in seems to be gone, too
18:38:02funmanpunt2: nice, now we must reduce the diff to find which change exactly was needed
18:38:46saratogathe CodecPerformanceComparison results on the wiki are pretty old for Coldfire, some CF owner should update them
18:39:04punt2alright, just lead the way :)
18:39:34funmanpunt2: can you do that?
18:39:57n1ssaratoga: i've been meaning to do so but there's always one more patch to commit first
18:40:25saratogai understand
18:40:36n1sbtw, as you suspected the codeclib bitstream stuff doesn't work with libwma out of the box
18:40:44 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
18:41:09saratogayeah I gutted a 2007 era version of it for libwma
18:41:10kugeln1s: maybe you could have another go at gcc 4.4.4 for cf after you finished your recent codec stuff? :)
18:41:23n1skugel: maybe
18:41:34saratogaso at very least the function arguments and the init are likely different
18:41:59n1ssaratoga: in fact, it built fine but crashed in the sim
18:43:43balintx_Finally I have to say RBUtil is completely crazy. I get http://imagerz.com/QEJGCktvAwJRBF1EFQVQ for every operation. This not happened before.
18:44:30n1sbalintx_: could there be any crazy firewalls or cache things between you and the server that corrups stuff?
18:44:44saratogan1s: did you have to change the init_vlc calls?
18:45:18balintx_firewall is only the Windows Firewall (which does not do anything, of course).
18:45:26n1ssaratoga: no, the args match the current version
18:47:00n1siirc it segfaulted in GET_VLC
18:48:28funmanfuzev2 2.3.33 has a new 'firmware letter' : D
18:50:33punt2funman: I think I can, if you give me some basic instructions and tell me which changes on lcd-ssd1303.c I should omit first to trial & error
18:51:31funmanpunt2: first try to revert the changes to sansaclip.h and to lcd_set_contrast() (you know how to use svn diff / svn status ?)
18:53:07punt2nope, but I am going to check the crash course now :)
18:53:48funmaneh, the 'fuzpD.bin' removed the "Settings" menu in the OF
18:55:10 Join kugel_ [0] (~kugel@g231109252.adsl.alicedsl.de)
18:55:32 Quit kugel (Disconnected by services)
18:55:35saratogan1s: can you give me your diff?
18:55:36 Nick kugel_ is now known as kugel (~kugel@g231109252.adsl.alicedsl.de)
18:55:40 Quit kugel (Changing host)
18:55:40 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:56:35gevaertsJdGordon1: are conditional viewports supposed to work with patch 7?
18:57:15n1ssaratoga: i just removed the bitstream.[ch] files in libwma and used lib/ffmpeg_bitsream.c and lib/ffmpeg_get_bits.h instead afair
18:57:23bertrikfunman, weird. Maybe 'D' stands for demo?
18:59:11gevaertsJdGordon1: my theme (druppel for 240x320) has two viewports if there's AA (one with track info and one with AA), and one in case of no AA (track info, positioned a bit lower than the same info in case of AA). With your patch, all three seem to be shown
18:59:36funmanbertrik: dunno, but it doesn't help patchign the fuzev2 so i won't look for the details (you could check the individual enable/disable bits if you want)
19:00
19:02:08 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
19:03:43Bagdersaratoga: done!
19:03:50saratogathanks
19:03:54 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
19:04:46 Join dfkt [0] (dfkt@unaffiliated/dfkt)
19:04:55funmanbertrik: did you look into the details of AMS OF patching? (creation of UPGRADE.FIN, that kind of things)
19:07:02funmanalso do you remember seeing a "Sansa Fu.ze" file on the device?
19:09:09 Join Jerom [0] (~heidi@79.132.46.225)
19:10:19punt2refreshed my memory on svn diff/revert/status and compiling a patched build without your sansaclipplus.h and lcd_set_contrast modifications now
19:11:34 Quit flydutch (Quit: /* empty */)
19:13:20 Quit hebz0rl (Ping timeout: 264 seconds)
19:14:31saratogathe codec lib bitstream code seems to be working for wma std
19:14:58CIA-6New commit by saratoga (r27432): Use codeclib version of the ffmpeg bitstream code for wma rather then ancient hacked up version.
19:15:18gevaertsJdGordon1: scrolling also seems to be broken
19:15:22CIA-6New commit by nls (r27433): Properly fix warning by copying the swap16 and swap32 inline functions from system.h to mips/ingenic_jz47xx/system-target.h to replace the defines.
19:15:49funmann1s: duplicating code?
19:16:34CIA-6r27432 build result: 0 errors, 3 warnings (saratoga committed)
19:17:09n1sfunman: yes, i chose that over further #ifdef madness in system.h, also the functions are 1 and 2 lines and very unlikely to change
19:18:07saratogawhy did updating the thai translation increase binsize?
19:18:16CIA-6r27433 build result: All green
19:18:18funmansaratoga: thai is the langauge with the biggest binsize
19:18:34funmanwe declare a buffer with this size (max_language_size.h iirc)
19:18:36saratogaweird
19:18:46n1ssaratoga: and the lang buffer is automatically made the size of the biggest lang
19:19:47n1ssaratoga: ah, not much changes needed at all in libwma, nice
19:19:53 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
19:20:52saratogayeah just had to set the use_static flag
19:21:25saratogais wma pro enabled on the clipv2?
19:21:45n1syeah, enabled for everything with more than 2MB ram
19:22:59 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:25:01n1sbtw, i want to try the "A32_BITSTREAM_READER" for arm targets, since it should be a lot faster, maybe for cf too in fact
19:25:15saratogamy pro test files just freeze on the clipv2
19:25:26 Join kugel_ [0] (~kugel@g231106124.adsl.alicedsl.de)
19:25:48 Quit kugel (Disconnected by services)
19:25:51 Nick kugel_ is now known as kugel (~kugel@g231106124.adsl.alicedsl.de)
19:25:56 Quit kugel (Changing host)
19:25:56 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:26:19saratogan1s: I tried them with wma std years ago and picked the fastest one
19:27:16n1soh
19:28:27saratogadonno if they've changed since then though
19:28:45 Join kugel_ [0] (~kugel@g231235071.adsl.alicedsl.de)
19:29:01funmangcc could've changed that
19:29:07 Quit kugel (Disconnected by services)
19:29:11 Nick kugel_ is now known as kugel (~kugel@g231235071.adsl.alicedsl.de)
19:29:15 Quit kugel (Changing host)
19:29:15 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:29:35n1si cant help but think reading 4 bytes is a lot slower than reading a long word and byteswapping it
19:31:00 Quit antil33t (Ping timeout: 240 seconds)
19:31:05n1snot that this has a very large impact on most lower bitrate files
19:31:34n1si'll experiment if i get bored enough some time
19:31:42punt2funman: after reverting, contrast and fading issues are still fixed
19:31:57punt2now just revert on diff after the other until the bugs occur again?
19:32:02punt2*one
19:33:17 Join kugel_ [0] (~kugel@g231108009.adsl.alicedsl.de)
19:33:31saratogan1s: http://archives.free.net.ph/message/20081227.194902.84b97745.el.html
19:33:39 Quit kugel (Disconnected by services)
19:33:43 Nick kugel_ is now known as kugel (~kugel@g231108009.adsl.alicedsl.de)
19:33:47 Quit kugel (Changing host)
19:33:47 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:33:49saratogaperhaps its been optimize since that test?
19:34:36saratogacan anyone try a wma pro file on AMS and tell me if it crashes? http://duke.edu/~mgg6/rockbox/wmapro/
19:35:23 Join anewuser [0] (anewuser@unaffiliated/anewuser)
19:36:21funmanpunt2: yes, the difference is in the lcd controller init procedure
19:36:50funmansome commands were added, some were removed, the order was modified a bit
19:37:14funmani think first thing is re-order it like the original so we can pin the difference down to added/removed commands and modified arguments
19:39:25punt2okay, just give me the diff or tell me what exactly to remove
19:39:38punt2once you are done, that is :)
19:40:39n1ssaratoga: it freezes on CF
19:42:24saratogaok we'll let mt worry about it then :)
19:43:51 Join captainkewlll [0] (2669ecc2@gateway/web/freenode/ip.38.105.236.194)
19:44:34bertrikfunman, no
19:45:20 Join antil33t [0] (~Mudkips@124-197-51-80.callplus.net.nz)
19:45:20saratogamy commit made wma a bit faster
19:45:27bertrikfunman, well I did try to look into the fuze v2 OF to see if I could spot the conditions for upgrade, but couldn't find it
19:46:02n1ssaratoga: yeah, i think they optimized vlc_init and build_table some
19:46:06funmanso far i just found some references to fuzp*.bin and upgrade.fin
19:46:23funmanpunt2: well i was expecting you to reduce the diff, less work for me :)
19:46:36funmanpunt2: "svn diff" should tell you what i changed
19:48:44punt2so basically all I do is revert -> remove next diff from patch -> -> patch -> build -> test -> rinse & repeat with next diff?
19:49:13 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
19:49:28funmannot really because the diff is all condensed in one hunk
19:50:07funmanrun svn diff -> edit lcd-ssd1303.c in order to reduce the output of svn diff next time it's run -> build -> test
19:50:50punt2alright, the closer to the original the better, right?
19:50:54funmanexact
19:51:09 Join bunnyboi [0] (~androgyne@cpe-72-224-19-1.nycap.res.rr.com)
19:51:14funmanbut do it step by step because contrast will stop working at one point and you want to know exactly which change cause it to work or not
19:51:23saratogaflac and shn seem to depend on ALT_BITSTREAM_READER
19:51:26punt2understood
19:53:47saratogain shn: s->header_bits = s->gb.index;
19:53:52saratogapretty sure thats not a good idea
19:55:28 Join CGL [0] (~CGL@190.207.140.100)
19:55:40 Join Tarvis [0] (~Tarvis@106.203.100.97.cfl.res.rr.com)
20:00
20:00:18***Saving seen data "./dancer.seen"
20:01:57 Join kugel_ [0] (~kugel@e178122043.adsl.alicedsl.de)
20:02:19 Quit kugel (Disconnected by services)
20:02:23 Nick kugel_ is now known as kugel (~kugel@e178122043.adsl.alicedsl.de)
20:02:29 Quit kugel (Changing host)
20:02:29 Join kugel [0] (~kugel@rockbox/developer/kugel)
20:10:26funmanpunt2: any luck yet ?
20:11:22CIA-6New commit by Domonoky (r27434): check result of validatezip() call.
20:12:19punt2didn't start yet. need to finish up on some work first, commute home and then I'll have all night :)
20:12:31funmanok ^^
20:12:47funmanwhat is your timezone? (i'm in CEST)
20:12:55CIA-6r27434 build result: All green
20:13:01 Join Peter93 [0] (~IRC-Clien@port-92-202-66-102.dynamic.qsc.de)
20:13:29punt2same
20:13:43punt2gmt+1 :)
20:13:52 Quit CGL (Ping timeout: 276 seconds)
20:18:09 Quit storm` ()
20:22:17 Join MethoS- [0] (~clemens@134.102.106.250)
20:22:43 Join markun [0] (www-data@rockbox/developer/markun)
20:25:00domonokyfunman: i think i found out why the this clipplus user has problems uploading his theme. i have run the clipplus checkwps on his themefile, and checkwps segfaults :-)
20:25:13 Quit markun (Client Quit)
20:25:18 Quit panni_ (Read error: Connection reset by peer)
20:25:22 Join markun [0] (www-data@rockbox/developer/markun)
20:25:30funmannice ;)
20:26:17domonokyso its a broken checkwps at least for all clips
20:26:30 Join panni_ [0] (hannes@ip-95-222-52-93.unitymediagroup.de)
20:26:33bluebrotherscorche`: can you promote bieber to developer in the forums?
20:26:54funmandomonoky: did you find the checkwps problem?
20:27:21domonokycheckwps for e200/e200v2 it works, but for all clips it seqfaults..
20:27:32domonokyfunman: no i just found that out..
20:28:03domonokyno gdb on the themeserver.. so i cant tell you where it segfaults :-)
20:28:24funmani don't know how to use checkwps
20:29:00domonokyusing it is easy. Build it and run: checkwps <themefile.wps>
20:29:00 Quit stoffel (Read error: Connection reset by peer)
20:30:35funmanno segfault here: % ./checkwps.sansaclip /media/bordel/rockbox/wps/zezayer.128x64x1.wps
20:31:04domonokytry it with the wps file the user posted
20:31:30domonokyhttp://forums.rockbox.org/index.php?topic=25266.0
20:32:06*domonoky cant build checkwps on cygwin, so i cant debug that at moment :-/
20:34:05funmanworks fine
20:34:14funman"WPS parsed OK"
20:34:54 Join binaryhermit [0] (~binaryher@adsl-99-141-196-197.dsl.emhril.sbcglobal.net)
20:35:41 Join DerPapst [0] (~Alexander@p4FE8FE6E.dip.t-dialin.net)
20:39:23bieberJdGordon1: That doesn't crash anything on my system
20:42:39 Quit dfkt (Read error: Connection reset by peer)
20:45:09funmandomonoky: is checkwps rebuilt frequently on the theme server ?
20:45:19 Quit Peter93 (Ping timeout: 240 seconds)
20:45:23 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
20:45:35domonokyfunman: it should get rebuild daily i think
20:47:45domonokycurrent revision on the themesite is: r27426
20:51:35funmanah it segfaults when built in 32 bits
20:52:15 Quit binaryhermit (Quit: Leaving)
20:54:14funmanJdGordon: what's the status of apps/gui/skin_engine/skin_parser.c , should we debug it ?
20:55:28funmanJdGordon1: http://pastie.org/1046100
20:59:01CIA-6New commit by bieber (r27435): Theme Editor: Made toutouch area coordinates absolute rather than relative
21:00
21:00:26CIA-6r27435 build result: All green
21:03:08 Join dfkt [0] (dfkt@unaffiliated/dfkt)
21:04:03CIA-6New commit by bieber (r27436): Theme Editor: Corrected last commit. Touch areas should have been made relative, not absolute
21:04:39 Quit markun (Quit: CGI:IRC 0.5.9 (2006/06/06))
21:05:26CIA-6r27436 build result: All green
21:06:41 Join stoffel [0] (~quassel@62.237.32.210)
21:10:05 Part watto
21:16:40 Join storm` [0] (~lol@ip98-179-182-180.ri.ri.cox.net)
21:16:53funmandomonoky: which CPU does the themeserver have ?
21:18:17domonokyproc/cpu-info tells me intel xeon @ 2Ghz (quad-core)
21:18:55funmanand running a 32 bits kernel?
21:22:28 Join notlistening [0] (~tom@94-195-105-95.zone9.bethere.co.uk)
21:22:32 Quit funman (Quit: free(random());)
21:23:39notlisteninganyone know if the in rockbox charging has been reenabled on the clip+
21:24:08 Join halmi [0] (~netbook@93-82-32-1.adsl.highway.telekom.at)
21:24:30 Quit tchan (Read error: Connection reset by peer)
21:25:19 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
21:27:08 Quit halmi_ (Ping timeout: 264 seconds)
21:31:23 Join binaryhermit [0] (~binaryher@99.141.196.197)
21:37:21 Quit whydoubt (Read error: Operation timed out)
21:37:27 Join whydoubt [0] (~whydoubt@ip68-12-76-9.ok.ok.cox.net)
21:38:46n1ssaratoga: right, i skimmed ffmpeg history for flac and saw some commits regarding eliminating dependence on ALT_BITSTREAM_READER
21:38:51 Quit wombat23 (Ping timeout: 240 seconds)
21:38:55 Join wombat23 [0] (~beuteltie@adsl-76-230-21-225.dsl.pltn13.sbcglobal.net)
21:39:23n1sanother task for a rainy day, our flac decoder is based on 5 years old ffmpeg code
21:39:45 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
21:41:14 Join M3DLG [0] (~M3DLG@bb-87-81-252-83.ukonline.co.uk)
21:49:09saratogai guess we could disable flac/shn and just benchmark in libwma
21:49:15 Quit stoffel (Remote host closed the connection)
22:00
22:00:20***Saving seen data "./dancer.seen"
22:05:20 Join Kriblom [0] (~gregoire@cpy69-1-78-225-128-2.fbx.proxad.net)
22:11:14 Quit S_a_i_n_t ()
22:11:50 Part punt2
22:15:42 Quit Kriblom (Quit: Ex-Chat)
22:16:23 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.2.20)
22:23:22CIA-6New commit by bertrik (r27437): Correction of bug in Czech language - FS #11428 by Marek Salaba
22:24:51CIA-6r27437 build result: All green
22:41:57pixelmaok, people who are interested in the manual, I'd like other opinions: currently image names for the manual consist of "basename-resolution" or "basename-resolution-targetname", the latter must match exactly and if such a file exists takes priority over the former. All Clip manuals use "sansaclip" which is why the Clip+ one uses the wrong player image, if I make the Clip+ one use "sansaclipplus" it won't use the "sansaclip" screenshots. My idea is
22:41:57pixelma to add another "playertype" naming scheme in the "middle" between the two which would be true for all Clips but gives the possibility to specify the "sansaclip(plus)" for unique images. There is an existing option I could reuse for this but it doesn't seem the correct way - is it worth to add another option to the platform files?
22:42:50 Quit bmbl (Quit: Bye!)
22:43:12*pixelma hopes people could follow
22:45:48bluebrotherso you want something like basename-resolution-manufacturer-targetname?
22:48:04 Join Topy [0] (~Topy44@f048204025.adsl.alicedsl.de)
22:48:32pixelmano, for you: \specimg would be set correctly to "sansaclipplus" for the Clip+ and invent another option say "sansaclips" for use of images common to all Clips (or even Ondios for that matter)
22:49:58pixelmafor e.g. common screenshots. Priority if exists: \specimg > \playerfamily (or however it'll be called) > simple ressolution
22:51:16 Quit notlistening (Quit: Leaving)
22:51:35pixelmaI could already make it work with \playertype but am not so happy with reusing it for this as it is also used elsewhere
22:51:35 Quit T44 (Ping timeout: 245 seconds)
22:52:19 Quit Topy44 (Ping timeout: 240 seconds)
22:57:25 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
22:58:08pixelmabluebrother: ss-fm-radio-screen-128x64x1-Clip.png (currently with reusing \playertype) would be used for both; ss-fm-radio-screen-128x64x1-sansaclip.png would only be used for the non-plus Clips, ss-fm-radio-screen-128x64x1-sansaclipplus.png for the Clip+, ss-fm-radio-screen-128x64x1.png for all 128x64x1. Currently we could even use the latter because the Clips are the only 128x64x1 targets but that won't be true if e.g. an m200 manual comes
22:59:33 Join Topy44 [0] (~topy@2a01:608:ffff:2627:2e0:4dff:fe83:a45e)
22:59:48n1spixelma: i think it sounds like a good idea but how to do it, i don't know
23:00
23:00:53n1sit would be nice if you could do it like the features and have sansaclip* match all playernames starting with sansaclip but that it's part of a filename makes that kind of impossible
23:01:14pixelmathat's one of the reasons I'd like to have the third scheme because the Clips' screen is probably unique (and special). The other reason is that the same problem exists for the Ondios which we got around by duplicating screenshots, one named -ondiofm the other ondiosp but since it only applies to some screenshots it wasn't that important)
23:02:08n1ssaratoga: since all the bitstream reading code is in the .h file different codecs can use different bitstream readers by just defining the right *BITSTREAM_READER define before including ffmpeg_bitstream.h
23:02:25n1sflac/shn doesn't use the vlc stuff
23:04:35pixelman1s: I already know how to make it work (not automatic though) and wanted an opinion if it would be o.k. to reuse the already existing \playertype option for this because it is also used elsewhere, e.g. the footer - or if I should invent another option and add it to the affected targets' platform files
23:04:46bluebrotherhmm. I'm wondering how this could be done best.
23:05:33pixelma"this"?
23:05:36 Quit bluebrother (Disconnected by services)
23:05:39 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
23:05:40n1sheh
23:07:13n1spixelma: ah, well, if you chose to reuse the \playertype option would you still be able to do a more specific match, say for a theoretical future clipv3
23:09:28pixelmayes, but you that would be true for a new option too. \playertype is currently "Clip+" for the Plus and I would have to change it - which means there could be place that could be wrong or inaccurate now (I believe I saw it in the page footer)
23:10:37pixelmas/you//
23:10:38bluebroth3ryou could define a \typefallback for each manual, so for the clipplus that would be clip.
23:11:06bluebroth3ri.e. add a specific fallback to the platform file
23:11:27pixelmadoes that mean you would prefer a new option? :)
23:16:14 Join robin0800 [0] (~quassel@cpc2-brig8-0-0-cust964.brig.cable.ntl.com)
23:16:32 Join [Saint] [0] (S_a_i_n_t@203.184.3.19)
23:17:33 Quit S_a_i_n_t (Ping timeout: 264 seconds)
23:19:13bluebroth3rpixelma: basically, yes.
23:19:29bluebroth3rso as far as I understand the problem we now have a fallback path like
23:20:07bluebroth3rbasename-resolution-modelname > basename-resolution
23:20:42bluebroth3rso in the clip case a fallback basename-resolution-modelname > basename-resolution-fallbackmodelname > basename-resolution would be in place, right?
23:20:49CIA-6New commit by nls (r27438): Remove two unused macros.
23:20:54pixelmayes
23:21:42bluebroth3rideally that would mean to introduce a new macro that defines that fallbackmodelname, and extend the screenshot macro to handle the additional cause. Having the screenshot macro to gracefully handle the case that the fallbackmodelname isn't defined would be good.
23:22:09bluebroth3rso you only need to add that in the cases where it's needed. Like with the \specimg, but that is always defined right now.
23:22:24CIA-6r27438 build result: All green
23:22:41bluebroth3rone could of course set that fallbackmodelname to \specimg in the cases where it isn't needed. But not needing to set it at all might be nice.
23:22:42pixelmaI already did that - and it falss through to the next in the priority list as it is now
23:22:56bluebroth3rnice.
23:24:55pixelmamy question was "only" if I would need to invent a new option - in your explanation called typefallback - or can reuse the already existing \playertype which needs checking for other places it is used
23:26:10bluebroth3rwell, I would go with the new option as it seems safer to me :)
23:27:21pixelmaok, me too (took me a while to think it through though). n1s, what about you?
23:28:13n1si'm not quite sure i'm awake enough for this :)
23:29:34n1salthough, "reusing" things for something they were not intended makes me a little uneasy
23:30:31pixelmaindeed. I'll go with that then
23:30:51pixelmathanks you two for taking the time :)
23:31:15 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.2.226)
23:31:42kugeln1s: I think you could have simply removed it from mips' system-target.h
23:31:46 Quit [Saint] (Ping timeout: 276 seconds)
23:32:41n1skugel: where would those functions come from then?
23:33:04kugelsystem.h, it includes system-target (or the other way around, not sure)
23:33:41 Quit bunnyboi (Quit: IRC is just multiplayer notepad)
23:33:42n1syeah, but i wanted to avoid polluting system.h with more target specific #ifdefing
23:34:40n1sor do you think that would be better?
23:34:58kugelwhy should #ifdefs be needed?
23:36:21kugelyou never include system-target.h directly, only system.h which has the functions. they shouldn't even exist in system-target.h (unless asm optimized) IIUC
23:37:53n1syeah, but if you have the functions in system.h and that includes syste-target.h which implements the functions again, the compiler will complain for sure, so #ifdefs are needed
23:38:34n1sso, the sawp* functions in system.h are guarded by #if (CONFIG_PLATFORM & PLATFORM_HOSTED)
23:38:35kugelwhich is why I suggest removing them from system-target.h
23:38:47kugeloh, right
23:39:11CIA-6New commit by bieber (r27439): Theme Editor: Added New Project feature
23:39:18kugelwell, an #ifdef is better than code duplication still, imo
23:40:07n1sin fact, a better approach is for any system-target.h file that implement these functions to define a PP symbol and then system.h checks for that
23:40:26kugelI just thought the same
23:40:45CIA-6r27439 build result: All green
23:41:10n1sTARGET_OPTIMIZED_BYTESWAPS?
23:43:18 Join FrankJohn [0] (~Anon331@pool-74-96-76-145.washdc.fios.verizon.net)
23:43:28FrankJohnhi room = )
23:43:45kugelor, maybe reverted logic (#define NEED_GENERIC_BYTESWAP in system-target.h), that way a new system-target.h creator would be notified because the build would fail
23:43:52FrankJohnnew KoRn is a little cool so far
23:44:30FrankJohnanyOne like rock music specifically, or are We in a computer chat room ?
23:44:44n1skugel: yes, also less modifications
23:44:52bluebroth3rFrankJohn: this channel is about Rockbox, a replacement firmware for mp3 players.
23:45:01FrankJohnany single ladies want to chat? = P
23:45:05bluebroth3rplease read the channel guidelines linked in the topic
23:45:11BagderFrankJohn: see topic
23:45:11FrankJohnoh I see
23:45:27 Part FrankJohn
23:45:42*bluebroth3r wonders what's the point of channel topics if people only ignore them :/
23:45:56kugelthere's little point in them
23:46:20pixelmathe thing is you usually won't notice people who do read them
23:46:35bluebroth3rIMO there _is_ point in them. Like the subject line in mails. People just need to pay attention ...
23:50:24 Quit robin0800 (Remote host closed the connection)
23:54:16 Quit Battousai (Remote host closed the connection)
23:54:31 Join Battousai [0] (~bryan@gentoo/developer/battousai)

Previous day | Next day