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 2019-10-14

00:05:13 Quit FiletMignon (Ping timeout: 260 seconds)
00:09:51 Join MrZeus [0] (
00:30:13 Quit petur (Remote host closed the connection)
00:31:42 Quit TheSeven (Ping timeout: 276 seconds)
00:31:58 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
00:42:59 Join PimpiN8 [0] (
01:30:35***Saving seen data "./dancer.seen"
01:38:10 Quit PrinceKyle (Quit: In a flash of purple light, the prince disappears!)
01:38:59 Join PrinceKyle [0] (
02:12:07 Join MrZeus_ [0] (
02:14:01 Quit PimpiN8 (Quit: My MacBook has gone to sleep. ZZZzzz…)
02:14:35 Quit MrZeus (Ping timeout: 240 seconds)
02:16:47 Quit MrZeus_ (Ping timeout: 250 seconds)
03:10:07__builtinOK, I'm consistently getting filesystem corruption under Rockbox with the iFlash Solo
03:10:20__builtinparts of the FAT are being overwritten with zeros, evidently
03:13:29 Quit Strife89 (Quit: No Ping reply in 180 seconds.)
03:15:00 Join Strife89 [0] (
03:17:21 Quit lagfra (Ping timeout: 245 seconds)
03:19:10 Join lagfra [0] (
03:30:36***Saving seen data "./dancer.seen"
04:45:50__builtinthis is odd... building the ipod6g bootloader from HEAD results in a bootloader that doesn't work when installed
04:46:16__builtinthe install succeeds with the "alive" tone sequence, but then it just sits there with a black screen
04:46:24__builtintime to bisect...
04:55:48__builtinalright, offending commit is 744274220833e017f3cc3f1b61abd0fc7c3885ec
04:55:55__builtin"iPod Classic: disable IRAM1"
04:59:18__builtinuser890104: can you take a look at this?
05:23:11__builtinfixed, I think
05:23:45fs-bluebotBuild Server message: New build round started. Revision 3d6d903, 280 builds, 11 clients.
05:25:07PrinceKyleMan I wonder if I could still be a build client. I got me an Odroid XU4 here, and it's really not bad.
05:25:33__builtinwe've had RPi clients before
05:25:51PrinceKyleProblem is I've had a couple of glitches with the compiler. Not sure what's happening, but it's giving me error messages telling me to report the bug.
05:26:23PrinceKyleOh wow. I got a 3B+ here, but I really want a 4. I know it could do it.
05:26:26__builtinoh, that's a hairy problem
05:26:48__builtinI've got a build client Docker image floating around somewhere
05:27:42__builtinonly x86_64, unfortunately
05:27:44PrinceKyleIt's hard to figure out. It seems random, happening to different components at different times. My first 3 builds were fine, then I had 2 that bugged out and I had to run the build again, at which time it ran just fine.
05:27:57__builtinis this on the ODROID?
05:28:21PrinceKyleBut that makes me think I'd have a real problem as a client. But maybe my 64-bit C2 will work better.
05:28:34PrinceKyleIt's faster than the RPi.
05:29:21PrinceKyleThis XU4 is the best thing I got now, since my AMD megatower died. But it would be really quite a capable client if not for the random gcc error.
05:29:37__builtinis the gcc error on a rockbox compile?
05:29:44__builtinI'd report that, btw
05:30:40***No seen item changed, no save performed.
05:31:06PrinceKyleYeah on the Odroid. The toolchain built just fine, but I've gotten two random errors that said report the gcc bug when running Rockbox builds.
05:31:27PrinceKyleOne in part of rbcodec and another in one of the game plugins.
05:31:44__builtinpaste it?
05:31:56__builtinwhich gcc version, btw?
05:32:15PrinceKyleThe one that builds.
05:32:35__builtinhow about the host gcc?
05:33:05PrinceKyleOh that's 8.3.0.
05:33:26__builtinit might be a quirk of that host gcc
05:33:40PrinceKyleI haven't tried to build for anything other than sansa, using the ARM compiler.
05:34:05__builtincan you paste the errors?
05:34:11PrinceKyleIt could be. I don't seem to have problems building anything else, but one never knows.
05:35:13PrinceKyleI don't have it now. But let me try building again and see if I get it. I just helped find a ClipZip problem, so any other build wouldn't be anything I'd use for now.
05:35:39PrinceKyleI con't mind testing it. Maybe rebuild the toolchain if it can help.
05:36:38PrinceKyleOK here goes. I'm not sure if I'll get the error or not, but I'll paste it if I do.
05:37:01PrinceKyleOf course part of the problem could be that I'm running 8 threads.
05:37:22BilgusDo you have 8 cores?
05:37:43PrinceKyleYes. But 4 are A15 cores and 4 are A7.
05:38:05PrinceKyleIt's the big/little thing. I don't actually notice the A7 cores most of the time.
05:38:14Bilgushuh sounds interesting
05:38:20PrinceKyleThis thing is actually pretty fast.
05:38:43Bilgusalso I have an updated that I use for the newer gcc builds
05:39:05PrinceKyleOh. That may help. I'm running the one in git.
05:39:12Bilguslet me figure out what IO've done with it this time
05:39:58PrinceKyleDid you see my report on the ClipZip by the way? I found out that 3.13 is working, so it's something between 3.13 and 3.14 that broke this specific sd.
05:40:00__builtinBilgus: "newer build" as in newer host GCC, or newer output toolchain?
05:40:37Bilgushost I learned not to screw with the toolchains
05:40:52__builtinthat'd be nice to get into master
05:41:15*__builtin hasn't gotten a toolchain to compile successfully for a while
05:41:41PrinceKyleOops I caused the error this time. Had `make -J8' instead of `make -j8'.
05:41:47*__builtin thinks he was hitting a compiler bug with a gcc 9.2.0 host
05:41:49PrinceKyleIt's going now though.
05:42:55Bilgusonly thing that might mess you up is it is silent unless errors are present I got REALLY tired of having to parse errors
05:43:36fs-bluebotBuild Server message: Build round completed after 1191 seconds.
05:43:37fs-bluebotBuild Server message: Revision 3d6d903 result: All green
05:43:40Bilguseasy enough to change though
05:43:46PrinceKyleI actually like silence when things are going well. Using a screen reader, it helps to know only if something is wrong.
05:44:55PrinceKyleI'm in the game plugins now. No error so far.
05:47:59BilgusPrinceKyle, I saw that you said 3.13 was working but I'd still check the card to be sure I'll look at whats changed since then
05:48:43__builtin... bisect ...
05:49:04Bilgushere would be my first guess lol
05:49:24PrinceKyleOK I'll try another one. I think I have either a silicon power or a samsung somewhere. I really don't want to pull it out of my phone haha
05:49:50Bilgusif you have builds just try putting retry back to 100
05:50:03PrinceKyleStill not sure why it works on ClipPlus but not ClipZip.
05:50:05BilgusI bet you the card is just slow
05:51:04PrinceKyleOh it may be. As I recall, when I bought the Transcend, it was slower than either of the others. But I didn't think a read speed of I believe 80MB/s would be too slow for rb.
05:51:27Bilgusall the other commits look to me like they'd increase compatibility
05:51:50PrinceKyleThis build was perfect by the way. No errors to paste sadly. It's like that car that won't start that you tow to the mechanic and it starts right up.
05:52:26Bilguswell thats kinda why I'd like you to check if your card is real before we get too far into it I might just be crap flash
05:52:53Bilgusafter that bisect will be the way to go and figure out exactly where it breaks
05:53:26PrinceKyleYeah. Could be. Strange that it has no problems on 3.13 or on any build on ClipPlus. But fake flash can certainly work that way.
05:53:44PrinceKyleOr just a bad controler.
05:55:26PrinceKyleWell, I'll probably get back to this in the morning. Got to get an early start. If I get an error to paste with a build, I'll paste it, and I''ll run some tests on this sd as well.
05:55:29BilgusId conclude it was voltage scaling but we already tried that
05:56:21PrinceKyleYeah it's bad me thinks. But I'll definitely run some tests.
06:00:09Bilgus3.13 hqappened at the beginning of 2013 for reference
06:00:45Bilgussorry end 0f 2013 beginning of 2014
06:01:40Bilgushmm and the only changes for the sd code are voltage scaling
06:03:11Bilgushelps if you look at the right file smh
06:04:40PrinceKyleWeird indeed. I have to conclude it's a bad SD, which may possibly explain another problem I had. I copied all the files off to a networked drive, reformatted the SD in my computer and then copied all the files back. But after the copy, I found quite a few files and folders missing, but I got no error messages while copying. Yeah, it's probably bad, and I've just sent everyone on a wild goose chase. But I'll definitely
06:06:04Bilguswild goose chases are just a way to get re-familiarized with code you haven't looked at in a while
06:06:12PrinceKylehehaheha yeah.
06:07:09PrinceKyleI also see some changes to the bootloader since my 2015 build as well. Do I need to grab the OF again to install the new bootloader, or is there a way to do it without that?
06:07:28Bilgustheres enough changes that deal directly with that issue if the card isn't bad then I'd start bisecting it
06:07:58PrinceKyleOK so I don't need to worry about the bootloader?
06:08:04Bilgusthe new bootloader gives you the ability to boot off the sdcard
06:08:18Bilgusbut i'd hold off till you get to the root
06:08:28PrinceKyleAH good.
06:08:56Bilgusfigure it has the same sdcode in it if it turns out to be borked then there will be some stuff getting changed
06:09:14PrinceKyleOh yes.
06:09:35PrinceKyleSo I'll get back on this first thing in the morning here. Thanks much for the help.
06:09:57Bilgusnp ciao
07:17:06 Join ZincAlloy [0] (~Adium@2a02:8108:9440:dfc:604e:4fc3:5395:24d2)
07:21:22 Quit ZincAlloy (Ping timeout: 246 seconds)
07:30:23 Quit St3ak (Quit: Free ZNC ~ Powered by LunarBNC:
07:30:42***Saving seen data "./dancer.seen"
07:43:24 Join St3ak [0] (
07:52:50 Join deevious [0] (~Thunderbi@
07:54:56 Join ZincAlloy [0] (~Adium@2a02:8108:9440:dfc:604e:4fc3:5395:24d2)
08:00:26 Quit ZincAlloy (Ping timeout: 250 seconds)
08:31:05 Quit Strife89 (Ping timeout: 268 seconds)
08:58:53 Join PimpiN8 [0] (~textual@
09:00:16 Join petur [0] (~petur@
09:00:16 Quit petur (Changing host)
09:00:16 Join petur [0] (~petur@rockbox/developer/petur)
09:15:00 Join dys [0] (~dys@2003:5b:203b:100:a64c:c8ff:fef4:13a6)
09:28:56 Quit PimpiN8 (Quit: Textual IRC Client:
09:30:43***Saving seen data "./dancer.seen"
09:47:00 Join Strife89 [0] (
10:03:34 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:38:45 Quit St3ak (Quit: Free ZNC ~ Powered by LunarBNC:
10:45:39 Join St3ak [0] (
10:58:27 Quit Strife89 (Quit: No Ping reply in 180 seconds.)
10:59:20 Join jdarnley [0] (
10:59:49 Join Strife89 [0] (
11:01:13 Quit J_Darnley (Ping timeout: 240 seconds)
11:12:22 Join PimpiN8 [0] (~textual@
11:30:44***Saving seen data "./dancer.seen"
11:58:35 Quit St3ak (Read error: Connection reset by peer)
12:04:33 Join St3ak [0] (
12:10:43 Quit PimpiN8 (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:11:45 Join PimpiN8 [0] (~textual@
12:32:28 Quit PimpiN8 (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:52:09 Join vmx [0] (
13:30:47***Saving seen data "./dancer.seen"
13:53:02 Join massiveH [0] (
13:57:02 Join ZincAlloy [0] (
14:01:13 Quit ZincAlloy (Ping timeout: 240 seconds)
14:27:19 Join PimpiN8 [0] (~textual@
14:30:50 Quit petur (Read error: Connection reset by peer)
14:37:09 Join petur [0] (~petur@rockbox/developer/petur)
15:18:49 Join speachy [0] (40eebded@
15:30:50***Saving seen data "./dancer.seen"
15:39:30 Quit massiveH (Quit: Leaving)
15:59:57 Quit vmx (Ping timeout: 265 seconds)
16:09:02 Quit PimpiN8 (Quit: Textual IRC Client:
16:11:39 Join PimpiN8 [0] (~textual@
16:38:11PrinceKyleI think there are still tests I can run, but this sd is being reported as the real deal. Could be just a bad controller or something, but it's hard to say at this point, and it seems I don't have another 128GB handy at the moment. I do have 64GB, so I'll try that with the latest build.
16:42:04PrinceKyleI also ran 2 builds for ClipZip, but still couldn't get the error. I'll still update the toolchain using the updated and see what that does.
16:57:15 Quit PimpiN8 (Quit: My MacBook has gone to sleep. ZZZzzz…)
17:30:53***Saving seen data "./dancer.seen"
17:32:02 Join Rower [0] (
17:32:18 Quit Rower- (Read error: Connection reset by peer)
17:33:05 Quit petur (Read error: Connection reset by peer)
17:39:39 Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA)
17:55:09 Join Moarc [0] (
17:57:53 Join ZincAlloy [0] (
18:01:04 Join vmx [0] (
18:01:55 Quit ZincAlloy (Ping timeout: 240 seconds)
18:22:56 Join ZincAlloy [0] (~Adium@2a02:8108:9440:dfc:147e:d782:d4dc:a793)
18:27:42 Quit pamaury (Quit: Konversation terminated!)
18:54:04 Join textmate [0] (
18:59:57 Quit PrinceKyle (Ping timeout: 250 seconds)
19:09:06 Join PrinceKyle [0] (
19:15:25 Join MrZeus_ [0] (
19:30:54***Saving seen data "./dancer.seen"
19:41:33 Quit dys (Ping timeout: 250 seconds)
19:57:15 Quit MrZeus_ (Ping timeout: 240 seconds)
20:07:10 Join lebellium [0] (
20:24:42textmatehey goys
20:24:48textmateSo i have this issue
20:25:11textmatePretty much; ive got this ipod video (5th gen)
20:25:35textmateand i bought a 3000mAh thicc battery upgrade for it
20:26:08textmateIt unfortunately looks like the battery simply isnt getting recognised and doesnt werk
20:28:32speachyhmm. don't think that has anything to do with rockbox; IIRC those batteries are all "dumb"
20:29:32textmateI know it doesnt, but im still asking here since im not sure where else
20:29:51textmateso I dont believe this really *that* offtopic
20:31:29speachy3rd-party batteries are notoriously untrustworthy
20:31:52*__builtin has had good luck with ifixit
20:32:09textmatewell it is an oem LG battery speachY
20:32:19textmatejust sold in a kit with the connector
20:33:57speachyconnector? thought that generation used a ribbon cable.. stock was something like 1000mAh too.
20:34:25textmateyeah well i meant that
20:37:20speachymost likely it's a defective battery.
20:37:38speachyis it simply not powering up, not holding a charge, or what?
20:38:04textmateIt doesnt power up
20:38:22textmateits just stuck at the "Please Wait, Very Low Battery"
20:38:42textmateive left it charging a few times overnight, waking up to that same message
20:38:57textmateI have a few other batteries and they work
20:39:32gevaertsThe ifixit replacement battery for those is 850mAh. 3000 sounds like an exaggeration at least, a scam at worst
20:39:40speachyso it's not charging. Wonder if there's some sort of protection circuit that's being confused.
20:40:14textmategevaerts: The ali express seller said it was 4000mAh
20:40:18speachygevaerts, there are high-capacity batteries but they require case mods and/or solid state storage.
20:40:23textmatebut its printed on the battery as 3000mAh
20:40:40speachyheh, that's like "each sticker on the car is good for 5HP!"
20:41:17speachyif the battery is the same size physically as the stock battery (ie no case mod required) it's a scam...
20:47:37Bilgusspeachy I know you did some voice updates, and gevaerts might know as well...
20:48:12BilgusI was messing with voice in menus and I noticed the voice prompts are out of sync
20:49:03Bilgusso like the battery level is voiced but it says buffer size 20%
20:49:12Bilgusbuffer size 10%
20:49:16*gevaerts tries to figure out the physical size of a thick video battery
20:51:05Bilgusthe first few menu prompts are right 'Files' 'Database' 'Resume Playback' 'Settings' 'Recording' 'Fm Radio' then..
20:51:24Bilgus'TRACK SAVE' instead of playlist catalogue
20:51:47Bilgus'Plugins' (correct)
20:52:09Bilgus'System' (correct'
20:52:37Bilgus'Save Changes' instead of 'Shortcuts'
20:52:52gevaertsAccording to wikipedia, LiPo batteries are 250 to 730 Wh/L, which (assuming 3.7V average) is 70 to 200 Ah/L, which is 70 to 200 mAh/cm^3, and as far as I can figure out the thick ipod video battery is around 6cm^3, so it's going to be between 420mAh and 1800mAh
20:53:19textmatespeachy: Bilgus: i can tell you did last quite a while−− while it lasted
20:53:21gevaertsBilgus: I don't know a lot about this, but are the voices for the same build?
20:53:40textmatei could easily listen on one charge for many weeks with 1 hour a day
20:53:52Bilgusin te settings menu you get to Theme Settings and it says 'Central European'
20:53:57textmateAround a month of nearly everyday 1 hour listening
20:54:20textmateIt was definitely better than the stock battery at least
20:54:27BilgusI grabbed them off the build page I assume they are current with the dev build or is there some way to re-generate them
20:55:33BilgusIt really isn't an issue for me but I bet it'd drive blind users batty
20:57:16gevaertsYou can generate voice files with rbutil I believe
20:57:39gevaertsAnd you can do a voice build from configure if you have the right stuff installed
20:58:00speachybefore I disappeared for a while I recall there being something up with voice files on the site not being generated properly.. don't recall the details
20:58:03Bilgusdoes the build system do that automatically as well?
20:58:30Bilgusah that sounds like a likely cause speachy
20:59:21gevaertsNo. IIRC the daily build script on the server does that, but I have no idea how that works. It uses festival or espeak I think, which many "real" users aren't happy with anyway, so AFAIK most people who use this seriously build their own with rbutil
20:59:31Bilgusmaybe they just haven't been regenerated in a long while
20:59:50gevaertsThat way they can use whatever voices and system their OS has available, with the settings they like
21:01:25BilgusI'll try generating them with rbutil tonight and hopefully they line up properly (but maybe there should be a note about this on the extras page)
21:02:25Bilgusthis is clearly inaccurate 'we have daily updated English voice files for download'
21:06:26Bilgusthe time stamps make it seem as if the voice files are current so maybe we are missing something that needs updated in the lang files?
21:18:45 Quit speachy (Ping timeout: 260 seconds)
21:30:59***Saving seen data "./dancer.seen"
21:38:43 Join J_Darnley [0] (
21:41:36 Quit jdarnley (Ping timeout: 240 seconds)
22:23:20Bilgusok so even generating them locally causes the same issue
22:23:53Bilgusbuildfuze+voice$ ../tools/configure −−type=avs −−target=sansafuzeplus −−ram=-1 −−language=0 −−tts=f −−voice=-1
22:25:37Bilgusall I can figure is the Voice IDs don't match
22:55:31__builtinah, the joy of legacy codebases :)
22:56:00*__builtin wonders how many other things are broken without us realizing
23:01:01gevaertsIf I had to guess, I'd speculate on the plugin translations?
23:01:25gevaertsBut that's really only because those are numbered text things too
23:02:50__builtinbootloaders too
23:06:33__builtinhow are those built by the autobuilder?
23:10:34Bilguswell thats the last thing added so it seems like a likely culprit but it looks as if the IDs are getting generated in the right order
23:11:38 Quit lebellium (Quit: Leaving)
23:12:07Bilgushmmm mp3 file ./LANG_CHESSBOX_GAMES.mp3 not found!
23:13:05Bilgusmp3 file ./LANG_BOOKMARK_SHUFFLE.mp3 not found!
23:14:05BilgusI wonder how it handles if the file doesn't exist it seems to skip over but I bet it doesn't have any manner of correcting it in the file index
23:15:50Bilgusno they are just blank strings so probably not it they just happen to be near the strings that are wrong :/
23:15:54__builtinwhat's the "score" field (6th field) in www.git/rbmaster/builds?
23:16:02Bilgusmaybe I should listen to the generated clips
23:16:49__builtin(actually, I think I know what it *is* −− just not how it's generated)
23:21:36__builtinwow, this is quite the kludge...
23:22:44__builtinapparently the score is (the total user/kernel time)*100, rounded to the nearest integer
23:23:24__builtinthe only problem here being that the time is completely unreproducible between systems... :(
23:31:01***Saving seen data "./dancer.seen"
23:32:30Bilgusok so the right voice file is being generated still unsure how voice Ids get mapped to the .voice file though
23:43:31 Join MrZeus [0] (
23:49:27Bilguslang_enum.h doesn't match the generated voice IDs
23:50:32Bilgusnow the question becomes WHY

Previous day | Next day