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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2020-07-13

00:11:55 Quit ac_laptop (Ping timeout: 246 seconds)
00:28:03***Saving seen data "./dancer.seen"
00:34:38 Join Natch [0] (
01:13:53blbro[m]speachy rbutil doesn't use the genlang script on the server if the strings are on the player.
01:14:40blbro[m]There's a file iirc that has them. If that is missing it queries the server.
01:15:39blbro[m]Since we have that in place I was wondering before if we should kill the genlang www stuff ... Forgot about it though :)
01:23:13 Join ZincAlloy [0] (
01:27:23 Quit ZincAlloy (Ping timeout: 240 seconds)
01:39:19 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:b96f:9cb1:eee3:2171)
01:43:59 Quit ZincAlloy (Ping timeout: 272 seconds)
02:24:27 Join mendelmunkis [0] (
02:24:30 Quit mendel_munkis (Remote host closed the connection)
02:28:07***Saving seen data "./dancer.seen"
02:34:10 Join advcomp2019 [0] (
02:34:10 Quit advcomp2019 (Changing host)
02:34:10 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
02:36:37 Quit advcomp2019__ (Ping timeout: 265 seconds)
02:37:09 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
03:09:31 Quit Oksana (Read error: Connection reset by peer)
03:28:36 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
03:32:42 Join petur [0] (~petur@
03:32:42 Quit petur (Changing host)
03:32:43 Join petur [0] (~petur@rockbox/developer/petur)
03:58:20 Quit S|h|a|w|n (Read error: Connection reset by peer)
04:00:38 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
04:28:09***Saving seen data "./dancer.seen"
04:28:22 Quit pamaury (Ping timeout: 256 seconds)
04:30:57 Quit S|h|a|w|n (Read error: Connection reset by peer)
04:46:26 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
05:57:56 Join kann [0] (kann@fsf/member/kann)
06:28:10***Saving seen data "./dancer.seen"
06:59:01 Nick mendelmunkis is now known as mendel_munkis (
06:59:23mendel_munkisTIL: updating while in middle of testing debug builds is a bad idea.
06:59:49mendel_munkis(thankfully I didn't overwrite my production build with a test)
07:05:17mendel_munkisBilgus: I'm not sure if it's related to your pictureflow fixes but I just updated and pictureflow no longer finds any album artwork. (the previous version was based on HEAD sep 2019)
07:08:46mendel_munkisturning off resize album art fixes that.
07:30:59speachyblbro[m], as long as rbutil calls that URL it'll have to stay on the www site.
07:32:28speachyand.. the stuff in has gaps vs English.
07:40:21 Join massive_H [0] (
07:41:20blbro[m]That shouldn't have gaps ... Rbutil uses that for voice file creation if present.
07:42:53 Join Ckat [0] (~Ckat@xn--z7x.xn--6frz82g)
07:42:54 Quit massiveH (Ping timeout: 256 seconds)
07:43:10speachyno gaps, just missing entries at the end
07:43:40blbro[m]Though ... Maybe it falls back to english if entries are missing? Can't remember.
07:43:42speachygenlang acts differently in its various modes.
07:44:10 Join __Bilgus_ [0] (41ba23be@
07:44:38__Bilgus_mendel_munkis Thanks, I'm sure its related at least in proximity if nothing else
07:44:54__Bilgus_so no artwork found while resize is enabled
07:45:46speachyapparently in the binary blob mode (ie for the player firmware to consume) genlang removes all strings not present in English, and anything missing gets left out.
07:46:22speachyso the player firmware must be the one falling back to English for missing text strings.
07:48:37blbro[m]Hmm? I always thought genlang does the fallback. Much more complicated on the player ...
07:49:16 Join ac_laptop [0] (~ac_laptop@
07:49:25speachythe way it seems to work on the player is that there's table of pointers for each string ID, and that gets updated for each string present in the new language file.
07:50:04speachyand it always loads English first (as it's built into the binary) so that's how the fallback is done
07:50:57__Bilgus_mendel_munkis not sure whats going on there I can't repro and don't see what resizing the art has to do with not finding it eithe
07:52:41__Bilgus_I assume you let it finish building the cache at the beginning w/ progressbar did it give any error messages like 'bad artwork'
07:55:33speachyhere's the fun bit −− genlang generates different voicestrings if you're in binary vs voicebinary vs voicelist mode. only binary + voicebin did the right thing.
07:55:51speachy(by stripping out strings not present in English)
07:56:08speachyvoicebin-only and voicelist-only did not.
08:01:20speachytrying to figure out if always stripping not-present-in-English will break anything. (the -u option, maybe?)
08:04:03speachyso, the big question −− is it better to have English-string voice fallback, or no voice at all?
08:07:38speachy(and that's "english as pronounced by $lang's tts rules")
08:15:47__Bilgus_oh mendel could you try setting resize back to true and see if it works now it could be that the cfg didn't rebuild the cache till you choose to not resize artwork
08:16:46__Bilgus_speachy thats probably still preferable to silent failure
08:17:53speachyso I've fixed genlang to filter vs english for all voice modes
08:18:26speachynow comes filling in the missing strings for
08:18:36__Bilgus_weird I thought it already did that not just in special circumstances
08:19:13speachyit only filtered if "binary" mode was on (ie what generates the player-ready .lng from the source .lang files)
08:19:52speachysince the makefiles generated binary .voicelist at the same time, it accidentally worked
08:20:22speachybut it didn't work right if you generated non-binary voicelists (ie what an AV build with 'make voice' does)
08:20:31*speachy shakes his head.
08:21:26__Bilgus_well at least it fails loudly now!
08:22:12speachyat least it's fixed in 2/3 use cases now. :)
08:22:52__Bilgus_makes me wonder how many times misformed voice files were the fault of a non reproducible crash
08:23:44speachy...all I wanted to do was generate some non-english voices nightly.
08:23:51speachy... serves me right
08:24:08__Bilgus_I think thats how cans of worms generally get opened
08:25:53speachyokay, is now done right.
08:26:35speachyonly added something like 70K to the zip file.
08:27:34__Bilgus_BB was saying it grabbed it off the server if it didn't exist so same amount of bandwidth used I presume
08:28:12***Saving seen data "./dancer.seen"
08:28:30speachyeach binary file will be larger, so there's definitely going to be a hit
08:29:29 Quit massive_H (Quit: Leaving)
08:29:39 Join vmx [0] (
08:29:52speachyg#2526 andg#2527
08:29:55 Join massiveH [0] (
08:29:55fs-bluebotGerrit review #2526 at : genlang: Skip strings not present in english in all voicelist modes by Solomon Peachy
08:29:55fs-bluebotGerrit review #2527 at : voice: Ensure all voicelist files in are complete by Solomon Peachy
08:34:32fs-bluebotBuild Server message: New build round started. Revision ca31bad, 295 builds, 10 clients.
08:53:21fs-bluebotBuild Server message: Build round completed after 1129 seconds.
08:53:22fs-bluebotBuild Server message: Revision ca31bad result: 6 errors 83 warnings
09:12:04mendel_munkis__Bilgus_: yes I waited for it to rebuild the cache with no errors (several times.)
09:12:22mendel_munkisIt's rebuilding the cache with resize reenabled now
09:12:25speachyestonian had a couple of issues, fixed..
09:12:59mendel_munkisyeah apparently it lied about rebuilding the cache
09:13:11mendel_munkisit works fine now with resize on.
09:14:34mendel_munkisdoes logf to serial break dumping log to file?
09:15:41speachydamnit. voicestring generation still not complete −− voice-only strings aren't being backfilled.
09:22:16speachyno, they are, but the ordering is different.
09:22:16speachythe ID is correct but the output order is different. I hope that doesn't actually matter.
09:26:39fs-bluebotBuild Server message: New build round started. Revision cb3363f, 295 builds, 10 clients.
09:27:43 Quit vmx (*.net *.split)
09:27:43 Quit petur (*.net *.split)
09:27:43 Quit lemon_jesus (*.net *.split)
09:27:43 Quit klock (*.net *.split)
09:27:44 Quit flabrus (*.net *.split)
09:27:44 Quit gevaerts (*.net *.split)
09:27:51speachyblbro[m], your builder is still hiccupping on simulator builds. can you investigate why?
09:28:12 Join vmx [0] (
09:28:12 Join petur [0] (~petur@rockbox/developer/petur)
09:28:12 Join lemon_jesus [0] (
09:28:12 Join klock [0] (~freeklock@unaffiliated/klock)
09:28:12 Join flabrus [0] (~beard@unaffiliated/flabrus)
09:28:12 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
09:30:44 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:6dd8:2f55:896c:90ed)
09:51:49 Quit ac_laptop (Ping timeout: 264 seconds)
09:54:04*speachy gently prods fs-bluebot
09:57:56 Quit massiveH (Quit: Leaving)
10:06:21Strife89Lol, so much for my desktop being "decent"
10:06:44Strife89Fourth-lowest score
10:06:50speachythe upload speed hurts you. but don't forget the others all have hot ccaches
10:07:48Strife89Good point
10:09:23 Quit MrZeus (Ping timeout: 272 seconds)
10:09:37gevaertsMy builders are a core2 and an atom. You must win against those! :)
10:13:51Strife89By points/sec., I do! Barely!
10:15:04gevaertsAh, I lied. Not an Atom, an AMD thing of similar status
10:15:13Strife89Ahhh, neat
10:15:24speachyone of the Bobcat cores?
10:15:26Strife89I forgot AMD even participated in that class
10:15:33gevaerts"Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz" and "AMD Athlon(tm) II Neo N36L Dual-Core Processor" according to /proc/cpuinfo
10:18:02Strife89Oh man, I remember the desktop I used to have. Called it "tsubotsubo", and for most of it's life, it's had a Celeron 450
10:18:36Strife89And I tried running it as a builder
10:19:14 Quit vmx (Ping timeout: 272 seconds)
10:28:13***Saving seen data "./dancer.seen"
10:28:37 Join ac_laptop [0] (~ac_laptop@
10:57:01Strife89speachy: I know you've been busy with genlang, but I was wondering whether you looked at the batterybench I pasted last night.
10:57:22*speachy not-so-politely pokes fs-bluebot.
10:59:21Strife89You might have to smack it with a large trout ...
11:01:34speachynope, haven't checked into that
11:01:39speachyone fire at at a time.
11:02:49speachyI just enabled the code that builds multiple voices nightly. Only have three enabled that were positively confirmed as sane (english, english-us, and polish)
11:04:03*speachy kicks off another Vivado run.
11:04:05 Quit livvy (Remote host closed the connection)
11:05:31 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
11:06:33speachyIt's a shame that there's no easy way to optimize the post-commit tasks to not run if nothing relevant changed.
11:10:22bluebrotherspeachy: I disabled sim builds on that box for now.
11:13:07bluebrotherno idea what's going wrong. When trying to build manually everything's fine :/
11:14:24fs-bluebotBuild Server message: Build round completed after 1151 seconds.
11:14:26fs-bluebotBuild Server message: Revision ead398a result: All green
11:23:54speachygotta love those heisenbugs
11:25:12speachywtf, vivado segfaulted? that's... new.
11:39:40 Quit petur (Quit: Connection reset by beer)
11:56:35 Quit livvy (Remote host closed the connection)
12:03:59 Join vmx [0] (
12:06:00 Quit sakax (Remote host closed the connection)
12:06:51 Join livvy [0] (~livvy@gateway/tor-sasl/livvy)
12:07:51 Join ZincAlloy [0] (
12:12:45 Quit ZincAlloy (Ping timeout: 265 seconds)
12:13:23speachyrather than attempting to recreate make in shell scripts... why not just do it in make to begin with..
12:24:07 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:45bf:c622:2f72:4a3e)
12:28:15***Saving seen data "./dancer.seen"
12:28:43 Quit ZincAlloy (Ping timeout: 272 seconds)
12:42:17Strife89My laptop joins the build farm
12:42:35Strife89While on my work's fast(er) Internet :D
12:55:23 Join Jiri [0] (b2ffa81d@
12:57:41 Quit Jiri (Remote host closed the connection)
12:59:18 Join man150 [0] (b2ffa81d@
13:01:47 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:b03b:93ef:af2c:7a44)
13:05:07man150 Hello Speachi, I have tested your new dayily generated voices, I have tested Slovak and English languages, Both are OK, I can test the Czech language too.
13:13:20speachyman150, thanks for the sanity check. I did not generate czech voices though, as it fell below the "good" coverage target (95%) I used.
13:15:12speachyCzech is at 91% coverage
13:15:18man150I can translate new strings in the Czech language
13:15:31speachythat would be appreciated!
13:16:12 Quit man150 (Remote host closed the connection)
13:19:03 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:4d18:ea6:8745:f155)
13:22:03speachyworth mentioning that there are global "voice corrections" that do things like sound out acronyms ("MP3" -> "M P 3") either globally or per-language
13:31:35speachyStrife89, if you'd like to try another newer-toolchain-compiled build on your ipodcolor, let me know. I'm curious if the startup issues you were having are resolved, or if there's still something wrong with that target (in a way that the mini2g doesn't demonstrate)
13:32:14 Join lebellium [0] (
13:35:47 Join ShockTohp [0] (~kevin@
13:55:12 Quit ShockTohp (Ping timeout: 256 seconds)
14:21:16 Quit vmx (Ping timeout: 256 seconds)
14:28:19***Saving seen data "./dancer.seen"
14:45:04Strife89speachy: Sure, I'm willing to try
14:50:18Strife89Thanks, I'll try it out once I get to a stopping point
14:53:36 Join vmx [0] (
15:00:01 Quit vmx (Quit: Leaving)
15:04:15Strife89speachy: Still get "failed reading .voice" on startup, and the device appears to freeze when I attempt to shut down
15:05:33Strife89Scratch that, now I think it's just freezing shortly after it reaches the main menu
15:05:54Strife89The backlight is stuck on, so I don't think it's an I/O wait issue
15:06:33speachyokay, so it's no good. good data point, will have to look into whre the mini2g and color diverge
15:07:09Strife89Third boot attempt: freeze on Apple logo
15:07:35speachythe only difference between that and the build you were previously running is being compiled with '-Os'
15:07:37Strife89Fourth boot: Definitely freezes even if I don't attempt to go anywhere from the main menu
15:07:50speachyso something's clearly being miscompiled.
15:08:58Strife89Reverting to my previous build :)
15:49:20 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
15:55:04 Quit __Bilgus_ (Ping timeout: 245 seconds)
15:55:37 Join ShockTohp [0] (~kevin@
15:59:46fs-bluebotBuild Server message: New build round started. Revision 9adfab9, 295 builds, 11 clients.
16:06:14 Quit ShockTohp (Ping timeout: 240 seconds)
16:12:24fs-bluebotBuild Server message: Build round completed after 758 seconds.
16:12:25fs-bluebotBuild Server message: Revision 9adfab9 result: All green
16:13:13 Join ShockTohp [0] (~kevin@
16:28:23***Saving seen data "./dancer.seen"
16:45:18 Quit rogeliodh (Ping timeout: 256 seconds)
16:57:06 Join reductum [0] (
16:58:52 Join rogeliodh [0] (~rogeliodh@
17:00:25 Quit MrZeus (Ping timeout: 272 seconds)
17:10:20 Quit efqw (Quit: Connection closed for inactivity)
17:10:25 Quit ShockTohp (Ping timeout: 264 seconds)
17:57:27 Quit lebellium (Quit: Leaving)
18:09:20 Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach)
18:21:48 Quit ZincAlloy (Quit: Leaving.)
18:24:37 Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide)
18:27:12 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:e95a:cb84:4ae7:4c32)
18:28:26***Saving seen data "./dancer.seen"
18:50:34 Quit sakax (Remote host closed the connection)
18:57:05 Join jdarnley [0] (
18:58:55 Quit J_Darnley (Ping timeout: 246 seconds)
19:03:35 Quit ac_laptop (Ping timeout: 265 seconds)
19:04:50 Join efqw [0] (uid412670@gateway/web/
19:40:36 Quit MrZeus (Read error: No route to host)
19:42:09 Join MrZeus [0] (
20:10:17 Quit MrZeus (Ping timeout: 265 seconds)
20:22:48speachyyay, got the multiboot loader installed on my old clip+. no more internal flash wear..
20:27:53Strife89I forgot that was a thing! How's it coming along?
20:28:22Strife89I have a Clip+, Clip Zip, c250, and e260 - all of which could benefit ...
20:28:29speachywell, means I can once again use it again without wondering if each bootup will be the last
20:28:30***Saving seen data "./dancer.seen"
20:28:51speachythe multiboot loader has been live on the download site (and rbutil) since the end of 2020
20:29:11 Join TheEaterOfSouls [0] (~souls@unaffiliated/theeaterofsouls)
20:31:34 Quit reductum (Quit: WeeChat 2.8)
20:31:45speachyI think it's just the clip+, fuze+, clipzip, and fuzev2 −− the c250/e260 aren't covered.
20:31:55speachyfuzev1 too
20:32:35Strife89So is this a thing built into current builds (for supported players)? Other than putting .rockbox and the "player" stub on the microSD, do I need to do anything else (such as updating the bootloader)?
20:33:05speachyyou have to re-flash the bootloader (via rbutil or whatever)
20:33:38Strife89Also, are you a time traveler?
20:33:43speachyand a dev/daily build.
20:34:00Strife89"the multiboot loader has been live [...] since the end of 2020"
20:34:23speachyeh, it's just a typical off-by-one programmer error.
20:34:32speachy(that should have read ...2021)
20:37:34Strife89I miss having a Fuze v1 ... those things are freakin' expensive everywhere I've looked
20:37:51speachyI have a dead v1 and several v2s in my pile of broken dreams.
20:38:01Strife89I went through two while I was in college. Both died of unknown causes
20:38:13speachythey were awesome but physically pretty fragile. plus with that awful ribbon connector that would break if you breathed on it funny
20:38:32Strife89Guh. No wonder.
20:38:44TheEaterOfSoulsMy Clip v1 still works
20:38:52TheEaterOfSoulsBought in 2008
20:39:20speachyif left in vehicles the heat cycles would make the case more brittle, combined with battery swelling, broke the cases on more than one of mine
20:39:45Strife89I once had an iPod Video which once flew out of my jacket pocket, tumbled on the concrete, *came open*, and it was still fine. (What killed it was me, several years later when the Select button broke and I took it apart to take a look - and accidentally snapped off the power connector)
20:40:34Strife89Meanwhile the Fuze could die if you glared at it
20:41:01speachythey didn't like being inside a black car in florida summers.
20:41:17speachybut even beyond that I seem to be pretty destructive to my players even though they rarely leave the car
20:41:38Strife89Do you have butterfingers like me?
20:41:58speachyyes, but with high luck rolls
20:42:40speachythe x3 I've been using has proven to be physically indestrucitble. I could probably drive over it.
20:43:01TheEaterOfSoulsSo I just bought an XDuoo X3
20:43:10Strife89Also nice
20:43:28Strife89I hope to be able to buy one someday
20:43:41TheEaterOfSoulsI'm blind though, so have no idea what I'm doing until I can get Rockbox booted with a voice file
20:43:46speachythe v1 is hard to find. v2s seem to be going out too.
20:43:49Strife89Oh, ouch.
20:44:14Strife89To both your blindness and the scarcity of X3s
20:44:15speachyTheEaterOfSouls, getting the initial setup will be a challenge without being able to work the original firmware menus
20:44:28TheEaterOfSoulsWould anyone mind telling me what keys I need to press to get to the firmware update option in settings?
20:44:43TheEaterOfSoulsYeah, I figured
20:44:48speachysure, I can walk you through that.
20:45:14Strife89I should take a look at the manual for the X3 (if one exists) to make sure it can walk one through that.
20:45:49speachyTheEaterOfSouls, you will need three things to start:
20:45:50Strife89The manuals have been written so that a blind person can gather enough info to perform an install on any given player.
20:46:34speachy1) the original XVortex-patched flash image (since we still can't recreate it yet..)
20:46:44speachy2) the most recent daily build
20:46:53speachyand 3) the most recent voice file
20:47:27speachyonce all of that is on the sd card, the rest should be automagic save for the OF button pressing.
20:48:28TheEaterOfSoulsI generate voice files, but yeah
20:48:50TheEaterOfSoulsWait, what about the bootloader listed on the wiki page? Do we not use that?
20:48:57speachy(I've had "Write the X3 manual" on my todo list for at least a year)
20:49:07speachyTheEaterOfSouls, is this the original X3, or the X3ii?
20:49:15TheEaterOfSoulsThe II
20:49:24TheEaterOfSoulsSorry, should've been more clear
20:49:56speachyOk, then you'll just need the pre-patched bootloader image linked off the wiki page
20:50:12TheEaterOfSoulsAnd now I can't find my SD card
20:51:40speachyI am not able to walk you through the X3ii's button presses as I only have the original X3
20:52:15TheEaterOfSoulsAh well
20:56:04Strife89There's an app on the Google Play Store called "Be My Eyes" (without spaces, alas), which I've thought about trying out
20:57:07Strife89As you may guess, the app has a blind user send their phone's camera feed to a sighted user to have them describe objects in front of them
20:58:11TheEaterOfSoulsYeah, I could use that. I'll probably just ask one of my roommates, though
20:59:23Strife89Glad to hear that you have one
20:59:39TheEaterOfSoulsThanks :D
21:03:23Strife89Following up on "thought about trying out": I haven't actually tried it because (1) my Internet sucks and I probably wouldn't be very helpful as a result, and (2) I don't enjoy video calls
21:03:55TheEaterOfSoulsI don't either, honestly
21:05:02Strife89Oh gosh, I can only imagine
21:05:27Strife89Having to position your face, etc. for the other party's camera when you can't even see it?
21:05:52Strife89I'm picturing the experience of a regular voice call on speakerphone, except worse
21:05:57TheEaterOfSoulsEh, it's not really that. I just don't do well talking to people
21:06:51Strife89Well, that I think I understand, too, at least as someone who leans introverted
21:08:24Strife89But I beg your pardon and your swift rebuttal if I make an inconsiderate remark
21:10:36Strife89speachy: So, rbutil 1.4.1, right?
21:10:57speachyI built it out of git, but yeah, 1.4.1 should be fine
21:11:05TheEaterOfSoulsHa, found it
21:11:38TheEaterOfSoulsWhat inconsiderate remark? I don't follow
21:12:07Strife89I'm hoping that I haven't made and do not make one! I'm a worry-wart though
21:12:26TheEaterOfSoulsNah, it's fine
21:14:45Strife89speachy: Guess what - I've run into trouble again
21:15:31speachystuff like that is why I just build it myself
21:15:56speachyI don't know what debian calls its libcrypto++
21:16:05Strife89package search for libcrypto and what I have installed
21:16:10speachyor what version it is (in case you do have it installed)
21:16:29Strife89I do, version 6
21:16:41Strife89Er, 5.6
21:17:35Strife89Guess I'll try to compile
21:17:44speachydo you have libcryptopp-dev installed?
21:18:19Strife89Nope, it doesn't appear to exist in Debian's repos
21:18:28Strife89Closest is libcrypto++-dev
21:18:37speachyyeah, that
21:18:54Strife89Installing that now
21:19:10speachywe've talked about scrapping libcrypto++ in favor of embedding a few bits of libtomcrypt.
21:20:12Strife89Do you mind giving me the gist for building rbutil? My first thought was ./rockbox/tools/configure, but of course it's not there.
21:20:27speachycd tools/rbutil/rbutilqt ; make
21:23:03Strife89The checkout I'm working in must be borked or something. tools/rbutil doesn't exist, even after a git pull
21:23:31TheEaterOfSoulsThis X3ii has a weird key layout
21:24:16speachythe X3ii is less odd than the original X3. though IMO the original actually works very well for operation without seeing it.
21:24:19Strife89No makefil
21:26:43Strife89I ... may need to install qt5 packages
21:29:25Strife89"qmake-qt5" is "installed", so I think I'm fine. I try to tab-complete "qmake" and get "qmake" and "qmake-qt4".
21:29:42speachymight be just plain qmake then
21:29:43Strife89You wanna know what "qmake" by itself is on my system? Q. T. 3.
21:29:49speachyah neverind
21:29:53Strife89Check the end of that paste
21:32:09Strife89Should I just - I dunno - set up an Ubuntu VM or chroot and try to build in *that*?
21:32:37Strife89Or Arch, Debian testing, something similar
21:33:02speachyyou need at least qt4 to build rbutil
21:33:16speachy(so qmake-qt4)
21:33:52Strife89Unfortunately, it spits out errors right away. Very similar errors.
21:35:14Strife89You know, I just had an idea
21:37:35Strife89Nope, didn't really help
21:38:01Strife89What I tried:
21:38:23speachyno, that's no good
21:39:10Strife89Yeah, I kind of figured it was dumb to begin with
21:39:15speachythere's always the likes of 'apt install *qt5*
21:39:50speachyor apt install *qt5*dev
21:40:27Strife89... yeah, my brain isn't firing on all cylinders tonight
21:40:39Strife89And that's a feat, because it doesn't have many to start with
21:42:17Strife89But I note that "Qt_5.13" is in the error in my last paste, and the Debian repos appear to have Qt 5.11 ... I might be up against a brick wall without switching branches, if not distros
21:44:33TheEaterOfSoulsIs there a tool for generating voice files that isn't rbutil?
21:44:56speachyyou can generate stuff via the git repo
21:45:03Strife89Trying to search the backports available
21:46:43speachyTheEaterOfSouls, I assume you're wanting to generate clips for the files you're loading, as opposed to voiced menus?
21:47:34speachythe x3ii doesn't exist as far as rbutil is concerned, so you won't have much luck with the latter. I think it should be okay with the former.
21:49:24speachyto be honest I've been procrastinating from digging into rbutil. I'd rather pour over asm listings and crash dumps than hack on C++.
21:52:17Strife89Well, the ethos of Rockbox's devs has always been that the devs work on what *they* want to work on, not what the users want. :)
21:52:50Strife89And I've always been just a user
21:53:39TheEaterOfSoulsNo, I was thinking voice menus
21:54:08TheEaterOfSoulsThe autogenerated ones use Festival which is not great, but I'll live
21:55:57 Quit cockroach (Quit: leaving)
21:56:03Strife89(reply #13 is what I'm trying to point to)
21:56:57speachyvoice operation is probably rockbox's only true killer feature.
21:58:08Strife89I wouldn't say "only", especially given the current mobile device market
21:58:18speachymost of my work on rockbox has been to improve that, and rbutil's lack of support for the only commercially-available players we support is a considerable speedbump for blind users.
21:58:28TheEaterOfSoulsYou think? I always found all the plugins and stuff pretty cool, even though I can't use most of them
21:58:37TheEaterOfSoulsFormat support,in the past
21:59:06speachyin modern times, we're compared to smartphones and their app ecosystem.
21:59:26Strife89And their blind-unfriendly touchscreens
21:59:53speachyyeah. and always-streaming music.
21:59:55TheEaterOfSoulsI love these devices Rockbox runs on that are now kind of niche. They do exactly what they're intended for with no nonsense
22:00:06Strife89Almost everything Rockbox runs on is controlled only by buttons
22:00:09speachyTheEaterOfSouls, exactly.
22:00:54TheEaterOfSoulsYeah, not a fan of streaming services and DRM. Give me what I paid for, please
22:01:04speachybut it's a small niche that's getting harder to play in.
22:01:06TheEaterOfSoulsActually, touchscreens can and do work for the blind
22:01:48TheEaterOfSoulsBut I much prefer a player running Rockbox to listening to music on my phone. The UI is just more practical, and the battery life is much longer
22:02:12speachyDAPs only really sell at the ultra-low-end that will never run rockbox, and those catering to "audiophiles" who buy oxygen-free audio cables.
22:02:24Strife89Yeah, I miss the days where most electronics were single-purpose
22:02:41Strife89(Not enough to go BACK to those days entirely, but I still miss them)
22:02:54TheEaterOfSoulsI also don't want to waste my phone/laptop battery when I have a device that can play music easily three times as long
22:02:55speachyand even then mostly for the Chinese domestic market.
22:03:54speachyso we get scraps with zero documentation available, and (usually) no source code for the GPL'd bits, so at best all we can really accmplish is running as a Linux application on a sorta-good-enough platform.
22:04:07TheEaterOfSoulsSame. It's not a fashionable view, but I'm really tired of devices just trying to do and be everything now
22:04:44speachythe hardware is more complex than ever and the product lifecycles are so short that by the time we can get things working well (if at all) stuff is already discontinued.
22:05:16Strife89Ooof, that last point is a big one.
22:05:18speachyas that forum thread Strife89 put it, the skills needed to do this stuff are pretty specialized.
22:06:10speachyand (for example) the only reason I've had so much time to put into Rockbox lately is because $dayjob build/test cycles are multi-hour affairs.
22:06:25speachyreverse engineering on a (very) part time basis is a pretty slow process
22:06:57TheEaterOfSoulsI love what you guys do, though, even if it is just a hosted application now
22:07:39speachyI've been trying to get rid of the roadblocks to new contributors −− be it the infrastructure, build system, and whatnot, and also trying to re-encourage old contributors that gave up while their patches bitrotted
22:08:46 Quit rogeliodh (Ping timeout: 256 seconds)
22:09:01speachyall of this to try to build up some momentum again.. support on some modern players brings new users in, fixing up warts with multilingual voices (ie most of the past week) greatly expands the userbase for whom rockbox will JustWork(tm).
22:10:28Strife89I wish you luck, speachy.
22:10:45speachybecause damnit, I don't want to see rockbox bitrot away into irrelevance.
22:11:03Strife89That scares me, too
22:11:06speachywhen everything else sucks so badly
22:11:13TheEaterOfSoulsSeriously, thanks. It's good to know there are people who still believe in it
22:12:09 Join rogeliodh [0] (~rogeliodh@
22:12:22speachyIf I didn't step up and contribute my expertise (which lines up pretty well with the skills needed) then how could I ask (or expect) anyone else to do so?
22:12:41speachyand I say that as someone who's been using rockbox since the original Archos player way back when.
22:12:55speachyonly took me 1.5 decades to meaningfully contribute back. :D
22:13:21TheEaterOfSoulsI intend to contribute as well, if I ever get skilled in that area
22:15:56TheEaterOfSoulsIs there a potential for rockbox to run natively on players like the X3, do you think?
22:16:08TheEaterOfSoulsEventually, that is
22:18:43speachyI think so; the datasheets for the major components are available I think.
22:19:15speachythe hardware in the X3ii is pretty kick-ass and we'd get a lot more usable RAM out of a native port
22:28:33***Saving seen data "./dancer.seen"
22:29:27Strife89If I ever get out of this funk, I want to help work on the manual again
22:29:50speachyon the other hand, simply being able to really customize the linux load (especially the kernel) would probably be much more beneficial.
22:30:00speachyfrom a reward-for-effort perspective.
22:30:15Strife89I have zero coding skills, but I managed to pick up LaTeX well enough the one time I made a contribution
22:30:24speachythe AGPTek Rocker has some really annoying quirks that could be trivially fixed if we could rebuild the kernel.
22:32:15speachywith that source we would also have much better insight into how the hardware is actually put togher; greatly reducing the amount of reverse-engineering needed to do a native port to the X1000E
22:34:50speachyand (to take this conversation back full circle) the reason HW makers don't want to try and ship rockbox is that we're missing the various fancy UI bells-n-whistles and DRM and integration into external music services and the myriad other things supposedly needed to appeal to the modern (mostly Chinese) market.
22:35:19speachybut that's fine, we'll outlast them all. :)
22:35:54speachyas long as we manage to keep developer interest going.
22:37:29*speachy gets off his soapbox and returns to his normal introverted self.
22:39:25Strife89Whoever managed to convince the world that all modern tech had to be Pretty™ ought to be slapped
22:40:25Strife89It's incredibly subjective anyway
22:40:25speachyI disagree, "pretty" has always outsold not-pretty, all else being equal.
22:40:58speachythe ultimate goal of "tech" is to become invisible.
22:41:18speachybut advertising requires "engagment" which is at odds with being invisible.
22:43:38Strife89Yeah, good points
22:46:01speachyI fear that by the time modern DAP makers realize the folly of mostly-touch interfaces, the market will probably be gone completely.
22:47:11Strife89Suppose that they already have, but that they also believe that trying to go back is foolhardy
22:47:24speachybecause it's no longer as pretty and thus won't sell as well
22:47:37mendel_munkisI would suggest suing agptek for the source if not for the fact that they did send pamaury a player to work on.
22:48:02Strife89On the other hand, the nostalgia market makes beau coup riches
22:48:04speachyexcept in the o2-depleted-cables high-end DAP+DAC+amp+whatever fanboy equipment that's usually immune to reason
22:48:37Strife89So if Apple ever decides to make an "iPod Classic Edition", well ... the devices - and the money- practically make themselves
22:48:58speachyI'm amazed they kept the "classic" ipods going as long as they did.
22:49:28Strife89I'm amazed that iTunes still supports them
22:50:08speachysomehow, their replacement for itunes is even more of a dumpster fire in spite of only having a fraciton of the functionality
22:50:11Strife89The nature of tech these days seems to be to not support anything greater than five years old. (Windows' general software compatibility being the exception.)
22:51:15Strife89See: millions of Android devices.
22:51:22speachythe ipod was never intended to operate untethered. it needed itunes to do anything.
22:52:06Strife89I know, but even the newest non-Touch iPods are over ten years old.
22:52:35Strife89And it surely costs Apple money to keep that support in
22:53:41speachyOldApple understood the value of the long tail.
22:54:36speachyNewApple is a phone company that happens to make a few other things.
22:54:57speachy(IMO the only reason anyone cares about Macs any more is because they're required in order to develop for the iOS ecosystem)
22:55:58speachythey've turned into a cargo cult around Jobs' ideas without actually _understainding_ why they mattered
22:57:03speachyin the end the iPods will probably be the last holdout of Rockbox, if only because so darn many were made so there's still a huge secondhand inventory with plenty of replacement parts.
22:57:16speachybut I'm getting all soapbox-y again.
22:58:51mendel_munkisspeachy: remind to see if it's possible to mod one of my old fuze+ to use a non touchpad directional setup.
23:04:23speachywill try
23:07:12Strife89$ at 09:00 \ > echo "Remind mendel" \ ^D
23:08:59speachyand my vivado synth run errored out at the last step 10.5 hours after it started.
23:10:40speachyfailed timing anyway. but the errors that went boom should have come up much earlier.
23:11:41speachythe joys of living on the bleeding edge.
23:12:43TheEaterOfSoulsI have two Clip Zips and Plus(es?) that I grabbed in 2015
23:12:56TheEaterOfSoulsVery expensive, but worth it
23:13:09TheEaterOfSoulsI will continue using Rockbox until they're all dead, including my Clip v1
23:14:10TheEaterOfSoulsOr buy more, if I can find them
23:15:29TheEaterOfSoulsspeachy: why do you think touch interfaces are foolish?
23:15:38speachythe Clips are pretty robust devices. the weakest point is the headphone jack but that's easily repaired.
23:16:11speachythey're not inherently foolish, but instead require a very different approach to user interaction and thus interface design.
23:16:40TheEaterOfSoulsYeah, I have a third Zip that went through the wash twice, I think (I was a teenager, bare with me) the battery swelled so I pried open the case and removed it, and then I couldn't get it back together
23:16:47TheEaterOfSoulsSo it's just the top board
23:16:53TheEaterOfSoulsStill works
23:16:55speachytouch-only interfaces are nearly always unusable without being able to see what you're doing.
23:17:22speachyI should look into getting a spare battery or two for mine.
23:17:53TheEaterOfSoulsYou can use them audibly
23:18:33TheEaterOfSoulsBut IMO the screen readers on phones make things like simple music playback/selection of options less efficient than buttons
23:18:43*speachy nods.
23:19:15TheEaterOfSoulsDon't get me wrong, I'm glad they exist. I just prefer a DAP for listening to music/books/whatever
23:20:13speachydedicated controls that allow you to self-locate
23:20:46TheEaterOfSoulsIf I'm running and want to switch tracks, I don't want to mess with the touch interface
23:21:31TheEaterOfSoulsDo you see it getting better, realistically?
23:22:01TheEaterOfSoulsI kind of feel like the trends of DRM and connectivity and bloat will continue
23:22:30Strife89I'd suggest the real problem with touch interfaces is that most designers simply don't consider the experience for anyone with less knowledge or fewer senses
23:23:00Strife89I feel the same way, TheEaterOfSouls
23:23:17TheEaterOfSoulsI kind of disagree on the first point. User interfaces are trying to appeal to people with absolutely no skill
23:23:37TheEaterOfSoulsIt seems to be the thing to do, if you want it to sell, and ruins it for the rest of us
23:23:51Strife89Music publishers got pressured into selling DRM-free music, so when streaming started to look viable they practically salivated over it
23:24:01TheEaterOfSoulsAs for the second, that's an issue everywhere. People generally don't think/care about people with disabilities
23:24:09speachyyeah. they make more money from streaming than they do "Selling" stuff now.
23:24:32Strife89I've seen that sentiment on Twitter a lot, lately ("people don't care about disabilities", I mean)
23:24:46Strife89Oh God I made a terrible typo
23:24:55Strife89But you know what I meant, I hope
23:25:19speachyif you look at one of the modern DAPs, how much effort was put into the actual music functionality vs the "apps" they threw on there?
23:25:26TheEaterOfSoulsBandcamp still sells DRM-free music, and I love them for that
23:25:30TheEaterOfSoulsAnd Downpour for books
23:25:38TheEaterOfSoulsBoth have awful websites, but I use them
23:25:40Strife89Bandcamp forever!
23:25:48TheEaterOfSoulsYes :D
23:25:51Strife89And GOG, too
23:26:11Strife89Ah, yes
23:28:48speachyDAPs are a niche. The makers need to understand the reasons one would buy one as opposed to using a smartphone.
23:29:13speachyonce you make the DAP into a crappy, less-useful smartphone, why not just stick with the smartphone you already own?
23:29:57speachydon't get me wrong, Rockbox has needed a touch-ish interface since the fuze+
23:30:50speachyie gesture input, etc etc.
23:31:46speachyrockbox is also unusual in that it doesn't tie into some sort of storefront or external music library manager.
23:31:46 Join __Bilgus_ [0] (41ba23be@
23:32:46__Bilgus_speachy apparently yahoo EOL'd their oauth framework any idea how I can switch the provider without being able to login?
23:33:06TheEaterOfSoulsOne of the great things about it
23:34:00speachy__Bilgus_, create a new account with the provider you need and I can manually associate the credentials with your old account.
23:34:37speachyif possible, use the same email address in gerrit.
23:34:49speachyit'll make it easier for me to find things via the sql console.
23:39:28speachyI wonder how many users are affected by this.
23:42:06__Bilgus_speachy does it have to be the same email? I'd rather change that from yahoo as well
23:42:29speachyI guess not. when you're registerd, just tell me what to look for
23:42:57 Quit TheSeven (Ping timeout: 260 seconds)
23:43:16 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
23:45:00speachy15 yahoo emails registered...
23:48:03speachy50 folks registered with yahoo openid credentials.
23:54:35 Quit rogeliodh (Quit: Ping timeout (120 seconds))
23:54:48speachyso I guess I can expect to have to do this several more times, unless a newer version of gerrit fixed it.
23:54:48 Join rogeliodh2 [0] (~rogeliodh@
23:54:49 Nick rogeliodh2 is now known as rogeliodh (~rogeliodh@
23:57:59fs-bluebotBuild Server message: New build round started. Revision 4591510, 295 builds, 11 clients.

Previous day | Next day