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 2020-10-21

00:41:19***Saving seen data "./dancer.seen"
01:14:45 Join advcomp2019__ [0] (~advcomp20@unaffiliated/advcomp2019)
01:17:25 Quit advcomp2019_ (Ping timeout: 240 seconds)
01:51:46 Join Rower [0] (
01:53:44 Quit efqw (Changing host)
01:53:44 Join efqw [0] (uid412670@unaffiliated/efqw)
01:53:44 Quit efqw (Changing host)
01:53:44 Join efqw [0] (uid412670@gateway/web/
01:55:12 Part efqw
01:55:13 Join efqw [0] (uid412670@gateway/web/
02:02:38 Quit Rower (Ping timeout: 260 seconds)
02:03:03 Join Rower [0] (
02:24:34 Join petur [0] (
02:24:34 Quit petur (Changing host)
02:24:34 Join petur [0] (~petur@rockbox/developer/petur)
02:41:21***Saving seen data "./dancer.seen"
02:52:44 Quit petur (Quit: Connection reset by beer)
02:57:44 Join bzed_ [0] (
02:59:13 Quit bzed (Ping timeout: 260 seconds)
02:59:18 Nick bzed_ is now known as bzed (
04:03:20 Quit APLU (Ping timeout: 246 seconds)
04:04:19 Join APLU [0] (~mulx@2a03:7220:8081:2900::1)
04:41:25***Saving seen data "./dancer.seen"
05:08:44 Quit Oksana (Ping timeout: 256 seconds)
06:41:28***Saving seen data "./dancer.seen"
07:39:17 Join wodz [0] (
07:42:52wodzspeachy: I am thinking about resume working on bt support. Are any of new targets you added have BT hardware?
07:43:15speachywodz: all of them. :)
07:43:35speachy(everything but the m3k, anyway. but that's a pathological basket case atm)
07:44:13wodzspeachy: are they still using bluez4 like agptek or upgraded to bluez5?
07:44:30speachypresumably bluez4.
07:44:42speachygimme a moment, I'll check..
07:45:40speachybluez 4.101
07:45:53wodzanyway bluetooth programing is f*** irritating. Dbus is devils invention. And resistance to provide sane bluetooth C api (with docs) blows my mind
07:47:16wodzWe have a few options. There is custom system_server which in theory could be used. But it relies on writing json files in system partition.
07:47:26speachyI had started down that path too.. was actually hoping I could get away with disabling bluez completely and using an embedded stack directly within rockbox, but alas there's nothing suitable that speaks BT Classic.
07:47:53speachyall the system_server seems to do is route commands to the cmdline bluez tools
07:48:20speachyI'd like to keep any BT code we write to be sorta portable..
07:48:22wodzWe could use libbluetooth but it is deprecated and lacks any documentation. There is only one example online showing how to perform discovery
07:49:33wodzAnd finally we could use blessed dbus api. BUT it is pretty complex, and we need to speek bluez4 variant which is not documented very well.
07:50:33speachy...or just invoke the bluez cmdline tools directly?
07:51:03wodzthats the option too but I really don't like it
07:51:04speachy(though I don't think that will let us take advantage of AVRCP, for example)
07:51:37speachywell, we could also scrap bluez4 and build/install our own bluez5 stuff.
07:51:53wodzpretty complicated
07:52:20speachyhacking bluez5 into something directly linkable is our only "viable" option for bt on native targets anyway.
07:52:26wodzgetting glib and dbus crosscompiled for rocker took me whole day
07:52:49wodzwhich native targets?
07:53:58speachywell, none at the moment, but I really want to do a native X1000 port due to the general level of crappiness that is the Ingenic Linux platform.
07:54:26speachy(And there are discussions ongoing to get a new, made-for-rockbox player built)
07:54:58wodzwell having native port is fun but very time consuming
07:55:01speachy(with schematics and documentation for the port to not be a 95% reverse-engineering exercise)
07:55:16wodzhave fun
07:55:48speachyIt's one of the many fireballs I'm juggling
07:56:51wodzIsn't it better to improve X1000 linux port instead?
07:58:09speachyI've done a ton of work recently along those lines, yes, but without the kernel sources there's not a whole lot we can ultimately do to make the platform suck less.
07:59:09speachyEverything else can just be scrapped in favor of mainline/upstream
07:59:38wodzI saw you are fiddling with alsa driver for hosted target. Please make alsa dev handler settable. So once bt will land we could call function to switch alsa to different handler
08:00:00wodzcurrently it is hardcoded
08:00:12speachyit's worse than that −− we have to make the hwcaps settable too. :)
08:00:47wodzthats part of the problem, yes
08:01:54speachyit also looks like these bt-capable players are using userspace bits from CSR's SDK.
08:08:36wodzThis is the most complete example of using libbluetooth
08:18:55 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
08:20:14 Quit Stanley00 ()
08:26:14pacmanthere's also the alsa plugin for bluetooth output
08:26:19pacmanto consider as welll
08:26:22 Nick pacman is now known as braewoods (~braewoods@learnprogramming/staff/braewoods)
08:27:49speachybraewoods: it's either the bluez 4 bundled plugin or it's from CSR's sdk.
08:29:05speachy(the one on these ingenic targets, I mean)
08:39:56wodzok, crosscompiled glib, dbus and bluez4 so once I get my rocker back from my son I'll play with this
08:41:02braewoodsman i must have my head in networking
08:41:04speachybtw I recommend re-installing the "Bootloader" with what's sitting on d.r.o now
08:41:05braewoodsi thought you said router
08:41:29speachyit has some nice quality-of-life improvements
08:41:32***Saving seen data "./dancer.seen"
08:41:38speachy(though still that long countdown)
09:02:07 Join massiveH [0] (
09:24:05fs-bluebot_Build Server message: New build round started. Revision 6b13437, 293 builds, 9 clients.
09:26:54speachywodz: did a little bit of alsa driver rearrangement; the driver should allow the output dev to be changed at runtime now. exposed caps are another matter.
09:27:56 Quit pamaury (Ping timeout: 256 seconds)
09:39:30fs-bluebot_Build Server message: Build round completed after 926 seconds.
09:39:33fs-bluebot_Build Server message: Revision 6b13437 result: All green
09:39:37fs-bluebot_Build Server message: New build round started. Revision 884d892, 293 builds, 8 clients.
09:55:21fs-bluebot_Build Server message: Build round completed after 944 seconds.
09:55:26fs-bluebot_Build Server message: Revision 884d892 result: All green
10:02:06 Quit beencubed (Quit: Leaving)
10:16:56 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:19:59braewoodsspeachy: i love these parts or repair units. they're good fodder for bootloader testing if i can get them repaired enough for that at least
10:22:33speachybraewoods: that they are.
10:22:37 Quit massiveH (Quit: Leaving)
10:29:45 Quit speachy (Quit: WeeChat 2.8)
10:32:10 Join speachy [0] (~speachy@
10:41:33***Saving seen data "./dancer.seen"
11:48:39 Quit pamaury (Ping timeout: 260 seconds)
12:22:09 Join beencubed [0] (~beencubed@
12:35:50 Join MrZeus_ [0] (~MrZeus@
12:41:36***Saving seen data "./dancer.seen"
13:42:12 Join johnb7 [0] (
13:42:31 Join lebellium [0] (
14:02:32 Quit johnb7 (Ping timeout: 256 seconds)
14:02:45_bilgus__Oh I hit most of the rocker bugs today, plug on startup device boots but no USB unplug Panic Mount 0, sluggish UI on boot clears with menu access
14:10:57_bilgus__speachy I think I have the update on WPS fixed with g#2811
14:10:59fs-bluebot_Gerrit review #2811 at : LCD core move buf ptr and address look up function viewport struct by William Wilgus
14:12:29_bilgus__what I did is add a Viewport_ex function that allows passing flags and marking viewports that get switched out as dirty then the WPS is able to check for the flag and update accordingly and clear the DIRTY flag when it is finished with the vp
14:13:12_bilgus__lebellium, might fix your issue too but haven't specifically looked yet
14:13:25speachynow was that panic with the latest bootloader
14:13:48speachy(ie edbdefa09a-201018)
14:13:55_bilgus__it looks like the functionality was planned for the WPS originally as there is a unused viewport_switched flag
14:14:29_bilgus__nope 201016
14:14:37efqwI'll run a battery bench on the m3k in a bit once I have it fully charged.
14:15:24efqwAll I need to do is to set it to loop through something continuously and then run the battery bench plugin, right?
14:15:26speachy(there were two nasty bugs in the '16 version)
14:15:31speachyefqw: yep
14:15:33_bilgus__I'll try again with the newest BL tonight but it was very random anyway
14:16:23efqwcool, I'll plug in some low impedance earbuds too
14:16:29speachythe sluggish menu has to do with something going on in the background. We're actually only single-threaded.
14:16:44speachyeffectively using sw threads instead of native posix threading.
14:17:09speachy(why? who knows. we use proper SDL threads in the UISim/SDLApp)
14:17:53_bilgus__IDK i kinda like our "threading' system
14:18:32_bilgus__braewoods, did you get a chance to see if your device was displaying properly?
14:20:13_bilgus__speachy I don't think FS 12927 has anything to do with the display corruption on nonstd strides I'm guessing alignment issue or a macro SOMEWHERE!
14:20:50_bilgus__I intend to get this done soon I'm already 3 days past when I wanted to have it done lol
14:22:13 Join johnb7 [0] (
14:29:29speachy_bilgus__: so is this 2bit corruption technically a regression, or has it, like true friendship, been there in front of us all along?
14:32:34speachyand bad news, doesn't seem to have fixed the problem
14:35:53 Quit johnb7 (Ping timeout: 258 seconds)
14:37:36speachy_bilgus__: oh, it's clearly visible in the UISim, provided you have a track with embedded art.
14:37:51speachy(on the x3ii or erosq targets)
14:40:04speachywant a screenshot?
14:40:49_bilgus__well the lcd drivers were always set up with device stride in mind its the fastest way after all
14:41:37***Saving seen data "./dancer.seen"
14:42:34_bilgus__ah I take it you only see it when there is album art though correct?
14:44:37speachyyes, because otherwise one of the text lines repaints directly over the (former) splash
14:45:45_bilgus__tahts a separate vp though so its probably repainted separately I just need to make it aware of the dirty/clean flag if so
14:46:34speachythere's basically nothing but the background under those particular pixels.
14:47:30_bilgus__the other option is to have splash keep track of what it overwrites but I'd much prefer to make this agnostic
14:51:06 Quit bluebrother^ (Read error: No route to host)
14:51:07 Quit fs-bluebot_ (Read error: Connection reset by peer)
14:51:24 Join fs-bluebot [0] (
14:51:26 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
14:58:59 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
15:11:31efqwok, battery bench started, I enabled "repeat all", this should be sufficient, right?
15:11:48efqwvol is set at -50db
15:15:31_bilgus__I don't remember what we set the levels at I wanna say half scale sounds right
15:17:19efqwfor this target I really don't think the LCD is going to be a significant battery hog, its battery is 1000mAh+
15:18:16efqwI did mess with the menus for a bit after starting the bench but it shouldn't affect runtime too much.
15:21:14speachywhat matters more is the curve we get for calibrating the fuel gauge.
15:21:53speachythe actual runtime will allow us to roughly remaining runtime
15:22:23efqwTheoretically we should run the battery down, clear it at the very last moment before automatic shutdown, and then charge the device back up.
15:23:09efqwBut I'm not entirely sure if the hacky axp173 driver is up for that task
15:25:29efqwI really want to nuke absolutely everything on this device and start from scratch, but this isn't a feasible option for us right now. I'm afraid fiio is going to continue the m3k and effectively kill the cheapest x1000e target :/
15:26:35speachythat's the risk for all of these things
15:33:09efqwI tried to tail the txt file but it seems like the plugin crashed, lol
15:33:23efqwI'll charge it back up and try again
15:34:04fs-bluebotBuild Server message: New build round started. Revision ce3da62, 293 builds, 9 clients.
15:34:31efqwstatus: charging, voltage: 4221
15:34:39efqwthis is odd
15:35:05efqw(I'm not 100% sure if the status thing ever changes btw)
15:40:10speachyhuh, the fiio x3ii uses the same soc and audio path as the original xduoo x3. and most of the fiio x1/x3/x5 variants share the same soc and dac.
15:46:46 Join johnb7 [0] (
15:49:24fs-bluebotBuild Server message: Build round completed after 920 seconds.
15:49:27fs-bluebotBuild Server message: Revision ce3da62 result: All green
15:51:51_bilgus__weird I made it simpler and just check the current viewport to see if its dirty and yes it knows when it loses the viewport yet if album art is displayed it is still there
15:52:08_bilgus__wonder if the AA makes a copy of the current vp earlier
15:52:27speachywell, the bit under the albumart is repainted...
15:52:59speachyI wonder if the problem is that the backdrop itself is getting overwritten
15:53:04_bilgus__yes but its calling display clear when the dirty bit is set
15:53:23_bilgus__try a song w/o AA and it works properly
15:53:34speachyyes, because the layout of the WPS is different
15:53:50speachythere's a line of WPS text painted over where the splash lands
15:53:51_bilgus__oh the backdrop makes sense
15:53:59_bilgus__i'll try turning it off
15:54:54_bilgus__damn nope same diff
15:55:17_bilgus__I made a much bigger splash
15:55:29_bilgus__its not an overwrite clearing this one
15:56:18speachyfrom what could tell the viewport dirty code only repained the actual viewports overlaid on top of the buffer, ignoring pixels that weren't part of the viewports.
15:56:26speachybut the backdrop was supposed to be painted first..
15:57:46_bilgus__I do like the idea though it should simplify some aspects of the UI if reliable..
16:10:47 Quit johnb7 (Quit: Nettalk6 -
16:11:16efqwfor some reason the battery bench plugin just writes the txt and never updates it any further
16:11:19efqwis this normal?
16:11:35speachyit's supposed to write once a minute
16:12:05efqwI can cat that file in console and the "Rockbox has been running for" is static
16:12:18efqwI get no new lines about the battery either
16:22:32 Quit lebellium (Quit: Leaving)
16:24:21 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
16:35:13blbro[m]Yay, GitHub actions now also has the Rockbox Utility Windows version building :) Hopefully also fixes the native Windows build issues I had.
16:35:41blbro[m]But cleanup and proper commit to master another day.
16:40:45_bilgus__UM i think BB only updates the file on shutdown
16:41:06_bilgus__it doesn't do it incremental AFAIR
16:41:41***Saving seen data "./dancer.seen"
16:42:34 Quit MrZeus_ (Ping timeout: 256 seconds)
16:44:37braewoods_bilgus__: no not yet
17:20:03 Join MrZeus_ [0] (~MrZeus@
17:27:15 Quit simpleOne (Quit: Textual IRC Client:
18:28:28 Quit pamaury (Ping timeout: 260 seconds)
18:40:22 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
18:41:05 Quit prof_wolfff (Ping timeout: 240 seconds)
18:41:42***Saving seen data "./dancer.seen"
19:09:46 Quit Jack87 (Quit: Jack has left the building.)
19:14:21 Join Jack87 [0] (Jack87@nasadmin/admin/jack87)
19:14:37 Quit Jack87 (Client Quit)
19:18:05 Join Jack87 [0] (Jack87@nasadmin/admin/jack87)
19:18:15 Quit Jack87 (Client Quit)
19:21:27 Join fs-bluebot_ [0] (
19:21:31 Quit bluebrother (Disconnected by services)
19:21:34 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
19:23:54 Quit fs-bluebot (Ping timeout: 256 seconds)
19:26:13 Join Jack87 [0] (Jack87@nasadmin/admin/jack87)
19:29:44 Quit Moarc (Ping timeout: 256 seconds)
19:29:45 Quit Jack87 (Quit: Jack has left the building.)
19:47:50 Quit APLU (Quit: !suicide)
19:49:23 Quit paulk-leonov (Ping timeout: 246 seconds)
19:51:40 Join paulk-leonov [0] (
19:53:35 Join APLU [0] (~mulx@2a03:7220:8081:2900::1)
19:55:18 Quit rogeliodh (Quit: The Lounge -
19:55:49 Join rogeliodh [0] (
20:03:15 Quit rogeliodh (Quit: The Lounge -
20:03:45 Join rogeliodh [0] (
20:06:44 Join prof_wolfff [0] (
20:11:11braewoodsthe first H320 arrived but my H120 adapter does not fit it. i guess i need a slightly larger one.
20:11:24braewoodsi ordered some usb to barrel jack adapters to try out
20:19:41_bilgus__#g2811 should be working right now speachy
20:22:21 Quit MrZeus_ (Ping timeout: 265 seconds)
20:38:02speachythank you, mr. somebody
20:41:45***Saving seen data "./dancer.seen"
20:55:30speachy_bilgus__: hmm... seems to still leave the artifact on the uisim
21:10:38speachydoes it on the real deal too
21:10:54speachy(with patch set 24)
21:21:52 Quit ps-auxw (Ping timeout: 246 seconds)
21:33:30 Join ps-auxw [0] (
21:34:39_bilgus__smh wtf
21:38:02_bilgus__sorry speachy I was rushing out the door earlier
21:41:24_bilgus__and duh don't know wtf I was thinking clearing the dirty bit prior to checking it seems to be my MO
21:55:11speachyI did that too in some of the hp/lo/muting code. :D
21:59:11_bilgus__I'm gonna get rid of the rest of the plugin compability stuff before I do another commit but if you want you just need to delete the clear flag stuff at the top of set_viewport in lcd_bitmap common /* when vp is set iclear dirty bit */
22:00:04_bilgus__lol I wonder sometimes about me
22:00:28 Quit cockroach (Quit: leaving)
22:11:01 Quit prof_wolfff (Ping timeout: 264 seconds)
22:40:31braewoodsyawn. these fakes on ebay get old.
22:40:43braewoods1TB micro sd? For less than $50? fake.
22:41:09braewoodsthe brand names cost a lot more than that...
22:41:48***Saving seen data "./dancer.seen"
22:57:40 Join Moarc [0] (
23:08:24_bilgus__gotta get rid of this bad silicon somehow more expensive than gold that sand is
23:45:47 Quit TheSeven (Ping timeout: 260 seconds)
23:46:07 Join [7] [0] (~quassel@rockbox/developer/TheSeven)

Previous day | Next day