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 2015-01-08

00:04:54 Join Provel [0] (
00:11:55 Quit bp0 (Quit: Leaving)
00:11:55 Quit lebellium (Quit: ChatZilla [Firefox 35.0/20141229214612])
00:12:30 Quit Provel (Ping timeout: 244 seconds)
00:16:38 Quit AlexP (Remote host closed the connection)
00:24:04 Quit uwe_ (Ping timeout: 244 seconds)
00:28:00 Quit xorly (Ping timeout: 244 seconds)
00:31:05 Join uwe_ [0] (
00:34:56***Saving seen data "./dancer.seen"
01:02:46 Quit bertrik (Remote host closed the connection)
01:21:57 Join AlexP [0] (~alex@rockbox/staff/AlexP)
01:38:33[Franklin]foolsh: see G#1124
01:38:37fs-bluebot3Gerrit review #1124 at : 3[WIP] Racing game by Franklin Wei
01:38:57[Franklin]really basic right now
01:39:18[Franklin]I was hoping that someone with a bit more knowledge of 3d/2.5d rendering could give me a bit of help
01:40:45[Franklin]I found this:
01:41:10foolshjust saw an artical today let try to dig it up
01:47:00 Join d33tah [0] (
01:47:19d33tahi'm trying to build rockbox for creative zen
01:47:21[Franklin]hello d33tah!
01:47:31d33tahgot the error: bitmaps/rockboxlogo.h: No such file or directory
01:47:52d33tahgoogle mentions modifying SOURCES in the makefile, but i'm kind of lost on how exactly
01:47:57d33tah(latest git)
01:49:10d33tahbootloader doesn't build either: /home/d33tah/workspace/rockbox/firmware/common/version.c:22:23: error: rbversion.h: No such file or directory
01:49:25[Franklin]how are you building?
01:49:31[Franklin]what command?
01:49:46d33tahused to build the toolchain
01:49:52[Franklin]and then?
01:49:54d33tahthen made "build" directory, cd'ed there and ran
01:50:17[Franklin]that's causing the error?
01:50:24d33tahlooks like.
01:50:32[Franklin]you didn't run make or anything?
01:50:40d33tahsorry, yeah, i did run make
01:50:41 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
01:50:43d33tahgot -j8 in MAKEFLAGS
01:50:49d33tahbut with -j1 it fails as well
01:51:07[Franklin]have you moved your build dir?
01:51:22d33tahwhen? i don't think i did
01:51:49d33tahFedora 21
01:51:58d33tahcommit f91434cc7b4204ca47c86090928cdd454dd52606
01:52:20[Franklin]that's my HEAD too
01:52:48d33tahgot ccache on and my $SHELL is zsh
01:53:07[Franklin]can you ls in the build dir?
01:53:42d33tahhold on, i think i found the issue
01:53:55d33tahlooks like it's leftover from my previous toolchain
01:54:27d33tahi had arm-linux-gnu-gcc and tried hacking configure to use it, after recreating the build dir it seems to go on
01:54:32d33tahsorry [Franklin], my fault
01:54:38[Franklin]np ;)
01:55:36 Quit AlexP (Remote host closed the connection)
02:07:39foolsh[Franklin]: hmm, the horizon isn't starting at screen_height/2 on the fuze+ (at least not yet) ;)
02:09:01[Franklin]yeah, it's very much a WIP :)
02:14:51foolshTo talk a bit about the pseudo part of putting cars and trees on the screen, the Y position of an object is calulated by it Z value, you'll most likely have window of visible "sprites" who's Z value places them between screen_height/2 and screen_height, a Z index falls (or increses which ever you way you want) with acceleration, if they don't fall in between those two limits they get skipped
02:15:53foolshand allong the way you draw them shorter Y/Z and thiner X/Z the farther they are away
02:17:56 Join bp0 [0] (~bp@unaffiliated/bp0)
02:22:58[Franklin]thanks foolsh
02:27:58[Franklin]I'm just not sure if my current code is the right way to do it
02:28:32[Franklin]should I abandon the scanline method that I'm using now and move to an poly-by-poly method?
02:29:06foolshIt would make it more portable
02:29:44foolshif you don't you could calulate that on the fly with screen_hieght _width etc
02:34:48[Franklin]I don't know
02:35:00***Saving seen data "./dancer.seen"
02:35:12[Franklin]scanline is easier, I think
02:37:16foolsheven so, it can still be ambiguous asfar as screen size goes
02:37:32foolsh#define GROUND_LEVEL (LCD_HEIGHT/2)
02:37:42foolshand other such things
02:39:16[Franklin]but then poly-by-poly is much more extensible
02:39:38[Franklin]I think that's what this does:
02:39:59*[Franklin] thinks he'll base the rockbox game off of that one
02:44:00*foolsh reminisces about F-zero on the SNES
02:44:53foolshI really liked the courses with opponents game play style
02:46:36[Franklin]this has got a looong way to go until it's playable ;)
02:46:53Ivoah[Franklin]: What'cha workin' on?
02:47:05[Franklin]Ivoah: G#1124
02:47:09fs-bluebot3Gerrit review #1124 at : 3[WIP] Racing game by Franklin Wei
02:48:23Ivoah[Franklin]: How many projects do you have going right now? xworld, z80e, this racing game...
02:49:17[Franklin]xworld is "done"
02:49:24[Franklin]z80e is in progress
02:49:39[Franklin]"racer" was just started
02:52:14[Franklin]foolsh: lookee here!
02:52:42[Franklin]too bad it's C++, otherwise I would just have copy+pasted it :)
02:53:01foolshThis one is nicely done but no code :(
02:53:06[Franklin]looks like 3-clause BSD license
02:54:14Ivoah[Franklin]: rockbox doesn't support C++?
02:55:49IvoahWow, that's surprising. Why not?
02:56:57[Franklin]too heavy and slow
03:03:18[Saint]While that's largely true, it isn't always the case. I think its more that the ++ bits aren't particularly relevant to the embedded use case, no?
03:05:36[Saint]more specifically _our_ embedded use case.
03:28:42JanCit might be possible to implement a subset of C++ that's useful for Rockbox & similar (embeded) use cases
03:29:39[Franklin]yes, it is completely possible
03:29:40[Franklin]but why?
03:29:53[Franklin]C is faster and more compact in general
03:34:52JanC[Franklin]: in theory, a well-chosen subset of C++ could be more compact even (more expressiveness => more potential optimisations)
03:35:48JanCof course that would exclude stuff like exceptions etc.
03:43:33 Quit [Franklin] (Ping timeout: 252 seconds)
03:44:38 Join chrisb [0] (
03:52:42 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
04:01:56 Quit chrisb (Remote host closed the connection)
04:02:17 Join chrisb [0] (
04:35:04***Saving seen data "./dancer.seen"
04:38:00 Quit charlie (Read error: Connection reset by peer)
04:54:53 Quit krabador (Quit: Take the time.)
04:59:33 Quit williamtdr (Quit: Connection closed for inactivity)
05:10:18 Quit eeeeeta (Ping timeout: 250 seconds)
05:14:41 Quit TheSeven (Ping timeout: 265 seconds)
05:15:47 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:19:36 Quit saratoga (Ping timeout: 246 seconds)
06:30:06 Join benjabean1 [0] (
06:32:10 Part benjabean1 ("Leaving")
06:35:06***Saving seen data "./dancer.seen"
06:46:02 Nick cmhobbs is now known as jurdboyt (~cmhobbs@fsf/member/cmhobbs)
06:46:40 Nick jurdboyt is now known as cmhobbs (~cmhobbs@fsf/member/cmhobbs)
06:47:40 Nick cmhobbs is now known as jurdboyt (~cmhobbs@fsf/member/cmhobbs)
06:49:15 Nick jurdboyt is now known as cmhobbs (~cmhobbs@fsf/member/cmhobbs)
07:06:33 Join eeeeeta [0] (~eeeeeta@gateway/tor-sasl/eeeeeta)
07:22:45 Quit cmhobbs (Remote host closed the connection)
07:49:26 Quit chrisb (Ping timeout: 264 seconds)
08:03:44 Quit nialv7 (Remote host closed the connection)
08:16:52 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:25:18 Join ender` [0] (
08:25:44 Quit eeeeeta (Ping timeout: 250 seconds)
08:28:20 Join LinusN [0] (
08:33:33 Join Provel [0] (
08:35:07***Saving seen data "./dancer.seen"
08:44:59 Quit ukleinek (Quit: change hosts)
08:49:57 Quit Cinos (Ping timeout: 245 seconds)
08:56:26 Join Cinos [0] (
09:00:24 Quit pamaury (Ping timeout: 265 seconds)
09:04:05 Join xorly [0] (
09:06:33 Join petur [0] (5bb7304d@rockbox/developer/petur)
09:21:31 Quit the-kyle (Ping timeout: 245 seconds)
09:22:07 Join the-kyle [0] (
09:43:06 Quit shmibs (Ping timeout: 272 seconds)
09:47:13 Join AlexP [0] (~alex@rockbox/staff/AlexP)
09:48:21 Quit xorly (Ping timeout: 240 seconds)
09:48:57 Join edhelas [0] (~edhelas@
09:51:13 Join shmibs [0] (~shmibs@
10:07:05 Join wodz [0] (
10:35:09***Saving seen data "./dancer.seen"
10:45:05 Join maruk1 [0] (
10:50:36 Join eeeeeta [0] (~eeeeeta@gateway/tor-sasl/eeeeeta)
11:21:34 Join pamaury [0] (
11:21:35 Quit pamaury (Changing host)
11:21:35 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:35:36 Quit AlexP (Remote host closed the connection)
11:44:17 Join Scromple_ [0] (~Simon@
11:45:10 Quit Scr0mple (Read error: Connection reset by peer)
11:45:29 Join Ivoah_ [0] (
11:47:18 Quit Ivoah (Ping timeout: 245 seconds)
11:53:02wodzpamaury: ping
11:54:57 Join AlexP [0] (~alex@rockbox/staff/AlexP)
12:00:52alexbobp[Saint]: ah so it will randomly work with some devices and not others?
12:00:59alexbobpre: the ipod controls thing
12:35:10***Saving seen data "./dancer.seen"
12:49:45pamaurywodz: pong
12:53:05wodzpamaury: in hwstub .lds files the special symbol names are a bit strange. What is the reason of oc_ prefix other then historical?
12:53:28pamauryhistorical, means on-chip
12:55:22wodzpamaury: Are you going to push to gerrit current state of soc desc work?
12:56:01pamauryyeah, I just have a question for you. I stumbled a compatibility issue about hwstub_shell
12:56:21 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
12:56:49pamaury_I the current version, registers have a .read() and .write() method
12:57:27pamaury_and if they have a SCT variant, the .set(), .clr() and .tog() methods use the SCT variant; if they don't, they are emulated with RMW
12:58:21pamaury_with the new format, the concept of SCT is not hardwired anymore, and registers with SCT variants will naturally have a reg.SET.write() method (because of the SET subnode)
12:59:09pamaury_it means that to keep compatibility (ie .set() is using the SET variant when possible) I have to explicitely write code for it and assume some naming convention. What do you think ?
13:00:14wodzisn't it cleaner to do sed s/.set(/.SET.write(/g ?
13:00:29pamaury_yeah that's what I think too
13:01:35wodzanyway 'my' targets don't have SCT variants so do what you find more convenient from your POV
13:25:41 Quit kvieta (Ping timeout: 245 seconds)
13:33:57 Join kvieta [0] (~kvieta@
14:08:28pamaury_wodz: g#1073
14:08:33fs-bluebot3Gerrit review #1073 at : 3soc_desc: new version of the desc file format by Amaury Pouly
14:08:58wodzpamaury_: Ok, will look
14:09:15pamaury_at the moment my old lua scripts are broken because I need to do some renaming and maybe add a few fields, also hwstub_shell is mostly untested at this point except for the register tree generation
14:11:27wodzpamaury_: you mean SCT thingy is broken in lua, right?
14:13:40pamauryfor rk27xx I think it should work
14:14:09 Quit pamaury_ (Ping timeout: 244 seconds)
14:14:15pamauryat least it is implemented, although untested
14:14:34pamauryyou can do reg.set(x) and it will read-modify-write
14:20:39 Join einhirn [0] (
14:23:47 Quit wodz (Quit: Leaving)
14:35:14***Saving seen data "./dancer.seen"
14:36:19 Join xorly [0] (
14:51:18 Join amayer [0] (
15:32:30 Quit AlexP (Remote host closed the connection)
15:57:45 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
16:00:21 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
16:00:30 Quit dfkt (Disconnected by services)
16:00:45 Nick dfkt_ is now known as dfkt (dfkt@unaffiliated/dfkt)
16:07:34pamauryhere we go, let's see if the usb rework compiles on all targets...
16:09:15 Nick cmhobbs is now known as jurdboyt (~cmhobbs@fsf/member/cmhobbs)
16:09:34 Part LinusN
16:10:54 Join ZincAlloy [0] (
16:11:01 Quit jurdboyt (Quit: Leaving)
16:11:15 Join cmhobbs [0] (
16:11:15 Quit cmhobbs (Changing host)
16:11:15 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
16:12:50 Join chrisb [0] (
16:16:35pamauryah, a few yellows
16:17:03gevaertsWeird ones
16:17:42pamauryyeah I'm surprised by these warnings, I need to try to compile them to see
16:17:57pamauryI must say I forgot to grep in the bootloaders
16:19:38pamaurydamn the as3525 bootloader uses #ifdef USE_ROCKBOX_USB
16:20:29gevaertsShould be HAVE_BOOTLOADER_USB, I guess
16:21:05pamauryit's weird, based on pamaury/b3db692c1880ef97a0f2">
16:21:12pamaurythe as3525 bootloader don't do HAVE_BOOTLOADER_USB
16:21:22pamaurybut they have USE_ROCKBOX_USB...
16:21:53pamaurydo they have bootloader usb mode ? I don't think so ?!
16:22:37pamauryhum, from the code it seems they do a "Bootloader USB mode" for charging only
16:23:05pamaurywho can confirm this ?
16:24:45pamaurybased on this assumption, USE_ROCKBOX_USB is wrong too, it should be !USB_NONE
16:24:45 Quit xorly (Read error: Connection reset by peer)
16:26:37 Join xorly [0] (
16:27:22 Join Misanthropos [0] (
16:27:56gevaertspamaury: I'm not convinced. It looks strange to me
16:29:26pamaurywhy ? the #ifdef could even be removed since those have usb anyway
16:29:34pamaury(usb detection at least)
16:30:00gevaertsI mean, are you sure that's not full bootloader USB?
16:30:15gevaertsThey did get USB_FULL_INIT
16:30:24gevaerts(defined(USE_ROCKBOX_USB) && CONFIG_USBOTG == USBOTG_AS3525)
16:31:31pamauryhum, but the thread doesn't acknowledge usb connection so it cannot do mass storage in theory
16:32:13gevaertsNo, that's where I'm confused :)
16:32:47pamauryso yeah it has a full usb stack (which is bad because it should define HAVE_BOOTLOADER_USB instead of hacking usb.c) but I don't see how it can use it
16:32:52pamauryexcept for charging
16:35:15***Saving seen data "./dancer.seen"
16:35:56gevaertsI think I have a fuzev2 somewhere at home. I can check tonight what the bootloader actually does
16:36:15pamauryyeah I have a clip+, I can check tonight
16:36:32pamauryit doesn't impact the released bootloader anyway
16:37:05gevaertsIt doesn't, no, but I think that code hasn';t been changed since the bootloader release, so it will clarify things
16:37:24pamaurythe pp bl also uses USE_ROCKBOX_USB
16:38:09pamauryurg, what is this usb_drv_powered() thing ???
16:41:01pamaurylooks like another way to know if usb is inserted to me
16:41:42pamauryexcept the code seems to indicate it's not the case :(
16:43:10pamauryI'll have to get rid of this if possible
16:43:47pamaurydamn this bootloader is a mess
16:44:29pamaurygevaerts: I'm going to commit the other patches on gerrit about usb if don't mind ?
16:44:37pamauryI'll fix this tonight
16:44:49gevaertsI agree
16:45:09gevaertsThis bootloader thing needs more thinking and testing, and shouldn't block the other stuff
16:46:27pamauryyes, there is plenty of old code which does things the old way
16:46:32pamauryit needs a cleanup
16:59:38pamauryother commits didn't make it worse
17:06:22 Quit cmhobbs (Remote host closed the connection)
17:15:14 Quit Misanthropos (Ping timeout: 255 seconds)
17:23:30 Quit petur (Quit: Page closed)
17:29:39 Quit edhelas (Ping timeout: 244 seconds)
17:39:36 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
17:44:58pamauryyes ?
17:45:30gevaertsShould I fix the manual in a hacky way, or hold to principles and blame debian?
17:45:33 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
17:45:45pamaurywhat is the problem ?
17:46:03pamauryyou could do both ;)
17:47:20gevaertsSee g#1127
17:47:24fs-bluebot3Gerrit review #1127 at : 3Include scrreprt.4ht from by Frank Gevaerts
17:47:32gevaertsWith that, the html manual builds for me
17:48:01gevaertsIt's not really a clean fix though
17:48:52pamauryurg, I'm not sure I understand why this is necessary
17:49:12gevaertsBecause the scrreprt.4ht in debian is "broken"
17:49:28gevaertsIt depends on other stuff, and that other stuff has changed in apparently incompatible ways
17:50:34gevaertsYou can also overwrite your packaged version, but that's not really much better :)
17:51:39gevaertsAnd reporting the bug and waiting until it's fixed in stable releases might take a while...
17:52:05pamauryis it broken in stable only or all versions ?
17:52:05 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
17:52:20pamauryI'd say we could add this file to fix it and report the bug if needed
17:52:34gevaertsIt's broken in all current debian-related distros as far as I can see
17:52:47gevaertsIt might be OK in some others
18:02:40*gevaerts wants to add the debian bug number to the commit message before he pushes, so now he has to wait for that to come back...
18:12:55 Quit amayer (Quit: Leaving)
18:16:12 Join amayer [0] (
18:17:55 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:21:27 Quit cmhobbs (Remote host closed the connection)
18:34:36 Join edhelas [0] (
18:35:19***Saving seen data "./dancer.seen"
18:40:00 Quit maruk1 (Quit: Leaving.)
18:41:56gevaertsSo it's still a bit more complicated
18:43:33 Join GeekShadow [0] (
18:43:33 Quit GeekShadow (Changing host)
18:43:33 Join GeekShadow [0] (~antoine@reactos/tester/GeekShadow)
18:43:54 Quit d33tah (Ping timeout: 255 seconds)
18:46:19 Join d33tah [0] (
18:49:21 Quit pamaury (Remote host closed the connection)
19:17:11gevaertsOK, so I don't really get the manual thing
19:18:07gevaertsOn my laptop, putting that scrreprt.4ht in manual/ fixes things, on TheSeven's build client it doesn't, and on older setups (such as my build client) where the manual still usewd to build it breaks things
19:18:29gevaertsI'm going to hope it fixes things on the server so the manuals on the website are back...
19:25:51 Join AlexP [0] (~alex@rockbox/staff/AlexP)
19:31:35 Join RiD [0] (
19:34:38 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
19:46:46pamaurygevaerts: my sansa clip+ doesn't seem to do USB bootloader mode
19:47:20pamauryah wait, I need to press select
19:48:09pamaurydoesn't seem to have any effect, maybe my bootloader isn't new enough
19:48:32 Quit d33tah (Ping timeout: 265 seconds)
19:51:47 Join d33tah [0] (
19:52:00gevaertspamaury: my fuzev2 does do it
19:52:14pamaurywhen you press select ?.
19:52:21pamaurywhat happen then ?
19:53:29gevaertsIf I hold select when plugging it in while it's off, the screen says "Bootloader USB mode" and I have USB
19:53:55gevaertsWhich works
19:54:03pamaurydo you have the storage ?
19:54:20gevaertsI can mount it and see files
19:56:08pamauryI really don't understand how this is possible...
20:01:07pamaurythe main thread needs to acknowledge the connection
20:02:58pamauryI will check on my fuze+ what happen
20:14:38gevaertspamaury: the main thread doesn't actually have to ack it
20:14:55gevaertsEvery event handler needs to ack it, but if you don't have an event handler...
20:21:31pamauryah that's right
20:22:13 Quit chrisb (Ping timeout: 244 seconds)
20:27:48 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
20:29:18 Join krabador [0] (
20:29:32 Quit krabador (Changing host)
20:29:32 Join krabador [0] (~krabador_@unaffiliated/krabador)
20:29:47 Quit JdGordon (Ping timeout: 264 seconds)
20:35:20***Saving seen data "./dancer.seen"
20:46:50 Join lorenzo92 [0] (
20:47:34fs-bluebot3Gerrit review #1120 at : 3ypr0/ypr1: GPIO handling API refactoring by Lorenzo Miori
20:48:19fs-bluebot3Gerrit review #707 at : 3yp-r0: improve the charging code by Lorenzo Miori
21:05:06[Saint]lorenzo92: JSYK, fs-bluebot can deal with multiple gerrit task instances on the same line.
21:05:29[Saint]for example: g#1120 g#707
21:05:56[Saint]Oh...well, isn't that just hilarious. Heh. :)
21:06:14*[Saint] prods fs-bluebot
21:07:06[Saint]Well, I just made a fool of myself. Up yours, fs-bluebot. It /should/ deal with that just fine.
21:07:38fs-bluebot3Gerrit review #1120 at : 3ypr0/ypr1: GPIO handling API refactoring by Lorenzo Miori
21:07:38fs-bluebot3Gerrit review #707 at : 3yp-r0: improve the charging code by Lorenzo Miori
21:08:08[Saint]Aha. He was taking a nap. :)
21:24:04lorenzo92haha indeed, anyway spammers are not 1 liners :D
21:38:06 Join chrisb [0] (
21:39:38 Quit edhelas (Ping timeout: 264 seconds)
21:45:52 Join lebellium [0] (
22:00:33 Join saratoga [0] (123e1f2d@gateway/web/freenode/ip.
22:00:58saratogahas anyone built the compilers on macos recently? just trying it now fails building gcc with some compiler errors
22:01:23saratoga../../gcc-4.4.4/gcc/toplev.c:571:1: error: redefinition of a 'extern inline' function 'exact_log2' is not supported in C99 mode
22:04:34pamaurygevaerts: I'm puzzled, on my fuze+ bootloader USB mode won't work without the made thread explicitely acknowledging the connect
22:05:09pamauryI must be missing something...
22:06:44gevaertspamaury: button_init()
22:07:10gevaertsThe as3525 bootloader doesn't do that, and it sets up an event queue
22:07:27pamauryindeed, and how/when is this queue polled ?
22:08:16 Join edhelas [0] (
22:09:51pamaurybut I already call it regularly...
22:10:03gevaertsWell yes
22:10:09*gevaerts is confused
22:10:17 Join _wodz_ [0] (594b6add@gateway/web/freenode/ip.
22:10:21gevaertsWhat's the question?
22:11:00pamaurywhy do I need the explicit call to usb_acknowledge(SYS_USB_CONNECTED_ACK);
22:11:12gevaertsBecause there's an event queue
22:11:23gevaertsas3525 doesn't do tha
22:11:52pamauryah you mean that the button function don't do the ack...
22:11:56pamauryyeah that makes sense
22:12:03_wodz_gevaerts: Could it be that db->quote() thing causes clients to think that build is always sucessfull?
22:12:19gevaerts_wodz_: I think so, yes
22:12:40_wodz_gevaerts: would be nice to test this somehow
22:13:37pamauryhum, I think I spotted the bug in the imx233 bootloader
22:13:58gevaerts_wodz_: that's Zagor's job I'd say :)
22:14:00pamaurybefore loading the firmware, I never stop the threads or the timer or the scheduler, something nasty could happen
22:15:58pamauryhum, I disable interrupts though
22:16:16pamauryOne day I'll find why the bootloader *sometimes* freezes after usb mode
22:16:45_wodz_with disabled interrupts I don't see what can go wrong
22:18:03pamaurygevaerts: so about sansa as3525 bootloader, I think we should define HAVE_BOOTLOADER_USB and ifdef HAVE_BOOTLOADER_USB in the bl
22:18:17*gevaerts nods
22:19:34pamauryalso I don't see why the includes are conditional
22:20:50 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
22:21:27pamaurygevaerts: on which target did you say you can bootloader usb mode ?
22:21:38gevaertsThat was fuzev2
22:21:39pamauryIt is a as3525 target right ? not as3525v2 ?
22:22:14gevaertshmmm, I *think* that's v2
22:22:17saratogav2 is amsv2
22:22:27saratogaits the same guts as the clipv2
22:22:45pamauryhum, ah yeah there is also a special case in usb.c for USBOTG_S3C6400X, so ugly
22:22:59 Quit edhelas (Ping timeout: 252 seconds)
22:23:21pamauryand USBOTG_S3C6400X is used ipodnano2g and ipod6g so I need to be careful not to break those
22:24:04 Join RiD [0] (
22:24:09_wodz_pamaury: ipod6g doesn't have bootloader yet, nano2g have diskmode
22:24:32_wodz_pamaury: and both have broken usb driver in HEAD
22:25:49pamauryyeah I know that but for what concerns me, I just want to not change anything, just cleanup
22:26:29_wodz_cut with axe :P
22:27:22pamaurywell more precisely I want to rationalise a little bit the usb code
22:28:37_wodz_pamaury: can I somehow convert v1->v2 desc file now? swiss_knife or something?
22:29:03pamaury_wodz_: yeah, use swiss_knife convert
22:30:07_wodz_pamaury: any estimated time for qeditor supporting v2 format?
22:31:43pamaurygevaerts: g#1128
22:31:48fs-bluebot3Gerrit review #1128 at : 3usb: finally get rid of USE_ROCKBOX_USB in Sansa bootloader, use bootloader usb by Amaury Pouly
22:32:09pamaury_wodz_: not really, tonight I want to finish hwstub_shell and then I'll work on qeditor, I need to work a bit on it before any giving an estimation
22:32:50pamaurymy impression is that it should be quick, I spent quite a bit of time writing generic helpers which will considerably simply the code of qeditor I think
22:33:25gevaertspamaury: looks ok to me
22:33:50pamauryok let's go
22:34:38_wodz_pamaury: I saw in logs you wrote some tool to see what defines combination is used for particular build. I think you should commit this. This is something I was wondering a few times how to get this data
22:35:05pamauryah yes of course, I'll put it in utils/analysis
22:35:24***Saving seen data "./dancer.seen"
22:45:18fs-bluebot3Gerrit review #1129 at : 3Introduce a new analysis tool to detect which macros are defined in each build. by Amaury Pouly
22:45:30pamauryif you have any suggestions, otherwise I'll commit it
22:46:36thomasjfoxpamaury: nice tool. I wanted something for my cppcheck stuff, too
22:47:05pamauryfor the curious, the output looks like this: pamaury/b3db692c1880ef97a0f2/raw/f4a2740a351bf8304b6abcb9d5666c4978b1d9e4/usb_defines.txt">
22:48:11_wodz_pamaury: neat
22:48:15thomasjfoxdid you consider invoking "configure" locally and then grep the results (makefile / autoconf.h)?
22:48:29pamaurygevaerts: apparently all the yellows are gone
22:48:55pamaurythomasjfox: it will only work for a very small list
22:48:58gevaertspamaury: now fix the Recorder and we'll be all green again :)
22:49:02pamaurythomasjfox: this is much more general
22:49:23pamaurythomasjfox: but this will run "configure" locally yes
22:49:43pamaurygevaerts: hehe, not sure I'm going to do this ;)
22:52:23 Quit _wodz_ (Quit: Page closed)
22:52:38thomasjfoxpamaury: ah, I was fooled by "www repository build list". I thought it gets the list online
22:53:16pamaurythomasjfox: no, it gets the list for the www/ repository, which you need to have a local checkout
22:53:27pamauryalthough it wouldn't be hard to get it online I guess
22:53:29thomasjfoxyes, exactly. All is fine :)
22:54:25pamaurymaybe I'll put the link in usage()
22:54:26 Join wodz [0] (
22:55:17thomasjfoxjust don't download it automatically, especially if it's via http and there is no way to verify it
22:55:43pamauryyeah that's right, especially given it will happily run the commands in it ;)
22:58:43 Join cmhobbs [0] (
22:58:43 Quit cmhobbs (Changing host)
22:58:43 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
22:58:45 Join [Franklin] [0] (~franklin@unaffiliated/franklin)
22:59:52wodzFollowing the wave of cleanups I'd like to get rid of those stupid function names _backlight_init(), _backlight_on(), _backlight_off() and _backlight_set_brightness(). What do you think about backlight_hw_init(), backlight_hw_on(), backlight_hw_off() and backlight_hw_brightness()?
23:02:20pamaurythat's fine with me, I was thinking about backlight_target_* but backlight_hw_* is fine maybe even better
23:02:47pamauryand I agree the old names are just stupid
23:03:02wodzor backlighthw_{init, on, off, brightness} just like audiohw_ things
23:03:50thomasjfoxbacklight_hw_xxx somehow feels more readable than backlighthw_xxx
23:03:59 Quit lorenzo92 (Ping timeout: 255 seconds)
23:04:28pamauryyeah I don't really like the backlighthw_*, it's okay for audiohw because it's short and readable
23:05:31thomasjfoxwild guess: the brain stops parsing after "audio" automatically, that's why audiohw is just fine regarding readability
23:08:11thomasjfoxhmm. in the plugins or codecs, a good bunch of functions can be made static. Is it worth the effort?
23:08:47[Franklin]thomasjfox: almost everything can
23:08:47pamauryprobably not I'd say
23:09:03thomasjfoxit's more interesting for the main app code I guess
23:09:12[Franklin]everything but plugin_start, I think
23:10:07thomasjfox[Franklin]: making a function status just reduces the visibility for the linker. If you call functions from other .c modules, the function can't be made static
23:10:33[Franklin]oh, yes
23:10:47[Franklin]still, you should easily be able to automate it
23:10:58pamaurywodz: I have hwstub_shell working on imx233
23:11:04[Franklin]at least for single-file plugins
23:11:14pamauryat least it can run my nwze360 script which inits the lcd and display some pixels, so it's a pretty good test
23:11:56pamauryI have updated the gerrit task
23:12:10thomasjfox[Franklin]: yes, but it's not worth the effort for plugins since they get loaded in the static plugin buffer and flushed out when exited
23:12:18wodzpamaury: I'll test atj and rk27xx when I find a bit bigger chunk of spare time to play with hardware
23:12:32pamauryok, in the mean time I'll work on qeditor
23:12:43pamauryprobably not tonight though
23:13:16[Franklin]oh yeh
23:13:37 Quit bluebrother (Disconnected by services)
23:13:42 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
23:14:29 Quit chrisb (Ping timeout: 256 seconds)
23:15:27thomasjfoxpamaury: how about this one: [rockbox/utils/regtools/qeditor/utils.cpp:892] -> [rockbox/utils/regtools/qeditor/utils.cpp:890]: (style, inconclusive) Found duplicate branches for 'if' and 'else'.
23:16:20 Quit fs-bluebot (Ping timeout: 244 seconds)
23:16:34pamauryhum, interesting :)
23:17:32pamauryI'll fix it in the gerrit patch, thanks :)
23:21:14 Quit saratoga (Ping timeout: 246 seconds)
23:21:20 Join chrisb [0] (
23:22:46 Quit amayer (Quit: Leaving)
23:29:05[Franklin]someone want to commit G#1122?
23:29:29[Franklin]no bot :(
23:31:54thomasjfox[Franklin]: why was the playback control removed?
23:32:20 Quit ender` (Quit: I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth.-- Umberto Eco)
23:32:56thomasjfoxthe commit message doesn't give a reason
23:33:30[Franklin]thomasjfox: xworld grabs the audiobuf, so there can be no playback
23:34:06thomasjfoxah, I thought it was referring to the playback control of the game
23:34:27[Franklin]lol no :P
23:35:11 Quit pamaury (Ping timeout: 264 seconds)
23:35:14thomasjfoxcan you add that reason to the commit message?
23:35:26[Franklin]oh sure
23:35:32thomasjfoxthen I'll commit it
23:41:34[Franklin]foolsh: I think I've come up with a way to implement fast pseudo-3d in rockbox
23:41:54[Franklin]it's going to essentially be a polygonal engine, except in 2 dimensions
23:42:29[Franklin]I'll use xlcd to do fast tri fills
23:47:08thomasjfox[Franklin]: submitted
23:47:26[Franklin]ok, shouldn't break anything, I'd hope :)
23:48:18thomasjfoxgevaerts: ah, now I got it, you fixed that damn red for the HTML manual that was stuck forever. Nice.
23:48:33[Franklin]yay! :D
23:50:44[Franklin]can I read them online now?
23:52:33 Join fs-bluebot [0] (
23:52:34 Quit cmhobbs (Remote host closed the connection)
23:54:10fs-bluebotBuild Server message: 3Build round completed after 504 seconds.
23:54:47 Quit lebellium (Quit: ChatZilla [Firefox 35.0/20141229214612])

Previous day | Next day