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 2010-12-06

00:00:06pixelmaabout the rockbox calendar in gmail
00:00:20pixelmaeven without account?
00:00:22n1sbut flac needs to read much more data from the flash which also costs energy
00:00:37AlexPOh, I don't know if there is a web interface
00:00:48gevaertsmusepack seems to be better at some bitrates than mp3. Not worth switching for that though I think
00:01:04marinesok, thanks
00:01:19gevaertsEspecially since it's worse at other bitrates
00:01:32n1slloking at the gigabeat F numbers mpc is ~9MHz faster than mp3
00:01:32AlexPpixelma: But I think there is a way
00:01:59AlexPpixelma: yes you can, one mo
00:02:22n1son the as3525 that page claims the 320kbpsmp3 is faster than all other mp3 files
00:02:40pixelmaAlexP: I think other google "services" like spreadsheet are available without the account (maybe setable by the starteer) but am not sure about others
00:03:02the_KyleDoes the UI simulator support spoken menus and .talk files? Since I'm unable to use the utility at this point, I'd like to be able to test a command line script to create proper talk files.
00:03:17n1sthe_Kyle: yes
00:03:29AlexPpixelma: Let me know if that works for you, then I'll add to my mail
00:03:30pixelmaAlexP: ok, that works. Maybe this should be sent to the ml too?
00:03:53the_Kylen1s: Thanks.
00:04:16pixelmaAlexP: I only see a blank month though. I assumed nothing was set for December yet
00:04:39AlexPpixelma: try february, there should be something there
00:05:00pixelmayes, there are dates for 3.8
00:05:17AlexPIf you go the other way, 3.7 should be there too
00:05:45AlexPmaybe only 3.7.1
00:05:50AlexPI can't remember :)
00:05:56pixelmaI see 3.7.1 but not the one before
00:06:05gevaertsWe'd add the point releases, but they can change quickly
00:06:13gevaertsI don't think we added 3.7
00:06:27gevaertsI added 3.7.1 to see if it worked
00:06:37AlexPNo, I think I added 3.7 in my previous demonstration, but didn't bother for this
00:06:55*pixelma bookmarks
00:08:17***Saving seen data "./dancer.seen"
00:09:03 Quit Kupop (Ping timeout: 245 seconds)
00:12:00 Quit froggyman (Quit: Bye)
00:15:06 Quit bertrik (Quit: :tiuQ)
00:15:14 Quit n1s (Quit: Lmnar)
00:15:42 Part marines
00:24:38 Join JdGordon| [0] (
00:24:38 Quit JdGordon| (Changing host)
00:24:38 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
00:28:25kugelgevaerts: I extracted pth's aproach and it somehow works
00:28:44kugelI understand what they're doing, and according to the comments in their code I'm now a elite hacker :P
00:28:54gevaertsgood :)
00:28:56JdGordon|kugel: it does sound backwards, but we could do a "get_plugindata_dir()" function for just plugins and do it a bit simpler than what is there now
00:28:59 Quit Kitar|st (Ping timeout: 255 seconds)
00:29:06JdGordon|char* get_plugin_data_dir(void)
00:29:24the_KyleIs it possible to load a prebuilt Rockbox release into a simulator for its target, or do I need to checkout everything from svn and build? If I need to checkout from svn, do I just use the 20101126 date to get the 3.7.1 release from trunk, or does the Rockbox svn have a branch for each release?
00:31:06gevaertsthe_Kyle: there are pre-built simulators (for windows) at">
00:31:06JdGordon|you have to build the sim for the target
00:31:12JdGordon|its a simulator not an emulator
00:31:36gevaertsThe changes between 3.7 and 3.7.1 are minor and won't affect voice files
00:32:00the_Kylegevaerts: Problem is I'm running on Linux and need to build the sim there.
00:32:04gevaertsah, ok
00:32:28gevaertsthe_Kyle: you need the svn:// branch then
00:32:55 Quit mt (Read error: No route to host)
00:33:29 Join Kitar|st [0] (
00:34:51the_Kylegevaerts: Thanks. And I believe I'm understanding correctly that I will need to build the entire firmware and not just the simulator, since it's not an emulator, and my prebuilt firmware was built for a specific target and not for the simulator?
00:35:13 Quit pamaury (Remote host closed the connection)
00:35:33gevaertsthe_Kyle: you need to build (and install) the simulator. That's it
00:36:48the_Kylegevaerts: Ah. Got it. Thanks for the help. And n1s as well.
00:40:57kugelgevaerts:, does that work on your n900?
00:46:15 Quit kevku (Quit: KVIrc 4.0.2 Insomnia
00:46:39kugelJdGordon|: can we change PLUGIN_DIR for plugins (so that it contains /.rockbox) and use another macro for the core? that'd be less changes I think
00:47:23kugelunless plugins need to know the path where the .rocks are..
00:48:44 Quit ender` (Quit: error compiling committee.c: too many arguments to function)
00:51:45 Join designate72 [0] (
00:53:15JdGordon|kugel: I'd prefer PLUGINS_DIR and PLUGINDATA_DIR personally
00:53:21JdGordon|even though it is slightly more changes
01:07:19 Join user890104 [0] (
01:07:55the_KyleIs it possible to create a new voice file without using the utility? I'm getting it to speak, but I'd rather have it speak with a different synthesizer.
01:15:51DreamxtremeSo is noone creating a port for the classic then ?
01:18:49 Join shai [0] (
01:24:56kugelgevaerts: success \o/
01:27:06kugelgevaerts: I uploaded a second zip (same url), it now has a *context()-like API
01:31:45 Quit shai (Ping timeout: 259 seconds)
01:33:04 Join shai [0] (
01:45:32rasherthe_Kyle: Yeah, you can build voice files as part of the build system
01:46:02rasherAssuming you can use your synthesizer from perl somehow, you could add support for another synthesizer
01:46:35rasherSeems the wiki has been scrubbed of mention of how to do this
01:46:38*rasher frowns
01:47:35kugelgevaerts: I suppose I should rather use that instead of *context then
01:48:32gevaertsHave you tested it in rockbox?
01:48:40the_Kylerasher: Not sure if I can use it from Perl or not, although I know I can use system commands. I don't think either of the synthesizers I'd like to use have perl bindings directly.
01:49:00rasherthe_Kyle: system commands can be called from perl :) That's good enough
01:49:06kugelgevaerts: no
01:49:08rasherthe_Kyle: check tools/
01:49:20gevaertsBetter do that first then :)
01:49:49kugelgevaerts: I think the signaling stuff requires larger stacks, at least sigaltstack returned ENOMEM with a small stacks
01:49:58*the_Kyle checks the voice tool. I'm sure I can at least get eSpeak to work. Kinda like SVox Pico for this, but either will do.
01:50:18 Quit DerPapst (Quit: Leaving.)
01:50:26rasherthe_Kyle: espeak is already available (and should be in rbutility as well, in fact)
01:50:37rasherthe_Kyle: the build system has a path for selecting a voice build, which you then end by "make voice" which should give you a voice file
01:52:31kugeland it looks like a large hack too :)
01:57:46 Quit designate72 (Read error: Connection reset by peer)
01:58:23kugelah, MINSIGSTKSZ
01:59:25 Quit kugel (Remote host closed the connection)
02:04:47 Join mystica555_ [0] (
02:08:21***Saving seen data "./dancer.seen"
02:17:44 Quit robin0800 (Remote host closed the connection)
02:20:58the_Kylerasher: Any pointers to docs for making this work with eSpeak?
02:21:38rasherthe_Kyle: simply running tools/configure as if you wanted to build Rockbox should let you do that already
02:22:33the_Kylerasher: Oh yes, must be in the questions. I think I missed it, since when I ran it I was focused on building the simulator.
02:22:53*the_Kyle checks the build system for the voice file build
02:23:04rasherI forget how it goes, you may have to select advanced
02:23:13rasherAnd then "make voice", I think
02:23:16rasherIt's been a while
02:24:11the_KyleBy the way, how do I turn off my clip+ simulation? I can't seem to find the key that is mapped to the power button.
02:24:58JdGordon|just close it
02:25:54 Join tvelocity [0] (
02:25:55the_KyleJdGordon: Thanks. Crazy I didn't think of that.
02:33:09 Quit liar (Quit: Leaving)
02:38:39 Quit tvelocity (Quit: Αποχώρησε)
02:40:05 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
02:40:54 Join designate72 [0] (
02:41:35 Join Keripo [0] (
02:57:53the_KyleI thought I got the voice build started: tools/configure, advanced, voice. But it said "voice build selected" and it's just sitting here with a cursor. There's no prompt, and the process seems to be frozen.
02:59:31 Nick evilnick_ is now known as evilnick (
02:59:39 Quit evilnick (Changing host)
02:59:39 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
03:02:43 Quit mystica555_ (Ping timeout: 260 seconds)
03:03:25 Join mystica555_ [0] (
03:06:18the_KyleVoice build process is still frozen, so I think it's waiting for input. Just not sure what I need to tell it. It hasn't asked for anything. There's no hard disk activity of any kind.
03:08:11the_KyleI can confirm that it's waiting for input. There's no help, however. I tried "help" and "?" with only an error message that says the option is unsupported.
03:10:04the_KyleOh. I think I got it. I pressed enter, and now it asks for a language.
03:12:57the_KyleFor future reference: tools/configure, target, advanced, voice, enter, language number, TTS option. Then make voice. If someone could please add that to the wiki documentation. Thanks.
03:16:15the_KyleI would like the ability to change the speed, pitch and gender of the eSpeak synthesizer. Is this possible? Is it just a simple edit of the voice build tool?
03:19:47 Quit Keripo (Quit: Leaving.)
03:21:27rasherthe_Kyle: well you can enter commandline options for espeak during configure.
03:23:13*the_Kyle checks the configure options.
03:26:49 Join froggyman [0] (~seth@unaffiliated/froggyman)
03:27:27the_KyleI need more information about −−encopts. It says it sets encoder manual options. Is this speex options?
03:27:55rasherThough I'm not sure you should be modifying those
03:28:05rasherI think the voice system may make some assumptions
03:28:21rasherCan't hurt to try, if you want to lower the filesize or whatever
03:29:32the_Kylerasher: I want to speed up the voice, which should reduce the size enough. I was thinking of trying to increase quality, since I don't have to worry so much about size if speech is faster.
03:30:17the_KyleSince I'm only running a simulation at this point, I'll go ahead and try it. If it breaks, I can just build again without bricking a Sansa.
03:30:46rasherWell there's no risck of bricking by trying out different voice files
03:30:58rasherBut using the sim is far easier of course
03:31:47the_KyleWell, being blind and having a screen reader on the computer definitely helps. Then I can just move the voice file that I like to the clip+ when I get it working.
03:39:08 Nick noamsml_ is now known as noamsml (~noamsml@
03:44:44the_KyleAre rbspeexenc options different from the standard speex encoder?
03:45:53 Quit designate72 (Ping timeout: 260 seconds)
03:46:11the_KyleMy voice files only use 8KB of disk space if I change the speex options at configure time. I didn't try the simulation, because I think I can safely assume it won't talk if it's that small.
03:49:30the_KyleOh. I think this is going to work. Looks like rbspeexenc only takes the short speex options, and I was trying to set long options as in the speex manpage.
03:49:52the_KyleIt's taking a lot longer to encode this time, which seems to be a good sign.
03:50:59the_KyleOuch! 1.5 MB? Well, my player will be 8 GB, so I'll try the simulation and see if it works. But that bitrate is a bit high, so I'll expect a slowdown.
03:53:27the_KyleCan I use my rbspeexenc to create .talk files?
03:53:53rasherthe_Kyle: I believe rbspeexenc is simply the standard speex encoder with something stripped out
03:54:04rasherSo the answer to your question would be yes
03:54:59rasherthe_Kyle: I don't really know much about that. has a mode for generating .talk files, but it never saw much testing I think. You'd have to invoke it manually
03:55:10the_Kylerasher: So then could I safely pipe espeak through sox to create the speex files and name them with the appropriate .talk extension also?
03:55:37rasherI think you should use rbspeexenc
03:56:16the_KyleThanks. I'll try rbspeexenc and see if it works, although I can't see why it wouldn't work.
03:59:23the_KyleDoes anybody know if rbspeexenc will take a pipe, or do I need to create a temporary wav file?
03:59:27rasherthe_Kyle: something like ./ -C -e=rbspeexenc -E="-rbspeexenc options" -s=espeak -S="-espeak options" /path/to/dir
04:00:55 Quit MethoS- (Remote host closed the connection)
04:01:07 Quit InsDel (Read error: Connection reset by peer)
04:08:24***Saving seen data "./dancer.seen"
04:11:46 Quit TheSeven (Ping timeout: 255 seconds)
04:12:28 Join saratoga_ [0] (9803c22e@gateway/web/freenode/ip.
04:12:32saratoga_AlexP: please ban chen4h8t9l
04:14:38scorchesaratoga_: bah - spammers are still getting through?
04:14:48 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
04:15:16 Join Barahir [0] (
04:16:29scorchesaratoga_: looks like they registered on the 2nd - pre-upgrade
04:16:43saratoga_ah makes sense
04:17:07 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:18:29 Quit Barahir_ (Ping timeout: 260 seconds)
04:19:28saratoga_scorche: guessing hjg76fghf, yNouNE89F, confidb61 and xsoft111231 are probably also bots
04:19:53scorchesaratoga_: sounds like it, but i prefer to wait until after they have actually spammed
04:20:39 Join Topy [0] (
04:20:45saratoga_lots of them have spam sigs: e.g. t.clark70
04:21:14saratoga_this page helpfully shows all the bots since they're the only zero posters with links in their sigs
04:22:28scorchespam sigs is a good enough reason to pre-emptively ban them
04:22:35the_KyleGot it. Works like a dream. CD into the folder and run the files through a for. Works using temporary wav file. Espeak the wav file and then rbspeexenc it. Again more painless than I thought it would be.
04:23:59 Quit T44 (Ping timeout: 245 seconds)
04:24:03*scorche bans a coupel users with spam sigs as listed there
04:24:13*scorche also attends typing class
04:24:29the_Kylefor f in *; do espeak ,voice options> /tmp/temp.wav "$f"; ~/path/to/rockbox/tools/rbspeexenc /tmp/temp.wav "$"; done
04:25:33saratoga_theres hundreds of them, going back weeks
04:25:38saratoga_most of them never activated though
04:25:41saratoga_broken script i guess
04:25:53 Join Keripo [0] (
04:26:14scorchesaratoga_: i guess it might be simpler to just ban them as they pop up then...
04:26:28the_KyleIf I typed that correctly, this is all I had to do to get all the song .talk files. Now I'll read the manual to remember where to put the .talk files for folders, and then I'll have Rockbox running, at least on the simulator, exactly how I want it.
04:26:44saratoga_scorche: apparently bots have been registering for months
04:26:55saratoga_theres way more of them then normal users
04:27:03the_KyleThen I'll be able to do the same on the clip+ itself.
04:27:43saratoga_seems it started in september
04:28:20saratoga_probably about 500 of them total, only a couple dozen of which ever seemed to figure out how to post :)
04:28:34scorchesaratoga_: well, hopefully they never pop up...not much i can do if there are that many of them - luckily, we have plenty of eyes on the forum to spot the moles as they pop up from their hole
04:33:52 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
04:39:24 Quit amiconn (Disconnected by services)
04:39:24 Quit pixelma (Disconnected by services)
04:39:24 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:39:26 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:39:29 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:39:44 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:44:46 Join Xerion_ [0] (
04:46:54 Quit Xerion (Ping timeout: 245 seconds)
04:49:17 Quit Xerion_ (Ping timeout: 264 seconds)
04:50:01 Quit sasquatch (Quit: WeeChat 0.3.2)
04:50:23 Quit fyrestorm (Read error: Connection reset by peer)
04:50:27 Join sasquatch [0] (
04:52:19 Join Xerion [0] (
04:58:42rasherscorche: aren't you still allowing their links though?
04:59:41scorcherasher: yeah - but i dont have the time to go through thousands of bots registered months back right now...
05:00:52rasherMaybe just delete un-activated accounts older than a few weeks?
05:01:31scorchedont they have to be activated in order to have a signature?
05:01:53rasherdunno, I thought that's what saratoga_ was saying
05:02:31scorchethere are currently 788 users awaiting activation from march of 2008 to today
05:03:08saratoga_you could do an SQL query that deleted users with a sig and zero posts, and that would get them, but since they never post no one ever sees their sigs
05:03:49scorchei suppose that is an idea
05:04:17rashersaratoga_: it's still gaining them inbound links
05:04:19the_KyleI'm not able to make Rockbox speak a directory name. I created a file in the directory, just as the voice howto on the wiki says to do, but Rockbox is still spelling the name.
05:04:49rasherthe_Kyle: did you turn on talk clips for folders? I blieve it's a setting seperate from the one for files
05:05:51the_Kylerasher: I did. But I just found my problem I think. I called the file and it needs to be called Sorry, my speech didn't speak the underscore, and somehow I missed it the first time I spelled the name of the file.
05:07:00the_Kylerasher: Yes. That did fix the problem. Sorry for the mixup.
05:08:24*scorche wonders if deleting the entry in the members table is enough to delete the user and not make the forum kill itself
05:09:02JdGordon|doesnt sound very safe :p
05:09:12JdGordon|rasher: you goign to commit the metadata logger thing?
05:09:38rasher5oh er, I guess
05:10:56 Quit saratoga_ (Quit: Page closed)
05:15:57 Join evilnick_ [0] (
05:17:46 Join fyrestorm [0] (
05:18:22 Quit evilnick (Ping timeout: 250 seconds)
05:18:33 Join evilnick [0] (
05:18:38 Quit evilnick (Changing host)
05:18:38 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
05:18:51scorchei guess the easiest way (for me at least) to kill this is to adjust signature settings to include no links, apply it to all signatures, and then post a notice saying that if you had links in your sig, you need to re-add them - odd are, a bot is not going to do that...
05:19:27scorchethough, there still are the websiteUrl links
05:20:43 Quit evilnick_ (Ping timeout: 276 seconds)
05:21:49scorchethere are "only" 861 users that have a websiteUrl - 1347 users with a signature!=""
05:22:01scorchethough i would have though there are a lot more people with sigs than that...
05:22:10JdGordon|please dont
05:22:19JdGordon|is it really that big a deal?
05:25:09the_KyleIs the Clip+ context menu implemented in the simulator?
05:25:52JdGordon|oh lovely... scorche confidb61
05:27:32scorche733 users with no posts and a signature - 574 users with no posts and a website url
05:29:42the_KyleGot it. Hold down select.
05:32:22scorcheso how about if i UPDATE the members table nuking all websiteUrls and signatures in users with zero posts? - there are some valid users in here, but i think that is a decent solution...
05:35:42scorcheUPDATE 'smf_members' SET websiteUrl="", signature="" WHERE posts=0
05:38:36scorcheanyone see any reason why i shouldnt just do that? - the collateral damage shouldnt be *too* bad...
05:38:55JdGordon|whats the point?
05:39:52scorchegets rid of all of the spammy links in those fields from the hundreds of bots that signed up
05:40:52JdGordon|yes but what's the point of doing that? it wont noticaby speed anything up, or reclaim space and the links arent being displayed anywhere
05:41:27scorchethe links are being displayed on their user pages which is generally enough for crawlers to pick up on
05:45:05 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:50:19 Quit antil33t (Ping timeout: 240 seconds)
05:52:17 Join shodan45 [0] (
05:53:20shodan45"Sansa Fuze v1 and E200 v2 now have usb support enabled." <−− what does that mean, exactly?
05:54:03shodan45usb transfer? or usb charging? both?
05:59:34the_KyleIs is possible to simulate an installed microSD card in the Clip+ simulator?
06:01:20evilnickshodan45: At least usb transfer and I'm not sure about charging
06:01:46shodan45evilnick, ah thanks
06:04:25 Quit JesusFreak316 (Ping timeout: 245 seconds)
06:08:26***Saving seen data "./dancer.seen"
06:12:17 Quit Judas_PhD (Quit: This is a quitting message)
06:12:45 Quit Kitar|st ()
06:13:51 Join antil33t [0] (
06:17:10the_KyleI'm getting some strange dump files in my simdisk. Anyone know what these are? The file name looks like like dump 101206-001318.bmp
06:19:53 Quit LambdaCalculus37 (Quit: Sleep time is now)
06:21:29JdGordon|stop pressing f5
06:21:35JdGordon|or num-0
06:25:00the_KyleOops. That's it, although I don't recall pressing either of those.
06:27:01 Join evilnick_ [0] (
06:27:19 Join Judas_PhD [0] (
06:30:23 Quit evilnick (Ping timeout: 255 seconds)
06:38:10 Quit JdGordon (Ping timeout: 265 seconds)
06:40:45 Quit guymann (Quit: world updates)
06:48:29 Join [Saint] [0] (cbb80223@gateway/web/freenode/ip.
07:05:41 Join stoffel [0] (
07:34:15 Join clear_runway [0] (
07:35:54 Quit shodan45 (Remote host closed the connection)
07:37:37clear_runwayanyone home?
07:39:52[Saint]clear_runway: Just ask your question, there's plenty of people around
07:41:57clear_runwaynot really in for a question, just hanging
07:42:15clear_runwayjust found rockbox, incredibly awesome
07:42:19 Quit Judas_PhD (Quit: This is a quitting message)
07:42:47clear_runwayi mean, i had no idea a second gen ipod nano could emulate a game boy and play video
07:44:32[Saint]clear_runway: Is there a question there?
07:47:18clear_runwaynot really. just fan-worship or whatever
07:48:09[Saint]It's not really surprising from a technical point of view...the gameboy/gameboy are hardly power-houses when it comes to performance ;)
07:48:33[Saint]and, the Nano2G is one of the more powerful DAPs that Rockbox supports.
07:49:55clear_runwayjust saying. i'm actually a little more surprised about the video, and it not being supported without RB. apple has some 'splaining to do.
07:51:12[Saint]Not really, it's more of a novelty...watching video on a 176x132 screen is not something that can be done for long periods.
07:51:31[Saint]This I know, as I own several Nano2Gs.
07:52:19clear_runwaystill, I watched the third lotr movie with no problems. legally ripped from my collection and converted.
07:53:02[Saint]It is also very, very common for a manufacturer to produce a device that doesn't do even half the things it is capable of, as then there wouldn't be as much of a market for "the next big thing" later on.
07:53:15 Join MaxHR [0] (
07:53:23clear_runwayyeah, aint that the truth
07:55:14 Quit Keripo (Quit: Leaving.)
07:56:38clear_runwaybut thats what people like us are for
07:57:41timccchas anyone volunteered for the classic1g port?
07:58:13[Saint]timccc: It's not a matter of volunteering ;)
07:58:25[Saint]It's a matter of doing the work ;D
07:59:21[Saint]And yes, there is some work going on for the Classic presently.
07:59:40clear_runway...does anyone have a classic 1g anymore? the failure rate was pretty high, what with the hard drive and the mechanical scrollwheel
07:59:43timccci meant that while i expected many people would be ready to go and do the work, when the7 mentioned it recently seemingly no one responded
08:00:34[Saint]TheSeven was gifted a Classic fairly recently (3 weeks ago?), and is working on it.
08:00:47timcccbut thats great to hear saint. the large capacity ipods are perfect for rockbox :D can't wait.
08:01:33[Saint]clear_runway: I think you're getting your classics confused, not the iPod 1G...the the "Classic" iPod 6th gen
08:01:40MaxHRIs there a way to change the default behaviors for what happens when buttons are pressed in certain modes? I am using the sansa firmware
08:02:16[Saint]Apple dropped the confusion ball calling it a Classic, as the Apple site now refers to the older iPods as "classic" also.
08:02:58[Saint]MaxHR: This isn't a place to discuss the behaviour of the original firmware sorry.
08:03:18clear_runwaywikipedia calls it the "classic 6g" and retroactively applies this label to all previous models, iirc
08:04:00 Quit JdGordon| (Quit: leaving)
08:04:05MaxHR[Saint]: I am using the rockbox firmware
08:04:38MaxHRthe version for the sansa player
08:05:02[Saint]MaxHR: Aha..sorry "I am using the sansa firmware" confused me.
08:05:15[Saint]what is it exactly you're wanting to change?
08:06:02MaxHRfor example, during now playing mode, the down button will bring you back to main menu, I would like it to show current playlist
08:07:44[Saint]You can achieve that quite easily by editing the keymap for your player, but when I say "quite easily" I mean in my opinion...not necessarily in yours.
08:08:30***Saving seen data "./dancer.seen"
08:08:30[Saint]You'll need to download the Rockbox source code, and have a development environment set up to consider doing it.
08:08:52clear_runwaysound like fun...and by fun i mean hell.
08:10:40[Saint]editing a keymap is one of the simpler changes a user with no/limited experience in programming could only seems daunting as people sem to expect to open up the .rockbox folder on the device, change/delete a file...and it's magically done.
08:11:36[Saint]Whereas in reality you need to apply your changes to the source and compile your own build every time.
08:11:41MaxHR[Saint]: ok, so I need to recompile it from source? there isn't just a settings file to edit
08:12:00[Saint]correct, and no there is not.
08:12:22MaxHRmy responses may be delayed to your answer, I am on 28.8 dialup at the moment :p
08:14:20 Join Kupop [0] (
08:14:38clear_runway"magically done" he says. look, that really shouldnt be nessessary. I mean, RB is at version 3.7.1. there shouldnt even be a file you need to edit, it should be in a settings menu
08:15:23MaxHRare there any plans in future to make changing settings easier than having to recompile? Rockbox is nice, but much of the interface to me needs to be freindlier, I have a few ideas on this that I would change right away...
08:15:53clear_runwayof course, thats the beauty of open source
08:16:03clear_runwayif you want it, change it yourself
08:16:44 Join JdGord [0] (~jd@
08:17:06[Saint]MaxHR: The problem with that statement is, you're not changing a "setting"...settings are already configurable on the player.
08:18:20[Saint]I'm sure youi're thinking it's as simple as "make this key do X as opposed to Y", but you also need to consider what to do with the original functions of key X.
08:18:35MaxHRso keymap is not a setting? or the default behaviour of what to do when player first turns on?
08:19:22[Saint]keymaps are not classed as a setting, no.
08:20:30clear_runwayyes, it would be impractical, but it's better than recompiling the whole thing
08:21:10[Saint]No, certainly not.
08:22:35the_KyleListening to MIDI file playback, it seems to work better than I expected, except that it sounds like sustains aren't supported. Is this a known issue?
08:22:41clear_runwayit's just that, it scares people off. most people find it hard enough to edit a config file. text scares people. code doubly so.
08:22:46MaxHRI know nothing of how rockbox is done, but I remember using vdr in linux, changing keymap for the remote control just involved editing one file, no recompile
08:23:29[Saint]clear_runway: you're looking at it like Rockbox is a "product" though, it isn't.
08:23:42clear_runwayI mean, even freaking dwarf fortress has a gui (okay, sort of) for changing keybindings ingame
08:24:00[Saint]It is made by developers, *for* developers...and if everyday-joe uses it and gets a kick out of it, great. ;)
08:24:39clear_runwayalright, fair enough
08:25:15MaxHRdo developers encourage idea concept suggestions from joe users?
08:25:40JdGordDepends on the idea
08:25:52[Saint]I'm not sure they encourage it ;)
08:25:58[Saint]But, they'll listen ;p
08:27:16 Join TheLemonMan [0] (
08:27:31 Quit clear_runway (Quit: Leaving)
08:27:50 Join einhirn [0] (
08:28:14 Quit BHSPitMonkey (Read error: Connection reset by peer)
08:28:54MaxHRI assume forum "Feature Ideas" is the best place...
08:29:52[Saint]Here on IRC, or the mailing list is probably the "best place"...."Feature Ideas" is simply the correct place.
08:29:53 Join ender` [0] (
08:30:40JdGordHere is best
08:32:42MaxHRhmm, ok, well first idea would be to have a button that would bring up the current playlist directly from the now playing screen
08:33:17[Saint]can't you do that with Hotkey?
08:33:36[Saint]matter of fact...isn't playlistviewer the default for WPS in that instance?
08:34:16MaxHRIm not familiar with that (hotkey)... but currently, none of the buttons on my sansa will go directly to the current playlist from now playing screen
08:34:57[Saint]read your players should be Settings - General Settings - Hotkey IIRC
08:35:29[Saint]it *should* be enabled by default also..but you'll need to consult the manual for the correct keymapping
08:35:33MaxHRok, thankyou, will check that out
08:35:39the_KyleI didn't find a hotkey setting, but there is the quick screen, which may do what you want.
08:36:29MaxHRok, second idea would be to auto resume now playing when powering the player back up, assuming there is a current playlist that hasn't yet ended
08:37:35[Saint]Car Adapter Mode
08:38:00[Saint]Settings - General Settings - System - Car Adapter Mode
08:38:19 Quit TheSeven (Ping timeout: 240 seconds)
08:38:37[Saint]that does exactly what you want, playback will resume when the player powers back up again.
08:38:47MaxHRnice, thx
08:39:24[Saint]the_Kyle: you'll find Hotkey settings in Settings - General Settings - Hotkey (assuming you're not using a truly ancient build)
08:41:19MaxHRhmm, I see the wps hotkey setting, currently set to display current playlist... but none of the player buttons do this
08:41:56the_Kyle[Saint]: I have 3.7, but there/s no hotkey in general settings. Could this be because I'm running the simulator for now? Once I have my clip+ and install it there, will I have this setting?
08:42:02[Saint]it's a key combination...not a seperate key
08:42:23[Saint]consult the manual on this one sorry, I don;t have the target so I can't tell you offhand.
08:44:39MaxHRok, thats what I was hoping, that the down key function could be chaned (w/o recompile)
08:45:11[Saint]it will still retain it's original function.
08:45:34[Saint]Hotkey just adds a secondary function to a grouping of keys.
08:46:04 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:46:10 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
08:46:59 Part Zagor
08:47:14 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:47:26[Saint]the_Kyle: Hotkey seems to be disabled in the Clip+ builds for some reason.
08:47:34[Saint]I've no idea why personally.
08:48:06[Saint]None of the Clip(s) use it actually.
08:48:23the_KyleInteresting. Maybe keys can't be grouped on the clip+?
08:48:43the_Kyleor any of the clips?
08:49:12the_KyleSeems like a possible hardware limitation.
08:50:07[Saint]I'm not sure that's it...the Clip(s) all support hardware hold no?
08:50:56the_KyleAs far as I know, they do. Is there a way to enable it at buildtime?
08:51:44the_KyleSorry, I mean the hotkey setting.
08:52:01[Saint]you could simply uncomment #define HAVE_HOTKEY in the players config file.
08:52:26[Saint]yeah, you could uncomment #define HAVE_HOTKEY and see if it "magically works".
08:52:40the_KyleThat was easy. Well, here goes a test.
08:53:01[Saint]I'd assume it's disabled for a reason...but stranger things have happened.
08:53:21MaxHRanybody know where manual tells what the wps hotkey combo is? I am googling and also browsing manual, nothing yet
08:54:28the_KyleWell, I can test here, since it's just the simulator at the moment. Is there anything I need to do differently for that, or would uncommenting the hotkey setting in the clip+ target work on both the real player and the simulator?
08:55:29[Saint]should work for both.
08:55:37[Saint]the_Kyle: ^
08:56:11[Saint]MaxHR: Section 4.5
08:59:43MaxHRok, so for mine it should be the REC button
09:01:12MaxHRone button access to that is beautiful
09:01:13*[Saint] assumes MaxHR has it working now.
09:01:37[Saint]Glad I could be of service.
09:02:51MaxHRstrange, pressing it again (REC) while in the playlist does nothing, no button takes you back to now playing
09:03:07MaxHRplay button does
09:04:20 Quit ender` (Read error: Connection reset by peer)
09:04:27[Saint]It feels good to help someone get around an issue that's been bugging them.
09:05:31 Join ender` [0] (
09:05:53 Quit user890104 (Ping timeout: 272 seconds)
09:06:04 Join wodz [0] (
09:06:21MaxHRyes, rockbox 50% better to use after I got the HiVis theme, now it is another 75% better to use at least for me
09:06:26MaxHRThank you
09:06:56[Saint]125% \o/
09:07:51MaxHRwill get even better if I learn how to create skins, though prob over my head
09:09:46MaxHRis there a way to reset the volume to * certain level whenever player is powered off/on
09:10:42 Join Rob2223 [0] (
09:11:09[Saint]any settings you want to reset/persist after a power cycle you can put in "fixed.cfg"
09:12:12MaxHRnice, was looking to avoid getting blasted when switching from using player in car (w/ volume on max) to headphones, and not remembering to check vol ;p
09:12:47[Saint]to make it a lot easier, just take your config.cfg file and delete everything except for the volume string...set that value to the value you want to volume to be set to at each boot, then save it as "fixed.cfg" and place it in the .rockbox folder along with your config.cfg file
09:13:23[Saint]when I say "take your config.cfg file" I mean "take OF your config.cfg file" too ;)
09:13:36[Saint]well...that failed.
09:13:43the_Kyle /* Define this if a programmable hotkey is mapped */
09:13:49[Saint]"take A COPY OF your config.cfg file"
09:14:07 Quit Rob2222 (Ping timeout: 250 seconds)
09:14:21the_KyleSince the option is commented out, could that mean that there is no programmable hotkey mapped?
09:14:28MaxHRaww, ok, I understand, and I just discovered I can edit the config file right on the player, thought I would have to do it on computer via usb
09:15:21[Saint]the_Kyle: My guess is that there's no keymap for it, probably the reason it isn't defined.
09:15:58[Saint]IIRC the Clip(s) keymaps were pretty cramped up even before Hotkey was added.
09:16:17the_Kyle[Saint]: Probably no sense defining it in that case, since it won't do anything if no key is mapped for that function.
09:16:51the_KyleIt would seem that everything is already mapped to a function.
09:17:51the_KyleI personally wouldn't see a problem with putting things on the quick screen, but it seems that some options don't allow for that.
09:18:14the_KyleI couldn't add the playlist stuff for example.
09:19:00the_KyleShuffle and repeat are there, however, and I'll probably keep those as they are, considering I use them a lot.
09:20:12the_KyleBut playlist view does seem to be disabled, meaning that there is no context menu, either to put it on the quick screen or anything else.
09:22:28[Saint]what are you wanting on the quickscreen?
09:23:36the_KyleI thought the playlist view may be able to be put there.
09:24:30[Saint]Ah, no...there's a bunch of stuff that can't be.
09:25:01[Saint]It's not quite as simple as "anything in a menu/context menu"
09:26:01[Saint]I'm not sure if there's actually a list of what can be put in the quickscreen or not, I think it simply boils down to trial and error.
09:26:59the_KyleNot a problem. The playlist view is easy enough for me to find, I just thought I could make it even easier.
09:30:09the_KyleSeems at least to me that there's no need to make the playlist view easier to find. It's in the context menu while a song is playing.
09:30:55the_KyleWhich is also in the manual, I believe, but there's nothing like trial and error.
09:31:30 Quit Kupop (Ping timeout: 245 seconds)
09:36:49the_KyleIs voice access to the track info not yet implemented?
09:38:16[Saint]I'm not sure if it ever will be or not. There's some patches in the tracker along that line IIRC
09:39:48the_KyleI wouldn't consider it to be high priority, it would just be nice to be able to hear the info.
09:42:04[Saint]I believe there's some kind of proof of concept in the tracker, but iirc it was built around the WPS not much use to you.
09:42:54[Saint]Obviously there needs to be some form of trigger for it, as you probably wouldn't want to be constantly updated about the track state.
09:43:42the_KyleI was hoping to have it spoken whenever I select View Track Info from the context menu on the WPS.
09:44:12rasherthe_Kyle: I'm pretty sure you can get to it through the context menu while in the wps
09:44:41rasherthe_Kyle: oh, that's what you just said didn't work. Sorry
09:44:47the_Kylerasher: I can get to it, but the info isn't spoken.
09:44:52rasherI would indeed have expected that to work as well
09:47:07[Saint]yes...I would expect that to work.
09:47:24[Saint]I thought you were meaning some form of "voice on demand" feature.
09:47:51pixelma[Saint]: the Clip+ doesn't have a hold switch, it uses softhold
09:48:00pixelmaone key combo taken
09:48:06 Join evilnick__ [0] (
09:48:10pixelma(just an example)
09:48:46[Saint]pixelma: Yeah, I came to the conclusion that it was most likely a crowded keymap keeping it from being enabled.
09:49:41 Quit mystica555_ (Ping timeout: 240 seconds)
09:51:30 Quit evilnick_ (Ping timeout: 272 seconds)
09:54:07 Join Baw|BNC [0] (
09:54:17 Quit Bawitdaba (Read error: Connection reset by peer)
10:00:42 Join evilnick_ [0] (
10:00:48 Join Judas_PhD [0] (
10:04:04 Quit evilnick__ (Ping timeout: 265 seconds)
10:04:24 Quit JdGord (Quit: Bye)
10:07:10 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:07:21 Quit Baw|BNC (Read error: Connection reset by peer)
10:07:28 Join Bawitdaba [0] (
10:07:39 Join efyx [0] (
10:08:32***Saving seen data "./dancer.seen"
10:14:02 Quit MaxHR (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
10:19:03 Join Baw|BNC [0] (
10:19:15 Quit Bawitdaba (Read error: Connection reset by peer)
10:33:40 Join swilde [0] (
10:35:18 Join DerPapst [0] (
10:37:44 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
10:47:11 Quit [Saint] ()
10:47:14 Quit Dreamxtreme (Ping timeout: 272 seconds)
10:56:09 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:57:16 Join user890104 [0] (~Venci@
11:13:06 Join giovanni [0] (
11:24:31CIA-7New commit by wodz (r28739): HD300: Fix bootloader hang when RTC alarm fires. Make entering bootmenu more reliable. Add some comments.
11:24:52CIA-7New commit by wodz (r28740): RTC s35380a - add support for alarm function. Currently this powers up the device but doesn't start selected function (radio/recording/playback)
11:24:56CIA-7New commit by wodz (r28741): HD300: enable RTC alarm function (working partially for now)
11:26:48CIA-7r28739 build result: All green
11:28:55 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:29:01CIA-7r28741 build result: All green
11:31:19 Quit wodz (Quit: Leaving)
11:31:22 Join Dreamxtreme [0] (~Dre@
11:36:14 Quit Judas_PhD (Quit: This is a quitting message)
11:41:29 Quit ender` (Quit: Do not believe any statistic you didn't falsify yourself.)
11:53:21 Join The_Pwny [0] (
12:03:52 Quit TheSeven (Ping timeout: 272 seconds)
12:07:18 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:08:34***Saving seen data "./dancer.seen"
12:10:54 Quit mattymo (Ping timeout: 264 seconds)
12:18:20 Quit stoffel (Remote host closed the connection)
12:22:33 Join PurlingNayuki [0] (~PurlingNa@
12:22:38 Part PurlingNayuki
12:26:03 Join teru [0] (
12:26:47 Join InsDel [0] (
12:42:07 Join Judas_PhD [0] (
12:46:48 Join n1s [0] (~n1s@rockbox/developer/n1s)
12:54:46 Quit InsDel (Read error: Connection reset by peer)
12:57:09 Join ajb [0] (
13:00:04 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
13:16:12 Quit The_Pwny (Quit: Easy as 3.14159265358979323846...)
13:20:15 Quit GeekShadow (Ping timeout: 255 seconds)
13:21:37 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
13:30:09 Quit GeekShadow (Ping timeout: 255 seconds)
13:34:14 Join bluebroth3r [0] (
13:34:20 Quit bluebroth3r (Changing host)
13:34:20 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
13:34:20 Quit bluebrother (Disconnected by services)
13:42:56 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
13:59:04Tornemore forum spam:;u=28013
14:00:54gevaertsZagor: I just had the wiki ask me to confirm that it was me doing a change
14:01:07gevaertsWith a link to
14:01:24Zagoryes, I get that sometimes too. I don't know what they base it on.
14:03:59n1sit's been telling me my changes are suspicious lately too
14:07:08giovanniHi there, I've got a problem, I'd like to fix the bug in onda's players where you can't change volume in FM radio...but i need some tips, where do I have to look at?
14:07:22 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
14:08:35***Saving seen data "./dancer.seen"
14:18:46 Join designate72 [0] (
14:58:20 Quit tchan (Read error: Connection reset by peer)
14:58:39 Join MethoS- [0] (~clemens@
14:58:47 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
14:59:11 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
15:01:32 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:10:27n1sanyone around with a lowmem target that can test a patch for tremor?
15:15:04 Quit teru (Quit: Quit)
15:15:47n1sor maybe i can hack some config to get a tiny codec buffer...
15:16:09 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:24:33 Join panni_ [0] (
15:27:53 Quit designate72 (Remote host closed the connection)
15:30:47 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
15:31:06 Quit dfkt (Read error: Connection reset by peer)
15:36:57CIA-7New commit by nls (r28742): libtremor: ...
15:38:56CIA-7r28742 build result: All green
15:39:38n1seven though this patch fixed a crash i don't think it's suitable for backporting
15:41:22 Join dfkt [0] (dfkt@unaffiliated/dfkt)
15:41:41 Quit dfkt_ (Read error: Connection reset by peer)
15:42:54gevaertsI'd say a patch that touches 3000 lines can't be suitable for backporting ever :)
15:43:16AlexPn1s: It reads to me that the fix was a side effect of other changes, and the patch was massive anyway, so I'd agree
15:45:18n1sAlexP: i don't actually know how much of the patch is needed for the fix as the problem was that a return value was not checked
15:46:05gevaertsn1s: how easy is it to trigger the problem?
15:46:10 Join designate72 [0] (
15:47:01n1sgevaerts: rewind within the last ~2 seconds of a vorbis file (might need to be followed by another vorbis file in the playlist)
15:47:17n1sso, definitely doable but not very likely i'd say
15:48:11gevaertsok, maybe add it to the release notes then, but I'd say that's enough
15:48:55n1swait, i'll give it a quick test, might only need a single line in fact
15:51:16 Join kevku [0] (~kevku@2001:7d0:0:f000::135d)
15:51:20n1sah, yes, seems the fix is a one-liner
15:51:48n1si'll check that in to the branch then
15:52:10gevaertsYes, one liners are good :)
15:57:09 Quit Llorean (Read error: Connection reset by peer)
15:58:25CIA-7New commit by nls (r28743): libtremor: backport fix for FS #10484, fixing a crash
16:01:37 Quit dfkt (Read error: Connection reset by peer)
16:01:38 Quit linuxguy3 (Read error: Connection reset by peer)
16:01:48 Join dfkt [0] (dfkt@unaffiliated/dfkt)
16:02:18kugeln1s: no speed difference from that commit?
16:02:30 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
16:02:47 Join krabador [0] (
16:03:20n1skugel: +-0.1MHz
16:03:43n1si forgot to mention that
16:03:56 Quit antil33t (Read error: Connection reset by peer)
16:04:05 Join antil33t [0] (
16:04:32 Join Fuze2 [0] (
16:04:44n1si think it is *actually* faster (it saves about 0.1MHz on PP) but on coldfire the icache keeps changing the numbers so i'm not sure there
16:05:19Fuze2hello world
16:06:18Fuze2Is it normal that the Fuze2 allways boost?
16:07:00 Join linuxguy3 [0] (
16:07:19Fuze2I saw lot of good stuff for codecs, but for a target that always boost that's doesn't mind
16:07:36 Join DSStrife89 [0] (~Strife89@
16:07:49gevaertsWhat exactly does "always boost" mean?
16:08:37***Saving seen data "./dancer.seen"
16:08:38Fuze2Oh sorry, CPU boost
16:09:32gevaertsYes, but what exactly do you mean by that? Is it always running at maximum frequency, or does it boost for every track, or something else?
16:09:36 Quit designate72 (Ping timeout: 260 seconds)
16:10:29n1samsv2 doesn't have adjustable freq (yet), does it?
16:11:27gevaertsAh, right, indeed not
16:11:48Fuze2gevaerts: I wanted to ask if there is something to adjust frequencies of CPU to have a sort of idle
16:12:28Fuze2n1s: oh ok, what is the stoper? Lack of docs about the used chip?
16:12:54Fuze2if of course you do know :)
16:13:10saratogaamiconn: you around?
16:13:15n1si'd assume so, or that noone has worked much on it
16:13:46saratogai've got boosting enabled on mine, but some people reported occasional crashes when using it
16:14:02n1sFuze2: less cpu usage could still help a bit as the cpu can sleep more even if always running at the same freq
16:14:16Fuze2rockbox works very well on and have a good battery runtime, but pitty that the CPU is always to max even if not needed
16:14:51Fuze2Not even a bug or crash here :) good job folks
16:15:32Fuze2saratoga: are the fuze v2 all the same? Hardware wise
16:15:32AlexPI'm sure there are some bugs :)
16:16:30Fuze2AlexP: nothing in my daily use
16:17:51 Quit GodEater (Ping timeout: 272 seconds)
16:18:21 Quit saratoga (Ping timeout: 265 seconds)
16:18:36Fuze2saratoga: is there any informations about those crashes? Is there a patch somewhere to enable it (maybe just a matter of 1 line in the config file)?
16:18:54Fuze2oh, I was too slow
16:20:34Fuze2I found this: /* Define this if you have adjustable CPU frequency */ #define HAVE_ADJUSTABLE_CPU_FREQ
16:21:14 Join designate72 [0] (
16:21:17 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.
16:21:20Fuze2but I don't know if the low level part is still there? (saratoga mentioned that it was enabled before)
16:22:04saratogayes thats all you need to do to enable it
16:23:17Fuze2ok thanks saratogo, I guess that the low level part is there then? (maybe that's what needed to be fixed)
16:23:29Fuze2saratoga sorry :)
16:23:50saratogayes its present but disabled until someone figures out why it maybe crashes
16:24:13Fuze2is there bug reports about it somewhere?
16:24:31saratogai dont' think so
16:25:24Fuze2ohoh, maybe you have to enable it then to have people reports and to spot the problem;)
16:26:13gevaertsUnless those people also provide a fix, that's not too useful
16:27:54 Join jgarvey [0] (
16:28:18Fuze2gevaerts: maybe they could :p
16:29:18Fuze2is there someone around here that did work on this and that know a bit what's going on here?
16:30:05Fuze2on Fuzev2 CPU I meant
16:35:17Fuze2ok, it seems that r25924 disabled it: as3525v2: disable adjustable CPU frequency When the rest of the code is confirmed to work properly without CPU frequency changes, it will be possible to reliabily test set_cpu_frequency()
16:37:14 Quit giovanni (Quit: Sto andando via)
16:37:22 Quit kugel (Ping timeout: 245 seconds)
16:37:36Fuze2saratoga: apparently, judjing from svn logs, that was the Clip+ target that had problems
16:46:05Fuze2no eabi gcc on this cygwin install. I used h
16:47:43n1syes, you have to build it yourself, using like for any other development environment
16:48:04Fuze2but it didn't mention eabi, it seems that wasn't be updated since long :)
16:48:34TheLemonMantry using the codesourcery prebuilt gcc bins
16:48:49saratogathats probably not going to work
16:48:50gevaertsTheLemonMan: no
16:49:09TheLemonManwhy not? those are eabi compliant
16:49:09n1sthat is not the recommended version though an might cause trouble
16:50:04TorneIt won't work, almost certainly
16:50:11Torneour gcc is patched to use different compile options for libgcc
16:50:19Fuze2n1s: thanks for the help, I tried to run but complained as not found
16:50:30Tornethe default libgcc build will almost certainly fail to link
16:50:38Tornebecause of our lack of C++ support
16:50:45n1sFuze2: that means you need as (among other things)
16:51:07n1siow you need a full build environment to build the cross toolchain
16:51:35Fuze2n1s: a hint how to do please? I have Cygwin runing well here apparently
16:51:53Fuze2and I folowed the Wiki link
16:52:12saratogaunless you're an expert on cygwin you probably shouldn't use cygwin
16:52:51Fuze2saratoga: I'm under wine, and didn't want a vmware setup
16:53:14saratogayou're running unix? then why are you following the windows directions
16:53:32Fuze2indeed but...
16:53:42saratogayou realize cygwin is essentially a unix emulator right
16:54:11saratogawhat if you thought this through
16:54:58Fuze2I just followed your Wiki to be able to buil rockbox under Cygwin
16:55:37 Quit DSStrife89 (Read error: Connection reset by peer)
16:57:09Fuze2Why not? :)
16:58:20*Fuze2 don't have the point to have full steps how to set up a Cygwin dev env if that's not desired
16:59:19Fuze2or not encouraged
16:59:25gevaertsIt's for people who run windows
16:59:27 Join DSStrife89 [0] (~Strife89@
16:59:36Torneit's for people who run windows and ignore our advice to use vmware
17:01:16Tornesome people can't or won't virtualise linux, so they use cygwin.
17:02:30TorneFuze2: what is your actual operating system?
17:02:38 Quit GeekShadow (Quit: The cake is a lie !)
17:03:27 Quit krabador (Ping timeout: 272 seconds)
17:03:33Fuze2Torne: I do use Windows 7
17:03:34 Quit factor (Ping timeout: 240 seconds)
17:03:55Tornethen what does wine have to do with anything?
17:04:00Fuze2I did encounter in the past with samba when tried to build rockbox
17:04:22saratogado you know what wine is
17:04:22Torneanyway. we recommend that you install linux, either natively, or in vmware or virtualpc or some other emulator, and use the linux build instructions.
17:04:35Tornebut if you can't do that, then the cygwin instructoins should work if you actually follow them correctly
17:05:16saratogatechnically i don't think we have working cygwin directions anymore
17:05:20Tornesaratoga: Well, no
17:05:31TorneIgnore step 3 in the instructions; the rockbox mirror no longer has useful packages
17:05:37Lloreansaratoga: But you can get a standard cygwin development environment and follow the linux directions, I would assume.
17:05:39Tornebut the fact that you can't run means you didn't follow step 2 correctly
17:05:42saratogayou also need a bunch of other packages
17:05:59pixelmaTorne: afaik it has except for arm
17:06:00Tornesaratoga: you shouldn't, we fixed gcc to build with an internal copy of libmpfr/libgmp
17:06:14Tornethat fix should apply on cygwin as well
17:06:22Fuze2Torne: ok, and thanks for the advise. But the Wiki seems to not be up to date and doesn't mention arm eabi gcc, and don't provide it
17:06:26saratogaTorne: IIRC we don't include everything needed to actually run
17:06:28 Part Zagor
17:06:47 Join stripwax [0] (
17:06:49saratogaFuze2: yes that is what we are telling you
17:06:49Tornesaratoga: it should be sufficient, it includes enough to build the sim and that should (now) be the same set
17:07:03Torneif he doesn't have as, then he hasn't installed even those packages.
17:07:04 Quit panni_ (Read error: Connection reset by peer)
17:07:09Torneso, he hasn't followed the instructions :)
17:07:10saratogai'm too lazy to try but i'm 99% sure people have said it doesn't
17:07:18saratogai think its missing patch or something
17:07:24TorneOk, that's possible
17:07:29saratogaor some other weird lib needed by rockboxdev
17:07:32 Join panni_ [0] (
17:07:34Torneno, wait, it has patchutils
17:07:39Tornethere should be no weird libs needed any more
17:07:42Tornethat was only a temporary situation
17:07:50Tornegcc 4 needed libmpfr and libgmp
17:08:03Fuze2Torne; I did installed all the required packages. Where this should reside?
17:08:06Tornebut we changed rockboxdev so it downloads the source for those and sticks them in the gcc dir, so gcc builds them internally
17:08:16Tornesome time *after* we switched to gcc 4
17:08:20Torneso there was a period where you needed extra libs
17:08:23Tornebut you shouldn't any more.
17:08:25saratogathats probably what i'm thinking of then
17:08:35TorneFuze2: it's in the rockbox source
17:08:47Tornefollow the linux instructions
17:08:52Tornethey explain how to use it
17:08:54Fuze2I do have them I did checkout, I will see
17:09:22Tornesaratoga: it might still not work, i've not tried it, but I think it probably should
17:09:37 Join eWill [0] (
17:09:59Fuze2duh, error couldn't start
17:10:15eWillI'm not seeing the "RB forces Full-Speed USB mode" on the Fuze v2. How do I reproduce that problem?
17:10:26eWilli.e. I'm getting high-speed.
17:10:47 Join krabador [0] (
17:11:33 Quit krabador (Read error: Connection reset by peer)
17:13:34 Quit eWill (Client Quit)
17:13:42*Fuze2 gives up
17:14:00Fuze2I don't know why this wont start!!
17:14:11Torne*exactly* what are you typing?
17:14:13Torneand in what directory?
17:14:18Torneand exactly what does it say?
17:15:28Fuze2I'm on the tools directory, and I simply type "run"
17:15:49Tornethere is no command "run"
17:15:53Torneread the instructions
17:16:30Torneyou want ./
17:16:53Tornehm, ok, windows does have a command "run"
17:16:57Tornebut that's nothing to do with cygwin
17:18:51Fuze2wow, ok, that made it. Thanks !
17:19:23Fuze2I just need tp pick up a dowloader tool
17:21:48 Quit designate72 (Quit: Leaving)
17:22:06 Join designate72 [0] (
17:30:21 Quit antil33t (Read error: Connection reset by peer)
17:30:31 Join antil33t [0] (
17:34:22CIA-7New commit by nls (r28744): libtremor: merge upstream revision 17510.
17:35:23 Join slooopy [0] (
17:36:17CIA-7r28744 build result: All green
17:39:10 Join bmbl [0] (
17:39:10 Quit bmbl (Changing host)
17:39:10 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
17:40:07CIA-7New commit by nls (r28745): libtremor: merge upstream revision 17511 'Add checks/rejection for absurdly huge codebooks.'
17:41:56CIA-7r28745 build result: All green
17:44:42CIA-7New commit by nls (r28746): libtremor: merge upstream revision 17512 'Correctly handle the nonsensical codebook.dim==0 case.'
17:46:40CIA-7r28746 build result: All green
17:48:59CIA-7New commit by nls (r28747): libtremor: merge upstream revision 17513 'Add code to prevent heap attacks by exploiting dim==bignum and partition_codewords==partion_values^dim.'
17:50:44CIA-7r28747 build result: All green
17:52:13 Quit designate72 (Ping timeout: 272 seconds)
17:54:16Fuze2n1s: are those checks impacting speed by any rate?
17:54:37n1snot measurably
17:58:41 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:00:45 Join captainkewlllll [0] (2669ecc2@gateway/web/freenode/ip.
18:03:27Fuze2goodbye, and thank you for the help
18:03:32 Quit Fuze2 (Quit: CGI:IRC)
18:05:41 Join baptiste_ [0] (
18:07:50 Join designate72 [0] (
18:08:38***Saving seen data "./dancer.seen"
18:11:50 Join factor [0] (
18:12:02 Quit simonrvn (Ping timeout: 240 seconds)
18:13:51 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:14:43 Quit mortalscan (Remote host closed the connection)
18:14:57 Join mortalscan [0] (~mortalsca@
18:15:04CIA-7New commit by nls (r28748): libtremor: merge part of upstream revision 17514 adding some limit checking for allocations when unpacking comments.
18:16:07 Join simonrvn [0] (
18:16:44CIA-7r28748 build result: 246 errors, 82 warnings (nls committed)
18:16:48CIA-7New commit by nls (r28749): libtremor: export a libogg function that is now needed.
18:16:49 Quit Battousai (Read error: Operation timed out)
18:17:56 Quit DSStrife89 (Read error: Operation timed out)
18:18:40 Join Battousai [0] (~bryan@gentoo/developer/battousai)
18:18:43CIA-7r28749 build result: All green
18:20:45 Join DSStrife89 [0] (~Strife89@
18:24:34 Quit antil33t (Read error: Connection reset by peer)
18:24:43 Join antil33t [0] (
18:25:22 Quit simonrvn (Ping timeout: 240 seconds)
18:25:33 Join simonrvn_ [0] (~simon@
18:25:48 Nick simonrvn_ is now known as simonrvn (~simon@
18:26:50 Join JesusFreak316 [0] (
18:27:38 Quit stripwax (Quit:
18:29:36 Join simonrvn- [0] (
18:30:28 Quit ajb (Remote host closed the connection)
18:30:42 Join Xerion_ [0] (
18:32:02 Quit simonrvn (Ping timeout: 240 seconds)
18:32:03 Nick simonrvn- is now known as simonrvn (
18:32:33 Nick simonrvn is now known as Guest99616 (
18:32:47 Quit Xerion (Ping timeout: 245 seconds)
18:32:48 Nick Xerion_ is now known as Xerion (
18:35:55CIA-7New commit by nls (r28750): libtremor: merge upstream revision 17518 actually returning failure code from vorbis_synthesis_init
18:36:04 Join Xerion_ [0] (
18:36:07 Quit slooopy (Read error: Operation timed out)
18:37:16 Quit Xerion (Read error: Operation timed out)
18:37:16 Nick Xerion_ is now known as Xerion (
18:37:29CIA-7r28750 build result: All green
18:39:59 Join Jerom [0] (~jerome@
18:42:02 Quit Guest99616 (Ping timeout: 240 seconds)
18:42:21 Join simonrvn_ [0] (
18:43:34 Join LambdaCalculus37 [0] (~rmenes@
18:43:34 Quit LambdaCalculus37 (Changing host)
18:43:34 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
18:44:04 Quit elcan (Read error: Connection reset by peer)
18:44:10 Join elcan [0] (
18:44:43 Join liar [0] (
18:44:51 Quit amiconn (Remote host closed the connection)
18:44:51 Quit pixelma (Remote host closed the connection)
18:45:03 Join pixelma [0] (quassel@rockbox/staff/pixelma)
18:45:03 Join amiconn [0] (quassel@rockbox/developer/amiconn)
18:47:01 Join ppawel [0] (
18:47:14 Quit domonoky (Ping timeout: 240 seconds)
18:47:44 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:49:14 Quit Xerion (Read error: Operation timed out)
18:50:34 Quit TheSeven (Ping timeout: 240 seconds)
18:55:31 Join slooopy [0] (
19:02:43 Quit designate72 (Ping timeout: 255 seconds)
19:04:06 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:04:50 Quit baptiste_ (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20100106054534])
19:06:13 Join designate72 [0] (
19:06:31 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
19:07:17kugelandroid 2.3 adds awesome native APIs for us (input, audio, graphics)
19:07:45LambdaCalculus37kugel: Nice! :)
19:07:57kugelit's now possible to write entirely native apps, but I think we're still going to need some parts in java
19:09:04LambdaCalculus37kugel: Are you going to ditch support for older Android versions and rewrite entirely for Gingerbread, or keep one version with compatibility for most Android versions?
19:09:13 Quit MethoS- (Read error: Connection reset by peer)
19:10:28preglown1s: have you measured the performance difference after using libogg in tremor?
19:10:42preglowsupposedly, the ogg implementation in tremor is superior
19:11:05 Join MethoS- [0] (~clemens@
19:15:53 Nick simonrvn_ is now known as simonrvn (
19:19:23n1spreglow: yes, it got a tad faster on PP targets and a tad slower on coldfire (but might very well be due to the icache as disabling code that was never called had a pretty big effect (sometimes speeding up, sometimes slowing down)
19:20:14n1sbut still, it didn't change more than 0.1MHz for either PP or cf
19:21:00 Join toffe82 [0] (
19:21:11n1salso the ogg stuff was dropped from upstream tremor entirely so that code was unmaintained now and porting other upstream changes back to the old ogg stuff would be annoying
19:21:42 Quit Llorean (Quit: Leaving.)
19:21:56n1sand from looking at it i think this ogg implementation is somewhat simpler
19:22:08 Quit benedikt93 (Quit: Bye ;))
19:24:12 Quit DerPapst (Quit: Leaving.)
19:24:36 Quit MethoS- (Remote host closed the connection)
19:25:32 Join freddyb [0] (
19:26:22 Join MethoS- [0] (~clemens@
19:26:38 Quit LambdaCalculus37 (Quit: run!)
19:27:14n1sour special case in codebbok.c handles the most time consuming reading btw, it made about a 6MHz diff compared to how tremor does it usually
19:27:20n1son cf
19:27:56 Join ender` [0] (
19:28:08 Quit JesusFreak316 (Read error: Connection reset by peer)
19:29:18 Join JesusFreak316 [0] (
19:44:35 Quit krazykit (Ping timeout: 255 seconds)
19:46:39 Join krazykit [0] (
19:46:54 Quit toffe82 (Read error: Connection reset by peer)
19:47:57 Join sideral [0] (~sideral@unaffiliated/sideral)
19:49:06 Quit Judas_PhD (Quit: This is a quitting message)
19:52:17the_KyleHopefully patching tools/ to support Svox Pico. I don't hold out much hope, since the pico2wave utility is extremely limited and as far as I know, Perl bindings don't exist, but I may be able to implement the most basic support for it.
19:53:59 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:54:06 Quit designate72 (Remote host closed the connection)
19:58:39 Join baptiste_ [0] (
19:59:56 Join Judas_PhD [0] (
20:00:04preglown1s: very nice :)
20:05:28 Quit Jerom (Quit: Leaving.)
20:07:40 Quit factor (Ping timeout: 260 seconds)
20:08:39***Saving seen data "./dancer.seen"
20:09:35the_KyleOops. My simulator no longer talks. I'm not sure if it's the way I implemented Pico support or if I just didn't specify the right options for rbspeexend, or possibly one of the target options. I have a 1.2 MB English voice file, but no speech from the clip+ simulator.
20:09:47 Join DerPapst [0] (
20:09:54*the_Kyle tries again.
20:11:26*the_Kyle notes that he did get a screen full of segfaults, but those were ignored, because for some reason, they seem to be a problem with the pico2wave utility itself, and they have no effect on speech output.
20:14:43 Quit saratoga (Ping timeout: 265 seconds)
20:16:26 Join designate72 [0] (
20:17:28 Quit freddyb (Remote host closed the connection)
20:17:35the_KyleWhat is the difference in tools/ between -t=<target> and -i=<target_id>? The command seems to require both options, but during buildtime using tools/configure, only the numeric ID is needed.
20:22:02 Join Kupop [0] (
20:26:03 Join Llorean [0] (
20:26:19 Quit Llorean (Changing host)
20:26:19 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
20:28:04 Join Horscht [0] (~Horschti@xbmc/user/horscht)
20:32:39 Join freddyb [0] (
20:33:02 Quit DSStrife89 (Quit: Heading home.)
20:33:21 Join Keripo [0] (
20:39:25 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.
20:39:54 Quit MethoS- (Remote host closed the connection)
20:40:42the_KyleIt seems like my problem is in the way I'm calling tools/ I have something wrong in my options. I tried it with espeak, and I have the same problem with the voice file not speaking in the simulator. It's also interesting to note that my voice file built with is about 55 KB larger than it is when I use tools/configure to build the voice.
20:43:12 Quit freddyb (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723085406])
20:43:17the_KyleIs there any way outside of the simulator to test a voice file to determine whether or not there is even recognizable speech data in the file?
20:44:49 Join webguest98 [0] (
20:46:13 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
20:47:22 Quit webguest98 (Client Quit)
20:48:20 Join InsDel [0] (
20:51:19 Quit Keripo (Quit: Leaving.)
20:53:51evilnick_Bthe_Kyle: AFAIK, voice files are speex files, so preumably anything that can play those should be able to play voice files too.
20:55:30saratogathey're not standard speex files, they're not in the usual container
20:56:41the_Kyleevilnick_B: Speexdec doesn't recognize the file. Says it doesn't look like a speex file. File just says it's data.
20:57:02evilnick_BSorry, I was incorrect in my assumption then
20:57:10saratogayes they're not compatible with speex
20:57:14evilnick_Bgevaerts and saratoga know more than I do
20:57:35the_KyleNP. Any way to test for the presence of speech data?
20:58:23 Join clear_runway [0] (
21:00:09 Join DerPapst1 [0] (
21:00:11the_KyleThe strang thing is that I get different results using tools/ and tools/configure. I thought that confighre used default rbspeexenc options, but defaults in are not working.
21:00:35 Join lixxus99 [0] (
21:00:43 Quit DerPapst (Ping timeout: 245 seconds)
21:01:40 Join noamsml_ [0] (
21:01:47 Quit Rob2223 (Quit: Rob2223)
21:02:00lixxus99is the beastpatcher for gigabeat s compatible with v1.3 firmware or do i need to downgrade to 1.2 ?
21:02:15n1sthe_Kyle: the target id is a number unique for each target that is checked when loading the voice file, a voice file with the wrong number is rejected. this is *NOT* the same number that configure displays in the menu
21:02:48n1sthe "target" is used to determine what strings to use from the lang file
21:03:14n1syes the system is somewhat convoluted but this is not normally something you have to touch
21:03:14clear_runwayis there any way to disable the "fade out/in" effect on music? it's a little annoying
21:03:21n1sclear_runway: yes
21:03:28the_Kylen1s: Sounds like my problem exactly. How can I get this number?
21:03:38 Quit noamsml (Ping timeout: 245 seconds)
21:03:46 Join Rob2222 [0] (
21:03:46clear_runwayshort of messing with the source?
21:03:50the_KyleI was in fact using the number from the configuration table.
21:04:36n1sthe_Kyle: the number is set ny configure but is not the one displayed in the menu, clip+ has id 66
21:04:39 Join Keripo [0] (
21:05:00the_KyleIs sansaclipplus OK for the target string?
21:05:05saratogaclear_runway: search the manual
21:05:09n1sclear_runway: you can disable the fade in the settings
21:05:14clear_runwayokey dokey
21:05:37n1sthe_Kyle: um, wait it's all comming back to me now
21:06:02clear_runwayoh, thats what "crossfade" meant! durf du hurr
21:06:57clear_runwayor maybe not. ill stop bugging now
21:07:16 Part lixxus99
21:07:28evilnick_Bclear_runway: Fade on pause/stop is what you need
21:07:38clear_runwayoh. duh.
21:08:18 Join factor [0] (
21:11:05n1sthe_Kyle: no, the "target" needs to include all the "features" your voice file should have strings for in a colon separated list
21:12:07n1sif you make a regular voice build they can be found in the apps/features file in that build dir
21:13:08n1sthis stuff really isn't intended for people to run manually
21:14:11 Part docgnome ("Channel buffer killed")
21:17:31the_KyleLooks like I did get basic Svox Pico support implemented correctly in tools/ My feature list probably isn't correct, but it does speak in the simulator now.
21:18:53 Quit bmbl (Quit: Verlassend)
21:19:09the_KyleI think I'll look in the configuration scripts to try to patch it in there so I don't have to go through the entire feature list manually every time I want to build a voice file.
21:33:41kugelrockbox still builds and runs with the updated gcc from the ndk
21:39:53 Quit slooopy (Ping timeout: 255 seconds)
21:51:49 Join slooopy [0] (
21:56:08 Quit baptiste_ (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20100106054534])
22:01:58 Quit jfc (Read error: Connection reset by peer)
22:02:24 Join jfc [0] (
22:02:37the_KyleI got the Rockbox voice build to work with Svox Pico. The pico2wave utility is very limited, so sped, pitch and volume can't be changed, but it definitely works. It generates a 1.2 MB English voice file. Now to suppress the segmentation faults generated by pico2wave, but that's not a Rockbox problem, and it still produces perfect speech output.
22:03:38 Quit sideral (Ping timeout: 255 seconds)
22:04:59n1sthe_Kyle: at least volume should be pretty easy to change in the clips before encoding with some wav processing utility, pitch and speed should also be possible but might not give as good quality as if done in the speech engine directly
22:06:09 Quit ppawel (Quit: Leaving)
22:07:02the_Kylen1s: The Pico library is supposed to support setting volume, pitch, speed and voice directly. Unfortunately, the included utility, which I used to make my quick hack doesn't allow it.
22:07:36n1sis this librarcy accessible to use from your own utility?
22:07:48 Join Xerion [0] (
22:08:41***Saving seen data "./dancer.seen"
22:09:06the_KyleI don't have a utility for it other than the latest speech-dispatcher packages. And unfortunately, speech-dispatcher doesn't support .wav output. It only supports output to the sound device.
22:09:58the_KyleIt does, however, change all aspects of many speech synthesizers, including eSpeak and Pico.
22:11:09n1sbut i mean, would it be possible to write an app that uses this speech engine and that can control the parameters of it by interfacing with the library?
22:11:26 Quit factor (Read error: Connection reset by peer)
22:11:51 Join factor [0] (
22:12:10 Quit slooopy (Ping timeout: 276 seconds)
22:12:42the_KyleIt is definitely possible, and even the pico2wave utility itself could probably be made more extensive so that it could allow changes to the speech parameters.
22:13:48the_KyleOh, and any application that interfaces with the Pico library would also have to figure out how not to segfalt.
22:14:41bertrikIf a task did a wakeup_wait, will it get immediate attention when wakeup_signal is called from another thread, or can other threads get attention first?
22:14:58the_KyleIt doesn't affect output at all, but it does look bad and flud the screen with error messages.
22:20:26kugelgevaerts: my test program also works on android
22:21:22gevaertskugel: I guess the main question is whether it's faster than preemptive threads plus a lock
22:22:04TheSevenapparently there's still something that we don't know about the nano2g battery charger
22:22:09kugelpreemtive threads plus locks involve (several?) syscalls so should be a lot faster
22:22:15TheSeventhere must be ways to control the charge current and stop charging altogether
22:23:03TheSevendisk mode stops charging, but still draws the power it needs itself from USB after it received an eject command
22:23:03gevaertsIt's definitely worth trying
22:23:46TheSevenalso disk mode charges at ~177mA while rockbox/embios charges at ~110mA
22:24:24bertrikmaybe that's 100 mA charge current + 10 or 77 mA run current
22:24:40TheSevenno, i'm reading the battery current ADC
22:24:46TheSevenand i meant diagmode, not diskmode
22:24:58TheSeventotal usb current is ~230/270mA
22:25:40*TheSeven probably needs to do some more GPIO poking
22:26:25TheSevenand disk mode in the ejected state draws 24mA from the bus
22:27:58 Quit Keripo (Quit: Leaving.)
22:44:33the_KyleIs there a way to configure the amount of reduction of music volume when a voice clip is spoken? I can't find it in the manual.
22:45:29n1sno, don't think there's a way to do that
22:46:16saratogayou could probably acheive that with replaygain and the preamp setting
22:46:17the_KyleNot a major problem, but it does seem like anice feature to have in future.
22:47:03*the_Kyle checks the replaygain options. Maybe preamp will work, but I don't think it's connected.
22:47:07n1ssaratoga: do you know if voiced is mised in after dsp?
22:47:18bertrikthere's no reduction at all in the radio screen, so voice is pretty useless there IIRC
22:47:32n1sbertrik: yes that's a known issue
22:47:33saratogaspeex files won't have replaygain info anyway, so i think it won't matter
22:48:33n1sbertrik: probably just because noone implemented it, the volume reduction is done digitally for playback so the radio needs its own way
22:48:57n1sthe voice stuff is unfortunately somewhat unmaintained :/
22:49:10the_KyleIs the volume reduction har-coded?
22:50:13 Join Keripo [0] (
22:52:09saratogais rockbox USB on the clip+ stable enough to use in windows for development? getting annoyed by the db refresh while debugging
22:53:13bertrikthe_Kyle, I don't know
22:53:58bertrikFor the radio screen, I did once look into the possibility of the various radio chips to lower their output volume. Some can, but other can't.
22:54:34n1sbertrik: couldn't the codecs lower the input volume from the fm?
22:54:40bertrikI think the most feasible solution for volume reduction in the radio screen, is to reduce the volume at the codec
22:55:40the_KyleSounds like we need more coders on voice. Unfortunately, my skill level is basically good enough to do the quick Pico hack I did, and doesn't really go much rurther.
22:56:32the_KyleOtherwise, I'd help.
22:58:28 Quit kevku (Quit: KVIrc 4.0.2 Insomnia
23:01:15 Join GeekShadow [0] (
23:01:15 Quit GeekShadow (Changing host)
23:01:15 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
23:01:20 Quit jgarvey (Quit: Leaving)
23:04:59n1sthe_Kyle: afaict the reduction in volume is done by shifting the music samples right by 2
23:05:01kugelgevaerts: my program only works on 2.2+
23:05:38kugelbut I think that's not a problem since ARM is still the only officially supported architecture and we have the asm routines for context switches on arm
23:06:03kugel(or we could shit sigaltstack() ourself)
23:06:36the_Kylen1s: In that case, it sounds rather nontrivial to make volume reduction a configurable option.
23:07:36 Quit Horscht (Quit: Verlassend)
23:08:11n1sprobably, yes, if you want this to br reasonably adjustable you need to change it into a fixed point multiplication and probably add the value to the dsp config stuff but, no shouldn't be too hard
23:08:41bertrikn bits of shift right equals n*6 dB volume reduction IIRC
23:09:20n1sis 6, 12, 18dB fine grained enough then?
23:09:56the_KyleI'd have to hear it to be sure, but that actually seems good.
23:10:26n1sthen it should be even easier to implement
23:11:54the_KyleIt's a 2-bit shift now, right? In that case, 1, 2, 3 configurable would work well in my opinion.
23:15:12 Join JdGord [0] (~jd@
23:16:46 Quit bertrik (Quit: :tiuQ)
23:19:32 Quit captainkewlllll (Quit: Page closed)
23:19:35 Quit Kupop (Ping timeout: 250 seconds)
23:19:44n1syes, a configurable shift would work
23:20:03amiconnkugel: You know fibers only work on NT based windows? (I'm not saying we should support ages-old 9x/ME...)
23:20:08amiconnsaratoga: pong
23:20:08n1si wonder though if it wouldn't be nice with a more fine grained volume control
23:20:13kugelamiconn: yes
23:20:56 Quit n1s (Quit: Lmnar)
23:21:03saratogaamiconn: i'm going to turn on dircache by default on all >2MB targets since it seems to help the WPS screen noticeably on flash targets, is this going to impact archos any?
23:22:55amiconnNot stock ones, but those using the 8 MB mod
23:22:58 Join madalu [0] (~user@unaffiliated/madalu)
23:23:09amiconnI don't have such a modded archos myself
23:23:42saratogaok if someone mods their player they can probably be bothered to change a default setting
23:23:51*amiconn only uses dircache on 32MB+ hdd targets
23:24:30amiconnImo it's useless on flash based targets, and too much tradeoff on 16MB (and lower mem) hdd targets
23:25:23*TheSeven considers the huge audio buffer even more useless on flash based targets
23:25:36CIA-7New commit by saratoga (r28751): Enable dircache by default on all targets with enough memory to use it. Helps significantly on HD players, and is somewhat faster on most flash ...
23:25:36kugelsaratoga: it also makes track skipping faster in the playlist viewer
23:25:47saratogayes, significantly so
23:26:17*amiconn can't imagine how dircache helps the wps
23:26:29saratogaplaylist.c works differently if its enabled
23:26:34CIA-7New commit by kugel (r28752): Get rid of get_user_file_path and do the path handling in wrappers for open() and friends.
23:26:34saratogasorry playback.c
23:26:37saratogai have no idea why
23:26:57amiconnI mean, in what way does it help?
23:27:03kugelI didn't mean to commit that :\
23:27:12saratogaskipping tracks is faster
23:27:36amiconnHmm, that may explain it
23:27:41amiconnI rarely ever skip
23:27:43CIA-7r28751 build result: All green
23:27:44saratogae.g. if you start playing the first track in a folder and then keep hitting next, then its much faster with dircache
23:28:16CIA-7New commit by kugel (r28753): Fix configure and lib path
23:28:20CIA-7New commit by kugel (r28754): Fix extracting
23:28:24CIA-7New commit by kugel (r28755): Upgrade ndk requirement to r5 (it ships a better compiler) and cleanup configure accordingly.
23:28:28amiconnImo it's better not to add tracks you don't want to hear to the playlist (or even the device)...
23:29:02saratogai don't think its just skipping, i think all track changes are more responsive FWIW
23:29:30saratogabut skipping quickly through a playlist is the easiest way to notice that
23:30:00kugeltag lookup is vastly faster with dircache and tagcache ramcache both enabled even on flash
23:30:13CIA-7r28752 build result: All green
23:30:43amiconnWell, if you think it helps...
23:31:03*amiconn now needs to disable dirache explicitly on his flash targets
23:31:09 Join The_Pwny [0] (
23:31:13 Quit designate72 (Ping timeout: 260 seconds)
23:31:26kugelit doesn't hurt if you leave it on
23:32:04 Quit TheLemonMan (Quit: free(me))
23:32:29CIA-7r28755 build result: All green
23:32:41kugelthe accesses to flash/disk easily compensate the battery life lost due to smaller audio buffer, on flash it's marginal anyway
23:33:46saratogaalso amiconn, i'm somewhat annoyed that you undid my change that put test_codec on top of the viewers list
23:34:05saratogathat was quite handy for debugging codecs, which involves running test_codec over and over again as quickly as possible
23:34:16amiconnI didn't even notice that it was an intended change
23:34:23kugelsaratoga: autorock.rock?
23:34:25amiconnIt appeared to me as a (rather annoying) mistage
23:34:46saratogacan I autorock a viewer?
23:34:50kugelprobably not
23:35:12kugelbut you could hack the path to the test files into test_codec
23:35:16saratogaideally would be a way to script loading test codec on a file from the command line with no gui at all, but thats beyond my level of free time
23:35:42amiconnA test_* plugin should *never* end up as the default viewer for a file type, hence the test_* plugins have to come last for each extension
23:36:02saratogathey'll never be the default type for audio files though, right?
23:36:45kugelbut test_jpeg_mem for jpegs
23:36:55 Quit JdGord (Ping timeout: 250 seconds)
23:36:56amiconnCorrect for built-in types (i.e. audio files) - but not having a general rule makes misleading bugs like the one I fixed more likely
23:37:03 Part domonoky
23:37:34the_KyleI noticed there are no encoders for Ogg Vorbis, FLAC and Speex when recording. Is this because no fixed-point encoder yet exists?
23:37:56 Join mystica555_ [0] (
23:38:16saratogahmm yes i should not have put the jpeg ones on top when i moved test codec
23:38:28saratogathe_Kyle: correct
23:38:41saratogawell as far as i know, i suppose they could exist out there somewhere
23:38:42 Quit markun (Read error: Connection reset by peer)
23:39:01saratogaamiconn: do you mind if i put test_codec back on top then?
23:39:58saratogawhy is that?
23:40:02amiconnI mean, why should it go on top? It's a test plugin for development. Do you always develop, or do you just use your dap(s) as well?
23:41:32saratogaputting it on top means a lot less scrolling when i'm debugging
23:41:42saratogaduring normal use the test_codec isn't compiled, so it doesn't matter where it is
23:42:03amiconnDo you switch builds between development and normal use?
23:42:25saratogai'm not sure, i don't see where you're going with the question?
23:42:27 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
23:42:42*amiconn never does that, as it would take *a lot* longer than scrolling a little for calling test codec, even a dozen times or more
23:43:25saratogai'm just saying if the test_codec plugin is enabled, its probably the most likely viewer to use
23:43:47 Join JdGord [0] (~jd@
23:43:51amiconnAnd I'm saying it's not (and for a non-dev it's irrelevant anyway)
23:45:04saratogawhy would you enable test_codec if you didn't intend to use it?
23:45:13amiconnI'm always building with test plugins enabled for all my targets, but I'm not always using them
23:45:42amiconnBecause reconfiguring + rebuilding + installing takes quite some time (and I'm not talking cygwin)
23:46:06saratogado you actually use the test plugins that frequently?
23:46:15saratogathey're essential for codec work, but probably not most other things
23:46:17amiconnNo, but that's my point
23:46:31amiconnThere are more test plugins than just test_codec
23:46:44saratogawell why not just disable them if it bothers you having test_codec on top
23:46:56kugeldisabling doesn't get them off the dap
23:46:58*amiconn wonders whether saratoga reads
23:47:10amiconn[23:45:39] <amiconn> Because reconfiguring + rebuilding + installing takes quite some time (and I'm not talking cygwin)
23:47:19amiconnI *do* use test plugins, but not always
23:47:38saratogayes but if you seldom need them, then you seldom waste that time
23:47:43amiconnAnd even if I do, it's not always test_codec, but e.g. test_grey
23:48:22saratogaand really, all this to avoid having to scroll through one more menu item is silly
23:48:23amiconnNo, but having the test plugins around means they should come last (per extension), because otherwise they disturb normal operation
23:48:47saratogayes, but i'm talking about test_codec, which does not disturb them
23:49:06amiconnAnd I don't want to make an exception for test_codec because making an exception for one test plugins makes the rule less obvious
23:49:44saratogai don't think having this rule is worthwhile
23:49:58amiconnIt's not about the position in "Open with.." - that I don't care about
23:49:58saratogatheres no need to sort to the viewers file
23:50:18amiconnI have to search for the plugin I want to use anyway, because they are sorted erratically
23:50:31amiconnThere *is* a reason
23:51:12amiconnThe commit I did was the result of chasing a bug - starting in the completely wrong place because a test plugin (for .jpg) ended up as the default viewer
23:51:23saratogayes i saw that
23:51:33amiconnI thought the new image viewer was buggy
23:51:50saratogai'm saying you should have moved the jpeg related viewers since they were actually the problem
23:52:31amiconnWhen I found that the test plugin was the culprit, I saw that a lot of test plugins had been added at the top of viewers.config
23:53:10amiconnI think this was a mistake, caused by one or two test plugins already listed at the top when further ones were added
23:54:07amiconnThat's why I'm thinking test plugins always have to come after the normal plugins for each extension (if there is any)
23:54:18saratogayes i realize that
23:55:43saratogaok I am out of time, I will try to tease out of you why you don't want me to fix the test_codec issue when I have more time
23:57:26 Quit pamaury (Remote host closed the connection)

Previous day | Next day