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 2021-04-06

00:05:22***Saving seen data "./dancer.seen"
00:24:23 Quit dconrad (Remote host closed the connection)
01:31:58 Quit pixelma (Quit: .)
01:31:58 Quit amiconn (Quit: - Chat comfortably. Anywhere.)
01:32:57 Join pixelma [0] (marianne@rockbox/staff/pixelma)
01:32:58 Join amiconn [0] (jens@rockbox/developer/amiconn)
01:56:38 Join ZincAlloy [0] (
02:05:25***Saving seen data "./dancer.seen"
02:55:14 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
03:02:27 Quit ZincAlloy (Quit: Leaving.)
03:04:47 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:90db:5e98:ef13:4106)
03:16:07 Quit ZincAlloy (Quit: Leaving.)
03:34:16 Join ZincAlloy [0] (
04:05:28***Saving seen data "./dancer.seen"
04:35:39braewoodsspeachy: i was thinking of writing a different USB driver for basic remote control via the host
04:35:54braewoodsspeachy: kinda like reverse HID if you will
04:36:08braewoodsproprietary for rockbox i guess
04:36:29braewoodsMTP technically has these features but they're not really implemented by clients so
04:36:35braewoodsthis is probably better due to lower overhead
04:36:52braewoodsi envision the possibility of the ability to build your own remote control dock or so
04:37:02braewoodsnot for the average user for sure
04:37:30braewoodsimagine an arduino that can act as a host and instead send commands over usb
05:09:16braewoodshere's something you don't see too often
05:09:29braewoodsan x86 board with a public schematic
05:12:34 Join edhelas [0] (9d94237298@2a03:4000:51:f44:4e1:2ff:fe00:4257)
05:12:45 Part edhelas
05:12:59 Join edhelas [0] (9d94237298@2a03:4000:51:f44:4e1:2ff:fe00:4257)
05:14:02 Quit Moarc (Read error: Connection reset by peer)
05:15:41 Join Moarc [0] (
05:32:31 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
05:58:44speachybraewoods: The question I keep coming back to is: what would the utility be versus using a host-native media player?
05:59:51speachy(including native rockbox, I might add)
06:00:19speachyOf course, you're free to hack on whatever you'd like.
06:05:31***Saving seen data "./dancer.seen"
06:12:46speachy_bilgus: you use a clipzip as your primary device, correct? USB still reliable for you?
06:12:53fs-bluebot Clip Zip does not connect properly to PC on the first time (Windows 10) (bugs, unconfirmed)
06:34:26 Quit akaWolf (Ping timeout: 240 seconds)
07:11:25_bilgusdon't use windows but I'll try it
07:13:07speachyyeah, I figured as much.
07:18:25 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
07:19:30_bilgusIts actually faster than this last build?
07:19:42_bilguslike transfers 'seem' snappier
07:20:12_bilgusmight be something up but it looks to be doing just fine
07:20:57_bilgushahah and this is the key remap one so I can use the down button for the first time in like a year
07:21:15_bilgusit happen to be vol up but hey
07:23:51_bilgusTBH it wouldn't surprise me if it were a w10 thing
07:24:31_bilgusone of my kids hs a w10 laptop i'll try to get a confirmation on if the zip works there
07:48:47speachybased on experience it's far more likely to be a physical (port/cable) issue
07:57:20_bilgus& clipzip has a less robust connector than the plus
07:59:51 Quit pamaury (Ping timeout: 260 seconds)
08:02:10 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
08:05:32***Saving seen data "./dancer.seen"
08:10:39_bilgusFs#13282 sounds like someone needs an ipod
08:10:40fs-bluebot iPod Radio Remote functionality has regressed since 3.15 stable. (bugs, unconfirmed)
08:15:24speachyI have a mini2g that should be covered by that, but.. no remote
08:16:05speachybut the patch referenced _was_ tested by its author, so..
08:17:19_bilgusthe volume thing just needs a bit more finesse but without a device it sounds arduous
08:26:00 Join MrZeus [0] (~MrZeus@2a02:c7f:a0aa:4400:b027:e45f:4988:76bc)
08:35:10_bilgus13268 - Supports png format for album artwork we should support the format of their preferred music store to handle the data rationally?
08:35:45 Join eevan [0] (~eevan@
08:35:49_bilgusbest argument I've heard so far I guess
09:04:02speachywe already have PNG support in the imageviewer
09:04:53 Join massiveH [0] (
09:05:29speachyhuh, duplicated decoders
09:10:17edhelasspeachy duplicated code you mean ?
09:11:00speachyyeah, the album art code has its own jpg and bmp decoder, and there's another one in the imageviewer plugin
09:37:11_bilgusimageviewer already has a plugin overlay for those decoders itd probably make sense to do it that way
09:40:00_bilgusIOW its not really duplicated because it just gets thrown into the plugin buffer when needed itd be hard to tie that back to core being its already done
09:40:55_bilgusgoing the other direction would be nice.. loading the decoder from the imgviewer decoder ovcerlay
09:41:03_bilgusbut that seem slow
09:41:10_bilgusor sounds slow I should say
09:44:29 Quit ZincAlloy (Quit: Leaving.)
10:00:08 Quit atsampson (Ping timeout: 258 seconds)
10:02:05 Join atsampson [0] (
10:05:33***Saving seen data "./dancer.seen"
10:10:23 Quit eevan (Quit: Leaving)
10:28:19 Quit massiveH (Quit: Leaving)
10:34:37braewoodsspeachy: indeed. it wouldn't make sense for a normal host.
10:34:55braewoodsspeachy: but it could if you were using a periphereal pretending to be a host. like an IR remote bridge or so.
10:35:41braewoodsi've seen docks for HW specific units
10:35:45braewoodsand i thought
10:35:55braewoodsimagine if we could do that with a USB driver.
10:36:50braewoodsbut yea, it needs a client or so to make it useful.
10:37:20braewoodsthat can be provided by a regular host or something else even depending on your creativity.
10:37:54braewoodssince we only have a usb device stack
10:38:18braewoodsthis is the only real way to make it work
10:38:36braewoodsarduino has usb host stuff
10:39:47braewoodsthough i would probably end up using embedded linux
10:40:44braewoodsimagine being able to dock rockbox over usb :D
10:47:09 Quit DenimDev (Ping timeout: 252 seconds)
12:05:36***Saving seen data "./dancer.seen"
12:30:58 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:244d:5ea4:16be:cd26)
12:36:34speachyour png decoder doesn't let us scale on the fly. :/
12:39:28braewoodsspeachy: is there any existing RB functions to change the functional state?
12:39:34braewoodse.g., pause, alter volume, etc
12:39:43braewoodsor is it all button input controlled
12:42:49speachyno idea.. but one can always just inject events.
12:42:56speachylook at audio_playback_handler() in apps/playback.c
12:43:32braewoodsi'm planning to implement essentially a serial protocol for commands to rockbox
12:43:35braewoodsbasic stuff
12:43:37braewoodsno file IO
12:43:47braewoodsso it can be allowed to run in the BG
12:43:51braewoodswith minimal resource use
12:44:06speachywe have at least one implementation of this already, via IAP
12:44:43speachyand had another (that I think I removed as part of the archos stuff) that implemented the protocol used by Alpine automotive head units
12:44:56speachyapple's Ipod Accessory Protocol
12:45:09braewoodsisn't that limited to ipod targets only
12:45:30speachysure, but there's no inherent reason why it has to be
12:45:49braewoodswell given what's available
12:45:58braewoodsUSB serial is the best option i think
12:46:16braewoodshow did IAP work over?
12:46:18speachystick to ACM rather than "serial"
12:46:26braewoodsi thought it required
12:46:42braewoodsa bunch of pins in a special connector
12:47:03speachyultimately it's just a bidirectional serial protocol. (that can be tunneled over USB too)
12:47:51braewoodsdoes anything implement IAP on the host side?
12:47:55braewoodsi don't think so to my knowledge?
12:48:32braewoodsi just assumed IAP was utterly bound to their dock connector
12:48:57braewoodsif there's nothing else using IAP, how would tunneling it over USB help?
12:49:18speachygenetic linux implementation on gadget side, includes a client app
12:50:07edhelasI have a small iPod remote that can be tested with various iPod that I know if it can help
12:50:36edhelasiPod Mini 2G, iPod 3G, iPod Photo and iPod 4G
12:50:42braewoodsso you're saying we could reuse the IAP in rockbox for this?
12:50:46edhelas*I own
12:50:48braewoodsjust make it work over USB to be more generic?
12:51:29speachyedhelas: can you comment on fs#13282?
12:51:30fs-bluebot iPod Radio Remote functionality has regressed since 3.15 stable. (bugs, unconfirmed)
12:52:22braewoodsspeachy: when i think about it... adding a way to do this with other rockbox targets over their software usb stack...
12:52:25speachybraewoods: I'm not saying that reusing it is necessarily the CorrectAndProper(tm) approach to take, but it's worth at least looking into.
12:52:27edhelasI don't have the Radio Remote, the simple remote
12:52:36braewoodsspeachy: oh
12:52:43braewoodsspeachy: ok
12:52:50edhelasthat one
12:52:59braewoodsspeachy: it might be the best option though for compatibility with existing stuff
12:53:03edhelasthat actually never worked :D
12:53:09speachybraewoods: Apply has historically been quite pissy about other devices pretending to be iPods, but since they havent' sold any non-iOS devices for quite some time now...
12:53:59braewoodsspeachy: i see... my reasoning is just to allow use to (in theory) work with existing products
12:54:04braewoodson non-ipod target
12:54:16braewoodsassuming we can interface over USB
12:54:16speachyanyway. I keep coming back to.. if you have to tether the rockbox'd device to somethign capable of controlling it, that controller is probably a more attractive target to have play the audio to begin with.
12:54:39speachythe only remotely standard protocol for that sort of remote control is Apple's IAP.
12:54:45braewoodsyea exactly
12:54:59braewoodsi'll see if i can find a way to use it outside of ipods
12:55:09braewoodsotherwise not much point
12:55:25speachyvarious manufacturers of audio gear haave their own protocols for controlling car audio equipment (eg cd changers etc) but that's largely obsolete these days.
12:56:00speachyusing it outside of ipods is easy, but you'll have to have rockbox advertise the apple vid/pid and (probably) appropriate endpoints.
12:56:22braewoodsoh, we have to use their IDs?
12:56:36braewoodsthat's a problem
12:57:17speachyyep, that's how the widgets know they're talking to an ipod.
12:57:50speachyactually full IAP-over-USB would be _really_ useful, as pretty much any car audio system in the past decade will be able to control it.
12:58:14speachybut that requires streaming the audio out via USB (sort of like using an external DAC) rather than the internal DAC path.
12:58:30braewoodsentirely doable
12:58:37braewoodsand probably more useful than official usb audio
12:58:40speachywhich has its own uses independent of IAP.
12:58:54edhelasone small thing, IAP is also used using the top connector on old iPods
12:59:00edhelasnext to the jack one
12:59:30braewoodsedhelas: we're discussing the possibility of introducing it to non-ipods but ok
12:59:32speachyedhelas: but when they switched to the dock connector, they still supported serial IAP over that, unless I'm quite mistaken
12:59:58speachy(there are a _lot_ of docks out there that worked that way...)
13:00:13edhelasspeachy I see
13:00:47speachyanyway, rockbox supports IAPv1->v7.
13:01:03speachyfrom a protocol perspective. actual physical interfacing, no idea if any of that works
13:01:56braewoodsspeachy: yea, you're right.
13:02:15braewoodsi think we can bridge IAP to work on other rockbox targets that connect to regular hosts
13:02:30braewoodsprovided the accessory can be connected to via regular USB
13:02:57braewoodsthis sounds like a good choice for another USB mode
13:03:03braewoodsIAP mode
13:03:28braewoodsit pretends to be iPod so it can work with them
13:04:37braewoodsspeachy: you're right i think, this would basically make usb audio pointless
13:04:53speachyyes and no; different use cases.
13:04:59braewoodsi see
13:05:11speachyUSB audio is plug-n-play on pretty much any "real" OS out there
13:05:34speachyand you get an instant DAC (of varying quality)
13:06:07speachyIAP goes in the other direction.
13:06:58speachyfrom host-based remote controls to the host acting as the DAC and UI.
13:07:18speachy(with the advantage of being plug-n-play on hundreds of millionsof cars)
13:07:36speachy(but tbh bluetooth is a much better investment of time and effort)
13:08:30 Join lebellium [0] (
13:08:31braewoodsjust bluetooth has a few problems of its own
13:08:48braewoodseven if we had a BT stack right now
13:09:01braewoodsit's limited to devices that have a BT chip we could use
13:09:16braewoodsunless there's some USB bridge you know about
13:10:29speachythere are quite a few
13:10:44braewoodsaren't they all for hosts?
13:10:46speachygranted we'd have to implement USB host support
13:10:48braewoodswe aren't a USB host
13:11:33speachyno matter which way you slice it, a lot of code needs writing.
13:12:44braewoodsto really emulate an ipod we'd still need a way to bridge to the dock connector
13:13:19speachywhich in turn requires HW mods and/or USB host support. :)
13:14:06braewoodslooks like it exists.
13:15:36braewoodsstill interesting idea
13:15:37speachyso it's a USB DAC.
13:15:47braewoodsi guess so
13:16:20braewoodswe might be able to make a general subset of IAP support
13:16:41braewoodsif we can find an easy way to bridge to the dock connector's usb interface
13:16:46braewoodsassuming it does more than just provide power
13:17:46braewoodsanyway it's for another day
13:17:53braewoodsi have other projects to finish first
13:18:06braewoodsbertrik: what?
13:18:22bertrik"finish other projects first" is the universal excuse!
13:18:29braewoodsbut it's true
13:18:55bertrikI think that's why I have some many small side projects
13:19:07bertrikso I can avoid working on that other project
13:19:26braewoodsi'm in no condition to work on this right now
13:21:41 Join MrZeus_ [0] (~MrZeus@2a02:c7f:a0aa:4400:b027:e45f:4988:76bc)
13:24:39 Quit MrZeus (Ping timeout: 260 seconds)
13:34:37speachyfour more IP ranges blocked. overall bandwidth use nearly cut in half.
14:05:37***Saving seen data "./dancer.seen"
14:05:41fs-bluebotBuild Server message: New build round started. Revision 0fb333c85e, 298 builds, 10 clients.
14:20:50fs-bluebotBuild Server message: Build round completed after 909 seconds.
14:20:53fs-bluebotBuild Server message: Revision 0fb333c85e result: All green
14:21:01fs-bluebotBuild Server message: New build round started. Revision 1b314502c8, 298 builds, 10 clients.
14:35:14fs-bluebotBuild Server message: Build round completed after 853 seconds.
14:35:16fs-bluebotBuild Server message: Revision 1b314502c8 result: All green
14:35:17fs-bluebotBuild Server message: New build round started. Revision 28c89386af, 298 builds, 10 clients.
14:47:06fs-bluebotBuild Server message: Build round completed after 709 seconds.
14:47:08fs-bluebotBuild Server message: Revision 28c89386af result: All green
14:55:47 Quit CR0W (Ping timeout: 248 seconds)
15:05:39 Join CR0W [0] (
15:05:39 Quit CR0W (Changing host)
15:05:39 Join CR0W [0] (~narf@unaffiliated/em64t)
16:01:32 Join perricerdo [0] (
16:05:38***Saving seen data "./dancer.seen"
16:07:07 Quit perricerdo (Quit: Lost terminal)
16:24:21edhelasspeachy I just found that article, might help you :)
16:25:31speachywell, "speed" doesn't help us much when it relies on x86 SIMD for acceleration.. :D
16:25:50speachythe bigger concern for us is actually RAM usage.
16:28:03speachythey also achieve some of that speed/efficiency at the cost of potentially much higher RAM usage.
16:28:19edhelasI see :D
16:28:28edhelasso "thanks but not thanks"
16:29:00speachy(that said we actually decode the entire image at once currently, which is a major problem with often-quite-large embedded artwork..)
16:34:38 Quit kakaka (Remote host closed the connection)
16:35:15 Join kakaka [0] (~koniu@gateway/tor-sasl/koniu)
16:39:36 Quit ZincAlloy (Quit: Leaving.)
16:39:54 Join amachronic [0] (5284b95f@
16:42:08amachronichas anybody had issues with pictureflow lately? It's hanging for me on the Fiio M3K.
16:42:42braewoodsamachronic: native or hosted
16:42:44speachyOffhand I don't recall anyone reporting using it on any MIPS target.
16:43:12speachybilgus fixed up some issues with it a few months ago
16:43:19speachydoes it work in a sim?
16:43:32amachronicI'm compiling one now
16:44:57amachronichmm, seems to work but it tells me "Album Not Found"
16:45:37amachronicon the M3K it hangs at the end of removing duplicates.
16:48:07braewoodsnot too surprising, M3K is still really experimental
16:48:33braewoodseven so it's funny running into unresolved issues on stable ports too
16:50:50amachroniclol I know, I'm the author of the port. Just curious if the problem is M3K, MIPS, or more widespread (and hence worth a bug report)
16:52:46speachyI think I turned it on once on my X3 for giggles, it worked but was pretty useless given the awful screen.
16:57:35 Join jdarnley [0] (
17:03:27braewoodsyou know what might be useful for our rockbox utility packaging?
17:03:35braewoodsOBS support at some point
17:03:50braewoodswe could build specific builds for major distributions
17:03:54braewoodsusing the service
17:04:14braewoodsand have an external party host the binaries
17:04:50 Join ZincAlloy [0] (
17:06:28 Quit kakaka (Remote host closed the connection)
17:06:56 Join kakaka [0] (~koniu@gateway/tor-sasl/koniu)
17:35:25 Quit amachronic (Quit: Connection closed)
17:39:00 Quit ZincAlloy (Quit: Leaving.)
17:41:26 Join amachronic [0] (5284bbcf@
17:45:27 Quit lebellium (Quit: Leaving)
17:52:25 Quit jdarnley (Quit: ZNC 1.7.1 -
17:55:34 Join PimpiN8 [0] (~PimpiN8@
18:02:33 Quit PimpiN8 (Ping timeout: 265 seconds)
18:05:39***Saving seen data "./dancer.seen"
18:09:11 Quit Barlow (Ping timeout: 260 seconds)
18:10:05 Quit ArsenArsen (Changing host)
18:10:05 Join ArsenArsen [0] (~Arsen@managarm/dev/ArsenArsen)
18:51:56 Join tam [0] (
18:52:12 Quit tam (Client Quit)
19:01:58 Quit pamaury (Ping timeout: 240 seconds)
19:05:22amachronicokay confirmed there's an actual pictureflow bug here, FS #13283
19:05:23fs-bluebot Pictureflow segfault on targets with a portrait display orientation (bugs, unconfirmed)
19:07:15amachronic_bilgus: maybe you want to check it out
19:39:59 Quit amachronic (Quit: Connection closed)
19:50:31braewoodsbluebrother: does dual boot on the gigabeat S actually work? i'm reading conflicting reports about it.
19:50:40braewoodsthe bootloader code doesn't indicate it has any support at all
19:50:59braewoodsyet the manual says it does and even gives a method of how to trigger it
19:53:34braewoodsi'm going to explore the issue with the gigabeats i've got on order
19:53:50braewoodsbut if there is a way to dual boot... it should be documented
19:55:56braewoodsspeachy: were you looking for a agptek rocker?
19:56:16braewoodsspeachy: bilgus said he had one.
19:59:48 Join jdarnley [0] (
20:00:59 Quit MrZeus_ (Read error: Connection reset by peer)
20:01:39speachybraewoods: technically it's mine; I sent it to him many moons ago
20:02:14braewoodsspeachy: ok, i thought you'd forgotten about it perhaps
20:02:21 Join MrZeus_ [0] (~MrZeus@
20:02:54braewoodsspeachy: btw thanks for your advice. IAP is a much more useful solution.
20:03:10braewoodsadding support for the USB version of it will be useful albeit limited
20:05:41***Saving seen data "./dancer.seen"
20:08:06 Quit tomato (Ping timeout: 240 seconds)
20:08:42 Join tomato [0] (t0mato@gateway/vpn/mullvad/tomato)
20:09:35 Join Barlow [0] (
20:09:35 Quit Barlow (Changing host)
20:09:35 Join Barlow [0] (~barlow@unaffiliated/barlow)
20:11:01braewoodsspeachy: if this works out it means we can do a lot with many different apple compatible docks
20:11:13braewoodsbluetooth support indirectly to a degree
20:22:32 Quit kakaka (Remote host closed the connection)
20:22:59 Join kakaka [0] (~koniu@gateway/tor-sasl/koniu)
20:28:29braewoodshm. interesting.
20:28:39braewoodsi may end up building a new bootloader for these.
20:29:05braewoodsthere's unreleased improvements in the commit history
20:33:50 Quit MrZeus_ (Ping timeout: 265 seconds)
20:54:30 Quit ac_laptop (Ping timeout: 268 seconds)
21:05:59 Join ac_laptop [0] (~ac_laptop@
21:07:02 Join dconrad [0] (~dconrad@
21:08:58 Join Saijin_Naib [0] (
21:11:00 Quit Saijin_Naib (Disconnected by services)
21:11:04 Join Saijin-Naib [0] (
21:28:43 Quit Saijin-Naib (Read error: Connection reset by peer)
21:29:27 Join Saijin_Naib [0] (
21:53:27 Quit Saijin_Naib (Remote host closed the connection)
22:05:42***Saving seen data "./dancer.seen"
22:25:51 Join Saijin_Naib [0] (
22:28:11 Quit ufdm (Ping timeout: 240 seconds)
22:28:32 Join ufdm [0] (
22:30:42 Quit Saijin_Naib (Ping timeout: 258 seconds)
22:39:15braewoodsspeachy: can you see the problem with this line?
22:39:36braewoodsi was agast when i saw this
22:40:06 Join f1reflyylmao [0] (
22:40:22speachythe '!= 0' is superfluous?
22:40:36braewoodscheck sizeof and the type of the argument it is using
22:40:59braewoodsit's taking the size of a pointer
22:41:11braewoodseasy fix? convert to an array like it should be
22:41:21 Quit f1refly (Ping timeout: 250 seconds)
22:41:21 Nick f1reflyylmao is now known as f1refly (
22:41:37braewoodsthough given the intent
22:41:44braewoodsprobably sizeof then - 1
22:41:50braewoodsso the null terminator won't be included
22:42:08speachyit's a const array, so the compiler already does the right thing
22:42:18braewoodsdoes it now... it's declared as a pointer
22:42:22braewoodsi was getting a warning
22:43:41braewoodsas i thought
22:43:48braewoodsit's getting a pointer size instead
22:44:49speachythen by all means, fix the definition
22:45:00braewoodsi get this
22:45:06braewoods8 256
22:45:08braewoodsas my output
22:45:24braewoodsi've spent so much time around string arrays i saw this issue immediately
22:46:04speachystill, the net result is that we compare too little, but that's realistically enough for it to function "correctly"
22:46:20braewoodsbut still
22:46:31braewoodsincorrect logic is incorrect logic
23:22:21 Quit dconrad (Remote host closed the connection)
23:31:03 Join dconrad [0] (~dconrad@
23:31:52 Part dconrad

Previous day | Next day