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 2021-03-10

00:46:06 Quit Barlow (Ping timeout: 256 seconds)
00:46:45 Join Barlow [0] (
00:46:45 Quit Barlow (Changing host)
00:46:45 Join Barlow [0] (~barlow@unaffiliated/barlow)
01:19:54***Saving seen data "./dancer.seen"
01:32:32 Quit ufdm (Remote host closed the connection)
02:19:43 Join Rower [0] (
02:50:51 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
03:19:56***Saving seen data "./dancer.seen"
03:44:30braewoodsda fuck, this must be a mistake
03:45:29braewoodsdd says it has 36 MB/s IO speeds on the gogears on their new CF cards
03:45:35braewoodsover USB 2.0
03:45:40braewoodsseems too high to be real
03:45:48braewoodsoh well, time to format 'em so i can repair them
04:36:46 Quit koniu (Ping timeout: 268 seconds)
04:44:09braewoodskeypads don't work on either gogear now
04:44:19braewoodsguess i need to recheck the ribbon cables
04:50:54 Join koniu [0] (~koniu@gateway/tor-sasl/koniu)
04:55:17 Quit pamaury (Ping timeout: 272 seconds)
04:55:32 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
05:14:05braewoodsi think it's ready for merge, it works that i can tell
05:14:09braewoodsi finally worked out the issues
05:14:29braewoodsprobably linux only compatibiility but that's ok
05:14:44braewoodswe can probably probe for it via userspace drivers on the other OS
05:14:55braewoodsif we need it
05:15:28braewoodsbut this makes it possible to query for it from our own software too
05:16:02braewoodsbluebrother: you have any interest in using this feature in rbutil?
05:16:10braewoodsno idea what we would use it for but
05:16:26braewoodsyou can use it to probe the battery level via USB HID protocol
05:16:44braewoodswe can also add our own additions to the feature report if we wanted to
05:16:54braewoodsfor now i just have padding to make Linux happy
05:20:00***Saving seen data "./dancer.seen"
05:21:05 Join blbro[m] [0] (blbrostrat@gateway/shell/
05:36:23 Quit kadoban (Quit: Bridge terminating on SIGTERM)
05:36:28 Quit danielp3344 (Quit: Bridge terminating on SIGTERM)
05:36:28 Quit ccf-100 (Quit: Bridge terminating on SIGTERM)
05:36:29 Quit blbro[m] (Quit: Bridge terminating on SIGTERM)
05:36:30 Quit kadobanana (Quit: Bridge terminating on SIGTERM)
05:36:30 Quit M-iam-some0nee[m (Quit: Bridge terminating on SIGTERM)
05:36:39 Quit mud (Quit: Bridge terminating on SIGTERM)
05:47:26 Join blbro[m] [0] (blbrostrat@gateway/shell/
06:06:38 Join kadobanana [0] (kadoban9he@gateway/shell/
06:06:38 Join danielp3344 [0] (danielp334@gateway/shell/
06:06:38 Join mud [0] (kadobanmat@gateway/shell/
06:06:38 Join ccf-100 [0] (ccf-100mat@gateway/shell/
06:06:38 Join kadoban [0] (kadobanemp@gateway/shell/
06:22:33 Quit Saijin_Naib (Remote host closed the connection)
06:22:41 Join Saijin_Naib [0] (
06:48:04fs-bluebotBuild Server message: New build round started. Revision 0ba174789b, 293 builds, 9 clients.
07:01:46fs-bluebotBuild Server message: Build round completed after 822 seconds.
07:01:49fs-bluebotBuild Server message: Revision 0ba174789b result: All green
07:06:52braewoodsspeachy: appears linux probes for it every 2 minutes or so
07:06:58braewoodsso even better
07:08:01 Join Soap_ [0] (~Soap@rockbox/staff/soap)
07:08:18 Quit Soap_ (Client Quit)
07:09:37braewoodsso even better
07:20:04***Saving seen data "./dancer.seen"
07:49:10 Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach)
08:42:07_bilgusJFIF windows 10 , wtf is JFIF even still doing around?
08:42:34_bilgus^this guys failing coverart is JFIF
08:54:42 Join MrZeus_ [0] (~MrZeus@
09:12:54 Join massiveH [0] (
09:20:06***No seen item changed, no save performed.
09:50:08_bilgusBah nope my mistake I guess that is a valid jpeg encoding I was thinking EXIF
10:45:23 Quit massiveH (Quit: Leaving)
11:03:24 Join jdarnley [0] (
11:04:03 Quit J_Darnley (Ping timeout: 260 seconds)
11:14:08 Quit tomato (Ping timeout: 272 seconds)
11:20:10***Saving seen data "./dancer.seen"
11:27:00 Quit ender| (Quit: Have you heard of the reverse excorcism? It's when the devil tells the priest to exit the child's body.)
11:42:26 Quit Saijin_Naib (Read error: Connection reset by peer)
11:53:16 Join ender| [0] (~ender1@2a01:260:4094:1:6045:6fff:fedd:cbf3)
12:00:09 Join Saijin_Naib [0] (
12:49:05 Join lebellium [0] (
13:16:20 Join lebellium_ [0] (
13:16:45 Quit lebellium (Ping timeout: 264 seconds)
13:20:12***Saving seen data "./dancer.seen"
13:26:36 Quit XDjackieXD (Quit: bye o/)
13:31:31 Join XDjackieXD [0] (
13:54:36 Quit Saijin_Naib (Read error: Connection reset by peer)
14:21:30 Quit jdarnley (Ping timeout: 265 seconds)
14:24:12 Join Saijin_Naib [0] (
14:25:17 Quit Saijin_Naib (Read error: Connection reset by peer)
14:27:10 Join J_Darnley [0] (
14:33:13 Join dweeber` [0] (
14:33:17 Join Soap_ [0] (~Soap@rockbox/staff/soap)
14:33:37 Join lebellium__ [0] (
14:33:45 Join tchan1 [0] (
14:34:09 Quit tchan1 (Client Quit)
14:34:30 Join tchan1 [0] (
14:34:38 Quit tchan1 (Client Quit)
14:34:59 Join tchan1 [0] (
14:36:04 Quit tchan1 (Client Quit)
14:36:26 Join tchan1 [0] (
14:38:09 Quit tchan (Disconnected by services)
14:38:18 Quit tchan1 (Client Quit)
14:38:27 Quit gevaerts (Disconnected by services)
14:38:30 Join gevaerts_ [0] (~fg@rockbox/developer/gevaerts)
14:38:35 Join tchan [0] (
14:38:35 Quit tchan (Changing host)
14:38:35 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
14:41:22 Join CR0W_ [0] (
14:42:42 Quit lebellium_ (*.net *.split)
14:42:43 Quit Soap (*.net *.split)
14:42:43 Quit dweeber_ (*.net *.split)
14:42:43 Quit CR0W (*.net *.split)
14:42:43 Quit KalBot (*.net *.split)
14:50:31 Join KalBot [0] (
14:57:58 Quit J_Darnley (Ping timeout: 260 seconds)
14:58:10 Join J_Darnley [0] (
15:02:37fs-bluebotBuild Server message: New build round started. Revision 73ec3971f7, 293 builds, 9 clients.
15:03:20speachy_bilgus: what do you think about g#3206 ?
15:03:23fs-bluebotGerrit review #3206 at : Move "View" option for playlists up one level by Christian Soffke
15:04:45edhelas__bilgus yes progressive JPG support is a big thing for me as well
15:05:13edhelas_the tool i'm using to save the cover arts in my albums folders is saving all the covers in progressive jpg
15:06:12edhelas_i tried to patch the internal jpg decoder but i don't have the skill yet to do so
15:06:33 Quit Rower (Ping timeout: 264 seconds)
15:11:27_bilgusspeachy I've often wondered why its in there I can't imagine it'll cause too much angst?
15:12:10_bilgusso I'm chasing this jpg display bug and it smells so much like stack overflow but damned if I can pinpoint it
15:12:32speachyyeah, it seemed sane to me but that's not necessarily a plus in its favor..
15:12:47speachy_bilgus: hmm, maybe an interrupt is coming along at the wrong time?
15:15:13fs-bluebotBuild Server message: Build round completed after 754 seconds.
15:15:15fs-bluebotBuild Server message: Revision 73ec3971f7 result: All green
15:15:16fs-bluebotBuild Server message: New build round started. Revision b3b8310e4e, 293 builds, 8 clients.
15:18:19 Join Saijin_Naib [0] (
15:20:14***Saving seen data "./dancer.seen"
15:23:25_bilgusedhelas_, the jpeg decoder is probably never gonna support much more than already its pretty heavy on the ram as is but perhaps it could be expanded for device with more RAM
15:24:19speachyedhelas_: and progressive decoding is a _lot_ more ram-heavy.
15:24:40edhelas_what about stopping on the first frames ?
15:25:13edhelas_regarding the resolution of most of the device screens, one or two frames should be enough, no need to decode the full 500x500px one
15:27:10fs-bluebotBuild Server message: Build round completed after 714 seconds.
15:27:12fs-bluebotBuild Server message: Revision b3b8310e4e result: All green
15:33:52 Quit berber (Remote host closed the connection)
15:36:36_bilgusthta sounds like a possibly viable strategy
16:03:50speachyhah, the Cayin N3 that someone asked about on the forum is a fairly straightforward X1000+hibylinux port, but the N3Pro is a whole other beast. About 4x the price, it comes with a *tube amp*
16:17:40 Join berber [0] (
16:23:01 Quit berber (Quit: The Lounge -
16:23:26 Join berber [0] (
16:32:41 Join lebellium [0] (
16:33:33 Quit lebellium__ (Ping timeout: 264 seconds)
16:41:07 Join ubervison [0] (~ubervison@2a02:aa12:b106:1b80:6257:18ff:fe9b:d6a1)
16:41:12 Quit ubervison (Client Quit)
16:43:02 Quit berber (Quit: The Lounge -
16:50:57speachysweet, figured out how to enable the headphone audio controls on the x3ii
16:55:58speachyI don't have
16:56:46 Join ac_laptop [0] (~ac_laptop@
17:03:15speachyI don't have a remote that'll do more than play/pause
17:10:22braewoodsspeachy: i'm going to try to make most out of MTP... turns out we can do a lot more than file transfer.
17:10:29braewoodsit can act as a remote control over USB too.
17:10:41braewoodswe can even extend it with our own extensions if we want to
17:12:36 Quit pamaury (Ping timeout: 246 seconds)
17:12:38braewoodsi think i'm going to try adding support for the host to control rockbox remotely
17:12:43braewoodswhile connected to USB
17:12:53braewoodsmaybe via a custom application
17:12:55braewoodsif need be
17:13:18braewoodsMTP is more powerful than i realized
17:14:15braewoodsin any case i suspect mtp is the best investment for new drivers
17:20:17***Saving seen data "./dancer.seen"
17:36:03 Quit advcomp2019_ (Ping timeout: 260 seconds)
17:36:56 Join advcomp2019_ [0] (
17:36:56 Quit advcomp2019_ (Changing host)
17:36:56 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
17:39:53 Join jdarnley [0] (
17:42:05 Quit J_Darnley (Ping timeout: 256 seconds)
18:00:48 Quit lebellium (Quit: Leaving)
18:46:54fs-bluebotBuild Server message: New build round started. Revision 555299af9f, 293 builds, 9 clients.
18:47:28_bilgusspeachy any insight on this If I set a static pointer = NULL and later check if (!ptr) it never hits, but If I instead use if ptr != NULL it works and !ptr works too
18:48:13speachydoes anything set that pointer?
18:48:28_bilgusme after that
18:48:34_bilgusbut prior no
18:48:35speachycould be a helpfully aggressive optimizer
18:48:47_bilgusits almost like its being optmized out
18:51:21speachyhmm, the 'rockbox info' view of disks doesn't actually work for hosted (or simulator) builds.
18:51:30speachyalways comes up as not present.
18:52:11_bilgusI can see why
18:53:10speachyit relies on the 'size' property of the volume which isn't ever set, instead of using the storage is_present() api
18:55:55speachyit also tries to look up "/mnt/sd_0/<microSD0>"
18:56:01speachywhich is.. bogus.
18:59:15fs-bluebotBuild Server message: Build round completed after 741 seconds.
18:59:22fs-bluebotBuild Server message: Revision 555299af9f result: 0 errors 2 warnings
19:11:19fs-bluebotBuild Server message: New build round started. Revision a87abdb28f, 293 builds, 9 clients.
19:20:21***Saving seen data "./dancer.seen"
19:23:24fs-bluebotBuild Server message: Build round completed after 725 seconds.
19:23:30fs-bluebotBuild Server message: Revision a87abdb28f result: All green
19:25:41 Join berber [0] (
19:26:43 Quit Saijin_Naib (Disconnected by services)
19:26:47 Join Saijin-Naib [0] (
19:40:53 Quit jschwart (Ping timeout: 260 seconds)
19:48:02 Quit Saijin-Naib (Ping timeout: 264 seconds)
19:59:51_bilgushmm so the way imageviewer does the decoder it pre-compiles the decoder as a plugin then strips the CRT
20:01:56_bilgusI suspect that would make it dependent on the underlying plugin buffer
20:02:24_bilgusI'll try setting the plugin buffer to 0 first
20:08:57 Quit MrZeus_ (Ping timeout: 264 seconds)
20:24:05speachythe imageviewer plugin has so many past sins embodied within
20:27:14speachymost of which were due to the realities of archos targets
20:30:09 Join Saijin_Naib [0] (
20:47:56braewoods_bilgus: one suggestion would be to go through and convert the jpegs to the final image you needed and cache the result
20:48:06braewoodsso you might be able to do that if you use the full RAM available
20:57:38 Quit Saijin_Naib (Ping timeout: 264 seconds)
20:58:55 Join jschwart [0] (~quassel@2001:985:2c6e:0:b00b:32ff:fe28:5567)
21:09:17 Nick hook54321_ is now known as hook54321 (sid149355@gateway/web/
21:20:23***Saving seen data "./dancer.seen"
21:31:57 Quit rogeliodh (Quit: Ping timeout (120 seconds))
21:33:26 Quit trfl (Remote host closed the connection)
21:34:02 Quit user890104 (Remote host closed the connection)
21:36:17 Join rogeliodh9 [0] (
21:36:18 Join trfl_ [0] (
21:36:18 Join user890104_ [0] (
21:36:23speachyFINALLY. three rabbit holes later I finally accomplished my original task
21:36:23 Quit michaelni (Ping timeout: 260 seconds)
21:36:30 Nick rogeliodh9 is now known as rogeliodh (
21:36:36speachy g#3210
21:36:38fs-bluebotGerrit review #3210 at : xduoo_x3ii: Improvements in the meymappings! by Solomon Peachy
21:36:53 Join michaelni [0] (
21:44:15speachyinline headphone remotes are working on the x3ii. don't know if the x20 supports it or not
21:45:22speachy_bilgus: in theory the Rocker supports it too
21:49:57fs-bluebotBuild Server message: New build round started. Revision 87e37a4d48, 293 builds, 9 clients.
21:54:27 Quit user890104_ (Ping timeout: 260 seconds)
21:54:36 Join user890104 [0] (
22:02:02fs-bluebotBuild Server message: Build round completed after 725 seconds.
22:02:05fs-bluebotBuild Server message: Revision 87e37a4d48 result: All green
22:02:32_bilgusspeachy people have been asking for that for a long while
22:03:56speachynone of the other hiby-based players seem to support inline remotes, except possibly the Rocker.
22:07:28 Quit cockroach (Quit: leaving)
22:09:00speachyhmm, can detect the poweron reason too (KEY or VBUS only)
22:14:13_bilgus g#3211
22:14:15fs-bluebotGerrit review #3211 at : ImageViewer Fix jpeg decoder by William Wilgus
22:18:40speachyactually that makes sense
22:19:01speachyno wait, that doesn't.
22:19:42speachybut the resulting code is cleaner anyway
22:20:28speachymost likely the crash was happening because the viewport we wer intiailzed with went away
22:20:49speachyand there's no reason for that to be static, so pfft.
22:27:10 Join f1reflyylmao [0] (
22:27:54_bilgusno its never null its whatever the buffer it copied to was before
22:28:31 Quit f1refly (Ping timeout: 272 seconds)
22:28:31 Nick f1reflyylmao is now known as f1refly (
22:28:57_bilgusI wanted it static so it could just get set once
22:39:31braewoodsok with battery strength put to bed
22:39:39braewoodsit's time for me to get serious about MTP
22:40:18braewoodsit might be preferable to our UMS driver for supported targets
22:42:00braewoodsplus given the design we could also add support for stuff like remote controls of basic rockbox functionality
22:43:57fs-bluebotBuild Server message: New build round started. Revision eb9f05f835, 293 builds, 9 clients.
22:54:46fs-bluebotBuild Server message: Build round completed after 649 seconds.
22:54:49fs-bluebotBuild Server message: Revision eb9f05f835 result: All green
23:20:25***Saving seen data "./dancer.seen"
23:31:48speachy_bilgus: any idea how long it's been "broken" ?
23:32:04speachycould it be fallout from the toolchain bump? (and/or the viewport rewrite you did..)
23:34:31speachyhahahah the only Rocker listed on ebay is in the UK, asking GBP 300.
23:39:58braewoodsspeachy: i think my iriver h10 has a regression
23:40:09braewoodsit works on 3.15 but not the latest development
23:40:31braewoodsit just locks up during boot
23:41:41speachystock hdd, or ssd mod?
23:42:10 Quit koniu (Remote host closed the connection)
23:42:13_bilgusIts for sure my doing with the vp rewrite
23:42:35 Join koniu [0] (~koniu@gateway/tor-sasl/koniu)
23:42:55speachybut I can't really comment onwhat it could be other than to try going back in 3 month increments.
23:43:22braewoodsspeachy: cf mod card mod
23:43:24speachy(I'm failry sure that I had a report of success in the post-toolchain-bump era
23:43:39_bilgusI cookie cutter copy pastad all those old frame buffer grabs not realizing that particular file doesn't init\ statics
23:43:52braewoodsthe same CF card works fine on 3.15
23:44:03speachydue to crt0 getting stripped, eh?
23:44:05braewoodsbut not latest builds
23:44:12_bilgusoh sorry missed that
23:44:16speachy3.15 is ~18mo old now
23:44:23speachyso... yeah.
23:45:58_bilgusI think you tsted thge h10 after the vp stuff braewoods?
23:46:37braewoodsvp? the new toolchain?
23:46:41braewoodsyea, i tested it previously
23:46:46braewoodssame result
23:46:48_bilgusviewport stuff
23:47:20braewoods_bilgus: erm yea, but what does that have to do with this? the h10 doesn't have a remote afaik
23:47:59braewoodssounds like i would have to trace the issue
23:48:29_bilgusyes git bisect will work here luckily
23:48:33speachythe new toolchains landed in July 2015. So I'd suggest doing a checkout from roughly August 2020 and see if that works..
23:48:54speachyjust to see if it's from the toolchain, or something else.
23:49:12speachy(wtf... bad fingers..)
23:49:14braewoodsi'll try it later
23:49:39braewoodsbut i figured if it was a general PP problem
23:49:43braewoodsmy hdd1630 would have issues too
23:49:45braewoodsit does not
23:50:03braewoodsso something different with the H10
23:51:56speachystart just after the dust settling from the toolchian bump (b4865b05b0a3cc45f2873599c6411bef92f40b1c) to make sure you have a known good point post-toolchain
23:52:43speachyand if that worked, then git-bisect will probably be the most expedient approach to take.
23:53:32braewoodsi remember how much of a pain it was to find out what broke h300 bootloader
23:53:41speachy(there have been no H10-specific changes (other than a bootloader fix) since 2015.
23:54:36speachyanyway, off to bed for me. good luck. :D
23:55:22speachy...the h10 does have model-specific asm code for the lcd functions. I suppose it's possible that has a bad interaction with the viewport rewrite.
23:55:34braewoodswho knows
23:55:37braewoodsi'll look into it
23:55:55braewoodsbut it is important we try to keep ports working as long as we consider them supported
23:56:24braewoodsin any case i intend to mainly develop our USB drivers
23:56:57braewoodsthough some ports have become so rare i wonder if we have cause to keep them around
23:57:13braewoodsH10 isn't one of those yet
23:57:22braewoodsi still see fixer uppers show up on ebay
23:58:21braewoods_bilgus: how practical would it be to scale background IO to account for current load?
23:58:30braewoodsthat will be a factor in MTP development
23:58:58braewoodsi don't know what rockbox provides for trying to limit IO rates so i don't preclude the regular operations from working

Previous day | Next day