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 2015-01-11

00:03:38 Quit xorly (Ping timeout: 264 seconds)
00:09:15 Join xorly [0] (
00:15:52 Quit y4n (Quit: MOTHER EUROPA CALLING ME!)
00:24:21 Quit lorenzo92 (Ping timeout: 240 seconds)
00:36:29***Saving seen data "./dancer.seen"
00:40:56pamaurywodz (logs): I'm in the process of porting qeditor, see g#1131, at the moment it compiles (but I excluded the edit tab from compilation) and I can browse registers, I haven't tried reading some values yet but it's a good start :)
00:41:01fs-bluebot3Gerrit review #1131 at : 3regtoosl/qeditor: port to the new description format by Amaury Pouly
00:47:29 Quit pamaury (Ping timeout: 244 seconds)
00:59:41 Quit thomasjfox (Quit: Konversation terminated!)
01:07:22 Quit bertrik (Remote host closed the connection)
01:09:02 Quit Rower (Ping timeout: 264 seconds)
01:33:02TheSevenhm, not quite execute code on n5g, but rather bypass sigchecks on n7g or whatever the current one is called
01:33:22TheSeventhose sigchecks don't directly allow code execution, but open up a lot of additional attack surface
01:33:39TheSevenMarcAndersen: ^
01:47:58[Saint]Likely confusing 4 and 5G, I posit.
01:48:22[Saint]Where there is execution, but no access to storage, so its a very limited playground.
01:50:04 Join AlexP [0] (~alex@rockbox/staff/AlexP)
02:08:07 Quit xorly (Ping timeout: 245 seconds)
02:18:26 Quit ender` (Quit: It's nice to know that my launch to orbit won't have any pesky back-up systems weighing me down.-- Andy Weir: The Martian)
02:23:58 Quit AlexP (Remote host closed the connection)
02:34:51 Quit lebellium (Quit: ChatZilla [Firefox 35.0/20150106233618])
02:36:31***Saving seen data "./dancer.seen"
02:48:52 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
02:55:44 Join lunarl0n [0] (
03:03:59 Part lunarl0n
03:14:02 Quit ZincAlloy (Quit: Leaving.)
03:29:01 Join GreyGhost [0] (3bb71761@gateway/web/freenode/ip.
03:30:39 Quit krabador (Quit: Take the time.)
03:30:59 Quit GreyGhost (Client Quit)
03:32:23 Join GreyGhost [0] (3bb71761@gateway/web/freenode/ip.
03:34:36 Join krabador [0] (~krabador_@unaffiliated/krabador)
03:34:37GreyGhost[Saint], Hello.. I was wondering if the ipod 6g usb driver rewrite for the rb toolchain is in need of testers or does it need someone to fix it? Cos I could volunteer for the former :P
03:35:00[Saint]I needs no testing. No.
03:35:16[Saint]It needs to be completely re-written.
03:35:56TheSevenwodz fixed the bitfield stuff etc. a while ago, but apparently broke something else in the process
03:35:59TheSeventhat needs a closer look
03:36:29GreyGhost[Saint)], ok then.. Can't help there..
03:37:26MarcAndersenIs there any hope for rockbox on the nano 5g?
03:38:02GreyGhostTheSeven, btw.. Using the classic as OTG is working great now.. Copied over some 2.5 gb of data without freeze. With the old driver it froze even on a few hundred mbs..
03:38:23[Saint]MarcAndersen: I'll say this much, no one is working on it.
03:38:32[Saint]MarcAndersen: so, for the forseeable
03:38:42*GreyGhost is TheNewGuy whom you gave the patched rockbox amd emcore copy
03:38:46MarcAndersendo we need to take one apart?
03:39:12MarcAndersenok. so it's software problems?
03:39:14TheSevenGreyGhost: that's a good data point... now if only wodz's version would work...
03:39:39TheSevenMarcAndersen: nobody managed to defeat apple's mechanisms that are trying to prevent us from running any code on that device yet
03:40:02[Saint]MarcAndersen: its a "no viable or confirmwd exploit and no one who can do the work who actually cares or owns the device" problem.
03:40:09TheSeven(unlike some older generations, where we successfully managed to circumvent those)
03:40:13TheSevenit's kinda like iphone jailbreaking
03:40:31[Saint]These new iPods are more like iPhones than the older generation iPods.
03:40:44[Saint]bah - beat me to it again. Damn you 7! :p
03:41:34GreyGhostTheSeven, yeah.. That's why I came by to see if I can offer any help.. Sadly all I could do was test (so not much real help :P)
03:41:40[Saint]I used to care.
03:41:50MarcAndersenI would like to help if I can, but don't know where to start at this pone...
03:41:52[Saint]But most, if not all, of my development time now goes towards Android.
03:42:06[Saint]I mean, honestly, Android handsets are a LOT more interesting.
03:42:12[Saint]To me at least.
03:42:37[Saint]I could bust my ass working on a DAP that _does not want_ me inside it.
03:42:45[Saint]Or work on a largely free and open platform that does.
03:42:46MarcAndersenandroid is not as accessible to the blind as button based playiers I think
03:42:51[Saint]pretty easy choice.
03:43:20[Saint]MarcAndersen: that's largely untrue. Android is very highly usable with Google's TalkBack engine, no?
03:43:40[Saint]I've witnessed totally unsighted people use Android many a time.
03:43:51[Saint]even as power-users.
03:44:04[Saint]but, anyhoo - topic, sorry to devolve the conversation.
03:44:17GreyGhostAnyway I'll continue to lurk around the IRC logs and pop in of i see any driver related activity.. Thanks..
03:44:19MarcAndersenSure. But I still think running rockbox on a full operating system is a bit strange
03:44:51 Quit the-kyle (Remote host closed the connection)
03:45:23[Saint]MarcAndersen: aha, right, yeah I agree somewhat. My Android development involves practically zero Rockbox work these days.
03:45:46[Saint]I got sick of trying to shoot a moving target with the ART runtime, and then got bored with it.
03:45:50[Saint]I should pick it back up.
03:46:23[Saint]Rockbox being stuck on 4.4.4 means I have to deliberate NOT update at least one of my devices, which is a bit of a PITA.
03:46:30MarcAndersenBut if there is a rockbox for android, why not make it for ios and put it in cydia?
03:46:42[Saint]I got it to not crash *immediately* on 5.x...but I got bored and stopped.
03:46:57[Saint]MarcAndersen: because none of us care about iOS.
03:47:02[Saint]If you do, go for it.
03:47:17MarcAndersenhahaha I use symbian
03:47:18[Saint]We work for ourselves, not the public.
03:47:30[Saint]If the public gets a kick out of it, great, but its the developers project.
03:48:22[Saint]We don't align what is worked on with popular public opinion.
03:48:31[Saint]The supported device list should demonstrate that.
03:48:59MarcAndersen[Saint]: That's a good point
03:49:47MarcAndersenCan I get rockbox for android in the play store_
03:50:56[Saint]I only stick around these days tohandle support really.
03:51:07[Saint]My Rockbox development time is next to nil.
03:51:32[Saint]I work on what takes my interest at the time, and for better or worse, that hasn't been Rockbox for a while now.
03:52:00[Saint]MarcAndersen: due to the way we handle builds, no, no you can't.
03:52:17[Saint]you /could/, but we'd need to upload a build for _every_ supported resolution.
03:52:22[Saint]...which is plainly stupid.
03:52:53[Saint]there's no "one size fits all" build, and no clear and/or sane path to get there.
03:53:34[Saint]Nothing anyone can agree on at least. And a lack of anyone to do it if there were.
03:54:49MarcAndersen[Saint]: ok
03:55:15[Saint]I understand that's not what you wanted to hear, but that's the state of things as I understand it.
03:55:37MarcAndersenI only run android in virtual box at the moment
03:56:29 Join CaptainKewl [0] (
04:07:43 Quit GreyGhost (Ping timeout: 246 seconds)
04:27:13 Join ploco [0] (dce9b7f9@gateway/web/freenode/ip.
04:29:34ploco[Saint]: one Chinese user report back a non-crash case with 5.0 when using my old build which disable the power thread....but play will cause a immediate FC
04:30:17[Saint]I have a build that dies after 3 or 4 seconds. Instead of immediately.
04:30:18plocothat leads to two things. 1 notification handle, 2 Android threads.
04:30:20[Saint]...progress? ;)
04:30:56plocoI got nothing to debug. waiting for a CM12 port
04:31:25[Saint]Ah, that's right, you're on some legacy device yeah?
04:31:47plocoyes, st18i really old
04:32:32[Saint]Sadly, you'll never see an official cm-12 port either.
04:32:44[Saint]there's hope from the community, though, maybe.
04:33:47plocoactually I don't care, I could just use a Android 5.0 VM and debug from there. might do it when I have a little bit more time next month
04:36:35***Saving seen data "./dancer.seen"
04:43:52 Quit ploco (Quit: Page closed)
04:47:10 Join djruffkutz [0] (
04:47:35 Quit djruffkutz (Excess Flood)
05:11:31 Quit TheSeven (Ping timeout: 244 seconds)
05:12:40 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:28:52 Quit mshathlonxp (Ping timeout: 256 seconds)
05:36:12 Quit krabador (Quit: Take the time.)
06:36:39***Saving seen data "./dancer.seen"
06:44:55 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
07:33:42foolsh[Saint]: woohoo! I took a sample and did a count, 0 adults and two eggs, and the aggs look brown and shriveled, I wonder if the mixture I'm using inhibits the reproductive cycle?
07:34:02*foolsh is happier today
07:34:27[Saint]I don't think you're where you think you are.
07:34:34foolshlol opps :)
08:35:17 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
08:35:52 Quit JdGordon (Ping timeout: 256 seconds)
08:36:41***Saving seen data "./dancer.seen"
08:56:33 Join lorenzo92 [0] (
09:07:33 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
09:33:39 Quit rela (Quit: Leaving)
09:49:46 Quit bertrik (Remote host closed the connection)
10:19:10 Quit bzed (Remote host closed the connection)
10:36:43***Saving seen data "./dancer.seen"
10:42:02 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:16:54 Join MMlosh [0] (~MMlosh@2001:470:6f:23:24f:63ff:fe01:4900)
11:25:19 Join ender` [0] (
11:40:53 Join xorly [0] (
11:47:48 Quit lorenzo92 (Ping timeout: 244 seconds)
11:53:57 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
12:01:28 Join Babydylanw02 [0] (44697471@gateway/web/freenode/ip.
12:05:03 Quit Babydylanw02 (Client Quit)
12:14:06 Quit JanC (Ping timeout: 264 seconds)
12:19:48 Join lorenzo92 [0] (
12:23:09 Quit APLU (Quit: !sucide)
12:23:11pamaurywodz (logs): qeditor is working with hwstub and the new register desc, see the gerrit task ^^
12:27:20 Join JanC [0] (~janc@lugwv/member/JanC)
12:32:12 Join APLU [0] (
12:36:45***Saving seen data "./dancer.seen"
13:18:43 Join lebellium [0] (
13:20:06 Join TheLemonMan [0] (~lemonboy@unaffiliated/thelemonman)
13:45:53lorenzo92logf("logf") calls are opted out when not compiling with logf support, right?
13:47:32 Join RiD [0] (
13:50:56lorenzo92i.e. the string is not even in the final executable?
14:36:46***No seen item changed, no save performed.
15:48:34lorenzo92bluetooth implementation so far is going fine, but slowly ^^ anyway, rockbox is now able to initialize all the hardware, and specifically the bcm20xx with some vendor specific commands. plus a debug screen to toggle the basic controller state (pin-level)
15:49:13lorenzo92next, i would like to fix "firmware blob" update aka patchram. this one is not essential, but OF does provide one
16:06:02 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
16:07:18 Quit JdGordon_ (Ping timeout: 245 seconds)
16:11:31 Quit lorenzo92 (Ping timeout: 245 seconds)
16:15:03 Quit pystar89 (Ping timeout: 264 seconds)
16:21:02 Quit Zambezi (Ping timeout: 264 seconds)
16:21:46 Join Zambezi [0] (
16:22:15 Quit Galois (Ping timeout: 264 seconds)
16:36:47***Saving seen data "./dancer.seen"
16:42:24 Quit TheLemonMan (Quit: leaving)
16:57:43 Join Galois [0] (
17:15:36 Join ZincAlloy [0] (
18:02:09 Join AlexP [0] (~alex@rockbox/staff/AlexP)
18:03:23fs-bluebotBuild Server message: 3New build round started. Revision cfbd9cb, 255 builds, 25 clients.
18:05:31*gevaerts waits excitedly to see if everything is still green!
18:07:12thomasjfoxyou are not alone ;)
18:07:30gevaertsWe probably have different reasons though
18:07:38thomasjfoxthough I did grep those variables before modifying them...
18:07:54gevaertsIf everything goes as expected, fs-bluebot should now properly report yellows and reds
18:08:05thomasjfoxall of them?
18:08:32thomasjfoxsometimes the SDL app outputs over 120 yellow warnings. It looks like the compiler settings sometimes change...
18:08:35gevaertsWell, a total. The same data the CIA bot used to give us
18:09:00 Join bzed [0] (
18:09:21gevaertsWe'll see :)
18:11:22thomasjfoxCIA bot retired when the switch from svn to git was done?
18:11:33fs-bluebotBuild Server message: 3Build round completed after 490 seconds.
18:11:34 Join [Franklin] [0] (~franklin@unaffiliated/franklin)
18:11:39gevaertsNo. The server that ran on had a disk accident, and they lost all data
18:12:05gevaertsSo the people running that service decided not to bother restarting it
18:12:11 Quit bzed (Remote host closed the connection)
18:12:40thomasjfoxdoes fs-bluebot only output a diff in yellow/red or should it report the two reds for the "Recorder" build?
18:13:05gevaertsI'm confused!
18:14:01bluebrotherfs-bluebot outputs exactly what the build server reports. Since that report was wrong in the past its filtered out
18:14:14gevaertsRight :)
18:14:15bluebrotherI'll change that. Judging from the log it should be ok now
18:14:22gevaertsAnyway, my client didn't get the line...
18:14:29gevaertsThat's mostly what confuses me
18:14:45thomasjfoxI can do more static stuff if a commit is needed ;)
18:15:04gevaertsAnyway, yes, it should be fixed now. This is the first build with the fix though
18:16:37 Join pystar89 [0] (
18:17:48thomasjfoxbluebrother: let me know when the filter is removed and I'll commit g#1119
18:17:52fs-bluebot3Gerrit review #1119 at : 3Improve readability by adding parenthesis by Thomas Jarosch
18:18:39bluebrotherthomasjfox: should work now
18:18:52[Franklin]thomasjfox: have you looked at G#887?
18:18:52 Join bzed [0] (
18:18:56fs-bluebot3Gerrit review #887 at : 3Enhancement of the metronome plugin: by Thomas Orgis
18:19:02thomasjfox[Franklin]: no
18:20:08fs-bluebotBuild Server message: 3New build round started. Revision 85c98bc, 255 builds, 25 clients.
18:22:32 Quit bzed (Remote host closed the connection)
18:26:59fs-bluebotBuild Server message: 3Build round completed after 412 seconds.
18:27:00fs-bluebotBuild Server message: 3Revision 85c98bc result: 2 errors 0 warnings
18:27:15[Franklin]finally :D
18:27:33pixelmais someone running a fairly recent build on an Iaudio X5/M5 or maybe M3? I'm getting an STKOV voice (when enabled and voice file present) quite often with a build from beginning of last year or so and I fear it just got worse
18:28:03thomasjfoxhmm, would it be possible to mask the two errors for the Recorder build?
18:28:20[Franklin]or better yet, fix it? :P
18:28:29gevaerts[Franklin]: go ahead!
18:28:48gevaertsI'm wondering if it makes sense at all to keep trying to auto-build it
18:28:54[Franklin]perhaps a better compressor would do
18:29:47gevaertsIt's similar enough to recorderv2 and fmrecorder for breakages to slip through, I'd say
18:31:01*[Franklin] seems to have "xracer" working
18:31:12[Franklin]just that the poly fill function is not working :(
18:31:32gevaerts[Franklin]: that was also what you were missing for sgt-puzzles, wasn't it? :)
18:31:51[Franklin]but this has a set number of points
18:31:55[Franklin]4 max
18:32:15[Franklin]so just 2 tris
18:32:28[Franklin]but it's the ARGUMENTS to that function, to be more precise
18:32:35 Join einhirn [0] (
18:35:24bluebrothergevaerts: \o/
18:36:17bluebrothermasking them is problematic, at least for the bot: it doesn't know those are the recorder errors, it just mirrors the result it gets from the build master
18:36:37[Franklin]just subtract two ;)
18:36:46bluebrotherand simply extracting the number and substracting 2 is ... well.
18:36:49***Saving seen data "./dancer.seen"
18:37:00 Quit einhirn (Ping timeout: 264 seconds)
18:37:06[Franklin]sed 's/2 errors/0 errors/'
18:37:31bluebrotherplus, that means the bot had even more dependencies on the format of the messages the build master sends. Nothing I'd like to have there.
18:38:00bluebrother[Franklin]: and if it reports 3 errors because there is another one? Won't work that way :)
18:38:30[Franklin]when there's an error, there's a big number
18:39:04[Franklin]bluebrother: BTW, where does fs-bluebot get the messages from?
18:39:29bluebrother[Franklin]: from the build server. It interacts with a build client running on the same machine.
18:39:33thomasjfoxprobably it's best to say good bye to the Record build or ask one last time on the mailinglist if someone (with the hardware) is willing to fix it
18:40:01[Franklin]just "deprecate" it
18:40:15bluebrotherthomasjfox: as much as I hate to say so but I do agree with that. It simply doesn't make much sense anymore :(
18:40:32thomasjfoxgevaerts already asked on the ML :D
18:41:45thomasjfoxhe should get a hug for that
18:42:57 Join bzed [0] (
18:43:27bluebrotherlike that?
18:44:09thomasjfoxexactly like that :D
19:00:22 Quit d33tah (Quit: WeeChat 1.0.1 - the better Irssi! [that's not the default quit msg])
19:00:29 Join d33tah [0] (
19:02:20[Franklin]foolsh: take a look at "xracer" now
19:04:18[Franklin]barely working, but kind of working
19:05:38foolshah nice, let me see what it looks like on a target
19:06:29[Franklin]wait... I got alternating colors working now
19:06:33[Franklin]let me push first
19:08:39 Join Strife89 [0] (
19:16:20 Join Water255 [0] (~chatzilla@
19:16:28foolshI see where you're going with, neat, heh some z depth culling issues on the sim, it might be fixable by placing the view port slightly forward
19:18:12Water255Hello all. I just want to say that I've used a lot of software, and I've had some of my best experiences with open-source programs. God bless you guys and gals.
19:18:47 Quit Water255 (Client Quit)
19:21:33[Franklin]foolsh: it's /very/ messy code rightn ow
19:21:53foolsh:) yeah
19:22:07[Franklin]but it's scrolling
19:24:37[Franklin]not sure what's causing the grass on the road, though
19:26:42foolshxracer, the x makes it cooler :D
19:26:53[Franklin]lol XD
19:28:00[Franklin]ok yay borders working
19:28:11[Franklin]now just got to get the left side of the road working :D
19:28:23*[Franklin] suspects a signedness issue
19:30:01[Franklin]foolsh: I mainly chose "xracer" not to annoy [Saint]
19:30:16[Franklin]otherwise I would have done "rockracerbox" :P
19:30:36foolshthat's a hidious name :D
19:30:52[Franklin]or "racebox"
19:31:18[Franklin]that's actually not that bad
19:31:31foolshOn the real fuze+ target, it looks the same but runs slightly slower, of course
19:31:49[Franklin]try the latest patch set
19:31:58[Franklin]it moves the camera a bit higher and faster
19:32:04[Franklin]so the scrolling's more obvious
19:32:42 Join TheLemonMan [0] (~lemonboy@unaffiliated/thelemonman)
19:39:02[Franklin]foolsh: does it scroll smoothly on target?
19:39:09pixelmahmm, maybe too obvious... but boxracer?
19:39:42[Franklin][Saint] would murder someone for that :P
19:44:31foolshnice touch with the track border, so are you feeling "Pole Position" or "Outrun"?
19:44:56 Join petur [0] (~petur@rockbox/developer/petur)
19:46:16[Franklin]more Pole Position
19:46:30[Franklin]does it only draw the right side on target too?
19:47:05foolshhang on let me build
19:49:20foolshHmm, it's more of an actual 3D engine now than before?
19:49:20 Join wodz [0] (
19:49:33[Franklin]foolsh: yes
19:49:33 Quit MMlosh (Quit: Bye...)
19:49:45[Franklin]it could easily be extended to 3d
19:50:08[Franklin]currently it's only 1.5D projected hacked into 3D
19:50:32foolshyes on target renders just the same as the sim
19:50:39[Franklin]I'm reworking the scroll code now so it loops
19:50:41*[Saint] eye twitches
19:51:04[Saint]rock*.rock or *box.rock makes me want to kick puppies
19:51:22[Franklin]how about rock*box.rock?
19:51:36*foolsh kicks two puppies
19:51:45wodzgevaerts: I know we have more _ prefixed functions. I have nothing against internal functions like this but I think its silly to have target specific functions needed by the core to be named like this
19:52:40wodzgevaerts: and finally I touched _buttonlight_* also :-)
19:52:56 Join MMlosh [0] (~MMlosh@2001:470:6f:23:d4af:3e66:6efc:9273)
19:52:57gevaertswodz: if you feel strongly about it, I won't object :)
19:53:36wodzok, then I'll push the button when I have time to hang around to fix emerging reds/yellows
19:53:54 Quit eeeeeta (Remote host closed the connection)
19:55:15 Join eeeeeta [0] (~eeeeeta@gateway/tor-sasl/eeeeeta)
19:55:27 Quit eeeeeta (Remote host closed the connection)
19:56:32 Join eeeeeta [0] (~eeeeeta@gateway/tor-sasl/eeeeeta)
20:00:21foolsh[Franklin]: I see you have some procedural elements and some render elements, perhaps use a procedure to draw the grass first, and then render the track on top will stop the left side from doing that
20:00:46[Franklin]that's what it does now
20:01:06[Franklin]it seems it's just that the road's not covering the grass
20:03:07foolshsorry, I meant in a seperate render_road fuction perhaps, the are mixed in together and gets called all at once
20:03:19foolshthey* are mixed
20:03:20[Franklin]ah yes
20:04:58 Quit MMlosh (Quit: Bye...)
20:05:48foolshso every frame, draw_grass then render_raod, I guess would fix it
20:06:12[Franklin]no, it (shouldn't) wouldn't
20:06:20[Franklin]same code, just function-ized
20:06:30[Franklin]that's what's being done now
20:06:41[Franklin]draw_grass draw_border draw_road
20:06:55[Franklin]just in one big function
20:07:53foolshAh its draw where we can see it drawing, needs a buffer to flip and flop,
20:08:23[Franklin]no, rockbox is double-buffered by default
20:08:29[Franklin]nothing shows until lcd_update() is called
20:14:27foolshAh, I have it this time, you call an extra lcd_update on line 209, only need one per frame
20:15:25foolshDang, nope it still doesn't draw the left side correctly
20:15:38[Franklin]I think it's a signedness issue
20:15:48[Franklin]probably something with project()
20:16:25 Join the-kyle [0] (
20:16:52[Franklin]bah... forget that
20:16:54[Franklin]I'll get lanes working
20:18:49foolshif it's a signedness problem, translate the origin far enough away so the scene never passes over zero
20:19:20[Franklin]signedness with x-coord
20:19:26[Franklin]so not that
20:26:30foolshneeds key some kind of handling, I have to manually kill the process for the sim each time
20:27:01[Franklin]too lazy :P
20:27:21[Franklin]I'll get to it ;)
20:28:12 Quit Strife89 (Ping timeout: 265 seconds)
20:29:15foolshI would look at cube.c it's basicly what you've done here but with backface culling, and see how they handled projectioning it from origin to screen
20:29:41*foolsh has sleepy finger again
20:29:52pixelmaguess I'm safe with boxracer then :P
20:31:15[Franklin]foolsh: changing SEGMENT_LENGTH to 1000 draws part of the left side of the road
20:32:13[Franklin]it also makes the border of the road jagged
20:36:50***Saving seen data "./dancer.seen"
21:03:03foolshI sort of fixed it
21:03:40foolshline 158 fill_poly(x1-w1, y1, x1+w1, y1, x2-w2, y2, x2+w2, y2, color);
21:03:54[Franklin]what's different?
21:04:06foolshline 158 was fill_poly(x1-w1, y1, x1+w1, y1, x2+w2, y2, x2-w2, y2, color);
21:04:41[Franklin]thanks foolsh
21:06:09foolshso it was a signedness error :)
21:07:30foolshnow the brder is jagged
21:17:43 Join krabador [0] (~krabador_@unaffiliated/krabador)
21:19:28 Join Ivoah [0] (
21:21:12 Quit eeeeeta (Remote host closed the connection)
21:22:54 Join eeeeeta [0] (~eeeeeta@gateway/tor-sasl/eeeeeta)
21:25:35[Franklin]foolsh: it doesn't seem to help
21:25:44[Franklin]I think it's the fill_poly function now
21:26:10 Part BobJonkman1
21:27:33[Franklin]foolsh: it's assuming that the coordinates are given in the correct order
21:27:41[Franklin]which they're not
21:27:57[Franklin]so I just need to reorganize them so that they are
21:30:25foolshopps I forgot about the change I did to line 122 pt_2d->w = scale * ROAD_WIDTH * LCD_WIDTH/2; shouldn't it be pt_2d->w = scale * ROAD_WIDTH * (LCD_WIDTH/2);
21:31:30foolshnope I see no difference
21:34:55[Franklin]this code is dizzying
21:35:09[Franklin]I'm getting a headache looking at it
21:36:54[Franklin]I guess I'll make fill_poly sort the coords
21:41:20*thomasjfox prepares for yellow
21:41:49fs-bluebotBuild Server message: 3New build round started. Revision 2a3e162, 255 builds, 25 clients.
21:41:59[Franklin].show HEAD
21:50:51Ivoah[Franklin]: Is there grayscale support in xworld?
21:51:00[Franklin]nope, sorry
21:51:04IvoahAww :(
21:51:11[Franklin]the game looks terrible in grayscale anyway
21:51:23[Franklin]Ivoah: in theory its totally possible, though
21:51:34IvoahAny more work on z80e?
21:51:41[Franklin]Ivoah: no, on xracer
21:52:00IvoahIs xracer going to support grayscale?
21:52:08[Franklin]Ivoah: it should
21:52:33[Franklin]nothing's color-specific yet
21:59:16thomasjfoxhmm, shouldn't there be a message from fs-bluebot?
21:59:24*[Franklin] slaps fs-bluebot
21:59:24fs-bluebot[Franklin]: ouch!
22:02:18bluebrotherhmm, interesting. My client has been temporarily disabled because of some error
22:02:56bluebrotheraaah. It tried to build the html manual on that machine, but it only supports pdf
22:03:09*bluebrother should fix this
22:03:51thomasjfoxstrange it worked before
22:05:04bluebrotherusually that machine gets the pdf manual
22:05:30bluebrotherseems the build master tried to give it the html version as well this time
22:08:35[Franklin]how can I fill a quad in rockbox?
22:11:05[Franklin]tris work to some extent, but they're dependent on input order
22:11:17[Franklin]sorting them at runtime is too expensive
22:16:00 Quit petur (Quit: Leaving)
22:18:54foolshYou'll have to use triangles, quads only work with 90 degree angles I believe
22:19:06foolshat least the last time I looked
22:19:30 Quit thomasjfox (Ping timeout: 264 seconds)
22:20:13foolshis this about the track border?
22:20:16[Franklin]w00t! it works!
22:20:36 Join choki [0] (5b729d16@gateway/web/freenode/ip.
22:20:47[Franklin]the border and the road
22:20:48chokianyone with a sansa clip + in here?
22:20:56chokibtw hello =)
22:20:58sobukusI wonder if I should do some more poking on G#887 to indicate that it's ready for merging (from my POV).
22:21:03sobukuschoki: Yes.
22:21:04fs-bluebot3Gerrit review #887 at : 3Enhancement of the metronome plugin: by Thomas Orgis
22:21:21chokisobukus: how is the sound quality compared to ipod touch 5g?
22:21:37chokisobukus: would you buy one even 2015?
22:21:40sobukuschoki: I cannot compare, but I have nothing to complain about with the sansa.
22:22:11sobukusSound quality lives in your headphones ...
22:22:47sobukuschoki: Yes, I did buy an additional one last year. Though, I'm using it as metronome for drum practice.
22:23:04sobukusEspecially with Rockbox, it's a darn useful device.
22:23:24chokisobukus: oh. can u recommend some headphones below 100$?
22:24:30sobukusWell, I got those Sony MDR Q68LW, but not because of sound (they're rather light on bass, but OK) but because I don't get entangled in cords!
22:24:57sobukusI liked some philips clip-ons some time way back, sound-wise.
22:25:08sobukusBut they fell apart eventually.
22:25:24sobukusI got some specific headphones for a relativfe some time back. Let me remember ...
22:25:40chokisobukus: do you think those headphones about 200$ are really worth the money? is it worth to get them for a sansa clip mp3 player?
22:26:30sobukusAh, yes, AKG K-450, those are fine, too.
22:27:14chokicheap and good? :D
22:27:26sobukusI didn't look into super-expensive headphones apart from Beyerdynamic DT-990 Pro (or DT-770) for studio work
22:27:42sobukusFor HiFi, I prefer by tower speakers.
22:28:23chokiu r a sound guy :D
22:28:31sobukusI'm a drummer.
22:28:41chokiwhy u up so late? ^^'
22:29:00sobukusIt's not that late.
22:29:14[Franklin]sobukus: speaking of G#887, I guess you'll have to wait a bit if no one sees it
22:29:20fs-bluebot3Gerrit review #887 at : 3Enhancement of the metronome plugin: by Thomas Orgis
22:29:25[Franklin]though hopefully gevaerts is here listening :)
22:29:48sobukus[Franklin]: It's fine, I was just wondering if there is some action missing on my part of if it's just down to waiting.
22:29:59[Franklin]just wait ;)
22:30:11[Franklin]poke people a bit if you want
22:30:37sobukuschoki: One point about headphones for me is if they are comfortable to wear. My ears don't like a lot of them.
22:30:49chokido you think standalone mp3 players are still alive today?
22:30:55sobukusThe beyerdynamic ones are super-comfy and have decent sound. win-win.
22:31:15sobukuschoki: Yes, but only if they have something to set them apart from smartphones.
22:31:22sobukuschoki: The Clip+ has.
22:31:34[Franklin]a phyiscal keypad to play games on :P
22:31:39sobukusIt's a lot smaller and lighter, has decent battery life.
22:31:48chokiin terms of sound quality?? i cant believe smartphones are so much bad O_o
22:31:57sobukusAnd, if the battery of the MP3 player is flat, you can still call folks on the phone;-)
22:32:15chokii need a feature phone :D
22:32:41sobukusI can switch out the SD card on the Clip+ without prying open some backside covering.
22:33:02sobukusAgain, it's small and light, so I can just clip it onto a T-Shirt for jogging/cycling.
22:33:03chokihahaha, it seems to be made exactly for doing this i guess
22:33:37sobukusI believe in devices for certain purposes.
22:34:11[Franklin]foolsh: fixed xracer
22:34:19sobukusI can place the clip somewhere at a venue to record a concert and not worry about some stage hand carrying off my pocket computer with all my contacts, messages, naked pictures, etc.
22:34:24[Franklin]not really a "fix" though
22:34:37[Franklin]just changed the poly drawing routine a bit
22:35:31sobukuschoki: Guitarists _could_ just use some apps on their smartphones for processing sound, but for some reason they insist on stomp-boxes;-)
22:36:52***Saving seen data "./dancer.seen"
22:36:58foolsh[Franklin]: yeah line 103 was the ticket
22:39:06[Franklin]now to make it a game
22:39:19[Franklin]first curves and hills though
22:39:32[Franklin]but maybe a drag-racing game wouldn't be so bad :D
22:40:15[Franklin]for a game, I'll need some libre sprites
22:40:40sobukuschoki: Coming back to the ipod touch ... for that I really have trouble seeing how it's not obsoleted, if even by your second smartphone that's still working but not in fashion anymore.
22:40:44[Franklin]possible but... really? :P
22:41:14foolshif you use bitmaps won't they have to be diffent sizes for the different targets?
22:41:31foolshlot's of extra work IMO
22:41:41[Franklin]gimp can scale pretty easily
22:43:38[Franklin]plus, if the game were to have multiple cars, they'd have to be scaled at run-time anyway
22:44:11foolsh[Franklin]: well then just a large one for each car, large enough to not crap out on a target with a large screen resolution, then scale it down onthe fly
22:44:25[Franklin]now to find one
22:45:51[Franklin]I suppose I could use Blender to make some
22:46:37[Franklin]but I'm not an expert blender artist
22:48:46wodzYou definitely didn't try scaling at such low resolutions. The output will look crap.
22:49:33[Franklin]I think I should get curves and hills working first, though
22:49:56wodzpamaury: ping
22:50:14pamaurywodz: pong but I'm going to bed
22:50:24pamauryso I can answer a quick question ^^
22:50:26chokisobukus: u think the ipod touch is end of life? i mean it is still a good alternative to iphone :D
22:50:56wodzpamaury: I am working on atj hwstub to survive bad things like unaligned access and such
22:51:12pamaurywodz: you can inspire from my patch I guess on ARM
22:51:34wodzpamaury: not really mips is quite different here
22:51:59wodzpamaury: anyway how do you signal in your ARM patch that recovery process took place?
22:52:20wodzpamaury: I mean do you expose this somehow?
22:52:26pamauryI just report read/write failure, by stalling
22:53:07pamauryby stalling
22:53:22wodzyes but how do you pass this information from exception?
22:53:46wodzI mean how upper level code knows something wrong happened?
22:54:29pamauryah, with some setjmp/longjmp like code, let me find the code
22:55:07fs-bluebot3Gerrit review #980 at : 3hwstub: implement read/write data abort recovery by Amaury Pouly
22:55:25pamaurybasically I call a function called set_data_abort_jmp()
22:55:27sobukuschoki: I think it's too similar to a phone.
22:55:36sobukusBut just my personal opinion.
22:55:40pamauryit returns 0 on the first call, and returns 1 on exception (it can returns twice)
22:55:50chokiyeah but only if u have smartphone :D
22:56:02pamauryit saves the current context and on exception it jumps back to the calling point and returns again
22:56:44pamaurythe implementation details are ARM specific of course but the principle should work on any platform I think
22:57:01wodzwill look at this
22:57:20pamaurylook at system.S, it quite commented :)
22:57:30[Saint]choki: are you aware what channel this is?
22:57:44chokisorry im silent now xD
22:57:59[Franklin]choki: -> #rockbox-communnity
22:58:01[Saint]I'm just curious, it wasn't me saying to be silent.
22:58:15[Saint]Its just the the iPod Touch has absolutely zero relevance to Rockbox.
22:58:40MarcAndersenIt could have if rockbox was ported to it or ios
22:59:01[Saint]That's a _huge_ if.
23:00:08pamaurywodz: I'm going to bed, feel free to ask tomorrow if you have more questions
23:00:20wodzpamaury: ok, thanks
23:00:22chokiim just think about if i get a sandisk clip+ and get rid of all that stuff in my pants :D
23:00:49MarcAndersenHey, is there any specific reason why the monkeys audio codec decodes so slowly?
23:03:21[Franklin]MarcAndersen: because it's an expensive algorithm, I'd think
23:03:59MarcAndersenyes, but no other codec is so slow to decode.
23:04:33 Quit pamaury (Ping timeout: 256 seconds)
23:04:47[Franklin]because no other codec is designed to such extreme specifications
23:04:56MarcAndersenI have a file where it plays one second of it, then there is 10 seconds of silance and then a second of audio again and so on
23:05:37MarcAndersenI will just stick to flac
23:05:59 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
23:06:11[Saint]If you can demonstrate an actual regression, its worth looking into.
23:06:19[Saint]But in general, its a rather insane codec.
23:06:39[Saint]And precisely nothing that should influence decoding speed has been changed in a long time.
23:06:45chokithis is not #codec :D
23:07:37Mode"#rockbox +o [Saint]" by ChanServ (ChanServ@services.)
23:07:44[Saint]we wanna play this game huh? ;)
23:07:56 Nick [Franklin] is now known as [ATFranklin] (~franklin@unaffiliated/franklin)
23:08:12 Nick [ATFranklin] is now known as [Franklin] (~franklin@unaffiliated/franklin)
23:08:12Mode"#rockbox -o [Saint]" by ChanServ (ChanServ@services.)
23:08:17 Nick [Franklin] is now known as [ImAnOperator] (~franklin@unaffiliated/franklin)
23:08:19 Nick [ImAnOperator] is now known as [Franklin] (~franklin@unaffiliated/franklin)
23:08:31MarcAndersenbut you only get a few more bytes of compression on monkeys audio insane besides preset 8 of flac
23:08:49[Franklin]it's stupid, APE
23:09:36[Saint]Anyway, as stated, if you can demonstrate an actual regression. I mean, if we've _actually_ gotten worse in decoding APE, which I doubt, it is definitely worth looking into.
23:10:01[Saint]In general its terrible on all but the fastest devices.
23:10:18MarcAndersenI just think it's the codec. my computer needs a second before even beginning to play, or when i seek in the file...
23:10:22[Saint]There's really no good reason to use it over more conventional lossless codecs.
23:10:50[Franklin]disk is cheap
23:12:10bertrikdisk may be cheap, but you have to backup too
23:12:21[Saint]it really isn't the best choice at all.
23:12:24[Saint]by a wide margin.
23:12:44[Franklin]foolsh: I got curves working now
23:12:47[Franklin]faked curves, that is
23:14:33 Quit bluebrother (Disconnected by services)
23:14:34 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
23:15:11MarcAndersenwow! a target on around 8% realtime with ape... that is insane
23:15:35 Join williamtdr [0] (uid27909@gateway/web/
23:16:03[Saint]compared to the gigabeat with flac:
23:16:03 Quit fs-bluebot (Ping timeout: 245 seconds)
23:16:04[Saint]flac_5.flac5395.70% realtimeDecode time - 3.26s9.78MHz
23:16:04[Saint]flac_8.flac4997.15% realtimeDecode time - 3.52s10.56MHz
23:16:12[Franklin]SPAM! :P
23:17:48chokido u listen to flac or mp3 guys?
23:18:03 Join fs-bluebot [0] (
23:18:18[Saint]Cook RA on the MSM7227 (ARMv6) is nuts...I wonder why its _so_ efficient to decode on that SoC in particular?
23:18:27[Saint]pity is an awful codec, but, wow.
23:18:47bertrikI have tin ears, and I only listen to compressed audio
23:19:09[Franklin]forget sound
23:19:21[Franklin]rockbox is the ultimate mobile gaming experience :D
23:19:21[Saint]If I'm streaming from my home server (which I usually do), I se mp3@320
23:19:30[Saint]on-device, I use flac8
23:20:22[Saint]I should so some flac decode numbers on the Snapdragon 801
23:20:29[Saint]I bet that's be damn impressive.
23:20:47[Saint]likely somewhere in the order of 10,000% realtime
23:20:52[Saint]if not more.
23:21:50[Saint]if a wee single core MSM7227 ARMv6 can do it at ~7000% realtime, imagine what four times the CPU and four times the cores can do.
23:22:12[Saint]there's a point, though, where it just gets silly.
23:23:03[Franklin]I doubt that you need 40,000% realtime
23:23:31[Saint]I doubt it'd scale linearly like that, but, yeah.
23:24:24[Saint]There's a lot more factors than just how many MHz and how many cores you can throw at decoding that make this difference.
23:25:03[Saint]but, yeah, I would posit the Sanpdragon 801 and 805 would barely break a sweat at flac or wav at all.
23:25:48[Franklin]even a z80 could decode wav at 1000% realtime
23:26:06[Franklin]what "decoding" is there to do? :P
23:26:26*[Saint] posits that number is more than a slight exaggeration.
23:26:36[Saint]drastically so, but, whatevs.
23:28:13[Saint]I also wonder if you're talking about the 2.5, 4, or 20MHz variants.
23:28:34[Saint]Though I sincerely doubt any of them could do 1000% realtime flac decode.
23:28:41[Saint]quite sincerely.
23:30:34[Saint]If you can get a z80 to do 1000% realtime decode of flac8, hell, I'll let you away with flac5, on a z80, any z80 variant (2.5/4/20MHz), there's hats to be eaten.
23:31:02[Saint]gah, sorry, wavpack.
23:31:19[Franklin]yeah... ;P
23:32:42 Quit wodz (Quit: Leaving)
23:32:57[Saint]The Classic can't even do it. ;)
23:33:47[Saint]The Gigabeats can, just.
23:34:53*[Saint] should really get around to fixing his Beast
23:35:37[Saint](Gigabeat S - aka. The Badass, DAP of Doom, Destroyer of Codecs)
23:35:47[Franklin]DAP of DooM?
23:36:04*[Franklin] imagines that DooM is quite nice on the DAP of Doom
23:36:32*choki is listening to E.S. Posthumus :3
23:37:09 Join saratoga_ [0] (123e1c4f@gateway/web/freenode/ip.
23:37:20saratoga_WAV isn't decoded, its just copied directly from the disk to the DAC
23:37:33saratoga_if you benchmark wav decoding its really just a hard disk benchmark
23:38:01 Quit ender` (Quit: Two possibilities exist: Either we are alone in the Universe, or we are not. Both are equally terrifying.-- Arthur C. Clarke)
23:38:28[Franklin]saratoga_: sobukus thinks that G#887 is "ready"
23:38:32fs-bluebot3Gerrit review #887 at : 3Enhancement of the metronome plugin: by Thomas Orgis
23:38:45saratoga_pretty much anything with a hardware 32 bit multiply instruction will decode flac
23:39:04[Franklin]though you may want to read the latest review
23:39:49saratoga_could someone else look at this? i'm kind of busy this week and don't have time to fool around in case it breaks on some target
23:40:12[Franklin]I've just glanced through it
23:43:51 Quit xorly (Ping timeout: 240 seconds)
23:51:23 Quit lebellium (Quit: ChatZilla [Firefox 35.0/20150108202552])
23:55:12 Join xorly [0] (
23:56:40 Join Karmadealer [0] (59844aaa@gateway/web/freenode/ip.
23:57:47 Quit choki (Quit: Page closed)

Previous day | Next day