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 2010-10-15

00:01:27n1sthat's the one i meant, the animated part is a slightly different shade of gray from the background but it is indeed hard to see
00:03:02Luca_Snot a big deal however, that's just ui candy, and the fuze recharges quite quickly :)
00:05:58 Join T44 [0] (~Topy44@cable-78-34-73-174.netcologne.de)
00:06:14 Quit Topy44 (Ping timeout: 252 seconds)
00:07:59 Quit n1s (Quit: Lmnar)
00:09:00S_a_i_n_tStill, it shouldn;t be soo hard to see that one needs to question if they're actually seeing it or not.
00:09:14S_a_i_n_tIt indeed wasn;t what I thought it was, but I'll still look at it.
00:09:28S_a_i_n_tIf the colours need to be more defined, then, they should be.
00:09:39 Quit captainkewllllll (Quit: Page closed)
00:11:10 Quit leavittx_ (Ping timeout: 240 seconds)
00:11:55Luca_Sgotta go now, goodnight
00:12:06 Quit Luca_S (Quit: CGI:IRC)
00:12:21*alexbobP is glad he uses boxamp theme
00:12:50alexbobPthe best thing about the boxamp theme is I can tell people I'm running winamp on my mp3 player and they will believe me just by seeing it
00:12:56alexbobPmost people don't do a whole lot of critical thinking
00:20:52 Quit domonoky (Read error: Connection reset by peer)
00:25:04 Join Zambezi_ [0] (Zulu@80.67.9.2)
00:25:48 Quit Zambezi_ (Changing host)
00:25:48 Join Zambezi_ [0] (Zulu@unaffiliated/zambezi)
00:26:48 Join Ypsy_ [0] (~ypsy@geekpadawan.de)
00:27:30 Quit YPSY (Ping timeout: 260 seconds)
00:27:30 Quit Zambezi (Read error: Connection reset by peer)
00:27:30 Quit drizztbsd (Read error: Connection reset by peer)
00:27:47 Nick Zambezi_ is now known as Zambezi (Zulu@unaffiliated/zambezi)
00:28:21 Join drizztbsd [0] (~quassel@unaffiliated/drizztbsd)
00:36:18 Quit mirak_ (Quit: Ex-Chat)
00:49:35 Quit S_a_i_n_t (Ping timeout: 240 seconds)
00:52:43 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.2.165)
00:57:28 Join JdGordon| [0] (~jonno@vl10.gw.ok-labs.com)
00:57:28 Quit JdGordon| (Changing host)
00:57:28 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
01:00
01:01:57 Join drizztbsd_ [0] (~quassel@unaffiliated/drizztbsd)
01:03:18 Quit drizztbsd (Read error: Connection reset by peer)
01:04:09 Nick drizztbsd_ is now known as drizztbsd (~quassel@unaffiliated/drizztbsd)
01:06:10***Saving seen data "./dancer.seen"
01:11:01 Join tvelocity [0] (~tony@weg38-1-82-237-37-150.fbx.proxad.net)
01:11:44 Quit pamaury (Remote host closed the connection)
01:14:00 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
01:23:19JdGordon|anyone awake who uses the playlist catalog?
01:26:41 Quit tvelocity (Quit: Αποχώρησε)
01:27:43JdGordon|meh, ok, apparently one of the reasons the catalog doesnt use the file browser is because it puts the most recent playlist at the top of the list. I wonder if it would be acceptable to add a new catalog menu item for "Add to: <last m3u filename>" so the browser doesnt need to display it
01:28:11JdGordon|that could even cut 2 key presses of that action
01:29:18 Part toffe82
01:30:54 Quit DerPapst (Quit: Leaving.)
01:33:45 Join CGL [0] (~CGL@190.207.235.80)
01:36:59 Join anewuser [0] (kvirc@unaffiliated/anewuser)
01:39:09 Quit S_a_i_n_t (Ping timeout: 276 seconds)
01:39:37 Join elinenbe_ [0] (~elinenbe@m4b2d36d0.tmodns.net)
01:41:18 Join soap [0] (~soap@rockbox/staff/soap)
01:42:33 Quit elinenbe (Ping timeout: 276 seconds)
01:42:33 Nick elinenbe_ is now known as elinenbe (~elinenbe@m4b2d36d0.tmodns.net)
01:52:15 Quit soap (Ping timeout: 272 seconds)
01:56:11 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.9)
02:00
02:09:09 Quit RoC_MasterMind (Ping timeout: 255 seconds)
02:10:15 Join RoC_MasterMind [0] (~Free@c-76-122-36-52.hsd1.fl.comcast.net)
02:24:47 Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
02:35:53 Join elinenbe_ [0] (~elinenbe@207-237-241-192.c3-0.80w-ubr1.nyr-80w.ny.cable.rcn.com)
02:38:27 Quit elinenbe (Ping timeout: 276 seconds)
02:38:27 Nick elinenbe_ is now known as elinenbe (~elinenbe@207-237-241-192.c3-0.80w-ubr1.nyr-80w.ny.cable.rcn.com)
02:41:05 Quit panni_ (Quit: ( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de ))
02:46:29 Quit stripwax (Read error: Connection reset by peer)
02:52:48 Join panni_ [0] (hannes@ip-178-203-81-220.unitymediagroup.de)
03:00
03:02:02 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
03:06:13***Saving seen data "./dancer.seen"
03:13:39 Quit Zarggg (Quit: Zarggg)
03:24:53 Quit Kitar|st ()
03:28:08 Join Zarggg [0] (~zarggg@24.229.139.169.res-cmts.sm.ptd.net)
04:00
04:03:48 Quit pixelma (Disconnected by services)
04:03:51 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:03:53 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:04:53 Quit amiconn (Disconnected by services)
04:04:54 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:05:14 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:08:19 Quit TheSeven (Ping timeout: 265 seconds)
04:11:58 Quit RoC_MasterMind (Quit: Leaving)
04:12:30 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:22:57 Quit Judas_PhD (Quit: This is a quitting message)
04:27:21 Quit edboyer93 ()
04:27:23 Quit Llorean (Read error: Connection reset by peer)
04:37:24 Join Barahir_ [0] (~jonathan@frnk-590fe366.pool.mediaWays.net)
04:41:03 Quit Barahir (Ping timeout: 264 seconds)
04:45:52 Quit bluebrother (Disconnected by services)
04:45:53 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
04:48:22 Quit dys (Ping timeout: 276 seconds)
04:48:28 Join dys` [0] (~andreas@krlh-5f726709.pool.mediaWays.net)
04:49:06 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
04:55:16 Quit panni_ (Quit: ( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de ))
05:00
05:06:15***Saving seen data "./dancer.seen"
05:26:57S_a_i_n_tHas the new Nano2G bootloader been added to RBUtil?
05:27:56S_a_i_n_tI've installed the latest build, but I can't boot the OF using menu so I'm guessing RBUtil is installing an older bootloader
05:28:05 Quit S00row1 (Read error: Connection reset by peer)
05:29:31 Join S00row [0] (~Administr@27-33-98-164.static.tpgi.com.au)
05:33:36 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
05:33:43 Quit ps-auxw (Ping timeout: 276 seconds)
05:39:45 Quit tchan (Quit: WeeChat 0.3.3-dev)
05:44:54 Join ps-auxw [0] (~arneb@p4FF7EB37.dip.t-dialin.net)
05:45:18 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
06:00
06:11:29 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
06:18:40 Quit shai (Quit: Leaving)
06:18:42 Join leavittx_ [0] (~lev@89.221.199.187)
06:21:19 Quit anewuser (Quit: omg unicode is harder than the sharp tip of a diamond drilled to the head ;()
06:26:18 Quit leavittx_ (Ping timeout: 240 seconds)
06:27:15 Quit Horscht (Quit: Verlassend)
06:39:40 Quit BHSPitMonkey (Quit: Ex-Chat)
06:42:40 Quit antil33t ()
06:43:45 Nick dys` is now known as dys (~andreas@krlh-5f726709.pool.mediaWays.net)
06:47:31 Join hebz0rl [0] (~hebz0rl@dslb-088-065-049-194.pools.arcor-ip.net)
06:53:26 Quit S_a_i_n_t (Disconnected by services)
06:53:29 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.208)
07:00
07:06:16***Saving seen data "./dancer.seen"
07:10:55 Quit Judas_PhD (Remote host closed the connection)
07:12:34 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
07:13:18 Quit Judas_PhD (Client Quit)
07:21:13 Quit sasquatch (Ping timeout: 252 seconds)
07:27:18 Quit mc2739 (Ping timeout: 240 seconds)
07:29:22 Join mc2739 [0] (~mc2739@71.20.73.59)
07:29:22 Quit mc2739 (Changing host)
07:29:22 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
07:37:57 Join sasquatch [0] (~username@p4FF2CB7A.dip.t-dialin.net)
07:49:44 Quit hebz0rl (Quit: Leaving)
07:56:06 Quit fyrestorm (Read error: Connection reset by peer)
07:56:32 Join fyrestorm [0] (~nnscript@cpe-69-203-144-35.si.res.rr.com)
07:57:43 Quit JdGordon| (Quit: leaving)
08:00
08:01:58 Join Kohli [0] (~kohlrabi@kohlio.de)
08:02:14 Quit Kohlrabi (Ping timeout: 240 seconds)
08:02:28 Quit dionoea (Ping timeout: 252 seconds)
08:03:55 Join dionoea [0] (~dionoea@yop.chewa.net)
08:06:48 Join JdGord [0] (~jd@122.110.128.169)
08:10:39 Join kugel [0] (~kugel@rockbox/developer/kugel)
08:17:22 Quit sasquatch (Ping timeout: 265 seconds)
08:21:16 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:28:03 Quit linuxstb (Ping timeout: 245 seconds)
08:30:40 Join sasquatch [0] (~username@p4FF2D591.dip.t-dialin.net)
08:35:39 Join leavittx_ [0] (~lev@host-97-146-66-217.spbmts.ru)
08:38:19 Quit TheSeven (Ping timeout: 240 seconds)
08:38:59 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
08:42:23 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
08:43:00 Join ender` [0] (krneki@foo.eternallybored.org)
08:43:43 Quit Llorean (Quit: Leaving.)
08:46:56 Quit Judas_PhD (Quit: This is a quitting message)
08:50:33 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
08:55:14 Quit JdGord (Quit: Bye)
09:00
09:05:40 Quit preglow (Ping timeout: 264 seconds)
09:06:17***Saving seen data "./dancer.seen"
09:06:43 Join LinusN [0] (~linus@giant.haxx.se)
09:06:44 Quit LinusN (Changing host)
09:06:44 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:09:27 Join Rob2223 [0] (~Miranda@p4FFF06BE.dip.t-dialin.net)
09:13:27 Quit Rob2222 (Ping timeout: 265 seconds)
09:16:13 Join petur [0] (d408b802@rockbox/developer/petur)
09:17:02 Quit elinenbe (Read error: Connection reset by peer)
09:23:06JdGordonwould anyone find it too strange if a menu item (in the catalog viewer) had the text "Add to: <last playlist filename>" and spoke "Add to last playlist" (i.e the text had the actual filename, the voice didnt)
09:23:19 Join Kitar|st [0] (Kitarist@BSN-143-101-140.dial-up.dsl.siol.net)
09:27:23 Join kevku [0] (~kevku@arch.tunnel.ipv6.estpak.ee)
09:33:58 Quit kevku (Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/)
09:38:36 Quit leavittx_ (Ping timeout: 240 seconds)
09:49:40 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
09:57:15 Part helmut
09:59:03 Quit sasquatch (Quit: WeeChat 0.3.2)
09:59:29 Join sasquatch [0] (~username@p4FF2D6B4.dip.t-dialin.net)
10:00
10:04:14 Join swilde [0] (~wilde@aktaia.intevation.org)
10:26:57 Join DerPapst [0] (~Alexander@p4FE8F695.dip.t-dialin.net)
10:27:26 Join kevku [0] (~kevku@mi10065.gprs.emt.ee)
10:37:10 Join casainho [0] (~chatzilla@pal-213-228-181-14.netvisao.pt)
10:39:43 Quit shai (Ping timeout: 245 seconds)
10:46:00 Quit DerPapst (Quit: Leaving.)
10:47:20 Quit kevku (Ping timeout: 250 seconds)
10:52:27 Quit casainho (Read error: Connection reset by peer)
10:54:59 Quit markun (Read error: Connection reset by peer)
10:57:22kugelI updated my test codec stats with ARMv6 results: http://www.alice-dsl.net/simonemartitz/rockbox/test_codec_stats.pdf
10:59:08kugelthe speed up from building for ARMv6 is rather marginal; except for ape with asm enabled
11:00
11:00:14 Join markun [0] (~markun@5ED33C2C.cm-7-4a.dynamic.ziggo.nl)
11:00:15 Quit markun (Changing host)
11:00:15 Join markun [0] (~markun@rockbox/developer/markun)
11:04:05 Quit kugel (Ping timeout: 276 seconds)
11:04:55 Quit TheSeven (Remote host closed the connection)
11:06:18***Saving seen data "./dancer.seen"
11:09:12 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:09:42kugelhowever, in asm enabled builds a armv5 one is sometimes even faster than armv6, I assume test_codec on a hosted environment isn't entirely accurate (due to task switching?)
11:14:12 Quit kugel (Remote host closed the connection)
11:14:34 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
11:17:20 Quit saratoga (Ping timeout: 265 seconds)
11:21:47 Join DerPapst [0] (~Alexander@188.107.197.224)
11:28:15 Quit JdGordon (Ping timeout: 240 seconds)
11:33:02 Join esperegu [0] (~quassel@145.116.15.244)
12:00
12:06:53 Part LinusN
12:08:39 Join LinusN [0] (~linus@rockbox/developer/LinusN)
12:11:56 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:34:46 Nick Kohli is now known as Kohlrabi (~kohlrabi@kohlio.de)
12:44:53 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:49:40 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
13:00
13:06:19***Saving seen data "./dancer.seen"
13:16:52 Join piotrekm [0] (~piotrek@unaffiliated/piotrekm)
13:24:19 Join elinenbe [0] (~elinenbe@207-237-241-192.c3-0.80w-ubr1.nyr-80w.ny.cable.rcn.com)
13:38:12 Part LinusN
13:49:32 Quit kugel (Ping timeout: 252 seconds)
13:50:38 Quit piotrekm (Quit: piotrekm)
14:00
14:18:22 Join leavittx_ [0] (~lev@89.221.199.187)
14:23:35 Join teru [0] (~teru@KD059133111160.ppp.dion.ne.jp)
14:25:26 Quit Dreamxtreme (Quit: I was raided by the FBI and all I got to keep was this lousy quit message!)
14:36:37 Quit petur (Quit: messing with work pc to get it working again :(((()
15:00
15:03:03 Join pSiKO [0] (~tbox@213.56.158.253)
15:06:22***Saving seen data "./dancer.seen"
15:07:50 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:14:46 Join n1s [0] (~n1s@nl118-174-240.student.uu.se)
15:14:47 Quit n1s (Changing host)
15:14:47 Join n1s [0] (~n1s@rockbox/developer/n1s)
15:35:01 Quit user890104 ()
15:35:45 Part pSiKO
15:41:44 Join user890104 [0] (Venci@Venci-Notebook-LAN.ipv6.6bez10.info)
15:46:38 Join linuxstb [0] (~linuxstb@59.230.19.95.dynamic.jazztel.es)
15:46:39 Quit linuxstb (Changing host)
15:46:39 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
15:47:46 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
15:49:48 Quit teru (Quit: Quit)
15:51:00S_a_i_n_tHmmm, I've tried it twice now...and it seems that rockboxdev.sh chokes building the arm-elf-eabi toolchain on CygWin with line 198 "rm -rf $builddir", it produces the error "rm: cannot remove `/tmp/rbdev-build/build-binutils': Device or resource busy"
15:51:26S_a_i_n_tAnother user saw this the other day also, is it possible to cod ethis to skip on CygWin machines?
15:51:35S_a_i_n_t*code
15:52:55n1sprobably :)
15:53:23Tornealternatively you could fix the actual problem? :)
15:53:24S_a_i_n_terrr....s/ist it possible/will it happen/ ;)
15:53:46S_a_i_n_twell, I have no idea of the actual problem.
15:53:49Tornesomething is presumably still running with its cwd inside there, or similar
15:53:57Torneprocess explorer's handle search will tell you
15:54:25 Join panni_ [0] (hannes@ip-178-203-81-220.unitymediagroup.de)
15:54:30 Join teru [0] (~teru@KD059133111160.ppp.dion.ne.jp)
15:54:31S_a_i_n_tas soon as it errors, it quits...so I can't tell to much from it.
15:54:59S_a_i_n_tit's just cleanup BS anyway, no reason why it can't be left there to be cleaned up manually.
15:55:22Tornethat's questionable logic
15:55:35Torneif the build and install is really done then nothing should have any files open in there any more
15:55:42Torneso maybe something *else* has gone wrong
15:55:55Tornesolving the real problem is better
15:55:56S_a_i_n_tBut, it's not *done*
15:56:01S_a_i_n_tit's pretty far from the end.
15:56:11Tornethe build and install must be done if it's deleting the build directory
15:56:13S_a_i_n_tit's some intermediary cleanup.
15:56:21S_a_i_n_tNope.
15:56:26S_a_i_n_tit's far, far from done.
15:56:34Tornethat doesn't make any sense
15:56:36S_a_i_n_t(at that stage)
15:56:57S_a_i_n_twhich is possibly why its failing.
15:57:01TorneIt *is* done
15:57:09Torneif you mean the last line in build()
15:57:17Torneit has built and installed the tool it was asked to build and install.
15:57:25Torneso then it deletes the build directory for that tool
15:57:36Torneso, there is no legitimate reason for anything to be holding that directory open
15:57:42Tornewhich means something has already gone wrong :)
15:57:59S_a_i_n_tbut, commenting out that line
15:58:02S_a_i_n_tit completes.
15:58:07Tornethat doesn't mean it did it properly
15:58:11S_a_i_n_tthere's still a LOT of shit to build after that.
15:58:15S_a_i_n_t(apparently)
15:58:19Torneit's *probably* fine, but you can't know that for sure
15:58:29S_a_i_n_tthe script continues for AGES after that if commented.
15:58:30Torneeach part of the toolchain is a totally seperate package
15:58:41Torneeach one is downloaded, unpacked, built, installed, and then the build directory completed
15:58:50Tornethe fact that the script isn't done then is not a surprise
15:58:53Tornebinutils is the first thing to be installed
15:59:00Tornebut it *is* done installing binutils
15:59:07Torneso it's perfectly reasonable for it to delete the directory it built binutils in
16:00
16:00:17Tornerockboxdev.sh is a script that builds two seperate things, binutils and gcc.
16:00:28S_a_i_n_tbut then it goes on to create the same dir again.
16:00:32Torneno it doesn't.
16:00:37S_a_i_n_tto build more stuff in it.
16:00:41Torneif it did, that would be broken
16:00:41amiconnS_a_i_n_t: It worked fine for me. Are you running rockboxdev.sh with administrative permissions?
16:00:46S_a_i_n_tNo reason to remove it at that stage really.
16:00:54S_a_i_n_tamiconn: yes.
16:01:15Torneit removes build-binutils
16:01:18Torneand creates build-gcc instead
16:02:07S_a_i_n_tamiconn: Hillshum(sp?) had exactly the same error approximately 2 days ago
16:02:33S_a_i_n_tif I comment line 198, it completes.
16:02:45S_a_i_n_tif not it will always fail at that point.
16:03:50 Join noamsml_ [0] (~noamsml@adsl-75-45-238-177.dsl.sfldmi.sbcglobal.net)
16:06:17 Quit noamsml (Ping timeout: 250 seconds)
16:08:09 Quit noamsml_ (Ping timeout: 252 seconds)
16:08:50 Join noamsml [0] (~noamsml@adsl-75-45-243-168.dsl.sfldmi.sbcglobal.net)
16:16:33TorneS_a_i_n_t: anyway, the problem is that it's the build script itself holding the directory open :)
16:16:53S_a_i_n_tAha....so?
16:17:03Torneso it should leave the directory first before it tries to delete it :)
16:17:16S_a_i_n_tI've just been commenting it out, but I assume there's a better fix than that ;)
16:17:32Torneyes
16:17:36Tornegimme a sec
16:17:45S_a_i_n_ttimes up!
16:18:48Tornehush you
16:19:25Torneokay i can't fix it right now, my dsl has gone down at home
16:19:45S_a_i_n_tNo worries.
16:20:03S_a_i_n_tShould I bother adding a bug report, or do you think you'll remember about it?
16:21:13Tornei'll do it later, it's ok
16:21:44Tornezagor slightly changed the order it does stuff when he refactored out some of the duplicated code in r26600
16:21:57Torneon unix it's not a problem to delete your current working directory :
16:21:59S_a_i_n_tSweet as. And, thanks. It's good to know I'm not going insane doing something wrong, and it is indeed borked ;)
16:22:04Torneit just doesn't actually go away until you leave
16:22:08Torneon windows it fails
16:22:19S_a_i_n_tYeah, coll.
16:22:25S_a_i_n_tThanks for taking a look at it.
16:22:27 Join kugel [0] (~kugel@rockbox/developer/kugel)
16:22:28Torneright
16:22:28S_a_i_n_t*cool
16:23:25Tornethe reason to remove it at that stage is just to save disk space, really
16:23:52Tornewhich is a reasonable thing to do, it just needs to cd to a different directory first
16:32:58n1swhatever happened to the freeze/release decision?
16:34:39 Part blithe
16:34:46 Quit ranma (Remote host closed the connection)
16:38:52 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
16:39:28 Join kazaik [0] (~kazaik@pool-71-166-29-66.bltmmd.east.verizon.net)
16:45:06amiconnTorne: That explains it. I certainly ran that script before r26600
16:46:00 Quit kugel (Ping timeout: 245 seconds)
16:47:12 Quit TheSeven (Ping timeout: 252 seconds)
16:51:11 Join gc5 [0] (~4141dfdd@giant.haxx.se)
16:56:54 Join toffe82 [0] (~chatzilla@maf.wirelesstcp.net)
17:00
17:01:31gc5a app shoulde made for iphone ipod ipad
17:02:01CIA-7New commit by 03nls (r28283): Pitch detector: do not break profiling builds.
17:02:48gc5what?
17:04:02CIA-7r28283 build result: All green
17:04:42 Part Zagor
17:04:54TorneCIA is a bot announcing changes to the code, it's not replying to you :)
17:05:04Torneif you want an iphone app, feel free to write one :)
17:05:22 Quit teru (Quit: Quit)
17:06:23***Saving seen data "./dancer.seen"
17:07:04gc5i dont have a mac so iant not yet anyways
17:08:37*n1s remembers something
17:09:33n1sprofiling doesn't work with test_codec
17:10:54n1sit does set up the function pointers in the api, so that *should* work...'
17:11:53 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:14:18n1sregular playback is slower than realtime though so no real gain from test_codec
17:16:02 Join Dreamxtreme [0] (~Dre@92.30.207.138)
17:16:22gc5my ipod 5 's baclight exploded while ising rockbox
17:17:05gc5it wasent rockboxes fault some idiot repaired it and used a contact explosive as glue
17:19:53 Join _s1gma [0] (~d.d.derp@77.107.164.131)
17:22:48S_a_i_n_tgc5: This channel is reserved for reality only. ;)
17:25:14gc5yes i was lying. the reality is the idiot (me) riped the bcklight out when i was replacing the screen :( now it is verry dark
17:26:15S_a_i_n_tAnd besides, not having a macintosh doesn;t stop you from porting RaaA to the iPhone/Pad
17:27:38gc5RaaA ?? what is RaaA
17:27:54S_a_i_n_tRockbox as an Application.
17:28:03S_a_i_n_tessentially what you are talking about.
17:29:04gc5oh does it allready exist as a iphone app
17:29:28 Quit gc5 (Quit: CGI:IRC)
17:29:32gevaertsno
17:29:33S_a_i_n_tNo, this is why I said not having a mac doesn;t stop you from porting it ;)
17:30:45 Join gc5 [0] (~4141dfdd@giant.haxx.se)
17:31:12S_a_i_n_tgc5: No, there is not a port for iPhone...only Android.
17:32:14gc5darn i have a ipod touch i want to be able to use rockbox on it becaus it was awsone in my ipod 5
17:32:17 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:32:59S_a_i_n_tWell, learn to code and write the port for RaaA to iPod touch then ;)
17:33:14S_a_i_n_tIt won't code itself.
17:33:45gc5how axn i make apps for a non jailbroken ipod with windows
17:34:13S_a_i_n_tNo idea, and that is completely offtopic here
17:34:49gc5oops sorry
17:35:04krazykitgc5, it's been discussed in here that it may be a violation of the GPL (rockbox's license) to distrube rockbox in the iphone app store (if it were even able to be approved)
17:35:17gc5i see
17:35:28gc5so maby cidea
17:36:11krazykitbut before that could even happen, someone with programming experience would have to port RaaA to the iOS platform, which would include lots of programming and time
17:37:09n1skrazykit: it won't be accepted into the app store
17:37:37gc5its not hapening anytime soon
17:37:48gc5maby never
17:37:50krazykitn1s, that's why i said "if it were" :)
17:38:04krazykitgc5, it's not happening as long as nobody is working on it
17:39:38gc5well i cant i mean i cant even program in qbasic
17:42:53 Quit krazykit (Quit: campus)
17:47:26 Quit gc5 (Quit: CGI:IRC)
17:55:22n1shmm, render_line_unrolled takes a whooping 13% of the total ticks in 128k decoding of vorbis on cf, of course profiling disables inlining, should probably test with mild inlining too
18:00
18:03:54TorneDid we come to a decision about what to do with the USB device ID changing thing?
18:04:06Tornethe issue peopl ehave with devices that expect to read the itunesdb over USB when they notice the device is an ipod
18:06:10S_a_i_n_tHmmmm? Oh, like iTunes opening with "Open iTunes when this iPod is detected" setting turned on, etc?
18:06:20Tornewell more for things like car head units and hifis
18:06:31Tornebecause on those you can't jus tuse a different media player program that's not dumb
18:06:32S_a_i_n_tOh, right.
18:06:43S_a_i_n_ttrue.
18:06:57Tornethey see the apple device IDs and they expect to be able to talk that weird SCSI-transported protocol where they can read the iTunesDB
18:07:02Tornewhich we don't implement :)
18:07:14Tornebecause they assume that on an apple device they won't be able to find any of the files otherwise.
18:07:20Torneand fail to give you the optoin of just browsing hte filesystem
18:07:25S_a_i_n_tSo, what is the solution?
18:07:44Tornethe "solution" for individual users is to compile a build that uses a different USB VID/PID
18:07:54Torneso the head unit doesn't recognise it as an ipod, and just mounts it as a usb flash drive
18:08:04Tornesince they do normalyl support non-ipod usb devices as well :)
18:08:21S_a_i_n_tCould there be a user setting for this on the DAP??
18:08:23Tornebut last time this was discussed people were reluctant to actually put the code to do that into the build
18:08:26S_a_i_n_t-?
18:08:26TorneWell, there could be, yes
18:08:37Tornebut the problem is that, well, you aren't supposed to use VIDs/PIDs that don't belong to you :)
18:08:54Torneand also, some OSes use those to identify things like what drive letter to mount as, so being able to change them can result in confusing behaviour to users
18:08:55S_a_i_n_tAh.
18:09:10Tornewe currently assume that nobody cares if we reuse the VID/PID that the device's OF uses
18:09:16Tornesince it is the same device, even if it's not hte same softrware :)
18:09:44Tornebut that's not so good in this case because our capabilities are not hte same as the device's normal software.
18:10:11S_a_i_n_tHow strict is the "aren't supposed to"? ;)
18:10:27Tornewell i don't know that the USB-IF can really *do* anything to us :)
18:10:44Tornethe usual way they enforce their rules is you are only allowed to use the USB trademarks (like the logo) if you follow their rules ;)
18:10:49S_a_i_n_t\o/
18:11:07Tornewe certainly can't afford to pay to join the USB-IF or to get a PID as an independant, costs loads
18:11:37Torneletting the user set it to any value kinda absolves us of responsibility but it's not very helpful to the people who actually need the feature
18:11:55Tornesince telling them "oh just pick some random number" is not very helpful, what if that clashes with something one of their devices/OSes recognises already?
18:12:21 Quit S00row (Read error: Connection reset by peer)
18:12:23Torne(you can kinda avoid that by using the VID/PID of a *different* rockbox device that isn't an ipod.. ;)
18:14:18 Join S00row [0] (~Administr@27-33-98-164.static.tpgi.com.au)
18:14:25alexbobPwell rockbox doesn't really need to use the usb logo anyways :P
18:14:38alexbobPmight as well just jack the id from some other player
18:17:20TornealexbobP: right, that's the point; it's fine for an individual user to build their own build that does that
18:17:35Tornebut it's not necessarily a good idea to just make that a setting, is the thing
18:17:55Tornethe potential for weird/bad stuff to happen is certainly present...
18:18:01alexbobPTorne: if we already know that the current state of affairs does the Wrong Thing (tm), I don't see why not
18:18:18Tornewha?
18:18:28TorneThe current behaviour is fine/correct for almost everyone
18:18:38alexbobP11:03:30 < Torne> the issue peopl ehave with devices that expect to read the itunesdb over USB when they notice the device is an ipod
18:18:57alexbobPisn't that a problem for anyone with a rockbox ipod?
18:19:02Torneno..
18:19:25Torneit's a problem for anyone with a USB host device which assumes that iPods should only be accessed via the itunesdb remote protocol
18:19:34Torneand doesn't give you any way of configuring that
18:19:41Tornewhich is a tiny number of people, tbh ;)
18:19:45alexbobPah, okay
18:19:55Tornesince it basically boils down to "some expensive car head units" and "some expensive hifis/amps"
18:20:05alexbobPI don't see what the potential is for problems if an ipod looks like it's a different msd device
18:20:59Tornewell, apart from anything else, people seem to like to fiddle with settings without knowing what they mean or do :)
18:21:15Torneand changing the device id will cause it to behave differently on many OSes the next time you plug it into the PC
18:21:15alexbobPof course, it's fun :D
18:21:29alexbobPwouldn't it just behave like a flash drive?
18:21:29Tornebe assigned a different drive letter, have its autoplay/whatever done differently, etc
18:21:38alexbobPhehe, drive letters...
18:21:42Torneit would behave like a *different* flash drive
18:21:52Tornewhereas normally it's always the same device :)
18:22:00alexbobPis it a problem when it gets a different drive letter? :P
18:22:20Tornemight be. users don't generally expect the identity of a device to change once they've already used it
18:22:29Tornei'm not saying it's a crippling issue or anything
18:22:32n1sah, restored inlining got that down to 2% :)
18:22:39Tornebut it's some potential confusion ;)
18:22:44alexbobPwell okay
18:22:54alexbobPit's of no consequence to me, since I tend to shirk away from buying apple products
18:23:23alexbobPSansa seems to be doing nothing to try to stop rockbox, and I like their players
18:23:51Tornenobody's ever really done anything to try to stop rockbox specifically :)
18:23:55 Nick m|c is now known as miceh (~mtq@h1439481.stratoserver.net)
18:23:58gevaertsTorne: actually, we don't just assume that nobody cares if we reuse the VID/PID that the device's OF uses. Llorean (IIRC it was him) sent an email to the USB-IF explaining the situation, and they said this is fine
18:23:59 Join casainho [0] (~chatzilla@bl16-85-185.dsl.telepac.pt)
18:24:05TorneOh, ok
18:24:09Tornewell that's even better :)
18:24:26Tornebut that doesn't speak to their opinoin on letting users set it to something else..
18:24:43alexbobPtheir opinion on letting *users* set it to something else is irrelevant :P
18:24:43gevaertsexactly
18:25:00gevaertsWell, technically lots of systems allow setting this
18:25:15gevaertsIf you have a windows mobile device, you can change VID and PID
18:25:35TorneWell, sure
18:25:36gevaertsAll you need is a registry editor :)
18:25:39alexbobPTorne: maybe nobody is specifically attacking rockbox, but doesn't apple design their players to only accept firmware that they signed or something?
18:25:55gevaertsalexbobP: yes, but that's not against rockbox
18:26:04TornealexbobP: yes. but that's just the way they do their next generation of hardware
18:26:11alexbobPgevaerts: it's against homebrew in general, and rockbox is homebrew
18:26:17gevaertsnot really
18:26:19Tornesecure boot chain is pretty much a standard thing you do on embedded hardware these days
18:26:23gevaertsThat's against people who modify the firmware to circumvent DRM on their music and apps
18:26:26Torneindeed
18:26:32alexbobPhahahahaha
18:26:42alexbobPI cannot fathom the concept of somebody who decides their best bet to break drm
18:26:42Torneyou laugh but people do that
18:26:45alexbobPis to tamper with device firmware
18:26:51Tornethere are loads of patched apple firmwares for ipods around
18:26:55alexbobPdrm is always so easy to break directly on your computer!
18:27:13Tornegevaerts: well, things like the FT232 let you just set it :)
18:27:21alexbobPanyways the concept of drm is retarded. I don't respect apple anymore because they're trying to "protect drm" instead of just stick it to rockbox :P
18:27:22TornealexbobP: you don't understand what we're talking about, then :)
18:27:48alexbobPTorne: um, what? I'm familiar with "fair"play.
18:27:52Tornegevaerts: actually FTDI give you a PID under their VID if you ask for one
18:28:09Tornegevaerts: but you are supposed to be a bona-fide FTDI customer using FTDI chips
18:28:09Torne:)
18:28:29TornealexbobP: see the iphone/ipodtouch/etc where you can install apps.. you can't break the DRM for those apps on the computer because they are checked on the device
18:28:33Torneyou have to do something to the device
18:28:51alexbobPTorne: oh, I was just thinking about music drm
18:28:53alexbobPTorne: I see what you mean
18:29:07Tornethere are other restrictions, yes
18:29:12alexbobPy'know, theoretically you could still crack any app on your computer
18:29:17Torneno you can't.
18:29:21alexbobPbut that's harder than modifying firmware
18:29:23alexbobPTorne: of course you can
18:29:26Tornethis is offtopic, anyway
18:29:31alexbobPTorne: any app is just a collection of data, and if you modify it you can do what you want
18:29:52gevaertsalexbobP: *if* you can get it to load
18:31:05Tornegevaerts: so, yeah, i feel like we should do *something* for the people who have these devices that need the VID/PID changing to be usable
18:31:11Tornegevaerts: but not sure what the sensible option is :)
18:31:27gevaertsTorne: I think the main blocker is that we need a good not too difficult way for users to set a different VID/PID pair, while at the same time not inviting people to play with that setting for the fun of it (because of the "interestin" effects it can have), *and* without us suggesting a VI/PID pair
18:31:49alexbobPare there any Apple brand flash drives? you should steal the VID/PID from that
18:31:55alexbobPwhy is it a problem if users want to play with the setting?
18:32:08Tornegevaerts: well, you could just have it as a config setting with no menu option
18:32:40gevaertsalexbobP: why would Apple-brand flash drives be better than anything else? And besides, *we* can't suggest any set of numbers
18:33:07S_a_i_n_tyou could just have a menu somewhere, with VID/PID1 and VID/PID2
18:33:29gevaertsTorne: that's one way, but most people don't really like file-only settings
18:33:49Tornegevaerts: by people do you mean users or us? :)
18:33:54gevaertsalexbobP: because it can cause all sorts of weirdness, leading to masses of support issues
18:33:57gevaertsTorne: devs :)
18:34:05Tornei was hoping in general that a user who needs this feature doesn't need to change it *back* ever
18:34:11Tornei.e. they don't also use anything that depends on it being the apple one
18:34:23Tornei don't think it's unreasonable to have a file-only setting for *this*
18:34:26Tornein general sure, that sucks
18:34:33Tornebut this is a pretty weird and obscure thing
18:34:39gevaertsI agree, but you know this "precedent" thing :)
18:34:54alexbobPgevaerts: if we don't suggest a set of numbers people are not likely to know which ones are even for MSD devices
18:35:07alexbobPgevaerts: and I just figured an ipod behaving as an apple flash drive would seem fitting...
18:35:21alexbobPat least it would show up as the correct color in the little icon popping up on your mac ;)
18:35:27Torneit doesn't matter what kind of device it's for
18:35:30 Join Strife89TX [0] (~cstrife89@207.144.201.128)
18:35:31Tornein general.
18:35:32gevaertsalexbobP: that's the whole point! Using VID/PID to decide how to talk to a device is *broken*!
18:35:40Tornethe VID/PID are not what's generally used to decide that :)
18:35:48Tornethe device seperately says "i am a mass storage device"
18:35:51Tornethat's a different thing
18:35:58alexbobPwell that's good, because that seems about as reasonable as paying attention to filename extensions
18:36:02gevaertsAnd using the Apple VID would likely not even fix the actual issue for some people
18:36:05Torneyou just need to pick a VID/PID that your OS doesn't already have a specific driver for.
18:36:14Tornebasically.
18:36:14alexbobPbut since the workaround is for devices that *do* pay attention to the VID, it seems like what VID you use might be important :P
18:36:25TornealexbobP: they are only looking for one vendor, though :)
18:36:25 Join s1gma_ [0] (~d.d.derp@77.107.164.131)
18:36:33Tornethey almost certainly treat every *other* vid identically
18:36:41Torneapple or not-apple
18:37:36gevaertsAnd we as a project can *not* pick a VID/PID pair. To do that, we either have to steal one from someone else which could well lead to trouble, or pay $2000
18:37:42gevaerts(unless prices have changed)
18:37:58alexbobPwhat if it's randomized?
18:38:00Strife89TXtimccc: For now the only way to change the font in the Stopwatch is to change the entire skin's font manually.
18:38:02TornealexbobP: same thing
18:38:11gevaertseven worse actually
18:38:17TornealexbobP: the point you are missing here is we don't want to do that because of a sense of professional responsibility
18:38:23Tornenot because we are afraid of getting in trouble
18:38:24timcccthanks Strife89TX
18:38:29Tornestandards bodies are a good thing and exist for a reason
18:38:44Strife89TXtimccc: Settings −−> Theme Settings −−> Font
18:38:50alexbobPTorne: to get $2000 from every device manufacturer ;)
18:38:52Strife89TXMight get annoying though.
18:38:52 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:39:12gevaertsalexbobP: developing USB costs money too
18:39:14TornealexbobP: you wouldn't be very happy with "USB" if it wasn't the USB tha tthe USB-IF define
18:39:33 Quit _s1gma (Ping timeout: 240 seconds)
18:39:34alexbobPheh, probably wouldn't be very universal...
18:39:37Torneexactly
18:40:23*Strife89TX reckons it would be neat if someone added a separate font setting for plugins like Stopwatch.
18:40:34gevaertsThese industry organisations with strict rules and big companies behind them (and fees...) are why you can buy a device these days and mostly expect it to work. Remember ISA?
18:40:54 Join _s1gma [0] (~d.d.derp@77.107.164.131)
18:41:53n1shmm, getting usefull results from profiling on coldfire is harder than i thought since function calls are so expensive inlining makes a very big difference but letting gcc inline as much as it wants means only a few functions remain...
18:42:18 Quit s1gma_ (Ping timeout: 250 seconds)
18:43:06timcccya. if you're trying to read while running a large font goes a long way. it's enough of a hassle to constantly play around with the theme settings that i don't bother.
18:45:44Strife89TXI don't know the code, but AFAIK it could be a nearly copy-paste process from the text viewer's font setting.
18:45:54S_a_i_n_tyou can easily just use two .cfgs
18:46:46Strife89TXS_a_i_n_t: That's a bit iffy, it'd only be for one setting and he'd still have to punch through a menu.
18:47:24S_a_i_n_tEasiest way a non-coder's going to be able to achieve it, though.
18:47:25Strife89TXPlus Rockbox will probably take longer to load such a .cfg than to just load another font.
18:47:51 Quit panni_ (Read error: Connection reset by peer)
18:48:31n1sStrife89TX: why would that take longer?
18:48:38timcccright, saint's only being practical in the interim. could i create a cfg that set fontsize 32 and another that set fontsize 12 and be done with it?
18:48:44 Quit casainho (Quit: ChatZilla 0.9.86 [Firefox 3.6.10/20100922073514])
18:48:51S_a_i_n_tI'm sure a seperate font setting could be integrated into the Stopwatch plugin, but someone has to actually do the work which I wouldn't hold my breath for unless someone takes a particular interest in who has the coding chops to actually be able to do it.
18:49:06n1stimccc: basically, but you need to select a 12pixel font and a 32 pixel one
18:49:11S_a_i_n_ttimccc: Yeah.
18:49:59timccccool, i'll give it a go now
18:50:36S_a_i_n_tthen, all you'll be doing is browsing a .cfg for Big/Small fonts. You can even use a "fixed.cfg" to make sure that your player always boots with a specific set of settings.
18:51:08S_a_i_n_tI use that (a fixed.cfg) to make sure that my player always defaults to my preferred settings every time it boots.
18:51:26Strife89TXS_a_i_n_t: The text viewer (a plugin) already has a separate-font setting. Forgive my short-sightedness, but does *that* much have to be done aside from copying that bit of code over to the stopwatch (also a plugin) and adding a menu?
18:52:04Tornesounds sufficient to me :)
18:52:40Strife89TXIt just doesn't sound like it would be that difficult for someone who has an idea of what they're doing.
18:52:45S_a_i_n_tStrife89TX: Not really, no...but my point wasn't really anything to do with how easy it is or isn't to do, it's getting someone to do it ;)
18:53:02timcccok, it's done. thank you :D
18:53:13Strife89TXS_a_i_n_t: I can't argue with that point. :)
18:53:31Tornethat doesn't mean you shouldn't ask though :)
18:54:07*Strife89TX asks. :)
18:54:34Tornewell not *me* ;)
18:54:59S_a_i_n_tTo the 'Feature Ideas Forum'!
18:55:14*Strife89TX can't head there right now. :/
18:57:41S_a_i_n_tStrife89TX: If I had a dev environment set up currently, I'd have a 'lil looksee at least, but I haven't got it up & running at the moment (been a slacker).
18:57:58 Join bertrik [0] (~bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
18:57:58 Quit bertrik (Changing host)
18:57:58 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:00
19:02:06n1shmmm, codebook decoding takes about 40% of the total time in vorbis decoding...
19:02:15 Quit kazaik (Ping timeout: 245 seconds)
19:02:23 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.152.3.198.221)
19:02:36n1son cf that is, i have a feeling this is a bit faster on arm
19:02:37saratogaTorne: I'd be ok with an option that flipped the USB IDs to that of the original Archos player
19:02:44saratogaseems fitting for our project
19:02:51saratogaso long as its off by default
19:02:52Tornehehe
19:03:15S_a_i_n_t"nostalgia mode"
19:03:20gevaertssaratoga: doesn't that one load special drivers in windows?
19:03:57saratogai have no idea actually
19:04:24saratogai doubt many people have the archos software installed anymore though
19:05:11pixelmaAFAIK there is a linux driver for it too, written by Zagor
19:05:47pixelmaand the Windows driver doesn't run in newer Windows versions
19:06:26***Saving seen data "./dancer.seen"
19:06:34saratogadidn't we have a patch that actually implemented that SCSI command various devices expected?
19:07:14 Quit Strife89TX (Quit: Job application get?)
19:08:22amiconn[17:14:16] <n1s> regular playback is slower than realtime though so no real gain from test_codec <== What are you working on?
19:08:27gevaertsWe have a patch that implements some ipod things, but that won't be sufficient. I assume the problematic hosts also assumes the original database, or even this USB audio streaming thing that apparently ipods can do
19:08:51Torneit's probably the direct track access over usb thing, yeah
19:08:51n1samiconn: i was testing vorbis with profiling on my h300
19:09:08amiconnah
19:09:11saratogathey stream the file, or actual PCM data over USB?
19:09:17Torneit means the host doesn't have to speak all versions of itunesdb
19:09:31Tornei think it just tells you the path of the file
19:09:37n1smainly to get an idea of where gcc 4.4 makes the code worse when i get around to test that again
19:09:46Tornenot sure though
19:09:53saratogaok that makes more sense
19:10:10Torneso the firmware does something mtp-like
19:10:20pixelmacould Rockbox detect if something wants to talk to it in an unknown protocol?
19:10:21n1sif i mark most of the inline asm functions etc to be always_inline and noinstrument_function the speed is about realtime
19:11:37Tornei think it's weird scsi commands
19:11:45Tornerather than a different usb profile
19:12:35gevaertsIf so, we can probably detect that, but this really needs investigation by someone who has such a device
19:13:08gevaertsBut then we're probably better off if we just reverse engineer the protocol :)
19:14:17Tornewe could maybe detect them but there is a lot of speculation here
19:14:21saratogaa message on the screen saying what happened might be nice
19:14:31saratogalike "unknown SCSI protocol detected"
19:15:22amiconnRegarding different VID/PID we have to be really careful, allocation issues aside
19:17:00amiconnSeveral of the existing usb msd solutions (both hw and sw) have quirks, and I know that at least some host usb stacks enable workarounds for those based on VID/PID
19:17:09saratogaFWIW it looks like the ffmpeg have done 90% of the NEON stuff we want already for us
19:17:21Torneamiconn: indeed
19:17:38saratogaso we should probably figure out if NEON helps battery life on Android, and then decide on a way to selectively enable it based on the host's ability
19:18:29gevaertsamiconn: actually, I wouldn't be surprised if we're already hitting some of those
19:20:37amiconnsaratoga: This is something I'd be interested in... do you have an idea how it would be possible to write the source for the codec libs in a way that they could be compiled for either runtime-selection or compile-time selection of code paths, depending on arm architecture version?
19:21:06saratogaamiconn: make the codeclib functions as function pointers, and then set them based on what the host says it has
19:21:19amiconnOf course without comproising performance for either case
19:22:40saratogaffmpeg has a huge library of NEON implemented functions for audio decoding, many of which are direct implementations of what we have taken from them in c
19:23:24 Quit DerPapst (Quit: Leaving.)
19:23:26saratogaand converting NEON to fixed point is often relatively easy since so many operations in neon can work with either float or fixed point variables
19:24:53amiconnHmm, we could compile all codecs multiple times, place them in subdirs of .rockbox/codecs/, and then select the subdir where to load codecs from at startup
19:25:03amiconnThat won't work for in-core code though
19:25:19saratogais that really better then just using function pointers?
19:25:35 Join Help [0] (~ae170bef@giant.haxx.se)
19:25:43amiconnAnd it may unnecessarily duplicate codecs (those without specific optimisations)
19:25:51amiconnIt's probably easier to implement
19:26:01 Nick Help is now known as Guest3976 (~ae170bef@giant.haxx.se)
19:26:02Guest3976hello
19:26:12saratogain the long term it might be nice to try and consolidate all these different ASM functions
19:26:33Guest3976Hey one question guys, where exactly is the download for rockboy for the e250v2 sansa?
19:26:39saratogaffmpeg has the right idea i think, where they try to provide generic DSP operations (e.g. vector_mul_add)
19:26:49saratogarockboy is built into rockbox
19:26:54saratogayou don't download it separately
19:26:56saratogasee manual
19:27:27saratogavarious things in the codecs are presently duplicated
19:27:47saratogafor instance the mdct windowing is currently implemented in ARM ASM 6 different times by my count
19:27:57amiconnYes, but then some optimisations are only effective when inlined
19:28:07saratogatrue
19:28:15Guest3976alright thak you
19:28:35saratogaalthough i wonder how many of those would be applicable to neon
19:28:52saratogait seems most suited for things like filterbanks and transforms
19:28:57 Quit Guest3976 (Client Quit)
19:29:08saratogai'm not sure it would make sense to inline short blocks of NEON like we do with ASM to work around gcc crap
19:29:26amiconnOne example are the various vector-multiply and combined vector-multiply/vector-add/sub implementations for the ape filters
19:29:58saratogayou wouldn't just call a NEON specific version of the entire filter?
19:30:13amiconnThen you have to build it twice
19:30:49amiconnThe C code is generic - mentioned functions are inlined from .h files selected by CPU architecture and variant
19:30:56 Quit jfc (Read error: Connection reset by peer)
19:31:18amiconns/twice/four times just for arm/
19:31:47 Join jfc [0] (~john@dpc6682208002.direcpc.com)
19:31:52 Quit Stummi (Excess Flood)
19:32:00amiconnv4, v5, v6 and v7
19:32:44 Join Guest70998 [0] (Stummi@doppeldenk.org)
19:34:35amiconnYet another detail is that the filter function isn't binary compatible across arm architecture versions
19:34:45 Quit bmbl (Ping timeout: 240 seconds)
19:35:01amiconnThe v4 version uses 32 bit ints (because it's faster there), v5 and higher use 16 bit ints
19:36:14 Nick Guest70998 is now known as Stummi (Stummi@doppeldenk.org)
19:39:08 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
19:41:04 Join jgarvey [0] (~jgarvey@cpe-065-190-066-089.nc.res.rr.com)
19:42:14 Join krazykit [0] (~kkit@wireless-165-124-204-45.nuwlan.northwestern.edu)
19:43:08bertrikn1s, I looked into the clip v1 firmware to check the BLNDADJ setting for the si4702. It's set to a setting that enables stereo at the lowest RSSI.
19:43:18 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:43:33bertrikWould you mind if I used the same setting in si4700.c ?
19:43:42n1snot at all
19:43:56bertrikok, great
19:44:18bertrikI'll commit that later tonight then
19:51:11 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
19:52:09 Quit krazykit (Quit: awe yeeeeeee)
19:58:20 Quit simonrvn (Ping timeout: 240 seconds)
20:00
20:13:34CIA-7New commit by 03bertrik (r28284): si4700 tuner: lower stereo switchover range, so it switches to stereo for weaker signals. Should fix FS #11405.
20:15:13CIA-7r28284 build result: All green
20:17:15linuxstbbertrik: Did the person who opened that task reply somewhere else? Or are you just assuming it's fixed?
20:18:01bertrikno, he didn't reply somewhere else
20:18:05*linuxstb sees that task was from June...
20:18:22bertrikI tested this myself and it works for me
20:23:53bertrikI don't know what our policy is on keeping tasks like that open, when the bug is a bit subjective and the bug submitter is non-responsive.
20:24:47n1si'd say close
20:28:24linuxstbYes, I agree. I just mistakenly thought you posted that final patch recently for some reason.
20:29:27 Join Strife89 [0] (a80dbf53@gateway/web/freenode/ip.168.13.191.83)
20:37:50 Quit linuxstb (Ping timeout: 240 seconds)
20:40:25 Quit esperegu (Ping timeout: 252 seconds)
20:41:00 Join esperegu [0] (~quassel@145.116.15.244)
20:50:42 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
20:55:26 Join krazykit [0] (~kkit@wireless-165-124-204-45.nuwlan.northwestern.edu)
20:55:27 Quit krazykit (Client Quit)
21:00
21:04:23 Quit thegeek (Ping timeout: 240 seconds)
21:06:29***Saving seen data "./dancer.seen"
21:08:20*Strife89 wonders whatever happened to "Rockbox Modern".
21:09:01bertrikI have no idea what you're talking about
21:09:21Strife89A proposed default theme, evidently. http://www.rockbox.org/wiki/DefaultWPS
21:11:04 Join thegeek [0] (~nnscript@172.80-203-148.nextgentel.com)
21:17:11 Join DerPapst [0] (~Alexander@p5797CC95.dip.t-dialin.net)
21:49:44 Quit thegeek (Read error: Connection reset by peer)
21:50:15 Join thegeek [0] (~nnscript@172.80-203-148.nextgentel.com)
21:57:26 Quit jgarvey (Quit: Leaving)
21:59:13 Quit jordan` (Quit: Coyote finally caught me)
22:00
22:10:46 Quit leavittx_ (Ping timeout: 240 seconds)
22:24:59 Join leavittx_ [0] (~lev@89.221.199.187)
22:31:19 Join krazykit [0] (~kkit@99-126-205-52.lightspeed.cicril.sbcglobal.net)
22:32:25 Join robin0800 [0] (~robin0800@149.254.60.157)
22:34:12 Quit robin0800 (Client Quit)
22:34:33 Join robin0800 [0] (~robin0800@149.254.60.29)
22:39:34 Quit robin0800 (Remote host closed the connection)
22:40:16 Join robin0800 [0] (~robin0800@149.254.61.157)
22:41:00 Quit Strife89 (Quit: Later!)
22:45:17 Quit robin0800 (Remote host closed the connection)
22:45:50 Join robin0800 [0] (~robin0800@149.254.61.157)
22:47:19 Join simonrvn [0] (simon@72.0.216.32)
22:48:39 Quit ender` (Quit: If I ever build a device to transfer the hero's energy into me, I will make sure it cannot operate in reverse.-- Evil Overlord List #180)
22:50:06 Join ender` [0] (krneki@foo.eternallybored.org)
22:54:23 Join jordan` [0] (~jordan@jem75-13-78-235-252-137.fbx.proxad.net)
22:57:53 Quit robin0800 (Remote host closed the connection)
22:58:22 Join robin0800 [0] (~robin0800@149.254.61.29)
22:59:00 Quit evilnick_B (Quit: Page closed)
23:00
23:03:21 Quit robin0800 (Remote host closed the connection)
23:04:17 Quit esperegu (Read error: Connection reset by peer)
23:05:37 Join simonrvn_ [0] (simon@210.215-ppp.3menatwork.com)
23:06:32***Saving seen data "./dancer.seen"
23:08:21 Quit simonrvn (Ping timeout: 240 seconds)
23:08:21 Nick simonrvn_ is now known as simonrvn (simon@210.215-ppp.3menatwork.com)
23:17:59 Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
23:18:04 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
23:24:22 Quit leavittx_ (Ping timeout: 240 seconds)
23:25:19 Join edboyer93 [0] (eboyer93@pool-71-185-65-59.phlapa.fios.verizon.net)
23:26:09 Join leavittx_ [0] (~lev@89.221.199.187)
23:36:41 Quit domonoky (Read error: Connection reset by peer)
23:50:41 Nick soap_ is now known as soap (~soap@rockbox/staff/soap)
23:54:25 Quit CGL (Remote host closed the connection)
23:55:23 Quit Llorean (Quit: Leaving.)
23:58:38 Quit komputes (Read error: Connection reset by peer)

Previous day | Next day