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 2020-08-07

00:08:24speachyyeah, I'm packing it in too for now; since my little reclocking experiment turned into another rabbit hole
00:14:38fs-bluebotBuild Server message: Build round completed after 903 seconds.
00:14:41fs-bluebotBuild Server message: Revision e0bb30a result: All green
00:14:42fs-bluebotBuild Server message: New build round started. Revision e0bcb0f, 280 builds, 8 clients.
00:27:05***Saving seen data "./dancer.seen"
00:29:56fs-bluebotBuild Server message: Build round completed after 915 seconds.
00:29:57fs-bluebotBuild Server message: Revision e0bcb0f result: All green
00:29:58fs-bluebotBuild Server message: New build round started. Revision 89c7a4e, 280 builds, 8 clients.
00:44:06fs-bluebotBuild Server message: Build round completed after 848 seconds.
00:44:07fs-bluebotBuild Server message: Revision 89c7a4e result: All green
00:58:59__builtinalright, just like that we have a (semi-working) sdl2 port
01:01:25__builtinalthough something appears to have broken the normal SDL sim builds...
01:03:01__builtinor far more likely, they were broken all along and I just never noticed
01:44:52 Quit ac_laptop (Ping timeout: 256 seconds)
02:21:56 Quit ufdm (Ping timeout: 256 seconds)
02:27:09***Saving seen data "./dancer.seen"
02:29:04 Join ufdm [0] (
03:25:13 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
04:03:05 Quit Oksana (Ping timeout: 240 seconds)
04:11:06 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
04:27:13***Saving seen data "./dancer.seen"
04:57:03 Quit koniu (Ping timeout: 240 seconds)
04:58:09 Join koniu [0] (~koniu@gateway/tor-sasl/koniu)
05:14:27 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
05:16:04 Join lemon_jesus2 [0] (
05:18:02 Quit lemon_jesus (Ping timeout: 260 seconds)
05:18:02 Nick lemon_jesus2 is now known as lemon_jesus (
05:18:54 Quit pamaury (Ping timeout: 244 seconds)
05:38:43 Quit scorche (Disconnected by services)
05:38:46 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
06:00:14 Quit S|h|a|w|n (Read error: Connection reset by peer)
06:27:14***Saving seen data "./dancer.seen"
06:31:19 Join heredoc [0] (heredoc@2a01:7e01::f03c:91ff:fec1:de1d)
06:57:43 Quit livvy (Ping timeout: 240 seconds)
06:58:23 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
08:13:20 Join massiveH [0] (
08:14:27mendel_munkisis logbot's code available?
08:26:21speachyyeah, it's pretty ancient though.
08:26:43speachyand under a funky license
08:27:17***Saving seen data "./dancer.seen"
08:45:00mendel_munkisspeachy: where would I find it?
08:54:15speachyI can gauaranteee you're going to have some issues finding its dependencies and getting it to build on modern systems
09:29:01 Quit massiveH (Quit: Leaving)
10:01:17 Join MrZeus [0] (
10:10:09 Join LambdaCalculus37 [0] (LambdaCalc@gateway/vpn/privateinternetaccess/lambdacalculus37)
10:17:49__Bilgus_oh, speachy did you ever check out that announce plugin?
10:20:34 Join ac_laptop [0] (~ac_laptop@
10:27:20***Saving seen data "./dancer.seen"
10:27:48LambdaCalculus37Found my Gigabeat T the other day. It still is stuck on the restore screen after 10 years of not having attempted to port Rockbox to it.
10:28:16__Bilgus_at least its consistent?
10:28:24LambdaCalculus37Maybe I should give it another crack one of these days?
10:29:18LambdaCalculus37The port itself should be relatively easy to do, as the hardware is the same as the Gigabeat S. It's just finding out the signature exploit to inject our code into it.
10:30:06LambdaCalculus37Heh, look at me saying "our code"... I'm not a Rockbox dev anymore. :P
10:36:30LambdaCalculus37But in all seriousness, I have been giving the Gigabeat T some thought. I need to look for all my notes and firmware updates to see if maybe I can tackle it again.
10:37:14__Bilgus_once a dev always a dev?
10:37:35LambdaCalculus37Guess so!
10:38:19LambdaCalculus37Speaking of firmware updates, I've noticed that a ton of linked firmware updates across the wiki pages, for both supported targets and unsupported targets, are now broken links. This can potentially put a major damper on people needing these firmwares for either research purposes or for restoring players.
10:38:59speachywe're going to have to mirror things ourselves
10:39:22speachyif doesn't have mirrors, anyway
10:39:26LambdaCalculus37At this rate, I doubt very much that most of these companies are even going to give a crap about software for devices that they don't even support anymore.
10:39:29__Bilgus_I'm not sure of the copyright aspect of that
10:39:44LambdaCalculus37That's the slippery slope.
10:39:46speachywe already have all manner of datasheets and whatnot linked as wiki attachments
10:40:10__Bilgus_probably better to host them ourselves and keep a baclkup on the infra
10:40:42speachyfor my printer work I have every bit of documentation/sdk/etc I've ever found stashed into a private git repo
10:42:29LambdaCalculus37That's a pretty good way of keeping stuff together and organized.
10:44:23speachya lot of it was provided under NDA
10:45:35speachybut even with proper relationships in place the companies are big and I've received nastygrams from one group while simultaneously being paid by another. :/
10:46:23speachybut these old firmware and datasheets for products that haven't been sold (or supported) for a decade, and/or were abandoned three corporate mergers ago..
10:50:46speachy$#%^#$@$ now I'm getting failures to read the SD card. even after I backed out the changes I'd made.
10:52:17LambdaCalculus37Anywho, I'm going to be getting another Sansa c200 from fleaBay soon (not sure if it's v1 or v2 yet), but I don't remember one detail: I know Rockbox allows use of SDHC card on some devices, but would SDXC (i.e. > 32GB) work on these as well?
10:52:50__Bilgus_they should
10:52:54speachyyeah, you'll need to reformat them as fat32 first
10:53:12LambdaCalculus37Okay, good to know. :)
10:54:20LambdaCalculus37I've got a couple of spare 32GB and 64GB cards I was using in my Raspberry Pis that I swapped for larger cards, so I can put one of those spares to use in my Sansa.
11:05:40speachywell well well, the stricter partition validation code worked
11:06:03speachywhen my player crashed the last time it partially trashed the first sector of the filesystem
11:11:52 Quit t0mato (Quit: Ping timeout (120 seconds))
11:21:42 Join t0mato [0] (~t0mato@
11:22:39 Quit sakax (Remote host closed the connection)
11:33:20 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
11:43:10speachysweeet. the X3 dynamically reclocks between 192 and 576 MHz
11:43:44speachycan probably clock lower on the low side
11:44:08speachythe audio codec PLL rejiggering has proven to be a lot messier
11:45:52speachythe jz4760's clocking tree is theoretically very flexible but there are so many restrictions and exceptions...
11:46:13 Quit pamaury (Ping timeout: 272 seconds)
11:49:41fs-bluebotBuild Server message: New build round started. Revision eb0e41c, 280 builds, 10 clients.
11:52:26speachythe x3 should see noticably better battery life with this. still working on the pll1/audio rejiggering.
11:58:03 Quit livvy (Ping timeout: 240 seconds)
12:06:47fs-bluebotBuild Server message: Build round completed after 1025 seconds.
12:06:50fs-bluebotBuild Server message: Revision eb0e41c result: All green
12:06:58__Bilgus_now we just need to push you to the rocker :p
12:18:37speachywithout kernel sources there's not a lot that can be done to improve the underlying platform
12:18:58__Bilgus_true, bastards.
12:19:00speachythat's one of the main reasons I abandoned mine for the x3
12:19:20speachythat, and much better physical robustness, and an actual like out
12:19:44speachythe back of my rocker is being held together by tape. stupid glass.
12:23:51__Bilgus_looks like its a bit hard to procure
12:24:03__Bilgus_I see the x3II
12:24:17speachyI have a spare brand new one I found on ebay for cheeeep
12:24:57__Bilgus_wanna trade for a NOS clip zip?
12:25:37speachynew old stock, ah
12:26:13__Bilgus_yeah circa 2010 I think
12:26:56speachysure, why not. since I managed to source some batteries via aliexpress the main reason I bought the spare is moot
12:27:23***Saving seen data "./dancer.seen"
12:27:30speachyits only weakness is the horrid screen. its funky key layout is actually really nice for using it by touch
12:28:12__Bilgus_I have the fuze+ in my truck its starting to show its age
12:28:53__Bilgus_probably a new battery might help but TBH its not my favorite
12:29:20__Bilgus_give me few days i'll get with you on trading mailing addresses
12:30:15speachyI still have my old clip+ that's usable (could use a battery swap) for ams hackery but daily-driver-wise, the x3 has been awesome.
12:30:53__Bilgus_I think I bought six gave 2 for gifts and the one I'm using now is missing button presses
12:31:20__Bilgus_at this rate I doubt I'll get to them before the batteries die from old age
12:31:42__Bilgus_I picked them up for $35
12:32:02speachyhuh, there's a brand-new x3 on ebay for $75+s/h
12:32:16speachyand a rockbox'd one for 175
12:32:52speachyand a HK seller with a bunch for 120, also new.
12:33:11speachythe one that has 'em for 75 has a bunch
12:34:14speachyman, this 4760 audio codec clocking crap is.. wtf.
12:35:22speachywas able to downclock the 48KHz mode at least, but 44KHz might be stuck with the PLL blasting @ over 500MHz due to the PLL and divisor restrictions.
12:36:17__builtinlicense lawyers: can we incorporate GPLv3 code into our codebase?
12:36:18__Bilgus_lol the lower quality uses more battery
12:36:26__Bilgus_cant write this shit :p
12:36:49 Join johnb2 [0] (
12:37:04speachywell, if I can come up with a different PLL base clock within the divisor restrictions..
12:37:21__builtinI'd suspect that the resulting combination would be covered by GPLv3
12:37:33speachy__builtin, yep.
12:38:00__builtinso covered by GPLv3 then?
12:38:01speachyI think rockbox is gpl2+ so it would be legally kosher
12:38:20speachybut we'd effectively end up moving rockbox gpl3+ if we did that
12:38:46__builtinyeah, that's the problem
12:38:51speachy(personally I much prefer gplv3)
12:39:19speachy(unless we have some sub-component that's GPL2-only..)
12:39:23__builtinperhaps we can build some technical means to isolate the GPLv3 portions
12:39:55__builtinI'm wanting to port SDLPoP so it'd just be a plugin
12:40:04gevaertsThe resulting *binaries* would be gplv3
12:40:24speachy__builtin, if it's a plugin we can probably treat the plugin boundary as a GPL barrier.
12:40:26gevaertsThat doesn't mean the source license has to be upgraded
12:41:02speachynot unlike linux's userspace boundary.
12:41:06__builtinspeachy: a barrier in what sense?
12:41:52gevaertsHmmm, given that you have to build everything together and there are very tight version constraints I'm not convinced that's true
12:42:02speachywe'd have to basically state that we don't consider rockbox plugins to be deriviate works of rockbox itself
12:42:39__builtinthat's a stretch
12:42:53gevaertsBut yes, I would have questions about declaring SDL to be a derivative of rockbox :)
12:43:16speachyplugins are independent binaries using a published system call interface.
12:43:46*gevaerts doesn't remember the wording about things being distributed together
12:43:47johnb2speachy : I used to login into gerrit using my yahoo ID, but this isn't working anymore. Is this a known issue?
12:44:00speachythe thing about GPL violations is that (1) someone has to care, and (2), that someone has to have the right to enforce it
12:44:19speachyjohnb2: yeah, yahoo disabled openid authentication
12:44:36johnb2what is a good alternative?
12:44:40speachythere are only two parties that can object −− us, and the authors of that plugin
12:45:53speachyjohnb2: I can't comment on generic openid providers. there are ones you can deploy and run yourself, or google and github are also enabled and work okay
12:46:10__builtinso they'd object over GPL3'd binaries being used in violation of GPL3 (but in compliance with v2)?
12:46:16johnb2yeah, just found that my github account works.
12:46:26speachyjohnb2: if you log in with a different provider I can fix things up so the duplicate accounts are linked
12:46:40speachyor rather, dedup things
12:46:51speachyfor any old stuff you have lying around
12:47:14gevaertsI don't actually think it matters at all. I mean, in the derivative work case, the entire binary distribution becomes gplv3, so you can get the source under gplv3. In the non-derivative case, you get the source as a gplv2+/gplv3 mix. You can *also* just get the source without invoking the "I got the binary, give me the source" clauses, in which case you also get the v2/v3 mix
12:48:10speachygevaerts: I'm speaking in general terms, but yes, you're essentially correct. But I go back to my original point about that effectively making rockbox gplv3
12:48:48speachythe original files are mostly v2+ but the final result is v3[+]
12:49:28gevaertsI think for rockbox the only case in which there would be a real difference would be if people hold patents on the stuff they contribute. I'm fairly confident in guessing that that hasn't happened yet
12:49:57__builtinperhaps the tivoization clauses too?
12:50:49__builtinin fact that's probably more relevant since we're largely a firmware project
12:51:23gevaertsWell, we only run on things that we managed to run things on
12:52:15gevaertsSo yes, in theory that might matter, and it could in theory block some official manufacturer ports, but I mean, how likely are those in the first place?
12:53:14__builtinheh, right
12:55:06gevaertsIf gplv3 code is committed though, I'd suggest putting something in docs/ that says which bits someone who really wants v2 binaries would have to disable (or fully committing to v3). I don't think it's worth the effort to go as far as things like ffmpeg, where you can pick the desired license compatibility in the build options
12:56:32gevaertsBut then that's all my opinion. I can't be reasonably considered "active" these days, so I don't think I'm entitled to a formal voice in this :)
13:00:23speachyeveryone who has code committed counts
13:01:56speachygevaerts: but yeah.. I agree about putting in a V2_ONLY build option. as long as it's just plugins though I don't think it'll be terribly big of a deal.
13:02:48gevaertsWell, in some ways, yes, (a) but we're already "2+", so switching to v3 is allowed, (b) I think the future should be decided by those who are active, and (c) I think there's no real effect anyway
13:03:30speachyhey, but we get to wank philisophical about licensing, that's an effect! :D
13:03:37gevaertsTrue :)
13:04:02gevaertsAnd not allowing v3 blocks some code, which could encourage people to work on things other people may find more important!
13:04:15gevaertsOr it could discourage them to work on things at all, of course
13:05:24__builtinI like the V2_ONLY idea
13:05:54__builtinprobably with an associated configure switch
13:06:04__Bilgus_because its the only easily tenable one?
13:06:27gevaertsI like the concept, but I think actual configure support can be delayed until someone actually needs it (or until someone really wants to work on it)
13:06:59__Bilgus_mmh I disagree
13:08:40__Bilgus_Can't change something like that and expect to push on the other work to the next guy
13:09:21__Bilgus_erm person
13:11:04__builtinwhat other work?
13:11:08 Join beencubed [0] (~beencubed@
13:11:26__builtinsomeone who can't comply with GPL3?
13:11:38__Bilgus_or doesn't want to
13:16:04gevaertsThe other aspect is do we want more #ifdefs in case someone who *really* wants a v2-only build appears?
13:18:23gevaertsAs long as it's just plugins, that's not too terrible of course (but then neither is manually removing that one directory or file from SOURCES or SUBDIRS. If it's more than that, yes, it needs to be done properly from the start, you don't want to tell people to comment out some lines in apps/somefile.c
13:20:05__builtinwe'd just have a couple #ifdefs in SUBDIRS
13:20:49__builtinand I think it's really unlikely that someone actually cares about v2-only
13:21:00gevaertsYes, I'm thinking adding this to configure probably takes less time than arguing about it :)
13:22:00__builtinI'll do it then
13:22:10__builtinand if people complain we'll hear about it
13:22:55gevaertsThe existence of the switch might also alleviate concerns from those who don't share my opinion that v3 doesn't actually change anything in practice
13:24:04gevaertsMaybe an email to -committers stating the intention to add v3 code would be useful? Carefully phrased so it's not an invitation to discuss things forever of course, you want to get on with it :)
13:24:13 Quit johnb2 (Ping timeout: 264 seconds)
13:26:17speachyWith my "I'm the one actually distributing this stuff" hat on, we'd definitely need to make sure all documentation properly states the GPLv3 nature of the binaries, with copies of the license, etc etc etc.
13:34:43gevaertsYes, agreed
13:41:47 Join lebellium [0] (
14:13:56 Quit LambdaCalculus37 (Quit: Leaving)
14:20:24 Join johnb2 [0] (
14:20:27 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
14:27:25***Saving seen data "./dancer.seen"
14:47:55 Quit craftyguy (Ping timeout: 246 seconds)
14:53:37 Quit johnb2 (Ping timeout: 264 seconds)
15:01:25 Join craftyguy [0] (~craftyguy@qutebrowser/dev/craftyguy)
15:03:31 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
15:03:36 Join craftyguy_ [0] (~craftyguy@qutebrowser/dev/craftyguy)
15:06:08 Quit craftyguy (Ping timeout: 256 seconds)
15:06:10speachyat 44.1KHz I have the X3 running _without_ PLL1
15:07:18 Join craftyguy [0] (~craftyguy@qutebrowser/dev/craftyguy)
15:08:19 Quit craftyguy_ (Ping timeout: 265 seconds)
15:14:17speachyhmm. wonder how hard it would be to get DSD working on this thing.
15:18:40speachyslkaving off the XTAL I can get it down to 0.04% freq error
15:19:02speachyat 44.1. at 48, the best I can do is 0.8%
15:23:53 Join craftyguy_ [0] (~craftyguy@qutebrowser/dev/craftyguy)
15:26:12 Quit craftyguy (Ping timeout: 256 seconds)
15:29:50 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
15:38:19 Join johnb2 [0] (
15:56:16fs-bluebotBuild Server message: New build round started. Revision b3a0187, 280 builds, 10 clients.
15:58:01 Quit johnb2 (Quit: Nettalk6 -
16:12:06fs-bluebotBuild Server message: Build round completed after 950 seconds.
16:12:09fs-bluebotBuild Server message: Revision b3a0187 result: All green
16:19:11fs-bluebotBuild Server message: New build round started. Revision adce547, 280 builds, 10 clients.
16:27:26***Saving seen data "./dancer.seen"
16:33:48fs-bluebotBuild Server message: Build round completed after 876 seconds.
16:33:49fs-bluebotBuild Server message: Revision adce547 result: All green
16:35:56Strife89The m200 OF is a fickle beast
16:36:14Strife89It keeps freezing while doing it's dreaded database refresh
16:37:22Strife89Can't really tell whether it's frozen or not, except to time it and guess at what's a reasonable point to decide it froze.
16:38:08 Nick craftyguy_ is now known as craftyguy (~craftyguy@qutebrowser/dev/craftyguy)
16:38:50 Join jdarnley [0] (
16:40:18 Quit J_Darnley (Ping timeout: 260 seconds)
16:41:02speachyrelatively high SD card loading
16:48:09Strife89Also, I think the Power button was pretty worn out when I got this thing, and I've made it worse from all the boot attempts
16:48:42Strife89I'm not finding it nearly impossible to boot the OF
16:49:37Strife89Plugging USB while holding Rew *should* work, but the OF is getting stuck on it's boot splash screen
16:51:41Strife89Net result: oops, I can't get files on or off the player without using the RB bootloader's (or Rockbox itself) USB mode
17:26:51 Quit lebellium (Quit: Leaving)
18:01:21 Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach)
18:04:11 Join J_Darnley [0] (
18:05:05 Quit jdarnley (Ping timeout: 240 seconds)
18:11:01 Quit Rower (Ping timeout: 264 seconds)
18:27:29***Saving seen data "./dancer.seen"
18:32:37 Quit sakax (Remote host closed the connection)
18:47:59 Quit __Bilgus_ (Ping timeout: 245 seconds)
19:08:37 Join __Bilgus_ [0] (41ba23be@
19:08:45 Join Jenetrix [0] (
19:08:48 Quit Jenetrix (Remote host closed the connection)
19:09:17 Join Jenetrix [0] (
19:09:27 Nick Jenetrix is now known as Guest5861 (
19:10:58 Quit Guest5861 (Client Quit)
19:11:30 Join Jen [0] (
19:14:07 Quit Jen (Client Quit)
19:20:20 Quit pamaury (Ping timeout: 256 seconds)
19:23:41 Join advcomp2019_ [0] (
19:23:41 Quit advcomp2019_ (Changing host)
19:23:41 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
19:24:04 Quit advcomp2019 (Ping timeout: 246 seconds)
19:51:43 Join massiveH [0] (
20:14:03Strife89__Bilgus_: My Sansa c250 also has a problem with battery bench on the current build, although it's a different problem.
20:14:50Strife89After I start a benchmark, it goes back to the plugins menu, and appears to freeze. I gave it five minutes, buttons still don't respond.
20:19:28Strife89I cleared my settings, deleted the bench text file, rebooted and tried again. Same problem. c200v1, revision eb0e41c1cc.
20:27:31***Saving seen data "./dancer.seen"
20:35:51Strife89Possibly having problems with it on my Clip+ now, but it could also have to do with having multiboot set up.
21:07:14 Quit MrZeus (Ping timeout: 260 seconds)
21:38:05__builtinladies and gents, we have video via SDL2.0! :)
22:19:12 Quit cockroach (Quit: leaving)
22:19:36__builtinthis is going much faster than 1.2
22:19:41__builtinprobably because it's basically the same code...
22:26:18 Quit marex-cloud (Read error: Connection reset by peer)
22:27:35***Saving seen data "./dancer.seen"
22:29:07 Quit ac_laptop (Ping timeout: 265 seconds)
22:33:45 Join marex-cloud [0] (sid137234@gateway/web/
22:49:52__Bilgus_Strife89 Clearly there is an issue let me see if I can repro
22:59:25 Quit dweeber_ (Ping timeout: 240 seconds)
23:07:36 Join ac_laptop [0] (~ac_laptop@
23:36:40__Bilgus_Strife by chance are you doing theser test with multiboot through a ROLO? rolo doesn't currently work right with multiboot because it bypasses the bootloader
23:43:21 Quit [7] (Ping timeout: 244 seconds)
23:43:45 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
23:48:02__builtinnote to self: the sim should fail gracefully if it can't open an audio device
23:48:36__builtinrather than crashing out in SDL_BuildAudioCVT as it does now
23:53:46__Bilgus_there are a lot of funny quirks like that in the sim it seems
23:54:11 Join massive_H [0] (
23:54:45__builtinfunnily enough, because the sim runs on SDL, I'm great at debugging it :)
23:55:03__builtinI'm literally running SDL in SDL!
23:56:39__Bilgus_the one wth the button buffer crashing when full is an interesting one but I've come to love it
23:56:54__builtinoh yeah, queue_post ovf?
23:57:05 Quit massiveH (Ping timeout: 240 seconds)
23:57:29__Bilgus_oh you never get to see it, its a handy way to stop hung plugins lol

Previous day | Next day