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 2020-07-12

00:20:52 Join JanC_ [0] (~janc@lugwv/member/JanC)
00:21:00 Nick JanC is now known as Guest54012 (~janc@lugwv/member/JanC)
00:21:00 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC)
00:22:43 Quit Guest54012 (Ping timeout: 272 seconds)
00:27:34***Saving seen data "./dancer.seen"
02:27:37***No seen item changed, no save performed.
02:53:02 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
02:57:55 Join lebellium [0] (
03:00:01 Quit Oksana (Read error: Connection reset by peer)
04:27:38***Saving seen data "./dancer.seen"
04:34:25 Join ZincAlloy [0] (
04:47:42 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
05:05:33 Join petur [0] (~petur@rockbox/developer/petur)
05:42:33 Quit pamaury (Ping timeout: 258 seconds)
06:27:39***Saving seen data "./dancer.seen"
07:41:37 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
07:43:09 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
08:05:40 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
08:19:37 Join ac_laptop [0] (~ac_laptop@
08:27:40***No seen item changed, no save performed.
09:18:52 Join jdarnley [0] (
09:19:06 Quit J_Darnley (Ping timeout: 260 seconds)
09:28:50 Quit rogeliodh (Quit: The Lounge -
09:31:18 Join rogeliodh [0] (~rogeliodh@
09:47:23 Join J_Darnley [0] (
09:49:28Strife89speachy: Your build is still handling it all like a champ
09:50:30 Quit jdarnley (Ping timeout: 256 seconds)
09:50:31Strife89I had another jump-to-conclusions moment when I tried to load album and it didn't start playback immediately ... but it turns out that was because I had a 3000x3000 cover.jpg that was taking too long to process.
09:57:01 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:b8da:7140:c018:98f4)
10:05:46speachyI take it that period of successful operation is unusual?
10:22:46 Quit ac_laptop (Ping timeout: 258 seconds)
10:27:43***Saving seen data "./dancer.seen"
10:37:24 Join ac_laptop [0] (~ac_laptop@
11:56:31Strife89This is easily the best that Rockbox has worked since swapping in the iFlash
11:57:09Strife89The best since I received the iPod, really, since on the HDD it was randomly failing to boo
12:00:26Strife89130 minutes of running time so far, just actually listening to music and enjoying it
12:01:10Strife89And a much better experience (IMO) than doing the same with my phone
12:04:41fs-bluebotBuild Server message: New build round started. Revision 5462907, 295 builds, 10 clients.
12:04:45 Quit livvy (Remote host closed the connection)
12:06:34 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
12:08:02 Quit mendel_munkis (Ping timeout: 256 seconds)
12:17:11 Quit dys (Remote host closed the connection)
12:17:43 Join dys [0] (
12:25:05fs-bluebotBuild Server message: Build round completed after 1224 seconds.
12:25:27fs-bluebotBuild Server message: Revision 5462907 result: All green
12:27:45***Saving seen data "./dancer.seen"
12:30:12speachythat just-completed build was the ATA power management "fix" you've been testing.
13:08:40 Quit J_Darnley (Ping timeout: 246 seconds)
13:11:42 Join J_Darnley [0] (
13:14:57Strife89I have started to notice that there may be an unrelated problem: the power meter still reads 100%.
13:15:48Strife89After 3.5 hours of continuous playback
13:19:44Strife89(At least, I hope it's unrelated.)
13:21:43 Join johnb4 [0] (
13:23:23johnb4the german voice file I built with RB Utility doesn't load properly (on the current dev firmware): error while loading .voice file.
13:24:26johnb4I remember bilgus adding some checks, but I don't recall details.
13:42:42 Quit livvy (Remote host closed the connection)
13:45:19 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
13:48:51 Join __Bilgus_ [0] (41ba23be@
13:49:49__Bilgus_Anyone around intereseted in pictureflow wanna checkout g#2454?
13:49:50fs-bluebotGerrit review #2454 at : Pictureflow Bugfixes & Usability fixes by William Wilgus
13:49:58johnb4speachy: My original report of the distorted output used a dev build from July 4th and you voice file (fuze). Using today's dev build loading of your voice file fails both on the fuze and clip+ for me.
13:55:03johnb4When I build the voice file myself with RB Util on Windows I use "SAPI5 TTS-System", language "deutsch, voice "Microsoft Hedda Desktop, which used to work fine a couple of months ago.
13:56:52__Bilgus_johnb4 system>Debug>Talk_engine_stats any error messages?
13:57:32__Bilgus_Talk Status: ERR(?)
13:59:43 Quit johnb4 (Ping timeout: 265 seconds)
14:19:00 Join johnb4 [0] (
14:21:11__Bilgus_johnb4 system>Debug>Talk_engine_stats any error messages?
14:21:12__Bilgus_Talk Status: ERR(?)
14:27:34johnb4both on Fuze and Clip+
14:27:49***Saving seen data "./dancer.seen"
14:29:02 Join mendelmunkis [0] (
14:34:35johnb4I had a power outage in our house :-(
14:40:35speachyI definitely can't use a MS Windows API to synthesize stuff on the build server, so that's out.
14:42:46speachyand none of the voice strings changed for at least a few weeks so I don't know why somethign would start failing.
14:46:04speachyon the mini2g, the english and english-us voices seem to work, latest build and the voice files I generated in that big batch
14:58:56speachyjohnb4, ERR(4) is comlaining that it's an incompatible file. which doesn't make sense if you're using a current build and your .rockbox dir is sane. (does a disk check come back clean?)
15:09:22 Quit johnb4 (Ping timeout: 256 seconds)
15:14:42 Join johnb4 [0] (
15:15:11johnb4chkdsk says it's fine. I did a fresh unzip of the dev build (on both devices).
15:29:02 Quit livvy (Remote host closed the connection)
15:29:38 Join ShockTohp [0] (
15:30:19 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
15:31:08johnb4Can you try deutsch.voice on the clip+? is the one I just built using rb utility.
15:31:38johnb4You should see the same error if you switch language to Deutsch.
15:37:59speachyI need to install the multiboot loader onto my clip+ as its onboard flash is (nearly) toast.
15:38:11speachyso that's out of commission for the moment
15:39:10 Quit eevan (Quit: Gateway shutdown)
15:39:10speachywhich reminds me.. perhaps a stupid question, but did you confirm you're actually running the new build?
15:39:21johnb4should I try and build the voice file on Linux? Just with the configure script?
15:40:07johnb4but would yield the same file that you uploaded, wouldn't it?
15:42:37johnb4New stupid question during debugging ;-) ... yes, confirmed 546...-200712 in Rockbox Info.
15:43:56 Quit ShockTohp (Ping timeout: 256 seconds)
15:44:42johnb4*No stupid ..
15:48:36speachyif you have espeak installed then yeah, you should be able to rebuild the voice file. it'll default to the same parameters used to generate the voice file that's failing
15:53:41speachyokay, on the mini2g it also failed to load the .de voice file.
15:54:48johnb4THen I will quit for today.
15:55:22 Quit johnb4 (Quit: Nettalk6 -
15:57:39speachyso.. is the generation bad, or is the parsing bad, or both?
15:58:12speachyand what makes .de special?
16:07:12speachythere are 6 untranslated strings in .de, I wonder if that's what's throwing it off.
16:07:26speachyvs english and polish (which have 100% coverage)
16:08:18Strife89Powered on my iPod and it looks like it's figured out what the battery level is: 20%
16:09:02speachyyeah, it's not .de that's broken, it's the fact it's not a complete translation so TALK_FINAL_ID doesn't match.
16:11:39speachy__Bilgus_, thoughts on how to fix this? (I'd say we should leave the strict check on the player code, and fix up the voice file generation to insert pauses for missing strings.
16:12:20 Quit MrZeus (Ping timeout: 246 seconds)
16:15:02 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:b8da:7140:c018:98f4)
16:16:14Strife89This is ... odd. The status bar says the battery's at 20%, yet the debug area for battery info currently shows 9%
16:16:32Strife89So perhaps Rockbox isn't updating the status bar?
16:16:44Strife89Old bugs are new again
16:17:28__Bilgus_speachy will that mess up the fallback to english?
16:17:59speachyIIRC there are no voice fallbacks
16:18:14speachythere's only an English fallback because we insert the english strings into the binary lang files
16:18:35__Bilgus_oh I was under the impression it was automagic
16:19:24__Bilgus_in that case padding the file isn't going to hurt anything
16:23:27speachylooks like the template/fallback code isn't doing the right thing for the genlang output that is used to create voices
16:27:53***Saving seen data "./dancer.seen"
16:36:45speachythis is quite a mess tbh.
16:37:29speachymakes me wonder if it ever worked for non-perfect translations
16:37:50__Bilgus_by wink any prayer most likely
16:44:51speachyokay, the simplest "fix" is to pre-process the lang files through the genlang update process, so the missing strings get copied over from English.
16:50:01speachyokay, that works.
16:53:05speachyit's a simple patch/hack, g#2524
16:53:07fs-bluebotGerrit review #2524 at : genlang: Voiceout mode needs to ensure there are no missing strings by Solomon Peachy
16:54:01speachyif no objections I'll merge it.
16:54:18Strife89Hmmm, it's not charging.
16:55:00speachydoes it charge in the Apple firmware? (I mean, it could be a bad battery..)
16:55:08Strife89Trying that now
16:56:17speachya similar fix will need to be applied to rbutil
16:58:39__Bilgus_why the ws changes in the other file?
16:59:31speachywhoops, I hadn't intended to commit that.
16:59:38speachyit's just whitespace though
17:00:00 Quit lebellium (Quit: Leaving)
17:00:20speachyrbutil relies on a script running on the website to handle the genlang stuff
17:01:00speachyI nuked it last week, as I didn't know what it was, and it had been broken since the migration without anyone complaining anyway.
17:02:54speachynew patchset updated
17:03:06speachygotta fix the voice script too. whoops.
17:03:13speachys/voice script/commit msg/
17:20:30speachytrying to fix the genlang www stuff now.
17:22:56 Quit livvy (Remote host closed the connection)
17:27:30 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
17:53:56speachyokay, the www genlang is working again. Now to tweak it so it preprocesses things properly.
17:54:36 Quit livvy (Remote host closed the connection)
17:55:55 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
17:58:24 Quit MrZeus (Ping timeout: 246 seconds)
18:01:00 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:b8da:7140:c018:98f4)
18:02:00 Quit Oksana (Ping timeout: 258 seconds)
18:06:01speachyI _think_ www genlang is fixed too.
18:06:16Strife89speachy: it does appear to charge outside of Rockbox (I used disk mode).
18:06:29Strife89Same cable and power brick
18:07:20fs-bluebotBuild Server message: New build round started. Revision 980124e, 295 builds, 10 clients.
18:07:47Strife89Rockbox's debug claims it's at 92%, having charged for around 70 minutes.
18:09:26Strife89When the current build round finishes, I'll update all of my DAPs and check whether only the iPod exhibits the odd battery level behavior.
18:09:58speachythe mini2g, xduoox3, and agptekrocker all charge properly
18:10:09speachythe current round is purely voicegen-related, no code changes.
18:10:39Strife89Do they all show the correct percentage on the status bar?
18:11:25speachy...close enough, anyway. the mini2g and x3 have pretty ancient batteries that don't last very long
18:12:23speachyyou're using a 3rd-party battery, right? putting aside the percentage reporting (which could easily be wrong for that battery) things should still _charge_ though.
18:12:56Strife89Yeah, that iPod was sold to me with a third party battery already installed
18:13:07speachyiirc the charging is a hw circuit
18:13:43speachyand rockbox only monitors it, rather than having any controls.
18:15:06speachyoh, there is a choice to charge or not −− look in the settings/battery menu
18:16:20Strife89It's currently set to "yes"
18:16:39speachynot sure what 'force' would be in that context, but eh..
18:17:13speachydo you have the capacity setting set correctly for the new battery?
18:17:35Strife89I set it to 1200mAh, to match the label
18:18:54Strife89I imagine that the capacity setting only affects the "remaining time" estimates.
18:20:22Strife89I'll run batterybench
18:20:32speachywhat kind of charging setup is it? PC or wall-wart?
18:21:05speachyif it's the latter, change that mode from 'yes' to 'force'
18:21:05Strife89Wall wart for the most recent charge attempt.
18:21:36Strife89Might explain things
18:21:53speachy'yes' apparently tries to negotiate USB current supply, which obviously won't work with a wall wart
18:22:39speachybut the downside to leaving it on force is that if you try to connect to a PC/whatever that _can't_ supply a full 0.5a the port will self-disconnect.
18:22:55Strife89It was an Apple wall wart this time (I had it handy for the iTunes restore)
18:24:35speachyjohnb4, if you use rbutil to generate voice files, it should now work again. (It's been broken for a while... and nobody noticed)
18:25:32fs-bluebotBuild Server message: Build round completed after 1092 seconds.
18:25:33fs-bluebotBuild Server message: Revision 980124e result: All green
18:26:32Strife89When I reboot my desktop back into Debian, I should set up the build client
18:27:08Strife89Internet sucks but at least I have a halfway decent CPU
18:27:27Strife89(even if it if an i3)
18:27:37Strife89*it is
18:27:54***Saving seen data "./dancer.seen"
18:30:03speachyokay, kicked off. last time it took 74 minutes but I think it had a partial voice pool.
18:31:00speachy(81 targets, 10 languages)
18:31:03Strife89Any way I can help with that?
18:31:21speachy...completely singlethreaded.
18:31:33Strife89Oh hell
18:31:41Strife89I'm sorry
18:31:47 Quit livvy (Remote host closed the connection)
18:31:53Strife89That's gotta suck
18:32:31speachyit re-uses voice strings if possible, so it would be pretty easy to parallelize across languages
18:33:10 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
18:35:56speachynot practical to farm it out without some rewriting of how this stuff is done
18:58:05 Quit ZincAlloy (Quit: Leaving.)
19:12:35 Quit MrZeus (Ping timeout: 272 seconds)
19:15:11 Quit petur (Remote host closed the connection)
19:28:02 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
19:40:17mendelmunkisDecreasing DEFAULT_UPDATE_DELAY in statusbar-skinned.h makes a statusbar bug happen more often. increasing it seems not to have much of an effect. any advice?
19:40:23 Nick mendelmunkis is now known as mendel_munkis (
19:50:25speachyI got nothing, sorry.
20:15:18 Quit livvy (Remote host closed the connection)
20:16:43 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
20:22:06 Quit livvy (Remote host closed the connection)
20:24:46 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
20:25:07 Join dweeber_ [0] (
20:27:54 Quit dweeber (Ping timeout: 240 seconds)
20:27:57***Saving seen data "./dancer.seen"
20:31:12 Quit livvy (Remote host closed the connection)
20:32:35 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
20:41:54Strife89Anyone available to help me troubleshoot a failure in building the cross compiler? (Specifically with building gcc 4.9.4)
20:47:05Strife89Build log
20:52:06 Quit livvy (Remote host closed the connection)
20:53:32 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
21:07:59 Quit Soap (Read error: Connection reset by peer)
21:14:14 Quit Ckat (Ping timeout: 256 seconds)
21:20:57 Quit sakax (Quit: Leaving)
21:58:00speachyStrife89, what device are you wanting to build for? Or just trying to build everything to be a build client?
21:58:51Strife89The latter.
21:59:26speachywhat base os? Debian?
21:59:37Strife89Yep, stable (10).
21:59:48speachy...which I think has Make 4.3
22:03:42speachyis MAKEFLAGS or MAKELEVEL set in the environment?
22:10:10Strife89I don't think so
22:10:32Strife89echo "$VARNAME", right?
22:10:55Strife89I get a blank line for both
22:12:32Strife89(supporting screenshot)
22:16:08speachydo you have g++ installed?
22:16:41speachynevermind, the log claims you do
22:19:34speachyok, this is stupid.
22:19:57speachyyou built the full rack of other toolchains with this, right?
22:20:05speachymost notably including sh
22:22:14speachythat sets MAKEFLAGS=-r, and nothing clears it for the other toolchain builds
22:23:14Strife89I attempted to, I'm not totally sure that it actually happened
22:24:01speachyI'd bet that if you re-ran the toolchain build, selecting only 'x' , it would have worked.
22:25:05speachyor try applying g#2525
22:25:07fs-bluebotGerrit review #2525 at : build: We don't want '-r' in MAKEFLAGS for toolchain builds. by Solomon Peachy
22:25:34Strife89I'll try that patch
22:25:50Strife89Might take me a bit, I need to read up on patching with git
22:26:44speachyon the upper-right corner of the gerrit page, under 'download' you can cut-n-paste the 'cherry pick' line to do what you want
22:27:05speachyor download the patch and apply it yourself
22:27:59***Saving seen data "./dancer.seen"
22:29:48Strife89Awesome, that cherry pick line helps a bunch
22:30:11Strife89To be specific about what I picked to build, I entered "s m a i x y"
22:30:17Strife89Which I will retry now
22:30:46speachyyep. if you had left 's' out then you wouldn't have seen x (and y) fail.
22:30:59speachyyou can get away with just re-running x y
22:31:26speachyor redo the whole thing if you don't mind burning cpu cycles to prove the patch is good
22:31:51 Join Soap [0] (~Soap@rockbox/staff/soap)
22:31:54Strife89In the interest of time, I'll just do x and y
22:32:30Strife89ROCKBOXDEV: mkdir build-gcc-4.9.4
22:32:30Strife89mkdir: cannot create directory ‘build-gcc-4.9.4’: File exists
22:32:47Strife89I think the script needs a simple failsafe for this situation
22:33:20speachythat is the failsafe. :D
22:33:30speachyrm -Rf /tmp/rbdev-build
22:33:48speachyit won't trash the stuff you've already built, just the WIP stuff.
22:33:50Strife89The whole build dir? ... alright
22:34:02speachythe downloaded files are in /tmp/rbdev-dl
22:34:13mendel_munkisso hopefull next week the build farm will have increased by 1/5
22:34:14speachyand the x-tools end up wherever you set RBDEV_PREFIX to
22:34:45Strife89I .... haven't explicitly set such a variable
22:34:51speachyso /usr/local then
23:00:39Strife89Made it through the gcc build
23:02:43fs-bluebotBuild Server message: New build round started. Revision ab1cf20, 295 builds, 9 clients.
23:12:57 Join Soap_ [0] (~Soap@rockbox/staff/soap)
23:13:32Strife89Wait a sec. It seems to be building gcc 4.9.4 again
23:14:03Strife89ROCKBOXDEV: Starting step 'gcc'
23:14:03Strife89ROCKBOXDEV: logging to /tmp/rbdev-build/build-gcc-4.9.4.log
23:14:03Strife89ROCKBOXDEV: mkdir build-gcc-4.9.4
23:14:03Strife89ROCKBOXDEV: cd build-gcc-4.9.4
23:14:03Strife89ROCKBOXDEV: gcc-4.9.4/configure
23:14:03 Quit Strife89 (Excess Flood)
23:14:05speachythree of the toolchains are gcc494 (mips-baremetal, mips-linux, arm-linux)
23:14:11 Join Strife89 [0] (sid399903@gateway/web/
23:14:28speachysh is gcc404, arm is gcc444, m68k is gcc452
23:14:46Strife89Sorry about the flood
23:14:53speachyand to repeat the line from when you were gone, three of the toolchains are gcc494 (mips-baremetal, mips-linux, arm-linux)
23:14:58Strife89I made a terrible misclick
23:15:47 Quit Soap (Ping timeout: 240 seconds)
23:16:25speachyI'm still hoping to drag everyone up to gcc494 but there are still apparently miscompilation bugs with the ipodcolor..
23:17:35speachy(and from there iterate all targets to newer versions in lock-step)
23:18:28Strife89Idiot me left /tmp on the root partition and I just got a warning that it has 700 MB free
23:18:52Strife89Things might go south in a few minutes.
23:19:28speachyI'm going to hit the sack. good luck with this toolchain stuff
23:19:43Strife89Alright, good night
23:26:23fs-bluebotBuild Server message: Build round completed after 1420 seconds.
23:26:24fs-bluebotBuild Server message: Revision ab1cf20 result: All green
23:26:59Strife8923 minutes, jeez
23:30:04 Quit __Bilgus_ (Ping timeout: 245 seconds)
23:44:09 Quit [7] (Disconnected by services)
23:44:16 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
23:46:43 Quit Natch (Remote host closed the connection)

Previous day | Next day