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 2020-11-17

00:08:53***Saving seen data "./dancer.seen"
00:31:12 Quit ac_laptop (Ping timeout: 260 seconds)
01:10:49 Join Rower [0] (
01:13:55 Join Rower- [0] (
01:18:15 Quit Rower (Ping timeout: 272 seconds)
01:19:13 Quit CommunistWitchDr (Ping timeout: 246 seconds)
01:30:39 Join CommunistWitchDr [0] (
02:08:57***Saving seen data "./dancer.seen"
02:48:04 Join petur [0] (~petur@rockbox/developer/petur)
03:01:58 Quit Misanthropos (Ping timeout: 260 seconds)
03:20:44 Join Misanthropos [0] (~Misanthro@
04:09:00***Saving seen data "./dancer.seen"
04:26:58 Quit t0mato (Quit: Ping timeout (120 seconds))
04:40:38 Join t0mato [0] (t0mato@gateway/vpn/mullvad/t0mato)
04:43:12 Quit ArsenArsen (Changing host)
04:43:12 Join ArsenArsen [0] (~Arsen@fsf/member/ArsenArsen)
04:46:35 Quit funman (Changing host)
04:46:35 Join funman [0] (~fun@rockbox/developer/funman)
04:57:33braewoodsspeachy: what else needs to happen for these bootloaders? i think i've done it all except have a new rbutil release
05:15:24 Quit S|h|a|w|n (Read error: Connection reset by peer)
05:35:29 Join prof_wolfff [0] (
05:38:38braewoodsnow i've seen everything. an 18650 cell with its own usb charging port.
05:48:03 Join edhelas [0] (9d94237298@2a01:7c8:aab8:6b9:5054:ff:fec9:fd84)
06:00:03 Quit Moarc (Read error: Connection reset by peer)
06:00:06 Join Moarc_ [0] (
06:05:25 Quit Misanthropos (Ping timeout: 240 seconds)
06:09:03***Saving seen data "./dancer.seen"
06:20:20 Join Misanthropos [0] (~Misanthro@
06:37:00edhelassince the last updated (and USB changes) I cannot connect anymore Rockbox using USB to my computer
06:37:10edhelasit seems that i'm not the only one in that case,53652.0.html
06:55:16speachywhat's the player?
06:58:59fs-bluebotBuild Server message: New build round started. Revision 0400a39, 293 builds, 9 clients.
06:59:14speachythis next build should make that better.
07:04:17 Quit Stanley00 (Remote host closed the connection)
07:14:04 Quit Moarc_ (Quit: i znowu NADMUCHAƁ BALONA)
07:14:17 Join Moarc [0] (
07:14:52fs-bluebotBuild Server message: Build round completed after 953 seconds.
07:14:58fs-bluebotBuild Server message: Revision 0400a39 result: All green
07:26:26edhelasfor my case the iPod last gen
08:01:10 Quit braewoods (Quit: WeeChat 2.8)
08:07:00 Join pacman [0] (~braewoods@learnprogramming/staff/braewoods)
08:07:08 Nick pacman is now known as braewoods (~braewoods@learnprogramming/staff/braewoods)
08:09:04***Saving seen data "./dancer.seen"
08:29:45mendel_munkisI tested g#3035 for the fuze+ manual. is there any other manual I should confirm it doesn't break?
08:29:49fs-bluebotGerrit review #3035 at : Manual: remove most HWCODEC artifacts by Moshe Piekarski
09:25:55speachymendel_munkis: when I did the initial excising of hwcodec+charcell stuff, I intentionally left lcd_bitmap behind (and always enabled) in the manual
09:26:08speachyjust because there was of it.
09:26:22speachylcd_bitmap and swcodec I mean
09:26:49speachyso what are the things you haven't been able to track down?
09:33:15 Join massiveH [0] (
09:33:56speachylcd_bitmap and swcodec are also heavily referenced in the language files
09:38:24mendel_munkisthe things I haven't tracked down are A chapter 12 text_viewer encodings some of the line breaks have been moved slightly B chapter 9 show icons diffpdf claims sonething changed but I cant see it
09:40:52mendel_munkisalso I left 5 instances of swcodec which I wasn;t sure xactly what they where doing.
09:47:25braewoodsmade a notice about it
09:48:29braewoodsspeachy: i think we need to wait for a new release of rbutil before we can release them to the website.
09:48:37braewoodswhy? the checksums are embedded in rbutil.
09:49:01braewoodsunless you can find evidence they're on the website somewhere.
10:03:38speachyok. g#3036 cleans up the language files.
10:03:40fs-bluebotGerrit review #3036 at : lang: Make all swcodec &| lcd_bitmap strings default. by Solomon Peachy
10:04:05speachy~ +21K,-36K :)
10:09:08***No seen item changed, no save performed.
10:16:16mendel_munkishow much of that is whitespace fixes? :)
10:25:44speachyvery little; pretty sure all of the language files have already been given the once-over by the updatelang tool.
10:27:34speachyI stand corrected; one of the language files has whitespace fixes.
10:46:03 Quit massiveH (Quit: Leaving)
10:48:23speachy g#3037 is next, with a net -5.2K lines
10:48:24fs-bluebotGerrit review #3037 at : lang: Drop all deprecated strings, and rename LANG__NEVER to LANG_NEVER by Solomon Peachy
10:50:17speachyjust some spring cleaning
10:59:17mendel_munkisyou're making my 400 line effort look bad
11:00:32mendel_munkisI was thinking about the xrick assets, maybe we should do what we have done for xworld?
11:01:02mendel_munkis(or ask square-enix to let us host them)
11:04:25speachythat was nearly entirely tool-generated. just took a few tweaks to the tool that generates it.
11:04:30speachy(gotta love write amplification!)
11:05:12 Join Stanley00 [0] (~stanley00@unaffiliated/stanley00)
11:07:12fs-bluebotBuild Server message: New build round started. Revision 20774d8, 293 builds, 9 clients.
11:07:19speachylet's see what breaks!
11:08:52speachyI completely rewrote update_lang a few months ago, primarily to fix voicefile generation but it also made this sort of thing a lot easier.
11:09:00*braewoods watches a dog walk up to speachy and say, "You broke my paw."
11:09:16 Quit Stanley00 (Ping timeout: 240 seconds)
11:09:33speachywouldn't that dog be limping up instead?
11:09:42mendel_munkisit was a front paw
11:10:07mendel_munkisare there any particularly likely to break manuals I should test?
11:11:59braewoodsmendel_munkis: you can test them all... manuallly ;p
11:12:58mendel_munkisbraewoods: no thanks
11:28:18fs-bluebotBuild Server message: Build round completed after 1266 seconds.
11:28:25fs-bluebotBuild Server message: Revision 20774d8 result: All green
11:48:50speachysurprising amoutn of size variations, huh.
11:50:57_bilgus_did you add lang entries to the aigoerosq
11:51:40_bilgus_the name looks like a foreign lang btw
11:52:21speachyno, but those targets do see large-ish size swings from builder to builder.
11:54:30speachymendel_munkis: IIRC older texlive distributions barfed with the \ipdfoutput -> \Ipdfoutput change.
11:55:14speachyif we need to enforce a specific texlive version to keep the builds from barfing, that can be done.
12:02:34speachy_bilgus_: looks like a proper fix for the usb prompt UI is called for; I enlarged the stack size as a hopeful hack.
12:03:35_bilgus_can you elaborate on that a bit? you mean the out of band call to the prompt?
12:03:57speachyyeah, reports of stack overflows on clipzip and ipodclassic
12:05:17speachyI _think_ the reports were prior to my most recent bugfix to fix the defaults to be consistent (ie not prompt by default)
12:05:46_bilgus_mmm yeah saving all that context is exceeding their limits
12:06:30_bilgus_ok I'll take a look and decide if its getting its own or if it will become a mechanism
12:06:59_bilgus_depends on what else might be able to share
12:07:04mendel_munkisyes. but the hwcodec stuff shouldn't require that
12:07:57mendel_munkisunless I misunderstood some stuff about gerrit
12:09:09***Saving seen data "./dancer.seen"
12:13:39mendel_munkisHow can I etermine minimum blugin buffer size for a plugin?
12:20:08 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
12:28:57braewoodsspeachy: thanks for your efforts on that front. it'll be useful when i work on MTP.
12:30:03braewoodssince i envision MTP being one of the options
12:30:12braewoodsfor some ports anyway
12:30:46braewoodsnot sure how practical MTP will end up being but i think it will be worth exploring at least
12:35:14 Quit petur (Quit: Connection reset by beer)
12:36:24braewoodswow. seems some of the older users have been wanting this for a long time.
12:36:41braewoodswhoever LambdaCalculus is?
12:37:16_bilgus_mendel_munkis, the minimum size it needs or will recieve?
12:37:40_bilgus_watermark it
12:38:58_bilgus_fill the buffer with 0xdeadbeef and look at whats left
12:39:22_bilgus_I just did this a few weeks ago in lua if you'd like an almost applicable example
12:39:28 Join Guest17 [0] (
12:39:56 Quit Guest17 (Remote host closed the connection)
12:40:01_bilgus_ g#2806
12:40:03fs-bluebotGerrit review #2806 at : lua add better memory stats by William Wilgus
12:41:45mendel_munkiswhy did you chose 0xBA5EFAC7?
12:44:04speachybraewoods: pamaury's old mtp patch had a full menu, instead of my hacked yesno thing, but yeah, once there's more options then a proper menu will be needed.
12:44:26speachy(maybe one day expanding to USB DAC support too..)
12:46:41_bilgus_it just seemed unlikely to be used
12:47:23_bilgus_cafebabe deadbeef and a few others are already used in the code so it makes it easy to know I'm looking at something lua touched
12:47:41mendel_munkismakes sense
12:48:43_bilgus_I also got a initial count of zero the first time I ran it through (before I started filling the buffer with the byte string)
12:49:15_bilgus_you know just to be sure it wasn't used in the code extensively (unlikely)
12:49:37_bilgus_not that it can't happen but it shouldnt be much error if so
12:53:50braewoodsspeachy: you mean exposing a usb audio device when connected to a host PC or so?
12:53:57speachybraewoods: yeah
12:54:11braewoodsfor what purpose? streaming to the PC or from it?
12:54:35speachysince this is all fantasizing, either/both!
12:54:44braewoodsi see.
12:54:59braewoodswell it might be possible in software
12:55:02speachyuse of standard external USB audio class devices, and also acting as one
12:55:07braewoodsif USB 1.1 is fast enough for audio?
12:55:14speachyUSB1.1 is plenty fast
12:55:37speachyat least for standard 16-bit 48K audio
12:56:11braewoodsi'll stick to MTP for now. i think audio processing would require special experience I do not have.
12:56:55braewoodsi'll see what i can do there. it'll be awhile before i have anything to show off.
12:57:15braewoodsi need to do a lot of research on USB and how rockbox does it and all
12:57:36 Join lebellium [0] (
12:57:38gevaertsUSB DAC would be a lot of fun to have. I suspect not all devices might be able to do it, but still
12:57:42braewoodsthough if you only have USB 1.1 i don't think MTP will be of much interest
12:57:57braewoodseven if you could do i
12:58:27*gevaerts points out that all USB 1.1 speeds also exist in USB 2.0 :)
12:59:05gevaertsAlso, I think all devices rockbox runs on that can do USB at all have high speed
12:59:06braewoodsjust fair question to ask if USB X feature would be useful even if we only have 1.1 on a given target
12:59:40braewoodsMTP could be useful for safety reasons
12:59:56speachyMTP is useful, regardless of actual transfer speed.
13:00:03gevaertsAudio is fine on full speed I'd say. Fancy DACs want more because they want to do 192kHz at 24 bit on seventeen channels, but that won't happen here ever
13:00:35braewoodsyea. i'll see what i can do to make it happen.
13:00:53braewoodsi think RB is only going to need to support up to 2 storage contexts though
13:01:13braewoodsi've never seen a RB target with more than 2 at least in terms of physical storage
13:01:24braewoodsmy clip+ exposed the internal and external separately
13:01:42gevaertsThe one thing you *might* have problems with for MTP (not sure, there have been a lot of fixes since I looked) is stability. Back in my day (I feel old by saying that!) I tried to get some sort of speed graph thingy on the USB screen working, and the extra load (or bugs, or bad scheduling...) killed reliability
13:01:50gevaertsAnd MTP kind of implies you do that all the time
13:02:14speachyAlso MTP means dealing with client-side "quirks".
13:02:27braewoodsindeed which is why it will be experimental for awhile
13:02:39gevaertsThat was on PP502x targets and gigabeat S, so what in rockbox is the arc driver. No idea what other controllers do
13:02:48braewoodsUMS is going to remain the default
13:02:54braewoodsit's stable and mature
13:03:01braewoodsMTP will be an experimental option
13:03:18gevaertsIt bricked my sansa c250 for a few days in the early days. Stupid typo :)
13:03:44speachyone thing that does need to be considered is that right now, USB takes over the whole system. There's no "background" operations going on that you have to fight with. With MTP, that changes.
13:03:52gevaertsYes, exactly
13:03:57speachyMTP during playback?
13:04:09braewoodswe'll see.
13:04:22gevaertsSame with usb audio, really
13:04:24braewoodsi'm not sure what it realistic
13:04:29speachyI mean, the various OFs that I recall supporting MTP didn't allow the player to be used simultaneously either
13:04:44braewoodsthey probably couldn't handle it well
13:04:47braewoodslimited resources and all
13:04:53speachyit's a lot of resources, exactly.
13:05:02braewoodsphones can do both easily
13:05:09speachyfirst step will be simply s/UMS/MTP/ :)
13:05:30braewoodswe need to implement a functioning MTP and then worry about optimizing it later
13:05:30gevaertsI'd say it should be tried, if only to stress-test to find bugs.
13:05:34speachybraewoods: even the crappiest smarphones had an order of magnitude more resources than the best rockbox target
13:05:56braewoodsi plan to make it exclusive at first
13:05:58speachyplus they don't expose the whole filesystem
13:06:08braewoodsbut in time maybe that can change
13:06:20speachyhmm, that does give me an idea though
13:06:36gevaertsIdeally MTP during playback only leads to choppy sound and nothing worse. If you get that far, the thing is *stable*
13:07:02braewoodson the PP units though it has dual cores so
13:07:04speachygevaerts: and for giggles, delete the file that's being played back.
13:07:05_bilgus_keep in mind there are way to release bufferlib allocs on demand
13:07:13braewoodsmaybe you can delegate the MTP to the main threa
13:07:15gevaertsOh yes :)
13:07:17braewoodsand leave the playing in the background
13:07:22mendel_munkisI personally hate mtp but if you get it working nice.
13:07:56mendel_munkis_bilgus_: given that i plan on adding a .5k margin of error regardless I should be fine.
13:17:38mendel_munkiswell used=unused seems like a bad start.
13:25:35 Quit amiconn (Quit: - Chat comfortably. Anywhere.)
13:25:35 Quit pixelma (Quit: .)
13:28:11 Join pixelma [0] (marianne@rockbox/staff/pixelma)
13:28:12 Join amiconn [0] (jens@rockbox/developer/amiconn)
13:33:22speachyhere's a fun one. fs#13257
13:33:23fs-bluebot USB Mode setting is ignored and always asks till WPS is opened (bugs, waiting on customer)
13:33:49speachythe main bug is fixed, but the second comment is another issue
13:34:18speachyin short, when we're booted up plugged into USB, the USB insertion event fires and is processed _before_ we load our settings
13:36:37speachyindeed, it intentionally happens before we even mount our storage
13:36:59mendel_munkiswell apparently it runs entirely in the audio buffer instead of the plugin buffer.
13:40:11speachyI understand why it's done that way; we want to provide a way to export the disk in UMS mode rather than possibly exploding and preventing the system from proceeding into the mount)
13:40:24speachys/mount/export via UMS)
13:58:06edhelasSo you'll use a specific directory to store the musics ?
14:09:12***Saving seen data "./dancer.seen"
14:16:19braewoodsspeachy: how important is supporting older windows from an MTP perspective?
14:16:39braewoodslike pre-windows 7 or 10 even
14:19:41speachyI'd say Win7 + 10 matter. but actually _debugging_ operation is the problem, since there's no source code to reference.
14:21:02braewoodsperhaps there is
14:21:31braewoodsi've seen plenty of MTP responder implementations in C++. it may be possible to port their quirks, if any.
14:22:12braewoodsbut first we need something that works well enough to worry about that
14:23:27braewoodsspeachy: if i were going to allocate IO buffers in USB... what are my options? using the audio buffer interferes with playback...
14:25:25speachyhonestly.. I'd say just statically allocate a chunk.
14:25:42speachyget it working with that, and then we can worry about making it dynamic.
14:26:01braewoodsi would think a multiple of 4K would be good
14:26:11braewoodsbe sure i can align writes to disk sectors or so
14:26:54braewoodsok. i'm going to read the MTP spec.
14:27:00braewoodswow 282 pages
14:29:15braewoodsspeachy: do we care about PTP? i haven't heard of anyone using that in ages.
14:29:37braewoodsplus we're not really a camera lol
14:29:47braewoodsat best we can display images someone put on the drive
14:30:44speachyno point to PTP.
14:30:50braewoodsyea thought so
14:31:00speachyI mean, if you want to implement it anyway, feel free! :D
14:31:30braewoodsnot interested honestly. i can't see anyone really using it.
14:33:56braewoodsi'll probably start my tests using a read-only fake data set
14:34:00braewoodsonce i figure out how this works
14:35:35 Quit beencubed (Quit: Leaving)
14:36:14braewoodshm. apparently MTP uses UTF16 for strings.
14:36:20braewoodsdoes rockbox already support this?
14:37:25 Quit Rower- (Ping timeout: 240 seconds)
14:37:57braewoodsjoy. i'll look at what i can do about that later.
14:38:16 Join beencubed [0] (~beencubed@
14:41:12 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
14:47:51mendel_munkisdo we have a compendium of keymap tables?
14:56:30braewoodshuh. MTP is a lot like basic HTTP
14:56:42braewoodsunidirectional, request/response setup
15:30:04 Quit Oksana (Ping timeout: 246 seconds)
15:54:07 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
16:01:52 Join _bilgus__ [0] (~bilgus@
16:03:02 Quit _bilgus_ (Ping timeout: 264 seconds)
16:09:16***Saving seen data "./dancer.seen"
16:11:33 Quit t0mato (Quit: Ping timeout (120 seconds))
16:13:27 Join ac_laptop [0] (~ac_laptop@
16:21:18 Join t0mato [0] (t0mato@gateway/vpn/mullvad/t0mato)
16:38:28 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
16:53:49 Quit lebellium (Quit: Leaving)
17:15:25 Quit livvy (Remote host closed the connection)
17:17:57 Quit kugel (Ping timeout: 265 seconds)
17:18:04 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
17:22:46 Join petur [0] (~petur@rockbox/developer/petur)
17:39:13 Quit ac_laptop (Ping timeout: 264 seconds)
18:09:17***Saving seen data "./dancer.seen"
19:25:37 Quit petur (Remote host closed the connection)
19:37:43 Quit Acou_Bass (Quit: ZNC 1.8.1 -
19:40:09 Join Acou_Bass [0] (
19:41:59 Quit prof_wolfff (Ping timeout: 265 seconds)
19:49:47 Join Misanthr- [0] (~Misanthro@
19:51:25 Quit Misanthropos (Ping timeout: 240 seconds)
19:51:25 Nick Misanthr- is now known as Misanthropos (~Misanthro@
19:51:47 Join prof_wolfff [0] (
19:54:11 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
19:54:56 Join fs-bluebot_ [0] (
19:57:07 Quit bluebrother^ (Ping timeout: 246 seconds)
19:57:08 Quit fs-bluebot (Ping timeout: 256 seconds)
20:09:21***Saving seen data "./dancer.seen"
20:21:12 Quit amdj (Remote host closed the connection)
20:21:25 Join amdj [0] (~aaron@freenode/staff/atheme.amdj)
20:36:56 Quit prof_wolfff (Ping timeout: 240 seconds)
21:09:23 Quit livvy (Ping timeout: 240 seconds)
21:12:46 Quit edhelas (Remote host closed the connection)
21:14:17 Join Misanthr- [0] (~Misanthro@
21:15:05 Quit Misanthropos (Ping timeout: 240 seconds)
21:15:05 Nick Misanthr- is now known as Misanthropos (~Misanthro@
21:18:15 Join ac_laptop [0] (~ac_laptop@
21:54:20 Join Stanley00 [0] (~stanley00@unaffiliated/stanley00)
22:00:59 Join karinka [0] (
22:01:08karinkaHey guys
22:01:27karinkaJust got a HiFi Walker H2, & have been running the dev build
22:01:48karinkaI've noticed a few problems with it
22:02:41karinkaJust wanting to know the best way I can help, I have some coding experience, but none with embedded devices
22:09:22***Saving seen data "./dancer.seen"
22:11:37 Quit karinka (Quit: karinka)
22:11:52 Join karinka [0] (
22:13:26 Quit karinka (Client Quit)
23:12:20 Quit ender| (Ping timeout: 268 seconds)
23:36:13 Quit ac_laptop (Ping timeout: 264 seconds)

Previous day | Next day