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 2013-07-24

00:01:47 Quit Strife89 (Quit: Heading out.)
00:05:20 Quit onder` (Quit: leaving)
00:05:58 Join onder` [0] (~onder@
00:07:48 Quit EstebanMonge (Ping timeout: 240 seconds)
00:24:33 Quit efyx (Quit: Quitte)
00:29:22 Join EstebanMonge [0] (~emonge@
00:31:57 Join Strife89 [0] (
00:32:22 Quit pamaury (Ping timeout: 246 seconds)
00:45:09 Part EstebanMonge
00:50:02 Join arcaicofl [0] (
00:52:02 Quit arcaicofl (Max SendQ exceeded)
00:52:31 Join arcaicofl [0] (
00:53:56 Quit mindflash (Ping timeout: 276 seconds)
01:04:34***Saving seen data "./dancer.seen"
01:10:44 Part ZincAlloy
01:12:22 Quit gelraen (Read error: Connection reset by peer)
01:12:40 Join gelraen [0] (
01:20:06 Quit liar (Ping timeout: 245 seconds)
01:35:17 Join JdGord [0] (
01:47:16 Quit arcaicofl (Quit: Saindo)
01:49:50 Quit JdGord (Ping timeout: 276 seconds)
01:50:29 Join JdGord [0] (
02:00:14 Quit JdGord (Ping timeout: 276 seconds)
02:00:36 Join JdGord [0] (
02:02:22 Join AndChat|72624 [0] (
02:06:02 Quit JdGord (Ping timeout: 256 seconds)
02:11:31 Join belak [0] (~belak@facebook/engineering/belak)
02:30:58 Quit AndChat|72624 (Quit: Bye)
02:31:11 Join JdGord [0] (
02:31:14 Quit JdGord (Client Quit)
02:32:55 Quit Strife89 (Quit: This computer has gone to sleep)
02:33:28 Join tertu [0] (
02:36:57[Saint]copper: *required* to bitch about it? :)
02:37:22[Saint]It doesn't seem like you mind that requirement.
02:38:48[Saint]copper: please don't confuse the EQ and volume step, you may have missed it, but, the two are unrelated.
02:39:31[Saint]I first thought they were not, it seems as though the volume step change would've happened if the EQ resolution wasn't changed anyway.
02:48:44 Join uwe__ [0] (
02:52:27 Quit uwe_ (Ping timeout: 245 seconds)
03:04:38***Saving seen data "./dancer.seen"
03:43:25 Quit belak (Quit: belak)
04:02:33 Quit [Saint] (Remote host closed the connection)
04:03:38 Join [Saint] [0] (~saint@rockbox/user/saint)
04:25:05 Quit bzed (Remote host closed the connection)
04:25:05 Join bzed [0] (
04:33:22 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:33:22 Quit amiconn (Disconnected by services)
04:33:22 Quit pixelma (Disconnected by services)
04:33:22 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:33:23 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:33:27 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:49:42 Join dell [0] (
04:49:44 Quit dell (Client Quit)
04:49:57 Join dell [0] (
05:04:43***Saving seen data "./dancer.seen"
05:11:47 Join Strife89 [0] (~Strife89@
05:23:45 Join JdGord [0] (
05:28:17 Quit JdGord (Ping timeout: 264 seconds)
05:33:41 Quit dell (Quit: Ex-Chat)
05:34:24 Join skofo [0] (
05:39:13skofoI am having a hard time using a bitmap in my plugin
05:41:19 Quit Strife89 (Quit: Vamoose!)
05:41:43 Join Strife89 [0] (
05:44:40skofoAh, found the issue.
05:46:48 Join belak [0] (~belak@facebook/engineering/belak)
05:48:24 Join amayer [0] (~amayer@
05:49:19 Quit skofo (Quit: leaving)
05:51:43 Quit [7] (Disconnected by services)
05:51:52 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:15:55 Quit shamus (Read error: Connection reset by peer)
06:16:49 Join shamus [0] (
06:24:30amayeri get an error when trying to compile a simulator. i ran tools/configure. then make && make fullinstall. the error i get is tools/bmp2rb: not found. it happens right after the line that says BMP2RB rockboxlogo.320x240x16.bmp
06:24:48amayerdo i need to compile bmp2rb first?
06:29:44[Saint]As you said, "BMP2RB rockboxlogo.320x240x16.bmp" means that already happened.
06:30:12[Saint]You're not running make with -j *?
06:30:36amayerno i didnt use -j
06:31:04[Saint]which sim?
06:31:07amayerand the BMP2RB comes first then the error that says tools/bmp2rb: not found
06:31:11amayerIpod Video
06:32:25[Saint]just trying now, won't be long.
06:33:36amayerim running Ubuntu 13.04 ive compiled tons of times on 12.04 but the first time i try on 13.04 i get errors
06:34:35[Saint]13.10 here, no error.
06:34:42[Saint]"make veryclean", try again.
06:37:05amayeri believe its failing on "make" not "make fullinstall"
06:37:17amayerill try again
06:38:10[Saint]yes, it will be. try make veryclean, then re-run make.
06:44:42amayer[Saint], still generating dependencys... im not ignoring you
06:51:54amayer[Saint], it seems to be working now
06:52:00amayerwhat does make veryclean do?
06:53:26[Saint]trashes everything in the build dir except for the files generated by configure, iirc.
06:53:48[Saint]Its likely you had leftovers from a failed build, or a build from another target device.
06:54:59amayerhmm... i ran "rm -R ./*" before i did ../tools/configure
06:57:05amayerill say
06:57:12amayerwell thank you for your help :)
06:57:33Strife89[Saint], I had some difficulties with building the Android .apk earlier that I believe you've helped with before.
06:58:05Strife89If you have time later, I'd like to try to fix it
06:58:51[Saint]I should edit that post slightly so it isn't necessary to download API16
06:59:11Strife89I'll follow that in the morning, then.
06:59:22[Saint]just install the SDk as-is, and then symlink adnroid-16 to android-17 in the platforms dir
07:00:10[Saint]alternatively, download android-16, but it seems odd to do so if RaaA will build anyway and the SDK comes with android-17.
07:02:11[Saint]the long and the short of it is 2 files we rely changed their location, and one is completely missing. :)
07:04:47***Saving seen data "./dancer.seen"
07:19:41 Quit amayer (Quit: Leaving)
07:48:59 Join shai [0] (
07:49:30 Quit shai (Read error: Connection reset by peer)
07:57:59 Join ZincAlloy [0] (
07:59:58 Join [Saint_] [0] (~saint@rockbox/user/saint)
08:01:37 Quit [Saint] (Disconnected by services)
08:01:43 Nick [Saint_] is now known as [Saint] (~saint@rockbox/user/saint)
08:12:08 Join ender` [0] (
08:13:01 Quit Scall (Quit: Bye bye)
08:15:21 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
08:32:33copper[Saint]: volume step and EQ step are the same concern
08:33:20[Saint]No, they're not.
08:34:28[Saint]When I looked initially, I messed up, I thought I had a part in the volume mess - I didn't.
08:34:43[Saint]Whatever the EQ step is, the volume would still be broken at this point.
08:35:48 Quit bluebrother (Disconnected by services)
08:35:53 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
08:36:56[Saint]copper: you even responded to an email in the ML that states the two are unrelated.
08:36:59[Saint]Did you read it?
08:37:04copper[Saint]: they're the same concern, usability-wise
08:37:49copperWhy do you think they're not?
08:38:10[Saint]Its not the same concern at all. You seem to think it is deliberate. The volume resolution could be 0.0000000001dB, and it would be entirely transparent to the user.
08:38:28[Saint]The fact it isn't is an error. Not deliberate. Nor is the EQ concerned here.
08:38:51copper*I* am concerned with the EQ steps as well.
08:39:00 Quit fs-bluebot (Ping timeout: 248 seconds)
08:39:09[Saint]That doesn't make the two related.
08:39:18copperI'm not saying they are
08:39:29[Saint]As stated in the ML, bring it up in an appropriate thread.
08:39:30copperbut it's the same usability problem
08:39:51[Saint]I doubt you will, though, as you threw your toys in that thread.
08:40:00coppercan you blame me?
08:40:05[Saint]I'm willing to be surprised, though.
08:43:59 Join einhirn [0] (
08:44:07*GodEater tries to work out how to change his password on the forums
08:44:24[Saint]Worth noting is that if you updated your build, you wouldn't see any issue.
08:44:53[Saint]...well, besides the EQ, which you've stated you don't use any more than the config you already have there.
08:44:59GodEaterah - there we go
08:45:41[Saint]pamaury changed the volume resolution for the F+ back to what was expected, not really a fix, more a sidestep.
08:46:49[Saint]I would've prefered">g#519 be committed personaly, but I think pamaury just wanted people to stop thinking *he* broke this.
08:47:19*[Saint] whistles for fs-bluebot
08:47:30[Saint]...where is that lazy bastard? :P
08:47:47GodEaternot here ;)
08:47:58copper[Saint]: posted
08:49:30[Saint]I'm sure that it will be brought up in the ML, but the fact that the internal stepping == .1dB isn't the issue here.
08:49:39[Saint](re: volume)
08:49:50[Saint]That should be entirely transparent to the user.
08:49:55copperthe UI is the issue
08:50:17copperindeed, I don't care what you use "internally"
08:51:08[Saint]I'm sure this will be brought up in the ML as well, as it was in here, but the "magic config option" won't happen.
08:51:15[Saint]That's a very big "No No" here.
08:51:20copperwhat's that?
08:52:11[Saint]"<don't_want_to_needlessly_highlight> on IRC, proposed that such fine-grained values be accepted in the configuration file, but not reflected in the user interface. That sounds perfectly acceptable to me, and would probably satisfy the most "enthusiastic" users, without bugging everyone else"
08:52:32copperI'm not the one who would care about that
08:52:38[Saint]Yeah, I know it doesn't bother you, just saying.
08:52:40copperonly that crazy dude on the forum
08:52:44 Quit tertu (Ping timeout: 268 seconds)
08:53:13 Join fs-bluebot [0] (
08:53:38[Saint]It really surprises me that pamaury chose to revert the stepping on the F+ rather than push">g#519
08:53:40fs-bluebotGerrit review #519 at"> : Fix volume handling of steps in wps, list and radio by Amaury Pouly (changes/19/519/1)
08:54:34copper[Saint]: because that patch doesn't help the volume handling issue
08:55:09[Saint]It most certainly should.
08:55:25copperwould it make the volume keys deal with 1dB steps?
08:55:56copperit looks like it only makes %pv display the correct value, doesn't it?
08:56:30copperare you confused about what I'm bitching about?
09:00:14[Saint]Possibly. So you're not ranting about the hilariously inaccurate dB values?
09:01:00copperwhere did you get that idea?
09:01:34ZincAlloyhilariously inaccurate? that would explain why the volume steps appear to be much smaller at lower volumes :D
09:01:37coppernowhere do I mention that in either one of my ML posts
09:02:10*copper bangs his head against the wall
09:02:17[Saint]Yes, but you also mention things that aren't relevant...and you have mentioned the dB values being incorrect in the past.
09:03:17[Saint]I'm sure I'm not the only one that has soem difficulty trying to parse whatever the grievance de jour is.
09:03:38 Quit kevku (Ping timeout: 260 seconds)
09:03:44copperwhat did I mention that isn't relevant?
09:04:51***Saving seen data "./dancer.seen"
09:05:11coppermy ML post mentions volume keys and the graphical equalizer, explicitely
09:05:33copperand argues that 0.1dB steps are not only inaudible, they also adversely affect usability
09:05:43copperwhat is not clear in there?
09:06:18GodEaterpresumably they're only inaudible at the bottom end of the scale, since, ya know, db is logarithmic">
09:06:34GodEatercopper: chill out man - give him a chance to catch up
09:06:37copperGodEater: what?
09:07:01copperwhat do you mean by "bottom end of the scale"? Very low volumes?
09:07:03GodEateryou're ranting without giving [Saint] another opportunity to parse what you wrote
09:07:08GodEateroh that
09:07:11GodEateryes - low volumes
09:07:18[Saint].1 dB is like...what, ~3% total volume or so?
09:07:26[Saint]that _should_ be audible.
09:08:19GodEatercopper: you come across as a very angry man.
09:08:20copperI don't even know what a percentage means in that context
09:08:39copperGodEater: I'm annoyed
09:08:47ZincAlloy1dB should be barely audible. at any volume
09:08:52copperpeople have been stonewalling me on this for what, 2 days?
09:09:01copperand yeah, ABX logs please
09:09:07copper0.1dB audible? Come on.
09:09:28ZincAlloythe db scale is logarithmic, but so is human hearing
09:10:09copperI'll eat my hat and post a youtube video of it if anyone successfully ABXes a 0.1dB difference at either end of the scale
09:10:56[Saint]I worded that badly. .1dB is about a 2~3% difference, it _should_ be audible...
09:11:09copperhow do you figure that number?
09:11:18copper3% of what
09:13:00[Saint]in apparent volume between it and the next or previous step. But how audible it is probably depends on how far down, or up, you are on the scale.
09:13:39[Saint]at very low volumes, it almost certainly wouldn't make a difference. Higher volumes...not so sure.
09:15:27copperI don't think percentages in this context mean anything
09:15:30copperwhat's 100%?
09:15:46copper3dB? 6dB?
09:15:48ZincAlloy+10db is twice as loud
09:16:09copper6dB is
09:16:26copperin the context of digital audio, it is
09:16:40ZincAlloyyou're not listening to digital audio
09:17:05copper0.5 FS is -6.02dB
09:17:09copper1 FS is 0dB
09:17:23copper0.25 FS is -12.04dB
09:18:25ZincAlloythat's not important to the user. he wants to know the difference in dB SPL
09:19:06[Saint]ZincAlloy: while you're here - how friendly are you with Inkscape/SVG?
09:19:14copperin any case, let's drop the percentages please
09:19:20ZincAlloyI'm using it for some basic stuff
09:19:36[Saint]You ever fely like making some high-res cabbie masters? :)
09:19:55ZincAlloyWe do have rather high res cabbie masters
09:20:15[Saint]not in scalable vector, though. :)
09:20:16ZincAlloypsd, though
09:20:35ZincAlloyright. but do we need that?
09:21:14[Saint]I seem to recall the .psd files I managed to dig up being incomplete. Could be wrong.
09:21:35ZincAlloywhat was missing? touch screen icons?
09:22:02[Saint]yeah, the quickmenu stuff iirc.
09:22:49ZincAlloygotta put up a new one with the scan icon anyway
09:23:34*[Saint] has re-read, and re-read, and now feels as though he has a complete grasp on copper's concerns.
09:24:04[Saint]the apparent linking of EQ and volume stepping confuzzled the fuck out of me.
09:25:02copperyou started it :P
09:27:20 Quit Strife89 (Quit: This computer has gone to sleep)
09:28:01copperZincAlloy: +6dB is twice the voltage, +10dB is 3 times, and +12dB is 4 times
09:28:04ZincAlloy[Saint]: The current graphics package contains all the touch screen icons, but not the fm icons
09:28:39ZincAlloyyes, but +10dB is still percieved as twice as loud
09:29:03[Saint]Hmmmm. I was sure there was something missing. Sorry.
09:29:17[Saint]Perhaps whatever I was loking for was buried under some layer.
09:29:44ZincAlloywhat were you looking for?
09:31:01[Saint]I can't remember exactly now, but, it was on my ToDo list - so I need to assume it was /something/">
09:31:25[Saint]"ToDo: Talk to ZincAlloy about missing cabbie elements"
09:31:35ZincAlloyI'll put up a new one later
09:31:45[Saint]I assume I wrote that (over 18 months ago (:P)) for a reason...
09:32:21ZincAlloyand I should make png to use with other applications as well
09:34:49copperthat page, if I understand correctly, seems to suggest that a 6dB change in volume (double the voltage) is perceived as 10dB increase, which is also "double the perceived volume"
09:34:55copperit's confusing as hell
09:35:31ZincAlloyyeah. translating units can be quite a pain
09:35:42 Join lorenzo92 [0] (
09:36:34copperZincAlloy: but that means that +6dB in Rockbox is indeed twice the perceived loudness
09:36:52copper+6dB in Rockbox = ratio of 2 = twice the voltage
09:36:56ZincAlloylet me check..
09:39:45copperthat page is very confusing
09:40:02ZincAlloyit's not just that page
09:40:13ZincAlloyit's the whole matter
09:41:57ZincAlloyah ha! "Actually, the just- noticeable-difference [in audio level] varies from0.1 dB to about 5 dB, depending on bandwidth, frequency, programmaterial, and the individual."
09:42:20ZincAlloywhile at the same time "One dB is the smallestchange in level that most people can hear−−the just-noticeabledifference."
09:42:22[Saint]woo! I was partially right about something!
09:43:18ZincAlloy"A 6 to10 dB increase in level is considered by most listeners to be"twice as loud."
09:43:20copperabout what?
09:43:39ZincAlloyabout noticeable differences in level
09:43:57copperI meant [Saint]
09:44:01[Saint]about .1dB potentiality being audible
09:44:17copperZincAlloy's quote says 1dB
09:44:25ZincAlloyread again
09:44:35[Saint]"0.1 dB to about 5 dB"
09:45:01copperthe two quotes contradict each other
09:45:01[Saint]massive range, but, I'll take it - it agrees with me :)
09:45:14[Saint]No, they don't.
09:45:26[Saint] /most/ people.
09:45:30[Saint]not *all* people.
09:46:28ZincAlloyand the 1dB thing is a general statement. .1 to 5 depends on a lot of factors. On average you can expect 1dB steps to be barely audible
09:47:16[Saint]Yeah, this seems to depend on how high or low the volume is at the time it it increased or decreased by .1dB
09:47:36[Saint]*it is
09:49:33ZincAlloynot sure how it relates to sound pressure. the text doesn't mention a corelation
09:50:16ZincAlloythough the human hearing is the most linear at around 80dB SPL
09:53:13 Join kevku [0] (~kevku@2a01:d0:ffff:34a::8:3)
09:54:40ZincAlloywhat is it that rockbox is displaying? dBu? dBv?
10:00:06 Quit akaWolf (Ping timeout: 245 seconds)
10:00:07ZincAlloyah, right. of course. it's digital
10:00:13coppernot even
10:00:22copperit's analog on targets that support it
10:07:13lorenzo92kugel: ping
10:07:49copperhere are three 30 seconds samples of "Get Lucky" by Daft Punk:
10:08:02copper0.1dB quieter:
10:08:07copper0.5dB quieter:
10:08:40[Saint]Well holy shit...quassel is frickin' awesome.
10:08:48[Saint]hover-link plays those files :)
10:09:37[Saint]gah...the source can't keep up.
10:11:47ZincAlloynot quite noticeable
10:12:11copperI have trouble ABXing the 0.5 dB difference, but I think it can be done
10:13:01copperdoes it make a useful difference though? Not really.
10:13:08coppernever mind the 0.1dB difference
10:13:21ZincAlloytoo subtle a step to be useful for a mp3 payer volume control
10:14:00ZincAlloysome pro audio EQs have stepped pots - with 0.5 or 1db steps
10:14:26[Saint]volume step? yes. EQ - still unsure.
10:15:17ZincAlloyin some frequency bands changes are very easily noticeable
10:15:30[Saint]Yes, that's my point.
10:16:09lorenzo92kugel: or anyone else, can you give me some hint to copy the entire toolchain from a system to another? i.e. i compiled it in the virtual machine, but it would be better to run it on the host OS
10:16:09copperI'd be OK with 0.5dB steps in the EQ and 1dB steps for volume
10:16:23ZincAlloyI think the EQ feels good as is
10:16:34copperZincAlloy: with 0.1dB steps?
10:16:39[Saint]define "as is"
10:16:45copperZincAlloy: it's currently 0.1dB
10:16:51[Saint]because "as is" == .1dB
10:16:54ZincAlloythat would be too small
10:17:23kugellorenzo92: pong
10:17:36kugellorenzo92: actually, you asked me about the toolchain recently? it's still on my ftp
10:17:44ZincAlloy1dB steps are subtle enough imho
10:17:45[Saint]realistically, the EQ should present the stepping the hardware accepts.
10:17:55lorenzo92kugel: ah! great :)
10:17:59[Saint]since it doesn't, the only way to not "lose out" is .1dB stepping
10:18:14copperwhat if the hardware presented 0.001dB steps
10:18:20copperwould you want to reflect that in the UI?
10:18:31copperso as to "not lose out"?
10:18:41[Saint]seeing as the lowest I have ever seen is .2, I think we needn't worry there.
10:18:57coppermy point is that usability and audibility trump hardware support
10:19:00 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:19:20[Saint]This only affects usability because the "fast EQ stepping" doesn't work.
10:19:26[Saint]It should. But, doesn't.
10:19:40copperwhat is that and how would I use it if it did work?
10:20:13[Saint]if you held down the button, instead of individual .1dB steps, it would do, say, 1dB steps.
10:20:26[Saint]much like accellerated scrolling in lists.
10:20:30copperthat's a really good way to overshoot
10:20:31[Saint]eeek. spelling.
10:21:01ZincAlloyI'd rather do it the other way round. 1dB steps unless a button is held down
10:21:13ZincAlloyeasier to use for the casual user
10:21:23[Saint]it would, in theory, be possible to do that, yes.
10:22:37copper[Saint]: btw, what does hardware support have to do with the EQ at all?
10:22:51copperisn't the EQ totally in software?
10:23:45copperyou're not telling the hardware to lower a frequency band by the smallest amount that the hardware supports
10:23:50[Saint]Not in all cases, I don't think. Needs confirmation.
10:24:55copperinternally, the hardware supports all volume variations, since it is converting a digital signal with 16 bit precision into an analog signal
10:25:06copperhardware volume steps have nothing to do with that
10:25:29copperthe hardware does not care what Rockbox does to the signal, it's not involved
10:26:21coppereven if the hardware supported only 6dB volume steps, Rockbox could still alter some frequency bands by 0.001dB, and the hardware would faithfully convert that signal into analog
10:26:42 Join froggymana [0] (~froggyman@unaffiliated/froggyman)
10:27:39copperso unless you're using a hardware EQ built into the DAC chip, which matches the Rockbox EQ configuration EXACTLY (with 10 bands), then the Rockbox EQ steps are completely unrelated to the hardware
10:28:08 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
10:28:22copperif the hardware EQ is 5 bands, then the Rockbox EQ is still software
10:28:26 Join JdGordon [0] (
10:28:27 Quit JdGordon (Changing host)
10:28:27 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
10:28:50copperI would be very surprised if there was ANY Rockbox port in which the EQ is not 100% software
10:28:51 Join akaWolf [0] (~akaWolf@
10:28:51 Quit akaWolf (Changing host)
10:28:51 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
10:29:09marazcopper: what did you ABX on?
10:29:11copperso there's nothing to "lose out"
10:29:16coppermaraz: foobar2000
10:29:25ZincAlloyif rockbox displays dBFS values - why can there be positive values?
10:29:29[Saint]Hum. That's a good point. There's not really any point in my discussing this. I've lost count of the times I've stated I don't care a fuck about the stepping of the EQ by now. :)
10:29:40copperZincAlloy: 2FS is a valid value
10:29:52 Quit froggyman (Ping timeout: 240 seconds)
10:29:52 Quit x56 (Ping timeout: 240 seconds)
10:29:53 Quit tchan (Ping timeout: 240 seconds)
10:29:53 Quit bluebrother^ (Ping timeout: 240 seconds)
10:29:57 Quit JdGordon_ (Ping timeout: 240 seconds)
10:29:59 Nick froggymana is now known as froggyman (~froggyman@unaffiliated/froggyman)
10:29:59ZincAlloyI thought 0 was max for FS?
10:30:06copperit's not "max"
10:30:09copperit's the reference
10:30:22[Saint]0 == reference
10:30:24marazcopper: headphones or speakers?
10:30:38copper+6dB (2FS) would badly clip the signal if it peaks at 0dBFS
10:30:42 Join x56 [0] (
10:30:57copperbut it will produce a perfectly valid signal with content that never goes past 0.5FS
10:31:02ZincAlloy"0 dBFS is assigned to the maximum possible digital level." no?
10:31:14coppermaraz: IEMs
10:31:25marazah, that explains it.
10:31:31copperZincAlloy: it's just a mathematical representation
10:31:32coppermaraz: ?
10:31:53marazcopper: I tried on my desktop multimedia speakers and don't find it too hard to abx 0.0 and -0.5
10:32:03copperZincAlloy: 0.5FS = -6dB, 1FS = 0dB, 2FS = +6dB, etc…
10:32:32[Saint]maraz: wow, what type of speakers are those? I would assume that would make it a lot harder.
10:32:35coppermaraz: I'm sorry but I have to ask, do you know what ABX means?
10:32:40 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
10:32:46marazcopper: yes, I used the fb2k ABX comparator
10:32:52marazprobability that you were guessing: 3.5%
10:32:58coppernot low enough
10:33:05coppernever mind
10:33:09marazstill, 7/8 :)
10:33:12marazI think I failed the first one
10:33:36[Saint]7/8 is "verified".
10:33:40maraz[Saint]: something I rigged up from a pair I had laying around.
10:33:41[Saint]now...repeat it!
10:33:50marazshould I?
10:34:05[Saint]Nah. I don't care in the slightest ;)
10:34:06coppermaraz: my IEMs didn't prevent me from ABXing
10:34:15copperjust my sensitivity and concentration at the time
10:34:34[Saint]My assumption would be IEMs make this process a lot easier.
10:34:48copperI got like 4/6, and I stopped because I thought it could probably be ABXable
10:35:08copperso I'm not going to claim that a 0.5dB difference can't be ABXed
10:35:58copperbut I still maintain that it's not useful, as far as volume control is concerned
10:36:09marazyeah, probably not.
10:36:25kugelfwiw, I think 0.5 db would be a goog compromise
10:36:34marazhowever, it's easy to tell once you spot something that gives it away, like snare decay
10:36:49copperkugel: it would still make volume control twice slower
10:36:51lorenzo92kugel: so my idea is to patch R0's cramfs on the fly, by means of our patched cramfs that ONLY contains files to be put in the rom
10:37:12kugelcopper: we're talking about the eq which is currently 0.1 isnt it?
10:37:16lorenzo92kugel: i.e. without extracting the original cramfs, process that's painful in windows (cygwin etc)
10:37:23copperkugel: both the EQ and volume control
10:38:13copperthough the volume control problem was only on the Fuze+ apparently, and pamaury reverted it recently
10:38:14lorenzo92bluebrother: i was talkin with kugel about YP-R0 and rbutil. Firmware packing and unpacking tools are ready but I'm still missing how to patch cramfs on windows...
10:38:46[Saint]I'd be cool with q == .1dB and EQ == .5dB, I'm given to understand there may be genuine reasons why someone might want that fine control there.
10:38:53kugelnormal volume control is, and has ever been, in 1db steps, no?
10:39:04[Saint]Not anymore, apparently.
10:39:11[Saint]see dev-ML
10:39:17*kugel must have missed something
10:40:51kugelhe said the internal representation changed, but what about the UI?
10:41:06fs-bluebotGerrit review #519 at"> : Fix volume handling of steps in wps, list and radio by Amaury Pouly (changes/19/519/1)
10:42:07kugelhow's this specific to the fuze+?
10:42:35copperkugel: because only pamaury started adapting the change to volume control
10:43:01copperbtw, the semi consensus that we got here is 1dB for volume and 0.5dB for EQ… which is precisely the way it was before the change!
10:43:19[Saint]and then very recently reverted it, due to the shit-storm it created, and fear of people thinking he caused it.
10:43:23kugelcopper: sounds good to me as well
10:44:15lorenzo92copper: indeed, I don't find any reason to complicate things up
10:44:25ZincAlloyI'd have no problem with 0.5dB for EQ, but I think 1dB will do just fine
10:44:31*copper sighs in relief
10:44:53 Quit akaWolf (Ping timeout: 246 seconds)
10:45:08copperanother argument is that users might have to deal with much larger ranges in volume (when changing headphones, or when using it as a line out) than they would in the EQ
10:45:30copperso 0.5dB steps are much less problematic, usability-wise, in the EQ, than they are in volume control
10:46:08coppere.g a user could use Rockbox at -32dB with their highly sensitive IEMs, and then set it to 0dB when plugging into a hi-fi or a car stereo
10:46:29[Saint]as long as if/when EQ goes back to .5dB, q stays at .1dB.
10:46:31copperthat would be 320 steps of 0.1dB, 64 steps at 0.5dB
10:46:38ZincAlloydb display in the eq is weird
10:46:38copper[Saint]: what's q?
10:46:49[Saint]I'm willing to be proven wrong, but I'm given to believe there's a genuine use for the latter.
10:46:54ZincAlloy0dB and −0db?
10:48:44copperthat's mandarin to me
10:49:26[Saint]higher q, the less it bleeds into the ranges beside it, and vice versa.
10:49:41[Saint]high q, sharp peak, low q, wide peak.
10:49:51ZincAlloysimilar to bandwidth
10:50:33copperthat's the thingy that I don't use
10:50:55[Saint]virtually no one does, as you need to be a rocket scientist to do it on-the-fly
10:51:18[Saint]a nice graph to plot the EQ has been discussed a few times.
10:51:47ZincAlloyI was wrong about the eq on my fuze being in 1dB steps
10:51:56ZincAlloyclip zip, not fuse, sorry
10:52:04*pamaury doesn't want to be killed by the crowd but still think there is a massive misunderstanding about what happened, nevermind. Could I also remind people that nightly builds are not subject to same quality requirements as releases ?
10:52:07ZincAlloyit's 0.5, but only displays full dBs
10:52:08JdGordoncopper: if you frequently do that sort of range and find the stepping annoying, use the shortcut menu and 2 configs
10:52:45lorenzo92pamaury: absolutely !
10:52:46copperJdGordon: wrt volume control?
10:52:56[Saint]pamaury: by all means, if you want to speak, speak - a precise breakdown of what happened, and why, is welcomed.
10:53:32[Saint]both you and I have taken blame for this (though, I brought it on myself assuming I was responsible), so, go nuts. :)
10:53:56JdGordoncopper: yes
10:54:37copperJdGordon: I'll look into that, thanks. But my point about usability, audibility and usefulness remains
10:55:25pamauryI already explained too many times: the volume never was, never will be, never fucking is in 0.1dB steps, this is a UI bug. I'm pretty sure it's the same for EQ. People keep arguing about 0.5dB vs 0.1dB vs 1dB but the truth is that the change probably never intended to change it.
10:58:00copperpamaury: where is the UI step defined?
10:58:28copperis it defined at all, with regards to "internal" 0.1dB steps?
10:58:34copperis there a provision for that?
10:58:57coppersome code that says somewhere, the UI should always deal with 1dB steps, no matter the internal base value
11:00:43copperJdGordon: can I load a .cfg file with a "file" type shortcut?
11:00:56coppertype: file
11:01:06copperdata: /.rockbox/some.cfg
11:01:42pamaurythat's the point: the code just increases the volume, and don't care about its unit. The unit depends on the target, it might 1dB or 0.1dB. And it completely ignores hardware steps.
11:04:55***Saving seen data "./dancer.seen"
11:09:01copperJdGordon: wow, that makes my Fuze+ crash, hard
11:09:11copper"data abort at [etc]"
11:10:58copperit also makes the sim crash
11:11:18copperSegmentation fault (core dumped)
11:13:52 Join ukleinek [0] (~ukl@unaffiliated/ukleinek)
11:40:45kugellorenzo92: how would that work?
11:41:03kugelcramfs is compressed isnt it?
11:41:56lorenzo92exactly. I tought about modifiying the mkcramfs executable to read the original cramfs and at the same time another small cramfs that ONLY contains needed modifications (files)
11:42:15lorenzo92kugel: and then producing a new cramfs with both mixed in
11:42:45lorenzo92in this way we completely avoid the problem of attributes, filenames while extracting
11:43:47 Quit [Saint] (Quit: something is a sick puppy here...)
11:52:15 Join [Saint] [0] (~saint@rockbox/user/saint)
11:53:37 Join krabador [0] (~krabador_@unaffiliated/krabador)
11:57:56copperwhen rockbox saves the configuration to config.cfg, it ommits default values
11:58:20copperso when you load another config file with non default values, and then revert to the first config file, non-default values from the second config file remain
11:58:54amiconnThis is on purpose. If you save a config explicitly, it will contain all values
12:00:44lorenzo92very trivial patch here">g#321, can someone check it?
12:00:46fs-bluebotGerrit review #321 at"> : Cube plugin fix by Lorenzo Miori (changes/21/321/2)
12:01:39lorenzo92well wait now that i see there is also the atextit() function!! what the hell?
12:02:52kugellorenzo92: the patch isnt trivial at all :p exit() is supposed to work for plugins
12:03:01kugelit works in the sim doesnt it?
12:03:29kugelexit() is provided by plugin_crt0.c, thus plugins should *not* be able call the libc version
12:03:31lorenzo92kugel: ah! okay then there is another problem :) the point is that I did not find any other plugin using the exit(), so I assumed there was a problem
12:03:58Tornefrotz uses exit(), no?
12:04:01kugelmost call exit_on_usb() which is a small wrapper
12:04:21Tornepretty sure i added the proper exit support for that :)
12:04:30Torneor fixed it. or sometbing.
12:04:35[Saint]Torne: frotz probably isn't very high on the list of commonly tested plugins :)
12:05:07lorenzo92let me check
12:06:03lorenzo92interesting, frotzt is even *not* compiled on R0 lol
12:07:52 Nick uwe__ is now known as uwe_ (
12:10:15copperI can't use this thing
12:10:46[Saint]"this thing"?
12:11:30lorenzo92kugel: there there must be a problem with longjmp
12:22:43lorenzo92kugel: actually, commenting out the exit function does the same (either crash or real exit) hum
12:22:58lorenzo92in plugin_crt0 of course
12:33:47TheSevenhow important is ipod classic dualboot to us?
12:34:06TheSevenanyone willing to actually take care of getting that fixed?
12:36:08kugelTorne: I think I did it :)
12:37:04kugellorenzo92: well, plugin_crt0's exit has to be called
12:37:20Torneah, i added atexit() support :)
12:37:38Tornenot-quite-atexit support
12:37:46Tornebut you fixed a bunch of things :)
12:38:06lorenzo92kugel: then I don't understand....commenting out longjmp still crashes/closes
12:46:49TheSevenso I guess not very important...
12:50:27[Saint]willing != able
12:51:09[Saint]those that want it don't have the time or technical ability, evidently.
12:51:48[Saint]its no blocker for the ports sake, ...qoukd it be handy? yep.
12:52:10[Saint]could I do it? ability - maybe, time - nope.
12:52:23TheSevenroughly same here
12:54:29copperTheSeven: isn't Rockbox Util support the top priority?
12:54:50copperand a stripped down bootloader
12:58:12[Saint]I wasn't sure anyone mentioned priority.
13:00:05 Join akaWolf [0] (~akaWolf@
13:00:05 Quit akaWolf (Changing host)
13:00:05 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
13:00:50 Join mortalis [0] (~mortalis@
13:03:47*copper finally managed to make 4 proper config files for his Classic and Fuze+
13:04:59***Saving seen data "./dancer.seen"
13:05:48 Join liar [0] (
13:07:31 Join olspookishmagus [0] (~pookie@
13:27:04 Quit belak (Quit: belak)
13:30:09TheSevencopper: well, what's the way to go there?
13:30:17TheSevenwhat does it take?
13:30:41TheSevencan we teach the rockbox utility to talk to DFU devices on windows?
13:30:44TheSevencan it install drivers?
13:31:11copperwhat do I know
13:31:15TheSevenshould we go for a completely different approach?
13:31:25TheSevenget the notes exploit to work on the classic and install through that?
13:31:56copperare you actually asking ME?
13:32:01[Saint]that would be desirable.
13:32:07TheSevenI'm asking the channel...
13:32:18[Saint]re: noteboot
13:32:32TheSevenok, what are the dependencies of that?
13:32:35[Saint]it'd be a lot easier for the user at least.
13:33:02*[Saint] prods bluebrother
13:33:21TheSeven1. OF version dependent return address
13:33:21TheSeven2. we need to be able to read files from the HDD in ~3 kilobytes of code
13:33:32TheSeven3. we need to figure out that return address
13:33:51copperOF hasn't been updated in years
13:33:52TheSeven4. the firmware must be vulnerable to it in the first place, and people might need to downgrade
13:34:14TheSevenIIRC the latest firmware has already patched that vulnerability
13:37:46copperwhy downgrade? emCORE works with the latest OF
13:53:11ZincAlloy[Saint]: Graphic package updated
13:56:12 Quit kevku (Ping timeout: 245 seconds)
13:58:03 Quit liar (Read error: Connection reset by peer)
14:23:47 Join amayer [0] (
14:50:17TorneThis forum thread is weird:;topicseen#new">;topicseen#new <- bots i'm guessing?
14:53:19 Quit lorenzo92 (Quit: ChatZilla [Firefox 22.0/20130712032019])
14:54:21gevaertsHmm, indeed
14:55:44*gevaerts decides to delete the lot
14:56:13gevaertsNot sure if banning makes sense
14:57:31gevaertsAll of those accounts have been used exactly once
14:57:37*gevaerts decides not to bother
15:04:49TheSevencopper: currently emcore doesn't work with any OF version, but if we want to install through the notes exploit we need a vulnerable firmware, and the latest one is fixed
15:05:01***Saving seen data "./dancer.seen"
15:06:30TheSevenso we have 3 possible "attack vectors" to install our firmware:
15:06:51 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee)
15:07:43TheSeven1. pwnage 2.0 exploit in DFU (which is what emcore currently does)
15:07:43TheSeven2. pwnage 2.0 exploit in on-disk firmware (requires an old bootloader version, which IIUC only works on 1g models)
15:07:43TheSeven3. notes exploit within the main firmware (requires a specific exploit for each firmware version, the latest firmware is fixed so it doesn't work)
15:10:08TheSevenDFU is certainly the most robust, surefire way of booting this thing (apple can't possibly patch it), but it has two problems:
15:10:08TheSeven1. it requires the user to press buttons for precise amounts of time (many users fail to do that right, even though it is fairly easy, if you just follow the instructions literally)
15:10:08TheSeven2. on windows it requires a special driver for the DFU mode to be installed, which will interfere with itunes. it can also piggyback on itunes' driver, but then that has to be installed and itunes will interfere with the DFU boot process if you don't kill each and every apple process at the right time
15:10:51TheSeventhe pwnage in on-disk firmware exploit is not practical IMO because it will only work on 80gb and 160gb thick models which have never been updated
15:11:02TheSeventhe notes exploit could be an alternative
15:19:02 Join pretty_function [0] (~sigBART@
15:20:39[Saint]the only way they could really attempt to stop noteboot is if they released a new firmware which disallowed flashing an older version, and people updated.
15:21:05 Quit mortalis (Quit: Leaving)
15:21:27[Saint]its pretty safe to say they've stopped with updates to the current Classic line, though.
15:22:33 Join mortalis [0] (~mortalis@
15:30:32 Quit pretty_function (Remote host closed the connection)
15:44:35 Quit krabador (Quit: Sto andando via)
16:04:49 Quit mortalis (Quit: Leaving)
16:05:08TheSevenso you'd let the rockbox utility downgrade the firmware and install through noteboot?
16:12:07 Quit shamus (Read error: Connection reset by peer)
16:12:45 Join shamus [0] (
16:20:27[Saint]presumably. but that seems messy.
16:23:57 Quit [Saint] (Remote host closed the connection)
16:24:27 Join tertu [0] (
16:24:54 Join [Saint] [0] (~saint@rockbox/user/saint)
16:36:52TheSevenit does indeed.
16:37:11TheSeventhe question is if DFU is more or less messy. or if there's another route that I'm missing.
16:48:26 Join Scall [0] (~chat@unaffiliated/scall)
16:49:40[Saint]there's no route I'm aware of that you didn't cover.
16:51:25[Saint]the DFU method is super ugly on Windows, and oh so painless with Linux...that bugs me.
16:52:20[Saint]well, the "+iTunes" method is.
17:04:18 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
17:04:22 Join Strife89 [0] (~Strife89@2602:306:250e:c3c9:20a:95ff:fef3:ec5f)
17:05:05***Saving seen data "./dancer.seen"
17:11:00 Quit melmothX (Quit: ciao)
17:19:56 Quit GeekShadow (Read error: Connection reset by peer)
17:20:16 Join lebellium [0] (
17:20:22 Join GeekShadow [0] (~antoine@reactos/tester/GeekShadow)
17:22:18 Quit Strife89 (Quit: Vamoose!)
17:22:22 Join Strife1989 [0] (~Strife89@2602:306:250e:c3c9:f592:5f13:5b75:3336)
17:23:22 Join guymann [0] (~c@unaffiliated/guymann)
17:30:58 Quit benedikt93 (Quit: Bye ;))
17:49:27 Quit olspookishmagus (Read error: Operation timed out)
17:50:31TheSeven[Saint]: the non-itunes method is outright broken afaict
17:51:28 Join olspookishmagus [0] (~pookie@
17:51:51 Nick olspookishmagus is now known as Guest15077 (~pookie@
17:59:11copperTheSeven: I've never been able to put the iPod in DFU mode on Windows
17:59:20copperwith and without iTunes and other apple stuff installed
17:59:31copperit always fails on my install
18:05:07pixelmalebellium: if you would build the bootloader yourself from source you should get a dualboot bootloader
18:05:20pixelmafor the Iaudio X5 that is
18:09:19lebelliumpixelma: you mean it won't work with RButility?
18:10:58saratogai'm ok with just having rbutil tell the user to put the device in DFU mode
18:11:08pixelmait looks like no updated bootloaders were released back then - the available bootloaders on (which the Rockbox Utility uses) and in the wiki are from 2008 and the dualboot patch got committed in 2011
18:11:52 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:12:32lebelliumpixelma: ok sounds clear, thanks. I hope it's not too hard to build my own bootloader, I'll look at it
18:12:58pixelmaor give amiconn a nudge to test and release newer ones ;)
18:13:24lebelliumhe doesn't look active enough for that :)
18:14:01gevaertsThat's because he doesn't get enough nudges :)
18:14:12pixelmaif you already set up the environment then building the bootloader isn't hard, it's just an option in configure
18:14:14gevaertsAnyway, no, building a bootloader is not hard
18:14:34gevaertsIt's just the same as building a regular build, except you pick "b" in configure
18:15:05lebelliumok nice, so it should be easy then, I already have my environnement
18:15:05pixelmaand for the Iaudios testing and installing bootloaders is safe as the Cowon loader still will be there and takes care of the flashing
18:15:51pixelmaquite safe I mean, don't want to sound a 100% sure
18:16:16lebelliumthe dualboot patch in 2011 was only for X5 right?
18:17:01tertucopper apparently you have to make sure iTunesHelper isn't running for it to enter DFU mode properly
18:17:31coppertertu: I've tried everything
18:17:42copperI don't know, maybe my Win7 install is fucked up
18:17:54tertuthe iPod acts right, i assume
18:17:56copperno problem on my Linux install
18:18:07tertureboots and then goes black?
18:18:29copperI also can't get iTunes to restore the OF on Win7
18:18:37copperI always need to borrow someone's Mac to do it
18:18:52pixelmalebellium: it was at least for the X5 and the M5, I'm not sure if it also applied to the M3. I'm running one on my M5 which doesn't help me much because the OF freezes hard with the SSD I have (weirdly but luckily the Cowon loader deals with it just fine though)
18:19:25tertuthat's a bit bizarre... i've only had problems like that dealing with an HFS iPod
18:20:46saratogacan the installer just kill any itunes processes itself?
18:21:01lebelliumpixelma: how can I find this patch? (I won't use the OF much either, it's just that as a collector I need non-modded devices that work as they did originally. So I have to be able to show someone how was the OF back to 2005 :) )
18:21:58pixelmait's committed, no separate patch anymore. You only need a new enough checkout of the source
18:22:49coppertertu: actually after I restore the iPod on a Mac, I then connect it to my Win7 install, and THEN iTunes finally tells me it wants to format it for Windows
18:23:08tertuwhat kind of iPod are you using
18:23:13copperthough I think I'm done restoring the iPod
18:23:26copperI got Rockbox on the Classic to finally work well
18:23:29copperiPod Classic
18:23:35copperlast version
18:23:55lebelliumpixelma: found :);a=commit;h=91ce4b2a60c4cbe8e3568f79c3a73572461ff40d">;a=commit;h=91ce4b2a60c4cbe8e3568f79c3a73572461ff40d so apparently it's only for X5/M5, not M3 unfortunately.
18:24:55tertui have an 80GB classic and after I restored it on a Mac (the data partition got screwed up I think) basically everything worked from there first try
18:25:21pixelmalebellium: hmm, interesting. I didn't remember the define but I was better than amiconn himself who didn't remember that commit at all
18:27:07pixelmaso it looks like there is a bit more to it to get the dualboot bootloader but I'm not sure I understand the commit message completely
18:28:04lebelliumI'm not sure either :D I'll see that when charging is complete. A friend gifted me this X5, he hasn't used it for years so I don't know how's the battery now
18:30:21*pixelma thinks there's a high probability that the battery isn't in a good shape anymore, the Iaudio batteries seem to degenerate quite a bit with time
18:32:59lebelliumhe already replaced the battery once so that's not the original 2005 battery but that was in 2008 or something like that, already 5 years ago :S
18:36:28 Join belak [0] (~belak@facebook/engineering/belak)
18:37:24 Join krabador [0] (~krabador_@unaffiliated/krabador)
18:44:20 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
18:54:59 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
19:02:40*bluebrother wonders why lorenzo92 only shows up when he's at work :/
19:03:01bluebrotherlorenzo92 (logs): is there a patch around? How does this patching work on Linux?
19:05:09***Saving seen data "./dancer.seen"
19:15:34 Join dell [0] (
19:16:10 Quit tertu (Ping timeout: 240 seconds)
19:19:15 Quit amiconn (Read error: Operation timed out)
19:19:17 Quit TheSeven (Read error: Connection reset by peer)
19:19:42 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
19:19:45 Join amiconn [0] (amiconn@rockbox/developer/amiconn)
19:32:00 Quit dell (Quit: Ex-Chat)
19:33:05 Quit bzed (Remote host closed the connection)
19:33:13 Join bzed [0] (
19:35:33lebelliumI managed to build a normal X5 bootloader but I don't understand the instructions for dualboot here:;a=commit;h=91ce4b2a60c4cbe8e3568f79c3a73572461ff40d">;a=commit;h=91ce4b2a60c4cbe8e3568f79c3a73572461ff40d
19:38:45gevaertslebellium: first add that define somewhere (maybe in config/iaudiox5.h), then just build (i.e. run make)
19:38:55gevaertsThen find the OF's x5_fw.bin
19:39:55gevaertsAnd then, run ../../tools/mkboot -iax5 of_x5_fw.bin bootloader.bin output.bin
19:40:06gevaerts(assuming you named the OF's file of_x5_fw.bin)
19:40:35gevaertsThen output.bin should be the file you need. I don't know if that has to be named x5_fw.bin, but I wouldn't be surprised
19:40:53gevaertsJust make sure you don't get all the .bin files mixed up :)
19:47:02 Quit Gareth (Ping timeout: 245 seconds)
19:47:22gevaertsStrife1989: does /home/michael/android/sdk/build-tools/android-4.2.2 exist?
19:48:31gevaertsWhat does ls -l /home/michael/android/sdk/build-tools/android-4.2.2/aapt say?
19:49:03 Join lorenzo92 [0] (
19:49:55lorenzo92bluebrother: haha here I am ^^ well it's pretty simple: firmware gets unpacked, then the cramfs is extracted and patched (simply copying some files to the rootfs) and then the cramfs is again packed and so on
19:50:22Strife1989gevaerts: -rwxrwx−−- 1 michael michael 1122758 May 13 12:31 /home/michael/android/sdk/build-tools/android-4.2.2/aapt
19:50:27lebellium[19:38]gevaertsThen find the OF's x5_fw.bin > what do you mean here? I have to put the OF file in my build directory?
19:50:40gevaertslebellium: yes
19:50:58lorenzo92to say, in linux we can extract it without any problem, on windows, well...only cygwin quite solves the problem (quite, because I should modify the binary anyways to handle \\ characters)
19:51:12lorenzo92*in filenames
19:51:35 Quit bzed (Remote host closed the connection)
19:51:44lorenzo92bluebrother: so the solution I was thinking is to create a cramfs patcher, as described in previous messages
19:51:59 Join bzed [0] (
19:54:54lebelliumgevaerts: ../../tools/mkboot: No such file or directory
19:55:27gevaertslebellium: well, possibly ../tools/mkboot. Depends on your directory structure, i.e. where your build directory sits
19:55:42lebelliumI'm in rockbox/X5bootloader
19:56:03gevaertsJust a single .. then
19:56:33*gevaerts assumed a build/target structure, which is what he uses himself these days
19:57:20gevaertsStrife1989: does running "/home/michael/android/sdk/build-tools/android-4.2.2/aapt" on its own do anything?
19:57:57Strife1989"No such file or directory." :o
19:58:36gevaertsOK. This is going to be a library/arch issue
19:58:49gevaertsWhat does "file /home/michael/android/sdk/build-tools/android-4.2.2/aapt" say?
19:59:00 Join Gareth [0] (~gareth@2607:ff38:2:83::3)
19:59:05Strife1989gevaerts: aapt: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.8, not stripped
19:59:19gevaertsOK, do you have a 32 bit or a 64 bit system?
19:59:20Strife1989I'm on 64-bit Ubuntu 13.04
19:59:46lebelliumgevaerts: so the of_x5_fw.bin I put in the directory has been modified now and is the file I have to put on the X5 (after maybe renaming it to x5_fw.bin) ?
19:59:46gevaertsWhat does "ldd /home/michael/android/sdk/build-tools/android-4.2.2/aapt" say?
20:00:07Strife1989"Not a dynamic executable."
20:00:32gevaertslebellium: the file itself shouldn't be modified, unless you specified it as an output file too, IIUC
20:01:07lebelliumso I don't understand what ../tools/mkboot -iax5 of_x5_fw.bin bootloader.bin output.bin did?
20:01:47gevaertslebellium: do you now have a file named output.bin?
20:02:12lebelliumyes. That's the one I have to rename to x5_fw.bin and put onto the X5?
20:02:29gevaertsStrife1989: you seem to have a rather pure 64 bit system then :) Let me have a look...
20:02:29lebelliumok now it's clear
20:02:31lebelliumthanks :)
20:05:15gevaertsStrife1989: do you mind taking this to PM? getting multiarch to work is a bit off-topic here I'd say
20:05:29Strife1989gevaerts: That's fine.
20:06:13 Join einhirn [0] (
20:07:34 Quit FOAD (Quit: I'll be back)
20:07:49 Join FOAD [0] (~foad@unaffiliated/foad)
20:11:55 Join dell [0] (
20:22:52 Quit einhirn (Ping timeout: 264 seconds)
20:28:21 Quit lorenzo92 (Ping timeout: 256 seconds)
20:35:24 Quit dell (Write error: Broken pipe)
20:42:49 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
20:43:36 Join dell [0] (
20:56:02 Quit dell (Quit: Ex-Chat)
20:58:01 Join wodz [0] (
21:01:47 Quit Guest15077 (Quit: All for nothing)
21:04:24lebelliumgevaerts: hum firmware upgraded but it still boots the OF. It did not work
21:05:08gevaertslebellium: triple-check if all the .bin files are what you think they are
21:05:12***Saving seen data "./dancer.seen"
21:07:11lebelliumI have of_x5_fw.bin (OF 2.10 I downloaded on the cowon website), x5_fw.bin automatically created when building and output.bin. I used output.bin (and renamed it to x5_fw.bin once on device)
21:09:11 Join EstebanMonge [0] (~emonge@
21:13:13gevaertslebellium: what exact command did you run to make the new output.bin?
21:13:45lebellium../tools/mkboot -iax5 of_x5_fw.bin bootloader.bin output.bin
21:14:03*gevaerts nods
21:14:08 Join rockbox [0] (
21:14:28gevaertsI don't know then
21:16:01lebelliumnow it works
21:16:25lebelliumI just needed to plug it to the PC and unplug it after upgrading the firmware
21:16:33lebelliumI don't understand why but at least it works
21:17:00 Quit thegeek (Read error: Connection reset by peer)
21:19:29*gevaerts also tries
21:21:00gevaertsWorked the first time for me, but then I had the pure rockbox bootloader installed before. Maybe that makes a difference
21:21:49 Quit rockbox (Quit: AndroIRC - Android IRC Client ( ))
21:22:02lebelliumwell that's not a big deal. It's just strange than directly after installing the new firmware it still booted the OF as if it had the previous bootloder/firmware still in cache
21:22:32*gevaerts needs to take his x5 apart to clean it and try to make the joystick a bit more responsive
21:22:41gevaertsMaybe also new batteries...
21:24:34lebelliumunfortunately the battery is soldered
21:30:05gevaertsBut a dead battery is no better than a badly soldered new battery
21:31:10wodzgevaerts: what about debugging h300 bootloader through BDM?
21:31:25*gevaerts needs to get back to that...
21:32:03gevaertswodz: devcon is probably the best bet to make serious progress there
21:32:44gevaertsI'll be rather busy for the next month or so as well, so I'm unlikely to have much time before then anyway
21:33:03wodzoh, and by the way are we fixed with 14-15 Sep for DevCon?
21:33:14gevaertsYes, as far as I know
21:33:29gevaertsWe'll need to find hotels though
21:47:32 Join lorenzo92 [0] (
21:49:18 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
22:00:35 Quit pamaury (Quit: No Ping reply in 180 seconds.)
22:00:46 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
22:06:05wodzpamaury: ping
22:07:27 Quit Marex (Ping timeout: 248 seconds)
22:07:40 Join Marex [0] (~Marex@
22:10:33bluebrotherlorenzo92: how are you manipulating the cramfs? Using mkcramfs from the cramfs package at sf?
22:12:27lorenzo92bluebrother: yes, exactly
22:12:35lorenzo92the problem arises on windows
22:13:00bluebrotherhmm. That doesn't immediately compile on windows
22:13:09bluebrotherso it needs some porting I guess ...
22:15:08lorenzo92exactly...and moreover we cannot do any extraction (permissions...filenames...)
22:15:56bluebrotherwhat kind of attributes does cramfs handle?
22:16:12lorenzo92everything (owner, group, etc)
22:16:13bluebrothersame as Linux?
22:16:26lorenzo92it's a linux rootfs :)
22:16:36lorenzo92back in 10 mins
22:17:15micahis it normal for my ipod class to basically require a total boot up when I lock the device when it isn't playing?
22:18:14wodzmicah: it turns off after some time of inactivity, so basically yes
22:18:44wodzmicah: rockbox doesn't enter deep sleep as OF does
22:18:53 Quit melmothX (Quit: #)
22:19:37bluebrotherand we want to replace some files inside the cramfs image, right?
22:25:46micahwodz: ok, i guess that can be configured too
22:26:40wodzmicah: idle time is configurable AFAIK
22:29:59micahok cool
22:32:20[Saint]copper: that's....outright odd.
22:32:41lorenzo92bluebrother: yes, exactly
22:33:02[Saint]Are you sure you didn't confuse "not in DFU mode" with "actually in DFU mode, which is identical to "off" to the user without OF feedback"?
22:34:09copper"The display of your iPod should now stay black, and a new USB device called "Apple Recovery (DFU) USB Driver" should connect to your PC."
22:34:11[Saint]My understand is that "USB+hardware defined key-combo==can't possibly not enter DFU mode without hardware failure"
22:34:19lorenzo92bluebrother: so my idea was either to develop an "on-the-fly" patcher (i.e. I take content of one image an merge with another) or ...?
22:34:20copperthe display is black but Win7 doesn't register
22:34:28copper"driver installation" fails
22:34:39[Saint]so, yes, its in DFU mode...but your OS is being a dick.
22:34:39copperyeah sorry
22:34:44[Saint]that's quite different.
22:35:20[Saint]I just had to check that one, because actually not going into DFU mode would be remarkable.
22:36:27[Saint]Its so sad that the WIndows install is *so* painful, and the linux install is entirely painless, but users assume it isn't...'cos linux.
22:39:17bluebrotherlorenzo92: the format doesn't seem to be too complicated as far as I can tell right now
22:39:30bluebrotherhaven't found a detailed format description yet though
22:41:15bluebrotherlorenzo92: do you have a link to an example file we would be processing later?
22:41:36 Quit shamus (Ping timeout: 240 seconds)
22:44:00lorenzo92bluebrother: yeah the format isn't too cryptic ^^
22:45:22bluebrotherso I guess it shouldn't be too much work to write some cramfs patching tool
22:45:34bluebrotherespecially if it's just replacing files
22:45:51lorenzo92it's also about adding something...
22:46:03bluebrotherso the only thing we need to change in the inode structure is length / offsets
22:46:15bluebrotherhmm. Ok, then it will be a bit more work :)
22:46:20lorenzo92yeah :)
22:46:33lorenzo92eventually we could merge 2 images as I proposed...
22:46:34bluebrotherbut the format itself doesn't look like something that is likely to give problems.
22:46:52bluebrotherwould that make things easier?
22:46:57lorenzo92so that we keep inode structures
22:47:01lorenzo92I think so
22:47:13lorenzo92no permissions to handle on different OSes
22:47:36lorenzo92and we can keep our cramfs in the repo (no legal issues)
22:48:06bluebrotherwhat exactly are we doing to the image? Add a file and replace another one?
22:48:46lorenzo92replace a file and adding several ones (at least 2 + folder atm)
22:49:02lorenzo92but I'm planning to update the safe mode, so there will be 3 files or so
22:49:13bluebrotherok, so we need to handle folders as well :/
22:49:48*ZincAlloy just tried a much darker gradient for cabbiev2 for clip zip. no luck. looks just as shitty. the darkest shade of grey is just much too light compared with black.
22:50:44lorenzo92bluebrother: well, if that is a problem we may start in a more simple way avoiding that
22:50:55lorenzo92i.e. placing files in etc/ for example
22:51:03 Quit thomasjfox (Quit: Konversation terminated!)
22:52:12lorenzo92bluebrother: I just remind you about these tools">g#506 ... they will be the first step in patching the firmware
22:52:14fs-bluebotGerrit review #506 at"> : Firmware tools for Samsung YP-R0/YP-R1 (and possibly others) by Lorenzo Miori (changes/06/506/6)
22:52:41lorenzo92they *should* work also on windows, I have only used ansi c standard lib
22:53:30[Saint]Hmmmmmm, odd.
22:53:56[Saint]Something is making 'sync" hang on my machine with an iPod Color plugged in using Rockbox USB.
22:54:03[Saint]...doesn't happen in disk-mode.
22:54:14[Saint]Odd odd odd.
22:54:54 Join rockbox [0] (~androirc@
22:55:46bluebrotherhmm, openssl
22:56:05bluebrotherthat's not really a nice dependency on Windows :/
22:56:13bluebrotherwell, a native Windows build setup that it :)
22:56:18 Quit rockbox (Remote host closed the connection)
22:56:34 Join rockbox [0] (~androirc@
22:56:39 Quit rockbox (Remote host closed the connection)
22:56:52 Join rockbox [0] (~androirc@
22:57:20rockboxbluebrother: connection problems pff
22:59:05 Quit lorenzo92 (Ping timeout: 256 seconds)
22:59:18 Quit rockbox (Remote host closed the connection)
22:59:46 Join lorenzo92 [0] (
23:00:02lorenzo92bluebrother: back online
23:00:46lorenzo92we can get rid of openssl -> it's just used for md5 computation
23:00:58 Part micah
23:01:06lorenzo92i think there is an implementation somewhere for it in rb, right?
23:01:41[Saint]hum, not sure if its in-core or a plugin, though.
23:01:42 Quit amayer (Quit: Leaving)
23:01:43bluebrotherah, right.
23:01:47bluebrothera plugin
23:01:55bluebrotherbut I don't think that matters much :)
23:02:11bluebrotherjust make it some libmd5sum, then use it from ypr0tools
23:02:51[Saint]many users accidentally find that plugin and complain.
23:03:00[Saint]but I don't recall it happening recently.
23:03:06[Saint]Its amusing when it does happen.
23:03:20[Saint]"My player is doing...things! Oh noes!"
23:03:56bluebrothermight be related to users using their phones these days instead?
23:04:38[Saint]I think that directly relates to marketing, and what is "cool".
23:04:52[Saint]Dedicated DAPs are largely dead.
23:05:14***Saving seen data "./dancer.seen"
23:05:23bluebrotherwell, they are yet-another-pice-of-hardware to carry around
23:05:34bluebrotherand given that phones have enough capabilities these days ...
23:05:44bluebrotherstill, working with the old hardware is fun :)
23:05:59[Saint]Oh, agreeable. If only all of them had decent internal amps and SQ>
23:06:35[Saint]All of them can do audio. Not all of them are any good at it. :)
23:06:51bluebrotherwell, those kids these days don't care too much about quality.
23:07:01bluebrotherat least the marketing target group :)
23:07:27 Quit Strife1989 (Quit: Heading out.)
23:07:35bluebrotherlorenzo92: does the YP-R1 hardware do it as well or is that different to the R0?
23:07:36[Saint]everything is streamed now, no room for quality.
23:07:41[Saint]until maybe....opus.
23:08:01lorenzo92bluebrother: kinda similar, same cpu, same firmware format and same kernel
23:08:05lorenzo92i have it too
23:08:21[Saint]But widespread adoption would still be far off even if it was adopted by manufacturers today.
23:08:28bluebrotherI've seen a R1 on ebay. Maybe I should try getting it
23:08:37bluebrotherassuming it doesn't get too pricey
23:08:59lorenzo92i would buy an R0 instead, but just because it has micro sd and NO touch :D
23:09:13[Saint]Yeah, R0
23:09:28[Saint]ask lorenzo92, I'm pretty sure he farms DAPs.
23:09:56bluebrotherfor those I only get display protection thingies found on ebay
23:10:08lorenzo92haha well, I have 3 daps now ^^
23:10:14bluebrotherbut for hacking cramfs I don't need a device
23:10:22lorenzo92indeed haha
23:13:43lebelliumX5 added! I also see a Samsung player needing Rockbox here :D
23:16:37[Saint]Is RoLo known broken?
23:17:00ZincAlloyyou collect classic mp3 players? :D
23:18:16lebelliumZincAlloy: not only classic mp3 players but basically all mp3 players :) On the pic that's only 7 out of 78
23:18:29ZincAlloyoh my
23:18:43ZincAlloythat's plenty
23:18:46lorenzo92bluebrother: oh! i actually found several copies of md5.c/.h in different parts of rb repo, i will do the same hihi
23:20:08fs-bluebotBuild Server message: New build round started.">Revision 49bcf35, 217 builds, 18 clients.
23:20:53[Saint]what the hell?
23:21:04[Saint]Rockbox keeps stating the smae version number.
23:21:16[Saint]but the fs is fine, and the update *is* applying.
23:21:32[Saint]and RoLo doesn't seem to work.
23:22:02[Saint]My beautiful CF'd, micro-USB'ed, bluetooth'ed Color may be dying.
23:22:14*[Saint] weeps
23:22:57bluebrotherI have a h10 that is showing similar symptoms since years.
23:23:10bluebrotherit's mostly collecting dust though, so ... :)
23:23:53[Saint]I need to use this one today.
23:24:00[Saint]luckily it still has audio on it :)
23:24:10 Quit kevku (Ping timeout: 259 seconds)
23:24:11[Saint]and, the codecs appear fine, and /half/ of the UI works.
23:24:16bluebrotheruse a different one :)
23:25:14[Saint]That would require being in a place I'm not. :)
23:27:40fs-bluebotBuild Server message: Build round completed after 452 seconds.
23:28:17 Quit [Saint] (Remote host closed the connection)
23:29:19 Join [Saint] [0] (~saint@rockbox/user/saint)
23:32:45lorenzo92bluebrother: got rid of openssl ;)
23:33:24 Join wipt_ [0] (
23:33:59wipt_will rockbox ever support opus for recording?
23:35:25wipt_or does opus have too much complexity for most targets?
23:35:53wodzwipt_: probably the letter
23:36:30wipt_wodz: mp3 is that much simplier?
23:37:02 Quit Rower (Read error: Connection reset by peer)
23:39:47wodzwipt_: 1) it is simpler 2) it is much better optimized
23:43:29bluebrotherlorenzo92: nice :)
23:45:07 Quit pamaury (Ping timeout: 248 seconds)
23:49:53saratogai think opus encoding is not too slow, but i'm not too interested in working on it in rockbox since i don't see that it matters when we have lossless options and pcs for encoding
23:50:03saratogawould probably take a lot of work to get the code well optimized
23:57:43derfWell, we're going to be doing a bunch of that work anyway.
23:57:57derfBecause we need it to work on mobile for WebRTC.

Previous day | Next day