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 2017-05-07

00:14:17 Part robertd1
00:14:21 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 53.0/20170413192749])
00:28:36 Quit cc___ (Quit: WeeChat 1.6)
00:55:24 Quit dreamlayers (Quit: Leaving.)
00:56:33 Quit scorche` (Read error: Connection reset by peer)
00:56:46MarqueteurFWIW, on the locking behaviour, with the Clip Zip, if it's slid into a pocket the volume buttons are relatively hard to hit, but the face buttons are easy to hit. The selective lock to leave the volume accessible when locking the rest is quite nice on that target. Clip+ is pretty similar there.
01:09:03 Join scorche [0] (~scorche@rockbox/administrator/scorche)
01:22:52***Saving seen data "./dancer.seen"
01:43:22 Quit MrZeus (Ping timeout: 260 seconds)
01:59:12 Quit ender` (Quit: Common sense isn't.)
02:01:53 Join jhMikeS [0] (
02:06:45 Quit ZincAlloy (Quit: Leaving.)
02:09:34 Join dreamlayers [0] (
02:55:37 Quit Marqueteur (Remote host closed the connection)
02:56:23 Join Marqueteur [0] (~Tor@fsf/member/TorC)
03:15:27 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
03:22:56***Saving seen data "./dancer.seen"
03:51:03 Quit jhMikeS (Ping timeout: 240 seconds)
04:20:18 Quit dreamlayers (Quit: Leaving.)
04:41:51 Quit rudi_s (Quit: leaving)
04:44:38 Join rudi_s [0] (
04:49:28 Join jhMikeS [0] (
05:22:58***Saving seen data "./dancer.seen"
05:26:04 Quit jhMikeS ()
05:26:04 Join Strife89 [0] (
05:29:20 Quit Strife1989 (Ping timeout: 240 seconds)
06:54:25 Quit TheSeven (Ping timeout: 255 seconds)
06:54:50 Join dreamlayers [0] (
06:54:54 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
07:23:02***Saving seen data "./dancer.seen"
07:28:24 Join PurlingNayuki [0] (~Thunderbi@
07:32:26 Quit PurlingNayuki (Ping timeout: 240 seconds)
07:41:30*pamaury now has a somewhat nice tool to compare manual keymap and rockbox keymap
07:45:54dreamlayerspamaury, that sort of tool could have probably caught the bug I fixed in 1b4ee1f.
07:52:26 Quit dreamlayers (Quit: Leaving.)
07:55:14pamauryI am not sure, first my tool does not (yet) handle plugins, only core keymap (but I plan to add plugin support), also it is only based on #define and other stuff, so not sure it could catch 1b4ee1f
09:01:05 Quit APLU (Quit: !suicide)
09:02:41 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:8060:f15e:b0de:21fb)
09:05:49 Join APLU [0] (
09:23:03***Saving seen data "./dancer.seen"
09:44:43 Join ender` [0] (
09:48:23 Join jhMikeS [0] (
09:52:43 Join johnb2 [0] (
10:08:54user890104how do you deal with header files from other projects that are part of a rockbox plugin and have the same name as a firmware header (config.h, disk.h, audio.h, cpu.h, mutex.h) ?
10:12:52 Join lorenzo92_1 [0] (
10:16:52bluebrotherpamaury: wouldn't it be better to generate the manual keymaps from the code?
10:20:04 Quit jhMikeS (Ping timeout: 240 seconds)
10:21:43 Join PurlingNayuki [0] (~Thunderbi@
10:22:04pamauryuser890104: for now I don't handle plugins, but with a combination of parsing and magic makefiles you can go pretty far. It could generate it I guess, but since they might be errors/false positive, at this point I prefer to point out the differences so that a human can have a look. Possibly the tool may print suggestions
10:23:52user890104i'm trying to include fake86 as a plugin, but since it's a x86 emulator it has headers with similar names to those of rockbox's internals
10:24:20user890104so i renamed them to fake86_*.h, but i don't know if it's a good idea to have half the headers renamed this way
10:25:00user890104or i can make apps/plugins/fake86/include/fake86/, put them there and include them as: #include "fake86/config.h" ?
10:26:20pamauryuser890104: I think you second approach (#include "config.h" -> #include "fake86/config.h") is best, at least it avoids renaming files
10:26:41user890104ok, i'll try that
10:27:26user890104also, where do i put plugin's data files?
10:27:39user890104like video rom image, bios image, floppy disk image and so on
10:28:15pamauryif those are copyrighted (like doom), we don't ship them but put a link in the manual/wiki or say that the users have to get them on their own
10:28:43user890104the bios can be compiled from source using nasm
10:28:57user890104and is released with the same license as the emulator code (GPL2) i think
10:29:11 Quit ZincAlloy (Quit: Leaving.)
10:29:15pamauryhum, then I am not sure.
10:29:22pamaurygevaerts: any thoughts on this?
10:29:34user890104ROM BASIC is copyrighted by IBM :) so we're not going to ship it probably
10:29:44pamauryI still think it's simpler to ask the user to download them
10:30:10user890104rbutil installs a free version of the doom files, right?
10:30:24pamauryI guess so
10:30:46bluebrotheryes, the freedoom one from the download server
10:30:51*pamaury needs to build a list of all latex's keymap commands. That will probably involve a lot of grepping
10:30:54user890104in this case, can we make another option to install an open source version of the binaries needed by the emulator (built by us)
10:31:09bluebrother(see rbutil.ini, doom_url :)
10:31:10user890104so the user has a working setup (floppy disk image, bios)
10:31:22user890104and only if they want to
10:31:33pamauryyeah doing that in RbUtil (+doc on the wiki) sounds like a sane idea
10:31:47user890104ok thanks, i'll look into it later
10:31:57bluebrotherwe can install all kind of stuff. It just needs to be done :)
10:32:16bluebrotherbut while at it, we could also install the midi patchset from the download server. Currently we don't do that
10:32:29user890104is it free/opensource?
10:33:27user890104i guess i can prepare a patch for that too, after i figure out how to add additional type of data files to rockbox installation
10:34:17bluebrotherthe wiki calls it "frameware", assuming the file on the download server is the same that is linked in the wiki
10:35:05bluebrotherwhy do you need additional types? We usually have zip files that we simply extract to the root of the player.
10:35:29bluebrotherthat way we don't need to care for paths in Rockbox Utility. And it works for basically everything, as long as you package it into a zip :)
10:36:52user890104by type i mean something like the doom files
10:37:34 Join lebellium [0] (
10:37:42 Quit johnb2 (Ping timeout: 255 seconds)
10:37:44user890104so the menu looks like: [ ] bootloader [ ] Rockbox [ ] Themes [ ] Game files (Doom) [ ] MIDI patchset [ ] fake86 files
10:39:50bluebrothersee gui/selectiveinstallwidget.cpp
10:41:11bluebrotherwe need to repackage the patchset to zip though
10:41:31*bluebrother goes for that
10:58:05bluebrotherhmm, that could use a bit of refactoring.
11:23:05***Saving seen data "./dancer.seen"
11:28:40 Join paulk-elm [0] (~user@
11:51:12 Join jhMikeS [0] (
11:55:12pamaurykugel: here?
11:55:50 Join johnb2 [0] (
12:05:00kugelpamaury: for a moment
12:06:12pamaurykugel: quick question: I am trying to make phabricator work and it seems ./bin/phd is ignoring phd.user. I tried to set phd.user to 'phabricator' (that's the user I was planning for it) but when I run ./bin/phd (as phabricator) it dies saying Requested path '/var/tmp/phd/pid' is not writable, did you run into this problem?
12:10:30kugelfor a moment
12:10:33kugelyou need to create that dir
12:10:35kugelI can share our production systemd units
12:10:38kugelI used /var/run/ though but that shouldn't make a difference
12:12:04kugelthe unit runs mkdir as root and then drops to www-data for running the daemons
12:13:12pamauryah yeah systemd units would be cool. So you run daemons also as www-data and not as a different user?
12:14:58kugelI chose the same user as the web site
12:15:47 Join fs-bluebot_ [0] (
12:16:09 Quit fs-bluebot (Ping timeout: 240 seconds)
12:16:10pamauryok, ah it seems to work now that the directory exists
12:16:26 Quit bluebrother (Ping timeout: 240 seconds)
12:16:45kugelso that local.json can be restricted to that user (0600)
12:18:25pamaurywhat is the advantage of that?
12:18:52kugelother users cant read the db password
12:19:31pamauryah that's a good point. But you could also set permission to phd-user:www-data
12:19:37 Join MrZeus [0] (~MrZeus@2a02:c7f:7066:fb00:e06d:e4c9:b109:1bec)
12:20:00kugelthere are other sites on our server
12:20:23 Quit fs-bluebot_ (Ping timeout: 260 seconds)
12:21:12kugelour security guy wanted some extra precautions
12:21:51pamauryis it not possible to have per vhost user with apache?
12:22:30pamauryotherwise yeah that's a good point
12:22:47kugelapache runs as www-data, it cannot just su to another one
12:23:19 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
12:23:39kugelbut I dont know if it can do that as root
12:26:38pamauryapparently there exists mpm_itk that can do that, but it has to run as root to be able to su to the various users
12:28:04SammysHPIf you have to create that directory more than once (e.g. after reboot) you can use systemd tmpfiles
12:28:27kugelfor us it was important to a) run apache as non-root and that other users cant access the phabricator db (ie read all souce code)
12:29:12 Quit PurlingNayuki (Ping timeout: 260 seconds)
12:30:43kugelmkdir -p at unit start works fine too 😃
12:31:30SammysHPAnd regarding permissions: if you are crazy you could try ACLs ;)
12:34:19 Join fs-bluebot [0] (
12:42:21 Quit jhMikeS (Ping timeout: 246 seconds)
12:49:42 Quit lorenzo92_1 (Ping timeout: 260 seconds)
13:11:27 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan)
13:23:07***Saving seen data "./dancer.seen"
13:37:14 Join robertd1 [0] (~root@
14:12:16 Quit pamaury (Ping timeout: 255 seconds)
14:13:25 Join lorenzo92_1 [0] (
14:33:18kugelpamaury (logs):
14:33:48kugelthe second unit is for aphlict module which enables in-browser notifications
14:35:21 Join JanC_ [0] (~janc@lugwv/member/JanC)
14:36:36 Nick JanC is now known as Guest4823 (~janc@lugwv/member/JanC)
14:36:36 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC)
14:37:35 Quit Guest4823 (Ping timeout: 268 seconds)
14:40:49 Quit lorenzo92_1 (Ping timeout: 240 seconds)
14:50:56 Quit johnb2 (Ping timeout: 260 seconds)
15:19:47 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:b9d7:9c91:6038:ff51)
15:23:08***Saving seen data "./dancer.seen"
15:51:44 Join johnb2 [0] (
15:55:49 Quit johnb2 (Ping timeout: 240 seconds)
15:57:11 Join cc___ [0] (
16:02:12 Join PurlingNayuki [0] (~Thunderbi@
16:06:09 Join Strife1989 [0] (
16:08:56 Quit Strife89 (Ping timeout: 240 seconds)
16:15:04 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
16:17:06 Quit PurlingNayuki (Ping timeout: 240 seconds)
16:28:03 Join PimpiN8 [0] (
16:36:21 Join PurlingNayuki [0] (~Thunderbi@
16:40:42 Quit PurlingNayuki (Ping timeout: 260 seconds)
16:50:08PimpiN8is it possible to make a playlist with rockbox?
16:53:58[Saint]It's not possible not to.
16:54:54[Saint]All our playback is playlist based. Like it or not it's unavoidable.
16:55:25*[Saint] suggests our fine manual for further reading on many commonly asked topics
17:18:58 Join johnb2 [0] (
17:23:09***Saving seen data "./dancer.seen"
17:36:20 Quit MrZeus (Ping timeout: 264 seconds)
18:06:49 Quit robertd1 (Ping timeout: 240 seconds)
18:14:16 Join PurlingNayuki [0] (~Thunderbi@
18:22:05 Quit preglow (*.net *.split)
18:27:30 Quit PurlingNayuki (Ping timeout: 255 seconds)
18:28:04 Join preglow [0] (~thomj@2001:840:4243:3::101)
18:29:10 Join PurlingNayuki [0] (~Thunderbi@
18:33:33 Quit PurlingNayuki (Ping timeout: 240 seconds)
18:34:44 Join PurlingNayuki [0] (~Thunderbi@
18:52:32 Quit PimpiN8 (Quit: My MacBook has gone to sleep. ZZZzzz…)
18:54:46 Quit johnb2 (Ping timeout: 260 seconds)
19:14:28 Join PimpiN8 [0] (~textual@2a02:a454:38ea:1:98be:b491:16dd:e269)
19:23:11***Saving seen data "./dancer.seen"
19:24:51 Join krabador [0] (~krabador@unaffiliated/krabador)
19:58:42 Join johnb2 [0] (
20:05:25 Quit johnb2 (Ping timeout: 258 seconds)
20:08:00 Join johnb2 [0] (
20:18:20 Quit johnb2 (Quit: Nettalk6 -
20:29:44 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan)
20:36:12 Join Strife89 [0] (
20:37:06 Quit Strife1989 (Ping timeout: 240 seconds)
21:00:39user890104the mkdepfile make function has decided that a certain source file depends on a header that is located in a different directory
21:00:59user890104is there a way to tell why, and change it to the right path?
21:02:14user890104/rockbox/build-sim-win64/fake86/config.h needs to become /rockbox/build-sim-win64/apps/plugins/fake86/config.h, also for mutex.h
21:04:21user890104actually /rockbox/apps/plugins/fake86/include/fake86/config.h
21:21:08user890104also, what is the purpose of the generated dep file? can't we just describe which files depend on which by makefile rules?
21:23:15***Saving seen data "./dancer.seen"
21:23:21*user890104 suspects some path substitution, which didn't expect a folder with the same name as the plugin inside its source dir
21:28:01 Quit paulk-elm (Remote host closed the connection)
21:58:32 Quit PurlingNayuki (Ping timeout: 264 seconds)
22:08:29__builtinuser890104: its purpose is to save us from having to do just that
22:08:45__builtinnobody wants to spend all day writing dependencies by hand
22:11:23user890104ok, but is there a way to debug why it fails to provide the correct path?
22:12:05__builtinI suspect it's the funky name the header has
22:12:17__builtintry renaming it to fake86_config.h
22:13:26[Saint]That's precisely what he was avoiding and suggested not to do.
22:13:53[Saint]And I kinda agree should be needless.
22:22:11*user890104 learns how rockbox threads work
22:29:09*__builtin needs to decide whether to include a VID/PID switcher or not
22:30:22[Saint]Didn't we have that discussion?
22:30:55[Saint]The reason it didn't happen in the last decade wasn't for a lack of people suggesting it.
22:32:52__builtinI'm still not sure if there's a real reason to have it at all
22:34:18__builtinrbutil support is implemented
22:36:20user890104if it's going to break backwards compatibility with other software that relies on vid/pid matching, shouldn't the release be called 4.0?
22:38:38__builtinwe were planning on calling the next release 4.0 anyway because we're planning to drop HWCODEC too
22:40:20[Saint]It's worth noting a good amount of software that detects USB IDs is terribly broken by Rockbox anyway.
22:40:32[Saint]So kinda a moot point I think.
22:40:38user890104__builtin: and add ipod6g? :)
22:40:53[Saint]See: nearly anything that expects to deal with an ipod
22:41:16__builtinuser890104: rbutil support is the only thing left, I believe
22:41:16[Saint]Fucking things just can't understand the iTunes DB not existing.
22:41:28user890104did you finish the manual?
22:41:36__builtinand that...
22:42:33__builtinlatex is truly annoying, I can't seem to get it to work
22:43:02[Saint]Syntax, or compilation?
22:43:16[Saint]The latter is detailed in the wiki pretty clearly.
22:43:22user890104note: expected ‘void (*)(void)’ but argument is of type ‘void * (*)(void *)’
22:43:25__builtinwell, maybe syntax
22:43:35*user890104 needs help decoding this message...
22:44:00__builtin[Saint]: it just errors out for random reasons which are impossible to track down without commenting out sections and recompiling
22:45:22__builtinuser890104: the arguments you're passing to a function don't match with what it expects
22:45:35__builtinit's expecting a function that returns nothing and takes no arguments
22:45:46__builtinwhile right now you're passing a function that returns a pointer and takes a pointer
22:47:38user890104__builtin: thanks, the C function type syntax is always confusing me
22:49:50 Join saratoga_ [0] (86ae6e0e@gateway/web/freenode/ip.
22:50:29saratoga_[Saint]: changing the USB IDs will probably break things like itunes sync at least
22:51:28[Saint]I'm not convinced it will. Automatic detection, sure.
22:51:39__builtinthat's probably a good thing
22:51:48[Saint]Warrants testing I suppose. iPods unfortunately might be a special case here.
22:51:51__builtinno more itunes interference
22:52:17[Saint]That's kinda a core principle of our operation so I wouldn't celebrate it.
22:52:27[Saint]May as well not dual-boot at all then.
22:52:53[Saint]I'd like to test it first but I'll be damned if I'm installing Windows or iTunes to do so.
22:54:02*user890104 has both
22:55:00saratoga_you need a certain USB ID to work with iTunes
22:55:29 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
22:55:35 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 53.0.2/20170504105526])
23:06:53__builtinuser890104: g#1603 if you want to test
23:06:55fs-bluebot3Gerrit review #1603 at : 3Convert USB targets to Openmoko USB IDs (WIP) by Franklin Wei
23:23:13 Join robertd1 [0] (~root@
23:23:16***Saving seen data "./dancer.seen"
23:29:05 Quit ender` (Quit: Despite the delicious irony, I will not force two heroes to fight each other in the arena.— Evil Overlord List #129)
23:34:00 Part robertd1
23:34:52 Quit cc___ (Quit: WeeChat 1.6)

Previous day | Next day