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

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

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

#rockbox log for 2020-06-29

00:21:17***Saving seen data "./dancer.seen"
00:45:35 Nick mendelmunkis is now known as mendel_munkis (
01:21:23 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:a099:b328:3e74:7ce4)
01:25:59 Quit ZincAlloy (Ping timeout: 272 seconds)
01:42:30 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:a099:b328:3e74:7ce4)
01:46:57 Quit ZincAlloy (Ping timeout: 260 seconds)
01:54:43 Join ZincAlloy [0] (
01:59:19 Quit ZincAlloy (Ping timeout: 260 seconds)
02:21:18***Saving seen data "./dancer.seen"
02:59:20 Join petur [0] (~petur@
02:59:20 Quit petur (Changing host)
02:59:20 Join petur [0] (~petur@rockbox/developer/petur)
04:03:27 Quit dys (Ping timeout: 260 seconds)
04:05:10 Quit craftyguy (Ping timeout: 260 seconds)
04:18:35mendel_munkisspeachy: you may want to take a look at fs#9513
04:18:36fs-bluebot Smarter release scripts (patches, new)
04:21:20***Saving seen data "./dancer.seen"
04:39:36mendel_munkisdoes the plugin api have any functions for extracting basename?
04:47:55 Join dys [0] (~dys@2003:5b:203b:100:a64c:c8ff:fef4:13a6)
04:50:13 Join pcbBob [0] (
05:19:34 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
06:13:21 Quit emacsomancer (Read error: Connection reset by peer)
06:18:45 Join emacsomancer [0] (
06:21:22***Saving seen data "./dancer.seen"
07:11:41 Join ac_laptop [0] (~ac_laptop@
07:40:12mendel_munkisis anyone able to test FS #10089? the underlying cause seems to be gone
07:40:14fs-bluebot Scrolling doesn't pause at the top or bottom of a list on 2nd Gen iPod (bugs, unconfirmed)
08:14:52 Join massiveH [0] (
08:21:25***No seen item changed, no save performed.
08:43:46__builtinmendel_munkis: funny you asked. Quake actually has its own, see apps/plugins/sdl/progs/quake/common.c
08:44:12mendel_munkisSo just do it from scratch. got it
08:45:00__builtinhmm, actually: firmware/export/pathfuncs.h:size_t path_basename(const char *name, const char **nameptr);
08:45:23__builtinit's not exposed via the api though
08:45:35mendel_munkisI just realized what I want isn't basename its basename+ext
08:45:48mendel_munkiswhich is easy enough to do
08:45:57mendel_munkisjust read bac intill /
08:46:55__builtinhmm, exporting those functions via the api isn't a bad idea
08:47:12__builtinbut of course some plugin would have to use it for it to be worthwhile
08:47:16mendel_munkisand here I am busy trying to take things out of the api
08:47:44__builtinoh, why?
08:48:17mendel_munkisso that the core can do weird things without bothering plugins.
08:48:34mendel_munkisalso side benifit of freeing op plugin ram space
08:51:09__builtinah, like what?
08:51:29mendel_munkismostly make the screen stand up and dance.
08:51:36mendel_munkis(or turn on its side)
08:51:52__builtinlike compiz? :)
08:51:59mendel_munkisnot quite
08:52:24mendel_munkisalso I feel like plugin authors shouldn't have to account for screen differences as much as they do
09:13:00 Join skolem [0] (~skolem@
09:40:57 Quit massiveH (Quit: Leaving)
09:46:09 Quit zagor[m] (Quit: killed)
09:46:09 Quit nihilazo (Quit: killed)
09:46:24 Quit blbro[m] (Quit: killed)
09:46:25 Quit danielp3344 (Quit: killed)
09:46:40 Quit kadoban (Quit: killed)
09:58:54mendel_munkisspeachy: what do you think about adding a plugins/new category to flyspray?
10:01:56speachythis this a feature request sort of thing?
10:02:27mendel_munkisI was thinking about all the tasks which include most of a plugin already written
10:03:00mendel_munkisIt makes browsing the tracker easier to exclude or only view those.
10:04:06speachywhat would the text of that sub-category be?
10:05:08speachy(ie how would the reporter know to put stuff in that vs the existing 'plugins' category?)
10:05:30 Quit petur (Quit: Connection reset by beer)
10:07:29mendel_munkisplugins/new plugin?
10:08:03 Join danielp3344 [0] (danielp334@gateway/shell/
10:13:57mendel_munkishow do I go about adding a string to the language files?
10:17:17speachyI don't think there's a tool to do it automagically; when I've done it I just cut-n-pasted the last entry and changed it as needed.
10:17:39mendel_munkisso orginization is a pipe dream?
10:17:44speachythe most important thing is that each 'id' isstring is unique.
10:18:31speachyas that's used in the lang enumeration
10:21:29***Saving seen data "./dancer.seen"
10:21:50speachyI wouldn't call it a pipe dream.. just that there's been little point to organizing it as it's not really meant to be consumed by humans. non-English files are generally created by the translate web site too
10:23:33speachyoh yuck. big spammer registration push on the forum..
10:28:03mendel_munkiswhats user in the lang files?
10:28:14mendel_munkisthe documentation on the wiki is outof date
10:34:08speachyI'm sorry, I can't parse your querstion...
10:34:54mendel_munkisthe lang files have a field user which seems to be core in every entry. I am wondering what it's meeant to be used for
10:36:13speachyah, good question. I don't think anything actually references that.
10:36:42speachynon-core stuff couldn't use languages until very recently
10:39:42 Join kadoban [0] (kadobanmat@gateway/shell/
10:39:43 Join zagor[m] [0] (bjstmatrix@gateway/shell/
10:39:43 Join nihilazo [0] (nihilazoma@gateway/shell/
10:39:49 Join blbro[m] [0] (blbrostrat@gateway/shell/
10:41:19mendel_munkishust to be sure that means double tap left correct?
10:57:45speachyI _think_ so, but don't take my word for it.
11:18:44 Quit pcbBob (Remote host closed the connection)
11:23:35 Join pcbBob [0] (
11:31:06speachysweet, looks like I've managed to get clang's static analyzer to run on top of the rockbox codebase.
11:34:16speachyprobably should kicked it off on a diffeent system. load average is over 16 already.
11:40:49mendel_munkisanyone know if nwza860 keymap is complete?
11:54:37 Quit skolem (Quit: WeeChat 2.8)
11:55:05 Join skolem [0] (~skolem@
11:55:21 Quit skolem (Client Quit)
11:55:37 Join skolem [0] (~skolem@
11:56:01 Quit skolem (Client Quit)
11:56:17 Join skolem [0] (~skolem@
11:59:30 Quit skolem (Client Quit)
11:59:47 Join skolem [0] (~skolem@
12:04:48 Quit skolem (Quit: WeeChat 2.8)
12:21:30***Saving seen data "./dancer.seen"
12:21:58 Join MrZeus [0] (
12:29:14 Quit ac_laptop (Ping timeout: 240 seconds)
12:38:06 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:cf7:bc14:5a92:232)
12:44:04speachyfinally finished. it found over 500 issues:
12:45:54__builtin can you do a color build?
12:46:02__builtinfor sdl apps and puzzles
12:46:13speachysure, it'll take a while though.
12:50:42speachyat least it'll be a parallel build this time
12:51:48 Join ac_laptop [0] (~ac_laptop@
12:58:41mendel_munkisspeachy: some of those are not actually bugs (I think)
12:58:58speachyoh to be sure, there are very likely false positives
12:59:38speachyI did this to see if it might help find why the new gcc on PP targets is assploding so badly.
13:00:58speachythe failures it's reporting at the end are things that even GCC says "... cannot be depended upon to work reliably and are not supported"
13:03:52speachythen there's this construct:
13:04:29mendel_munkiswhats wrong with that?
13:04:45speachynon-static initialization of array length
13:05:21speachygcc accepts it but clang goes "oh hell no". :)
13:06:07__builtinah, FAMs strike back
13:06:11speachythe one time I was able to get a usable stack trace out of the PP crash, it was in the middel of dircache operations
13:09:27 Quit dys (Ping timeout: 260 seconds)
13:19:37 Quit ac_laptop (Ping timeout: 265 seconds)
13:25:29 Join ac_laptop [0] (~ac_laptop@
13:43:02 Join lebellium [0] (
14:00:28mendel_munkisIs adding a string to english-us.lang enough to use it?
14:02:39 Quit atsampson (Quit: Lost terminal)
14:03:05 Join atsampson [0] (
14:09:42speachy'english.lang' is the master, not 'english-us'
14:12:12mendel_munkisthat would explain the errors
14:13:57mendel_munkisby the way whats with the license for the headers from the linux kernel?
14:14:22speachyunless otherwise noted, GPLv2
14:14:51mendel_munkisbut isn't rockbox is released as gpl2+?
14:14:57speachyindividual files may be something different (GPLv2+, BSD, MIT, MPL, etc)
14:18:40speachy824 total "bugs"
14:19:26speachy__builtin: enjoy. :)
14:21:32***Saving seen data "./dancer.seen"
14:21:57__builtinplenty from quake, I see :)
14:22:35__builtinthat ought to keep me busy for a while
14:23:00gevaertsYes, part of the problem with those things is that at least historically we haven't really touched imported code unless really needed
14:25:35speachythere's plenty of issues inside "purely rockbox" code.
14:38:42 Join dys [0] (
14:41:44bluebrotherFS #13198 complains about ipodpatcher and Rockbox Utility not running on Catalina, which isn't surprising since Apple remove 32bit support.
14:41:45fs-bluebot ipodpatcher and Rockbox Utility do not work in macOS Catalina (10.15) (bugs, unconfirmed)
14:42:27bluebrotherAs for ipodpatcher, what's the best solution here? I did build a 64bit ipodpatcher (with the released bootloaders), but it's from current HEAD, thus a different version than we have right now.
14:42:41bluebrotherbut given that Rockbox Utility builds ipodpatcher HEAD I don't see an issue with that.
14:42:58speachyit's not just the 64-bit-only nature; there are also stricter code signing/etc requirements.
14:44:17bluebrotheryeah, but you can ignore that to some extent. You just can't simply left-click things anymore :)
14:44:29bluebrotherright-click, open, confirm warning, runs.
14:44:35speachyis there really any reason to not just tag and build 1.4.n+1 from HEAD?
14:45:31bluebrotherthere's one bug in the Makefile, so it doesn't build the dmg. But that's a dead simple fix. And ipodpatcher was last versioned as 5.0
14:45:45bluebrotherbut yeah, we could tag ipodpatcher 5.1 and build it.
14:51:45bluebrotherhmm, there seems to be a problem with attaching a link to a comment in Flyspray. When I paste a Google Drive link into the comment it works, when adding it using the "attach link" it doesn't. Seems it gets cut off.
14:53:05mendel_munkisany idea why lcd_puts segfaults in the sim with ID2P as the string arg?
15:02:13 Quit dweeber (Read error: Connection reset by peer)
15:10:49bluebrotherhmm, the same issue should be present for sansapatcher
15:13:37bluebrother... and beastpatcher
15:13:48bluebrotherok, not today anymore :)
15:39:35 Join petur [0] (~petur@rockbox/developer/petur)
15:41:35speachyhmm. I wonder if these naked-with-non-simple-asm functions are to blame for the PP screwups..
16:21:33***Saving seen data "./dancer.seen"
16:51:26 Quit dys (Ping timeout: 260 seconds)
16:51:26 Quit benedikt93_ (Ping timeout: 260 seconds)
16:52:13 Join benedikt93 [0] (~quassel@unaffiliated/benedikt93)
16:55:10 Quit ac_laptop (Ping timeout: 260 seconds)
17:44:23 Join ac_laptop [0] (~ac_laptop@
17:45:02 Quit Huntereb (Read error: Connection reset by peer)
17:47:23 Join Huntereb [0] (
17:50:09 Quit lebellium (Quit: Leaving)
17:55:54 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
18:21:34***Saving seen data "./dancer.seen"
18:28:12 Join __Bilgus_ [0] (41ba23be@
18:28:48__Bilgus_mendel_munkis you should start giving a snippet of code with these questions.
18:32:20 Quit pamaury (Ping timeout: 256 seconds)
18:41:43 Join MrZeus_ [0] (
18:44:34 Quit MrZeus (Ping timeout: 246 seconds)
18:46:01speachyI think, at minimum, the arm corelock asm code was broken. I rejiggered it so it no longer uses naked functions (and proper asm annotations) and now the gcc494 pp builds blow up in a different place.
18:47:06speachyinterestingly, I disabled the asm code and moved back to the C corelock code, and it now blows up the same way as the "fixed" asm code.
18:47:56 Quit ZincAlloy (Quit: Leaving.)
18:49:07speachynow I'm geting PANICs about the wrong thread.
18:49:15speachyfor mutex unlocks
18:55:52 Quit petur (Quit: Leaving)
18:59:04 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:cf7:bc14:5a92:232)
19:02:28 Quit ZincAlloy (Client Quit)
19:05:11 Quit pcbBob (Remote host closed the connection)
19:41:09speachysweet, no immediate regressions on the old toolchain.
19:41:33speachyso, anyone with arm asm skillz care to give some feedback on g#2459?
19:41:35fs-bluebotGerrit review #2459 at : ARM: Rejigger the asm corelock functions by Solomon Peachy
20:02:25 Quit ecs (Excess Flood)
20:02:32 Join ecs [0] (
20:21:38***Saving seen data "./dancer.seen"
20:35:15 Quit MrZeus_ (Ping timeout: 258 seconds)
20:47:50 Quit ac_laptop (Ping timeout: 256 seconds)
20:54:43 Join ac_laptop [0] (~ac_laptop@
21:20:17 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
21:20:25 Join craftygu1 [0] (~craftyguy@qutebrowser/dev/craftyguy)
21:20:42 Nick craftygu1 is now known as craftyguy (~craftyguy@qutebrowser/dev/craftyguy)
21:35:59speachyI'd appreciate it if someone else with a PP-based device could double-check that patch to make sure it doesn't cause regressions.
21:37:56 Quit ac_laptop (Ping timeout: 265 seconds)
22:21:39***Saving seen data "./dancer.seen"
22:32:54 Quit craftyguy (Ping timeout: 240 seconds)
22:37:48 Quit marex-cloud (Read error: Connection reset by peer)
22:39:16 Join marex-cloud [0] (sid137234@gateway/web/
22:49:27 Join ac_laptop [0] (~ac_laptop@
23:00:04 Quit mendel_munkis (Ping timeout: 246 seconds)
23:41:06 Quit TheSeven (Disconnected by services)
23:41:16 Join [7] [0] (~quassel@rockbox/developer/TheSeven)

Previous day | Next day