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 2006-08-06

00:14:14 Join lodesi [0] (
00:18:56 Join manegol [0] (
00:21:28ShadowdogMUIs a patch required to make album art work or something?
00:23:19ShadowdogMUUg, got a link?
00:23:22ShadowdogMUor name?
00:23:42dionoeai don't. You should be able to find info i nthe forum
00:23:43Paul_The_NerdTry searching for album art in the patch tracker...
00:24:17ShadowdogMUOk do I want one for ID3 tags or WPS?
00:24:38ShadowdogMUthe ID3 isn't even album art
00:25:38ShadowdogMUThere is no download link, this is confusing
00:25:56ShadowdogMUand before I go further, do I have to make my own, or will it be able to use the album art I added with iTunes
00:26:14Paul_The_NerdRead the description of the patch.
00:26:40Paul_The_NerdMany, if not most, of us do not use it.
00:26:48freqmodYou'll need to convert album art to bmp files,
00:26:54ShadowdogMUScrew that
00:26:58*ShadowdogMU goes back to skin making
00:27:02freqmodif you want the patch then you'll rebuild
00:27:15freqmodelse you will have to get an experimental build
00:28:40 Part Paul_The_Nerd
00:34:57 Quit manegol ("Chatzilla 0.9.75 [Firefox]")
00:39:05 Quit freqmod (Read error: 104 (Connection reset by peer))
00:55:49 Quit Bger (Read error: 113 (No route to host))
00:58:54 Quit bluebrother ("Leaving")
01:06:14***Saving seen data "./dancer.seen"
01:18:55Mikachudionoea: i am now
01:19:09Mikachudionoea: no clue there though
01:22:21 Quit secleinteer (Read error: 104 (Connection reset by peer))
01:23:05 Join secleinteer [0] (n=secleint@
01:25:30 Quit secleinteer (Read error: 104 (Connection reset by peer))
01:26:09 Join secleinteer [0] (n=secleint@
01:26:54 Quit mikearthur (Remote closed the connection)
01:29:25 Quit obo ("bye")
01:35:42 Join Nibbler [0] (
01:39:26 Part pixelma
01:50:18 Quit lodesi ("leaving")
02:01:21 Quit daurnimator (Read error: 110 (Connection timed out))
02:06:40 Quit Febs (Read error: 110 (Connection timed out))
02:20:10 Quit spiorf (Remote closed the connection)
02:30:32 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
02:32:46 Quit menollo ("CGI:IRC (Ping timeout)")
02:37:47 Quit Poka64 ("nite all")
02:41:08 Quit bagawk (Read error: 110 (Connection timed out))
02:57:26 Join JdGordon [0] (
03:00:30 Join spiorf [0] (
03:06:18***Saving seen data "./dancer.seen"
03:09:04 Join jhMikeS [0] (
03:15:35 Quit ^jhMikeS^ (Read error: 60 (Operation timed out))
03:19:10 Quit JdGordon (Read error: 110 (Connection timed out))
03:20:39 Quit damaki (Read error: 110 (Connection timed out))
03:30:36 Join Febs [0] (
03:41:36 Join daurnimator [0] (n=quae@
03:55:15 Join sankao [0] (
04:03:24 Nick jhMikeS is now known as _jhMikeS_ (
04:04:06 Join BigMac [0] (n=dummeska@
04:04:41BigMachi all
04:05:00ShadowdogMUSeen Futurama?
04:05:01 Nick _jhMikeS_ is now known as jhMikeS (
04:05:12BigMaci started coding for my contribution to the ipod 5g
04:05:19ShadowdogMUCoding what?
04:05:25BigMacShadowdogMU: yah a few times why?
04:05:29jhMikeSShadowdogMU: yes, here and ther
04:05:31ShadowdogMUBecause I just made this,
04:05:39ShadowdogMUI know his face is messed up, but what do you think?
04:06:03BigMacnice work man
04:06:18jhMikeSCan't view svg. What do you need?
04:06:28BigMaceverything looks could cept the face
04:06:34ShadowdogMUAnything Adobe pretty much, SVG viewer
04:06:46ShadowdogMUWhat browser?
04:07:33BigMacanyone any good at making sprite a sprtie sheet
04:07:45BigMacyou will be accredited
04:09:42ShadowdogMUjhMikeS, got another browser?
04:10:19 Nick bagawk_ is now known as bagawk (n=lee@unaffiliated/bagawk)
04:10:20midkayBigMac: a sprite sheet for what? literally could be anything. as well, "accredited" means something else entirely.
04:10:21jhMikeSI got it working already. It's not that messed up. Nice work.
04:11:16BigMacim making a parappa the rapper plugin
04:11:27BigMacso i need a sprite sheet of him
04:11:29ShadowdogMUThanks, how did you get it to work, I want to know, so many people ask me why it doesn't
04:11:39BigMacbecause im horrible at imagery
04:11:49BigMacand iif they do they will get credit
04:11:57jhMikeSI just went to
04:12:16jhMikeSClicked on my lang and os and ran the install. Done.
04:12:18ShadowdogMUHehe, thanks
04:12:19midkayBigMac: a "parappa the rapper plugin"? do i even want to know..
04:12:46BigMacmidgey34: have you ever played parappa?
04:12:54midkayi'm midkay, and no.
04:13:11midkayhaha, looking at some screenshots; good luck..
04:13:31BigMactried the tab thing
04:13:42midkaypress it twice.
04:13:47BigMacit should be easy coding
04:13:50BigMaci didnt notice
04:14:03midkayshould it now? how much have you done?
04:14:13BigMaci have the source written in asm
04:14:21midkaywritten yourself?
04:14:30BigMacand a few of my friends agreed to help
04:15:02midkayso you'll just kind of convert someone else's asm code into C.. sounds easy, riight...
04:15:03BigMacbut we just thought about doing it today
04:15:27BigMacwere going to rewrite it
04:15:30BigMacbut i know asm
04:15:33BigMacso i can compare
04:15:46BigMacso i can see what something does in asm and replicate it in c
04:15:55midkaygood luck..
04:16:07BigMaceveryone said trying something is the best way to learn, so im trying
04:16:21*jhMikeS has finished his "homework assignment" =^)
04:17:01BigMacill post screenies when i make some progress
04:17:40midkayBigMac: IRC troubles again, or have you actually given up on that social channel? :)
04:17:52BigMacgave up
04:18:59 Quit sankao (Read error: 104 (Connection reset by peer))
04:19:29 Join tinkerer [0] (
04:19:38tinkerercan rockbox play .rm files?
04:20:02BigMaci dont think so
04:20:07BigMacill check
04:20:24midkayare those video files?
04:20:48midkay"no" regardless.
04:22:21 Part tinkerer
04:39:37 Quit hannesd (Read error: 104 (Connection reset by peer))
04:39:51 Join sankao [0] (
04:47:02 Quit BigMac (Read error: 104 (Connection reset by peer))
04:49:10 Join hannesd [0] (
04:51:53 Join BigMac [0] (n=dummeska@
05:02:14 Join Rob2222_ [0] (
05:06:21***Saving seen data "./dancer.seen"
05:06:57 Part BigMac
05:16:34 Join damaki [0] (
05:18:48 Quit Rob2222 (Read error: 113 (No route to host))
05:41:00 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
05:41:22 Join JdGordon [0] (
05:44:41 Join Rob2222 [0] (
05:52:42 Join wei123 [0] (
06:04:43 Quit wei123 (Client Quit)
06:07:50 Join peejay [0] (
06:12:52 Quit peejay ("CGI:IRC (Ping timeout)")
06:13:49 Join peejay [0] (n=iparkmyc@
06:15:46peejayis anyone here?
06:16:17*peejay slaps _Veseliq_ around a bit with a large trout
06:16:46peejayhey wankers! ANWSER ME!!!
06:33:46 Quit JdGordon (Read error: 104 (Connection reset by peer))
06:47:48 Join Paul_The_Nerd [0] (
06:55:13 Join BHSPitLappy [0] (
07:00:30BHSPitMonkeycall me a wanker, will you
07:00:50peejaythat was meant to be banker
07:00:54BHSPitMonkeyyou never asked a question
07:01:05BHSPitMonkeyso, I can understand why nobody "anwsered" you
07:01:13peejayi did
07:01:21peejayi asked if anyone was here :P
07:01:38BHSPitMonkey"is anyone here" isn't a question anyone will get off their ass and answer
07:01:50BHSPitMonkeysince you can clearly see there are 100 users here
07:02:21peejayyea but alot of people leave their pcs
07:02:56Paul_The_NerdAnd a lot of people just work until something worth responding to comes up
07:03:19BHSPitMonkeyasking to ask your question is bad IRC practice
07:03:26peejayright ok
07:03:32peejayi'll know in future i guess
07:03:39Paul_The_NerdMe, I just sit around looking for excuses to tell people to read the manual.
07:03:54BHSPitMonkeylet's hope you're one of those excuses, peejay !
07:04:00*BHSPitMonkey grins to Paul_The_Nerd
07:04:05Paul_The_NerdSo, before you ask your question, whatever it may be, consider carefully whether the answer might be lurking in the depths of the manual.
07:04:17peejayi dont have the manual :D
07:04:44Paul_The_NerdYou're on the internet.
07:04:51peejayAh true
07:05:05BHSPitMonkeygot ya there.
07:05:08Paul_The_NerdThe manual is readily available from our site in both HTML and PDF forms.
07:05:17Paul_The_NerdWe're handly like that.
07:05:25peejaywhat manual are you talkin about?
07:05:27Paul_The_NerdWe won't even settle for "My PDF viewer causes my system to crash"
07:05:38Paul_The_NerdWell, this is #rockbox, so the Rockbox manual...
07:06:22***Saving seen data "./dancer.seen"
07:07:58 Join jn [0] (
07:08:03peejayi doubt this problem is in 'the manual' to be honest
07:08:18jni have an idea for you guys :) anyone familar with the security system homebrew for psp
07:08:25jni was thinking about something similar for rb
07:09:04jnsomething that will check the cpu device mounts too.. if it doesnt match then it sends email to specified address with ip of the cpu it is being used on
07:09:30peejayhow come when i start Rockbox up the menu looks like DOS?
07:09:47Paul_The_NerdYeah, that's in the manual.
07:10:01Paul_The_NerdIf by "Like DOS" you mean "It's text-based"
07:10:16peejayso where is 'the manual'?
07:10:17jn"Yeah, that's in the manual." talking to me Paul?
07:11:02Paul_The_Nerdpeejay: You could try the "Documentation" link on the website... it's also available on the Daily Build page I believe
07:11:21Paul_The_NerdI mean, there's not really a "why" to it, so much as a "That's how it is. You can change the font or background if you'd like."
07:12:13peejayyea but i've spent the past 2 hours searching the site and forums
07:12:21peejayand am none the wiser to it :(
07:12:29peejayi dunno how to change them
07:12:31Paul_The_Nerdjn: No. Your idea is silly and impossible. It would never work without removing the dual boot option, plus most Rockboxable devices only exist as a UMS storage device, and the firmware can't actually communicate with the PC really, over the USB. :-P
07:13:15Paul_The_Nerdpeejay: Read the manual. Come back if there are parts of it you don't get. Say "It told me to do X, but when I try it, it doesn't work" or "It says do Y, but I really don't get what it means"]
07:13:45peejayif the manual is the documentation i've done everythin it says
07:13:50jnall you need to do is get files to autorun upon mounting
07:14:47Paul_The_Nerdpeejay: The manual is the manual. It's specifically called the manual.
07:15:29Paul_The_Nerdjn: Out of curiosity, would you just put several different programs designed to run on different OSes, and just hope the user has autoplay enabled?
07:16:32Paul_The_Nerdpeejay: Changing a theme is as simple as extracting it, putting the files in the right place, choosing "Browse Themes" and picking it. If it doesn't work, you're either trying to use a theme for a different player, or you haven't put the files in the right place.
07:17:54jnno i would make a version for *nix and one for windows and include a man
07:18:26jnyoull think of how you mocked me one day if you come home to your ipod missing..
07:18:39Paul_The_NerdOut of curiosity, how would it work?
07:18:50 Quit Febs (Read error: 104 (Connection reset by peer))
07:18:52BHSPitMonkeyOut of curiosity, what does this have to do with rockbox?
07:18:55Paul_The_NerdIf I have autoplay disabled and plug in an iPod with your software on it, it doesn't run, I delete it, I'm done.
07:19:05jngoogle psp anti theft
07:19:31jnyou dont need autoplay on..
07:19:37jnstop talking shit and read
07:19:48jnat least back up your ego.. god damn
07:19:55Paul_The_NerdI did back it up.
07:19:59Paul_The_NerdWhile you've not explained anything
07:20:34jnNo actually your rabid assumtions are hardly support for your claims
07:20:35Paul_The_NerdThe PSP has control of the USB device, and does not ONLY serve as an UMS device.
07:20:48Paul_The_NerdWhich means that software running on the PSP has access to the USB port for its own uses
07:20:55Paul_The_NerdI explained to you that software running on Rockbox in most cases does not
07:20:59BHSPitMonkeyjn, did you read about the original program on digg?
07:21:01Paul_The_NerdWhich means the software would need to run on the host PC
07:21:22Paul_The_NerdYou made the assumption that just because it can be done on the PSP it can work here, without any real technical knowledge
07:21:32jnor in theory NOT be running on the host pc
07:21:45Paul_The_NerdIn what theory?
07:21:49Paul_The_NerdWhere would it run if not the host PC?
07:22:33*BHSPitMonkey decides that jn should quit while he/she is behind.
07:23:17Paul_The_NerdHeck, the PSP one is a rather cheap hack too, if it depends on a .bat file
07:23:26Paul_The_NerdSeriously, here's how I'd bypass the PSP one:
07:23:33Paul_The_NerdA) Remove Memory Stick Duo
07:23:39Paul_The_NerdB) Delete PATS
07:23:40Paul_The_NerdC) reinsert
07:24:08jnor format ms from psp
07:24:49jnsave you the trouble of removing it
07:25:09Paul_The_NerdI wasn't even aware the PSP had a format option. Of course, I'd assumed the PATS was actually decent until I read about it.
07:25:31Paul_The_NerdSince it's just a simple script and a file set to autoplay, a small change or two and I'm sure you could use it on your iPod.
07:25:52jnbut not on rb
07:26:02Paul_The_NerdIt has nothing to do with iPod
07:26:07Paul_The_NerdIt'll work on any storage device.
07:26:15Paul_The_NerdErr nothing to do with Rockbox even.
07:26:23jnah.. want to help make this happen?
07:26:34BHSPitMonkey"This is an App that will help you get back your stolen psp... or any other usb device that uses Autoruns :D"
07:26:34jnhaha.. ok
07:26:44Paul_The_NerdBecause it's stupid, as I said.
07:26:46BHSPitMonkeyin other words, extract the damn thing to your iPod. that's it.
07:26:47Paul_The_NerdAnyone who disables autoplay is safe from it.
07:26:53BHSPitMonkeythere's nothing to "make happen"
07:27:08Paul_The_NerdWell he might want to change the .reg file
07:27:08jnBHSPitMonkey: how bout nix version
07:27:17Paul_The_NerdSo that it creates a unique folder separate from the one for his PSP
07:27:40Paul_The_Nerdjn: No *nix system is likely to automatically run ANY file from a plugged in mass storage device without user intervention.
07:27:58BHSPitMonkeyjn, personally, I think nix users are less likely to steal from people :)
07:28:20jndude im talking about me
07:28:30jni have to get it to recognize the nix box
07:28:37BHSPitMonkeythe "Version" depends on the theif's hardware.
07:28:37Paul_The_NerdPersonally, I think *nix users are likely to steal something far more interesting than an iPod if looking for pocket sized gadgets.
07:28:51BHSPitMonkeywell, not really.
07:29:05BHSPitMonkeyhehe Paul_The_Nerd
07:29:10Paul_The_Nerdjn: If it doesn't run, it doesn't have to recognize your box, now does it?
07:29:26BHSPitMonkeypaul is correct.
07:29:34BHSPitMonkeyit already is a nix version, pretty much..
07:29:37jnIf it doesn't run?
07:29:47Paul_The_NerdIt's entirely flawed in its base theory
07:29:47BHSPitMonkeyit will catch any theif with windows and autorun enabled.
07:29:57Paul_The_NerdThe security ONLY happens if the script gets executed
07:30:10Paul_The_NerdThe script only gets executed by stupid computer users: Those who have autoplay enabled in windows.
07:30:12jnbut you add the reg entires so it recognizes the owners box as safe
07:30:20BHSPitMonkeyjn, it won't even RUN on your computer.
07:30:23Paul_The_NerdHere's how I'd create such a thing
07:30:26BHSPitMonkeyno run, no registry check.
07:30:28Paul_The_Nerd1) It would require a modchipped PSP
07:30:28 Join unknownhost [0] (
07:30:28jnyour right
07:30:43Paul_The_Nerd2) You "lock" your PSP when not using it.
07:30:55jn"lock" ?
07:31:00Paul_The_Nerd3) When a USB cable is plugged in, or someone attempts to enter USB mode, it tries to connect to any open wifi networks and sent the email
07:31:13BHSPitMonkey"lock" is a function paul just made up
07:31:14Paul_The_Nerdjn: It would be a program on the PSP, notice I said it would require a modded one.
07:31:24Paul_The_NerdIt is something I made up. Thank you. :)
07:31:27jnhomebrew does not req a modded psp
07:31:36jnreq psp with >2.8 fw
07:31:43BHSPitMonkeyjn, homebrew also isn't capable of this concept.
07:31:44jn2.8 is newest and will be exploited soon
07:31:45unknownhostI'm getting: gcc: installation problem, cannot exec 'cc1': No such file or directory
07:31:53Paul_The_Nerd4) If a WiFi network could not be found, it attempts to use the USB port to gain access to the internet by way of the host PC, rather than serving as a UMS device. Hopefully this is possible.
07:31:54unknownhostdo I change my path or move the file?
07:31:58jnok then it would req flashing a 1.5 fw
07:32:04BHSPitMonkeyooh, a real problem!
07:32:09Paul_The_NerdNo storage-type connection will ever be made.
07:32:13BHSPitMonkeylet's divert attention from jn to unknownhost
07:32:22unknownhosttrying to build a sim
07:32:26jnadmit im sexy tho
07:32:35Paul_The_Nerdunknownhost: Your paths are wrong somehow, though I couldn't say how.
07:32:44BHSPitMonkeyand jn, why don't you try the command "/server" and do a /list there
07:33:14jnunknownhost: is that the only error you are recieving?
07:33:16unknownhostI was followign the simple tutorial soemone was so gracious to write....
07:33:38unknownhostbut my default path didn't match what was written in the tutorial
07:33:41unknownhostso I added:
07:33:44jnmight need build-essentials
07:34:20BHSPitMonkeywe're not all using ubuntu, jn
07:34:30peejaydoes anyone know a good program for converting videos for an iRiver H320?
07:34:40Paul_The_Nerdpeejay: Rockbox doesn't support video.
07:34:55BHSPitMonkey(though I actually happen to be, you shouldn't assume everyone else is)
07:35:02Paul_The_Nerdunknownhost: Do you get a positive response when you do a sh1-elf-gcc −−version ?
07:35:14jnBHSPitMonkey: what do you think?
07:35:30unknownhostcommand not found... does the path look ok?
07:35:54BHSPitMonkeyjn, I think I could go for a burger from steak n' shake right about now.
07:36:09peejayno but cant i still watch videos by accessing the original firmware?
07:36:12Paul_The_Nerdunknownhost: Are the cross compilers actually installed to those locations?
07:36:21jnpz then, im out too.. nice arguing with you guys for 20 minutes :)
07:36:29unknownhostI have 5 copies of cc1.exe, but none are in any of those locations
07:36:32Paul_The_Nerdpeejay: Yeah, but this isn't a channel about the original firmware. In fact most people here don't even have that player.
07:36:44peejayok right sorry
07:36:51Paul_The_Nerdunknownhost: Are you using Cygwin?
07:36:52*BHSPitMonkey is one of those "most"
07:36:55jnpeejay: yes you can.. google.
07:37:30peejayi have, only gettin people complainin bout VirtualDub
07:37:34peejaydoesnt matter
07:37:36peejaythanks anyway
07:37:42Paul_The_Nerdunknownhost: You probably should've followed what the tutorial said, though I honestly don't know. The Cygwin method has always been a bit fuzzy at times.
07:38:24unknownhostshould I just start adding to the path? or copy the file(s) to another location? what's best?
07:38:25*Paul_The_Nerd strongly recommends using VMWare instead.
07:38:42Paul_The_NerdThe solution is almost never "move the files"
07:39:03unknownhosthehe .. to the trash maybe
07:39:04Paul_The_NerdBut the problem is that the path doesn't point to your crosscompilers.
07:39:06Paul_The_Nerdat all.
07:39:39unknownhostright .. there are many versions of cc1... from 3.4.4 to 4.0.3
07:39:43 Join JdGordon [0] (
07:39:58unknownhostI need the latest version lined up in a path?
07:40:20Paul_The_NerdWhere in what I said did you get *that* statement?
07:40:37unknownhostI just made it up
07:40:53Paul_The_NerdYou need to make sure the path actually points to where all three of the cross compilers are located.
07:40:59unknownhostthe word 'right' was a mistake
07:41:00Paul_The_NerdIt will *probably* work if you get that bit fixed.
07:41:52*peejay slaps Paul_The_Nerd around a bit with a large trout
07:42:40jnhey that idea you had paul is pretty good and with a hacked fw that already exists... it could work out
07:42:51 Quit peejay ()
07:43:32Paul_The_Nerdunknownhost: Seriously, if you get the paths pointing to whatever /bin/ folders arm-elf-gcc, sh1-elf-gcc and m68k-elf-gcc are hiding in, it'll probably work.
07:45:02unknownhostyeah.. I guess that's the part I can't figure out
07:45:06unknownhoston this page:
07:45:33unknownhostat the bottom.... the base path is totally different that what was installed in my cygwin
07:45:49Paul_The_NerdWell, when you downloaded the CrossCompiler packages from the Rockbox Mirror, it *should've* set up the paths itself...
07:46:06unknownhostI guess I'm not sure of the syntax. neither appending or replacing the path line seem to get me going.
07:46:53Paul_The_NerdWhat you simply needed to do was add this line: PATH=$PATH:$HOME/bin:/usr/sbin:/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin to the end of the file.
07:47:09Paul_The_NerdAssuming everything (the cross compilers) got installed to the right places
07:47:18Paul_The_NerdOut of curiosity does /opt/arm/bin actually exist?
07:47:41 Quit JdGordon (Remote closed the connection)
07:49:35Paul_The_NerdAlso, why do you have ${PATH} instead of $PATH ?
07:49:43daurnimatorwhy not?
07:50:02Paul_The_NerdIs Cygwin friendly with that?
07:50:12unknownhostThat was in my default cygwin install when I initially opened it...
07:50:27 Join JdGordon [0] (
07:50:31unknownhostI tried appending with - :/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin
07:50:45unknownhostthen I tried replacing with - PATH=$PATH:$HOME/bin:/usr/sbin:/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin
07:50:49Paul_The_NerdIt doesn't say to append it...
07:51:00Paul_The_NerdIt also doesn't say to replace it.
07:51:08unknownhostit says to add at the end of the line... isn't that append?
07:51:16unknownhost"Edit .bash_profile (found in your home directory) and add ":/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin" (without the quotation marks) to the end of the line that says:"
07:52:07unknownhost( in my install, the last line looks like: PATH=${HOME}/bin:${PATH} )
07:52:08Paul_The_NerdOh, it's changed slightly
07:52:50unknownhostnot sure what the curley brackets are all about.
07:53:02Paul_The_NerdCygwin has perhaps changed its format?
07:53:20unknownhostvars i guess for local install
07:53:23Paul_The_NerdAs it says, the path should've also automatically been set up when you installed the Cross Compilers from the Rockbox mirror.
07:53:45unknownhostanyone one else using cygwin with a path I could try?
07:54:04Paul_The_NerdThe one on the wiki worked in the past.
07:55:42Paul_The_NerdWhy not just scrap cygwin and try vmware?
07:56:31unknownhostkind of a lot of overhead for me... that's what I had going before. Cygwin seemed a little less bloated.
07:56:39unknownhostno services... etc...
08:00:31 Join damaki_ [0] (
08:08:38 Quit unknownhost ()
08:10:01 Join unknownhost [0] (
08:10:49unknownhostis there a display path command?
08:14:08 Join midkay_ [0] (
08:14:44 Quit midkay (Read error: 104 (Connection reset by peer))
08:15:07unknownhostwhat does !C: mean?
08:15:08 Quit damaki (Read error: 110 (Connection timed out))
08:16:54 Quit dongs (Remote closed the connection)
08:16:57 Join dongs [0] (
08:20:55 Quit Paul_The_Nerd (Read error: 110 (Connection timed out))
08:28:57 Quit unknownhost ()
08:33:34 Join RedBreva_ [0] (
08:33:36 Nick RedBreva_ is now known as RedBreva (
08:40:15JdGordondamn waiting for the cross compilers to build is boring :p
08:40:47JdGordon1/3 done...
08:40:47jni wish there was someway i could help speed up video playback in rb
08:42:52jnreally? that's good.
08:43:10jnwhats eta for the remaining 2/3?
08:43:25JdGordonsh took 25min, so another hour i guess
08:43:38JdGordondunno why its taking so long, last time was 15min each
08:45:19jnit will be done tonight?
08:45:31jnwith decent framerate? your joking
08:46:31JdGordonwhat the heck??
08:48:06 Join safetydan [0] (
08:50:36 Join RedBreva_ [0] (
08:56:50BHSPitLappymy crosscompilers only get like 17fps
08:56:53BHSPitLappythey suck
08:59:08 Quit RedBreva (Read error: 60 (Operation timed out))
08:59:08jnany of you guys watch the movie "Pulse" yet?
08:59:22BHSPitLappylooks like another piece of crap
08:59:28BHSPitLappylike that video game horror movie
08:59:33jnthe original is pretty good..
08:59:42BHSPitLappyisn't pulse the one with the computers?
08:59:49BHSPitLappyoh wait
09:00:04jnits a remake of a japanese movie.. like ringu to ring
09:00:26 Join Bger [0] (n=bager@rockbox/developer/Bger)
09:00:34BHSPitLappyhow can you make a movie like that, without it becoming laughably dated within months
09:00:54jnwargames is still good.
09:01:05BHSPitLappylike, every time an old movie with a computer or something in it shows up, you have to laugh
09:01:13BHSPitLappybut you're centering the plot around it...
09:01:28jnmy girlfriend is making funny noises lol
09:01:42BHSPitLappywargames could still be accurate, there's lots of CLI users out there :P
09:01:51jnshe's asleep and im playing with her feet, almost time for "hidden penis, sleeping vagina"
09:01:52BHSPitLappyare you doing something to her? o.O
09:02:04BHSPitLappyyeah, not appropriate in here. sorry.
09:02:17jnlol my bad..
09:02:31BHSPitLappyor on most of this network
09:02:41jnagain my appologies
09:03:17BHSPitLappynow you know
09:04:25jnis there a faq about video playback on rb and what the devolpers are up against...
09:04:32jnim curious
09:04:50BHSPitLappyumm, a lack of initiative?
09:05:15BHSPitLappywhen a developer comes along who wants the feature bad enough, he'll add it in
09:05:33jnbut there IS work being done on it currently is there not?
09:05:42BHSPitLappyanother obstacle is a lack of important data about the hardware
09:05:49BHSPitLappy*cough* PortalPlayer *cough*
09:05:59BHSPitLappyI don't know.
09:06:24***Saving seen data "./dancer.seen"
09:06:58jnwhat is the main focus on now?
09:10:51 Quit daurnimator ("daurnimator was using Daurnimator's Legendary Script")
09:12:05 Join jonno [0] (
09:12:29 Quit JdGordon (Nick collision from services.)
09:12:32 Nick jonno is now known as JdGordon (
09:12:48safetydanjn, whatever the dev wants to do
09:12:54safetydanthe joy and curse of opensource
09:12:58safetydanit's like herding cats really
09:13:15JdGordonwoohoo... cross compilers done!
09:13:21safetydannone of the devs really seem that interested in video support
09:13:41safetydangoing to be a hard slog getting a codec optimised enough to run well enough to decode at a reasonable frame rate
09:14:45jnis there anyway i can help?
09:15:06safetydanget coding :)
09:15:34safetydanor convince someone at PortalPlayer to give Rockbox the developer documentation
09:15:50safetydanAnd Broadcomm as well as I think the make the video chipset for the iPod VIdeo
09:16:24jnBroadcom agh
09:29:31 Quit damaki_ (Read error: 104 (Connection reset by peer))
09:31:08 Join damaki [0] (
09:32:45 Join mikearthur [0] (
09:48:41 Quit JdGordon ("Konversation terminated!")
09:50:04 Join MarcoPolo [0] (
09:52:49 Join powr-toc [0] (
09:54:16 Join RedBreva__ [0] (
09:54:18 Nick RedBreva__ is now known as RedBreva (
09:58:14 Nick midkay_ is now known as midkay (
09:58:27 Quit MarcoPolo ("Bye !")
09:59:22 Join MarcoPolo [0] (
10:01:21 Join matsl [0] (
10:02:48 Quit RedBreva_ (Read error: 60 (Operation timed out))
10:14:34 Join unknownhost [0] (
10:15:26unknownhostIs it possible to apply the daily CVS with the album art patch? I get errors. Does that mean the patch needs to be updated?
10:16:01Bgerunknownhost: probably
10:16:39unknownhostwhat's the best way of figuring out the last cvs that worked with a patch?
10:20:06Bgerunknownhost: it would be easier if you just see why the patch failed and correct it manually
10:20:43unknownhosti wish that were the case
10:21:45lexis there are rss feed for the cvs activities? :o
10:22:10 Join FOAD [0] (
10:22:29unknownhostbger ... i actually tried going thruogh the error file, but I couldn't figure it out.
10:25:03 Join sando [0] (i=lolsteam@
10:28:57 Quit unknownhost ()
10:29:56 Join lodesi [0] (
10:31:18 Join nilss [0] (
10:37:53 Quit FOAD_ (Read error: 110 (Connection timed out))
10:38:52 Join rudefyet [0] (n=bobdole@
10:41:17 Quit rudefyet ()
10:42:29SlasheriLinusN: What would you think about this bootloader patch:
10:43:04 Join damaki_ [0] (
10:43:41SlasheriLinusN: It enabled Rockbox to load like before, or load from flash if original firmware has been replaced by rockbox. I have tested it on H140 and it was working fine. It also includes the EEPROM settings and disk intact flag.
10:52:26jhMikeSsafetydan: You there? I saw what you said earlier. I'm _very_ interested in video support.
10:53:28 Quit lodesi ("leaving")
10:54:57 Join lodesi [0] (
10:56:40 Join damaki__ [0] (
11:00:05safetydanjhMikeS, well now you just need to write the code to do it
11:01:40 Quit damaki (Read error: 110 (Connection timed out))
11:02:49 Join Kohlrabi [0] (
11:04:18 Quit BHSPitLappy (Remote closed the connection)
11:04:32safetydanjhMikeS, AFAIK the big issues are performance. Brush up on you arm, m68k, and sh skillz :)
11:06:28***Saving seen data "./dancer.seen"
11:09:59 Quit sankao (Read error: 110 (Connection timed out))
11:12:34 Quit damaki_ (Read error: 110 (Connection timed out))
11:14:24 Join arkascha [0] (
11:15:00 Part arkascha ("Konversation terminated!")
11:15:20jhMikeSsafetydan: There's also the matter of writing directly to the vid chips instead of buffering first. Some chips can be made to produce overlays for OSD. Should videos have to be native screen size?
11:15:44safetydanwhich is going to vary from target to target
11:15:50safetydannative screen size is always going to be easiest
11:16:17 Quit MarcoPolo (Remote closed the connection)
11:16:19jhMikeSOf course, but it's been on my mind a lot and I figured a way to have it not matter.
11:16:55jhMikeSScaling is way too hard for most unless its simple integer ratios.
11:18:03safetydanSo what do you propose instead?
11:18:20jhMikeSInstead of what? Scaling? or buffering?
11:19:27jhMikeSI propose a video and overlay API set plus a way of inlining the driver code into the codec. It only has to work with one display chip at a time.
11:19:51jhMikeSI'm guessing either integer sizes or forced native size.
11:20:12safetydanah, sorry thought you had some magical way to make non-integer scaling go fast :)
11:20:29jhMikeSWell, yeah.
11:20:45jhMikeSThat too, its a periodic function.
11:21:30jhMikeSI don't know my codecs and I'm guessing a full decode is almost always nescessary.
11:22:36jhMikeSIt's really always integer scaling unless you want nice soft interpolation.
11:22:41chendomy h340 won't start cause my 2200mAh battery's too fat and the HD can't spin up
11:22:57chendowhen people remove the blue padding, is it both paddings? or just the one from the HD to the battery
11:24:31 Join ze [0] (n=ze@
11:25:16jhMikeSRight now I have definite things in mind for the X5 and will read the chip docs for other devices. What players will take it? I mean, x5 can do it, iPod video of course, and ??
11:26:14*jhMikeS wonders if he's bored and talked poor safetydan to death. ??:)
11:26:35safetydanprobably h300 and h100 (though only in grayscale)
11:26:49jhMikeSIsn't H300 color??
11:26:54safetydanyeah it is
11:26:57safetydanI was talking about the h100
11:27:14safetydanI think the h300 and h100 are the same apart from the USBOTG and the color screen
11:27:24jhMikeSWhat gray shade count?
11:27:30safetydan33 I think
11:27:31safetydanor thereabouts
11:28:11jhMikeS33! Lose one and it's faster pow 2 math. :)
11:28:35safetydanamiconn is the person to talk to about that
11:28:39safetydanI'm probably wrong on the 33 count
11:29:19safetydanit's not normal grey scale support though... some trick with temporal aliasing and lcd refresh rates
11:29:25jhMikeSI think I might prove some stuff with the jpeg viewer first: Direct write, OSD overlay, 18-bit color.
11:29:45 Join _FireFly_ [0] (
11:29:57 Quit matsl (Remote closed the connection)
11:31:25jhMikeSI'm not sure what that means re: H100 lcd.
11:32:33safetydanjhMikeS, it's not just h100 lcd, amiconn has the same trick working on all the grayscale lcds (or at least nearly all)
11:33:07safetydanit's probably not too relevant to video support as only crazy people would probably want to watch much grainy grayscale video on a tiny screen :)
11:33:34jhMikeSIt's a hack then so the chip is not natively grayscale? You mean flip-flopping B&W to make different grays?
11:34:00safetydanpretty much
11:34:35 Quit TomFelker (Read error: 110 (Connection timed out))
11:34:40jhMikeSI'll have to look at that code closely.
11:35:05jhMikeSSoftware PWM it sounds like.
11:37:51 Join ze__ [0] (n=ze@
11:40:33 Quit ze__ (Client Quit)
11:40:39 Join ze__ [0] (i=ze@
11:41:04 Quit ze (Nick collision from services.)
11:41:09 Nick ze__ is now known as ze (i=ze@
11:44:41jhMikeSthat's highly specialized it appears
11:46:26 Part amiconn
11:46:40linuxstbjhMikeS: Do you have any specific codec in mind for video?
11:47:32jhMikeSSomeone mentioned MPEG already. I think that's a given. I would think anything that can be added would be good.
11:48:34jhMikeSThere's a link to a decoder that could be adapted somewhere in the forum.
11:49:10linuxstbWhat kind of MPEG? MPEG-1 would seem to be the simpliest codec.
11:49:30jhMikeSI believe that's it lemme look.
11:51:55jhMikeSArrgh...can't find it. It was just there not long ago...
11:53:21jhMikeSlinuxstb: It says Llorean sent you to look at Core Pocket Media Player. That's what I saw.
11:53:45 Join obo [0] (
11:53:55 Join JdGordon [0] (
11:54:54linuxstbYep. It could be useful, but there are various other MPEG-1 decoders around as well worth looking at such as the one in ffmpeg and libmpeg2.
11:58:12 Quit safetydan (Read error: 110 (Connection timed out))
11:58:31 Join safetydan [0] (
12:00:35 Quit nudelyn ("At Argon, we're working to keep your money.")
12:01:35jhMikeSI'll go look around later I'm snoozing shortly. There's also an audio playback bug at least on the x5 but I don't know if anyone else has been able to reproduce it (I can at will) FS #5747. I'm still learning my way around and would like to know the best place to look in the source.
12:04:25safetydanjhMikeS, best places to look would be apps/playback.c and apps/pcmbuf.c
12:04:32jhMikeSNeed to know where you can detect a skip to something else from a paused state so the old data is flushed before starting playback again.
12:05:19jhMikeSthanks. How much audio does the TLV320 buffer?
12:05:23 Join nudelyn [0] (
12:05:30 Join damaki_ [0] (
12:05:46linuxstbjhMikeS: Also, the low-level pcm driver is in firmware/pcm_playback.c
12:06:03safetydanjhMikeS, not a clue
12:07:02jhMikeS^^ Guess it's datasheet time again. I hope it's not in so much Engrish (sic) like the LCD chip. :)
12:07:58safetydanjhMikeS, firmware/drivers/tlv320.c might help
12:08:17safetydanand firmware/export/tlv320.h
12:08:52jhMikeStlv320.c just seems to deal with the I2C aspect.
12:10:14safetydanthe hardware interface isn't really my area
12:12:05Bgerhow do i tell find to skip some dir ?
12:12:08jhMikeSI love the low level stuff. Much easier and cleaner than UI code and such.
12:13:47Bgerforget me
12:13:50jhMikeSsafetydan: BTW: What is your area?
12:14:16JdGordonBger: we oucld never forget you!!
12:14:20JdGordoncould even
12:14:31BgerJdGordon: wazz up ;)
12:15:05Bgerwhat happened with this sml thing ?
12:15:29JdGordongot it mostly finished, handed in on friday, next one is due in 3 weeks :'(
12:16:18Bgervery strange is this sml
12:16:22safetydanjhMikeS, I have many and none really. Mostly simulator, eq UI and the occasional burst of fixed point trig function madness
12:16:24Bger(yoda style)
12:17:35jhMikeSWhy does the simulator thread code create actual threads. At least in windows. dk about sdl.
12:18:31safetydanDoes the Windows sim still work?
12:18:47BgerjhMikeS: the only used sim is the sdl one
12:19:02jhMikeSEven under windows?
12:19:08safetydannot sure why Rockbox threads are mapped on to system threads... in some ways I think it might easier not to
12:19:12safetydanjhMikeS, yup
12:19:23jhMikeSIt that real threads or faux?
12:19:46safetydanreal threads
12:19:50jhMikeSCome to think of it, why's that code in the tree then?
12:20:02 Join aliask [0] (
12:20:06safetydanbecause the sdl sim only came about a few (6?) months back
12:20:15safetydanand no one has got around to removing it yet
12:20:34jhMikeSThey could use SwitchToFiber in windows to schedule them just like the real thing.
12:21:04safetydansubmit a patch :)
12:21:08 Quit damaki__ (Read error: 110 (Connection timed out))
12:21:40JdGordonna, we need working sound in the sim more than closer threads handling...
12:21:57jhMikeSWas pondering that. CovertThreadToFiber and SwitchToFiber is all you need. Just put those in the thread table and schedule them.
12:22:10 Join Poka64 [0] (
12:22:33jhMikeSWhat min windows version is the sim supposed to support BTW?
12:22:47JdGordonshould work on all.. as long as u have SDL.dll
12:23:04jhMikeSHmmm...fibers are Win98+.
12:23:11jhMikeSAll NT I believe
12:24:04jhMikeSThen my sick mind wants to covert thing into hardware emulators. =)
12:24:35JdGordonthere really is no need (imo) to not do something becuase it will kill win98 support
12:25:00jhMikeSIt will only kill Win95 support.
12:25:09JdGordoneven better
12:26:29jhMikeSI have a loooooong TODO/WANNADO list. Focus grasshopper.
12:27:41jhMikeSOverlays, viewports, video, etc. etc. etc.
12:28:52safetydanI think video support might be more wanted than accurate hardware emulation in the sim
12:29:21safetydanJdGordon, sound still not working for you in the sim? It mostly works here.
12:30:07JdGordoni havnt actually tried it since the most recent changes :p
12:31:12 Quit Kohlrabi (Nick collision from services.)
12:31:21 Join Kohlrabi [0] (
12:33:16jhMikeSI want to see at least NTSC frame rate at native size.
12:33:31JdGordonnot gonna happen
12:33:48jhMikeShehe...the codec?
12:34:08JdGordonthat and the proccessor isnt fast enough.. unless your tlaking about a non compressed format?
12:34:38jhMikeSno...MPEG1. Which processor. Any of them?
12:35:45linuxstbDon't forget you need audio decoding as well...
12:36:26jhMikeSI know that but what would the boost ratio be for the audio do you think?
12:36:34 Join barrywardell [0] (n=barrywar@
12:37:34safetydanmp3 decoding is mostly done with no-boost on coldfire I think
12:38:01 Join mkey [0] (
12:38:02jhMikeSI get about 18-20% boost for mp3.
12:38:22jhMikeSor maybe 15-20
12:39:56linuxstbjhMikeS: What do you mean by using overlays?
12:40:36jhMikeSThe LCD chip can hit about 132 FPS at boost with repacking 16->18 bit.
12:41:12jhMikeSlinuxstb: I mean having the codec write directly to the LCD. Compile the right bits of driver code directly into the codec.
12:41:41safetydanis the x5 not coldfire? the boost ratio for me on a h120 is nearly 0
12:41:55safetydanor is there something else going on?
12:42:19jhMikeSReally...I was using the EQ and all that and crossfeed
12:42:32jhMikeSIt's Coldfire, yes.
12:42:43safetydanah, no eq or crossfeed for me
12:43:42jhMikeSI want the full 18bit so the HW can dither it nicely.
12:44:34linuxstbjhMikeS: That sounds like a nasty hack. We have a lot of LCDs to support, and we would want it to work in the sim as well...
12:45:10JdGordon... and you can write straight to the lcd buffer anyway cant you?
12:45:15jhMikeSYou won't get the performance then. Definiely not with the framebuffer in DRAM. can't. lcd_blit is just for mono and gray I think
12:46:46linuxstbI'm not saying to use the existing LCD API, we definitely need to create a new high-performance one for video.
12:46:51jhMikeSBesides, you'd have to write to a buffer first which would kill it. Avoid that step if possible. It will take a lot of crafting but it may be worth it.
12:48:00linuxstbMPEG-1 for example decodes the video to YUV. Maybe we could implement something like lcd_blit_yuv() for each LCD that does an asm-optimised yuv to native framebuffer conversion, and writes that data directly to the LCD.
12:48:11jhMikeSThe craft is in doing it and NOT making it a nasty hack but consistent. A coded need only deal with one display chip.
12:49:30jhMikeSTake the YUV, covert it in registers and send it to the display, chopping off the lower 2 bits. Don't pack it first so it has to be repacked by the driver later either.
12:50:00linuxstbYes, that's what I mean.
12:50:38JdGordonis there really a need for this yet tho? i mean, the best video codec we have atm is 2fps without actually displaying anything
12:50:56jhMikeSJust send it to the display. Don't call a function or at least an inline one with the LCD write code.
12:51:02jhMikeS2fps??? On what?
12:51:05obocould always use huffyuv :)
12:51:08JdGordonh300 iirc
12:51:24jhMikeShuffyuv, what's that?
12:51:42oboa fairly simple loseless compression (huffman yuv I think)
12:52:09oboit's very fast, I think it's about 2.5:1 compression
12:52:21jhMikeSWhat codec for 2fps? Uncompressed video will not read off the HD fast enough.
12:53:35 Join bluebrother [0] (
12:54:21jhMikeS2.5:1 is still huge.
12:55:23safetydanjhMikeS, I believe it was xvid
12:56:02jhMikeSThat's supposed to be easy to decode, no?
12:56:19safetydanI wouldn't have thought so as it's based on MPEG4
12:56:58safetydannote that 2fps was literally just decoding the frames, not displaying anything
12:57:05safetydanthere's some detail here
12:57:08safetydansearch for 2fps
12:57:20jhMikeSA consensus needs to be reached about which player actually get video.
12:57:43JdGordonprobaly only the ipod vid, if the broadcom chip is figured out
12:57:56JdGordonuntill the codec is optimized
12:58:09jhMikeSH300 LCD is pretty much the same to work with as X5.
12:58:13JdGordonif the OF can do 15fps on the h300 then we should be able to also
12:58:32JdGordoncpu is the limiting factor here.. not lcd..
12:58:40safetydanthere's no reason not to support all players
12:58:41jhMikeSX5 does 12 on OF with MPEG4
12:58:49safetydanit just might not go very fast on some of them
12:58:59safetydanbroadcom chip based ones are another kettle of fish
12:59:13JdGordonhaha ye, on the 2 line char display one :D
12:59:16safetydanjhMikeS, you can probably skip archos as they already have a video plugin :)
12:59:21JdGordonthere is a video plugin for the archos
12:59:36jhMikeShow does that work? hehe
12:59:54safetydanuncompressed video and a custom video format I believe
13:00:13jhMikeSqualifier about 2fps: "That was with a large (640x352 I think) test file."
13:00:32safetydanapps/plugins/video.c for archos video plugin
13:01:31jhMikeSwhat codec for archos?
13:01:36 Join fejfighter [0] (
13:01:56 Quit fejfighter (Client Quit)
13:01:59 Join fejfighter [0] (
13:03:15Mikachuyou don't _have_ to display video at fullscreen
13:04:30jhMikeSoh...but I do...I do! =)
13:04:47jhMikeSIt's got that pixel flipper gadget in it too.
13:06:08JdGordonhmm.. soo many boring user requests in FS
13:06:31***Saving seen data "./dancer.seen"
13:07:22jhMikeSMy sense is that it can be worked out for the more powerful devices like the X5, H300, iPod Video...slower devices will have to make due with fewer FPS or smaller sizes. Why should everyone suffer at the lowest common denominator?
13:07:40JdGordonis there any chance of this getting commited if i do the patch?
13:07:53JdGordonthat sounds like a reasonable request imo
13:08:03JdGordoneven if it is 3 years old :p
13:08:08jhMikeSJdGordon: Let's make a top 10 list of Most Boring FS requests.
13:08:35MikachuJdGordon: i think the skipping should work like normal, pressing left once already just goes to the beginning even in other repeat modes
13:08:59JdGordonbut but pressing again while in repeat one should goto the next/prev song
13:09:06 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
13:09:20JdGordonbut does it
13:09:31Mikachuin repeat 1? i don't know
13:09:37*JdGordon checks
13:09:42linuxstbDoes "repeat one" create a playlist with only one file, or does it just repeat the current track in the playlist?
13:09:43jhMikeSI don't use repeat one so I have no idea how it behaves.
13:10:18JdGordonit doesnt
13:10:27jhMikeSI think there should be a "repeat two" option in the quick menu =)
13:10:41Mikachua-b repeat for playlists
13:11:10JdGordonlinuxstb: it looks like it just keeps repeating the one track in the playlist
13:11:23Mikachucan you select another song in View Playlist?
13:11:30amiconnjhMikeS: You cannot use the video overlay feature of the lcd controllers
13:11:44jhMikeSThat's not right. It should go to the next.
13:11:54jhMikeSamiconn: why not?
13:11:56JdGordonMikachu: just tried that, no it doesnt like you changing tracks
13:11:58amiconnIt would require cpu control for some of the lcd controller pins
13:12:41amiconnThe mode pins, to be precise
13:12:51jhMikeSThere not connected to anything?
13:13:21amiconnTheir wiring is fixed
13:13:25jhMikeSWell, I'm just talking about bypassing framebuffers and writing directly. Not chip modes.
13:13:43amiconnYou can do that, e.g. doom on H300 does it
13:13:48jhMikeSHD66773R doesn't have overlays
13:13:57jhMikeSHow much does it help?
13:13:59amiconnBut don't expect too much on other platforms
13:14:15jhMikeSThat's why your capabilities will vary
13:14:34amiconnThe HD66789R does have overlay, but we cannot use it
13:15:04jhMikeS66789 is a more powerful chip IIRC
13:15:35amiconnOn the low-depth platforms, the core has lcd_blit() to provide direct-writing to the controller without direct hardware access in the plugin itself
13:15:52amiconnThis is used in video.rock on archos, and for the grayscale library
13:15:54jhMikeSThat's such a bummer, just one trace of copper would change that
13:16:44amiconnI doubt that video overlay would help video performance much
13:16:54amiconnThe most demanding task is decoding
13:17:07jhMikeSAre the bus writes pretty much as fast as they can go?
13:17:22Mikachuwhat about using something really old like smacker video?
13:17:53jhMikeSsmacker? how old is that cause I never heard of it
13:18:03Mikachui know it's used in the intro for Creatures
13:18:37Mikachui suppose it's not free though
13:19:01jhMikeSamiconn: you mean the HW overlay you can't use right?
13:19:24Mikachuheh, the license is $1500 so forget that
13:19:29Mikachubut i know it's fast
13:19:36jhMikeSMostly a convienience
13:20:39*JdGordon is bored
13:22:43jhMikeSjust looking at the audio thread make my boost ratio go up from 0%
13:24:11 Join XavierGr [0] (
13:26:34amiconnjhMikeS: The audio debug screen boost ratio is the average taken while you are in that screen
13:26:59amiconnSo if you enter it while the cpu is unboosted, the value will start at 0% and the go up
13:27:29amiconn(unless you're using a codec that is completely content at CPUFREQ_NORMAL
13:27:31jhMikeSI'm kind of in qm? flac)
13:29:10nudelynhow come flac decodes so fast? isn't flac a lossy codec + a lossless diff on top of that? is it really optimized, or is the lossy codec really simple?
13:29:29safetydannudelyn, I think you're thinking of wavpack
13:30:04safetydanflac is just lossless from the start
13:30:42jhMikeSFLAC uses prediction + diffs. The higher the comp the better it tries to fit it's predictors to the actual waveform, I believe.
13:30:43nudelynoh, i'm sure i read that flac was lossy+diff... maybe what i read was wrong
13:31:13jhMikeSIt uses Mid/Side stero too
13:32:01nudelyni guess it'd sound pretty awful without the diffs
13:32:36nudelyn(and/or still be large compared to mp3/ogg/etc)
13:32:53safetydanwavpack does alright in hybrid mode
13:33:04jhMikeSI'd like to hear that. It could be argued that the predictors are lossy becuase they are not perfect, hence the differ.
13:33:04 Quit mkey (""Welcome to IRC; Where men are men, women are men and little girls are FBI agents!"")
13:35:38jhMikeSIs my player getting better bat life or did someone change the batt scale?
13:36:49safetydanhrm... really need to stop leaving the nokia charger and the iriver charger plugs next to each other...
13:37:29*jhMikeS hears sizzling and smells smoke coming from safetydan's way
13:38:29safetydannah that's just my brain after reading some more astronomical algorithms stuff
13:38:31 Quit safetydan ("Ex-Chat")
13:42:00 Join lee-qid [0] (
13:47:08JdGordonwho has an idea how to kill some time?
13:48:11 Quit fejfighter ()
13:49:33JdGordon$ ./rockboxui &
13:49:33JdGordon[1] 2050
13:49:33JdGordonfatal: No I/O port permissions
13:50:05Mikachutry without the &
13:50:18JdGordonsame thing
13:50:21 Join Paul_The_Nerd [0] (
13:50:30Mikachudunno then
13:51:24JdGordon:'( sdl is installed... stupid bloody ubuntu
13:52:55JdGordonwow that was wierd :p ran it as sudo and it went fullscreen and very screwed up
13:53:12Mikachuoh heh
13:53:17Mikachumaybe sdl tries to use svgalib or something funny
13:56:39crwltry export SDL_VIDEODRIVER=x11 or something like that
13:58:12JdGordonX is using mergedFB if that makes a difference
13:58:26JdGordoni havnt got my dualhead setup properly yet
13:59:16JdGordoncrwl: fatal: No available video device
13:59:45 Quit XavierGr ()
14:00:26crwlJdGordon, sounds pretty strange
14:00:34crwli don't have the simulator installed, though, but still
14:01:02crwlI "fatal: No available video device" if I run SDL app with SDL_VIDEODRIVER=svgalib, for example
14:01:08crwlwith SDL_VIDEODRIVER=x11, everything works
14:01:46JdGordonhmm.. this sux
14:01:50crwland I'm running ubuntu too
14:07:14 Join adiamas [0] (
14:08:39 Part Paul_The_Nerd
14:28:22aliaskJdGordon: If you're that bored, you can get scrolling working on the lists in my plugin :D
14:28:45JdGordonscrolling in lists is buggered full stop!
14:28:59JdGordonits supposed to work up to 600odd chars but crashes after 90odd
14:29:04 Join MarcoPolo [0] (
14:29:20aliaskOh - so it could just be the string I was using... *goes to test*
14:32:25aliaskNope, it works in the text_editor plugin, but not mine - funny thing is: I copied the list code FROM there...
14:32:42JdGordonhow long is the line?
14:32:56aliaskOnly just off screen using nimbus12
14:33:03aliaskI'll actually count now.
14:33:24aliaskI'm pretty sure my list code must be screwed
14:34:54JdGordonif you paste it somewhere i can look.. but if u just copyied it from text_editor u should be ok
14:35:40aliask is what I have at the moment
14:37:36 Quit adiamas ("Chatzilla 0.9.75 [Firefox]")
14:37:50aliaskThere's a flag in gui_synclist_init called "scroll_all" in the header file, but setting it true doesn't help, and on top of that text_editor sets it false, and has scrolling working.
14:38:00JdGordonrb->gui_synclist_init(&properties_lists, &get_props, file, false, 2); <- shouldnt that be get_props without the & ?
14:38:31aliaskPossibly, that part's stolen from the id3 function (yes, it's all stolen code)
14:38:51aliaskLet's see what happens.
14:39:23JdGordonalso, i think your supposed to actually copy the text into the buffer, and return the buffer
14:39:30JdGordonthat could be a problem also...
14:40:16JdGordonwhats your plugin do? looks interesting
14:41:27aliaskBasically at the moment it just gets the filesize
14:41:43aliaskI was supprised there was nowhere in rockbox that actually did it.
14:42:08*JdGordon wanted to do something like that for folders and files
14:43:11aliaskFolders might be a bit harder - I thought of your dirwalker function though.
14:43:36JdGordoni wanted to add it to stats, but havnt been bothred...
14:43:53aliaskWhat, pass stats a folder or something?
14:43:59JdGordonye, but you cant
14:44:03JdGordonso i havnt
14:44:20*JdGordon doesnt know if dirwalker will ever make it into cvs.. but it would make life easier
14:44:32aliaskDid it actually end up reducing code size?
14:44:58JdGordonna, tiny increase.. but i rekon its worth it.. coz it makes coding easier
14:45:35 Join Paul_The_Nerd [0] (
14:46:40JdGordonktorrent is a horrible BT client :p
14:47:05aliaskDon't you use torrentflux?
14:47:24aliaskOr did I see that somewhere else?
14:47:35JdGordoni did, but i havnt reinstalled it after format
14:48:12aliaskOnly thing it's missing is priorities... I started using it after seeing it on your server :)
14:49:05aliaskDitching the & didn't help, but it didn't stop it working, and it saves 1 byte in the code, so out it stays
14:52:26JdGordonodd... i would havee thought that was a pointer error...
14:52:36JdGordontry actually copying the return strings into the buffer
14:52:55aliaskIn which function?
14:53:22 Join lostlogic_ [0] (
14:53:24 Quit lostlogic_ (Read error: 104 (Connection reset by peer))
14:56:23aliaskYeah, sorry, stupid question :p
14:57:41JdGordonalso, you should cache the filesize.. because that will recheck it every time you move in the list, but the file will stay the same size so there is no point doing that
14:58:07aliaskI just noticed that then, wondering why it was lagging (just after boot) every keypress.
14:59:45 Part Paul_The_Nerd
15:01:07aliaskOk, sticking that stuff in buffer didn't really do anything. I'm giving up for the night. School tomorrow.
15:01:43aliaskYay school :(
15:02:02JdGordonye, same, gotta leave in 12hrs 30min :p
15:02:18aliaskDamn you.
15:02:39aliaskWhich uni you at?
15:03:00aliaskI was there today, eating free saussages at open day :D
15:03:12aliaskI'll probably be there next year.
15:03:12JdGordonyour in yr12?
15:03:23JdGordonyour nuts.. take the year off
15:03:50 Join lamed [0] (
15:04:07aliaskThats still an option, I can just defer if I get in
15:04:18JdGordonwhat u want to get into?
15:05:22aliaskJdGordon: Science/Engineering double, or just engineering.
15:05:48aliaskhello lamed
15:06:33***Saving seen data "./dancer.seen"
15:07:30JdGordonwell good luck for it
15:08:05aliaskThanks. I better head off anyway - g'night
15:08:13 Quit lodesi ("leaving")
15:08:20 Quit aliask ("sleep")
15:16:59lamedanyone else is having trouble with cvs update?
15:17:34JdGordonwhat sort of trouble?
15:17:50lamed...this is weird. I can download a fresh repository but I can't update or status.
15:20:18lamedit just stands and do nothing.
15:21:49 Join damaki__ [0] (
15:22:39amiconndionoea: around?
15:23:16 Join webguest17 [0] (
15:23:16 Quit lamed ("CGI:IRC (EOF)")
15:23:24 Join lamed [0] (
15:24:43 Quit lamed (Client Quit)
15:24:43webguest17how do I tell grep to use a list of files made by say ls?
15:24:51 Quit webguest17 (Client Quit)
15:24:56 Join lamed [0] (
15:25:01lamedhow do I tell grep to use a list of files made by say ls?
15:25:11 Quit Nibbler ("Leaving")
15:25:47 Join lodesi [0] (
15:25:55 Join pike [0] (
15:26:05crwldo you mean something like ls | grep something
15:26:38lamedbut ls |grep something will cause grep to work on the filename, not on it's content.
15:27:12crwlgrep something `ls something`
15:27:13Mikachulamed: ls|xargs grep hi
15:27:21crwlgrep something *
15:27:33crwlsome greps probably don't take more than one filename argument
15:28:00crwlxargs helps then
15:28:42lamedmikachu always understands me :)
15:29:08 Join Nibbler [0] (
15:30:02Mikachui am of course assuming the command isn't just "ls"
15:30:17lamedthe idea was to run grep over grep
15:30:43lamedi guess there"s a simpler syntax for that as well
15:35:31 Quit Nibbler ("Leaving")
15:35:31 Quit lamed ("CGI:IRC (EOF)")
15:40:22 Quit damaki_ (Read error: 110 (Connection timed out))
15:42:38 Quit Poka64 ("bye for now")
15:45:24 Join lamed [0] (
15:47:10 Join sven_ [0] (
15:47:51lamedI think I found my cvs update problem. how can I make sure i'm logged in using my cvs name?
15:49:38lamedsomeone please change the topic to 'No answering question day'
15:50:14Mikachuor maybe change it to 'Sunday 4am'
15:51:55lamedit's 5 pm here :) and btw, ususally I join in in about 2-4 am, which would be like what? 1-3 pm in your time, and still not many people are around answering :D
15:53:36Mikachuuh oops, 4pm
15:54:30Mikachui hate am/pm
15:54:49lamedi hate cvs update
15:59:39 Quit sven_ (Read error: 54 (Connection reset by peer))
16:02:25 Join sven_ [0] (
16:05:00S0apI hate time zones.
16:08:34lamedda&%n it! I can update one repository but not the other.
16:09:11 Join dropandho [0] (
16:15:17 Quit damaki__ (Read error: 104 (Connection reset by peer))
16:19:07 Quit joe2 (Read error: 110 (Connection timed out))
16:19:20 Join damaki [0] (
16:23:47 Quit damaki (Remote closed the connection)
16:23:49 Quit lodesi ("leaving")
16:24:26 Quit barrywardell ()
16:28:13 Join barrywardell [0] (n=barrywar@
16:29:34 Join damaki [0] (
16:34:38 Quit Mikachu (Remote closed the connection)
16:35:22 Join lodesi [0] (
16:37:55 Quit damaki (Read error: 104 (Connection reset by peer))
16:38:58 Join damaki [0] (
16:39:07 Join Mikachu [0] (
16:39:41 Quit damaki (Read error: 104 (Connection reset by peer))
16:46:37 Join gapz [0] (n=gapz@
16:49:41 Part gapz ("init 0 - ||")
16:54:34 Quit pike ()
16:54:46 Join pike [0] (
17:04:28 Join Criamos [0] (
17:06:34***Saving seen data "./dancer.seen"
17:06:59 Quit Mikachu (Remote closed the connection)
17:08:58 Join Mikachu [0] (
17:30:56 Join Hansmaulwurf [0] (
17:47:33 Quit humulus ("leaving")
17:48:22 Join thegeek_ [0] (
17:48:22 Quit thegeek_ (Read error: 104 (Connection reset by peer))
17:48:25 Quit thegeek (Read error: 104 (Connection reset by peer))
17:48:36 Join thegeek [0] (
18:08:27 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
18:09:49 Quit powr-toc (Remote closed the connection)
18:10:27 Quit RedBreva (Read error: 54 (Connection reset by peer))
18:12:43 Join Psilonaut [0] (
18:15:13 Quit Criamos ("( :: NoNameScript 4.02 :: )")
18:18:01 Join Criamos [0] (
18:24:15 Quit lamed ("CGI:IRC (EOF)")
18:48:32 Quit mikearthur (Read error: 54 (Connection reset by peer))
18:52:22 Join dan_a [0] (n=dan-mirc@
18:59:02 Join sharpe [0] (
19:02:47 Join webguest67 [0] (
19:03:02 Nick webguest67 is now known as holyshit (
19:04:22holyshitI have some question..
19:04:37holyshithello~ any body~?
19:05:25 Quit barrywardell ()
19:05:32S0apask the question
19:05:54holyshitI have a iriver ifp-999
19:06:05dionoeaamiconn: ?
19:06:18 Join barrywardell [0] (n=barrywar@
19:06:37***Saving seen data "./dancer.seen"
19:06:40 Quit holyshit (Client Quit)
19:20:35 Join petur [0] (n=petur@rockbox/developer/petur)
19:27:53 Join eGen [0] (
19:29:29 Join LandusMikain [0] (
19:30:16 Quit Landus (Nick collision from services.)
19:30:22 Quit LandusMikain (Client Quit)
19:30:57 Join Landus [0] (
19:32:22 Quit Landus (Client Quit)
19:32:39 Join Landus [0] (
19:33:41*linuxstb watches some MPEG video on his ipod
19:34:33dionoeaw00t :)
19:34:45dionoeahow many fps do you have ?
19:34:58dan_alinuxstb: Well done!
19:35:04 Quit eGen_ (Read error: 113 (No route to host))
19:35:04linuxstbI wouldn't go as far as w00t...
19:35:30linuxstbIt's around 9fps - but that's just a first implementation with no optimisations, not even any use of IRAM.
19:35:38 Join bmac2 [0] (
19:35:57dionoeasounds great :) is it based on the work from "the other guy who's name i forgot" ?
19:36:21linuxstbBoth MPEG-1 and MPEG-2 seem to perform the same, which seems odd to me, but I don't know the details of what differs between the two.
19:36:42linuxstbNo. I took the same library he tried (libmpeg2) and ported it myself, rather than using his port.
19:37:25dionoeaif you need help on libmpeg2 you can ask sam and Meuuh on freenode. They're the current libmpeg2 mantainers
19:38:29dionoeaThey usually hang arround in #videolan
19:38:50linuxstbI'm just starting work with it. But I'm sure it's worth asking them for optimisation tips at some point. Although there are some functions with ASM versions for various platforms (X86, Sparc), so the obvious place to start would be ARM (and Coldfire) versions of those.
19:39:36midgey34dionoea: do you plan on changing the numbers in solitaire back into color or leaving them as black?
19:40:02dionoeai plan on changing those :)
19:40:17midgey34all right :)
19:42:02dionoeamidgey34: i was thinking about using stuff like for color screens
19:42:08dionoeai don't know if it's a good idea though
19:42:16bmac2is there a way I can set my ipod to automatically choose rockbox on boot instead of the apple software? I have ipodlinux and the bootloader installed, and everything works great. Just when it shuts down, I have to manually choose rockbox
19:42:37midgey34oh those look nice
19:42:45bmac2is that a function of the ipodlinux boot loader?
19:43:01dionoeamidgey34: the only issue is finding out if they look good on really small screens :p
19:43:17linuxstbbmac2: Yes. You can use a config file to set a default firmware and a timeout - check the wiki.
19:43:19dionoeaand i think that i'll also move the card drawing routines to a plugin lib so that other plugins can use it
19:43:29bmac2ok linuxstb thanks
19:43:44midgey34well, I renamed the solitaire_numbers etc to card_*
19:44:05midgey34since they overlap in blackjack, but I used a different cardback
19:47:30dionoeamidgey34: i don't know when you last had a look at the solitaire source code, but i changed the display stuff quite a bit this week to have simple card drawing functions. The idea would be to have a lib exporting these card drawing functions (the plugins wouldn't have any knowledge of the graphics themselves)
19:53:15amiconndionoea: There is still the rather strange bug with vanishing cards in solitaire
19:53:54dionoeathe loop vanish thing ?
19:54:06 Quit Ribs (".")
19:54:41 Join Ribs [0] (
19:54:42amiconnA new observation is that when starting with a fresh stack (at start of game or when "wrapping" around), you have to draw twice in order to see card(s) drawn
19:55:24amiconnWhat's interesting now is that the first draw actually draws a card, but this card is invisible(!)
19:55:31dionoeathat's weird. I'll fix it ... once my solitaire.c compiles again
19:56:04amiconnYou can put this card elsewhere if it matches, which causes the card to become visible
19:56:35amiconnIf this is possible and you put this invisible card away, the next drawn card is again invisible...
19:56:56dionoeacould you provide a simple key sequence so i can reproduce this ?
19:59:03Slasherilinuxstb: hi, would you like to have the newest bootloader? :)
19:59:26Slasherijust a moment
20:00:31linuxstbamiconn: Did you write the yuv_bitmap_part() function in the jpeg viewer?
20:03:05 Join leftright [0] (
20:05:48 Join Daishi [0] (
20:06:17 Quit pike (Read error: 110 (Connection timed out))
20:12:31Slasherilinuxstb: and rockbox.iriver
20:12:42Slasheriapply that patch to the latest cvs (revert the previous patch first)
20:12:55Slasherithat new bootloader uses the eeprom memory
20:13:41Slasheriups, bootloader.iriver
20:14:51leftrightSlasheri, what are the advantages of this eeprom bootloader
20:15:47Slasherileftright: putting rockbox to flash makes bootup a little faster and it allows for example dircache and tagcache to be serialized on the disk, so the last state can be loaded immediately thus saving some battery on bootup
20:16:37Slasheriuusing the eeprom configuration memory makes it possible for bootloader and rockbox to communicate with each other
20:17:02leftrightthanks :)
20:17:14Slasheriso bootloader can tell for example if it's safe to assume the disk is still intact from changes
20:17:43crwlwhen doing a battery benchmark, should the headphones be connected? does it make a noticeable difference?
20:17:59leftrightso changes can be done to the bootloader from disk ?
20:18:51Slasherileftright: no, bootloader still needs to be flashed (it can be now updated inside rockbox)
20:19:05 Part bmac2 ("Leaving")
20:19:14leftrightah so no need to reflash it then ?
20:19:38linuxstbSlasheri: eb6204a140f7c1ee5e0b61135eab84f2 bootloader.iriver ?
20:19:41Slasherihmm, what do you mean with reflashing?
20:20:01Slasherilinuxstb: seems to be correct
20:21:06leftrightonce your botloader has been installed, future changes to bootloader can be done without reflashing with a new bootloader
20:21:49Slasherileftright: hmm, not really. bootloader always needs to be flashed if there is a need to upgrade it
20:22:00Slasheribut the flashing can be done inside a rockbox conviently
20:22:20leftrightactually ignore me, i'm just shooting the breeze here, :>
20:22:26Slasherihehe :)
20:23:07linuxstbSlasheri: Seems to be working nicely.
20:23:19Slasherinice :) are you using dircache?
20:23:20amiconnlinuxstb: yes
20:23:27Slasheritry to boot twice
20:23:43Slasherithen you should see it's no more scanning the disk
20:24:40linuxstbamiconn: I'm testing it my libmpeg2 based video player - I wanted to ask you what the csub_x and csub_y parameters are for. It seems to work if I set them both to 2.
20:25:16Slasherilinuxstb: with current bootloader it's even possible to configure an eeprom setting so bootloader can load firmware by default either from disk or from flash
20:25:18amiconnThese parameters indicate the chroma subsampling
20:25:34Slasheribut i haven't added that "bootloader configuration" menu to rockbox yet
20:26:20Slasherii was even thinking that as we have plenty of flash, user could put two images in the flash. Other could be ram image and the other a ROM image
20:26:20amiconnu and v bitmaps can have a lower resolution than the luma bitmap
20:26:31linuxstbSlasheri: Could it somehow work automatically if the Rockbox version on disk is newer than the one in flash. And maybe the flashing plugin could set the "boot from flash" flag when it flashes a firmware.
20:26:49amiconnjpeg allows various combinations, so these parameters are variable
20:27:25Slasherilinuxstb: hmm, i don't know how to do that because bootloader has already gone when disk is still spinning up
20:27:40 Quit lee-qid ("Trillian (")
20:28:03dionoeampeg2 usually is 4:1:1
20:28:30Slasherilinuxstb: but maybe rockbox could change that flag to boot from disk if it's found current flash image is older than the rockbox.iriver available on disk
20:28:47dionoeaerr ... 4:2:0 sorry
20:28:52Slasherihowever, .iriver -files don't have a version flag included
20:29:30leftrightSlasheri: I delete songs frequently from albums and resync often with my pc, will that cope
20:30:09Slasherileftright: yes, that is why eeprom support is necessary so even if you use bootloader usb mode, rockbox will know that you have used it
20:30:21Slasheribecause now bootloader and rockbox can communicate
20:31:13linuxstbamiconn: My MPEG test files are 224x176, and the chroma bitmap seems to be 112x88. So am I right in thinking csub_x and csub_y should both be 2? (it at least looks like it's working)
20:31:44Slasheriamiconn: btw, what do you think about the firmware_settings struct in firmware/export/eeprom_settings.h?
20:31:54 Join JosM [0] (
20:32:03JosMhi all
20:32:18amiconnI didn't look at it at all yet
20:32:31JosMcan I ask a question about flashing?
20:32:51leftrightjust ask
20:33:24JosMwell, the link in the documentation to the files doesn't seem to work anymore. Where can I get the files? (Rec V1)
20:35:15dropandhois there anywhere on the wiki that i can read what this EEPROM support is all about?
20:35:31Slasheridropandho: hmm, not yet
20:35:46dropandhook- good...i wanted to make sure i wasn't loosing my mind- thanks!
20:35:49JosMaddendum, on
20:35:52 Part amiconn
20:35:55Slasheridropandho: but i am sure some information will be added there
20:36:02dropandhogreat- thanks guys!
20:37:04Slasheridropandho: the code is just not yet ready for regular users to start using it
20:37:12peturSlasheri: did you include debug code to read the eeprom?
20:37:23Slasheripetur: yes, should be in cvs
20:37:28*petur is looking at eeprom support for h300
20:37:36Slasheripetur: the rom dump feature should now dump eeprom contents also
20:37:40Slasheri(but only on H1xx)
20:38:05peturalready enabled eeprom here ;)
20:38:10Slasherihehe :)
20:38:30peturgood you wrote the driver, I seem to have deleted my half-working version :(
20:38:45Slasherioh :/
20:39:02peturwas in some test build directory which I cleaned up some time ago... grrrr
20:39:07Slasherihmm, but the H300 version was a bit different with hardware I2C-bus?
20:39:23Slasheriwith H100 that bus was implemented by software
20:40:01peturafaik it's on the same bus as the pcf
20:40:14Slasheriand the H300 has also bigger chip (more than 128 bytes)
20:40:24Slasheriok, so it needs a different driver then
20:40:44 Quit barrywardell ()
20:41:01Slasheribut maybe you can get some clue from the working eeprom code and the pcf driver
20:41:18Slasherijust combine those and basic single byte read/write operations should work :)
20:41:32Slasherii haven't yet implemented page write / sequential read for better performance
20:41:33JosMhmm, it seems that all the download links on don't work...
20:42:52leftrightwhere's Badger
20:43:44 Quit Bger ("brb")
20:44:40 Join [1]leftright [0] (
20:46:57 Part [1]leftright
20:55:24 Join gtkspert_ [0] (
20:55:58 Quit MarcoPolo (Remote closed the connection)
20:58:34dionoealinuxstb: gibalou on #videolan suggested having a look at . It's a pocket pc media player and it features arm optimisations.
20:58:38dionoeaIt's also GPL
20:59:03linuxstbDoes that use libmpeg2 then?
20:59:08dionoeai doubt it
20:59:30dionoea20:57:59 < gibalou> dionoea: yeah, it is a pocketpc player and most of these use arm processors
20:59:33dionoea20:58:09 < gibalou> it actually is the best pocketpc player around
21:00:00 Join Poka64 [0] (
21:00:21dionoeathe only issue is that you have to email the guy to get the source code ... but that shouldn't be too complicated
21:00:32linuxstbI've already got the source - google found it for me.
21:00:48dionoeahave fun :p
21:00:50linuxstbIt's GPL, so once one person gets the source...
21:01:40 Quit leftright (Read error: 110 (Connection timed out))
21:04:19 Quit Hansmaulwurf ("( :: NoNameScript 3.81 :: )")
21:05:07peturSlasheri: I wished you didn't couple eeprom driver with eeprom settings. Or does it keep out of the eeprom if not flashed?
21:06:39***Saving seen data "./dancer.seen"
21:14:10 Quit gtkspert (Read error: 110 (Connection timed out))
21:16:33Slasheripetur: yes, eeprom is not touched unless rockbox detects firmware flashed
21:17:05peturdoes weird things here
21:17:14peturmust enable step by step
21:17:20Slasheriwhat kind of things?
21:17:24 Join TomFelker [0] (
21:17:55peturthe init does something bad with the pcf communication
21:18:20Slasheriah, well that software i2c driver cannot work with the hardware driver
21:18:22peturIt jumps to a quicmenu and battery indicator goes to zero
21:18:48Slasheriso you need to replace all of the sw_i2c_* -functions with correct code taken from the pcf driver
21:20:06peturpcf50606 doesn't use hardware I2c it seems
21:20:12Slasherihmm, interesting
21:20:24Slasheriwell, then it might work
21:20:37peturbut there's stuff running in other tasks - will have to prevent that
21:20:44Slasheribut there is no need to init the i2c bus twice.. so just remove the init code from the eeprom driver
21:21:18*petur slaps himself hard for deleting his code
21:21:36Slasheriand you don't need the eeprom_settings.c for testing the driver
21:22:39peturyes, I'm that far already, just creating myself a debug screen to test some stuff
21:22:47Slasherinice :)
21:38:22 Join Bger [0] (n=bager@rockbox/developer/Bger)
21:41:05 Join Massa [0] (n=Massa@
21:44:18 Join pike [0] (
21:45:35dionoealinuxstb: 21:07:07 < Meuuh> dionoea : AFAIR BBPlay had an ARM-optimized YUV to RGV
21:47:00linuxstbdionoea: OK, thanks. I don't think the yuv2rgb conversion is much of a bottleneck though.
21:47:56 Join gtkspert [0] (
21:49:27dionoealinuxstb: 21:07:13 < Meuuh> which helped a lot actually
21:49:37dionoeayou do yuv to rgb for every frame
21:49:53dionoeathat's quite a lot
21:50:19petur30 times per second?
21:50:31dionoeait's one 3x3 matrix time one 3 item vector for each pixel
21:50:42dionoeapetur: 25 or 29.97 times per second yes :)
21:51:02Mikachuin mplayer, they bothered to write a yuv2rgb function in the pixel shader in the gl output driver
21:51:29dionoeaunfortunately i doubt that the ipod has such hardware :)
21:51:42Mikachuyes just pointing out that it might make a difference
21:51:58zealso the video output methods that mplayer can skip conversion with tend to be the fastest (afaik)
21:52:01dionoeamost x11 / opengl /directx graphics card indeed handle the conversion themselves
21:53:41 Join sankao [0] (
21:57:22 Join sharpe` [0] (
21:57:32 Quit sharpe (Read error: 104 (Connection reset by peer))
22:01:33linuxstbMy test file is now playing at around 11.3fps. If I disable the actual display of the frames (the yuv2rgb conversion, and copying to the framebuffer), then it's running at 20fps. So I guess that makes it a bottleneck...
22:03:28dionoeawhat's the ipod's LCD's max framerate atm ?
22:03:36linuxstbSo I think an asm optimised "yuv_blit" function which does the conversion from yuv2rgb and writes the data directly to the lcd hardware is the way to go.
22:03:44*petur whispers a little w00t
22:05:02linuxstbThis is on my Photo - test_fps is giving me 48.5fps
22:05:41 Quit gtkspert_ (Read error: 113 (No route to host))
22:05:51linuxstbSo with further optimisation of the decoder itself, 25fps doesn't look completely unrealistic. It would be even easier on the Nano's smaller LCD.
22:06:05dionoeasounds great :D
22:06:31linuxstbAnd that leaves the COP to take care of reading the data from disk and decoding the audio...
22:06:39linuxstb(or vice-versa)
22:06:43dionoeayou're decoding LCD size video ?
22:06:45 Quit netmasta10bt ("My damn controlling terminal disappeared!")
22:06:51 Join hexreel [0] (
22:06:58linuxstbWell, 4 pixels wider than the LCD - 224x176
22:07:36 Quit gtkspert (Read error: 110 (Connection timed out))
22:07:56dionoeais you fps_test plugin available for download somewhere ?
22:08:08*dan_a gets confused by pointers
22:08:31linuxstbIt's in CVS - but you need to add it to apps/plugins/SOURCES to compile it.
22:08:41dionoeaok, thanks
22:09:29*linuxstb encodes a Nano-sized file to test
22:09:30dan_aIf "const char *thread_name[MAXTHREADS];" works fine, why doesn't "const char *thread_name[NUM_CPUS][MAXTHREADS];"?
22:10:00 Join gtkspert [0] (
22:12:24dan_aAh, found it!
22:12:24 Join Paul_The_Nerd [0] (
22:16:42 Join MarcoPolo [0] (
22:18:02 Join barrywardell [0] (n=barrywar@
22:18:57linuxstbGood news for Nano owners - without rendering, the decoder is managing about 30fps for a 176x128 file on my Photo. With the existing unoptimised rendering, it's already about 18fps.
22:20:58midgey34linuxstb: iriver only manages 10fps on the h300 so your 11.3 on the photo is already impressive
22:21:17linuxstbThis is without sound...
22:21:45midgey34good point
22:22:14linuxstbi would be interested to see how it runs on a 124MHz coldfire compared to a 75MHz ARM.
22:26:36 Quit obo ("bye")
22:29:42Bagderit seems I might get a package tomorrow
22:30:27dan_aBagder: Oooh, a Sansa shaped package?
22:30:46dan_aWith JTAG?
22:30:52Bagdersupposedly, yes
22:30:55*linuxstb wonders if they've accidentally dropped some documentation in the box...
22:31:06peturdream on
22:31:17dionoeawhat cpu does the sansa use ?
22:31:21midgey34Badger: that didn't take very long
22:31:23Bgerlinuxstb: "accidentally" ...
22:31:26Bagderdionoea: PP5024
22:31:40dionoeahum ... documentation would indeed be nice :)
22:31:43Bagdermidgey34: UPS express apparently
22:31:51midgey34how nice
22:32:07linuxstbdionoea: It's similar to the PP5020/1/2 used in the newer ipods, but with some extra hardware (e.g. the audio codec) and minus some hardware (e.g. an ATA controller)
22:32:57linuxstbAny H300 owners want to test my mpeg player?
22:33:18Spidalinuxstb: yes
22:33:23Bgerlinuxstb: yes :)
22:33:27midgey34I can do it, but I have no way of making a test file
22:33:29linuxstbYou'll need an MPEG-2 "elementary stream" the same size (or smaller) than your LCD.
22:33:38dionoea(/me asks a question while many people are on the chan) has anyone already tried having fun with the ipod deck's infrared port ?
22:33:39linuxstbI suppose I could upload a test file as well...
22:34:03Bgerlinuxstb: it would be easier if you put such file, yes ...
22:35:02linuxstbOK, I'll encode a short test file now.
22:35:21Bgeror if you could tell mencoder opts .. :)
22:35:36S0apcan your decoder handle MPEG-1 in its current state?
22:36:09S0apfor the free version of TMPEGenc gives excellent fine-grained control over the creation of MPEG-1
22:36:32SpidaS0ap: which OS is TMPEGenc?
22:36:44 Join pixelma [0] (i=pixelma@
22:36:47Spidaok, doesn't work here...
22:36:56linuxstbS0ap: Yes, it handles both MPEG-1 and MPEG-2
22:36:58S0apthe pay version gives equally fine-grained control over MPEG-2.
22:37:19linuxstbI've been encoding at around 600kbps for 224x176 files.
22:37:45S0apIts encoder is considered a generation behind, but I don't know of another tool which allows the tweaking of so many nuances.
22:37:49 Quit pike (Read error: 110 (Connection timed out))
22:38:26 Quit Massa (Read error: 113 (No route to host))
22:38:37Bgerlinuxstb: patch ?
22:38:51S0apif I did my math right, 600kb@224x176 is even higher than SVCD bitrates.
22:38:52dionoeaS0ap: ffmpeg should allow testing of many nuances :)
22:39:11 Quit Poka64 ("XChat 2.6.6 -")
22:39:57S0apis there reason to believe a bitrate closer to 400 would give an apreciable increase in decoding framerate?
22:40:08 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
22:40:08linuxstbS0ap: Yes, it probably would.
22:40:30 Join mikearthur [0] (
22:41:06linuxstbPlugin is here:
22:41:27linuxstbExtract it in apps/plugins and then add "mpegplayer" to apps/plugins/SUBDIRS
22:42:12linuxstbAlso add the line ".m2v,viewers/mpegplayer.rock" to apps/plugins/viewers.config (or whatever extension you want to use)
22:42:51amiconnlinuxstb: Btw, yuv_bitmap_part() isn't optimised for speed
22:43:36amiconnWell, in fact it is, but I think there's room for more
22:44:14linuxstbI was going to ask you about that. What do you think of the idea of a "yuv_blit" function?
22:44:32amiconnIt does several divisions and modulo operations per pixel row
22:44:45 Quit JosM ()
22:45:16amiconnProvided we only need to support subsampling by powers of 2, these could be replaced by shifting and masking
22:45:54dionoeawell you could use only 1 sort of subsampling.
22:46:16amiconndionoea: Not for jpeg.
22:46:20dionoeaas 99% of mpeg2 files use 4:2:0
22:46:47dionoeaamiconn: well the mpeg2 decoder would have an optimised version i guess
22:47:02amiconnCould be done, yes
22:47:30Bgerlinuxstb: sh.t
22:47:44Bger/home/Bager/Rockbox/clr/rockbox-devel/apps/plugins/mpegplayer/header.c:364: undefined reference to `memcpy'
22:47:47S0apis this MPEG-2 SP or MP?
22:48:15midgey34Bger: I got the same
22:48:26Bgersome kind of var init...
22:48:34linuxstbI guess that's gcc needing memcpy...
22:48:48linuxstbJust add a simple wrapper to rb->memcpy at the top of header.c
22:49:12linuxstbStrange it doesn't complain for ARM.
22:49:13TomFelkerdoes gcc sometimes call memcpy on its own? i was having a problem with that when i tried compiling with -fno-inline
22:49:37Bagderyes it does
22:49:39amiconnTomFelker: Yes it does
22:50:41TomFelker(thanks to gcc3.4 cross compiling being broken and nobody will fix it on amd64, i'm trying to get it to compile on 4.0.1)
22:51:11amiconnTomFelker: You can build m68k-elf-gcc 3.4.x on amd64
22:51:25amiconnIt requires a somewhat hackish patch, but it works
22:51:28 Join blackvd [0] (
22:51:31linuxstbS0ap: I'm not sure what libmpeg2 can handle - but this plugin should support everything libmpeg2 does.
22:51:46TomFelkeramiconn: the patch gets around the ICE while compiling GCC?
22:52:05blackvdsay is there a option in the menu to turn on your album art or should it just show up?
22:52:16Bgerlinuxstb: ok, compiled
22:52:26 Quit MarcoPolo (Remote closed the connection)
22:52:32peturblackvd: needs a patch to work
22:52:34linuxstbBger: Do you still want a test file.
22:52:50Spidawhat happened to the rockbox-optimized patche for h3X0?
22:53:00Bgerlinuxstb: yep :(
22:53:03peturblackvd: or get a custom build in the forum
22:53:18markunlinuxstb: is your plugin based on mirak's patch?
22:53:21Bgerlinuxstb: or params for mencoder
22:53:24blackvdhm guess I'll just get a patch
22:53:26linuxstbmarkun: No, I started from scratch.
22:54:00peturblackvd: and build it yourself...
22:54:12S0aplinuxstb - did your test sample have B frames? Last question, I promise.
22:54:40BagderSpida: there's a whole range of them on
22:54:50linuxstbS0ap: I used mpeg2enc from mjpegtools - I've got a feeling that doesn't create B frames, but I may be wrong.
22:55:16markunlinuxstb: what kind of framerate are you getting?
22:55:52midgey34linuxstb: a test file would be helpful over here
22:56:49Paul_The_NerdSpida: Be aware, they aren't actually "optimized" but really "A whole bunch of patches that may or may not work consistently together".
22:56:55dionoeadid you try compiling with -Os ?
22:57:00amiconnlinuxstb, jhMikeS: Currently, lcd_blit() is for low-depth LCDs only because on the others there was no need for it. There's nothing stopping us from adding it
22:57:27amiconnI would try to stay away from 18bit though. More data for little benefit, plus either alignment problems or even more data
22:57:28SpidaBagder: doesn't resolve for me
22:57:35linuxstbmarkun: With rendering the decoded yuv data to the LCD, about 11.3fps. Without that rendering (and copying to the framebuffer), about 20fps.
22:57:53linuxstbmarkun: That's with a 224x176 mpeg2 file on my ipod Photo (and no audio)
22:57:57amiconnlcd_blit() could do 16->18bit on the fly the same way as lcd_update()
22:57:58SpidaPaul_The_Nerd: yeah, I know them, and had them working for some time.
22:58:03BagderSpida: then your dns plays tricks on you
22:58:10amiconnI will convert that stuff to asm as soon as I receive my X5
22:58:39linuxstbamiconn: I'm thinking about an optimised yuv_blit - to do the conversion and lcd_update() together.
22:59:30amiconnlinuxstb: A related question that came to mind - do you think it would be useful to add a little on-the-fly processing to the color ipod lcd updates as well?
22:59:51amiconnThen we coul duse the standard RGB565, and the lcd update would do the byte swap
23:00:12amiconnBut unlike the X5, it means a slight performance hit
23:00:52linuxstbMaybe it wouldn't be a performance hit - the lcd_update() function seems to need to wait after performing each write.
23:00:55Bgerlinuxstb: what happened with te test file ? :(
23:01:24linuxstbBger: Uploading...
23:01:27Bgerah, k
23:01:27amiconnlinuxstb: The question is whether it would be enough of an advantage to have the standard RGB565 on ipod
23:01:30linuxstbETA 3 minutes.
23:02:11amiconnFor the lcd drawing code it makes no difference at all
23:03:12amiconnX5 is a different story; for half-decent performance it would be necessary to make fb_data 32bit in order to use 18bit colour, and even then it woould only be pseudo-18bit
23:04:48linuxstbI can't think of any immediate advantage - LCD_RGBPACK doesn't seem to be called that frequently.
23:05:17linuxstbBg3r: (about 15MB)
23:05:35linuxstb(apologies for it being a cheesy pop video)
23:06:40***Saving seen data "./dancer.seen"
23:07:45Bgermencoder input_file -vf scale=224:176 -of rawvideo -ovc lavc -lavcopts vcodec=mpeg2video -o output.m2v <= this should work
23:07:54Bger(didn't tried it)
23:08:14amiconnjhMikeS: As you were talking about device performance and video support earlier: Our targets split in 2 categories, (1) hwcodec. (2) swcodec
23:08:18linuxstbBg3r: Maybe...
23:08:27sankaocan someone confirm that in arm 7 asm ldrned means load double words register if not equal ?
23:08:40amiconnAll currently swcodec targets have similar cpu power, so they can all get compressed video support
23:09:41amiconnHwcodec (archos) already does have video support, uncompressed b&w frames using temporal dithering to produce greyscale, and mp3 audio decoded by the hwcodec as normal
23:09:50Bger Don't panic. If you find one, report it to us, <= haha
23:10:04linuxstbsankao: I didn't think ldrd/strd was supported on arm7 - it's an Architecture 5 extension
23:10:28sankaoreally ?
23:10:36linuxstbarm7tdmi is ARMv4
23:10:57amiconnThe data rate needed for this is low enough to make compression unnecessary (480kbps for the video plus the mp2/mp3 audio track which can use all possible mp2/mp3 parameter combinations
23:11:03sankaoit comes from disassembled binary code of the sansa firmware
23:11:27linuxstbAre you sure it's actual code? i.e. not data, and not thumb code?
23:11:39sankaoI thought it was arm7 asm, but I have few knowledge of asm
23:11:51sankaocould be data
23:12:56sankaoit comes just before what the hex editor sees as ASCII
23:13:52amiconnlinuxstb: LCD_RGBPACK for constants doesn't matter at all - all the constant folding is done at compile time. There are even fewer (if any) places which use LCD_RGBPACK() with variables
23:14:06*Bger is silly
23:14:22Bgerbmp viewer didn't like this video :D
23:14:40midgey34linuxstb: what is the line supposed to be in viewers.config?
23:14:40midgey34m2v,rocks/mpegplayer,00 00 00 00 00 00 is what I have
23:14:49linuxstbYes, that will be fine.
23:15:36linuxstbamiconn: Yes, that's what I mean. pacbox uses it with variables, but only at init-time to create a lookup table based on palette data in the ROMs.
23:15:41dionoeahum ... "Undefined instruction at 01FEADC4" :)
23:15:41jhMikeSamiconn: sorry. wasn't here. I'm busy at the moment.
23:15:51Bgershitty OF
23:16:01*petur reads eeprom on his h340
23:16:30*amiconn is still trying to get his head around arm asm :/
23:17:00peturSlasheri: the pcf driver didn't like the state you left the lines in after calling your driver
23:17:35amiconnpetur: Why 2 drivers?
23:17:37Bgerlinuxstb: hm
23:17:46linuxstbBg3r: hm?
23:17:51Bgerruns and exits
23:17:58Bgerafter 2-3 secs
23:18:09Bgerblack screen
23:18:15peturamiconn: will merge... wa just testing with Slasheri's code for now
23:18:15linuxstbIs that with your file, or mine?
23:19:12linuxstbBger: It first of all fills a 25MB file with the input file, and then decodes it. So it seems to be exiting when it starts to decode it...
23:19:21midgey34dionoea: I have a a patch for color numbers in solitaire if you're interested
23:19:37dionoeamidgey34: sure :) i won't have time to look at it now though
23:19:54Bgermidgey34: what happens on your side ?
23:20:35midgey34eh, I can't get it to show up in the open with.. menu
23:20:46midgey34nothing after mp3_encoder is shown
23:21:01dionoea"Undefined instruction at 01FEADC4" here on video ipod :/ (if anybody cares :p )
23:21:44linuxstbmidgey34: Did you add it to the viewers.config file in apps/plugins/, or the version on your player?
23:21:53peturamiconn: my initial code (lost) used part of the pcf driver but didn't work ok. So I first want to use the separate driver and verify the eeprom
23:21:59dionoeacompiling it with -Os it now shows a white screen
23:22:05dionoeais that expected ?
23:22:34Bgerit should be black
23:22:42Bger rb->lcd_set_foreground(LCD_WHITE);
23:22:42Bger rb->lcd_set_background(LCD_BLACK);
23:22:42Bger rb->lcd_clear_display();
23:22:42DBUGEnqueued KICK Bger
23:22:42Bger rb->lcd_update();
23:22:46dionoeait was black for like 10 seconds and it's now white
23:23:05Bgerhah, it seems you went further than me
23:23:24dionoeai'll now try getting colors :D
23:23:31 Part blackvd
23:24:23dionoeamidgey34: thanks
23:24:29midgey34no problem
23:24:50 Join mynimal [0] (
23:25:01mynimalI'm having quite a bit of trouble.
23:25:14CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:25:14*Bger will go to bed
23:25:58mynimalBasically, I'm stuck in an infinite loop.
23:26:31mynimalIt's repeating "Rockbox error: -1" and "Loading original firmware..."
23:26:38Paul_The_Nerdmynimal: And then it reboots again?
23:26:39mynimalI'm assuming the iPod is pretty much screwed.
23:26:44Paul_The_NerdDo you have a 5G?
23:26:46mynimalNope, it's just in that loop.
23:26:51mynimalI have a 60GB iPod Photo.
23:26:56Paul_The_NerdWell, anyway
23:26:59Paul_The_NerdHold Menu+Select
23:27:05Paul_The_NerdAs SOON as the screen clears, hold Play+Select
23:27:11Paul_The_NerdIf you do it right, it will go into disk mode.
23:27:20Paul_The_NerdExtract a daily build onto the player, and it'll be fine.
23:27:39mynimalOh, great, thanks. I must have been pressing the wrong key combination.
23:27:52mynimalThanks a lot, saved me $300 haha
23:28:10Paul_The_NerdThe iPod is pretty much always recoverable.
23:29:15amiconnHmm, bad. ARM effecively has one register less than sh or coldfire
23:30:01 Quit mynimal ("CGI:IRC (EOF)")
23:30:04amiconnBut otoh the instruction set is quite powerful (but hard to memorise)
23:32:02 Quit _FireFly_ ("Leaving")
23:34:08 Join mynimal [0] (
23:34:12 Quit mynimal (Client Quit)
23:37:23 Quit Bger ("Leaving")
23:41:40dan_aYay! I've got a kernel running on the second processor on PP50xx machines!
23:42:21dan_aIt's useless at the moment, but it's still a result
23:44:36 Quit midkay ("Leaving")
23:47:35 Quit Rob2222 ()
23:52:12peturBagder, would it be possible to put a ' between the D and H in my last name in the cvs log?
23:53:05 Join Rob2222 [0] (
23:53:24Bagderlet's try and see
23:57:12Mikachuif you have to commit something to test,

Previous day | Next day