Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2012-10-31

00:00:03*scorche` would think it would be better to have B, C, A
00:00:31scorche`not like the ratings really matter - someone could just come along and skew the ratings quite easily, no?
00:00:45AlexP_at the same time I don't think it is massively important
00:00:47amayerI think so
00:01:08amayerAlexP_: I noticed it
00:01:18amayeri figured it wouldnt be that hard to fix
00:01:49AlexP_sure, I'm not objecting
00:05:44 Quit JdGord (Read error: Connection reset by peer)
00:13:34 Quit pamaury (Ping timeout: 260 seconds)
00:16:03 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
00:16:03 Quit LambdaCalculus37 (Client Quit)
00:20:46 Join JdGord [0] (~AndChat80@
00:22:10 Join Clear_runway [0] (~Clear_run@
00:23:59 Join JdG [0] (
00:25:02 Quit JdG (Client Quit)
00:27:35 Quit JdGord (Ping timeout: 240 seconds)
00:31:23 Quit ender1 (Quit: Knowledge is power. Power corrupts.)
00:42:08 Quit prof_wolfff (Ping timeout: 276 seconds)
00:58:58 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
01:07:51***Saving seen data "./dancer.seen"
01:13:27 Quit scorche (Disconnected by services)
01:13:30 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
01:19:25 Quit Horschti (Quit: Verlassend)
01:46:52 Part the-kyle ("PART #opus :PING 1351644413")
01:49:49 Join the-kyle [0] (
01:59:31 Quit tjb0607 (Ping timeout: 248 seconds)
02:01:22 Join tjb0607 [0] (~tjb0607@
02:04:47 Quit lebellium (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121023124120])
02:19:06 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
02:44:52 Quit guymann (Quit: hmm)
02:44:58 Join yuriks [0] (~yuriks@
02:48:59 Join guymann [0] (~c@unaffiliated/guymann)
02:53:49yuriksI was working on the glyph fallback patch a few months ago
02:54:09yuriksI picked it up again
02:54:22yuriksJust want to know, is a change in the .fnt format much frowned upon?
02:54:39yuriksexisting .fnt files would work on the new version
02:55:06JdGordonas long as all the tools are updated it should be fine
02:55:13JdGordonespecially if the change is backwards comapatible
02:55:40yuriksI want to change it so that missing glyphs get -1 on the offset/width table, instead of pointing to a dummy character
02:59:10JdGordonthat sounds ok
03:02:37 Quit efyx__ (Ping timeout: 245 seconds)
03:07:54***Saving seen data "./dancer.seen"
03:16:16 Join efyx__ [0] (~efyx@
03:23:55 Part amayer
04:25:15 Join TheSphinX_ [0] (
04:28:59 Quit TheSphinX^ (Ping timeout: 276 seconds)
04:47:27 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:47:28 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:47:28 Quit pixelma (Disconnected by services)
04:47:28 Quit amiconn (Disconnected by services)
04:47:30 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:47:30 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:49:15 Join Rower85 [0] (
05:07:58***Saving seen data "./dancer.seen"
05:29:59 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
05:30:52 Quit TheSeven (Disconnected by services)
05:31:01 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
05:55:08JdGordon[Saint]: you wanted something last night?
06:01:20[Saint]not like the ratings really matter - someone could just come along and skew the ratings quite easily, no?
06:01:50[Saint]scorche`: ^ that *could* happen, if you were rather dedicated, or had a large amount of IPs to throw around.
06:02:09[Saint]I believe it is limited to 1 rating, per theme, per IP, per day.
06:02:29[Saint]JdGordon: I was wondering something about alternating sublines.
06:04:09[Saint]I wondered if the %t(time_out) tag just made everything before a ";" a subline, or if you could do "Some_text_that_doesn't_alternate %t(time_out)some_text_that_does; %t(time_out)other_text_that_alternates"
06:07:20JdGordonthat quote was all on one line?
06:07:31JdGordonthat cant be done
06:07:41[Saint]right, thanks. all I needed to know.
06:07:43JdGordonthough, hmm
06:08:03[Saint]I assumed so, and I looked at the code and I was pretty damn sure...but I just wanted to confirm.
06:08:04JdGordonwhen/how would it make sense anyway?
06:10:11[Saint]I wanted to do "%Sx(Next Track:) %t(foo)some_metadata; %t(foo)some_more_metadata" instead of "%t(foo)%Sx(Next Track:) some_metadata; %t(foo)%Sx(Next Track:) some_more_metadata"
06:10:30[Saint](to avoid a small amount of needless repetition)
06:10:49[Saint]It's no big deal though, I can certainly deal with it.
06:17:45 Quit shamus (Read error: Connection reset by peer)
06:18:05 Join shamus [0] (
06:21:55JdGordon[Saint]: autosizing+layouts for viewports would be more useful
06:22:01JdGordonand would make that possible
06:22:19[Saint]I don't really see how...
06:23:36[Saint]All I wan't to be able to do is "some_static_text some_text_that_alternates" without weird viewport fuckery or having to reproduce the static text in each and every subline.
06:23:45[Saint]but, it really isn't important ;)
06:24:28JdGordonyou'd have 2 (child)viewports... one has "%Sx(Next Track:) ", the other has the sublines and would be glued to the first
06:24:59JdGordonAlso, i think the way we do metadata display is just stupid
06:25:08[Saint]yeah, that's way messier (imo) than just reproducing a bit of text that adds a few b to the theme total.
06:25:17JdGordonwhy should the themer decide what to display? surely it should be up to the user
06:25:24[Saint]and, what about it would you like to change?
06:25:37JdGordoni.e theme says "show 5 lines of metadata", user says "great, do these:"
06:26:32JdGordoni.e there would be some file which has just text tags which we'd import
06:26:35[Saint]Oh, that...yeah. Just, don't do any breaking changes before I get this cabbie for native targets finished ;)
06:26:43[Saint]You can break it all you like after that :P
06:26:54JdGordonthis would mean you wouldnt need to do that at all
06:27:04JdGordonwe'd ship 1 file all cabbies would use
06:27:30JdGordonand dont worry, my rockbox time has dropped to Not-Very-Much-At-All :(
06:29:29[Saint]When I do eventually push this to gerrit (my time is pretty much nil with the family invasion at my apartment presently), is it too much to push the code re-work (mainly cosmetic/unification work with very little changes) and change the default font as well?
06:29:37[Saint]Or, better as two seperate reviews?
06:30:09[Saint]I want everything that can to use an antialiased GNU Unifont so langs "Just Work"
06:30:53[Saint]A user changing the lang and finding out exactly how limited the default font (mainly Helvetica iirc) is on their target really isn't too nice.
06:31:34[Saint]if you want Hebrew or Arabic with cabbie on some targets it's just a joke :)
06:32:33[Saint]And it's non-obvious that the user would need to change the font, and in many cases changing to the only GNU variant we have will break the theme terribly as we only have a 16px version and many targets use a much smaller font.
06:33:12[Saint]....luckily, I have a fat stack of antialised GNU unifonts in my github, plus the recipes to reproduce them ;)
06:33:31JdGordon[Saint]: did you RaaAa cabbie changes ever get pushed?
06:34:21[Saint](the slight issue is that we build the fonts "on demand", and each font needs a slightly different recipe to make it look right. So it would be better to retrieve pre-made fonts from the source instead of build them on demand IMO.
06:34:49[Saint]and no, but I plan to do so either at the same time as, or slightly thereafter, the native cabbie stuff I'm working on.
06:36:25[Saint]the only issue with retrieving the pre-made fonts is that if ever the font's source updates, the committer needs to remember to generate the new version of the font.
06:36:49[Saint]but, I looked at the commit history and the fonts pretty much never do get updated.
06:37:35*JdGordon wonders how the fuck this user metadata choice thing would work
06:37:55JdGordonI don't tihnk it would be easy to hook into the tag parser to do it
06:38:31*JdGordon goes to contemplate this idea on the dunny
06:38:47[Saint]I was going to bring up the font thing on the ml tonight, but I'm not sure I have the energy, so...tomorrow.
06:51:01JdGordon[Saint]: what do you tihnk of the idea of a which is a wps with a bunch of special viewports in it, each only allowing text based tags which get used from other skins?
06:51:34JdGordoni.e there would be a viewport named "this_track_1_line" which expects exactly one line of text
06:52:22[Saint]define "bunch of special viewports", ideally, I would think it should just be reading metadata display code that is verbatim to that which you would find in any other theme.
06:52:41[Saint]essentially just throwing the existing metadata portion of cabbie out into its own file.
06:52:56JdGordonso in cabbie you'd have a viewport for the "this track" data and you'd call %XX(4) - which would then pull in the text from the this_track_4_lines viewport
06:53:08[Saint]I'd prefer it not have any crazy magic or special formatting, the theme itself should define the viewport.
06:53:22JdGordonyes, the theme itself would
06:53:25JdGordononly the text would be used
06:53:47JdGordonit might even be a new tag all together
06:53:55JdGordonI'm not sure if this would work, but it might
06:54:14[Saint]Oooooooohhhh, right, your initial explanation confused me. But, we're on the same track, yes.
06:54:20JdGordonwe already allow the wps to draw inside the sbs, so this owuldnt be so different
06:54:28[Saint]I wouldn't like the viewport to define how many lines it has, though.
06:54:40[Saint]it should display as many full lines as the font allows it to.
06:55:01JdGordonthe themer will know how many lines will fit
06:55:13[Saint]but some themes dont hardcode a font.
06:55:35JdGordonneed to think about this some more, and find time to code it
07:05:56*JdGordon thinks this might actually be very doable
07:07:28JdGordonmaybe we should just switch to css and java :p
07:08:00***Saving seen data "./dancer.seen"
07:22:39 Join derk0pf [0] (
07:39:19 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
07:40:30 Join mortalis [0] (
08:13:29 Join kevku [0] (
08:17:53 Quit bertrik (Ping timeout: 240 seconds)
08:28:35 Join Zagor [0] (
08:28:35 Quit Zagor (Changing host)
08:28:35 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:32:20Clear_runwayanyone there?
08:32:29*[Saint] just confirmed his theory that post; edit post immediately on the forum == akismet trigger
08:32:47[Saint]Clear_runway: multiple someones, in fact.
08:33:04Clear_runwaycool. I'm getting a weird problem on my ipod touch 2g.
08:33:22Clear_runwaywhat happens is the the top of the screen sort of flashes on the bottom
08:33:31[Saint]well, seeing as we don;t support that device an any way shape or form, I'm pretty sure you should talk to Apple.
08:33:31Clear_runwaythis only happens in rockbox
08:33:40Clear_runwayoh shit
08:33:41[Saint]we do NOT support the iPod Touch
08:33:44Clear_runwayi meant ipod nano
08:34:02[Saint]Aha, that's better ;)
08:34:10Clear_runwayI think its some sort of driver glitch
08:34:24[Saint]running what revision of Rockbox?
08:34:27Clear_runwayi tried rebooting. i tried reinstalling rockbox
08:34:30Clear_runwayer, just a sec
08:38:06[Saint]Is the same visual glitch present in the current release?
08:38:44Clear_runwaydunno. it looks faded, like it's flashing every other frame.
08:39:11[Saint]by "current release' I don't mean the current development snapshot, but the release build.
08:40:51Clear_runwayoh, and one other thing. if i leave it alone and the display turns off, it just crashes. white screen with blue lines running across.
08:42:12[Saint].uuuuugh, shit on it. that's right, the n2g release is ~2 releases behind.
08:42:31[Saint]dammit, I was hoping for something slightly closer to git HEAD as a "known good".
08:43:58[Saint]I can't do anything about it right now, but I can test later this evening or tomorrow afternoon with a slew on n2g (with various LCD types).
08:44:01Clear_runwaybut yeah, it's not borken hardware, only happens in rockbox.
08:45:04[Saint]well, as I can't look at confirming or denying this right now, it would help if you could outline your findings in
08:45:37[Saint]include steps to reproduce the problem, the revision used, and any other details you find applicable.
08:47:47 Join einhirn [0] (~Miranda@2001:638:605:4:a8de:19fa:59e9:2512)
09:07:03 Join LinusN [0] (
09:08:01***Saving seen data "./dancer.seen"
09:09:16 Join |akaWolf| [0] (~akaWolf@
09:10:35 Quit akaWolf (Ping timeout: 240 seconds)
09:22:20 Quit Clear_runway (Read error: Connection reset by peer)
09:22:43 Join Clear_runway [0] (~Clear_run@
09:44:17 Join pamaury [0] (
09:44:18 Quit pamaury (Changing host)
09:44:18 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:06:08 Join max3 [0] (~Administr@
10:06:10 Join lovasoa [0] (~Thunderbi@2a01:e35:8a2e:8080:6939:7f24:2508:9b05)
10:06:51 Join speckmade [0] (
10:07:02max3can someone help me out? i have an old school mini with rockbox on it with a cf card and it's on the fritz. i get the !folder icon on boot and then it resets. if i try to put it in disk mode it'll reset out as soon as it's recognized by my computer. i can access diag mode
10:23:01 Quit einhirn (Ping timeout: 260 seconds)
10:28:48 Quit lovasoa (Quit: lovasoa)
10:30:04 Part LinusN
10:31:48 Join LinusN [0] (
10:35:04max3can someone help me out? i have an old school mini with rockbox on it with a cf card and it's on the fritz. i get the !folder icon on boot and then it resets. if i try to put it in disk mode it'll reset out as soon as it's recognized by my computer. i can access diag mode
10:35:52 Join perrikwp_ [0] (
10:37:57 Join ender1 [0] (
10:38:05 Quit perrikwp (Ping timeout: 240 seconds)
10:39:12gevaerts[Saint]: post/edit is presumably submitted twice to akismet, which I suspect sees it as mass posting
10:47:42[Saint]gevaerts: aha, I was thinking something along those lines.
10:48:08gevaertsSo if you edit, just translate it to French at the same time or something :)
10:48:15[Saint]I was thinking there was some "X post per/s" rule, or so. But thatnks for (half) confirming my suspicions.
10:48:16 Join lebellium [0] (
10:48:56[Saint]gevaerts: were you able to dig my last post out of the spam bucket, please?
10:48:59 Quit ender` (Read error: Operation timed out)
10:49:02[Saint](if you've not done so already)
10:49:45[Saint]I really wish there was some form of "trusted user" setting.
10:49:55[Saint]like a post count of 100+ or something.
10:50:37[Saint]I would say at that stage, posts from said user are likely legit ;)
10:53:57JdGordon[Saint]: what are you doing on cabbie now?
10:54:05JdGordonand rpely to my pm!
10:54:35[Saint]JdGordon: syntax updates, code formatting/commenting, and unification.
10:54:46*[Saint] doesn't seem to have one.
10:55:55[Saint]a few cabbies still predate alignment tags and viewports...uuugh.
10:56:00copperWhen does System -> Running Time reset?
10:56:01[Saint]so, I guess I'll do them too.
10:57:18[Saint]copper: time between a full charge cycle iirc.
10:57:47[Saint](disclaimer: running off memory)
10:58:00copperit doesn't seem to reset after I charge it
10:58:48[Saint]I think it's the time between the last full charge and the last full discharge
10:58:49copperon an iPod Classic btw, not sure whether that's relevant or not
10:59:12[Saint]I _think_
10:59:23[Saint]not really in a position to poke at the source right now.
10:59:38copperok, thanks
11:00:00[Saint]The manual should explain this.
11:00:09[Saint](I can't really check that now either)
11:00:38[Saint]"This item shows the cumulative overall runtime of your player since you either disconnected it from charging (in Rockbox) or manually reset this item. A manual reset is done through pressing any button, followed by pressing Select or Next."
11:01:00coppernot sure it's working properly then
11:01:09[Saint]so, it should be "time since disconnected from charge"
11:02:25[Saint]there's also (or should be) "top time": "This item shows the cumulative overall runtime of your player since you last manually reset this item. A manual reset is done through pressing any button, followed by pressing Select or Next."
11:05:50copperyeah I reset them both
11:05:57coppertrying to determine battery life
11:06:07 Join ender` [0] (~ender1@2a01:260:4094:1:42:42:42:42)
11:06:12 Join wodz [0] (
11:06:23[Saint]run a battery benchmark.
11:07:41[Saint]also, check if the runtime is displayed correctly ina current development build:
11:07:50[Saint]if not, please report it on flyspray
11:08:03***Saving seen data "./dancer.seen"
11:09:09[Saint]copper: ^
11:10:17copperI'm using a custom build that's supposedly current
11:11:01wodzAFAIK classic doesn't have CURRENT* defines set so at least remining time will be -1 or something
11:12:03[Saint]copper: then, please don't report bugs on said build ;)
11:12:18[Saint]please, test and verify with the current *rockbox* build.
11:13:16[Saint]Mr. Jimmy Johnson's random build *may* be clean git HEAD, but, it likely isn't...if it was, it wouldn't need to exist :)
11:13:37copperit's got a 10 band EQ patch which I really like
11:13:42[Saint]the link above always points to the current build no matter what.
11:15:08pamauryJdGordon: ping
11:16:43copper[Saint]: bookmarked
11:22:56[Saint]copper: yeah, it's not publicly facing anywhere afaik, due to the port status, but it is a handy link for those that have already "taken the plunge" to have.
11:23:29[Saint]I really need to poke those emCORE boys into putting it up somewhere in the installation instructions
11:23:53[Saint](since, you guys *must* be coming from emCORE...)
11:24:41max3can someone help me out? i have an old school mini with rockbox on it with a cf card and it's on the fritz. i get the !folder icon on boot and then it resets. if i try to put it in disk mode it'll reset out as soon as it's recognized by my computer. i can access diag mode
11:25:10copper[Saint]: I thought there was talk about promoting the iPod Classic port to "unstable"
11:25:49copperfrom my personal experience, it's mostly the USB part that needs some work
11:26:46Torneit's not promoted because it's not possible to install it using our tools or following our instructions
11:27:10[Saint]copper: it needs a Rockbox bootloader for that.
11:27:24[Saint]Which, also means RbUtil support.
11:27:38copperI see
11:27:39[Saint]emCORE is far to non-user-friendly presently.
11:28:20[Saint]a "rockbox bootloader" would likely be a stub of emCORE itself, sand "facy-stuffs"
11:28:32coppersans* :)
11:28:53[Saint]hey, the s and the d are like *right* there, ok? :P
11:30:00[Saint]ideally, all the end user would need to do is place the device in DFU mode and point the installer (RbUtil) could take over from there.
11:30:08copperthat would be nice
11:30:10 Quit derk0pf ()
11:30:34[Saint]it's just one of those 'round-toit things ;)
11:30:35copperwhy hide the rockbox builds though
11:30:51copperemcore installation is easy enough for those who really want it
11:31:03[Saint]copper: there's no reason to make them publically available, really.
11:31:12[Saint]there's no way of installing them officially ;)
11:31:53[Saint]keep in mind that although emCORE and Rockbox are closely related, and share developers, they are two entirely seperate projects.
11:31:53copperwhat is it? You don't want people to come here and complain?
11:32:35copperI'm not complaining personally
11:32:47copperquite the opposite
11:34:27[Saint]Oh, I know. I'm just explaining. There's no way to install the port "officially", so the build needn't be publicly facing.
11:35:06[Saint]and, it isn't "hidden", it just isn't public. ;)
11:35:26*[Saint] takes off the pedantic hat
11:43:55soapI just removed a subnet ban on *
11:44:29soapI had /intended/ it to be a 14 day ban, 8 months ago. Clearly I had checked the wrong box.
11:45:23*[Saint] vaguely remembers that ban.
11:45:24soapI'll be traveling the next 48 hours, but I just want to point out there have been 15,197 hits on that trigger. While it seems harsh to perm ban an entire subnet, we may want to actually talk about it.
11:46:05[Saint]15,197 total, or, since "just now"?!?
11:46:21evilnick[Saint]: total, I hope!
11:46:28*[Saint] too...
11:46:43soap15K since January
11:47:06soapmind you that's a thousand times the average hits on one of my experimental IP bans from the same period.
11:48:34 Join derk0pf [0] (
11:48:35[Saint]yay for hundreds of thousands of unpatched windows machines!
11:49:35soapI'm late. Does someone mind posting that tidbit to the forum thread so it doesn't get lost? And so people know what's going on if I unleased a torrent of new spam?
11:49:57soapif not I'll do so when I land
11:49:57[7][Saint]: the reason why I didn't release an updated rockbox build on the emcore site yet is that those have broken USB :/
11:50:25[Saint]...they do?
11:50:32[Saint]it "just works" here.
11:50:47[Saint]should it not?
11:53:39copperrockbox often freezes up after unplugging the iPod from USB
11:54:01copperor crashes
11:54:50[Saint]that's odd, mine regularly has contact with various linux distros (narry a single windows box, though), and USB "just works" in git HEAD
11:55:01[Saint]I had no idea this was the case, apologies.
11:55:17 Join draft [0] (
11:55:38 Nick draft is now known as Guest92429 (
11:56:33Guest92429i have the latest rockbox on my iPod Classic 160GB... when i play S3M (it uses MikMod I think) the songs are played too fast.
11:56:39copperI have yet to determine if USB2 works better than USB3
11:56:52Guest92429like 2x to normal or something. has anyone noticed the same
11:57:58Guest92429eg. from here
11:58:10Guest92429my friend Jonne did those songs :)
12:00:25gevaertscopper: no rockbox player does USB3
12:01:05coppergevaerts: I meant plugging it into a USB3 port
12:01:28gevaertsYes, but that *shouldn't* make any difference
12:01:33coppershouldn't :)
12:01:47[Saint]that would essentially be exactly the same as plugging it into a USB 2.0 port.
12:02:22[Saint]USB 3 just rolls back to 2.0 when a device can't do USB 3
12:02:52wodzgevaerts: I guess you didn't have a chance to look into elfloader makefiles cleanup?
12:02:52copperno chance my hardware/software could make rockbox fail?
12:03:14gevaertswodz: sorry, no. I forgot about it... :\
12:04:07wodzgevaerts: you have a great chance still... and testing also
12:04:49wodzgevaerts: besides it would be nice if you could express your feeling about such feature on ML
12:06:29*gevaerts nods
12:09:47 Quit fs-bluebot (Ping timeout: 256 seconds)
12:10:52 Quit bluebrother (Ping timeout: 264 seconds)
12:11:11 Join fs-bluebot [0] (
12:12:40 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
12:13:20 Join T44 [0] (
12:14:30 Join einhirn [0] (
12:17:08 Quit Topy44 (Ping timeout: 244 seconds)
12:23:40wodzspam hit in General Discussion
12:30:44user890104is there a way to make an official translation of the rockbox user manual, and add it to the repository?
12:30:59user890104i only see an english one available
12:33:47wodzsure - translate and post to gerrit
12:34:13wodzthe truth is that it will fast go out of sync though
12:39:15 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
12:40:36Guest92429can anyone help me out with the module player? it is playing the modules too fast. is this a known bug? or is there workaround that?
12:48:13pamauryGuest92429: I don't know, have a look in the bug tracker and if not file a bug report with the file if possible
12:53:37wodzpamaury: what are you working currently on?
12:54:40pamauryseveral things: remove the fuze+ audio pop, reworking the touchpad handling and from time to time hacking/reverse engineering/dismantling a sony nwz player
12:55:00pamaurydo you need help for something ?
12:55:41wodzrk27xx usb as usual :-) Seriously, no I was just curious.
12:56:23pamauryI've not tried again rk27xx, still not able to get the setup interrupt. I'll go back to it at some point but I'm kind of bored :)
12:57:08pamaurymy priority now is really to remove this fuze+ pop because lots of people complained about it
12:59:24pamauryI hope the elfloaded will make it into the trunk soon :)
12:59:47wodzI doubt it considering lack of feedback
13:00:40wodzI am kinda discouraged by the lack of interest
13:03:33pamauryyeah, it happened to me several times, that's not a nice situation when you have done some work and nobody seems to care
13:08:05***Saving seen data "./dancer.seen"
13:14:58pixelmauser890104: not easily, as the manual is also source code (latex). This gives us the possibility to have the complete manual as one code and set the pieces together for the various players to match their hardware capabilities. I'm not sure how you could have a translation of that
13:21:22 Join amayer [0] (
13:28:33 Quit max3 (Ping timeout: 256 seconds)
13:35:22 Join dfkt [0] (dfkt@unaffiliated/dfkt)
13:37:35 Join nateloaf [0] (
13:39:32kugelwodz: I'm very interested, but my lack of time prevents me to show it better
13:42:30mortalishow could I determine battery_level_dangerous?
13:44:17wodzmortalis: This should be more or less similar on all targets utilizing li-ion/lipo batteries
13:45:48wodzanyway it should be the level which allows clean shutdown still
14:02:31mortalisit's bit annoying when device doesn't save last track and position. There is no big difference between dangerous & shutoff for hm601 (3350 and 3300 respectively), so I thought maybe I could just make them equal.
14:04:02mortalisguess i need to test it.
14:05:30wodzthis are kinda low
14:06:01amiconnwodz: Just to mention it - I *am* interested in the elf loader stuff
14:06:44wodzmortalis: dangerous level should be around 3500 - 3600
14:06:44amiconnI would have expected something more lightweight than elf, but your binsize comparisons show that elf is apparently lightweight enough (which I didn't exect)
14:07:17wodzamiconn: elfloader on SH is about +1kB in core
14:08:34wodzI mean with patched 4.6.3 and -flto
14:08:37wodzand -Os
14:10:25kugeldo you have un-cheated numbers? :)
14:10:33kugele.g. with the same set of options
14:11:17mortaliswodz: 3500 - 3600 is about 20%-40%. And there are several targets with 3380 dangerous level (tcc77x, ipod 1g, 2g, ... )
14:13:24wodzkugel: I comparision master - elf is with the same options. Different targets use different toolchains though
14:14:25kugelwhat is the +1kB compared against? I assumed git master (which doesnt use -flto or -Os)
14:14:26wodzmortalis: It is more of philosophic dispute now but you should not generally drain li battery so much - it is destructive to internal structure
14:15:05kugelamiconn: he tried something more lightweight before, it turned out too leightweight
14:15:37kugelit didn't support or dram/iram schemes and architecture details very well
14:15:44Torneamiconn: we don't have to handle the vast majority of the features of ELF. The size increase to the actual plugins on disk is considerable (though not that bad), but the code is not that complex
14:15:54Torneour loader is nowhere near as capable as :)
14:16:21wodzkugel: delta +1kB is the comparision of master and elf compiled with toolchain foreseen to be used if elfloader is merged
14:16:43Tornethere are theoretical hacks we could do to reduce the plugin on-disk sizes more, but they require toolchain trickery
14:16:53Tornewe can always do them later if we decide it's worthwhile, they wouldn't require any changes to the loader
14:17:16kugelbut I maintain that if sh cannot use the new loader (i.e. it needs another set of ifdefs and maintanance of legacy code) it should be finally dropped
14:18:09wodzkugel: I didn't test it yet but with V18 you can build *any* target with either loader you like
14:18:24wodzI mean on SH
14:18:42kugeli don't care
14:18:45 Quit derk0pf (Ping timeout: 245 seconds)
14:18:59kugelsh/hwcodec/charcell is a huge burden already without the loader
14:19:26kugelit just gets worse if we still include the legacy loader just for those
14:20:21kugelI don't actually want that we keep the old loader
14:20:24wodzThats why its good that amiconn foresee new loader potentially for SH
14:20:55 Join derk0pf [0] (
14:21:06 Quit wodz (Quit: Leaving)
14:22:16 Join petur [0] (~petur@rockbox/developer/petur)
14:22:26mortaliswodz (log): shutoff approximately same as in OF. And for now hm601 has about 8 hours runtime. If I set it to 3500-3600 it would be 5-6.5 hours. I don't think that users would be happy with that.
14:24:34amiconnkugel: git master *does* use -Os on SH for ages
14:27:02funmanbtw we could have a build with debug symbols
14:27:59funmanPanic: 0x12345678 -> arm-...-addr2line -e .rockbox/rocks/foo.rock 0x12345678
14:28:24 Quit |akaWolf| (Quit: my exit)
14:33:08 Quit derk0pf ()
14:34:34 Quit melmothX (Quit: #)
15:04:35 Quit kugel (Remote host closed the connection)
15:04:39 Quit mortalis (Quit: Leaving)
15:08:07***Saving seen data "./dancer.seen"
15:15:58 Join sentriz [0] (~Senan@
15:34:13 Join wodz [0] (
15:34:59wodzfunman: We could but address from panic will not be something you could pass to addr2line since elf rocks are 0 based
15:35:40wodzYou will have to know base address where plugin was loaded
15:35:49Torneyou could allow for that in the panic handler fairly easily
15:36:04Tornejust look to see where the plugin was loaded and if it's inside that range
15:36:11Torneand if so subtract the base address
15:36:17Torneor just display the base address as well
15:36:21Tornemaybe even filename? :)
15:36:58Tornei dunno how useful it would be unless we are also going to have symbols for core.
15:41:00wodzkugel: For SH I compared master compiled with -Os -flto with gcc-4.6.3 with elfloader build compiled with -Os -flto with gcc-4.6.3 - I don't feel I cheat here.
15:42:37 Part LinusN
15:44:40 Join LinusN [0] (
15:46:27[Saint]wodz: fwiw, I don't think it is a lack of interest
15:47:00[Saint]I think you're just playing in vastly unfamiliar territory.
15:48:34wodz[Saint]: I am not asking how to implement something - I am asking IF we are interested to have it. Thats big difference and only 3 people answered
15:51:35Tornewodz: well, you also know that i think we should have it
15:52:11wodzok, thats 4
15:53:28[Saint]if it encourages forking charcell and hwcodec, count me in too ;)
15:54:28[Saint]heh...but, seriously. I am interested, I just had the feeling you were looking for technical input, not "should we do this/do we want this?"
15:55:49[Saint]from what I am aware of with benefits vs. drawbacks, I wasn't even sure it was in any dispute.
15:57:30wodzyou are exactly right - there was no any dispute :-)
15:59:51wodzok I stop complaining
16:00:34[Saint]In a way, no input is good input.
16:00:56 Join XavierGr [0] (XavierGr@rockbox/staff/XavierGr)
16:01:19[Saint]If only simply for the fact that no input != negative input.
16:01:55wodzI don't feel thats the way to communicate within the project
16:03:38[Saint]well, no, of course not. but perhaps people feel that if they're not qualified to give technical input that "hey, good idea" will be of little use.
16:04:20 Quit sentriz (Ping timeout: 252 seconds)
16:04:36[Saint]...I know I do. Perhaps I shouldn't.
16:05:13[Saint]I haven't overlooked the work, it is just quite overwhelming for me and I'm sure a few others.
16:06:08 Quit tchan (Ping timeout: 260 seconds)
16:07:31 Quit Clear_runway (Ping timeout: 255 seconds)
16:15:58 Join tchan [0] (
16:15:59 Quit tchan (Changing host)
16:15:59 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
16:22:49 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
16:31:40 Join RAThomas [0] (
16:35:33 Quit wodz (Quit: Leaving)
16:41:47 Quit zoktar (Quit: -)
16:50:15 Join zoktar [0] (~zoktar@unaffiliated/zoktar)
16:51:49 Join pretty_function [0] (~sigBART@
16:54:10 Quit the-kyle (Ping timeout: 268 seconds)
16:57:51 Quit pretty_function (Remote host closed the connection)
16:58:36pamauryTorne: there ?
17:00:35pamauryour arm-elf-eabi-as doesn't accept this instruction on arm1176zfs-s: LDRPC, [SP+8+var_8_11874],#8 It says something about ']' expected, any hint ?
17:02:02Torneyou can't have an offset and a postincrement
17:02:20pamauryhum, this instruction is produced by ida-pro when disassembling :-/
17:02:25 Join mortalis [0] (~mortalis@
17:03:14Torneah. is var_8_11874 -8 by any chance?
17:03:16Tornei.e. zero
17:03:20Tornei.e. there is no offset
17:03:36Torneit's just labelling it like that so it looks like a local variable ref
17:03:55pamauryyes it's -89
17:04:15pamauryapparently the assembler is too stupid :(
17:05:11pamauryit also doesn't like the "ret" instruction, I thought it would have some kind of macros for that
17:05:29Tornethose macros are in arm's realview toolchain, i think
17:05:37Torneif you switch IDA to gnu assembly it won't emit them
17:05:45Tornei never have any of those enabled in IDA
17:06:00pamauryit's already emiting for gny assembly !
17:06:30pamauryI don't have a very recent version of ida though
17:07:14 Join the-kyle [0] (
17:08:03TorneIDA's assembly is kinda shit anyway
17:08:09Torneand is not really meant to be assembleable
17:08:11Tornejust readable
17:08:13***Saving seen data "./dancer.seen"
17:09:31 Join pretty_function [0] (~sigBART@
17:12:53 Quit KiwiCam (Ping timeout: 240 seconds)
17:13:48 Quit [Saint] (Ping timeout: 260 seconds)
17:14:46pamauryi'm kind of stuck, I don't what is wrong: my decrypting program or the key or my assumption about the underlying hardware encryption algorithm
17:14:53pamaury*don't know
17:18:28 Quit the-kyle (Quit: Leaving.)
17:19:33 Join the-kyle [0] (
17:21:26 Join [Saint] [0] (~quassel@rockbox/user/saint)
17:22:03 Join KiwiCam [0] (~quassel@
17:22:54 Quit petur (Quit: *plop*)
17:23:20 Part LinusN
17:25:50 Join lebellium_ [0] (
17:26:45 Quit Zagor (Quit: Clint excited)
17:27:13 Quit Guest92429 (Ping timeout: 260 seconds)
17:27:41 Quit lebellium (Ping timeout: 256 seconds)
17:27:45 Nick lebellium_ is now known as lebellium (
17:31:49 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
17:38:52 Join lebellium_ [0] (
17:41:51 Quit lebellium (Ping timeout: 256 seconds)
17:42:01 Nick lebellium_ is now known as lebellium (
17:52:56 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:00:34 Quit nateloaf (Quit: Leaving.)
18:02:01 Join simonnn [0] (
18:08:28 Quit [Deathrow] (Ping timeout: 260 seconds)
18:11:21 Quit ender` (Ping timeout: 246 seconds)
18:12:22 Quit pretty_function (Remote host closed the connection)
18:14:39 Quit ser (Ping timeout: 246 seconds)
18:15:29 Join ser [0] (~ser@2610:28:3090:3001:a:a:a:beef)
18:16:50 Join ender` [0] (~ender1@2a01:260:4094:1:42:42:42:42)
18:37:19 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
18:38:01 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
18:43:40 Quit mortalis (Quit: KVIrc 4.1.3 Equilibrium
18:48:16amayerwhat does the red bar on this patch mean?
18:48:20amayerit says something about rebasing. but i dont want to push another amend and lose the vote that is there.
18:49:12Torneit means the patch you based it on is out of date.
18:49:52Torneand therefore there's no guarantee that it can be submitted without conflicts
18:49:56Torneyou should rebase it
18:50:38Tornewell, at some point before it's submitted.
18:50:39amayerso "git rebase master"?
18:50:41amayerthat patch is in another branch on my computer.
18:51:06amayerwill you review it if i rebase it quick?
18:51:07Torneer, they apparently weren't in different branches, because one is based on the other
18:51:08 Join teachkindness20 [0] (
18:51:26amayeri pushed them from seperate branches
18:51:34amayerthey are 2 branches on my computer
18:51:45Torneyes, but the branch with the second patch contains both patches
18:52:00 Quit teachkindness20 (Client Quit)
18:52:12Torneyou started the second branch from the first patch, not from master :)
18:52:21amayerohh... i made the first change in master and pushed that
18:52:23amayerthen started a new branch
18:52:35Torneyou need to tell it what you want the branch to be based on
18:52:41 Join prof_wolfff [0] (
18:52:43Torneotherwise it uses whatever you have checked out right now
18:52:57amayerdamn it... i will get a hold of this whole git thing.
18:52:58Torneto make a branch for an isolated patch from the current latest code you want:
18:52:59amayeri see what i did wrong now
18:53:05Tornegit checkout -b some_name -t origin/master
18:53:16amayerahh. that seems handy
18:53:23Torneto make a new branch called some_name that starts from whatever origin/master currently is (i.e. the latest upstream) and tracks it
18:53:34Tornethen git status will tell you how many patches ahead/behind origin/master you are, also
18:53:41Torneso you can see how many local patches you have easilt
18:53:53amayerwell i will rebase and resubmit
18:54:14Torneto make them independant now you need to do something more complicated
18:54:24Tornenow that the other patch is part of that branch's history, rebasing will not remove it
18:54:39Tornerebase doesn't delete things
18:54:46Torneit just moves you to a later version of the branch you were based on
18:54:53amayerim not trying to fix it. i will just do it the right way in the future
18:55:06Torneit's not hard to fix, you just need a different command
18:55:15amayerif i rebase and resubmit will it fix the current problem?
18:55:17Tornegit rebase −−onto=origin/master HEAD^
18:55:27Tornewill take just the top patch on the current branch and rebase it onto master
18:55:31Torneand throw any other changes away
18:56:09Torneamayer: it's not a problem as such
18:56:12amayerso i checkout NEW_FEATURE then run that ^ git code
18:56:32amayerand the patch will be independent of master
18:56:51Torneof your local master branch, yes
18:56:58Torne"master" is just a name
18:57:05Torneit's the same as any other local branch
18:57:14Torneorigin/master is the actual code on the server
18:57:23Torne(as of the last time you ran a command that updated it)
18:58:42Tornewe should maybe try and get the trivial rebase hook set up on our gerrit, which will let you carry forward approvals when you rebase if the diff is the same
18:58:47Tornebut we don't habve that in place right now.
18:59:09Torneit doesn't matter a great deal; +1 is not sufficient to commit anyway and it's still visible in the change's history
19:00:25Tornei'm just reviewing your other patch (the "get latest theme" one
19:01:51amayerTorne: the git code you posted didnt work. it output the "proper syntax" of how to use the git command
19:03:06Tornethen read it :)
19:03:50amayeroh... just no "=
19:04:08amayer" inbetween −−onto and origin/master
19:04:09Tornei'm not quite entirely infallible :p
19:05:18amayerthanks for your help tho
19:05:44Tornei've committed that one
19:05:52 Quit kevku (Quit: KVIrc 4.2.0 Equilibrium
19:06:00Tornethe other one i'll look at later if i have time, i need to leave shortly and i can't get familiar with that bit of code so fast :p
19:08:16***Saving seen data "./dancer.seen"
19:08:36 Join derk0pf [0] (
19:27:22amayerTorne: its up when ever you get a chance
19:38:11 Quit tjb0607 (Ping timeout: 256 seconds)
19:39:13 Join tjb0607 [0] (~tjb0607@
19:40:44 Quit jm (Changing host)
19:40:44 Join jm [0] (jm@unaffiliated/nsx)
19:45:34 Quit akaWolf (Ping timeout: 240 seconds)
19:46:03 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
19:50:43 Join Wardo [0] (
20:04:08 Join Deathrow [0] (
20:19:29 Join speckmade2 [0] (
20:19:30 Join pretty_function [0] (~sigBART@
20:21:18 Quit speckmade (Ping timeout: 255 seconds)
20:22:59 Join pamaury_ [0] (
20:23:01 Quit pamaury (Read error: Operation timed out)
20:27:26 Quit Wardo (Quit: Blarglarg)
20:29:56 Quit pretty_function (Ping timeout: 265 seconds)
20:34:27 Quit tjb0607 (Ping timeout: 260 seconds)
20:35:16 Join tjb0607 [0] (~tjb0607@
20:42:59 Quit jm (Ping timeout: 252 seconds)
20:48:02 Quit benedikt93 (Quit: Bye ;))
20:52:11 Quit akaWolf (Quit: my exit)
20:54:54simonnnMy clip+ froze and the screen went black; at some point I got an ata error -2, but then nothing again... fdisk gives a 4mb device (without partition table)
20:55:03simonnnand unbricking didn't seem to help either
20:55:24simonnnDoes anyone have a hint what else I could try?
20:57:33 Quit factor (Read error: Connection reset by peer)
20:59:42 Quit y4n (Quit: PANTS OFF!)
21:08:18***Saving seen data "./dancer.seen"
21:10:17 Quit XavierGr (Read error: Connection reset by peer)
21:10:57 Join XavierGr [0] (XavierGr@rockbox/staff/XavierGr)
21:13:10funmanlet it sit alone for a while and try again in a few weeks, -2 probably means flash storage is dead
21:13:25funmandid you remove the sd card?
21:14:22simonnnI don't have one
21:14:30 Join scorche [0] (~scorche@rockbox/administrator/scorche)
21:14:43simonnnok I'll try to wait for a while then... thanks!
21:15:49 Join factor [0] (
21:16:37 Quit scorche` (Ping timeout: 256 seconds)
21:34:20 Join n1s [0] (
21:34:20 Quit n1s (Changing host)
21:34:20 Join n1s [0] (~n1s@rockbox/developer/n1s)
21:43:37 Join simonlnu_ [0] (HiXCzxIWDi@unaffiliated/simonrvn)
21:44:06 Quit n1s (Read error: Operation timed out)
21:46:49 Nick pamaury_ is now known as pamaury (
21:47:02 Quit simonlnu_ (Client Quit)
21:47:06 Quit pamaury (Changing host)
21:47:06 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
21:55:40 Join nateloaf [0] (
22:10:07 Quit amayer (Ping timeout: 260 seconds)
22:13:32 Quit derf (Remote host closed the connection)
22:20:09 Nick AlexP_ is now known as AlexP (~alex@rockbox/staff/AlexP)
22:30:12 Join n1s [0] (
22:30:13 Quit n1s (Changing host)
22:30:13 Join n1s [0] (~n1s@rockbox/developer/n1s)
22:38:52 Join jm [0] (
22:41:19 Quit mystica555 (Remote host closed the connection)
22:48:13yuriksJdGordon: what do you think of it now?
22:48:34yuriksmain problem I have is that missing characters on the main font still occupy it's cache space
22:48:41yuriks(they need to be loaded to know that they're missing)
22:50:37 Join mystica555 [0] (
22:53:13yuriksit's also inneficient, lot's of duplicate cache searching in the name of code deduplication, but I'll work on that if you think the way it works now is good enough
22:56:35 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
23:08:20***Saving seen data "./dancer.seen"
23:15:59 Quit ender` (Read error: Operation timed out)
23:16:40 Quit XavierGr (Disconnected by services)
23:16:41 Join XavierGr [0] (XavierGr@rockbox/staff/XavierGr)
23:16:52 Join Guest44874 [0] (
23:17:11 Quit factor (Read error: Connection reset by peer)
23:21:25 Part speckmade2
23:25:22 Quit Gallomimia (Quit: Gallomimia)
23:26:52[Saint]JdGordon: ping
23:28:36[Saint]+anyone else that feels affectionate towards, tied to, or qualified to talk of the subject of: cabbiev2
23:29:08[Saint]pixelma: (perhaps? ^) ping
23:30:35[Saint]I have decided that I really hat the fact cabbie draws the backgrounds for icons in the backdrop itself, would anyone have anything against me changing this behaviour?
23:30:44[Saint]*really hat/really hate/
23:31:52 Quit ender1 (Quit: Why do people give each other flowers? To celebrate various important occasions, they're killing living creatures? Why restrict it to plants? "Sweetheart, let's make up. Have this deceased squirrel.")
23:31:56[Saint]sorry, backgrounds for icons and progress/volume bars
23:33:08 Join ender` [0] (~ender1@2a01:260:4094:1:42:42:42:42)
23:33:32 Join factor [0] (
23:33:42 Quit factor (Read error: Connection reset by peer)
23:34:07 Join factor [0] (
23:34:25[Saint]changing this behavior, with the addition of splitting each theme element out to its own viewport instead of defining everything at offsets reletive to the global fullscreen viewport (which I think is very ugly), would allow for:
23:35:03[Saint]1 - make the tansition to themes that can resize dynamically easier, and
23:35:39[Saint]2 - make it easier to implement an idea I'm playing with where theme elements can be toggled on/off by a "magic file" or <some_other_trigger>
23:37:55[Saint]Other people interested in the default theme and the theme engine in general should also loog at the logs for the conversation JdGordon and I had last night (my time) about splitting out metadata parsing (optionally) to a plain text file to allow the user to decide what gets displayed throughout their themes more easily.
23:39:30[Saint]see here:
23:43:08AlexP[Saint]: Sounds a good idea
23:43:45[Saint]which? (I'm assuming the latter, linked one)
23:43:55AlexPwell, sort of both
23:44:01AlexPThe former certainly
23:44:03[Saint]Ah, cool. :)
23:44:17AlexPThe latter I haven't really thought about, but at first impression seems reasonable
23:45:22[Saint]what I'd like to happen is that we have one file for all the cabbies (the cabbie ports can display between 3 and 11 lines of metadata in theory) and we order them in such a way that the most needed data is presented first.
23:46:29[Saint]I could /probably/ make some magic to decide what to display based on how many lines can be displayed, but I really don't want the theme to be aware of how many lines of metadata it could/should display
23:46:40AlexPI'm not too keen on information overload, but I guess you'd sort that by adjusting the font/spacing
23:46:50[Saint]rather I would like it to just display as many full lines as the viewport calling it can muster.
23:47:01[Saint]AlexP: indeed, yes.
23:47:45 Quit bertrik (Ping timeout: 244 seconds)
23:48:14[Saint]I don't want to jam pack all possible metadata in there, but I want some form of ordered list so that the devices that *can* display some more info will, without affecting those that can't.
23:48:39AlexPAnything that works towards dynamic resizing is good IMO
23:49:17AlexPBring on the 2560x1600 theme :)
23:49:31[Saint]is that the Nexus 7?
23:49:33 Join lebellium_ [0] (
23:49:35AlexPnexus 10
23:50:29[Saint]actually, there's a change I have stashed around here that I would like to push to gerrit re: RaaA, need to find it.
23:50:44AlexPIt'd be good to get stuff on there in general
23:51:05AlexPThen people can see it, I for one am rubbish at going to external sites
23:51:08[Saint]I got theme centering for hosted targets working.
23:52:09AlexPback in a mo, need to reboot server
23:52:10[Saint]so that instead of aligning to 0,0 if the theme is smaller than the device resolution, it attepts to center it (with a preference towards the top and left in the case where it can't be centered exactly)
23:52:36[Saint]need to find what brach it's stashed away in.
23:52:38 Quit AlexP (Read error: Connection reset by peer)
23:52:47 Quit lebellium (Ping timeout: 260 seconds)
23:52:56 Nick lebellium_ is now known as lebellium (
23:58:44 Join AlexP [0] (~alex@rockbox/staff/AlexP)

Previous day | Next day