00:23:38 | | Quit ac_laptop (Ping timeout: 272 seconds) |
01:00 |
01:50:47 | *** | Saving seen data "./dancer.seen" |
02:00 |
02:47:53 | | Join dys [0] (~dys@2003:5b:203b:100:a64c:c8ff:fef4:13a6) |
03:00 |
03:19:10 | | Join lebellium [0] (~lebellium@89-92-253-148.hfc.dyn.abo.bbox.fr) |
03:35:39 | | Quit koniu (Read error: Connection reset by peer) |
03:35:39 | | Quit livvy (Read error: Connection reset by peer) |
03:39:01 | | Join livvy [0] (~livvy@gateway/tor-sasl/livvy) |
03:48:27 | | Join lebellium_ [0] (~lebellium@89-92-253-148.hfc.dyn.abo.bbox.fr) |
03:50:51 | *** | Saving seen data "./dancer.seen" |
03:52:22 | | Quit lebellium (Ping timeout: 256 seconds) |
03:56:40 | | Join koniu [0] (~koniu@gateway/tor-sasl/koniu) |
04:00 |
04:37:14 | | Quit Jinx` (Ping timeout: 240 seconds) |
04:39:05 | bluebrother | speachy: ok, there are now 2 problems with Rockbox Utility: |
04:39:53 | bluebrother | (1) it doesn't know about this new "retried" status, and will not display it correctly. Shouldn't affect functionality though. At least right now the status info from the download server still indicates stable. |
04:41:39 | bluebrother | (2) the forced upgrade to ssl on download.rockbox.org breaks things for some people. Seems this guy is affected http://forums.rockbox.org/index.php/topic,53326.msg246215.html We had this issue in the past with Windows binaries too. Those didn't have ssl support enabled for license reasons. Back then the forced upgrade to ssl on download.rockbox.org was simply disabled. |
04:52:11 | bluebrother | the main ssl issue here is that Qt uses OpenSSL, and that has a license that was considered ... problematic for static builds. mxe links it statically, so the latest Windows version should simply work. Which raises the question how mxe addresses the license issue ... or if it isn't an issue anymore these days. |
05:00 |
05:30:41 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
05:50:54 | *** | Saving seen data "./dancer.seen" |
07:00 |
07:17:31 | | Join livvy_ [0] (~livvy@gateway/tor-sasl/livvy) |
07:17:42 | | Quit livvy (Remote host closed the connection) |
07:19:11 | | Join livvy [0] (~livvy@gateway/tor-sasl/livvy) |
07:19:22 | | Quit livvy_ (Remote host closed the connection) |
07:25:52 | speachy | bluebrother, I turned off the forced SSL to d.r.o |
07:27:20 | | Quit Acou_Bass (Ping timeout: 272 seconds) |
07:36:57 | speachy | and "retired" is just status=0, and I think rbutil only uses that for an informational tag. |
07:39:19 | speachy | Nothing has generated a new release build-info file yet, so that hasn't gone anywhere.. |
07:43:10 | fs-bluebot_ | Build Server message: New build round started. Revision 74f358c, 292 builds, 10 clients. |
07:45:02 | Bilgus | Ok after this build round completes there will now be a status item in debug>talk engine stats |
07:45:35 | | Join Acou_Bass [0] (~eddie@cpc97736-bolt17-2-0-cust152.10-3.cable.virginm.net) |
07:46:10 | Bilgus | logf will still need enabled for anything more than just a coarse error but better than nothing |
07:47:01 | Bilgus | It recognizes OOM, Alloc, incompatible header / ver |
07:50:57 | *** | Saving seen data "./dancer.seen" |
07:52:53 | speachy | oh that will be useful |
07:55:41 | | Join massiveH [0] (~massiveH@ool-18e4eaeb.dyn.optonline.net) |
07:57:15 | Bilgus | tis the hope :P |
07:57:36 | speachy | bluebrother, g#2382 is an attempt to make rbutil handle "retired" statuses. |
07:57:38 | fs-bluebot_ | Gerrit review #2382 at http://gerrit.rockbox.org/r/2382 : rbutil: Handle the new 'Retired' target status. by Solomon Peachy |
08:00 |
08:03:43 | Bilgus | it there a way to get the Qt packages to compile that doesn't require jumping through hoops & giving them all my info? |
08:03:58 | speachy | "them"? |
08:04:07 | Bilgus | Qt |
08:04:56 | speachy | download the source/binaries from a 3rd party? |
08:05:39 | speachy | doing this for osx or windows, I assume? |
08:06:36 | Bilgus | I hate doing any programming in windows it'll be linux |
08:06:46 | fs-bluebot_ | Build Server message: Build round completed after 1415 seconds. |
08:06:55 | speachy | under Linux the qt stuff tends to be already packaged by the distro. |
08:07:17 | Bilgus | ah hopefully that the case then |
08:08:48 | speachy | ok, looks like homepc-petur had a little hiccup on a build. |
08:09:43 | Bilgus | o_O |
08:10:32 | speachy | that was a larger binary bump than I expeced from that change.. |
08:11:20 | speachy | and it's strange how some targets never seem to get a binary size change either. |
08:11:55 | speachy | the x3ii changed, but rocker and x20 didn't. That makes no sense to me. |
08:12:22 | Bilgus | agreed I'll try and compress that a bit |
08:15:03 | Bilgus | just having it as an error # will knock out a bit a little more opaque but OOM and alloc errors weren't spelled out either |
08:15:24 | speachy | and what, the strings aren't added to the lang files? :D |
08:16:13 | Bilgus | never do in the debug menu |
08:16:25 | Bilgus | its english only |
08:16:29 | * | speachy nods. |
08:16:57 | speachy | just thinking of an actual blind user, how will they know the voice is busted? |
08:17:04 | speachy | besides it not working, I mean. |
08:17:15 | speachy | it's a bit of a chicken-n-egg problem |
08:19:25 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
08:20:19 | Bilgus | IKR! I debated about that for a bit |
08:22:08 | Bilgus | I thought about making it say something or beeping but at least theres some indication this way even if they need a sighted user to tell them |
08:22:47 | speachy | there's a good argument for making the "voice failed to load" a splash message. |
08:24:53 | Bilgus | then it needs to be a localized string to be actually useful and bigger binary bump |
08:27:44 | Bilgus | lets see if there is a sutible string already in lang |
08:28:46 | Bilgus | LANG_READ_FAILED |
08:30:07 | speachy | kinda generic but it's better than having to go into the debug menu to find out why voice stopped working... |
08:30:15 | Bilgus | that sounds good "Failed reading %s", lang.voice |
08:34:12 | Bilgus | its going to depend on how early the lang file is loaded if this works or not |
08:34:23 | Bilgus | rather the voice file |
08:34:30 | speachy | yea. |
08:40:19 | | Join MrZeus [0] (~MrZeus@89-168-118-53.dynamic.dsl.as9105.com) |
08:50:20 | Bilgus | Works ok just going to be 'Failed Reading .voice' |
08:55:31 | fs-bluebot_ | Build Server message: New build round started. Revision dcf8154, 292 builds, 10 clients. |
08:56:08 | Bilgus | comon petur 'you can doit' |
08:58:48 | speachy | sweet, the rework to make the build.r.o site static succeeded |
08:59:18 | speachy | it's now regenerated upon completion of a build round. |
09:00 |
09:00:06 | Bilgus | oh instead of constantly refreshing? |
09:00:50 | speachy | no, the dev.cgi "autobuilder" page was already static |
09:01:08 | speachy | this is the main build page, with the links to the last successful builds for each target |
09:01:32 | Bilgus | ah the dev builds page |
09:01:41 | speachy | to display the git revision for each, it was opening each build's zip file and parsing out rockbox-info.txt |
09:01:59 | speachy | so what, 50-odd zipfiles on eacn and every page load... |
09:02:04 | Bilgus | ick |
09:02:35 | speachy | completely unnecessary. :) |
09:02:44 | Bilgus | wow thats a lot of overhead at several 1000 views |
09:04:16 | speachy | yeah, I was alway irked at how slow the build.r.o pages loaded. Much better now. :) |
09:04:44 | | Join petur [0] (~petur@rockbox/developer/petur) |
09:04:51 | Bilgus | waiting on builds to complete makes me sing that Tom Petty song ; The waiting is the hardest part |
09:06:13 | Bilgus | I still run them all locally on the really big stuff but that takes like 2 hours+ |
09:07:04 | Bilgus | 20-30 mins much preferable :D |
09:08:01 | speachy | so I take it the voice load happens late enough for the slash to be useful? |
09:09:18 | speachy | since I have buildmaster under teh same roof now, I was also able to eliminate some of the cron jobs I set up to keep a lot of the other stuff synchronized. |
09:09:31 | Bilgus | yeah its real early but it pops just fter the fw title splash |
09:09:37 | speachy | excellent |
09:11:33 | Bilgus | it definitely beats no indication whatsoever which I should have thought of when I did the original patch but I was more concerned with the crash than the non working voice |
09:11:36 | | Quit massiveH (Quit: Leaving) |
09:12:47 | speachy | such is the joy of F/OSS, drunkenly stumbling towards perfection |
09:13:37 | Bilgus | BTW speachy your take over of the site very bueno it was def in need of some TLC which i'm sure HAXX guys realized they just weren't able to provide |
09:13:52 | speachy | now, next question. is this something that could/should be documented? |
09:14:13 | speachy | in the voice section of the manual, mention that there are compatibility tests and if they fail, it'll splash.. |
09:14:25 | Bilgus | IDK ill look in the voice part of the manual |
09:14:55 | speachy | I was frustrated by the state of rockbox.org site/infra long before I ever thought taking over could be an option |
09:15:14 | fs-bluebot_ | Build Server message: Build round completed after 1183 seconds. |
09:15:16 | speachy | so many "pebbles in the shoe" sorts of problems. |
09:16:22 | Bilgus | hmm no all green / all teh errors message? |
09:16:24 | speachy | wtf.. the autobuilder's h120 manual is tied to the sim build, not the normal build. |
09:16:37 | speachy | the bot isn't picking up errors for some reason. |
09:16:55 | Bilgus | well save a bit of space I was hoping for more like -100 |
09:16:58 | speachy | or rather, the buildmaster isn't. |
09:21:38 | Bilgus | speachy did you add voice to rtc and the wake alarm? |
09:22:07 | speachy | I don't recall. |
09:22:36 | Bilgus | just noticed the voice section mentions plugins RTC and Alarm don't support voice I'll change that to Some plugins |
09:22:52 | speachy | there are still a lot of voice-enabling patches out of poretsky's tree left to sort out. |
09:23:30 | speachy | but it doesn't appear that the alarclock is voice-enabled. |
09:25:03 | speachy | ...there's still an outstanding patch that adds voice to the alarm timer, but nothing for the alarm clock plugin itself. |
09:25:49 | speachy | it's probably more accurate to say that "most plugins do not currently support voice" |
09:26:19 | speachy | a whole bunch do now but... there are a lot of plugins. |
09:28:59 | | Join vmx [0] (~vmx@ip5f5ac60f.dynamic.kabel-deutschland.de) |
09:32:50 | | Quit fauweh (Ping timeout: 246 seconds) |
09:33:50 | | Quit craftyguy (Remote host closed the connection) |
09:34:41 | | Join craftyguy [0] (~craftyguy@qutebrowser/dev/craftyguy) |
09:36:13 | | Quit CommunistWitchDr (Ping timeout: 265 seconds) |
09:37:45 | | Join CommunistWitchDr [0] (quassel@024-217-039-226.res.spectrum.com) |
09:42:30 | Bilgus | g#2384 |
09:42:40 | fs-bluebot_ | Gerrit review #2384 at http://gerrit.rockbox.org/r/2384 : talk.h Add manual entry by William Wilgus |
09:43:18 | speachy | buildserver generated the result, but got a sigpipe that resulted in all clients getting dropped.. |
09:44:18 | speachy | so perhaps one of the clients isn't behaving properly, blocked on the message, and the sigpipe handling force-booted all connected clients. |
09:45:05 | speachy | looks like it's probably deepthought-ender. |
09:45:18 | Bilgus | you have something you want to do before I push this? |
09:48:05 | speachy | nah, go for it. don't know what's going on with the results logging but there weren't any server-side changes. |
09:48:16 | speachy | data's being properly captured |
09:51:00 | *** | Saving seen data "./dancer.seen" |
09:52:16 | fs-bluebot_ | Build Server message: New build round started. Revision ce29bec, 292 builds, 10 clients. |
09:58:31 | Bilgus | be interesting to see if the bin size still jumps around it always seemed to kinda jump up and down randomly maybe based on which client builds it |
09:59:30 | speachy | I don't expect any change in that regard. I may have changed how the data gets stored/visualized but the instability is the source. |
10:00 |
10:00:56 | speachy | probably due to randomness in object linking order. I bet it would go away if we stopped allowing the builders to parallelize their builds. |
10:04:25 | Bilgus | makes sense |
10:09:30 | fs-bluebot_ | Build Server message: Build round completed after 1034 seconds. |
10:13:14 | Bilgus | nope all bin deltas are +- 0 :) |
10:16:23 | speachy | hmm, need to investigate how "avg ul speed" and "round ul speed" are computed |
10:20:44 | | Join fauweh [0] (~root@ithaqua.unzane.com) |
10:29:41 | | Quit Moarc (Quit: i znowu NADMUCHAĆ BALONA) |
10:32:00 | | Join Moarc [0] (~chujko@a105.net128.okay.pl) |
11:00 |
11:08:36 | | Join johnb4 [0] (~johnb2@p5b3af381.dip0.t-ipconnect.de) |
11:13:25 | | Quit johnb4 (Client Quit) |
11:17:17 | | Quit dys (Ping timeout: 260 seconds) |
11:51:02 | *** | Saving seen data "./dancer.seen" |
11:54:32 | | Quit ufdm (Remote host closed the connection) |
11:55:15 | | Join ufdm [0] (~ufdm@c-73-164-63-214.hsd1.mn.comcast.net) |
12:00 |
12:02:16 | bluebrother | speachy: an incompatible voice file could make the player play a couple of beeps on startup. That would be recognizable even if the user cannot see the screen (for whatever reason) |
12:04:48 | bluebrother | speachy: configuring a manual build as sim build has the nice side effect that it also builds if you don't have the cross compilers installed. |
12:05:32 | bluebrother | hmm, this cryptopp thing is really a pain. Finally got Rockbox Utility build on OS X again but now it doesn't link due to cryptopp, and that having a different target SDK. |
12:10:47 | | Join sakax [0] (~r0b0t@unaffiliated/r0b0t) |
12:15:25 | Bilgus | bluebrother, I too thought of a beep but figured it was too cryptic but maybe along with the message it would be ok |
12:22:54 | Bilgus | ah nevermind its too early to play a beep |
12:23:37 | bluebrother | another option: Add a pre-rendered file to .rockbox and play that. |
12:23:44 | bluebrother | hmm, playback not ready at that point? |
12:25:41 | Bilgus | yeah IDK I think its good enough at least it doesn't crash anymore |
12:25:59 | bluebrother | :) |
12:36:52 | * | bluebrother wonders if there is a simple implementation that could be used to replace libcryptopp |
12:50:34 | | Join krabador [0] (~krabador@unaffiliated/krabador) |
12:57:53 | Bilgus | depends on how you define simple :p |
13:00 |
13:04:49 | | Quit krabador (Remote host closed the connection) |
13:06:10 | bluebrother | well, simple enough so that it can be easily included in the Rockbox tools, or if it's an external library used without much hassle. And portable. |
13:06:29 | bluebrother | cryptopp is nice on Linux. But on all other OSes it's a nightmare, at least for our setup. |
13:07:09 | bluebrother | finally managed to get somewhere on OS X again, but it fails spectacularly when linking with libcryptopp. |
13:07:38 | bluebrother | and on Windows we have a problem as well, since it's nothing we can expect to be present system-wide. |
13:08:01 | bluebrother | and doing all this on our own is ... additional work I'm seeing less and less benefit in. |
13:08:12 | bluebrother | maybe simply go back to the old implementation. |
13:09:37 | bluebrother | cryptopp is pretty huge, and as far as I can see we only use it for aes encoding / decoding. |
13:10:55 | Bilgus | yeah agreed on the AES part looking through the source but TBH I don't remeber what the old way was |
13:12:45 | Bilgus | https://github.com/kokke/tiny-AES-c |
13:13:00 | bluebrother | crap, nwzboot also uses it. |
13:13:26 | bluebrother | for md5 calculation. |
13:16:11 | bluebrother | old way = our own implementation. |
13:19:00 | bluebrother | fs-bluebot_: .describe 8b3f5a8ad74 |
13:19:01 | fs-bluebot_ | imxtools/sbtools: switch AES implementation to Crypto++ by Amaury Pouly |
13:19:08 | bluebrother | that's when cryptopp came in |
13:24:55 | Bilgus | g#1452 yeah quite a while ago |
13:24:57 | fs-bluebot_ | Gerrit review #1452 at http://gerrit.rockbox.org/r/1452 : imxtools/sbtools: switch AES implementation to Crypto++ by Amaury Pouly |
13:51:06 | *** | Saving seen data "./dancer.seen" |
14:00 |
14:52:43 | | Quit koniu (Ping timeout: 240 seconds) |
14:55:44 | | Join koniu [0] (~koniu@gateway/tor-sasl/koniu) |
15:00 |
15:26:57 | pamaury | bluebrother: Bilgus: just noticed that you were ranting about crypto++ ;) The main reason I switched is that the entired imx image is encrypted with AES, that's pretty large and the self-contained implementation was really slow |
15:28:26 | pamaury | so my advise before you make any switch is to at least benchmark it |
15:50:44 | bluebrother | pamaury: it's not crypto++ per se, it's using it in our applications on non-Linux systems. On Linux it's simple, you use pkg-config to find it and things are well. On Windows ... not a chance, things get complicated. At least mxe provides a package for crypto++, so one can use that to cross-compile a Windows binary and avoid the hassle. |
15:51:07 | *** | Saving seen data "./dancer.seen" |
15:51:54 | bluebrother | I've decided to try looking into creating a recent binary for macOS. On that things become even more convoluted. Build crypto++, build Rockbox Utility, try to link, fails. Because it uses different SDK / libc++ :/ |
15:52:38 | | Join jdarnley [0] (~J_Darnley@d51A44418.access.telenet.be) |
15:53:12 | bluebrother | (and that's after figuring how to install XCode in 2020 ... on 10.11 (my hardware isn't supported by a newer macOS version) you cannot install XCode anymore due to an expired certificate. Had to set the clock back to 2019 :o ) |
15:54:43 | | Quit J_Darnley (Ping timeout: 260 seconds) |
16:00 |
16:25:18 | | Quit vmx (Quit: Leaving) |
16:45:40 | blbro[m] | speachy: themes.r.o doesn't work with ssl - I get a website, but not the themes one. |
17:00 |
17:11:26 | pamaury | bluebrother: yes I understand the problem, it's crypto++ is particularly tricky though |
17:33:13 | | Quit TorC (Ping timeout: 264 seconds) |
17:46:48 | | Join krabador [0] (~krabador@unaffiliated/krabador) |
17:50:20 | | Join TorC [0] (~Tor@fsf/member/TorC) |
17:51:10 | *** | Saving seen data "./dancer.seen" |
18:00 |
18:08:59 | speachy | there are plenty of standalone aes implemenmtations |
18:09:56 | speachy | if nothing else, libtom.. |
18:10:17 | speachy | but yes, fixing themes.r.o and translate.r.o is on my list for this weekend/next week |
18:10:29 | speachy | with respect to ssl |
18:10:58 | speachy | thought the links off the main page were not fixed to http:// not https but maybe I accidentally undid that |
18:12:26 | speachy | s/not //; |
18:20:04 | blbro[m] | I typed the url myself :) |
18:21:34 | blbro[m] | Actually I tried to change some urls in Rockbox Utility to https when I stumbled across that. Wasn't aware this is a known issue :) |
18:23:44 | | Quit petur (Quit: Leaving) |
18:45:44 | | Join Marqueteur [0] (~Tor@fsf/member/TorC) |
18:45:44 | | Nick TorC is now known as Guest55252 (~Tor@fsf/member/TorC) |
18:45:44 | | Nick Marqueteur is now known as TorC (~Tor@fsf/member/TorC) |
18:52:25 | | Quit Guest55252 (Read error: Connection reset by peer) |
18:52:25 | | Quit TorC (Read error: Connection reset by peer) |
18:52:44 | | Join Marqueteur [0] (~Tor@fsf/member/TorC) |
18:52:44 | | Nick Marqueteur is now known as TorC (~Tor@fsf/member/TorC) |
18:53:11 | | Nick TorC is now known as Guest58576 (~Tor@fsf/member/TorC) |
18:53:13 | | Join TorC [0] (~Tor@fsf/member/TorC) |
18:53:33 | | Quit TorC (Client Quit) |
18:54:29 | | Quit lebellium_ (Quit: Leaving) |
19:00 |
19:03:54 | | Nick Guest58576 is now known as TorC (~Tor@fsf/member/TorC) |
19:17:28 | | Quit pamaury (Quit: this->disconnect()) |
19:19:13 | | Quit krabador (Read error: Connection reset by peer) |
19:33:39 | | Join krabador [0] (~krabador@unaffiliated/krabador) |
19:40:37 | | Join massiveH [0] (~massiveH@ool-18e4eaeb.dyn.optonline.net) |
19:49:18 | | Join skolem [0] (~skolem@47.61.119.15) |
19:51:15 | *** | Saving seen data "./dancer.seen" |
19:57:06 | | Quit skolem (Quit: WeeChat 2.8) |
20:00 |
20:05:19 | | Quit livvy (Remote host closed the connection) |
20:17:21 | | Join livvy [0] (~livvy@gateway/tor-sasl/livvy) |
21:00 |
21:23:11 | | Join advcomp2019_ [0] (~advcomp20@63-229-188-49.sxct.qwest.net) |
21:23:11 | | Quit advcomp2019_ (Changing host) |
21:23:11 | | Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019) |
21:26:43 | | Quit advcomp2019 (Ping timeout: 265 seconds) |
21:27:48 | | Quit krabador (Remote host closed the connection) |
21:34:59 | | Quit MrZeus (Ping timeout: 258 seconds) |
21:51:17 | *** | Saving seen data "./dancer.seen" |
23:00 |
23:32:34 | | Quit ac_laptop (Ping timeout: 246 seconds) |
23:51:18 | *** | Saving seen data "./dancer.seen" |
23:54:08 | | Quit beencubed (Ping timeout: 265 seconds) |