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

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

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

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

#rockbox log for 2011-10-05

00:00:28 Quit robin0800 (Ping timeout: 260 seconds)
00:03:38 Quit benedikt93 (Quit: Bye ;))
00:07:13bertrikah, the nano2g doesn't even support radio yet, because IAP does not work on that target yet
00:07:55 Join Thra11_ [0] (
00:08:16 Join T44 [0] (
00:08:55wodzbertrik: do serial work on nano2g?
00:09:31bertrikwodz, I made a patch for IAP over UART0 for the nano2g, but I have no idea if it actually works (probably doesn't)
00:10:15bertrikI remember someone working on serial output for the nano2g recently, was it Lalufu perhaps?
00:10:27 Quit Thra11 (Read error: Operation timed out)
00:11:07pixelmaaha, looks like the "buttons locked" splash is actually handled by the button action system
00:11:58 Quit Topy44 (Ping timeout: 248 seconds)
00:13:52wodzpolemon: ok if you want to compile nano2g sim with FM screen you need hack a bit firmware/config/ipodnano2g.h file 1) uncoment CONFIG_TUNER line 2) add SRC_CAP_FMRADIO to INPUT_SRC_CAPS
00:14:01wodzjust checked this
00:14:20pixelmadoes a 1st gen Nano sim show the radio screen?
00:14:47wodzit should
00:15:31pixelmathen that's enough for theme making - both have the same screen size and bit depth which is the only thing that counts
00:16:42pixelmaI believe that's what polemon wanted as the question was about viewing FMSs
00:16:51wodzI recall someone stated today that there is no other target with screen dimentions as nano2g
00:17:26pixelmaI said there were no others like the Nanos
00:17:39wodzah, I misunderstood than
00:19:03 Quit wodz (Quit: Leaving)
00:22:48saratogawhats the max clock on the Fuze+?
00:23:17JdGordonfreddyb: whyd you close the glyphcache patch?
00:23:27JdGordonyour one which rejigs the load size
00:24:25 Quit bertrik (Quit: And That, My Liege, Is How We Know the Earth to Be Banana Shaped)
00:29:03*JdGordon does like this idea boaut The Fund buying him a sports car though
00:40:10freddyb(You mean allocate by glyphs.) I thought it was a good idea but i'm not going to bat for it.
00:40:40 Join krey_ [0] (
00:41:01krey_hello, I can't remount my ipod since I installed dropbox
00:43:30saratogaprobably worth mentioning what ipod you have
00:49:51krey_video (5g)
00:49:57krey_the block device doesn't exist
00:56:20 Join Scromple [0] (
01:03:41 Join robin0800 [0] (~robin0800@
01:04:48JdGordonfreddyb: i dont tihnk anyone else had any thoughts on it at all..
01:04:59JdGordonand actually tihnking about it i tinhk its a good idea
01:09:54 Join ReimuHakurei [0] (
01:10:14freddybIf there's no resistance I might reopen it but part of the reason I did it was to not starve international users of glyphs and the only likely way to get this in is if reducing memory usage is the only possibility.
01:14:33JdGordon99% of users wont realise there is a difference
01:18:53 Quit Thra11_ (Quit: kthxbai)
01:27:10 Quit KiwiCam (Ping timeout: 252 seconds)
01:28:34pixelmakrey_: which computer OS?
01:28:48krey_pixelma: linux
01:30:10 Quit robin0800 (Read error: Connection timed out)
01:30:49pixelmahmm, I just knew that some versions of MacOS had trouble with our USB HID mode so turning that off would have been a suggestion but I believe there's no such known problem under linux. Still worth a try maybe but other than that I need to pass
01:32:18pixelmayou should always be able to use the Apple firmware for USB (manually started or so) but Rockbox USB should work on the Video
01:32:23 Join ChickeNES [0] (
01:34:16krey_pixelma: even if I reboot into disk mode, nothing happens. is this normal?
01:35:42 Quit RansomTime (Ping timeout: 248 seconds)
01:36:43 Join RansomTime [0] (~RT@wikia/vstf/Randomtime)
01:41:20 Join robin0800 [0] (~robin0800@
01:43:10 Join bobby2hat [0] (
01:44:45bobby2hathello all, why have I got a write error on a (v. old!) 3g ipod classic when installing the bootloader with utility or ipodpatcher when run as administrator, windows 7, in disk mode...
01:45:26 Quit AlexP (Read error: Connection reset by peer)
01:46:24JdGordonmaybe the disk is dead
01:47:46 Quit krey_ (Remote host closed the connection)
01:49:08bobby2hattried moving 200mb mp3 it worked, checked disk for errors...
01:49:54bobby2hatwould anything restrict access?
01:52:51 Quit RansomTime (Ping timeout: 245 seconds)
01:54:09***Saving seen data "./dancer.seen"
01:56:42 Quit robin0800 (Ping timeout: 256 seconds)
02:01:44 Join chelz [0] (~iridesenc@unaffiliated/chelz)
02:02:32chelzis there an article/guide/etc on the differences/similarities between Rockbox and the iPod 5G / 5.5G / 6G native firmware?
02:05:09 Quit hilbert (Quit: Textual IRC Client:
02:05:48saratogachelz: if you just mean an overview of rockbox features and the apple firmware features, try wikipedia
02:08:09chelzsaratoga: well i'm trying to decide if i should get an iPod 5G / 5.5G that's only 80GB that can dualboot or iPod 6G that's 160GB but can only run rockbox. so i'm curious what, if anything, rockbox lacks that might be worth only 80GB
02:08:43 Join RiotGear [0] (
02:09:38saratogachelz: the 6G port isn't in particularly good shape at the moment, so thats something to consider
02:14:45 Quit bobby2hat ()
02:20:01 Quit fatline (Quit: This computer has gone to sleep)
02:33:09[Saint_AndChat]Dammit...who highlighted me?
02:33:25[Saint_AndChat]My scrollback doesn't go that far.
02:34:21[Saint_AndChat]Bah...just dismissed the notification without reading it.
02:34:30krazykit[Saint_AndChat], so why not look in the logs for your nick ;)
02:35:00[Saint_AndChat]Because its painful on a phone :)
02:36:58 Quit soap (Read error: Connection reset by peer)
02:37:35JdGordonthe logviewer is fine on the phone
02:37:37 Quit [Saint_AndChat] (Quit: AndChat - Android IRC Client -
02:37:56 Join [Saint_AndChat] [0] (~Saint]@
02:42:32 Quit tjb0607 (Ping timeout: 252 seconds)
02:44:10 Quit ReimuHakurei (Remote host closed the connection)
02:45:01[Saint_AndChat]Well...that was boring :)
02:45:39[Saint_AndChat]alexbobp: weird your build hated my theme.
02:45:55 Join soap [0] (~soap@rockbox/staff/soap)
02:46:05[Saint_AndChat]You don't have resources on sdcard running f, do you? It hates that.
02:46:23[Saint_AndChat]It'll copy over the. wps
02:47:02 Join Keripo [0] (
02:50:51 Join milk [0] (
02:52:14 Join tjb0607 [0] (
02:59:58 Join soap__ [0] (~soap@
03:00:39 Quit freddyb (Quit: Leaving)
03:00:40 Quit soap (Ping timeout: 258 seconds)
03:00:52 Nick soap__ is now known as soap (~soap@
03:01:11 Quit soap (Changing host)
03:01:11 Join soap [0] (~soap@rockbox/staff/soap)
03:03:10 Quit kadoban (Ping timeout: 258 seconds)
03:11:23 Quit ChickeNES (Quit: Computer has gone to sleep.)
03:18:16 Join ChickeNES [0] (
03:26:02 Join ChickeNE_ [0] (
03:26:27 Quit ChickeNES (Read error: Connection reset by peer)
03:26:43 Quit Keripo (Ping timeout: 276 seconds)
03:26:56 Join Keripo [0] (
03:30:26polemonwhich fonts get installed when installing minimal installation, or without the extra font pack
03:31:15[Saint_AndChat]Only whatever your cabbie port uses, and sysfont.
03:31:19JdGordononly the one(s) needed by the default theme
03:32:21polemonand which are those?
03:32:49polemonI installed the font pack, now I have many, but I'd like to minimize the space .rockbox takes up
03:33:18[Saint_AndChat]Look at cabbiev2 cfg
03:33:25[Saint_AndChat]It'll tell you.
03:33:41polemonok, I'll check
03:33:44JdGordonit actually may not :)
03:33:57JdGordonpolemon: whats the point? save 10mb or less?
03:34:13[Saint_AndChat]It may not?
03:34:15chkktri_btw, why not to ad some ani-aliased fonts?
03:34:39[Saint_AndChat]Feel free to submit some.
03:35:00chkktri_hmm i could
03:35:01[Saint_AndChat]Ones with acceptable licenses, obviously.
03:35:13chkktri_but they're propiretary
03:35:27[Saint_AndChat]Then, no, we can't use them.
03:35:28chkktri_gejavu fonts should be fine
03:36:17[Saint_AndChat]I supplied a font pack ages ago...its not a lack of fonts, its a lack of....effort.
03:37:08chkktri_you mean bdf fonts?
03:37:21[Saint_AndChat]One sec, ill drop a new link here from my dropbox, huge aa font pack.
03:38:39 Join Jak_o_Shadows [0] (
03:38:55[Saint_AndChat]Lots of sizes, bold, regular, light, acceptable license.
03:38:58polemonJdGordon: just so I know what font's I should use in my themes
03:39:19polemonanyway, I'm looking at cabbiev2, but there's just one font in the .cfg
03:39:37[Saint_AndChat]polemon: how many did you expect?
03:39:46[Saint_AndChat]Cabbie only uses one font.
03:40:11[Saint_AndChat]But, that font differs between targets.
03:40:17polemonyeah, but that wasn't what I was looking for
03:40:41polemonwhat font's are always installed, no matter if it is minimal installation, or something else.
03:40:47polemonwhich fonts are those?
03:41:02[Saint_AndChat]It differs between targets.
03:41:10[Saint_AndChat]You can't do it that way.
03:41:41polemonWell, the theme I made, is just for the iPod nano
03:41:47[Saint_AndChat]The only way is to state that the user must download the font pack and use it in order to use your theme.
03:42:05[Saint_AndChat]That way your not limited to using a specific font.
03:42:18polemonso there is not a single font, that is consistent on all installs?
03:42:56[Saint_AndChat]Well, there is...but its 8px high, and ugly as hell :)
03:43:18[Saint_AndChat]Its only a fallback, though.
03:43:23polemonthere's this helv*.fnt fonts, they seem to be like a fallback or something
03:43:37polemonhelvR*.fnt, even
03:43:39JdGordoneven then, sysfont doesnt have many glyphs
03:43:44[Saint_AndChat]No, they're not. That's what cabbie uses.
03:44:03[Saint_AndChat]Really really.
03:44:31[Saint_AndChat]The size used, however, differs between target.
03:44:52[Saint_AndChat]There's no one font present on all targets other than sysfont.
03:45:41[Saint_AndChat]Use whatever font you want from the font pack, and state in the theme description that the font pack MUST be used.
03:46:02[Saint_AndChat]You can't account for all cases. Its not possible.
03:46:13polemonwell, then pretty much every theme needs the font pack...
03:46:31[Saint_AndChat]Yes. Correct.
03:47:52[Saint_AndChat]You can't include any font that's in the font pack in your theme, it'll wont be uploaded. I guess you discovered that.
03:48:44[Saint_AndChat]I did mention this some days ago...
03:54:10***Saving seen data "./dancer.seen"
03:56:34 Quit Keripo (Ping timeout: 240 seconds)
03:59:44 Join Jak_o_Shadows1 [0] (
04:02:29 Quit Jak_o_Shadows (Ping timeout: 276 seconds)
04:03:11 Quit [7] (Disconnected by services)
04:03:20 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:07:03 Quit FoH (Quit: ¡ooʇ ‘ǝןdoǝd ǝɹɐ sʇɐq)
04:13:14polemonerm, I haven't discovered that, but I wouldn't do it anyway...
04:14:11polemonok, I'm gonna package the themes shortly, sadly, I can't test the .fms, so I can't include that one...
04:15:13 Quit Galois (Ping timeout: 244 seconds)
04:15:16[Saint_AndChat]Ummmm...why can't you test it?
04:15:24[Saint_AndChat]I assure you you can.
04:15:32polemonbecause I can't select the fms display in the UI
04:15:37polemonUI simulator, that is
04:15:55[Saint_AndChat]Use the nano1g sim.
04:16:21polemonok, I'll try
04:16:37JdGordonthe nano1g always shows the radio?
04:16:51[Saint_AndChat]In the sim, yes.
04:17:02JdGordonthe sim shuold actually be changed to have a magic key to toggle fm-present
04:18:53polemonyeah, that would be nice, but then again, I had to find out the hotkeys by trial and error
04:19:07polemonthere's help file, ot text file or somethnig
04:19:23[Saint_AndChat]You can just click on the ui buttons.
04:19:49[Saint_AndChat]Pretty simple...
04:20:23[Saint_AndChat]It works the same as the physical device does.
04:21:14JdGordonwe shuold actually have a IAP simulator program to talk to the sim(or real target even!)
04:21:43[Saint_AndChat]That'd be nice.
04:25:12 Quit GeekShadow (Ping timeout: 252 seconds)
04:26:15polemonok, I've installed the Nano 1G UI sim and there's the FM radio!
04:26:22polemonand the recording screen
04:26:49polemonok, it seems, I'm not finished with the theme afterall
04:27:18 Join GeekShadow [0] (
04:28:21polemonfor the recording screen, how do I do that?
04:28:38polemonis there another file for it?
04:29:35JdGordonthe recording screen cant really be skinned...
04:29:38polemonit just uses the .wps?
04:29:47JdGordonyou can do it in the sbs if you're really adventurous though
04:29:51 Quit pixelma (Disconnected by services)
04:29:53 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:29:55 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:30:08JdGordonthat was on my todo list, and probably wont happen for quite a while now
04:30:18*JdGordon has no interest in recording
04:30:53polemonyewell, there's other things that can't be skinned, seemingly
04:30:59polemonVU level meters, for instance
04:31:16polemonthose that appear when the skin is not working, or when there is none at all..
04:31:31 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:31:31 Quit amiconn (Disconnected by services)
04:31:37[Saint_AndChat]%?cs and the .sbs == a full recording screen skin.
04:31:38JdGordonvu meters (called the peak meters) *can* be skinned
04:31:49JdGordonno [Saint_AndChat]
04:31:50[Saint_AndChat]Beat me to it JdGordon :)
04:31:53 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:32:17polemonyeah, but in the docs, I couldn't find the tags for it...
04:32:26[Saint_AndChat]JdGordon: no?
04:32:39[Saint_AndChat]polemon: looking in the wrong place, probably.
04:32:51JdGordonpolemon: %pL and %pR will let you display a bar for the left and right channels
04:33:02 Quit fiesh (Ping timeout: 276 seconds)
04:33:14JdGordon[Saint_AndChat]: no, the rec screen does... err.. actually...
04:33:21[Saint_AndChat]polemon: search for "bar tags"
04:34:12JdGordonyeah, the rec screen hardcodes all its info, the sbs wont work too well with it
04:34:45[Saint_AndChat]Yeah, I was a bit can skin around it :)
04:34:51 Quit ChickeNE_ (Quit: Computer has gone to sleep.)
04:34:51[Saint_AndChat]And I've done so.
04:35:10polemonAh! OK, I found the peak meters, must have overlooked them
04:35:27[Saint_AndChat]polemon: you really should use the manual, also.
04:35:32 Join fiesh [0] (
04:35:48polemonwell, all I could find, was on
04:35:48[Saint_AndChat]Only use customWPS for tags that aren't in the manual.
04:35:54polemonI'm reading the Wiki, basically
04:36:08[Saint_AndChat]The manual CLEARLY lists the skin tags.
04:36:16[Saint_AndChat]Its even in the index.
04:36:25[Saint_AndChat]You didn't look very well :)
04:36:37JdGordonbe nice....
04:36:43polemonit's quite confusing...
04:36:49JdGordonit certainly is
04:36:50polemonI mean the Wiki and page layout
04:36:54JdGordonthe wiki is pretty shit
04:37:05polemonlooks like a DokuWiki a bity
04:37:26[Saint_AndChat]Yeah, that's why I recommended the manual. Its laid out a bit better.
04:37:40[Saint_AndChat]There's tags misssing from it, though.
04:38:24JdGordonthe tags don't belong in the manual, they should be in a seperate document autogened every time skin_engine/ is touched
04:38:37JdGordonand not in fucking TeX, something simpler
04:38:44[Saint_AndChat]There's less tags missing from the manual than there is tags missing from CUSTOMWPS
04:38:47polemonyou guys should do it the Motorola way...
04:39:11[Saint_AndChat]Everything Motorola does is wrong by default :)
04:39:28[Saint_AndChat]I know, I own a Defy :p
04:40:35JdGordonhaha snap
04:44:25polemonMotorola has nice documentation policies
04:44:43polemonyou can get all documentation on paper for free
04:44:54polemonyou don't even have to pay for shipping
04:45:13polemona friend of mine ordered the full documentation of the M68k
04:45:23polemonover 6000 pages in total
04:45:30[Saint_AndChat]Make a sizable donation, and we'll talk :)
04:46:19[Saint_AndChat]"The Fund" would go broke /pretty/ quickly with printing and shipping.
04:46:26polemonyou could start by making a nice LaTeX documents
04:46:37[Saint_AndChat]We have those.
04:46:40JdGordondocumentation is boring...
04:46:44[Saint_AndChat]Its called the manual.
04:46:45JdGordonhence the lack of them
04:47:04polemonwell, I'm not a big fan of doing documentation work myself...
04:47:24[Saint_AndChat]And LaTex != "nice"
04:47:33[Saint_AndChat]Its horrible to work Anthony.
04:47:36polemoncompanies usually have lower tier programmers or so called "technical writers" for those tasks
04:47:42polemon[Saint_AndChat]: nope
04:47:59[Saint_AndChat] is.
04:48:01JdGordonthe tech writer here isnt techy at all :)
04:48:04polemonI like LaTeX, I just shunned PDFlatex a year ago, or so
04:48:13JdGordonshe struggles with the command line even :)
04:48:30polemonomg, EVERY single technical writer I've known in the companies I'
04:48:39polemonve worked for, were female
04:48:46JdGordonpolemon: like I said, if you want to help here it would be hugely useful to nuke CustomWPS and the manual and start a new seperate document
04:49:00polemonbad timing
04:49:07[Saint_AndChat]polemon: perhaps you just volunteered to maintain our manual :)
04:49:15polemonI'll be working on my masters thesis coming two semesters...
04:49:25[Saint_AndChat]Since you like latex so much ....
04:49:44polemonI do like it, and I have to use it even more in the upcomping months and years
04:49:56[Saint_AndChat]Poor boy.
04:50:15[Saint_AndChat]Another soul twisted by the horrors of latex :)
04:50:16polemonI wasn't a big fan of LaTeX until I discovered XeLaTeX
04:50:44polemonXeTeX is certainly easier to use and let's you produce nicer documents
04:52:00fieshin what sense nicer?
04:52:08polemonfor instance, you have OTF support
04:52:24JdGordongoing very offtopic now
04:52:29polemonyou're not bound to those shitty Computer Modern fonts, etc, you can easily swap them
04:52:29fieshI don't see how that makes it nicer ;)
04:52:41fieshyou're not bound to them in latex either
04:52:46polemonwell, I have no problem shutting up just now...
04:52:46fiesh\usepackage{txfonts} e.g.
04:52:49fieshoh yes
04:52:52polemonfiesh: poop
04:52:56JdGordonstop the TeX talk...
04:53:05JdGordonmove to #rockbox-community
04:53:23[Saint_AndChat]But yeah, JdGordon is right. It'd be a huge benefit to nuke customWPS and the manual and have one port of call for all such docs. If you know latex, and Luke it, ans are good with it...I almost guarantee it'll still need fixing when your semesters of study are up :)
04:53:26fieshcouple days ago I asked about the Bauer stereophonic binaural DSP but didn't get any answer
04:53:44fieshhas anyone ever considered implementing it? I'm kind of tempted but alas have zero time right now, so it might take a long time until I get to it
04:54:22polemonWell, I could take a look at it, and maybe do it on the side, since writing LaTeX documents, doesn't require too much cognitive workout
04:54:35JdGordon[Saint_AndChat]: I *did* even attempt to have the help doc and the tag table (for the parser) generated from the same source xml file remember.... but that didnt happen :/
04:54:56JdGordonfiesh: yes
04:55:17JdGordoni.e sounds like at least *you* have contemplated implementing it :)
04:56:07[Saint_AndChat]JdGordon: yeah, I remember that. Our problem is having ...what, like 2 "doc gurus".
04:56:26[Saint_AndChat]I can only think of 2...
04:56:30fieshJdGordon: ok, let me rephrase that to "has anybody except me ever..." ;)
04:56:41fieshI was actually surprised it isn't implemented in rockbox yet
04:56:49fieshwould seem very natural to have a dsp for headsets
04:56:49[Saint_AndChat]Likely no.
04:57:06JdGordonthere are apparently 4 people online at the moment, so its hardly a sensible question...
04:57:22*JdGordon is being a bit of an ass :)
04:57:34JdGordonyou sure you cant do the same affect with the current dsp?
04:57:42fieshwell it could be they have been part of a discussion etc., so it's not just 4 people
04:57:44 Join Keripo [0] (
04:58:07fieshpretty sure,
04:58:28[Saint_AndChat]fiesh: if there's nothing on flyspray, then it was *only* discussion, and therefore not relevant. :)
04:58:54[Saint_AndChat]Discussion, and implementation ans motivation are very different things.
04:58:56fieshok, thanks
04:59:59[Saint_AndChat]Things are often discussed, but that's no indication anyone intended doing anything about it.
05:00:52[Saint_AndChat]Seems a nice idea, but it should be possible already with custom configs and our existing settings.
05:01:34fieshit should?
05:01:44[Saint_AndChat]It should, yes.
05:02:18[Saint_AndChat]Just have an "in car" configuration, a "headset" configuration, etc.
05:02:26[Saint_AndChat]Run when needed.
05:02:38fieshyes that's what I want, but how would I get the DSP?
05:03:00saratogafiesh: is that much different then the crossfeed DSPs we support?
05:03:11 Join Rob2223 [0] (
05:04:29fieshsaratoga: to my understsanding, the bs2b is more sophisticated
05:04:43fieshbut I haven't done any direct listening comparison, I should do that, good idea
05:04:47saratogawhat does it do differently?
05:06:37fieshhmm good question, I never looked at its details I must admit, I was always more than happy with the bs2b ;)
05:07:00 Quit Rob2222 (Ping timeout: 258 seconds)
05:07:10alexbobp[Saint_AndChat]: it was AlexP, who is not me :P
05:07:38fieshhmm maybe I'm wrong actually and it doesn't give anything extra, thanks saratoga, I'll try some direct comparisons
05:07:57[Saint_AndChat]"Its different! Different is better...can haz moar differents!" :)
05:08:19[Saint_AndChat]alexbobp: yeah, sorry...Nick complete fail.
05:08:31alexbobpno worries
05:08:31saratogaskimming that link its cites someone named "Meier" who is suspiciously similar in name to the "Meier crossfeed" on the patch tracker
05:09:04[Saint_AndChat]saratoga: I'd like that in svn ...I've been using it a long time.
05:09:08[Saint_AndChat]I like it.
05:09:26[Saint_AndChat]I'm not sure why it wasn't donutted.
05:09:37[Saint_AndChat]Lol, *committed
05:10:33saratogabug bertrik, hes probably just been working on other stuff
05:14:06 Quit T44 (Read error: Connection reset by peer)
05:21:25 Join Topy44 [0] (
05:28:23 Quit [Saint_AndChat] (Remote host closed the connection)
05:32:32 Part chelz
05:40:18Unhelpfulfiesh: i've honestly never noted this fatiguing effect of headphone listening with pure stereophonic sources. but, if something like bs2b is going to do what it should do to work like it ought, i would imagine it must perform an approximation of the hrtf.
05:42:14fieshUnhelpful: I think the crossfeed filter really does a decent job, I gave it a couple of tries and can't complain at all, although I tested it vs. bs2b on my laptop, so the output device was not the same
05:42:18fieshUnhelpful: but thanks
05:42:56 Quit Horscht (Quit: Verlassend)
05:43:45Unhelpfulfiesh: well, a simple crossfeed probably does *enough*. if you want to simulate listening on speakers, though, there is a well-known method for converting "some audio" to "what this would sound like in each ear if coming from a specific point in space"
05:44:14Unhelpfulwhich, if you want to really sound like two speakers, is what you would aim for
05:54:13***Saving seen data "./dancer.seen"
05:56:38 Join [Saint_AndChat] [0] (~Saint]@
06:06:27 Quit niekie (Quit: No Ping reply in 180 seconds.)
06:06:45 Join niekie [0] (~niek@CAcert/Assurer/niekie)
06:12:49 Join Galois [0] (
06:19:01CIA-14New commit by 03fredwbauer (r30714): Do not font_unload() failed fonts when aborting a skin load
06:23:43CIA-14r30714 build result: 0 errors, 7 warnings (fredwbauer committed)
06:27:53 Join fatline [0] (
06:37:12[Saint_AndChat]*-community made me think....does, or could, rbutil collect stats on successful installs, and if possible drop a file to count updates as well?
06:37:30[Saint_AndChat](So updates aren't counted as installs)
06:41:42[Saint_AndChat]That would be useful, no?
06:42:06[Saint_AndChat]So we have even a ballpark guess of the userbase?
06:44:11 Quit fatline (Quit: This computer has gone to sleep)
06:44:57CIA-14New commit by 03Blue_Dude (r30715): Clean up compressor setting code
06:47:04 Quit Keripo (Ping timeout: 260 seconds)
06:53:22CIA-14r30715 build result: 0 errors, 7 warnings (Blue_Dude committed)
06:54:12 Quit Zharf (Read error: No route to host)
06:58:37 Quit Scromple (Ping timeout: 260 seconds)
06:59:43 Join Scromple [0] (
07:00:20 Quit [Saint_AndChat] (Remote host closed the connection)
07:02:45 Join [Saint_AndChat] [0] (~Saint]@
07:04:27 Quit Scromple (Ping timeout: 260 seconds)
07:09:35 Quit [Saint_AndChat] (Read error: Connection reset by peer)
07:11:45 Join Scromple [0] (
07:22:06 Quit powell14ski (Quit: powell14ski)
07:35:14 Join jludwig_ [0] (
07:38:58 Quit jludwig (Ping timeout: 256 seconds)
07:40:15 Quit milk (Quit: baaaiiii)
07:41:54 Join kadoban [0] (
07:48:34 Join [Saint_AndChat] [0] (~Saint]@
07:54:17***Saving seen data "./dancer.seen"
07:54:23 Join Keripo [0] (
07:55:34 Join RansomTime [0] (~RT@wikia/vstf/Randomtime)
08:22:52 Quit Scromple (Read error: Connection reset by peer)
08:23:28chkktri_looks like i'm having problems with D2 since new item in debug menu
08:23:30 Quit Keripo (Quit: Leaving.)
08:24:04chkktri_to be stright: rockbox freezes while playing, then changing track and before playing track
08:24:49chkktri_some mp3 files don't play at all, just wps screen with readen tags and calculated time
08:25:16chkktri_though sometimes they play but with huge (~30sec) delay
08:25:42chkktri_also there's ~5-10 sec delay then changing tracks
08:27:43 Join Scromple [0] (
08:28:21 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:34:02[Saint_AndChat]There's no active maintainer for the D2, and hasn't been for some time, its also not "supported".
08:34:14[Saint_AndChat]I'm not surprised at all to be honest.
08:34:28[Saint_AndChat]Nor should anyone be :)
08:35:21[Saint_AndChat]You can help by bisecting which commit introduced the issues you're seeing.
08:35:30 Join mudd1 [0] (
08:35:42[Saint_AndChat]Since you apparently have this target.
08:38:05 Quit RansomTime (Ping timeout: 276 seconds)
08:40:23 Quit [Saint_AndChat] (Remote host closed the connection)
08:40:54 Join [Saint_AndChat] [0] (~Saint]@
08:41:42 Join ender` [0] (
08:46:53chkktri_well i'll try to find from where it starts once again
08:47:38kugelZagor: we talked about this before
08:47:50kugelI think we want a dynamic tick for all targets, not only RaaA
08:48:44Zagordynamic as in pause/resume, or on-demand tick lengths?
08:49:14kugelthe latter would be nice
08:50:43kugelthe target independent part could call a timer_reschedule() function and fix up current_tick as appropriate
08:51:12kugel(this way HZ needn't be dynamic)
08:51:23 Quit Scromple (Ping timeout: 252 seconds)
08:52:51kugelyou can relax the tick rate when no music is playing (could possibly even do it depending on the codec) and no recent ui interaction took place
08:53:28Zagorraaa is still special since we really go into "don't even exist" mode, which never happens on hardware targets
08:53:51Zagorbut sure, I'd love having it everywhere
08:54:42kugelbtw, the tick rate isn't a problem on my phone (0% cpu usage)#
08:55:08Zagorwhich phone model is that?
08:55:24kugelhtc legend
08:55:43ZagorI get pretty solid 2% cpu use
08:55:47Zagoron my nexus s
08:56:51Zagorbtw, where did you get the march and mtune gcc flags from? they're rather strange :-)
08:57:24kugelthat was the ndk default when I did the port
08:57:35[Saint_AndChat]1% while backgrounded here on moto Defy, 2% on galaxy 5
08:57:37kugeli.e. what the standard build system of the ndk used
08:58:06kugelI'm fine with tuning for armv6 even
09:02:09kugelZagor: I'm not sure I like the solution to control the tick from java. I think the native code should control the java code (as per target tree), not the other way around. plus, the foreground/background switch is initiated from native, going back to native seems backward to me
09:02:58kugelbtw, you can have underscores in native methods. the separator is _1 then (or something like this)
09:03:17kugele.g. Java_org_rockbox_RockboxService_tick_1resume()
09:03:38kugeleventually we should get rid of the automatic linkage, and register methods explicitely
09:06:11 Join einhirn [0] (
09:06:59 Join simonlnu_ [0] (u2FJYR2uze@unaffiliated/simonrvn)
09:10:35 Quit simonlnu (Ping timeout: 276 seconds)
09:10:36 Nick simonlnu_ is now known as simonlnu (u2FJYR2uze@unaffiliated/simonrvn)
09:15:45 Join petur [0] (~petur@rockbox/developer/petur)
09:24:59 Join God_Eater [0] (93722cc8@gateway/web/freenode/ip.
09:25:06 Quit God_Eater (Changing host)
09:25:06 Join God_Eater [0] (93722cc8@rockbox/staff/GodEater)
09:40:21Zagorkugel: yeah you're probably right about native vs java. I was in the mindset that "app being minimized" is a java thing.
09:40:47 Join KiwiCam [0] (
09:44:13kugelZagor: yes, that separation isn't always easy. I try to solve this kind of thinks in an interrupt-like fashion and respond to it in native (app minimized is handled in lcd-android.c btw)
09:46:51KiwiCamI hope this is clear enough: I have a bitmap that's 4 pixels across and three down. The top row and bottom row are black. (It's for a Sansa Clip so there are only two colours as such). The centre row is four white pixels across. Or, is it better to have just one row of black pixels four across, and duplicate them to give the impression of the block I mentioned first? What uses less memory? I'll be using a lot of this block in a them
09:46:51KiwiCame that seems to be taking forever :)
09:53:08[Saint_AndChat]Bitmap strips use less memory than individual bitmaps to comprise the same image.
09:53:22[Saint_AndChat]I hope that's what you were asking.
09:54:20***Saving seen data "./dancer.seen"
10:05:18KiwiCamSo, it's better to have one long bitmap say 3 x 27 pixels, than 9 bitmaps, each 3 by 1 making 27, joined together to look the same as the long bitmap?
10:26:36 Join LinusN [0] (
10:31:51[Saint_AndChat]KiwiCam: sorry, I was busy for a bit. Yeah, it is better to do it with a bitmap strip than it would be to use individual images. Easier to/ better code too :)
10:40:43KiwiCamTrue. Thanks again.
10:40:50 Quit KiwiCam (Quit: Leaving)
10:52:46 Quit [Saint_AndChat] (Read error: Connection reset by peer)
11:07:49JdGordonhas there been any more movement on the git move?
11:08:22Zagornot since the gerrit installation.
11:08:38JdGordonis supposed to show anything?
11:08:50JdGordonthats where gerrit is going to end up no?
11:08:59Zagorno it's at
11:09:13JdGordonah, okey dokey
11:15:13 Quit God_Eater (Quit: Page closed)
11:16:59JdGordonBuschel: esp ping! :)
11:27:58 Join KiwiCam [0] (
11:38:21 Quit bluebrother (Disconnected by services)
11:38:22 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
11:39:58 Quit fs-bluebot (Ping timeout: 248 seconds)
11:41:06 Join fs-bluebot [0] (
11:47:44kugelZagor: RE tick rate: What I think would work well: 3 states, a) idle tick rate 1Hz, b) music, tick rate 25Hz, c) user activity, 100Hz. on native targets c) would be defined as via backlight on, on RaaA whether an activity is open or not
11:49:23kugel(targets without backlight are tricky in that scheme, it's not goign to be worse for them)
11:49:34Zagorwhy have tick on at all in idle?
11:50:03kugelthere's lots of code depending on the tick
11:50:21 Join MethoS- [0] (~clemens@
11:51:06kugelmost importantly the thread scheduler, and various timers (e.g. idle poweroff)
11:51:29Zagorthread scheduler doesn't need to be running in idle
11:51:35Zagoridle poweroff does though
11:52:18Zagorunless we rather solve that with a single-shot timer
11:52:44Zagormy worry is that if we design for 1Hz we're going to be more lazy than if we aim for true zero-tick idle.
11:52:53kugelI feel turning it off entirely is a bit drastic
11:53:18JdGordonanyone wanna test a patch to fix glyphcache?
11:53:54ZagorI can't test stuff at work, I'm afraid
11:54:07kugelyou need the tick to respond to outside events too, so that the button queue is checked
11:54:22***Saving seen data "./dancer.seen"
11:54:41Zagoroutside events should start the tick rather
11:54:42kugel(e.g. start playback via remote button/widget when the app is in the background)
11:55:07JdGordonI'm playing with a new system to save a different .glyphcache file for each font to better cache glyphs, currently i save them to /.rockbox/<fontname>.glyhpcache. shold I put it in .rockbox/fonts (they'll be hidden in rockbox anyway), or some other place/name?
11:55:07Zagorif those events are polled, 1Hz isn't going to cut it anyway
11:55:35*kugel doesn't want TIMER_AFTER/_BEFORE to be wrong too
11:56:48kugelJdGordon: are you back? that would be great
11:57:01Zagorthat's what I mean. if we have to do polling in idle, we're going to want to up it to 10 Hz or more and suddenly we don't have much of an idle state at all
11:57:56kugelI don't want to suggest polliong in idle state is good, but measuring time should still work
11:58:49Zagorchange the idle timer to a one-shot firing at shutdown time
11:59:01kugelJdGordon: oh well, I was curious since you want our opinion
11:59:32kugelZagor: timers work via tick :)
11:59:53Zagorkugel: we welcome everyone who contributes. long-term commitment is necessary.
11:59:57Zagoris *not* necessary
12:00:07Zagorthe tick timer doesn't
12:00:39 Join Jak_o_Shadows [0] (
12:00:40kugelyes, but that one drives all other timers
12:01:13Tornethe fix for that is to not have the tick drive all other timers
12:01:17Zagoruh, yes. and we are talking about how we want to change that to avoid wasting power running empty timer loops.
12:01:30Torneyou set a timer interrupt to go off when the next timer is due
12:01:35Torneet voila, no tick
12:01:52kugelright, on RaaA
12:01:59kugelit's not that simple for native targets
12:02:02Torneyes it is
12:02:09kugelno it isn't
12:02:15kugelmost targets have only 2 timers
12:02:51Torneyou only need one
12:03:00Torneset to go off when the next timer is due.
12:03:12kugelyes, but's it's not simple to do
12:03:21Torne..yes it is!
12:03:26 Quit Jak_o_Shadows1 (Ping timeout: 255 seconds)
12:03:34kugelTorne: patch here in 5min=? :)
12:03:37Torne<- wrote tickless kernels before
12:03:45kugelwe would have that by now if it was that easy
12:04:14JdGordonyou just keep a list of pending timers and scedule the next one each time the timer goes off.... not hard at all
12:04:25kugelit's sure doable, and it's not rocket-science, but it
12:04:31TorneYeah. Interrupt goes off, pop timer off priority heap, read next delta, set interrupt again
12:04:34kugelit's harder to implement that what we have now
12:04:34Tornethen execute timer
12:04:36JdGordonjust dont queue up a million tick times at boot
12:04:50TorneIt is more complicated than having a tick, yes
12:04:56Tornebut it's still simple
12:06:16kugelthe tickless kernel is only part of it, the otehr code needs to be adapted. e.g. our responsiveness to user actions depends on the tick
12:06:27LalufuWould that imply new interfaces for the timeout* functions, too?
12:06:30kugeland threading
12:07:01JdGordonthat would all be done transparently
12:07:13JdGordonqueue_wait() would fix that
12:07:23JdGordon(or queue_send() or whatever does the timeout)
12:07:50*JdGordon naivly assumed thats how our ticks were already being done!
12:08:17kugelyou need to decide whether latency is important currently or not, and stuff
12:08:24 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
12:15:19 Quit kadoban (Ping timeout: 260 seconds)
12:28:23 Quit KiwiCam (Quit: Leaving)
12:38:07 Join mortalis [0] (
12:42:03 Join Jerom [0] (~jerome@
12:49:11Zagorkugel: the latency does not change. we just do away with all the ticks that do nothing more than increment the timer variable.
12:49:54 Join wodz [0] (
12:51:24wodzTorne: am I correct that gdb-stubs exploit undefined instruction or swi on arm for single stepping?
12:51:35Torneno idea.
12:52:04Tornei have never debugged on non-emulated hardware without using an ICE
12:52:16Torneor with gdb, really :)
12:52:40wodzwe don't use swi in rockbox right?
12:52:47Tornewe don't, no
12:52:50Tornebut we also can't sensibly
12:52:58Tornewe run in svc mode all the time, so swi will corrupt state
12:53:03Torne(specifically r14)
12:53:22Tornewe should probably run in sys mode instead :)
12:53:58Tornethere is also, yaknow, BKPT instruction :)
12:55:09wodzisn't it for v5+ ?
12:55:23Tornesys mode exists in V4+
12:55:48TorneOh. Yeah.
12:55:59Torneundef works just as well
12:56:13Torneif you're only using a software monitor
12:56:22Tornei have no idea what real code actually does, thoguh
12:59:59wodzWould there be any side effects of running in sys mode? Change should be trivial (only crt0.S probably)
13:00:23Tornewodz: Nope. If you never use SWI it makes absolutely no difference at all, other than anywhere you actually inspect/set the mode bits (e.g. stack setup)
13:00:32TorneIt just enables you to use swi later.
13:01:17 Join Jak_o_Shadows1 [0] (
13:02:02JdGordonwodz: you're trying to debug the font boot crash?.
13:02:16Tornewodz: the other other option for debugging is, yaknow, actual debug hardware in cp14
13:02:33Tornebut not everything has that :)
13:02:42wodzJdGordon: I don't have target affected
13:03:41Lalufuwodz: Yes, I've done some work with the serial port on the ipod2g.
13:03:42 Quit Jak_o_Shadows (Ping timeout: 248 seconds)
13:03:42 Quit soap (Ping timeout: 248 seconds)
13:03:47Lalufuanything in particular?
13:03:57Lalufumake that mini2g
13:04:31wodzTorne: I am still thinking about extend our crash capabilities - gdb-stub should be quite easy option
13:04:59wodzmini or nano?
13:05:05Lalufuthe mini
13:05:10wodzmini has working serial AFAIK
13:05:31Lalufuyes, it does. It even has two working serials, which is a bit of a problem
13:05:40LalufuI suspect all the other ipods are similar
13:06:01wodzanyway Torne has my mini
13:07:22 Quit Jerom (Quit: Leaving.)
13:07:51LalufuSER0 is the serial port next to the headphone jack, SER1 is the serial port in the dock connector
13:09:55Lalufubeing able to use both (not at the same time, but either one or the other) is somewhere on my todo-list
13:13:30 Join hilbert [0] (
13:14:55JdGordonis there any reason GPIOD_PIN(7) = (1<<7); wont work too early in the boot sequence?
13:15:15JdGordonthat turns the button light on, I;ve put that pretty much at the strat of init() and it isnt happening
13:15:33JdGordonso, is there any memory issues which could cause the binary to just cause the loader to crash?>
13:15:54kugelZagor:the tickless system needs to react to changing latency demands, though
13:16:16wodzJdGordon: You need to configure pin direction most probably
13:16:28Zagorkugel: of course. it's just a more efficient way to run a timer.
13:16:50kugelit needs to be implemented :)
13:17:41JdGordonwodz: that macro isnt enough?
13:19:14Zagorkugel: and, more importantly, thoroughly tested. that's why I thought adding a quick fix for raaa was a quicker solution. it's also much more important on raaa than hardware targets.
13:20:11JdGordon\o/ progress
13:20:23kugelthe 3 states I proposed should be relatively simple to implement, but it's not truly tickless
13:20:43kugelyour quick fix breaks code that depends on the tick/TIMER_AFTER/TIME_BEFORE etc
13:20:59wodzJdGordon: usually you need to setup pin direction first (in separate register associated with the port) and than you are free to change pin level
13:20:59Zagorwhat raaa code does that?
13:21:24Zagorwe're idle. nothing happens or is going to happen. we can just as well just freeze time.
13:21:32*JdGordon got a visible data abort
13:21:37kugeldid you double check that assumption?
13:21:55Zagorcan you think of a case?
13:22:07kugelthere is code that depends on this. I can't tell you all the places or if they're active in RaaA
13:22:09Zagorwe don't do much mathematical code proofs here :)
13:22:32kugelwe shouldn't have misbehaving macros, still
13:23:21ZagorI don't see how that is relevant to my fix
13:23:44kugelstopping the tick makes e.g. TIME_AFTER wrong
13:24:01Zagortime has *stopped*
13:24:24kugeltime doesn't stop
13:24:36ZagorI just did
13:24:53Zagorwe don't do wall-clock in rockbox, other than in the status bar.
13:25:37kugelto be clear: I'm against this hack, because it's a) a hack which breaks sensible assumptions and b) it's android specific
13:26:13ZagorI disagree your assumptions are sensible
13:26:15kugelI don't think we need this kind of code in our repo
13:26:58Zagoruh, what kind of statement is that???
13:27:34kugelthe problem it tries to solve isn't bad enough to warrant such hacks IMO
13:28:07Zagor"the rockbox app drains battery" isn't bad enough to fix?
13:28:26JdGordonyou're wrong, TIME_BEFORE/AFTER isnt used anywhere tick time matters
13:28:30kugelnobody so far reported stronger battery drain, in contrast the battery life is better since rockbox doesn't eat as much for decoding
13:28:54Zagorkugel: I reported it. and proved it.
13:29:14Zagorand whole problem is that we drain battery when the user thinks he has turned *off* the app.
13:29:20kugelproved by looking at the 2% cpu usage?
13:29:34kugeldid you actually measure the battery life?
13:29:42Zagorwhat is wrong with you?
13:30:03Zagorthe app is closed. we should not be running *at all*!
13:31:01kugelthen we should just exit entirely
13:31:21kugelthough, that's also not the best solution because of the large startup time
13:31:32Zagorexactly. we should just stop running.
13:39:24JdGordonok, so not having commit access does suck... can someone quickly commit a major fuckup fix?
13:39:49Zagorsure, just pastebin it
13:40:02 Join Thra11 [0] (
13:41:13JdGordonI'm not sure that will be hit but it for sure a bug :)
13:44:24CIA-14New commit by 03zagor (r30716): Only alloc if necessary. Patch by Jonathan Gordon.
13:45:33 Quit hilbert (Quit: Computer has gone to sleep.)
13:45:47 Quit mortalis (Ping timeout: 240 seconds)
13:46:15 Join hilbert [0] (
13:46:42JdGordonnot quite :) That stopped it crashing if someone gave a font_id which hadnt been allocated, or during searching a font id was unloaded and the font requested didnt have that glyph
13:47:23Zagorheh, almost :)
13:47:45JdGordonso yeah, unlikely to be hit but good to check for anyway
13:48:04JdGordonchar *foo[10] would be equivilant to char foo[40] right?
13:48:22JdGordonthe * was there accidentaly... hmm..
13:48:31Zagorsize-wise, that is
13:48:43JdGordonim getting a crash somewhere after show_splash, so getting somewhere at least
13:48:46JdGordonnot ery helpful though
13:50:15CIA-14r30716 build result: 0 errors, 7 warnings (zagor committed)
13:50:30amiconnkugel, Zagor: I don't think a tickless design (or even variable tick) would work on native targets, especially not on the older ones
13:50:52Zagoramiconn: due to the polling needed?
13:51:06amiconnThe tick is our only timekeeping method, and the SoCs don't have alternative methods of timekeeping like desktop CPUs
13:51:27Zagortickless timer doesn't mean we lose track of time
13:51:38Zagoronly that we increment different amounts each "tick"
13:52:01amiconnVariable timer will also be difficult and prone to drift/ losing time
13:52:45Zagoractually there is a case to be made that variable timer is *more* accurate since we don't do as many
13:53:45Zagorbut as I've said I think it's more a "nice to have" thing on hardware targets. we don't sit idle for long times on them, as opposed to raaa.
13:53:55amiconnAtm we're using the timers in repeat mode. This is inherently reliable and quite stable. Allowed interrupt latency is up to a little less than a tick interval
13:54:18amiconnIf you want it variable, you need single shot, and then latencies add up
13:54:26***Saving seen data "./dancer.seen"
13:54:46Zagorwhy would you have more latency with single shot?
13:55:24LinusNbecause the time between the irq and the setup of the new timer interval is unknown
13:55:46amiconnYou get the interrupt, enter the isr (possibly after some delay if either a higher prio interrupt is being serviced, or interrupts were disabled), then reprogram the timer
13:55:55kugelin repeat mode the time spent in the handler doesn't add to the timer interval
13:55:55Zagorright, very true
13:56:17amiconnYou have no means to track the time between the interrupt being triggered and actually being processed
13:57:12kugelI dont think that drift is a problem
13:58:02amiconnThat depends how often you reprogram the tick
13:58:18amiconnIf you do it every cycle, it very much is
13:58:44amiconnIf you're running in repeat mode and only change the tick frequency every now and then, it might be okay
13:58:56LinusNwe can minimize the drift by always using reload, and compensate for how long the timer has run when reprogramming it. then we lose only a few timer clocks
13:59:03kugelamiconn: that was my proposal
13:59:57kugelbut even in one-shot, a drift of a few nanoseconds once every few (worst case being ten) milliseconds doesnt sound problematic
13:59:58amiconnWe're already doing something similar on coldfire when boosting/ unboosting. Sinec the timer is running on core clock, we have to reprogram the prescaler in order to keep the timer frequency constant
14:00:26amiconnBut there we're using repeat mode, and we don't reset the timer itself
14:01:21kugelLinusN: on newer targets the timer auto-reload even
14:01:33amiconnLinusN: That only works reliably if you can read the timer value in an atomic operation. Not sure whether this is possible on all SoCs
14:01:56 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
14:03:39Zagortickless on hw is a red herring. I only want to stop the idle cpu use on raaa.
14:05:13amiconnIdle poweroff?
14:06:15Zagoralmost. but we only want to do it when the app is closed/minimised
14:06:49Zagorand it should be able to resume without having to reload/restart
14:06:53kugelidlepoweroff works as epected
14:07:23kugelZagor: you think that's better than exiting entirely and restarting?
14:07:35Zagorwhat is?
14:07:41kugelI think exiting entirely would be best combined with faster startup time
14:08:07Zagorno, we must be able to resume to where the app was closed. and fast.
14:08:16Zagorthink phone call
14:08:39kugelphone call works right nwo
14:08:51Zagoryes, but is using 2% cpu
14:10:06Zagorunloading the app from ram as soon as the user goes to the home screen or anohter app is just silly
14:10:30Zagorwe should just stop running until the user brings us back (or android unloads us)
14:10:52Zagorwhy is that so strange to you?
14:11:04kugelit's not
14:12:01kugelbut with really fast startup time (like 200ms) we can even unload from ram
14:12:17Zagorbut we don't want to. there is no reason why we'd do that.
14:12:32kugelwhy wouldn't we want to do that?
14:12:36kugelfreeing unused ram is a good thing
14:12:46kugelif it comes without drawback
14:12:48Zagorbecause it is a lot of unnecessary work and wait?
14:13:23kugelwhat wait?
14:13:28Zagorplus we don't save the state enough to come back exactly where we were in menus, database or file browser for instance
14:13:41Zagoryou just said 200ms
14:14:09kugel200ms is effectively instant
14:14:30Zagorand completely unnecessary
14:14:37Zagorwill you stop this nonsense?
14:14:44kugelthe screen transistion animation takes longer than this
14:15:29 Join matze` [0] (
14:15:47kugelwill do
14:15:51LinusNstill, it's 200ms that the cpu can spend doing better things
14:18:35Zagoron another note, I'm unhappy about the current trend of ignoring warnings introduced in commits. we want a clean green build table, people!
14:19:46JdGordondisable that build client :)
14:20:21Zagoryeah, punish every build client that dares produce warnings. that'll tech em! ;)
14:20:25 Quit petur (Read error: Connection reset by peer)
14:25:00 Quit factor (Quit: Leaving)
14:25:22JdGordonwas buflib ever properly tested on low mem targets?
14:25:38 Quit Jak_o_Shadows1 (Remote host closed the connection)
14:26:12kugelwell, I tested a bit on my clipv1
14:26:33kugelanyway, buflib alone didn't increase memory usage
14:27:01kugeli also tested on hwcodec
14:27:36 Join [Saint_AndChat] [0] (~Saint]@
14:27:47JdGordonok, was allocing largeish blocks and having things moved around tested properly on low mem targets?
14:28:52 Nick [Saint_AndChat] is now known as [Saint_] (~Saint]@
14:28:58kugelI think there's a problem in the shrink_callback if it actually runs oom. are you hitting this?
14:29:19kugelbut you'd need to alloc like 500k on lowmem to hit this
14:29:21 Nick [Saint_] is now known as [Saint__] (~Saint]@
14:29:28JdGordoni dont know what I'm hitting
14:29:44JdGordonfor some reason im crahsing in viewportmnaager_init() during boot
14:30:18kugeldid you try a sim configured with very low memory?
14:31:58JdGordonsi works fine
14:36:44 Join petur [0] (~petur@rockbox/developer/petur)
14:37:29 Quit [Saint__] (Quit: AndChat - Android IRC Client -
14:40:47 Join ScaV [0] (~ScaV@
14:42:46ScaVhi, where i can see RockBox daily build's changelog?
14:43:04wodzScaV: svn changelog is on the frontpage
14:44:01ScaVwodz: ah, sorry for stupid question..
14:45:01 Join TheLemonMan [0] (
14:48:46 Join WalkGood [0] (~4@unaffiliated/walkgood)
14:52:01ScaVMy iPod nano 2g hangs when I select an album from the picture flow. But if you open the album, which is now played, then everything is OK.
14:52:24 Join Llorean1 [0] (
14:54:38 Quit Llorean (Ping timeout: 248 seconds)
14:58:00 Nick Llorean1 is now known as Llorean (
14:58:11 Quit Llorean (Changing host)
14:58:11 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
15:03:18kugelZagor: sorry if I was harsh or something. it's just that I would like to see a solution which is beneficial for more targets than android
15:05:20 Quit antil33t (Read error: Connection reset by peer)
15:05:44 Join antil33t [0] (
15:09:32 Quit wodz (Quit: Leaving)
15:13:05JdGordonanyone done an ipod manual restore recently?
15:16:06linuxstbJdGordon: Why is "recently" relevant?
15:16:57JdGordonI guess it isnt, just figured they'd be more likely to remember the steps
15:17:25JdGordonthe wiki instructions dont seem to be working for me
15:29:01JdGordonwoohoo! took long enough, back in buisness
15:30:19LloreanSo, was there something wrong with the steps?
15:33:22 Join God_Eater [0] (93722cc8@rockbox/staff/GodEater)
15:44:23 Quit TheLemonMan (Quit: WeeChat 0.3.5)
15:45:08 Join fatline [0] (
15:52:06 Join [Saint_] [0] (~Saint]@
15:54:30***Saving seen data "./dancer.seen"
15:56:29 Quit antil33t (Read error: Connection reset by peer)
15:56:55 Join antil33t [0] (
16:01:02 Quit hilbert (Ping timeout: 255 seconds)
16:08:07 Join hilbert [0] (~hilbert@
16:12:57 Join mortalis [0] (
16:13:28 Quit fatline (Quit: This computer has gone to sleep)
16:24:59 Join krey_ [0] (
16:26:16 Quit hilbert (Quit: Computer has gone to sleep.)
16:26:54 Join hilbert [0] (~hilbert@
16:28:52krey_hello, I've submitted a bug report (, and I read that chances of it getting any attention are greatly increased by irc presence, so here I am, anyone interested in helping?
16:32:05 Quit ScaV (Remote host closed the connection)
16:32:28 Quit [Saint_] (Quit: AndChat - Android IRC Client -
16:32:37 Part LinusN
16:34:58 Quit mortalis (Quit: CGI:IRC (Ping timeout))
16:37:25 Join [Saint_] [0] (~Saint]@
16:38:38 Quit Topy44 (Ping timeout: 248 seconds)
16:40:30Lloreankrey_: When you say "the problem persists in disk mode" do you mean when booting into the dedicated disk mode?
16:41:04 Join othniel [0] (
16:41:43 Join Topy44 [0] (
16:42:19CIA-14New commit by 03zagor (r30717): Changed #warning to a TODO comment.
16:43:56 Quit othniel (Client Quit)
16:44:29 Join othniel [0] (
16:44:29krey_Llorean: yes
16:45:56 Join mortalis [0] (
16:45:58 Quit othniel (Client Quit)
16:46:55 Quit mikroflops (Ping timeout: 256 seconds)
16:47:46CIA-14r30717 build result: 0 errors, 6 warnings (zagor committed)
16:48:21 Part Zagor
16:50:28 Join mikroflops [0] (
16:51:50Lloreankrey_: That's Apple code. If that's not working, it's not a Rockbox bug you're experiencing.
16:52:48 Join othniel [0] (
17:02:17 Quit [Saint_] (Quit: AndChat - Android IRC Client -
17:02:35 Join [Saint] [0] (~Saint]@
17:09:03 Join 31NAACFM5 [0] (
17:10:13krey_Llorean: but it worked before I installed rockbox. should I uninstall it and see if I can mount in disk-mode?
17:11:18krey_Llorean: I think it might be connected to this issue:
17:11:23Lloreankrey_: You can do that, but disk mode really is something that's run outside of Rockbox. If it's not working, there's not something Rockbox can do about it.
17:11:52Lloreankrey_: That report notes it *does* work in disk mode.
17:12:37krey_Llorean: "the device does work as a HID" <−− I think this might be the problem
17:13:07LloreanDisk Mode doesn't involve any Rockbox code. Including the HID code, which isn't even loaded off the disk at that point.
17:21:15krey_Llorean: I fixed the issue by reinstalling rockbox. thanks for you for helping
17:22:02 Join y4n [0] (y4n@unaffiliated/y4ndexx)
17:22:12 Join Rob2222 [0] (
17:22:54 Part WalkGood
17:24:04 Quit Rob2223 (Ping timeout: 260 seconds)
17:24:05 Join fatline [0] (
17:26:10 Quit Thra11 (Read error: Operation timed out)
17:27:00 Quit GeekShadow (Remote host closed the connection)
17:30:03 Join Thra11 [0] (
17:38:37 Quit mortalis (Quit: CGI:IRC)
17:43:16 Quit [Saint] (Quit: AndChat - Android IRC Client -
17:44:11 Join [Saint] [0] (~Saint]@
17:46:35 Nick [Saint] is now known as Guest10213 (~Saint]@
17:54:04 Quit Guest10213 (Quit: AndChat - Android IRC Client -
17:54:15 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:54:32***Saving seen data "./dancer.seen"
17:55:02 Quit krey_ (Remote host closed the connection)
17:55:53 Quit petur (Quit: *plop*)
17:56:51 Quit tchan (Quit: WeeChat 0.3.5)
17:57:05 Join swilde [0] (
17:57:10othnielOn the FeatureComparison wiki page will someone add "FM Recording codec/samplerate/kbps" in Other features section −− after "Radio Recording" (should be FM Radio Recording imho) My Stock e280v2 records FM in .WAV the codec is pcm_s16le/24000 stereo/768kbps according to ffmpeg. 353,894,400 bytes per hour. Whereas my e280v1 has rockbox and can record in MP3. Or you may want to break the change out into several lines: FM re
17:57:12othnielcord file format, FM record codec, FM record samplerate, FM record kpbs, This would help people compute Maximum FM record time considering the DAP capacity/record velocity −− not battery life. Or your could put a line for max time also. An 8GB stock firmware e280v2 then gives 22.6 hours FM record time at 768 kbps. Not good. Rockbox FM record time at 128 kbps MP3 gives 135 hours. For lecture recording, I prefer the 8kHz s
17:57:14othnielampled adpcm_ima_wav codec stereo (64 kbps), but the Sansa e280 has better battery life than the Chinese S1mp3 that records in that format. FFmpeg can be used to convert recording into SPEEX or AMR or G.729.
17:57:16othnielAlso, who can I talk to about .m4b support in rockbox? Those containers have an AAC audio stream and PNG artwork that changes every few minutes. I can provide samples or a download link. Wouldn't think it would be much effort.
17:58:30 Join n1s [0] (~quassel@rockbox/developer/n1s)
18:01:35 Join GermanMushroom [0] (
18:02:00 Join pamaury [0] (
18:02:00 Quit pamaury (Changing host)
18:02:00 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
18:02:04othnielI don't have write access and I suggest a footnote on "FM Radio Recording" It should say only on Iriver H3x0, Sansa e200v1/v2, Sansa c200/m, etc. I found a FAQ with that info but it is way buried.
18:05:44othnielMaybe MP3 Realtime Encoding and Wavpack Realtime Encoding should be moved down under FM Radio and Voice Recording?
18:08:22 Join mortalis [0] (
18:08:24n1sothniel: you can get access to change the wiki if you want
18:09:12othnielok but I dont want to get fussed at for making a lot of changes.
18:09:23othnielI want to run them by somebody first.
18:09:45othnielI'll register
18:10:13othnielwhat do I do / where?
18:11:44n1sothniel: click the deit link on a wiki page and cancel the login prompt and then fill out the reg form, come in here and promise not to spam and someone will give you write access
18:12:17othnielFilling out Wiki main Registration now.
18:14:32n1sthe feature comparison page has not been updated for most our targets btw
18:17:23n1smuch of the info in the first table seems very archos centric
18:17:43n1sand a lot of it is outdated for even those
18:18:35n1shaha, we have "Corrects reported bugs" checked as "Yes" for rockbox :)
18:19:24n1sthere's also duplication between the two tables
18:20:12n1sIMO the first table should go away and if anything usefull is in it that's missing from the other table it should be added in there instead
18:20:12 Join AlexP [0] (~alex@rockbox/staff/AlexP)
18:20:41saratogawe should probably remove that feature comparison until someone makes a new one
18:21:18othnielI like the feature comparison
18:21:35othnielIt is for folks who are wanting to get into rockbox
18:21:49othnieland learn what it is −− quickly
18:21:51saratogayeah but its so old its basically useless :)
18:22:00othnielI will fix it.
18:22:11othnieland organize it
18:22:11 Quit GermanMushroom (Ping timeout: 240 seconds)
18:22:15LloreanThis is a better page to improve, I think:
18:22:17n1ssaratoga: any comments on the patch i posted in FS #11856
18:22:18fs-bluebot 3Set of ogg vorbis files will not play (bugs, unconfirmed)
18:22:34saratogan1s: i never figured out a good work around, so if yours works go for it
18:23:07n1sit's a fairly ugly hack but seems to work fine
18:23:14saratogaothniel: should probably split it into an archos comparison page, and a comparison page for everything else
18:23:24saratogaand then don't mention specific models
18:23:47saratogasince aside from archos, most everything else is pretty similar
18:25:11 Join ChickeNES [0] (
18:26:49othnielHow long until I receive the email confirmation?
18:29:10othnielI am looking for a place to detail FM recording.
18:29:25othnielI will put some on 'WhyRockbox'
18:29:59othnielbut the other seemed as good as any place I could find.
18:30:04n1suh, the email ,ight not come there's been some trouble with that
18:30:20othnielI am interested not in recording music, but lectures/seminars
18:30:43othnielrecording it in AMR, SPEEX or similar voice codecs and
18:30:50othnielof course playback.
18:31:03othnielnot 320Kbps mp3
18:32:23othnielThe Buyers Guide is the only place that says what hardware you need to purchase to do FM recording with rockbox.
18:32:28n1sdoes anyone remember if the wiki registration works even without the mail?
18:32:40othnielI put my email in.
18:33:04n1syeah, but the registration confirmations never arrive
18:33:44n1sothniel: try now
18:33:49n1sto log in
18:34:19 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:34:40othnieldont know where
18:34:54n1sothniel: click edit on any wiki page
18:37:09othnielI'm in
18:37:20 Join Buschel [0] (
18:37:25 Quit Thra11 (Ping timeout: 256 seconds)
18:37:25othnielI'm good with HTML
18:45:30 Join Jerom [0] (~jerome@
18:49:30 Quit antil33t (Read error: Connection reset by peer)
18:49:55 Join antil33t [0] (
18:50:00 Quit ej0rge (Ping timeout: 258 seconds)
18:50:10 Join [Saint] [0] (~Saint]@
18:50:53 Join ej0rge [0] (~alhaz@
18:51:30 Join Thra11 [0] (~thrall@
18:52:23 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:55:21 Join freddyb [0] (
18:58:34 Join GermanMushroom [0] (
18:59:51 Quit mortalis (Quit: CGI:IRC)
19:00:40 Join kadoban [0] (
19:02:05 Quit n1s (Ping timeout: 276 seconds)
19:04:23freddybHow bad is it to block all font relocations while reloading the skin data? It fixes some crashes for me related to switching themes... I put a patch on FS #12298.
19:04:23fs-bluebot 3no display / fatal error on Sansa Clip (bugs, new)
19:04:43freddybIt didn't fix the Clip problem, tho.
19:05:42kugelfreddyb: that means the move callback is still incorrect, no?
19:08:21 Join RansomTime [0] (~RT@wikia/vstf/Randomtime)
19:08:33freddybI don't think so. I think it has more to do with some (struct font *) that's not forgotten outside of font.c but I haven't found it. I don't know much about viewports, scrolling text, etc.
19:09:30kugelI see
19:09:48freddybOr maybe some other memory pointer is not updated and a font is moved into its area.
19:10:10kugelwhat's left using struct font*? wasn't most access just for the font height and that replaced by a wrapper function?
19:10:42kugelI guess I shall commit the debug alloc patch
19:14:02freddybI haven't really found much that uses *pf's but what i found I locked in the patch on the FS. If I knew the skin engine better I would try delaying the repaint until the allocations are done.
19:14:56kugelfreddyb: pretty sure it doesn't draw during allocations
19:15:09kugelallocs are made in the parse-phase
19:15:39 Join bertrik [0] (
19:15:39 Quit bertrik (Changing host)
19:15:39 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:19:13 Quit ChickeNES (Quit: Computer has gone to sleep.)
19:21:06 Quit [Saint] (Remote host closed the connection)
19:30:48 Join Horscht [0] (
19:30:49 Quit Horscht (Changing host)
19:30:49 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:31:16 Join fyrestorm [0] (
19:34:37 Quit GermanMushroom (Quit: Ik ga weg)
19:37:14othnielok I'm having problems saving my changes.
19:37:24othnielFirst set of changes worked fine
19:37:40othnielI tried to save up a few and do them together.
19:38:03othnielbut preview could not render html links properly.
19:38:24othnielI saved and tried to make more changes −− now it rejects my attempts.
19:38:42AlexPLinks on the wiki don't work like normal html IIRC
19:39:04AlexPIt is something like [[Text][http://link...]] IIRC
19:40:11LalufuHm. I just noticed that rockbox does not seem to know the concept of a null pointer exception. That'll take some getting used to.
19:40:28AlexPOh, other way around
19:40:39othnielSo I have to put [] around the whole thing and alternate text inside and link inside with extra sets of []
19:40:50AlexPYou do [[link][text]]
19:40:58AlexPHave a look at the source of another page
19:41:08AlexPBut it isn't html
19:41:29othnielYeah this page is full of <a href="#footnote22" etc
19:41:36 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20110928134238])
19:42:03AlexPSorry, I didn't mean not at all
19:42:08AlexPbut have a look at
19:43:54CIA-14New commit by 03kugel (r30718): Revert r30599 "Check for the magic file "/.rockbox/skin_buffer_size.txt" [...]" ...
19:44:27othniel/usr/bin/rcs -o&#37;REVISION&#124;N&#37; &#37;FILENAME&#124;F&#37; of .../Main/FeatureComparison.txt failed:
19:44:31othnielGo back in your browser and save your changes locally.
19:46:41CIA-14r30718 build result: 0 errors, 2 warnings (kugel committed)
19:47:26kugelhm, my build client quite bumped in the ranking
19:49:10othnielAttention: Error saving topic During save of Main.FeatureComparison an error was found by the version control system. Please notify your Wiki administrator.
19:49:18othnielWhat do I do now?
19:50:25AlexPothniel: Try again, or wait until Zagor is around and ask him if he has any idea
19:54:29 Quit antil33t (Read error: Connection reset by peer)
19:54:34***Saving seen data "./dancer.seen"
19:54:56 Join antil33t [0] (
19:55:24othnielWhen you say try again. Do I just hit the back button on the browser and press save changes. or start over and hit edit make the change and press save...
19:56:09AlexPI don't know
19:58:08 Join Strife89 [0] (~Strife89@
19:58:42kugeldidnt JdGordon have a lcd_puts* overhaul patch somewhere?
20:01:27 Quit hilbert (Quit: Computer has gone to sleep.)
20:02:54 Quit bluefoxx (Ping timeout: 248 seconds)
20:05:09 Join hilbert [0] (~hilbert@
20:18:42 Quit freddyb (Quit: Leaving)
20:25:37 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
20:27:56 Quit mudd1 (Quit: Ex-Chat)
20:28:25 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
20:29:05 Quit saratoga (Ping timeout: 252 seconds)
20:32:21CIA-14New commit by 03kugel (r30719): core_alloc: Provide a tiny test allocation, which can be freed for debug purposes. ...
20:32:45kugelthe above exposes a crash in the font code
20:34:51CIA-14r30719 build result: 0 errors, 1 warnings (kugel committed)
20:38:07kugelin an sdl-app 320x480 build at least
20:38:26 Join ChickeNES [0] (
20:43:20 Quit ChickeNES (Client Quit)
20:46:16 Join ChickeNES [0] (~ChickeNES@
20:54:54kugelright. the font move callback is fine, but something else corrupts the cache
21:01:16 Quit Jerom (Quit: Leaving.)
21:06:26 Join Thra11_ [0] (~thrall@
21:09:35 Quit Thra11 (Ping timeout: 260 seconds)
21:12:39 Quit ChickeNES (Quit: Computer has gone to sleep.)
21:21:47 Quit soap_ (Ping timeout: 240 seconds)
21:23:21 Join GermanMushroom [0] (
21:23:26 Quit pamaury (Ping timeout: 248 seconds)
21:23:27 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
21:24:31 Join robin0800 [0] (~robin0800@
21:25:12othnielZagor I was told to talk to you about wiki updates that are not being accepted by rcs
21:25:46Zagorthat sounds bad!
21:25:56Zagorwhat does it say?
21:26:14othnielI make some cnanges and try to save.
21:26:43othnielit works sometimes and sometimes it dont. Is it me or the system?
21:27:15ZagorI would like the exact error message, if possible
21:27:36othnielothniel: /usr/bin/rcs -o&#37;REVISION&#124;N&#37; &#37;FILENAME&#124;F&#37; of .../Main/FeatureComparison.txt failed:
21:27:40othnielothniel: Go back in your browser and save your changes locally.
21:28:11othnielAnd then it decides to work
21:28:16ukleinekZagor: that makes me yike just because it includes the word "rcs" :-)
21:28:18othnielso I make some more changes a
21:28:24Zagorukleinek: hehe
21:28:35othnieli thought it was svn
21:28:53othnieland when I save no workee
21:34:53othnielIt just worked for me.
21:35:13othnielWhat did you do Zagor?
21:35:32Zagorthinking about upgrading :)
21:35:44othnielThanks −− next time I ask I want you to do the same thing...
21:36:09Zagorgreat. I'll keep thinking about it :)
21:36:21othnielIts just intermittent
21:36:44othnielI wonder if checkins have something locked.
21:36:50Zagoryeah. there's a number of odd issues with the wiki. maybe some are fixed in the later version.
21:36:52othnielin the WIKI
21:37:18othnielLike my changes in the Firmware Feature Comparison Page?
21:37:47 Quit Thra11_ (Ping timeout: 240 seconds)
21:39:57 Quit GermanMushroom (Quit: Ik ga weg)
21:45:14 Quit kadoban (Ping timeout: 276 seconds)
21:46:19AlexPothniel: Has it told you that you are suspicious yet?
21:46:35othnielOh yeah
21:46:51othnielI dont think I am.
21:47:01AlexPNo, it tends to tell everyone they are :)
21:47:15othnielI thought it was funny....
21:47:27AlexPOf course it is probably right in many cases, although for the wrong reasons
21:47:35othnielDo YOU like my changes?
21:47:45othnielIts a bot
21:47:52othnielwhat do they know?
21:49:10othnielI just keep trying out new changes in the preview until the commit takes.
21:52:52 Join Thra11_ [0] (~thrall@
21:54:37***Saving seen data "./dancer.seen"
21:57:18 Join ChickeNES [0] (~ChickeNES@
21:57:35Zagordon't we support m4b files? I thought that was just m4a flagged "resume enable"?
21:57:57LloreanZagor: Some of them have chapters, we don't support the chapters.
21:58:20othnielThe m4b files I have contain artwork
21:58:45othnielThey are aac audio with png graphics which changes every 5 minutes
21:58:50othnielor so
21:59:12othnielYes they are bookmarkable but when rockbox hits the png in the data stream
21:59:40othnielit doesnt throw it away or display it −− it aborts playback and goes to the next track.
22:00:20othnielUsually the artwork is early in the podcast/magazine
22:00:27othnielso no listen
22:01:10LloreanRockbox is rather picky with the AAC files it supports. They need to be optimized for streaming (which involves moving all the metadata to the front). It sounds like those aren't.
22:01:31 Quit Llorean (Read error: Connection reset by peer)
22:01:40othnielOk so I still have a use for my Ipod classic.
22:02:29 Join Llorean [0] (
22:02:37 Quit Llorean (Changing host)
22:02:37 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
22:02:52othnielyeah my m4b have chapters to separate the articles in the magazine.
22:02:54 Quit Horscht (Quit: Verlassend)
22:03:15 Quit hilbert (Quit: Textual IRC Client:
22:05:02Zagorwohoo, added the symlink
22:05:49Zagornow works again
22:06:01LloreanGood news!
22:06:12 Join Horscht [0] (~Horscht@xbmc/user/horscht)
22:06:51othnielI guess you were commenting on my flagging of no .m4b support in the FeatureComparison −− changed to "No chapter support"
22:07:03 Quit y4n (Quit: Today is the perfect day for a perfect day.)
22:08:48othnielno PNG support either Right?
22:09:32 Join kadoban [0] (
22:10:05Zagorno support for anything in the middle of the files
22:11:04LloreanRockbox only supports BMP and JPEG album art, and only JPEG as embedded art.
22:11:13LloreanThat's covered in the manual.
22:11:57othnielWould the mp4player handle it?
22:12:15LloreanI don't even know what you mean by "the mp4player"
22:12:29othnielIts a plugin for some targets
22:13:27othnielSee rockbox extras
22:14:10LloreanI don't see anything called "mp4player" on the extras page.
22:14:10 Quit kadoban (Ping timeout: 255 seconds)
22:14:26LloreanDo you mean mpeg player?
22:14:38LloreanThat's completely unrelated to mp4.
22:14:48LloreanIt's a video player that plays mpeg 1 and mpeg 2 video.
22:15:02LloreanIt really helps if you use the proper names of things.
22:15:43othnielYes that is one of my pet peeves also, But an m4b is just like a mpeg movie.
22:15:57othnielExcept it is mostly audio with pictures
22:16:09othnielnot moving pictures either.
22:16:37scorche|shthere are many different types of "mpeg movies
22:16:38LloreanAn m4b is an AAC stream in an MP4 container.
22:16:49 Quit thomasjfox (Remote host closed the connection)
22:17:07LloreanThat's a different type of audio, a different type of container, and no video whatsoever compared to MPEG1/MPEG2 movies.
22:17:20LloreanThe only similarity is that both are part of the MPEG umbrella.
22:17:28othnielok but from where I sit they are similar.
22:17:56othnielYet rockbox supports one and not the other
22:18:14 Quit user890104 (Read error: Operation timed out)
22:18:21LloreanMP3 is more similar to MP4 than MPEG1 video files are.
22:18:28 Join user890104 [0] (
22:19:00 Join RedWork [0] (~dmaker@
22:19:11othnielThe .m4b files I was asking about have an AAC audio stream and PNG images every 3 to 5 minutes.
22:19:15LloreanRockbox plays AAC audio in MP4 containers just fine as long as the metadata is optimized for streaming.
22:19:25othnielIt would be nice to listen to them
22:19:33Llorean.m4b files are just AAC in MP4, but the ones you're describing have metadata that gets in the way
22:19:45othnielBetter to actually show the images as they are encountered in the stream.
22:19:48LloreanRun it through an app like MP4box that fixes the metadata and they ought to play
22:20:48othnielI know how to use ffmpeg command line, but I'd like to see the images as they are encountered.
22:20:54LloreanThen fix it.
22:20:56LloreanIt's open source.
22:21:15othnielI'd love to.
22:21:18LloreanIt's more work than you seem to think it is, and there's very little interest in actually *doing* the work from the people, like you, who want it done.
22:21:18RedWorkafternoon folks
22:21:24othnielI'm trying to get up to speed.
22:21:44RedWorkI'm attempting to load RB on a Toshiba GigaBeat S, following the instructions here:
22:21:58 Join LinusN [0] (
22:22:08AlexPothniel: Part of the trouble is is that while there is a work around, not many of the devs are going to put the time into something they don't use
22:22:24RedWorkI get as far as beastloader saying it's complete and rebooting, but it comes up in toshiba recovery mode, not BR mass storage
22:22:35Lloreanothniel: First, someone would need to add support for PNG album art in general.
22:22:38RedWorkI've tried hard rebooting with the battery switch
22:22:45Lloreanothniel: Then they'd need to fix the MP4 parser to deal with mid-file metadata
22:22:49othnielRight. I'm getting a Sansa Fuze+ this week
22:22:52RedWorkalexp, yes, beastpatcher.exe
22:22:53Lloreanothniel: Then they'd probably also have to fix it to deal with exceedingly long files.
22:22:59AlexPLlorean: image viewer supports png IIRC, so at least we have some code
22:23:05AlexPBut not in core
22:23:08LloreanAlexP: Yes.
22:23:10othnielI've got e280v1 (3) and 2 e280v2
22:23:30AlexPRedWork: Are you trying to use the single or dual boot bootloader?
22:23:43othnielAn Ipod Classic 80GB
22:23:47RedWorksingle, I've no use for the original firmware
22:23:54 Quit robin0800 (Ping timeout: 258 seconds)
22:24:09AlexPRedWork: Try dual
22:24:21RedWorkalexp, alright, I'll give it a try
22:24:23AlexPRedWork: The Toshiba bootloader can be very randomly picky
22:24:33AlexPRedWork: And it seems to have fewer problems with dual
22:24:49LloreanYeah, a lot of people seem to have better luck with dual boot than single boot, with certain versions of the Toshiba firmware.
22:24:59AlexPRedWork: For some people single is fine, others have problems. For all those that have had problems with single, dual has worked
22:25:00RedWorkalexP, i've gotten that impression (picky), but didn't know dual was less so
22:25:13RedWorkthanks for the help, I'll give it a try
22:25:20AlexPfingers crossed :)
22:30:15 Join robin0800 [0] (
22:33:55RedWorkalright, the link for is 404, so nk.bin is looking to be difficult to get
22:34:51 Quit Stummi (Quit: Bye!)
22:40:19othnielAlexP you still there? I've reorganized the Feature Comparison page to present information more clearly. Any feedback?
22:40:27AlexPone mo
22:41:42 Quit benedikt93 (Quit: Bye ;))
22:42:22 Join Jerom [0] (~jerome@
22:42:27AlexPothniel: looks fine to me
22:42:52 Quit ChickeNES (Quit: Computer has gone to sleep.)
22:42:54AlexPI think the biggest issue there is the difficulty of comparing RB to the vast number of different OFs
22:43:03othnielLot of missing tags on footnote refs
22:43:18AlexPIt might be even worth dropping the comparison part and just expanding/merging with WhyRockbox
22:43:18 Join GeekShadow [0] (
22:43:45othnielYes but this page is just a general info for newer users to get them interested in Rockbox.
22:43:59LloreanThat's the intent of the WhyRockbox page.
22:44:29AlexPWhyRockbox is the one we link from the front page
22:44:43AlexPQuite a lot of it is duplicated between the two too
22:44:54AlexPothniel: Thanks for working on this by the way
22:44:57AlexPThe wiki is a mess
22:45:34othnielHey I'm an unemployed software developer that uses Rockbox. I should have helped a long time ago.
22:46:08AlexPwell late is better than never :)
22:47:18othnielI will continue to help. But I'll probably help with the Fuze+ port when I get mine.
22:49:01 Join hilbert [0] (~hilbert@
22:50:21 Quit robin0800 (Ping timeout: 245 seconds)
22:53:50 Join n1s [0] (~quassel@rockbox/developer/n1s)
22:57:26othnielA question on 10.1 in the manual. What are the kbps for PCM wave, AIFF, wavpack and mp3 the last being a function of sampling rate?
22:58:44LloreanI'm not sure what you mean
22:58:51AlexPothniel: I don't follow
22:59:24LloreanPCM and AIFF have a bitrate determined by bit depth and sample rate. On most players I guess that'll be 1411kbps.
22:59:25othnielwhen you type in ffmpeg -i file it tells you what the kbps for the file
22:59:36Lloreanwavpack's bitrate is entirely dependent on the content being recorded, since it's lossless compression.
22:59:46AlexPand wav and aiff are fixed
22:59:51AlexPmp3 you can chose
22:59:52othnielalso tells you the codec container stereo or not sampling rate etc
22:59:56LloreanMP3's bitrate is up to the user.
23:00:09othnielWav is not fixed
23:00:10AlexPyou are recording here
23:00:20Lloreanothniel: WAV is fixed based on sample rate and bit depth.
23:00:27AlexPYou set the bitrate you want the mp3 to record at
23:00:30LloreanFor each combination thereof, there is only one bitrate a wave file will be.
23:00:31othnielI have one recorder which does 768kbps
23:00:38othnieland another that does 64kbps
23:00:44AlexPothniel: It is, you can't chose wav bitrate
23:00:56othnielAgain nope it depends on the codec
23:00:58AlexPIt is uncompressed
23:01:07AlexPwav is the "codec"
23:01:11LloreanWell, PCM is
23:01:37othnielwav is the container inside is a code for which codec was used
23:01:38LloreanBut with PCM wav files, there literally is only one bitrate per sample rate / bit depth / channels combination
23:01:59LloreanA 44.1 khz, 16-bit, stereo PCM wav file will always be 1411kbps
23:02:07LloreanIf you halve the sample rate, it'll halve the kbps.
23:02:22LloreanBecause it's uncompressed it's literally samples * bits per sample * channels
23:02:33n1sthere are some weird compressed bitstreams in some wav files though
23:02:51AlexPPerhaps we should say PCM WAV or something
23:02:52n1sothniel: if you just say wav file everyone will assume you mean it's a pcm file
23:02:58Llorean44100 * 16 * 2 = 1411200 or 1411.2kbps
23:03:22othnielrecode a wav file using the adpcm_ima_wav using ffmpeg and you will see that it is not true
23:03:35n1ssee, adpcm != pcm
23:03:55AlexPadpcm is different to pcm
23:03:58Lloreanothniel: Rockbox doesn't do adpcm.
23:04:00othnielI used to think that before I bought a Chinese MP3 player that stores very efficient wav recordings
23:04:16n1sso if you are talking about different ecodings, speaking of wavfiles is entirely unhelpful
23:04:16AlexPIt isn't PCM, which is what everybody means when they say wav
23:04:28Lloreanothniel: The manual explicitly says PCM wave. Other formats that can be in the WAV container are irrelevant
23:04:30LloreanThey aren't supported.
23:04:47gevaertsThere's of course also the fact that some chinese mp3 files encode as mp3 but use .wav as the file extension...
23:04:47othnielbut it is wav and is supported by WMP
23:04:51othnielI'm learning.
23:04:54AlexPoh, we already specify :)
23:04:58n1si think we play some of those weird wavs actually but don't record them
23:05:03Lloreanothniel: What does WMP have to do with questions about Rockbox?
23:05:05AlexPothniel: but not PCM WAV
23:05:18Lloreanothniel: You asked about the bitrate for Rockbox. I answered that it's a fixed function of sample rate, bit depth, and channels
23:05:24othnielOk im beat down. I learn fast.
23:06:10AlexPheh :)
23:06:39LloreanAIFF is uncompressed as well, and follows the same rules as PCM WAV in respect to Rockbox.
23:06:45LloreanBecause wavpack is a lossless format, its bitrate is variable, and can't be predicted in advance.
23:06:54LloreanAnd MP3 as said, is selectable.
23:07:51othnielI thought the Sansa e280v2 was bad!
23:08:02AlexPhow so?
23:08:41othnielYou are saying that if I record in rockbox using WAV I will generate 1411k bps over a megabit per second?
23:09:14LloreanYou're generating that no matter what you record in. The other formats either compress it, or throw some of it away and compress it.
23:09:16AlexPdepending on sample rate etc
23:09:16othnielThat is twice as large as the e280v2 OF record.
23:09:28othnielIts not selectable
23:09:29AlexPI think the e200 is limited to 22kHz
23:09:38AlexPSo it will be 768 kbps
23:09:38LloreanThe e200v1 is. I don't know about the v2
23:09:43LloreanBut they're also mono, so that'd halve it again.
23:09:54LloreanUnless it stupidly records to stereo from a mono input.
23:10:01AlexPIt might :)
23:10:10othnielAccording to manual 10.2
23:10:10gevaertsYou could be recording from FM
23:10:56othnielThe file produced by OF is stereo.
23:11:12othnielThat's what I'm talking about.
23:11:37 Join saratoga [0] (98034408@gateway/web/freenode/ip.
23:11:58AlexPSo if it is stereo, limited to 22 kHz, and 16 bit it'll be half 1441 kbps
23:12:05AlexPuncompressed audio is large
23:12:22AlexPThis is why if you want to really keep everything, wavpack is a great idea
23:12:24Lloreanothniel: I don't see what you're referring to in the e200v2 manual section 10.2
23:12:41othnielrockbox manual 10.2
23:12:46othnielfor e200v2
23:12:48LloreanLiterally the entirety of that section of the Rockbox manual is "This sets the bitrate when using the MPEG Layer 3 format. And has no settings for the other formats"
23:13:15othnielso wav has no settings for sample rate
23:13:28LloreanNo setting for bitrate.
23:13:36LloreanBitrate is not sample rate.
23:13:40othnielOnly mono/stereo
23:13:51Llorean"Sample Rate" would be section 10.3
23:14:00LloreanFrequency = frequency of samples, or sample rate.
23:14:08Llorean10.5 is channels.
23:14:12LloreanI see no mention of the OF anywhere anyway.
23:14:13saratogashould probably be a comma rather then a period between those clauses
23:14:17AlexPFrequency you can set depending on what the hardware can do
23:14:26AlexPsaratoga: yes, not great English
23:14:37AlexPI'll change it
23:14:41LloreanAlexP: According to the e200v2 manual we support 96khz down through 8khz.
23:14:44LloreanA pretty respectable range
23:14:50AlexPyeah, pretty good
23:14:54saratogaothniel: aside from grammar, that seems quite clear to me, do you actually have some question about what its explaining?
23:15:00AlexPLlorean: It must be the v1 that is limited
23:15:17othnielSo when I read it I assumed that wav had no settings for frequency.
23:15:23LloreanAlexP: Yeah, the V1 is definitely limited.
23:15:27saratoga"Note: The 11.025 kHz setting is not available when using MPEG Layer 3 format." -< niether is anything above 48k
23:15:40Lloreansaratoga: General limits of MP3 I assume?
23:15:57saratogaalthough i think mp3 can do 11k too, but maybe our encoder can't
23:15:58Lloreanothniel: Why would something about bitrate choices lead you to assumptions about frequency choices?
23:16:15Lloreanothniel: The section on frequency doesn't mention any limitations in regard to PCM Wave.
23:16:37 Part LinusN
23:16:48saratogaperhaps 10.2 should be renamed "10.2 Encoder Settings (MP3 only)"
23:16:49 Quit ender` (Quit: "I'm the head wizard now. I've only got to give an order and a thousand wizards will ... uh ... disobey, come to think of it, or say 'What?', or start to argue. But they have to take notice."-- Archchancellor (Terry Pratchett: Lords and Ladies))
23:16:49othnielok if I'm wrong I'm glad.
23:17:19Lloreansaratoga: At least for now, that may be wise.
23:17:39Lloreansaratoga: Or perhaps since the only encoder-related setting is currently "Bitrate" just rename it to "Bitrate (MP3 only)"
23:17:54saratogayeah makes sense
23:17:59LloreanSince some people think of frequency, channels, etc as encoder settings (even though they're really source settings) and that seems to have caused confusion here.
23:18:30 Join pamaury [0] (
23:18:30 Quit pamaury (Changing host)
23:18:30 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
23:18:38saratogaanother idea might be to make that setting only appear in the menu when the encoder is set to mp3
23:18:47 Part Zagor
23:18:49saratogatrying it now it actually still exists for wav, but you can't set it, which is a bit weird
23:18:57pamaurybertrik: I finished the two battery benches
23:19:16AlexPsaratoga: What does it look like if you have one of the other formats selected?
23:19:39saratogaits listed but if you click it you get a splash saying "no settings"
23:19:48Lloreansaratoga: It might be nice if WAV could replace it with expected output bitrate based on the other options, but yeah, probably just hiding it for everything but MP3 might be nice
23:20:05AlexPIf I change the manual does someone want to change the menu?
23:20:19AlexPI like "Bitrate (MP3 only)" myself
23:20:32saratogaanyone familiar with the menu code? i can never figure out if my changes will break something
23:20:44othnielAre you thinking to make Format 1st, Frequency 2nd and Encoder 3rd on the menu?
23:21:18othnielOr should Source come 1st?
23:22:33RedWorkAlexP, that did it, thanks.
23:22:40AlexPRedWork: cool :)
23:22:49AlexPYou managed to find nk.bin somewhere?
23:23:05RedWorka buddy of mine had it stashed somewhere
23:23:20RedWorkactually had several older versions
23:23:21AlexPIt is a shame we can't distribute it, but there you go
23:23:27othniel8k Hz sample rate * 16 bit samples * stereo gives 256k bps wav recordings?
23:23:58LloreanShould be, yes.
23:24:08RedWorkthanks again, I'm out of here
23:24:12 Quit RedWork (Quit: Leaving)
23:25:37othnielI can live with that. Interesting discussion. Did not know that ADPCM was not supported in rockbox. It would cut the size from 256k bps to 64k bps
23:25:57othnielCan do that conversion with ffmpeg
23:26:00saratogaadpcm is not really that useful in recent times
23:26:10othnielThen to SPEEX
23:26:16saratogaif you want lossy mp3 is a lot better
23:26:19AlexPsaratoga: To change the menu name, it'd just be the lang files, no?
23:26:20othnielOr straight to speex
23:26:41bertrikpamaury, and what is the result of the battery benches?
23:26:43AlexPif you want lossy use mp3, if you want lossless use wavpack
23:26:48saratogaAlexP: oh i was thinking of making the menu not appear if it wasn't needed
23:26:57AlexPwav and aiff don't have a lot of uses really
23:27:01AlexPsaratoga: ah right :)
23:27:13AlexPWell renaming would be a good start I think
23:27:25othnielWell I'm recording lectures so low sampling WAV is fine for me.
23:27:33AlexPwhy not use mp3 then?
23:27:36AlexPIt'll be smaller
23:27:39LloreanYeah, using wave or AIFF is just a waste of space unless you simply do not have a tool that can deal with wavpack.
23:27:43pamaurybertrik: 14:52:39 with svn VS 14:27:43 with the #if 1
23:27:52othnielWhat is the battery life of mp3 recording vs Wav recording?
23:28:05pamauryso half an hour of difference :-/
23:28:12saratogaprobably depends on if you have flash or a hard disk
23:28:22othnielon e280v1 or e280v2
23:28:58othnielMy recording are about 3.5 hours long each
23:29:29LloreanI don't think anyone's done any real testing on the encoder efficiencies.
23:29:32othnielso bitrate is important to me to not fill up DAR
23:29:57othnielMaybe I can test and post on the WIKI.
23:30:18CIA-14New commit by 03alex (r30720): Manual: Small English improvement
23:30:38AlexPI just removed the second sentence, it was superfluous
23:32:38CIA-14r30720 build result: 0 errors, 1 warnings (alex committed)
23:32:53othnielIs the online manual for e280v2 updated AlexP?
23:33:12AlexPothniel: It will be when it is rebuilt in a few hours
23:33:27AlexPIt's done once at day at about 06:00 IIRC
23:33:33pamauryI don't know if r30717 was a good idea since the warning was to have to know you had to implement it
23:33:34othnielYou were speaking of editing section 10.2?
23:33:42pamaury*was nice
23:33:47AlexPI did, but only very slightly
23:39:26 Quit n1s (Remote host closed the connection)
23:40:48 Join robin0800 [0] (~robin0800@
23:46:11othnielWhy is the Ipod Classic mentioned on the HardDriveReplacement wiki page?
23:46:27othnielIs it a supported rocbox platform ?
23:46:50 Quit pamaury (Remote host closed the connection)
23:47:20Lloreanothniel: It's a port in development
23:47:42othnielLike the Sansa Fuze+?
23:47:57LloreanMore or less
23:48:03LloreanYou'll see them both in the same area on the front page.
23:50:26othnielDo you mean dev.cgi or some other Front Page?
23:50:53 Quit hilbert (Quit: Textual IRC Client:
23:51:09 Join hilbert [0] (~hilbert@
23:54:40***Saving seen data "./dancer.seen"
23:56:29 Join soap [0] (
23:56:30 Quit soap (Changing host)
23:56:30 Join soap [0] (~soap@rockbox/staff/soap)
23:56:45*gevaerts never even considered that anyone could think of dev.cgi as the front page...
23:58:34 Quit domonoky (Read error: Connection reset by peer)

Previous day | Next day