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 2006-09-19

00:01:46Paul_The_NerdI think the Apple one still only clicks when the cursor moves, but I like the idea of it clicking each time the wheel *wants* the cursor to move, since then you know how fast you're actually triggering the wheel. More like physically turning something. Does that make any sense?
00:02:03 Join dpassen [0] (n=dpassen1@resnet-236-163.resnet.UMBC.EDU)
00:02:29dionoeaPaul_The_Nerd: it does
00:03:42preglowi agree completely
00:04:07*amiconn wants adjustable wheel sensitivity, and the hard-wired repeat when keeping the finger on the wheel to be removed
00:07:24liniobo: setting thread priority 4 (same as UI thread) works fine
00:08:13lininow to make a build with unplug + piezo and i can finally go to sleep
00:08:57linino sound skipping nomatter how fast or long i scroll
00:10:00KalthareI figure I'm probably in way over my head, but how would I go about debugging crashes on the ipod mini 1g?
00:10:54Paul_The_NerdCrashes as in "Error message" or freezes?
00:11:17KalthareData abort at hex. I figure for the freezes I'm out of luck.
00:13:23Paul_The_NerdWell, the hex values can be looked up in the .map file for that build (available when compiled)
00:14:18KalthareYeah, it looks like the crash I got just now was in set_cpu_frequency(). But isn't there a way I can get a disassembly or something that shows where in the function it crashed?
00:14:22pixelmaamiconn: still there?
00:14:30 Quit CriamosAndy ("( :: NoNameScript 4.03 :: )")
00:14:50mirakamiconn: I would like acceleration control of the scrolling iriver too
00:15:04linigood night all
00:16:38amiconnmirak: ??
00:18:19obonite lini
00:20:46 Quit petur ("here today, gone tomorrow")
00:23:31 Quit lee-qid (Read error: 110 (Connection timed out))
00:24:40 Quit aegray (Read error: 110 (Connection timed out))
00:25:13mirakamiconn: control the select bar scrolling
00:25:18mirakin folder browsing
00:25:52amiconnWhat so you want to control there?
00:25:55 Quit dpassen1 (Read error: 110 (Connection timed out))
00:26:02 Quit obo ("bye")
00:26:17amiconnI think the acceleration is perfect for button-scroll targets and doesn't need any adjustment
00:26:20miraka setting that let you choose when you want the bar to accelerate and how much
00:26:44amiconnThe ipod scroll wheel is an entirely different thing
00:27:27mirakI think it's too slow to start to accelerate
00:27:31mirakfor me
00:27:54mirakI admmit people have different tastes on that matter
00:28:48 Quit Nibbier (Read error: 104 (Connection reset by peer))
00:30:16 Join RoC_MM [0] (
00:30:33 Quit linuxstb_ ("CGI:IRC")
00:31:39 Quit RoC_MM (Client Quit)
00:36:44 Quit Pyromancer (Remote closed the connection)
00:41:40 Join webguest61 [0] (i=917408a3@gateway/web/cgi-irc/
00:44:45 Quit lodesi ("leaving")
00:47:54 Join Nibbier [0] (
00:48:48 Join Pyromancer [0] (
00:50:09KalthareGenerally speaking, is there a way a newbie like me can help stabilize the mini1g port?
00:59:52 Join italys [0] (
01:01:40 Quit webguest61 ("CGI:IRC")
01:01:57linuxstbKalthare: One useful thing is to find reliable ways to recreate bugs. Being able to easily recreate a crash/freeze is half-way to fixing it.
01:02:07 Join rotator [0] (n=e@rockbox/developer/rotator)
01:02:55italysRockbox is incredible
01:03:32 Quit matsl (Remote closed the connection)
01:03:48Kaltharelinuxstb: Hm. Well, hopefully I can, but I've just been running into random crashes and freezes during playback. :/
01:04:12Paul_The_NerdIt's the CPU boosting bug. Again.
01:04:51KaltharePaul_The_Nerd: Oh. What's known about it?
01:05:05Paul_The_NerdDisabling CPU scaling prevents it.
01:05:14Paul_The_NerdBoosting the cpu by hand prevents it.
01:05:22*midkay hopes playback and config loading are fixed when he upgrades to the latest CVS build..
01:05:29midkaytoday wasn't particularly fun in those aspects. :)
01:05:45Paul_The_NerdAnd as far as I know, it only happens on 1G Mini, 4G Photo and Grayscale.
01:06:11Paul_The_NerdThat's about as much as I've got "reliably"
01:06:27SoapHas a standard way of refering to the two different "generations" of the Nanos been decided upon yet?
01:06:37Paul_The_NerdNano, and Nano 2G
01:06:48SoapI think the wiki needs updated to reflect that all the info refers ONLY to the Nano 1G
01:06:50Paul_The_NerdI guess the old ones could be referred to as 1G nanos now
01:07:09midkaysigh.. i think it was that new.. er.. scheduler.
01:07:13linuxstbI think "1st gen" and "2nd gen" is better - too many people confuse "1G" with "1GB"...
01:07:24midkayconfig loading is broken, and playlists seem to play in reverse order or something..
01:07:43KaltharePaul_The_Nerd: I don't suppose there's anything I can do to help?
01:08:08Soaplinuxstb - that makes sense, should I start changing "Ipod 4G" to "Ipod 4th gen" also?
01:08:37 Quit spiorf (Read error: 60 (Operation timed out))
01:08:52 Quit klrspz ()
01:09:49Paul_The_NerdKalthare: I'm really not sure, honestly.
01:10:30italysi just installed rockbox on my 20gig 4gen grayscale pod
01:11:39KaltharePaul_The_Nerd: Well, I'll see if I can figure something out. I don't expect much, tho −− this is way beyond my ability.
01:11:58 Quit Pyromancer ("Ex-Chat")
01:12:21Paul_The_NerdKalthare: If you can come up with a reliable way to cause the crash, that would be, as linuxstb said, one of the best possible things
01:14:05KaltharePaul_The_Nerd: Okay, I'll poke at it for a while.
01:14:10***Saving seen data "./dancer.seen"
01:19:00 Join Pyromancer [0] (
01:21:11KalthareHmm. I updated Rockbox today, and since then, the wheel's been really unresponsive.
01:22:05linuxstbThat will be the new scheduler...
01:22:46KalthareOh. Does it have crashes, too?
01:22:51midkayit's glitchy, i just had to get an older daily..
01:23:26midkayit causes playlists to be played in reverse, apparently.. as well as breaking config loading.. and when i changed crossfade settings it got stuck on "restarting playback".
01:23:55linuxstbSurely you can't blame the scheduler for everything...
01:24:12midkaythe problems are introduced on that day, and i didn't see anything else that would cause it?
01:24:27KalthareHmm, it freezes if I run doom and then sudoku.
01:24:59preglowsounds really dubious that the scheduler could cause that
01:25:08midkaythere's the scheduler, and then.. a button mapping adjustment.. a bugfix.. and a menu adjustment.
01:25:12midkayit comes in on exactly that day..
01:25:38midkaythis build is the one that causes those things to happen, so.. i'm running the day before now and it's fine.
01:26:20italysis there anyway to remove the apple OS?
01:26:25midkayi'd far prefer the new scheduler, i'd love having slower buttons or LCD updates sometimes instead of cutting out music. but it's not so nice when things are played out of order etc..
01:26:59 Quit Kohlrabi ("Quit")
01:30:29 Quit jhMikeS (Read error: 104 (Connection reset by peer))
01:30:45 Join Gatechy260 [0] (
01:31:33linuxstbitalys: Write this file to your firmware partition instead of the rockboot.bin you created - it doesn't include the Apple firmware and will greatly speed up booting on your 4g:
01:31:52italysI should still be able to mount my device in windows, right?
01:31:53linuxstbitalys: You can also delete the iPod_Control folder if you haven't already.
01:32:23linuxstbitalys: Yes, it doesn't affect the contents of your ipod's flash - which is where the disk mode code is.
01:33:18KalthareWell, I don't know enough to blame any particular code change, but it definitely reliably crashes if I run Sudoku after RockDoom.
01:33:55linuxstbKalthare: Have you set up a development environment and compiled your own version of Rockbox yet?
01:34:04 Join O11 [0] (
01:34:26O11Hmm.. I had a general thought about rockbox as a whole..
01:34:31Kaltharelinuxstb: Yeah, this is one I compiled myself.
01:35:02O11has rockbox ever approached a mp3 manufacturer to ask if they can have details on how the player works and the firmware?
01:35:19linuxstbKalthare: In which case, you could try older versions from CVS and see which change broke it.
01:35:35Kaltharelinuxstb: okay, I'll try that.
01:35:38 Join EV|Toshiba [0] (
01:35:42O11because it seems like some people see the fact that they can put rockbox on their player as an advantage..
01:35:51O11so it could be good for sales..
01:35:57linuxstbO11: Yes. Generally mp3 manufacturer's don't have the info we need - it's the chip manufactuerers, and a growing number don't make their datasheets public.
01:36:12O11and it's not like it's going to make any negative effects on their sales..
01:36:18KalthareOh −− when I built it, it complained that the apps/plugins/zxbox/ directory was missing. What's the story there?
01:36:32 Quit JoeBorn ("Leaving")
01:36:39O11linuxstb: oh.. okay..
01:36:41linuxstbYou need to do "cvs update -dP" - zxbox is a new directory
01:36:44Paul_The_NerdKalthare: Did you download from CVS, or download a source archive?
01:37:02KalthareOkay, that'd be why. I don't know much about CVS, either. ;p
01:37:44 Quit O11 (Client Quit)
01:40:36 Quit TeaSea (Remote closed the connection)
01:41:07Kaltharelogf() doesn't cause problems, does it? I've been enabling it.
01:42:03linuxstbIt shouldn't do, but that could be worth testing.
01:42:42Kaltharelinuxstb: Fair enough.
01:43:26Paul_The_NerdApparently ipod_fw won't extract the firmware from a 5.5G, according to a recent post on our forums.
01:43:34Paul_The_NerdThink maybe the 5.5G is encrypted now too?
01:46:24Gatechy260can anyone help me set up cygwin to be able to use arm-elf-objdump?
01:46:33preglowPaul_The_Nerd: not unlikely
01:46:49 Quit Nibbier (Read error: 110 (Connection timed out))
01:47:35 Join Nibbier [0] (
01:48:44KalthareWhich ones are the 5.5G?
01:49:15Paul_The_NerdThe new Video with brighter screen
01:50:20 Join jhMikeS [0] (
01:50:51linuxstbI don't think ipod_fw would complain if the firmware is encrypted - there must at least be other things different.
01:51:10LandusDoes anyone know if the audioscrobbler plugin will keep logging, even if I restart my DAP, or connect it to a computer?
01:54:14italyshwo do I use tagcache
01:54:18italysI initialied it
01:55:21LandusHave the RAM cache option on.
01:55:34LandusWhen it's done initializing, reboot your DAP.
01:55:52Paul_The_NerdThe ram cache option is not necessary.
01:56:09Paul_The_NerdDigital Audio Player.
01:56:13LandusNo, but if your going to be going through the database, it'll drain battery life less.
01:56:14linuxstband "reboot your DAP" means "shut down cleanly by holding PLAY for a few seconds, then turn it back on"
01:56:31LandusAnd then go to General Settings -> File View -> Show Files and choose ID3 Database.
01:56:35italysI don't see RAM cache option
01:56:45 Quit Landus ("Leaving")
01:56:59scorchewell, just got done reading through ipl's forums to see if they had any tidbits on the 2g nano......all i can say, is that they have *a lot* of idiots with over 600 posts...
01:57:21 Join Landus [0] (
01:57:32Paul_The_NerdLandus: It will only save battery if you browse a lot
01:57:41LandusMan, I hate when I accidently hit Alt - F4.
01:57:47LandusAnyway, it's Load to RAM.
01:57:49italysmy iPod just locked up
01:57:51italyswhat do I do
01:57:57Paul_The_NerdLandus: If you like to set up a long playlist and then just listen to it, then it'll actually make your battery life worse.
01:59:14italyshow do I select RAM option, Landus
01:59:17italysI see Load Into Ram
01:59:18linuxstbscorche: I'm more interested about anyone who's tried IPL on an 80GB 5g
01:59:25LandusThat's it.
01:59:27LandusTurn it on.
01:59:30LandusEither yes or no.
01:59:44scorcheyeah...i was looking for that post...couldnt find it
01:59:51italysI rebooted
02:00:07LandusDid you switch the file view to ID3 Database?
02:00:50italysit's working, thanks
02:03:52italysI need to get a new iPod bat
02:05:11SoapBefore I butcher more of the wiki, could one of the elder gods tell me if they approve of
02:06:25italysgoing forward in a song list goes backwards
02:06:55 Quit [xt] (Read error: 60 (Operation timed out))
02:07:00 Quit Gatechy260 (Read error: 60 (Operation timed out))
02:08:15midkayitalys: use a few days older daily build, something broke it and i'm confident it's the new scheduler.
02:09:16 Join webguest14 [0] (i=47cdcd57@gateway/web/cgi-irc/
02:09:44 Quit mooch (Read error: 101 (Network is unreachable))
02:10:30webguest14what would be the better choice for rockbox - video ipod or iaudio x5l ? in terms of stability, etc...
02:11:34webguest14im leaning towards iaudio x5l because the native os supports flac and ogg . .. though that really wouldnt make much of a difference if i have rockbox on each . . .
02:12:33 Join ismo__ [0] (
02:13:20 Join mooch [0] (n=data@
02:13:53Soapstability is not an issue.
02:14:18SoapRockBox has not been tested on the /new/ ipod videos, and might not work at the moment.
02:14:45SoapThe assumption was it would work, but reports are leaking in that it may very well not.
02:15:07SoapBut regardless of that, the major difference is buttons, screen, battery life, and features.
02:15:56italysit locked up again :(
02:16:03 Join [xt] [0] (n=[xx]@wikipedia/xx)
02:18:18KalthareHm. I cleaned out .rockbox/ and the doom/sudoku crash went away. Grr. I should've tried that _first_.
02:18:36 Quit ismo_ (Read error: 113 (No route to host))
02:19:11 Part italys
02:22:22KalthareAnyway, I'll be back later.
02:22:26 Quit Kalthare ("Download Gaim:")
02:27:56linuxstbSoap: Didn't we just agree on "1st gen" instead of "1G" ?
02:28:39midgey34hmm anyone familiar with configuring SDL on Mac?
02:28:46midgey34I get this error: Unsupported system: Darwin, fix configure and retry
02:29:07midgey34im trying to build the sim
02:29:49linuxstbI thought that worked now... Check the patch tracker - I think there is a patch to make the sim work on the Mac.
02:30:09midgey34alright i'll check it out
02:30:12Soaplinuxstb - I can do that, but it would involve changing LOTS of appearances of 1G, 2G, 3G, 4G, 5G throughout the wiki.
02:30:59SoapI understand what and why you prefer that, but the overwhelming majority of the players are called NG
02:31:32linuxstbI think the others are less important - for the Nano, we have both 1st gen and 2nd gen, and also 1GB and 2GB versions.
02:32:18linuxstbBut also, I would like to see the wiki install instructions disappear and replaced by links to the manual.
02:32:20SoapThat is an argument I hadn't considered, I was just trying to keep the naming conventions consistant.
02:32:54SoapI was just starting there because they are the first linked pages in the ipod port section.
02:33:15 Quit ^BeN^ (Read error: 113 (No route to host))
02:33:16linuxstbAnother useful page would be one that answered the question "What ipod do I have?"
02:33:34Soapand yea, the *nix install page is weak, and the OS/X one needs updated like was done to the Windows one regarding the Nano with firmware 1.2
02:33:57linuxstbYes, I was too lazy to update all three pages with the 1.2 instructions...
02:34:05SoapCould we use the images from the manuals, or should photos be used?
02:34:36linuxstbYou mean the new svg images?
02:37:07 Join JdGordon [0] (
02:37:19linuxstbYes, I think the svg images would be fine.
02:37:44SoapThe only problems I imagine with a photo-lineup ID system is: A-is there a visual difference between the Mini 1st and 2nd, and what is the visual difference between a video 30GB 5G and a video 30GB 5.5G?
02:38:07Soap60/80GB answers itself.
02:38:38 Join davidc__ [0] (n=davidc@
02:38:47Paul_The_Nerd1st and 2nd gen minis have two key differences
02:38:59linuxstbI think you would need a textual description as well. Other information such as the date/year of release and the status in Rockbox would be nice in that page.
02:39:00Paul_The_NerdThe text color, (white vs colored) and whether or not they have their capacity on the back
02:43:37SoapOk, fixed my "fix" of the Windows instructions.
02:49:01midkayPaul_The_Nerd: "the text color" as in what?
02:49:12Paul_The_NerdI believe the labels on the wheel
02:49:36linuxstbThe IPL wiki has a good description -
02:50:41 Quit ismo__ (Read error: 110 (Connection timed out))
02:51:32*linuxstb wonders if Rockbox works on a 1st or 2nd generation ipod.
02:51:34 Quit mooch (Read error: 113 (No route to host))
02:52:00davidc__linuxstb: I can test for you when I get home...
02:53:13 Join daurnimator [0] (i=quae@unaffiliated/daurnimator)
02:53:49 Join aliask [0] (
02:55:00linuxstbdavidc__: Thanks. Feel free to fix any problems you might find :)
02:55:40*linuxstb goes to bed
02:55:51davidc__linuxstb: k.. is rockbox still fat only? [no hfs?]
02:56:06scorchedavidc__: yes
02:56:32desrtwhat about ext3?
02:56:42davidc__linuxstb: feel free to steal our hfs driver from the bootloader :P
02:56:49scorcheno support for that either
02:56:49davidc__it should be fairly simple to kludge on
02:56:58desrtext3 is more important than hfs :)
02:57:26davidc__desrt: eh, fine. You can steal both drivers then ;)
02:57:40desrtmy ipod won't even boot
02:58:49Soapplug it in to USB for an hour or so.
02:59:01Soapthey die hard when the battery is drained.
02:59:01davidc__make sure its a powered usb port
02:59:03desrtthat won't help
02:59:06davidc__charger is best
02:59:10davidc__desrt: yeah, it will.
02:59:16desrti assure you it won't
02:59:18davidc__desrt: unless ya hit it with an hammer ;)
02:59:26davidc__desrt: what did ya do to it then?
02:59:28desrtit's a new 8gb one
02:59:31davidc__linuxstb: seen
02:59:35desrtit doesn't work yet
02:59:36davidc__desrt: ah, what'd you do to it?
02:59:45desrttried to install rockbox :p
02:59:54davidc__desrt: ah, I thought you were talking about hard boot
02:59:55davidc__not os :P
02:59:58scorchewe told you it wouldnt work ;)
03:00:20desrtscorche; i made backups :)
03:00:35scorcheand if you hadnt, i would have called you names
03:00:39synicscorche: wtf.
03:00:42davidc__theres some info for a S5l8700
03:00:54scorchesynic: ??
03:01:20synicscorche: what are you running rockbox on?
03:01:42scorchesynic: are you the synic i think you are?
03:01:47synicscorche: I am.
03:01:58scorchesince when did you start coming in here?
03:02:13synicI dunno... since like 3 minutes after I found out you could run rockbox on an ipod
03:02:26 Quit jaebird ("Ex-Chat")
03:02:28synic(maybe last Wednesday or something)
03:02:29desrtsynic; DO NOT BE DECEIVED
03:02:33 Quit daurnimator ("I wonder where that fish could be")
03:02:34desrtit doesn't work :p
03:02:39synicworks for me :)
03:02:39desrtunless it's an old ipod
03:03:13synicscorche: no, but really, what are you running it on?
03:03:31desrt00000000 7b 7b 7e 7e 20 20 2f 2d 2d 2d 2d 2d 5c 20 20 20 |{{~~ /−−−−-\ |
03:03:31desrt00000010 7b 7b 7e 7e 20 2f 20 20 20 20 20 20 20 5c 20 20 |{{~~ / \ |
03:03:31desrt00000020 7b 7b 7e 7e 7c 20 20 20 20 20 20 20 20 20 7c 20 |{{~~| | |
03:03:31DBUGEnqueued KICK desrt
03:03:31desrt00000030 7b 7b 7e 7e 7c 20 53 20 54 20 4f 20 50 20 7c 20 |{{~~| S T O P | |
03:03:31desrt00000040 7b 7b 7e 7e 7c 20 20 20 20 20 20 20 20 20 7c 20 |{{~~| | |
03:03:33desrt00000050 7b 7b 7e 7e 20 5c 20 20 20 20 20 20 20 2f 20 20 |{{~~ \ / |
03:03:36desrt00000060 7b 7b 7e 7e 20 20 5c 2d 2d 2d 2d 2d 2f 20 20 20 |{{~~ \−−−−-/ |
03:03:41desrtoh dear lord
03:03:50desrtperfectly formatted for a texteditor
03:03:51davidc__desrt: hehe... I remember that since long ago
03:04:08scorchesynic: ajr
03:04:12linuxstbdavidc__: No, I hadn't found that page on the 8700 - thanks for the link.
03:04:19synicwtf is ajr
03:04:33davidc__linuxstb: no ds'es yet...
03:04:35scorchearchos jukebox recorder
03:04:38davidc__ but at least we have some specs
03:04:42linuxstbRegarding HFS support - Rockbox needs read/write, and IIUC, loader2 only has a read-only HFS driver.
03:05:40 Part Paul_The_Nerd
03:06:20CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
03:06:20*desrt finds some xml in here
03:06:42desrtuh. 100 calories?
03:07:18desrtthis must be the nike crap
03:07:41desrt <!−− Lance Armstrong attaboy voice file−−>
03:08:10desrtwhy do they store this crap in the firmware?
03:08:27desrt <!−− Paula Radcliffe attaboy voice file−−>
03:08:35desrtseriously. what a waste :p
03:08:36Soapwhy not?
03:08:49linuxstbdavidc_: Some more publicity info:
03:08:53desrtbecause i have to upload this crap when i flash my ipod now and usb is slow :p
03:08:57aliaskIt'd be stored on the hard drive either way - it may as well be hidden.
03:09:11Soapwhat a pedantic answer.
03:09:13desrtaliask; actually, i'd prefer to be able to delete it :)
03:09:36aliaskBut then you'd never be able to hear Lance Armstrong give you moral support while running!
03:10:52desrtso if not arm, what were the last round of ipods running on?
03:11:05linuxstbdesrt: There's a FAT16 filesystem hidden in that partition - that contains those xml files and the voice clips as .wav files.
03:11:16linuxstbdesrt: arm.
03:11:25scorchedesrt: they are arm....but samsung, not portalplayer
03:11:32desrt IPODRESOURCFAT16
03:11:33linuxstbSpecifically PortalPlayer SoCs containing two arm7tdmi cores.
03:11:35desrtyou speak the truth.
03:13:00linuxstbdesrt: Try running this program on the contents of the firmware paritition:
03:13:18linuxstbIt will extract the three images contained in it to files.
03:13:46linuxstbYou then need to strip the first 512 bytes from "image2.bin" and you can then use "mount -o loop" to access it.
03:14:11***Saving seen data "./dancer.seen"
03:14:19desrtoh. good. it's short.
03:14:33*desrt looks it over before executing code from random websites linked to by strangers on irc :)
03:15:11linuxstbI would be worried if you didn't...
03:15:19amiganSo I am convinced that rockbox is the greatest thing since sliced bread.
03:15:31desrtmy ipod is not a toaster. :(
03:15:32amiganOr at least it made my x5 a lot funner to use =]
03:15:39amigandesrt, it works about as well as one
03:16:30desrtlinuxstb; is this your code?
03:16:44desrtit has a bug
03:16:51linuxstb:) What?
03:16:53desrt fwbuf=malloc(len);
03:16:56desrtlen might be negative
03:17:14desrt(or insanely large, for that matter)
03:18:09linuxstbI'll take the risk...
03:18:30desrtit's insecure against rogue firmwares!
03:18:33linuxstbBut feel free to add a safety check - it shouldn't be more than 20MB
03:18:44desrti'm good, thanks :)
03:18:47amiganAllocating a negative amount of memory is fun to do.
03:19:03 Join daurnimator [0] (i=quae@unaffiliated/daurnimator)
03:19:22desrtyou need <stdlib.h> for malloc, btw and <unistd.h> for the various file functions you use
03:19:43*linuxstb makes a note never to share quick hacks again... :)
03:19:57desrtalso... your call to open() is incorrect
03:20:09desrtwhen creating files open() takes a 3rd argument −− the mode of the file to create
03:20:16desrtyou probably want 0666
03:20:22desrt(it gets modified by the process umask)
03:21:26 Quit bytie (Read error: 110 (Connection timed out))
03:22:25desrtso uh... i want to do a port :)
03:22:35linuxstbAnother interesting link relevant to the new Nano's processor:
03:23:21desrthey. not bad.
03:23:46desrt3 years old. i wonder if it's still at all accurate
03:24:10scorchelinuxstb: where did you find that one?
03:24:13linuxstbIt's talking about the CalmADM DSP
03:24:28desrtlinuxstb; it's a research paper...
03:24:29linuxstbscorche: googling for "calmadm"
03:24:38desrtat the time they published this they probably had implemented it in VHDL or something
03:26:53scorchelinuxstb: so is the new processor ARM9?
03:27:24linuxstbYes - see the first PDF I linked to (or the html page davidc__ linked to)
03:27:56 Join Kalthare [0] (
03:28:01desrtso what needs to be done, anyway?
03:28:17desrtmodify the firmware to watch what interesting things happen?
03:28:36linuxstbFirstly work out the encryption used in the firmware...
03:29:04desrtexcessively lame.
03:29:07scorchelinuxstb: ah...i see...somehow, i missed that link on my tour of samsung's site
03:29:32linuxstbUnless I'm mistaken, image1.bin (the main firmware) is encrypted. At least it's not ARM code, and there are no visible strings.
03:30:16scorche...not really
03:30:25desrtheh. the mbr looks like the start of a FAT filesystem
03:30:30desrthow odd
03:30:46linuxstbscorche: You were probably searching for "8701" - that page is for the 8700.
03:31:25scorchelinuxstb: i was, but i also was searching for anything of a similar name, which i must have glazed over that one in the list
03:31:49linuxstbA google for S5L8700 brings up a few other MP3 players using that chip...
03:32:07scorchesuch as? (lazy me)
03:32:11desrtlinuxstb; what do you use to decompile arm?
03:32:13linuxstbCould be interesting to get other firmware files.
03:32:26linuxstbscorche: :)
03:32:34scorchebah you =)
03:33:09Kaltharerrr, and I was just gonna get an ipod video, too
03:33:24linuxstbIn fact, it just seems to be one player:
03:33:49scorcheKalthare: the new ipod video does not have the new processor
03:34:18scorche"the unofficial ipod video"
03:34:50Kaltharescorche: but the bootloader doesn't work on it, right?
03:35:15scorcheKalthare: is still beginning to be investigated though
03:35:40KalthareI guess you'll have it figured out by the time I have the money.
03:36:02linuxstbdesrt: I've just been using arm-elf-objdump up to now, but should really get myself a copy of IDA Pro.
03:36:19scorcheif you wish to place a time estimate on when it will be available, go ahead =)
03:36:43*linuxstb predicts it will be available when it's done.
03:37:00*scorche votes to stick with linuxstb's prediction
03:37:07 Join Davide-NYC [0] (
03:37:15*desrt reads fw.c a bit closer
03:37:43linuxstbdesrt: The firmware partition format is documented here:
03:37:46KalthareOh, say −− does ejecting the ipod do anything in particular? Or can I just unmount it and disconnect it?
03:38:08linuxstbdesrt: So it looks like Apple have kept that part of the ipod almost the same at least.
03:38:29scorchelinuxstb: well, you might not have much luck getting a copy of the meizu's bootloader....unless you happen to have some chinese friends
03:38:32desrtso this says starts execution at 0x0
03:38:57desrti have to assume there is some other lower-level firmware (ie: not the disk) at this address
03:39:01desrtsince 0x0 on the disk is x86 code
03:39:04linuxstbscorche: It's a weird device - flash-based (1GB, 2GB or 4GB) with a 320x240 screen and xvid playback...
03:39:12scorcheyeah...i noticed
03:39:17scorcheand werent you off to bed? =P
03:40:08linuxstbThanks for reminding me. I should go there...
03:42:07desrthm. littleendian
03:42:54linuxstbdesrt: There will be some code in flash RAM which is executed when the Nano is powered up or reset. That will load (and decrypt) the firmware from the flash disk and execute it. The "aupd" image on disk is that bootstrap code and is flashed when you upgrade the firmware.
03:43:13linuxstbBut now I really need to go to bed. Goodnight.
03:44:18desrti guess if you send a new aupd then you risk for-real bricking your ipod
03:48:45 Quit daurnimator ("I wonder where that fish could be")
03:53:16 Join Quazgaa [0] (
03:53:57Davide-NYCdesrt: is this the same unit you were describing before?
03:57:00scorcheDavide-NYC: that is the placer with the samsung chip in it also
03:57:59Davide-NYCMy wants something with video playback for long flights. An 8GB version of that player looks like it could be endorsed.
03:58:07Davide-NYC*My wife wants
03:58:41scorcheDavide-NYC: i beleive it only comes in 1, 2, and 4
03:59:03Davide-NYCSo far yes. But maybe soon it'll be offered with larger capacity. (this was off topic, sorry)
03:59:10 Quit hardeep ("[BX] Become a BitchX Certified Systems Engineer today! Apply within!")
03:59:12desrtthey used mtools to create the resource filesystem
03:59:35Quazgaawow, people still use bitchx?
03:59:58 Quit Rob2222 (Read error: 110 (Connection timed out))
04:03:31KalthareLooks like I may be getting one of the new iPod things today. Any way I'd be able to help figure out the bootloader without putting it at risk?
04:03:35desrt"this is lance armstrong. congratulations. you've covered 250 miles"
04:03:51desrtthis is surreal
04:04:04synicdesrt: heh, that's in the firmware too?
04:04:45synicthat's really weird.
04:04:50desrti just sucked it out using mtools
04:04:51syniccrazy apple hackers
04:04:55 Quit gopp ("gop: god giveths and democarts taketh away!")
04:05:02desrti'm deeply amused
04:07:37synicbah, now I want to explore the 5g firmware
04:07:42synicthis is your fault.
04:08:00JdGordonfor zxbox... do need to get the .tap or .tzx?
04:08:23aliaskJdGordon: I think either is ok - just make sure you get the 48k one
04:08:30 Quit BHSPitLappy (Remote closed the connection)
04:09:13scorchesynic: not sure what else you might find of interest...that was added with the 1.2 firmware update for the nike-apple device
04:10:25 Join actionshrimp [0] (
04:13:36 Join ^BeN^ [0] (i=Paprica@
04:14:24 Join Paul_The_Nerd [0] (
04:17:01JdGordonsimcity seems to be the only game i cn get working...
04:29:54 Quit sucka (Read error: 101 (Network is unreachable))
04:32:42 Join vik [0] (
04:32:58 Part pixelma
04:33:54 Quit Kalthare ("Download Gaim:")
04:34:18 Join nudelyn [0] (
04:34:49vikwhat units does rockbox store volume in internally?
04:34:54 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
04:41:28 Quit XavierGr (Read error: 110 (Connection timed out))
04:45:35 Join mooch [0] (n=data@
04:47:51 Join ismo_ [0] (
04:50:19 Quit ^H (Read error: 110 (Connection timed out))
04:52:19midkayvik: you mean decibels or what?
04:52:32vikmidkay: yep
04:52:43midkayvik: then, decibels. :)
04:53:02 Quit amiconn (Read error: 110 (Connection timed out))
04:53:02 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
04:53:35vikwasn't sure if it was that or % or 0-255 or something.
04:53:52midkayah. used to be % a while ago.
04:57:36vikbut now I've found functions which give the min and max anyway, making it independent of units
05:06:36 Join mooch_ [0] (n=data@
05:07:15 Quit mooch (Read error: 101 (Network is unreachable))
05:10:01 Quit ismo_ (Read error: 110 (Connection timed out))
05:14:14***Saving seen data "./dancer.seen"
05:15:16 Join ismo_ [0] (
05:20:54 Quit EV|Toshiba (Read error: 104 (Connection reset by peer))
05:24:41 Quit _Veseliq_ (Remote closed the connection)
05:26:05 Join _Veseliq_ [0] (n=veseliq@
05:46:49 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
05:56:55 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
06:11:12 Join damaki [0] (
06:16:49 Quit Nibbier (Read error: 110 (Connection timed out))
06:16:57 Join Nibbier [0] (
06:23:58 Join JoeBorn [0] (
06:26:44 Quit damaki__ (Read error: 110 (Connection timed out))
06:31:51 Quit davidc__ (Read error: 54 (Connection reset by peer))
06:33:00 Join davidc__ [0] (n=davidc@
06:40:53 Quit rotator ("zzzzzzzzz")
06:54:00 Quit [xt] ("(madvirii): i thought the cosby show was programmed in linux")
07:08:01 Quit Xerion (Read error: 104 (Connection reset by peer))
07:09:54 Join aegray [0] (n=aegray@
07:11:38 Join Xerion [0] (
07:14:16***Saving seen data "./dancer.seen"
07:30:16 Join damaki_ [0] (
07:31:45 Join Kalthare [0] (
07:36:45 Join spiorf [0] (
07:39:14 Join rigel [0] (
07:39:20rigeldoes rockbox supprt flac?
07:39:42rigelwhat about a monkeys-audio plugin?
07:39:54JdGordonif you say so... id rather have 14 weeks of music on my DAP than 14 minutes
07:40:20rigel20 gigs is still 35-40 full length 74:00 cds
07:40:40Quazgaahehehe 14 minutes
07:41:06KalthareI think you'll get worse battery life, due to more disc accesses.. right?
07:41:19rigelah, probably
07:41:31scorcheyes...but it is a more efficient codec than others, so it balances out somewhat
07:41:38zei've got most of my stuff in mp3 and ogg
07:41:54zebut some things in flac that i like in particular and may be more prone to psychoacoustic artifacts
07:42:10zewhich those 2 properties fairly easily intersect for me
07:42:25Paul_The_NerdFLAC is one of the most efficient codecs, but MP3 is now, for all practical purposes, nearly as efficient since they both play unboosted.
07:42:41midkaythey do?
07:42:48KalthareAny news on the 5.5G?
07:42:48Paul_The_NerdOn coldfire
07:42:59scorcheKalthare: nope
07:43:03hcsha, I'm playing from ADPCM, beat that for efficiency
07:43:06Paul_The_NerdKalthare: The firmware partition contents have changed, so the tools need to be adjusted.
07:43:35Paul_The_NerdQuazgaa: The processor doesn't have to change to a faster speed (on the iRiver HXXX series players) during playback.
07:43:39scorchePaul_The_Nerd: about how long ago was it optimised for coldifre?
07:43:53Paul_The_Nerdscorche: It was actually a general optimization. It benefited the iPods as well.
07:43:57Quazgaai see
07:44:08scorcheah..i dont remember hearing such news
07:44:13KaltharePaul_The_Nerd: Okay, thanks −− that's what I figured. Is there anyone working on it?
07:44:21Paul_The_Nerdscorche: If I recall it was the one that made the iPods not *always* skip no matter what.
07:44:36Paul_The_NerdKalthare: I don't think anyone here has one yet.
07:44:40 Quit damaki (Read error: 110 (Connection timed out))
07:45:07KaltharePaul_The_Nerd: I'm getting one tomorrow. Any way I can help?
07:45:40Paul_The_NerdKalthare: Ask linuxstb, when he's around, if he'd be interested in a dump of the firmware partition from it. He might be.
07:46:11scorchei would say he is
07:46:14KaltharePaul_The_Nerd: Again, thank you.
07:46:23Paul_The_NerdI'll be he is.
07:46:28Paul_The_NerdBut I can't *promise* that he will be. :)
07:48:49KalthareWell, I'm pretty sure I'm getting one tomorrow, at least.
07:53:50 Quit JoeBorn ("Leaving")
08:01:50 Join LD [0] (
08:01:57 Quit davidc__ ()
08:02:03LDsup folk
08:03:54LDdoes anyone have any suggestions for a person using a pioneer ipod hookup?
08:04:05 Join marktaff [0] (
08:04:22 Join ^H [0] (
08:04:59 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
08:05:31 Quit linuxstb (Read error: 110 (Connection timed out))
08:05:34hcswhen rockbox freezes is there anything I should try besides rebooting?
08:06:35marktaffI'm thinking about buying an ipod nano at circut city. How can I tell if it is a generation 1 ipod nano that is supported by rockbox?
08:06:37^HI wihs MIRC wouldn't "helpfully" remember stupid nicks that you used as a one-off.
08:07:59Paul_The_Nerdmarktaff: If it's white, or if it's Black and NOT 8gb (and if it's not in an aluminum shell) it's a first gen. Best Buy, at least my Best Buy, doesn't have them any more.
08:08:02scorchemarktaff: the 1st gen nanos are either black or white...have a shell of plastic, not metal...come in 1, 2, and 4GB capacities
08:08:21Paul_The_Nerdscorche: The new ones come in black, but the black one is only 8gb
08:08:30 Nick ^H is now known as nudelyn2 (
08:08:32scorchei know
08:08:35marktaffThanks. this is a 2GB white one for $129US
08:08:55scorchedoes it come in a clear case or a square-ish box?
08:10:07marktaffHaven't see the box, just the display model
08:10:36scorchedoes it have curved edges on the sides?
08:10:41scorcheor just a beveled edge?
08:11:29marktaffIt was rectangular and thin with rounded corners, otherwise no curves
08:11:47scorchethen it was a 1st gen
08:11:49amiconnIiuc white should always be 1st gen
08:11:51 Join daurn [0] (n=quae@
08:12:19scorcheyeah...the new ones are silver...i just wanted someone to talk to =(
08:12:29marktaffand does rockbox work well on the gen 1 ipod nano?
08:13:03scorchedefine "well"
08:13:12scorcheit is well wnough by my standards
08:13:20marktaffI run cutting edge KDE, so I am OK with occasional bugs, just so that it usuallly works right, and bugs get fixed...
08:14:33marktaffThanks for all your help. I'll buy one on the way to work tomorrow. ;-)
08:14:41midkay<marktaff> It was rectangular and thin with rounded corners, otherwise no curves
08:14:41midkay<scorche> then it was a 1st gen
08:14:49midkayrounded corners means 2nd gen, no?
08:14:49LDI'm debating trying rockbox
08:14:53marktaffIt will make my flights *much* nicer
08:15:08midkaythey're certainly much more rounded than the 1st gen which are very barely rounded at each edge..
08:15:22scorchemarktaff: well, were they rounded, or a big curve?
08:15:42marktaffdon't recall for sure, just looked at the display for a minute or two...
08:15:51marktaffany way to tell based in model numbner?
08:16:09scorchewell, if you are sure it was white and not silver, it is a 1st gen anyway
08:16:12midkaycan't recall if it had some plastic on the front or if it was completely aluminum?
08:16:25scorchewhite on the front at least
08:16:36midkayscorche: alright, calm down. we heard you the third time.
08:16:41marktaffthis one is white plastic front, silver colored plastic back, I think
08:16:51scorchethen you should be fine
08:17:28midkaythat'd be a first gen, thus you *will* be fine. :)
08:17:34scorcheif all else fails, ask the employee if "it is a new one, or the old kind"
08:18:02scorcheterms they would understand ;)
08:18:05midkayhuh? if it fails he's already got it home and realized it doesn't work..
08:18:52scorchemidkay: not everything is literal....that is a common phrase meaning a variety of things including "if you are not sure about any of the means we have said"
08:19:41midkayscorche: i was just amused at how you start with "well, it's white, so it's a first gen" and end up at "well, it's white, and it's got plastic, so it's probably a first gen, and you could also ask if it's a new one or the old kind and maybe that would help reassure that it could be a 1g".. :)
08:19:41scorcheand if not, a phrase can be adapted to use
08:19:43marktaffThanks again guys.
08:20:01scorchemidkay: if you want to fight about it, i suggest uploading the map
08:20:12 Part marktaff ("Konversation terminated!")
08:20:28midkaydon't feel like owning you right now, anyways.
08:20:48scorcheplease...i have won a good majority of the time
08:21:46midkayusually when you wait to sv_lan 0 for a few minutes or disconnect and then argue that you were ahead when you "crashed" or "dropped out" even when there are screenshots to prove the opposite.
08:22:31midkayanyways, enough of your inconsistent drivel. i'm going to bed.
08:22:32scorcheexcuses...last time for example, we stopped after the first crash, and i i was ahead by at least 10
08:22:34 Quit nudelyn (Read error: 110 (Connection timed out))
08:22:38midkayriight, right.
08:22:59scorchedrivel?....that is insulting....that is usually a word we use to describe dongs =/
08:33:58 Join einhirn [0] (
08:33:58 Quit einhirn (Client Quit)
08:38:12 Join theli_ua [0] (
08:41:00 Quit LD ()
08:43:45 Join Mordov|hardlywo [0] (
08:48:57 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
08:49:38 Part Paul_The_Nerd
09:03:45 Join DarthShrine [0] (n=darth@pdpc/supporter/student/DarthShrine)
09:12:47 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:13:19 Quit vik ("Leaving")
09:13:56LinusNif any of you wonder where bagder is today, i can tell that he is now a proud father of a son
09:14:18***Saving seen data "./dancer.seen"
09:14:42peturwhen did he get the time to do that? must have been offline some minutes ;)
09:14:46rigelhey, the h10 20gb manual should be updated
09:14:51LinusNpetur: lol
09:14:56rigelbecause the controls are all messed up
09:15:03rigeli cant figure out how to get to these menus anymore
09:15:39preglowdamn, you people keep going at it
09:15:47preglowhow're you going to keep up the rockbox time if you keep fathering children!
09:17:09scorcheLinusN: is the child named sansa? =)
09:17:49preglowe200 would be a cooler name
09:18:28scorcheand as he grows up, we can start calling him e250
09:18:33scorchethen e280
09:18:42DarthShrineThen iPod *runs*
09:19:19scorche...that is after he has reached the top of the hill ;)
09:21:43 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
09:24:28 Quit gromit` (Read error: 104 (Connection reset by peer))
09:26:48 Join linuxstb [0] (i=5343d4aa@gateway/web/cgi-irc/
09:27:46linuxstbpreglow: Did you see this link in the IRC logs?
09:28:07preglowlinuxstb: i stopped reading irc logs after they became huge, heh
09:28:20preglowhaving a peek now
09:28:28linuxstbPlus here:
09:29:09peturwow... it has usbhost
09:29:28 Join gromit` [0] (
09:29:29peturin fact it has lots of neat stuff
09:29:36preglowis this the nano core or what?
09:29:48Kaltharelinuxstb: I'm probably going to be getting a 5.5G iPod tomorrow, and it was suggested to me you might be interested in a dump of the firmware partition. Is that the case?
09:30:11linuxstbpreglow We think the Nano has a S5L8701 - those are the specs for the S5L8700
09:30:26 Quit gromit` (Client Quit)
09:30:37petur"Digital data recording/playback system utilizing EEPROM and ROM memories as a storage medium"
09:30:46linuxstbKalthare: Yes, I'm always interested in firmwares.
09:30:59linuxstbpetur: Yes, patented..
09:31:38 Join pondlife [0] (
09:31:45linuxstbPlus a 120MHz DSP...
09:31:48Kaltharelinuxstb: Okay, cool. If I do get it, I'll get back in touch with you.
09:32:10preglowlinuxstb: looks very, very nice
09:32:24preglowlinuxstb: god knows how much fancy-ass stuff you can do with that rig
09:32:32amiconnpreglow: The calmrisc will be 'fun' to code for...
09:32:44preglowamiconn: using it as a dsp engine shouldn't be too hard
09:32:44linuxstbSeems major overkill for the Nano though - especially with Apple's track record of underusing hardware.
09:33:06preglowamiconn: and yeah, i probably WILL think it's fun :P
09:33:46linuxstbIf you google for "calmadm" you'll find a little info about it.
09:33:46amiconnYeah, maybe using it for dsp stuff exclusively will be easier than what the gmini 1x0 does
09:33:59preglowamiconn: by far
09:34:14preglowi know it's got a single cycle 24x24 mac engine
09:34:16preglowwhich is nice
09:34:41amiconnBut I wonder why they added the calmrisc+mac at all when there is a 200MHz arm core
09:34:49preglowamiconn: power consumption
09:35:21 Quit perplexity (Read error: 113 (No route to host))
09:38:16 Part BossG
09:39:26 Join BossG [0] (
09:40:48 Quit Kalthare ("Download Gaim:")
09:40:59amiconnLinusN: Did you upgrade meanwhile?
09:41:17LinusNdid that last night, after reading about it in the irc logs
09:41:51amiconnok :)
09:42:46preglow256kb sram...
09:43:04preglowif this thing is in the new nano's, i'm bloody getting one the second they're cracked
09:43:29*amiconn wonders what the ipl people will do
09:43:44preglowre it, probably...
09:44:58preglowbut it'll probably be a bitch
09:45:46preglowwhat with the no docs, encrypted code and epoxied chips
09:46:43nudelyn2what's sram good for?
09:46:59amiconnThere must be some unencrypted loader (in eprom) that loads and decrypts the encrypted part(s)
09:47:13preglowamiconn: sure, but we'll need to read that first, then
09:47:39amiconnIs the eprom part of the soc, or is it separate chip like in th eolder ipods?
09:47:42 Quit Quazgaa ("leaving")
09:47:48preglowseparate, i think
09:47:52preglowbut nothing impossible these days
09:48:11amiconnA separate chip could be unsoldered and read out externally
09:48:39preglowproblem is they've epoxied all the important chips to the board
09:48:47amiconnIt's a question of equipment, and of the package. BGA is hard...
09:49:00preglowso we'd need to read the bus lines directly
09:49:05amiconnAre there photos somewhere?
09:49:14preglowyeah, linuxstb will no doubt supply you with urls
09:49:50markunnudelyn2: it's very fast ram
09:50:04preglowdoesn't look like this thing has any onboard mask rom
09:51:18nudelyn2ah ok :)
09:52:02preglowThe ARM9 core is used to run the operating system, task management and various control tasks.
09:52:05preglow200 mhz for that, yes
09:52:47preglowwe bloody run an entire os with decoder on 75mhz arm7s now, so that's not overkill at all, no
09:53:23preglowhmm, it does seem to have 50kb onboard rom
09:53:27 Part BossG
09:54:57rigelmaybe it needs the extra horsepower for decrypting the code? ;)
09:55:56rigelkind of like how code bloat in windows drives to some extent adoption of new higher powered processors
09:56:08rigelon the consumer end, i mean
09:56:57preglowlet's just hope they've put some fancy-ass decrypting code in that eeprom, then
09:57:08preglowin which case we'll have to read code off the data bus when they write it to ram
09:57:19 Quit jhMikeS (Read error: 110 (Connection timed out))
09:57:38rigeli was just talking shit
09:57:45rigeli have no idea what im saying
09:58:26amiconnpreglow: If the system design is half-decent, you won't be able to read the code from the data bus if it's stored in the internal rom
09:58:47amiconnThe internal data won't be driven on the external bus
09:58:55amiconnThat would be a waste of time and power
09:59:02preglowamiconn: of course not, but they'll have to decrypt the firmware from the external rom
09:59:26preglowamiconn: and that's only in the unlikely case that it's not decrypted before it's flashed
09:59:29amiconnYes, but that's only the input data for the decrypter
09:59:45amiconn...not the code to decrypt it
10:01:16preglowsure, but we don't care if we can read the decrypted data
10:01:21preglowwhich we can, unless they put everything in sram
10:02:23 Quit solarflare ("This computer has gone to sleep")
10:04:36 Quit linuxstb ("CGI:IRC")
10:16:49 Join safetydan [0] (
10:17:58 Quit ismo_ (Read error: 113 (No route to host))
10:19:18 Join gromit` [0] (
10:21:28 Join shadok [0] (n=shad@unaffiliated/shadok)
10:22:17 Join Zmanu [0] (
10:22:39Zmanuis rockbox under gpl ?
10:22:48Zmanuok thanks
10:22:53 Part Zmanu
10:23:19scorcheaaaaannnnd with 36 seconds, ITS A NEW WORLD RECORD!
10:23:29LinusNnice with some simple questions for a change
10:23:39 Join Zmanu [0] (
10:23:42Zmanui come back
10:23:51Zmanuthe licence is not on the website
10:24:07Zmanumaybe something to add
10:24:12markunLinusN: someone connected to his gigabeat over serial yesterday:
10:24:44Zmanuok tnaks
10:25:10 Quit scottder_ (Remote closed the connection)
10:25:12 Join scottder [0] (
10:25:14 Part Zmanu
10:25:59LinusNmarkun: nice
10:26:55 Join mirak8 [0] (i=5372a20c@gateway/web/cgi-irc/
10:26:56markunHe's more interested in a (better) linux port thouhg
10:27:02nudelyn2Has anyone worked out why the 5.5G iPod has better *video* battery life, but identical audio life? More efficient backlight or something? (Might explain why it's brighter as well.)
10:28:00mirak8markun for gigabeat ?
10:31:39 Join solarflare [0] (
10:32:09mirak8you want to run gnome ? :p
10:33:49linuxstb_nudelyn2: I think Apple have included a major update to the firmware that runs on the Broadcom chip - the new version now decodes higher resolution video and has lower power consumption.
10:34:14 Join Seed [0] (i=ben@
10:34:23nudelyn2i see, so that improvement may apply to older 5G ipods as well
10:35:23linuxstb_I would expect so.
10:35:45linuxstb_At least, older 5gs can now decode the same higher resolutions as the new 5gs.
10:39:17markunmirak8: maybe matchbox? :)
10:41:44 Join bytie [0] (
10:42:16 Quit solarflare (Read error: 54 (Connection reset by peer))
10:43:43markunmirak8: but I'm more interested in rockbox
10:44:15 Quit Seedy (Read error: 110 (Connection timed out))
10:48:10mirak8was it possible to implement all ffmpeg ?
10:48:28mirak8or does it produce to big binary ?
10:50:49 Quit lini (Read error: 60 (Operation timed out))
10:53:07 Join PaulJam [0] (
11:00:14*amiconn still finds it really odd that a dap runs linux and has a 300MHz cpu
11:00:31amiconnI mean that it runs linux for normal operation
11:00:56pondlifeSeeing as jhMikeS isn't around, can anyone with a device handy confirm (a) that recording in CVS works and (b) find out whether breaks recording
11:01:21pondlifeIt's the playback-stops-voice-and-playback-dies fix. Take 2.
11:04:18LinusNpondlife: which' target do you have?
11:05:14pondlifeH300, but it's not here with me
11:05:42LinusNi see
11:05:58pondlifeI was hoping to see if jhMikeS could help, but he's not here.
11:06:38pondlifeAnyway, I'm off to a meeting now, but I'd really like to progress this asap as the current CVS is a bit of a mess and I feel a bit responsible.
11:07:02LinusNi'll test and report back
11:07:30peturgood final note on that patch :)
11:07:44pondlifeI bear no ill will ;-)
11:08:04*pondlife has calmed down a bit over the past day
11:08:43pondlifep.s. how do I link entries on Flyspray? RaeNye mentioned it on the ML
11:08:53pondlifei.e. link a bug and a patch
11:09:46 Join p3tur [0] (i=d4efd6a6@gateway/web/cgi-irc/
11:09:46 Quit petur ("CGI:IRC (EOF)")
11:09:59 Nick p3tur is now known as petur (i=d4efd6a6@gateway/web/cgi-irc/
11:10:54peturpondlife: edit task, then click the Related Tasks tab
11:11:02 Quit scorche (Read error: 104 (Connection reset by peer))
11:11:19 Join scorche [0] (
11:11:44petursilly me, you don't even have to edit task
11:14:11 Quit pondlife (""Uh ok, back in a few hours"")
11:14:20***Saving seen data "./dancer.seen"
11:14:23*petur wonders why intel names 4-core chips Core 2 Quadro and not Core 4
11:15:16 Quit mirak8 ("CGI:IRC (EOF)")
11:22:48preglowbecause they'r eintel
11:30:21amiconnProbably the same reason why the clock-tripled 486 was named DX4 instead of DX3
11:32:23LinusNhmmm, recording works with the patch, but the recording monitoring doesn't work
11:33:02LinusNthe peak meters show the correct levels, but the headphones are silent
11:33:34LinusNvoice and audio playback works fine
11:35:21 Join lee-qid [0] (
11:36:20 Join Mmmm [0] (
11:37:00 Join JHChristMan [0] (n=CHRIS@
11:37:26 Join perplexity [0] (
11:39:12 Part JHChristMan
11:39:18 Quit newbyx86 (Read error: 110 (Connection timed out))
11:45:16 Quit lee-qid ("Trillian (")
11:45:56 Join lee-qid [0] (
11:58:15 Join Kohlrabi [0] (
11:58:32 Join Baerman [0] (i=431578a3@gateway/web/cgi-irc/
11:58:58Baermanwhats up
11:59:17DarthShrineThe sky, quite possibly.
12:01:16 Join baerman714 [0] (
12:01:19 Quit Baerman (Client Quit)
12:01:27baerman714is rockbox nice?
12:01:46DarthShrineWhat do you define as nice.
12:01:58baerman714better then the ipod firmware that i have now?
12:02:20DarthShrineWhat's better?
12:02:29baerman714which is better
12:02:35baerman714rockbox or ipod firmware
12:03:10baerman714how do i instal it?
12:03:37baerman714ok 1st can it kill my ipod?
12:03:56markun(at least I don't think that has happened yet)
12:04:02baerman714i have a nano
12:04:30markunAnd there are a lot of users who have tried:">
12:04:41DarthShrineI've got it working on my Nano. Be warned of the battery usage, though.
12:05:20DarthShrineNo, just short.
12:05:29baerman714lol iasnt it always short/
12:05:43markunbaerman714: do you have use for any of the rockbox features?
12:06:14baerman714i like the theme changer thing
12:06:33 Quit petur ("lunch!")
12:06:35baerman714does it make nano play videos?
12:06:50DarthShrineIf you like huge, soundless videos, sure.
12:07:01scorche(it is a work in progress)
12:07:32baerman714do stolen ipods have anything in them that phones home if there reported?
12:07:46scorchehow could they?
12:07:52baerman714no idea
12:07:55DarthShrineYou've stolen an iPod now?
12:07:58baerman714when u connect them to itunes or something
12:08:00scorchethen that would be a no ;)
12:08:14baerman714no i "found" it
12:08:30baerman714no found
12:09:07scorchei have seen many cases of "Found ipod"
12:09:15Soapthere is a screen of owner information.
12:09:25baerman714watched some person walk into the apple store in the mall and connect there ipod to the charger thing and walk back into the mall to shop
12:09:26Soapisn't there?
12:09:32scorchemake the requirements to get it, to list the music that is on the device
12:09:34baerman714so i picked it up
12:09:47Soapno, that is theft.
12:09:52Slasheriah, now the tagnavi.config syntax has been improved with support for sub menus and including user's custom menu :)
12:09:54baerman714lol they shold have left it there then
12:10:05scorcheand you should have returned it to them
12:10:20Soaplol fuckhead lol
12:10:26baerman714the only info the onscreen thigny gives is the serial and stuff
12:10:34Soaptake it back
12:10:37baerman714u can change the name the person used
12:10:49baerman714and u can get the password that locks the screeen
12:11:03baerman714u just connect it to the pc and unhide the folder
12:11:11scorchei am sure that the person went to the employees of the store...they should have some information that the person left
12:11:14baerman714and click the file that says locked
12:11:15Soapthe point is, the name they used could help you return it, but you are a miserable scum and wouldn't even try.
12:11:27baerman714and it shows the pass
12:11:36DarthShrineYeah. Rockbox will hax your computer because of that.
12:11:39 Quit Mmmm (Remote closed the connection)
12:11:44baerman714the name said Pod
12:11:53baerman714wasnt a human name im sure
12:12:01scorchehilarous indeed
12:12:08Soaptake it back to the store you stole it from, stop making excuses to justify your theft.
12:12:19baerman714take it back for wat
12:12:27Soaptake it back so the owner can claim it.
12:12:39baerman714im sure the owner knows it got snatched
12:12:54scorchei really dont see the humor in that
12:13:03*markun neither
12:13:25baerman714its 8 gigs and its bluer
12:13:39Soapit isn't yours, take it back,
12:13:48markunrockbox doesn't work with the 8 gb version
12:13:51scorchei personally would be pretty damn pissed if someone stole by brand new device that i paid $250 for
12:13:58baerman714yea me to
12:14:00DarthShrinebaerman714: Keep it and give the guy the money for iPod + music
12:14:21baerman714wat music?
12:14:26markunLet's hope they have a camera in the store..
12:14:46DarthShrineThe music that's meant to be on an iPod...
12:14:53DarthShrineThat's normally the point of one.
12:14:54baerman714lol u pay for the music?
12:15:04baerman714i just dl it on p2p and add it to the librrary
12:15:28 Join Rob2222 [0] (
12:15:28scorchehe steals an ipod...what makes you think he would not steal music as well?
12:15:40baerman714why would someoen pay for music from itunes
12:15:49DarthShrineI never suggested that he paid. Maybe the original owner would though.
12:15:51scorchethere are things called CDs
12:16:06baerman714wat about it
12:16:09markunbaerman714: you can't run rockbox on your nano
12:16:22baerman714i have a 4 gig one too
12:16:35baerman714work on those ones?
12:16:38DarthShrineThat one will self destruct.
12:16:56DarthShrineAnd report to Apple for stolen music.
12:17:05baerman714what one
12:17:13 Join Genre9mp3 [0] (
12:17:34baerman714someone told me they do that is that true if u have downloaded music on ur ipod apple makes it shut off and not work again?
12:18:10scorchebaerman714: i personally dont have any urge to help you any longer...i doubt anyone else in this channel feels that either...
12:18:11 Quit bytie (Read error: 113 (No route to host))
12:18:26baerman714wow just cuz i found a ipod
12:18:28SoapLake Winola?
12:18:32Soapyou STOLE an ipod.
12:18:34scorcheyou didnt find it
12:18:37scorchesee above
12:18:52baerman714someone left it behind and i found it
12:18:54DarthShrineYou found it in a store.
12:19:01baerman714yea that some person left
12:19:09Soapstill doesn't make it yours.
12:19:15markunlike 'finding' a car that is parked somewhere
12:19:18DarthShrinePeople leave food on shelves in the super market. Do you take it?
12:19:33 Join KN|stiff [0] (
12:19:34scorchehe was charging his device, and most likely, you nabbed it when the employees were not looking
12:19:48baerman714no the emplyes didnt know
12:19:53baerman714he didnt ask
12:19:57*Genre9mp3 reads the logs...
12:19:59baerman714he just walked in from shoppin somewhere
12:20:01scorcheit is a common thing now a days for people to come into retail stores and charge their devices
12:20:16Soapit still isn't yous.
12:20:19baerman714emplyes didnt know it was there
12:20:27Soapit still doesn't make it yours.
12:20:44baerman714well i waited 3 hrs b4 i walked by it
12:20:55Soapit still doesn't make it yours.
12:20:58baerman714and it was like 2 min b4 closeing time for the mall
12:21:09DarthShrineNow you change the story.
12:21:11Soapit still doesn't justify taking what is not yours.
12:21:11baerman714so idk wat happen to the dude that left it
12:21:23markunI must admit it's strange to leave your new toy in a store
12:21:40baerman714yea id never leave my ipod sitting in some store
12:21:51DarthShrineI wouldn't if you were around, baerman714
12:22:07SoapDid you show it to all your friends at high school?
12:22:22scorchemarkun: when i used to work retail, people would come in and charge their dead devices for a bit...i am guessing that this person left it, told the employees, and nabbed it when they werent looking
12:22:38DarthShrineI'm below highschool O_o
12:22:38Genre9mp3hehe... baerman714... you seem to have your "own" point of view at the word "ethics"!
12:22:42baerman714nope he didnt talk to workers
12:22:48baerman714he stuck it on and walked out
12:23:08scorchei am done with this....we have created enough spam
12:23:53baerman714yea time for me to get off this laptop in target now
12:24:02DarthShrineSo funny.
12:24:14baerman714bought ATI x1900 512mb pciE card
12:24:30DarthShrineBy bought you mean found?
12:24:37baerman714no bought
12:24:43scorcheDarthShrine: if you are going to say indeed, then stop also =P
12:24:51DarthShrinescorche: Indeed pwnz j00.
12:24:59*Soap with the /ignore
12:25:04DarthShrineMost intellectual-sounding word...Ever.
12:25:48baerman714There was this blind man right
12:25:48baerman714He was feelin' his way down the street with a stick right (yeah), hey
12:25:48baerman714He walked past this fish market, you know what I'm sayin' (fish market?)
12:25:48DBUGEnqueued KICK baerman714
12:25:48baerman714He stopped he took a deep breath he said
12:25:48baerman714"Snfffffff, woooo good morning ladies"
12:27:38 Quit DarthShrine ("Leaving.")
12:28:46 Join DarthShrine [0] (n=darth@pdpc/supporter/student/DarthShrine)
12:31:40 Part baerman714
12:38:47 Part DarthShrine
12:41:10 Join damaki [0] (
12:41:14 Join scorche` [0] (
12:42:05 Join bytie [0] (
12:42:13scorche`testing...nothing to see here
12:42:41 Part scorche`
12:43:20 Quit PaulJam (".")
12:46:47 Join ismo_ [0] (
12:55:02 Quit damaki_ (Read error: 110 (Connection timed out))
13:02:56 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
13:06:49 Join Criamos [0] (
13:11:44theli_uaanyone can recommend something to read on preprocessor macroses?
13:14:24***Saving seen data "./dancer.seen"
13:15:25LinusN"the c programming language"?
13:17:47nudelyn2 :-)
13:18:23nudelyn2(less applicable in straight C of course)
13:18:51LinusNtry using inline functions when parsing the linker command files
13:19:33nudelyn2sounds nasty
13:19:45 Nick nudelyn2 is now known as nudelyn (
13:19:55 Join jhMikeS [0] (n=jethead7@
13:20:06LinusNwe use the preprocessor for a lot of stuff
13:20:34nudelynRB is C rather than C++ right? Does C even have inline?
13:20:38LinusNbut mostly for simple conditional compilation
13:21:05LinusNnudelyn: c99 has inline
13:21:08nudelynYeah, no alternative for conditional compilation.
13:21:12nudelynAh, I didn't know that.
13:21:46LinusNi agree that you shouldn't use macros instead of inlined functions
13:32:45IltsuHmm, my ipod doesn't boot :P
13:34:17dwihnoLinusN: speaking of nothing... 5400 RPM disks spin a whole lot more than the "older" 4x00-something drives :)
13:35:12LinusNdwihno: figures...
13:35:50 Join Mmmm [0] (
13:36:15dwihnoLinusN: still, pretty happy about it though... I read somewhere on ebay where someone claimed it's possible to use rockbox with 200gb drives (not that it really matters)
13:36:21dwihnobut it's kinda false advertising
13:37:02scorchedwihno: on what DAP>
13:37:45dwihnoscorche: mhm
13:38:05scorchewell, you can use a 200gb drive, but you will only get 127 GB out of it =)
13:38:54dwihnoscorche: if people manually sync the content with the drive bypassing the isd300 perhaps it might be doable.. but I think LinusN once mentioned a limit in rockbox as well
13:39:41scorchedwihno: well, if it happens, let me know...i have a 160 gig perpendicular drive in there atm =P
13:41:05 Join XavierGr [0] (
13:41:06scorchemakes it terribly hard to feel an urge to upgrade to a non-ancient player
13:41:09dwihnoscorche: well, you could simply detach the drive and sync it using some other interfacing... :)
13:41:36dwihnoYou got a 160GB drive just to squeeze the last 7GB out of the possible amount?
13:41:40scorchedwihno: for the price i got it, i dont mind at all that i am not getting the full potential out of it
13:42:01scorchefor $60, when those seagate drives were still 300...
13:43:14dwihnoI would just be annoyed knowing a lot of storage wouldn't be used :)
13:43:28 Quit safetydan ("Ex-Chat")
13:44:14scorchewhen i finally get a laptop, i can always change them out anyways to have more for the lappy
13:44:55dwihnotrue, true.
13:45:40scorcheand yeah...5400 do spin more...i can balance it on my knee and see it quiver when the drive spins up =>
13:45:45dwihnoI actually needed a smaller 2.5" disk for a linux'ish work project... So I convinced my superiors I could use my old disk (and thus, I could upgrade)
13:46:42 Join t0mas_ [0] (
13:53:05Mmmmamiconn: I think that when a timer has finished you should be presented with a screen that tells you it has finished and asks if you wish to execute the timer task. What do you think?
13:53:09MmmmIf so, would the only way to do this be to have a thread waiting for the task to be passed via a queue from the tick task?
13:54:43LinusNi'm thinking of a dedicated cron thread, sending SYS_xxx messages to the gui thread
13:55:14LinusNand the default message handler does what you just mentioned
13:55:14 Join newbyx86 [0] (
13:56:34Slasherijust committed the improved tag browser menu configuration support =)
13:56:49Slasheriincluding even sub menu support for including of external files :)
13:57:09MmmmLinusN: Hmmm I'm begginning to get the feeling I'm getting a bit above myself here! :D what's a cron thread?
13:57:33LinusNMmmm: you know what the unix "cron" system is?
13:57:46Mmmmno.. maybe I should look it up...
13:57:56LinusNit's a system for scheduling events
13:58:11LinusNi'll call it a timer thread from now on :-)
13:58:21Mmmmright! :D
13:59:04 Join pondlife [0] (
13:59:22LinusNone approach could be to let the timer thread have a list of callbacks to call at specific times
13:59:46IltsuAnyideas how I can recovery my ipod? (nano) It doesn't boot :(
14:00:00MmmmLinusN: I've made a countdown timer for the recording screen which uses a tick task for the timer, but if you implement what you just said, I suppose this won't be needed right?
14:00:49LinusNwhat does the tick task do in your implementation?
14:01:27Mmmmit just counts down and then I thought when it reaches 0 posts a queue
14:01:43LinusNmy idea is to create a general event scheduling framework, which the recording timer could use
14:01:59 Join actionshrimp [0] (
14:02:04Mmmmyeah..that would be better...but harder to do! :)
14:02:06LinusNyour tick task is basically what i had in mind for the "raw" timing
14:02:29LinusNexcept it posted to the timer thread queue
14:03:48LinusNthe simple raw timer api could probably be "send me a message in X ticks"
14:04:47Mmmmright, but wouldn't it also need to know whatthe timer is for so that it could do something if you replied yes to the message?
14:05:18LinusNthe thread asking for the message will of course know what to do when the message arrives
14:05:51LinusNthe "complex" api could be "call this function in X ticks"
14:06:22LinusNbut the trick is that the function must be called in the gui thread, otherwise the display may be garbled
14:11:49LinusNanother think to think about is how to unify the event scheduling for both rtc and non-rtc targets
14:11:56LinusNthing, even
14:13:20Mmmmnon rtc targets would have to be with countdowns only right? whereas rtc targets could be on an alarm or both...
14:13:59LinusNone approach could be that the api is always using countdown
14:14:26LinusNbut the rtc targets can set an alarm to wake up the unit when the timer expires
14:14:49Mmmmcouldnt you just test compare the current time with the wanted time
14:15:03Mmmmor doesnt the clock work like that?
14:15:25LinusNyou could, but that would take lots of unnecessary i2c traffic
14:15:53LinusNand you still want to set an alarm to be able to turn the unit off while waiting
14:16:15LinusNyou use i2c to talk to the rtc chip
14:16:27Mmmmahh, I see...
14:17:28LinusNregardless of how the time is set, the message queue mechanism could work in the same way on all targets
14:18:32Mmmmyeah, that makes sense...
14:18:34LinusNwhen i think of it, i don't think it's necessary to have a separate timer thread
14:19:11Mmmmyou mean have them combined?
14:19:18LinusNwe could also let the SYS_TIMER event have an argument
14:20:24LinusNall events have a data field that is rarely used
14:21:02LinusNthis field could contain either some sort of ID to keep track of what the event is supposed to do, or even a function pointer
14:22:09Mmmmthat's what i was thinking earlier! :D
14:22:20 Join mirak8 [0] (i=5372a20c@gateway/web/cgi-irc/
14:22:20LinusNthe timer thread would be responsible for re-arming the timer with the next event in turn, but that could just as well be done by the message handling function
14:22:58 Join bluebrother [0] (
14:23:35mirak8linuxstb_ have you looked at ffmpeg mpeg2 decoder ?
14:24:02bluebrotheris the forums stylesheet broken?
14:24:04LinusNso it would basically be set_timed_event(long ticks, void *fn);
14:24:22LinusNbluebrother: in what way?
14:24:54bluebrotherthe forums get displayed strangely for me.
14:25:08LinusNbluebrother: they look allright to me
14:27:33LinusNbluebrother: wow!
14:27:45bluebrotherlooking at the source code links to this style sheet:
14:28:13 Quit webguest14 ("CGI:IRC (EOF)")
14:28:17bluebrotherwhich results in garbage when trying to open it with firefox. Omitting the ?rc3 gives me a readable stylesheet.
14:28:55LinusNi have the same style sheet and it works for me
14:29:36bluebrotherstrange, in konqueror it works.
14:29:56CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
14:29:56*bluebrother looks for that "empty cache" button
14:29:56LinusNi use firefox too, and it looks fine
14:30:30bluebrothernow it works again.
14:30:45bluebrotherstrange, then something with the browser cache did stupid things :(
14:30:49synicthe entire site does that to me occasionally.
14:30:50pondlifeLinusN: Don't suppose you had time to try my patch out with recording yet?
14:31:04 Quit lee-qid ("Trillian (")
14:31:26pondlifeIf not, I'll go for a drive and grab my H300
14:32:52peturpondlife: see irc log, 11:32
14:34:11pondlifeBut it worked without the patch ?
14:34:28pondlifeI know the playback is ok, the sim does a grand job with them
14:34:40pondlifes/playback/voice and playback
14:34:54*petur points to LinusN
14:35:12LinusNpondlife: i didn't try the cvs version without the patch
14:35:45LinusNwant me to do that?
14:35:47pondlifeI'm a little surprised it would influence recording at all, but then I know very little of recording (or indeed Rockbox)
14:35:49pondlifeYes please
14:39:24LinusNmonitoring works without the patch
14:39:36LinusNwhat was your patch supposed to fix?
14:40:06pondlifeThe problem where playback stops and then voice doesn't work, and then any attempt to playback crashes Rockbox
14:40:16pondlifei.e. if playback stops naturally
14:40:24pondlifeAt end of playlist with repeat off
14:40:57 Quit petur ("worrrk")
14:41:08pondlifeThe problem is caused by swap_codec being called when there is no audio codec running
14:41:22pondlifeWhich, of course, never returns
14:41:42pondlifeBug report is here:
14:43:25pondlifejhMikeS probably knows why it breaks recording.
14:45:09pondlifeHmm, I assumed that audio_codec_loaded would be true during encoding too - this is not the case. It's audio_codec_loaded, not audio_dec_loaded...
14:46:39 Nick mooch_ is now known as mooch (n=data@
14:50:47 Quit Bjoern-Erik (Read error: 110 (Connection timed out))
14:55:17 Quit ^BeN^ (Read error: 110 (Connection timed out))
14:55:42theli_uawhat is #define for the h300/h100 >|| (play/pause) button?
14:56:03theli_uaBUTTON_ON ?
14:56:09LinusNah, yes
14:56:52jhMikeShuh...I know something? :)
14:57:43*jhMikeS is just waking up
14:57:43pondlifeMore than me!
14:58:09jhMikeSI should read that bug report I take it? mkay
14:58:20pondlifeYes... I'm attempting a nice take 2 on my voice fix - one that doesn't break recording
14:58:52 Join PaulJam [0] (
14:58:54pondlifeFirst question is: Is playback.c used for monitoring?
14:59:00 Quit bluebrother ("out for real life")
15:00:42pondlifeOK. So it's reasonable that audio_codec_loaded is false.....
15:01:13pondlifeBut it seems that "playing" is set to true.
15:01:13jhMikeSfollow the thread execution in your head...that's what turned up the reason for voice hangs when entering the recording screen
15:01:54jhMikeSok...playing is true when finishing the playlist?
15:02:01pondlifeYes, likely
15:02:24pondlifeI'm a bit confused - need to do some more sim runs
15:02:25jhMikeSmust know for a fact everything its doing
15:03:12jhMikeSsim does this too? I never use the sim
15:03:35pondlifeIt does a good job of reproing the playback crash
15:03:55pondlifeI am attempting to add some dummy recording to it right now.
15:04:11pondlifei.e. encode will run, but input will always be 0.
15:04:42pondlifeI don't have an LCD remote, so real-time logf is tricky to do on the device
15:05:47jhMikeSthe report mentions nothing about recording or crashing. recording would only interfere when actually recording. if you don't go to the recording screen...worry not
15:06:08 Join Febs [0] (
15:06:41pondlifeNo, the problem is my patch. It resolves the reported bug by not calling swap_codec while there's no audio codec loaded. But appears to break recording monitoring. I am trying to work out why
15:06:44*jhMikeS being overly blatherous
15:06:50pondlifeBut don't know enough yet
15:06:56jhMikeSI know why
15:07:05jhMikeSThat's what I changed
15:07:10jhMikeSlook at the diff
15:08:04pondlifeI did, but this is just a 2-liner and I still don't understand.
15:08:53pondlifeAll I've added is a test on audio_codec_loaded, so as to yield rather than swap_codec
15:09:23pondlifeI need to educate myself a lot more
15:09:34pondlifeIf you can help that would be good....
15:09:37jhMikeSwaiting...for commit history to load
15:10:02 Part shadok ("Quit !")
15:10:21 Join MadDog011 [0] (
15:10:26 Part MadDog011
15:12:23jhMikeSThat yield is not needed
15:12:49pondlifeI thought it was, but I'll try again without
15:13:08 Quit Mmmm (Remote closed the connection)
15:13:14pondlifeWhy is it bad to enqueue voice messages when voice isn't the current codec BTW?
15:13:42pondlifeI would think they would just sit on the queue and shouldn't cause problems
15:13:52jhMikeSit can causes voice to retake the mutex before the codec thread get to grab it
15:14:15jhMikeSIt would be easier with preemptive threads
15:14:25***Saving seen data "./dancer.seen"
15:14:33pondlifeNot sure I disagree ;)
15:15:31pondlifeI was hoping that things could run as independently as possible - i.e. maybe the voice thread will take mutex again, but the codec thread should get another chance soon enough.
15:16:02pondlifeI won't remove any, but I do think there are too many ifs in there
15:16:03jhMikeSvoice when waiting on it's queue waits with the mutex locked
15:16:26pondlifeYes, but not forever.
15:16:34jhMikeSit will if there's no message
15:16:50jhMikeSor not playing anything iirc...I just look
15:17:03pondlifeThe problem is that if (playing) should not be there
15:17:20pondlifeVoice shouldn't depend on playing state at all really
15:17:40pondlifeJust like playing shouldn't depend on voice_is_playing!
15:18:44pondlifeI get the impression that over the past few years, there have been many while loops and ifs added to fix individual cases (all of which work and are safe) but which are just overcomplicating what should be simple.
15:19:43LinusNthe software playback engine isn't old enough to talk about "past few years"
15:19:47pondlifeMy previous patch was an attempt to simplify. It just got committed before it was finished.
15:20:06pondlifeSorry LinusN, just joking a bit
15:20:26LinusNbut we all agree that it should be a lot simpler
15:20:26pondlifeDo you not agree that these aspects should run independently?
15:20:27jhMikeSif(playing) probably doesn't need to be there if the message is only sent when loading the codec...I think...:)
15:21:38pondlifeYou mean in the voice queue reader?
15:21:44jhMikeSif current_codec == CODEC_IDX_VOICE then we know the voice codec has the lock
15:22:05jhMikeSthe converse is true with audio codecs too
15:23:14pondlifeYes, but why should enqueuing be affected?
15:23:30pondlifeI'm probably missing something basic here ;-)
15:24:17jhMikeSand (x)_codec_loaded is set outside the lock...
15:24:33jhMikeSHold on...I have to think it all through again! :)
15:26:30jhMikeSdon't do anything that calls swap_codec while the lock is not taken either since swap_codec assumes that the thread with the lock is what's calling it
15:29:18pondlifeYes, but that should be managed by the reading end of the queue
15:29:31pondlifeNot the enqueuing end
15:34:53LinusNthis whole locking/swapping stuff looks really shaky to me
15:35:05pondlifeIt's unpleasant at best
15:35:45pondlifeI assumed it was a necessary evil though.
15:37:39preglowwell, we have to swap, no?
15:37:55jhMikeSif the codec thread can own the mutex, the voice thread cannot process the sent messages anyway...another clue...if you queue a will happen later and voice will swap itself away
15:38:13LinusNpreglow: yes, we can only have one codec in memory at a time
15:39:09jhMikeSit won't be the correct swap though
15:39:48pondlifeWell, you are only really queuing instructions, not specifying anything about swapping....
15:40:10Mordov|hardlywoHELP :) I have been seartching forums all day looking for a xchange battery for my h120, can't find out what to buy,,, could someone point me in the right direction?PP
15:40:13jhMikeSsome instuctions perform swapping
15:40:25jhMikeSwhich is fine
15:40:26pondlifeWhen dequeued
15:41:32jhMikeSbut the message that swaps can be put in the queue when the voice thread cannot process it because it's in swap_codec looping around
15:41:53pondlifeBut it can pick it up next time it gets a chance
15:42:18pondlifeIsn't that the (main) point of the queues? To decouple the threads
15:42:32jhMikeSIt will be for the wrong request
15:42:59pondlifeFIFO... they will get there in the end...
15:43:41jhMikeSonce the codec thread leaves the mutex from playback ending...the voice thread will retake the mutex and find the queued swap message from the previous load
15:43:52jhMikeSnot the one you want
15:44:10pondlifeAre you referring to Q_AUDIO_PLAY?
15:44:33jhMikeSQ_AUDIO_PLAY and anything the would do what it does
15:45:06pondlifeThat's why I put the test in so as not to call swap_codec unless needed.
15:45:13 Quit Mordov|hardlywo ("The amount of experience you have is paralell with the amount of equipment you have ruined.")
15:46:23dongsno wai
15:47:54jhMikeSfor one thing the encoder doesn't flag itself with audio_codec_loaded
15:48:06pondlifeThat's the problem - my test is wrong
15:48:20pondlifeI had assumed the encoder was a codec
15:48:24jhMikeSif (!audio_codec_loaded && current_codec == CODEC_IDX_AUDIO) => is encoder codec
15:48:43 Join tvelocity [0] (
15:49:29jhMikeSthere we're some other reason spanning the entire engine as to why I did that but again I don't recall the specificities. It all seems obvious at the time then I just forget :P
15:49:29pondlifeAh... I thought there were only 2 codecs - voice and audio. And x_codec_loaded is a prerequisite for each to be used....
15:50:01pondlifeSo is there a good reason why the encoder shouldn't have audio_codec_loaded?
15:50:14jhMikeSthe recording hand was due to not swapping if !audio_codec_loaded
15:50:35pondlifei.e. it waits in voice forever
15:50:41jhMikeSI believe there was...yes
15:51:03pondlifeI'd rather have a simple test if we can get away with it.
15:51:51pondlifeThe main other use of audio_codec_loaded is to wait for the codec to stop and start
15:51:59jhMikeSI think queueing the swap should be avoided altogether if voice is not active
15:52:19jhMikeSBut the encoder has its own flag for that
15:52:51pondlifeBut the problem is that the voice may become active after the codec has processed tha
15:53:30pondlifeI maintain that all tests, especially timing sensitve ones, should be on the queue reading side.
15:53:35jhMikeSvoice won't be active during while the encoder is can leave swap_codec
15:53:56jhMikeS*can't leave
15:54:29 Quit daurn (Remote closed the connection)
15:54:58pondlifeI see you currently stop voice explicitly before encoding. Is this needed to stop voice being recorded? In the future it may not be...
15:55:20 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
15:55:30pondlifei.e. Only hardware restrictions should prevent voice during recording
15:55:44jhMikeSVoice can't be running
15:55:47pondlifeWhy not?
15:56:05pondlifeI mean, in theory
15:56:06jhMikeSDMA I believe...
15:56:34pondlifeOK - that's a general Rockbox restriction, not hardware dependent
15:56:38 Join daurn [0] (n=quae@
15:57:06tucozWhen applying replaygain to mp3-files, do I select track gain or album gain for an album?
15:57:53 Join charkins [0] (n=casey@pdpc/supporter/active/charkins)
15:59:00tucozI guess my question depends on how i listen to music. So it will be album gain for now
15:59:26jhMikeStucoz: either way really...your preference
15:59:39tucozI guess :)
16:00:10jhMikeSpondlife: Those stale swap messages have to be ignored for one
16:02:46 Part LinusN
16:02:48jhMikeSBut they should be ignored too if voice is already swapping. This isn't as simple as FIFO and let it be. :(
16:08:26pondlifeI don't see why that is a problem. If voice is already swapping it won't be reading the queue.
16:08:55jhMikeSexactly, the messages will back up
16:09:10pondlifeAnd then be read/discarded as soon as possible
16:09:12daurndoes anyone have Rock The Casbah in at least mp3 192?
16:09:25daurnmy copy is distorting (badly recorded from my LP)
16:09:59 Join rretzbach [0] (
16:10:43 Part Iltsu ("thanks<3")
16:11:32tucozdaurn: I have that song
16:11:56daurngood(ish) quality?
16:12:30tucozfrom the essential clash cd
16:12:37daurncould you dcc it?
16:13:12markuntucoz:in albumgain mode it will also add the trackgain tags
16:14:16tucozmarkun: ok. When I applied albumgain to an album, the volume got a lot lower
16:14:23daurntucoz: not working :S
16:14:26jhMikeSYeah...I think the voice problem is from the stale swap message after the audio codec unloaded.
16:14:37daurnmaybe upload to or such?
16:14:39tucozmaybe it is some firewall issues here
16:15:18lostlogicanyone else having problemw ith recent playback changes making the UI unresponsive until the first track is fully buffered?
16:15:43tucozI think people with ipods experience that
16:16:02lostlogicok, so it's not just me
16:16:15*lostlogic makes faces at slaheri and pondlife
16:18:48tucozdaurn: check your pm
16:18:53jhMikeSx_codec_loaded is not a good test for swapping...current_codec is
16:19:07daurntucoz: i got it
16:19:25daurni'm not idented
16:19:33 Nick daurn is now known as daurnimator (n=quae@
16:24:30pondlifelostlogic: I think that might be the scheduler rather than the playback changes
16:24:50pondlifeMy own build is pre-scheduler and seems ok (H300)
16:25:18jhMikeSit's becoming clearer again...audio_codec_loaded is a good test for swapping on the voice side
16:25:44pondlifeThat's what I concluded
16:26:08jhMikeSsince the voice thread can't run until the codec thread yields while waiting on the mutex
16:26:18pondlifeAnd I'm only really changing the voice side. You can comment out the #define PLAYBACK_VOICE to get a clean (no voice) version for comparison...
16:26:56pondlifeBut with the encoder this isn't true I guess.
16:27:08 Join nls [0] (
16:27:23*jhMikeS thinking through "out loud" :)
16:27:34*pondlife is too
16:29:17pondlifeIt's a bad idea for the voice thread to ever call swap_codec unless audio_codec_loaded=true.
16:30:29pondlifeBut this then breaks recording because audio_codec_;loaded = false and voice never swaps out?
16:31:18jhMikeSyou blocked the swap in Q_ENCODER_RECORD
16:31:46pondlifeI did, but that's been reverted now and is irrelevant to my problem
16:31:52pondlifeThat was just me trying to be clever
16:32:02*pondlife is not clever
16:32:25jhMikeScmon...this isn't really trivial but really should be simplified
16:33:07jhMikeSI'm wondering if audio_codec_loaded should be set to false before leaving the guarded section
16:33:16pondlifeWhich routine?
16:33:24 Join davidc__ [0] (n=davidc@
16:33:37jhMikeSSince it is not loaded anymore
16:33:48pondlifeIt should
16:34:08jhMikeSThe check after the switch will have to be changed
16:34:52pondlifeWhere does the guarded section end then?
16:34:53jhMikeSI guess it's ok if mutex_unlock doesn't yield
16:34:59pondlifeAh, I see
16:35:35jhMikeSno yield before setting it to false !!!
16:35:49pondlifeSorry was looking at the wrong switch()
16:36:49pondlifeSo you don't think audio_codec_loaded should be true during encoding?
16:37:26jhMikeSI'll have to scan the references to it again to see why
16:37:58jhMikeSthis was one reason:
16:37:59jhMikeS if (audio_codec_loaded)
16:37:59jhMikeS {
16:37:59DBUGEnqueued KICK jhMikeS
16:37:59jhMikeS if (ci.stop_codec)
16:37:59jhMikeS {
16:37:59***Alert Mode level 1
16:37:59jhMikeS status = CODEC_OK;
16:38:01jhMikeS if (!playing)
16:38:03jhMikeS pcmbuf_play_stop();
16:38:05jhMikeS }
16:38:07jhMikeS audio_codec_loaded = false;
16:38:09jhMikeS }
16:38:29jhMikeSit will call pcmbuf_play_stop if I set it true
16:39:13 Quit davidc__ (Read error: 104 (Connection reset by peer))
16:39:40pondlifeI have to reboot, back in a bit
16:39:49jhMikeSI put in separate encoder error handling to my next update. ci.enc_codec_loaded can be < 0 to indicate an error
16:40:17 Quit pondlife (Read error: 54 (Connection reset by peer))
16:43:45 Join jamwt [0] (i=jamwt@
16:44:21 Join pondlife [0] (
16:44:32jamwthi.. I was wondering if anyone knew whether anyone was working on supporting the serial interface and apple "accessory protocol" for rockbox/iPod
16:44:48jamwtif not, I might take a crack at it
16:46:04 Quit aliask ("Chatzilla 0.9.75 [Firefox]")
16:46:39markunjamwt: I think noone is working on it, so go ahead.
16:46:47theli_uaif someone is going to commit something soon can he commit also this: ?
16:46:59jamwtmarkun: k
16:48:00***Alert Mode OFF
16:52:39PaulJamSlasheri: do you have an example tagnavi_custom.config ? i have tried to create one, but the item "Custom view..." doesn't show up in the main menu (in the h300 uisim).
16:52:41 Join Mmmm [0] (
16:53:46 Join tucoz_ [0] (
16:54:02 Part tucoz_
16:54:25 Nick daurnimator is now known as daurnimator|afk (n=quae@unaffiliated/daurnimator)
16:55:35 Nick daurnimator|afk is now known as daurnimator (n=quae@
16:57:04daurnimatoranyone care to paste a log of what just happended?
16:57:24PaulJamSlasheri: ok, i got it to work. i didn't know, that i have to specify a root menu in the custom file too.
16:58:09daurnimatorPaulJam: ?
16:58:30 Quit Febs ()
16:59:17 Join _FireFly_ [0] (
16:59:24PaulJamdaurnimator: do you mean an irc-log?
16:59:32daurnimatori mean
16:59:38daurnimatorwhat did i just do/
16:59:44daurnimator- like 3 mins ago
16:59:48perplexity* daurnimator is now known as daurnimator|afk
16:59:48perplexity* daurnimator|afk is now known as daurnimator
16:59:48perplexity<daurnimator> uh
16:59:53perplexitynot much
17:00:42CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:00:42*daurnimator hugs dircproxy
17:00:56 Quit JdGordon ("Konversation terminated!")
17:02:14 Quit theli_ua ("Leaving")
17:02:19daurnimatori logged off & in
17:02:20daurnimatorseems it didn't restart
17:02:20daurnimator- when i rejoined, i got a log
17:02:21DBUGEnqueued KICK daurnimator
17:02:21daurnimator(with timestamps)
17:05:43 Join loco [0] (
17:06:27locohello everyone. is the simulator build (h120) supposed to work? there seems to be an error with the makefile for me.
17:06:42pondlifeThe H300 sim builds ok...
17:07:05perplexityaccording to the cvs build table, so does the h120 sim
17:07:28locoi get an make: *** zxbox: No such file or directory. Stop.
17:07:58perplexitysounds like you need to check out a fresh copy of cvs.. does your tree even have zxbox ?
17:08:15locowell, i cvs upped before.
17:09:05perplexitydo you have apps/plugins/zxbox ?
17:09:25locono, just apps/plugins/zxbox.c
17:09:31SlasheriPaulJam: hmm, weird.. there shouldn't be a need to specify a root menu
17:09:32amiconnloco: Did you cvs up -dP ?
17:09:45SlasheriPaulJam: but you need to name the custom menu as "custom" or it doesn't show up
17:09:46amiconnWithout the -d part, you won't get new directories
17:09:51locoargh... shame on me...
17:09:56Slasherisomething like %start_menu "custom" "My menu"
17:10:08locothanks a lot.
17:10:13Slasheriin fact it was %menu_start
17:11:22 Quit tucoz (Read error: 110 (Connection timed out))
17:11:22PaulJamSlasheri: that was what i tried first, but it didn't work. it worked when i added 'root_menu "custom"' at the end of the file.
17:11:27 Quit bytie ("Dana")
17:11:48 Quit XavierGr (Nick collision from services.)
17:11:51 Join XavierGr [0] (
17:12:48SlasheriPaulJam: hmm, interesting.. could you paste a copy of the file somewhere?
17:12:50 Quit loco ("bye and thanks for the help")
17:12:56jhMikeSwhy do we need a 512KB codec swap buffer for the voice codec when mpa_codec is only 73KB?
17:13:02SlasheriPaulJam: oh, and did you add the empty line after the menu items?
17:13:08Slasherithat is necessary for it to work
17:13:42SlasheriPaulJam: if it's the end of file, you need to empty lines there
17:13:49pondlifejhMikeS: because no-one thought about it?
17:14:19pondlifeI guess it could be that the voice codec may be something other than MP3 one day...
17:14:20jhMikeSI think it should check the file size before dividing up the buffer IMHO.
17:14:28***Saving seen data "./dancer.seen"
17:15:35jhMikeSpondlife: but it would probably always be the same codec so we'll know how much space it needs. The voice shouldn't be too elaborate in that regard.
17:15:43pondlifeYes, I agree
17:16:22amiconnjhMikeS: I won't necessarily always be the same codec
17:16:35CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:16:35*amiconn is waiting for preglow's magix speex decoder ;)
17:16:45amiconn*magic even
17:16:46jhMikeSHow many voice codecs do we need to use at once?
17:16:59pondlifeYes, but you'd use speex for alll voice in that case..
17:17:05jhMikeSI think a voice synth that sounds like SpeakNSpell would be cool or like HAL
17:17:09 Join bbroke [0] (
17:17:15jhMikeSThen we know the codec size
17:17:33pondlifeWell, you could get a robot arm to operate a SpeakNSpell, and generate voice clips that way... ;-)
17:17:36jhMikeSJust check the file size for voice codec and give it just enough
17:17:38pondlifeVery 1980s
17:17:53amiconnDuring transition, it might be necessary to load either mpa.codec or speex.codec depending on what voice file is found
17:17:58jhMikeSpondlife: of course! I still have mine
17:18:26pondlifeDoes it still mis-spell colour?
17:18:36jhMikeSok...then we have a list of possibilities and leave enought room for the largest of them...but they don't even approad 512KB
17:18:52jhMikeSIt's "color" stateside :)
17:18:58amiconnYes, but check how much ram mpa.codec actually takes
17:19:11amiconnThe 73KB of the binary image isn't everything
17:19:25amiconnThere's also the .bss segment
17:19:51PaulJamSlasheri: i had only a linebreak at the end of the file. i thought that conts as an empty line. with two linebraks it works (without specifiing a root) how much RAM does it take?
17:19:53pondlifeOf course, a smaller CODEC_SIZE would also reduce the memcpy-ing a bit.
17:20:05pondlifei.e. for voice swap-ins
17:20:23amiconnpondlife: Yes, but we could even save one of the swap buffers completely
17:20:32amiconn...with a special copy-swap routine
17:20:40jhMikeSamciconn: xor of course
17:20:44pondlifeGood idea
17:20:49amiconnjhMikeS: huh?
17:21:16pondlifeXOR for register swap...
17:21:18jhMikeSyou had something else in mind...both codecs get swapped on USB connect
17:21:28SlasheriPaulJam: nice :) but a good note, i think that is a bug and i will try to fix that
17:22:03amiconnJust load one memory line from the codec buffer into 4 registers, load one line from the swap buffer into another 4 registers, then write back in the opposite destination
17:22:16*jhMikeS was making a joke about his "amiconn loves eor/xor" :)
17:22:34pondlifeWell, who doesn't!
17:22:41preglowamiconn: it's on the tracker
17:22:51pondlifeA nwew sitcom - "Everyone Loves XOR"
17:22:52preglowamiconn: i don't have time to fix it to stop throwing warnings
17:23:13preglowfreqmod will do it once he's got time again
17:23:21amiconnjhMikeS: Xor is often very useful...
17:24:31jhMikeSYou can xor the codecs togther to swap them
17:24:58amiconnHehe, yes, but doing the read twice, write twice thing is faster
17:25:56pondlifeHmm, I don't like the loud beeping on skip tracks when you're at the end... is that new?
17:25:58jhMikeSWell...we have to watch the USB connect thing then
17:26:06pondlifeplayback.c - audio_next.
17:26:17amiconnjhMikeS: USB connect thing?
17:26:47amiconnpondlife: The beep isn't new. It's an option though, perhaps you enabled it by accident
17:27:21pondlifeNope, not the beep
17:27:51pondlifeSlasheri added a second beep (lower pitched) so you can tell if you're skipping and you've reached the start or end of a playlist
17:27:58pondlifeAdded in here:
17:28:01*amiconn doesn't use the swcodec targets enough to notice such things...
17:28:13pondlifeI prefer it not to beep at all in this case.
17:28:27pondlifeBeep should probably be a keyclick anyway - for feedback of a keypress.
17:28:40pondlifeI don't really think it's a playback thing per se.
17:29:00pondlifeThe lower beep is much louder sounding than the normal one, probably due to hearing non-linearity
17:29:17amiconnThe archoses run nice and stable, and there isn't a single feature of the newer targets that I actually need
17:29:39amiconnSo... all the newer targets are just there for the fun of coding
17:30:11pondlifeI love my Archos too, but need to fix it one day..
17:31:03amiconnOnly 50% of my targets are archoses, but I use them like 90% of the time I'm using daps
17:32:50jhMikeSamiconn: each codec swaps the other on USB connect...why? this seems like a collision given the way swap_codec works
17:33:43amiconnjhMikeS: Odd. So it constantly swaps codecs back and forth during USB connection?
17:34:35amiconnWhen playback is stopped, the main codec isn't needed, so the voice codec is what should be swapped in, and stay in until playback is started again thread swaps if voice_codec_loaded...and vice versa
17:35:45jhMikeSI mean codec thread not audio thread
17:36:06pondlifeOf course, the voice codec is also required sometimes during playback.
17:39:09amiconnYes, but the main codec isn't needed when playback is stopped
17:39:11jhMikeSwhy should they each swap the other if the other is loaded upon connect? ??
17:40:48 Quit aegray ("Lost terminal")
17:42:06amiconnpondlife: Hmm, maybe the second beep doesn't obey the beep volume setting?
17:42:26pondlifeIt sounds like it's playing two beeps - the low and high one together!
17:42:43pondlifeYou may need to listen to it and you'll see what I mean
17:42:51mirak8is there a way to keep the H300 connected to usb and having the device not in UMS ?
17:44:04amiconnmirak: Yes, hold A-B while plugging USB
17:44:17amiconnThen it just charges from USB (slowly)
17:44:49 Join bytie [0] (
17:44:52mirak8damn I didn't knew that :-/ It could be written on the screen while connected ^^
17:45:23mirak8amiconn while in charge mode can you listen music ?
17:46:52markunof course
17:47:47mirak8a-b ? doesn't work for me
17:49:06amiconnYou need to press & hold it while plugging. Once in USB mode it's too late
17:49:30amiconnMaybe it's another button for iriver... don't remember exactly
17:49:33markunmirak8: "Enable USB charging mode on H300, hold Rec when inserting the USB connector to not enter USB mode. This may not always work, since we don't yet know how to control the charging chip correctly."
17:49:41amiconnAh, so Rec it is
17:50:28preglowugh, beep
17:50:29preglowremove it!
17:51:09mirak8ok that's when inserting, what I am looking for is when rockbox is running, if I plug the usb I go to usbmode, then I want to go back to rockbox without unpluging the usb cable.
17:51:49markunmirak8: that doesn't work yet, but feel free to implement it
17:51:53PaulJamSlasheri: wouldn't it be better (and maybe simpler in the code) to completely ignore the tagnavi.config if a tagnavi_custom.config (with the correct version number) exists in the .rockbox folder? this way people who want to use (only) custom views and don't need entrys like for example "Composer" can set their custom view as root menu.
17:52:40 Quit XavierGr (Read error: 113 (No route to host))
17:53:11mirak8markun if it's technically feasible that's intersting. I would use that mostly when I build a version to test without removing the cable each time
17:53:31markunmirak8: but how to enter USB mode again..
17:55:23mirak8markun iriver firmware can go from charging to usb mode by pushing play.
17:55:28pondlifepreglow: Did you mean the extra beep I was referring to?
17:55:38SlasheriPaulJam: Hmm, maybe.. i have also thought that it would be possible to remember the last menu on boot
17:57:36preglowpondlife: all beeps!
17:57:49preglowpondlife: but especially ones you can't turn off, yes
17:58:08pondlifeHmm, does it not turn off?
17:58:13pondlifeI didn't realise that
17:58:24pondlifeI use beep=strong
17:58:33pondlifeGood when driving
17:58:37 Join solarflare [0] (
17:58:49preglowi might not know what you're talking about :>
17:59:55pondlifeI often have that effect...
18:00:20jhMikeSmaybe we should have a memswap function?
18:00:22pondlifeI meant the intentional (and configurable) beep on file skip
18:05:14amiconnSlasheri: I think it would be better to embed the default tagnavi.config layout, and not ship a tagnavi.config file at all
18:05:53amiconnSame thing as we do with the default wps
18:06:05Slasheriamiconn: hmm, true. but i am not sure how to do it.. just embed the file content to the code or something like that?
18:08:24PaulJamamiconn: but with a separate file it is easier to create custom views, because you have a good example file.
18:10:41 Join davidc__ [0] (n=davidc@
18:12:44 Nick daurnimator is now known as daurnimator|afk (n=quae@unaffiliated/daurnimator)
18:12:44DBUGEnqueued KICK daurnimator|afk
18:12:46 Nick daurnimator|afk is now known as daurnimator (n=quae@
18:12:46DBUGEnqueued KICK daurnimator
18:12:46 Join lodesi [0] (
18:16:26 Join merbanan [0] (
18:20:06 Join linuxstb__ [0] (
18:21:56 Quit linuxstb_ (Read error: 110 (Connection timed out))
18:22:57 Join bluebrother [0] (
18:25:12 Quit davidc__ ()
18:25:32 Quit Mmmm (Remote closed the connection)
18:27:41 Nick JBonBike is now known as JoeBorn (
18:30:00 Join MarcoPolo [0] (
18:36:15 Join ender` [0] (i=null@
18:36:40 Join pixelma [0] (i=pixelma@
18:41:49 Join lini [0] (i=pugsley@
18:45:10preglowcould anyone knowledgable with playback.c please add back the boost on read functionality?
18:45:50nlsIf I want to change a strig in english.lang do I edit <dest> and leave <source> as is?
18:52:12 Quit Strath (Read error: 54 (Connection reset by peer))
18:52:52pondlifepreglow: Can you explain more...
18:52:57pondlife"boost on read"
18:53:54 Join lightyear [0] (
18:54:13 Quit tvelocity ("Αποχώρησε")
18:54:26nlsboost when buffering from the disk?
18:54:46pondlifeAny idea when this was changed?
18:54:54 Quit ender` (Read error: 113 (No route to host))
18:55:43nlsafaics it still does it −− boost is 100% in audiothread while buffering...
18:55:49pondlifeShould do
18:57:11Slasherihmm, this is interesting. I was also able to reproduce the PDIR1FULL error at 0x3 with my h140 while building tagcache
18:57:29Slasherithis has _something_ to do with the new scheduler, not yet sure in what way
18:57:33 Join ender` [0] (i=null@
18:57:48preglowpondlife: well, boost on read
18:57:51preglowpondlife: boost while reading
18:57:54preglowpondlife: from the disk
18:59:02Slasherii wonder how it is possible to get such an error as PDIR1 shouldn't be enabled by default
18:59:05pondlifeYou implied it has changed... recently?
19:01:26preglowpondlife: talking to me or him?
19:02:13pondlifepreglow: You
19:03:21preglowpondlife: it used to boost some time ago, before lostlogic did some work on it
19:03:26Slasherithis seems good.. that pdir error always at same point, now at 0x2
19:03:29pondlifeAh, ok
19:03:42preglowpondlife: imho, it really should boost to minimise disk on time
19:03:54pondlifeI agree
19:04:01preglowpondlife: if the codec is very close to boosting on its own, rockbox can spend ages and ages reading
19:04:21preglowits gotten worse after the new scheduler as well, i think
19:05:27preglowyesterday i had a case where rockbox started switching the disk on and off continously when buffering was to take place
19:06:11preglowSlasheri: could that have been the new scheduler starving whatever thread is reading from the disk?
19:07:04Slasheripreglow: hmm, currently codec and buffering should have the same priority so they should get the equal amount of cpu by scheduler
19:07:34Slasheribut if buffering thread does not give codec enough cpu time, then codec thread's priority will be changed higher
19:07:39Slasheriand that will starve everything
19:07:49preglowwell, this wasn't exactly pretty
19:07:51pondlifepreglow: If you can reproduce, maybe try putting a cpu_boost(true) in at the start of audio_read_file(), and a corresponding cpu_boost(false) at the end?
19:08:07preglowcan't reproduce it, no
19:08:11pondlifeAh, ok
19:08:12preglowdon't even remember what codec i used
19:08:23preglowhmm, it probably was ogg
19:08:49amiconnHmm. Imho the codec thread should _never_ switch to realtime unless the CPU is already boosted
19:09:10preglow_realtime_ ?
19:09:13preglowthere is an actual realtime prio?
19:09:16Slasheriamiconn: cpu should be boosted before that switch
19:09:18 Quit bytie ("Dana")
19:09:40Slasheriat least boosting watermark is much higher than the priority switching watermark level
19:09:44 Join lowlight [0] (i=c730190a@gateway/web/cgi-irc/
19:09:57Slasheriand with iriver that seems to work correctly.. but maybe ipod behaves differently
19:10:02preglowthis was on iriver
19:10:09Slasherihmm, weird
19:10:15pondlifepreglow: I would aim to repro it if you can
19:10:17preglowbut yeah, do we actually use a realtime prio?
19:10:36Slasheriyep, when the pcm buffer data is really at critical
19:10:40preglowthat's a real no no, if you ask me
19:10:50Slasheriat the point unless we switch priority to realtime, audio will skip
19:10:54 Join obo [0] (
19:10:56preglowif we actually need to use realtime prio, then we're doing something seriously wrong
19:11:22preglowbetter to let it skip
19:11:32preglowi don't want to think about what my threads do when they're never called anymore
19:11:46Slasherihmm, then the priority scheduling would help only background tasks
19:11:49pondlifepreglow: Also pop a bug report on Flyspray?
19:11:53pondlifeI gotta go
19:11:54 Quit pondlife ()
19:11:58Slasherirealtime priority is switched back as soon as possible.. but other threads are still called
19:11:58preglowpondlife: i don't really have much details...
19:12:04Slasheribut just less often
19:12:12lowlightpreglow: boosting while buffering taken out around here... @ 19.53.26
19:12:19Slasherisomething like 1-3 times per second depending on the priority
19:12:35preglowSlasheri: well, then what i encountered pretty much has to be some other bug
19:12:39preglowbut i can't possibly imagine what
19:12:54Slasheriyes, that sounds like a bug
19:12:55lowlightpreglow: 19.53.54 # <amiconn> If so, that should besolved soon by a global boost-on-disk-read in the ata driver
19:13:13preglowthe file buffer was really low and it did try to read all the time
19:13:23 Join Arathis [0] (
19:13:27preglowbut it seemed to go really long between each read
19:13:52Slasheriplayback thread no longer boosts when buffering?
19:13:54Slasherithat is really bad
19:14:01preglowit hasn't done so since lostlogic did his stuff
19:14:11preglowagreed, very bad
19:14:19Slasherireally weird.. no wonder if that causes problems
19:14:31***Saving seen data "./dancer.seen"
19:14:33preglowi'm not convinved a boost in ata read is the correct solution, though
19:14:36preglowbut it doesn't sound very bad either
19:14:44amiconnWhy not?
19:15:02Slasheriand another place to do it would be the scheduler.. :)
19:15:07preglowamiconn: depends on how the reading is done
19:15:08Slasherii always experimented a little with that
19:15:22preglowamiconn: if lots of small blocks are read instead of one big, for example
19:15:27 Quit idnar (Nick collision from services.)
19:15:29 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
19:15:32preglowamiconn: we don't want to do a lot of boosting and unboosting
19:15:50amiconnOf course it shouldn't unboost immediately
19:15:53 Quit damaki (Read error: 104 (Connection reset by peer))
19:16:18Slasheripreglow: something i tried, was that unless cpu has been last in sleep withing 500ms, switch boost on. And then unboost as soon as we put core in sleep again
19:16:21amiconnIt should boost at the first read, and then start a timeout for unboosting
19:16:21preglowwell, if that can be made to work reliably, then that should be good
19:16:44 Join damaki [0] (
19:16:46Slasheriand boosting can be triggered manually also.. something like thread_trigger_boost();
19:16:48preglowSlasheri: i say we wait until we have removed some of the bugs already in the new scheduler until we start adding new stuff
19:16:52preglowit'll only make debugging harder
19:16:54Slasheriand no need to remember to unboost the cpu
19:17:04Slasheripreglow: yes, of course
19:17:12Slasherii am now investigating that PDIR issue
19:17:20amiconnI think auto-boosting in the scheduler would be bad
19:17:43amiconn(1) It would probably boost/unboost very often, defeating timer precision etc
19:18:07Slasherihmm, so boosting could affect the timer precision also?
19:18:15amiconnThere are also some places in the code where we must keep the cpu boosted for precision, so the manual boost has to stay anyway
19:18:36amiconnYes, on coldfire
19:18:39Slasheriamiconn: manual boost would always work as before too
19:18:51Slasheriif cpu is boosted manually, scheduler would not unboost it
19:18:54amiconnThat's the reason why backlight fading needs to keep the cpu boosted
19:18:57Slasheriit just calls the cpu_boost
19:19:02Slasheriyep, true
19:19:32amiconn(2) Boosting in the scheduler will probably boost more often than necessary, lowering battery runtime
19:20:09mirakany mpeg compression expert around ?
19:20:13preglowmirak: depends
19:20:25preglowi know something, but i'm not an expert
19:20:26Slasherihmm, really? for example if core has not been in sleep state during the last 500-1000ms and then boosting it..
19:20:47 Join MarcoPolo_ [0] (
19:20:52amiconnThe core not sleeping doesn't necessarily mean that it needs to be boosted
19:21:14Slasheritrue, but yet i can't think of any situation where it wouldn't mean that
19:21:19amiconnIt might just mean that the performance is just enough to keep up
19:21:47preglowin which case we'd definitely not want boosting
19:21:49 Join TeaSea [0] (
19:21:54amiconn...or there might be an unavoidable polling loop, or a data transfer loop...
19:22:25 Quit ScoTTie (Connection timed out)
19:22:25 Quit MarcoPolo (Read error: 113 (No route to host))
19:24:19mirakpreglow: this is about sparse DCT matrix. I was wondering what's the rate of zeros coefficient in this matrix. I know the goal in compression chain is at some point to have as much zeros as possible, however, I am not sure if quantification bring back some numbers.
19:25:53preglowbring back numbers?
19:26:04preglowquantisation can't bring back numbers
19:26:37mirakpreglow: if there is really a lot of zeros, then it could be valuable to skip 4 mac operations per 0 coefficient. however since 4 mac operations are 4 cycles, I am not sure a AND plus BRANCH would cost less in the end
19:26:42preglowit depends what matrix you're talking about. if you're talking i-frame transforms, then the ratio of zeros to non-zero numbers is usually very big
19:26:49mirakpreglow: quantisation is just about multiplying ?
19:27:27preglowthe matrices are actually coded as runs of 0s
19:27:33preglowbecause they're so many
19:27:52preglowit's a very common optimisation to check for zeros, but i don't know if it'll matter for coldfire
19:27:56mirakI know ffmpeg idct uses this skip if a coef is zero but, they don't have a mac to do the operations
19:28:10preglowif the data is in iram
19:28:14preglowi don't know if i'd bother
19:28:25amiconnI think that with emac, testing for zero would slow down things
19:28:29preglowi almsot certainly wouldn't
19:30:46 Quit PaulJam (".")
19:31:08 Quit charkins ("Leaving")
19:31:14mirak8 coefficients of a row are stored in upper and lower parts of 4 registers. So I would do eventually an AND 0000FFFF on upper part to cancel the upper part and then test lower part, however I am toasted for coef in adress registers ...
19:31:54mirakand the opposite to test upper part ...
19:32:53amiconnYes, and then you test, and if it's zero, you skip the mac.w with a branch... even the branch alone is slower than just performing the mac.w
19:33:24amiconnmac.w is single cycle, a conditional branch forward eats 3 cycles
19:33:45amiconnI'd say forget that
19:34:01mirakI would skip 4 mac
19:34:23amiconnYou still wouldn't gain anything
19:34:25mirakbut well if branch is 3
19:34:36amiconnmac/mac/mac/mac/movclr = 5 cycles
19:34:53amiconnand/test/branch is also 5 cycles
19:35:14mirakthe 4 skipped mac are on 4 different accumulators
19:35:34amiconn...and if you need to copy to a data register to do the check...
19:35:36mirakso the moveclr shouldn't be counted, but of course that's still 4
19:35:59mirakI can eventually not check that adress register ;)
19:36:14mirakok that's not worth, that's what I need to be sure
19:36:24amiconnThe zero row check is for CPUs with slow multipliers, or even no hardware multiplier at all
19:38:04CtcpIgnored 2 channel CTCP requests in 2 seconds at the last flood
19:38:04*amiconn wonders whether it would be worth it on arm
19:41:34mirakif zeros are really that much present I could eventually test if a full register is zero and skip 8 mac directly
19:42:40 Join Rondom [0] (
19:42:49mirakthat's the best I can think off ^^
19:42:50preglowamiconn: mla on arm would take about 3 cycles for an nn x 16 bit multiply, i think
19:44:31 Join klrspz [0] (n=klrSpz@
19:48:01markunpreglow: at least on ARM9 mla only takes 1 cycle if one of the operands is 0
19:52:54preglowahh, right
19:53:03pregloware you sure you don't mean 2?
19:53:17preglowi don't think the early termination triggers on the lowest byte
19:53:21preglowonly the top three bytes
19:56:05 Quit MarcoPolo_ (Read error: 60 (Operation timed out))
19:59:15 Join hardeep [0] (
20:01:15*amiconn wonders where preglow found those detais
20:03:26 Join S0ap [0] (n=Soap@unaffiliated/s0ap)
20:04:30Slasheriah, this is good. In fact the new scheduler makes rockbox to crash with bad mp3 files easier than previously
20:04:37Slasheriso now it's possible to fix that problem
20:07:42preglowamiconn: DDI0210C_7tdmi_r4p1_trm.pdf, section 6.6
20:07:52preglowSlasheri: how/why?
20:08:19Slasheripreglow: i don't know why, but that PDIR1FULL error is definately caused by a bad mp3 file (i have now found the file)
20:08:27Slasheritrying to reproduce it on simulator now..
20:08:45Slasheribefore the scheduler patch, no such files crashed the unit immediately
20:09:49 Nick daurnimator is now known as daurnimator|afk (n=quae@unaffiliated/daurnimator)
20:09:49DBUGEnqueued KICK daurnimator|afk
20:13:03 Join aegray [0] (
20:20:02 Quit Soap (Read error: 110 (Connection timed out))
20:20:54 Quit bbroke ("Chatzilla 0.9.75 [Firefox]")
20:29:03 Quit mirak8 ("CGI:IRC (Ping timeout)")
20:29:13 Join Paul_The_Nerd [0] (
20:29:31 Join Zagor [0] (
20:31:27 Quit damaki (Read error: 104 (Connection reset by peer))
20:32:37 Join damaki [0] (
20:32:38 Quit Nibbier (Read error: 104 (Connection reset by peer))
20:34:32 Join Nibbier [0] (
20:35:24 Join damaki_ [0] (
20:45:07 Quit mirak (Connection timed out)
20:45:20 Join mantono [0] (
20:46:54 Join darkless [0] (n=darkless@
20:48:15 Join mirak [0] (
20:48:45 Join bytie [0] (
20:51:30 Quit damaki (Read error: 110 (Connection timed out))
20:52:20CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
20:52:20*Bagder crawls in
20:52:20 Quit Rob2222 (Read error: 104 (Connection reset by peer))
20:53:28 Join Rob2222 [0] (
20:54:45*obo wishes his congratulations
20:55:18 Join XavierGr [0] (
20:57:21mirakhow can it be explained that writting to memory takes less cycles than reading from memory ?
20:57:40Paul_The_NerdWow, the tagnavi syntax just changed a lot.
20:58:00Paul_The_NerdWell, "increased"
21:00:50amiconnmirak: When reading, the CPU has to wait for the data
21:01:20*preglow hands bagder a cigar
21:01:25dpassenPaul_The_Nerd: are the changes documented?
21:02:17mirakamiconn: ok then optimising writes is pretty useless then ?
21:02:35amiconnWriting much data in sequence is still slow
21:03:32amiconnBut writing single data items is faster than reading, because there usually is a store buffer
21:03:53 Quit aegray (Read error: 110 (Connection timed out))
21:04:27amiconnOf course this only holds for memory writes; writing to an i/o area is usually unbuffered
21:04:45Paul_The_Nerddpassen: They're documented by way of "you can read the file and see how it works" really.
21:05:02dpassenWill try an updated build later
21:05:49mirakso in my code I need to write 4 bytes coming from 4 different computations, I put them in a register and do a left shift. Then move.l to memory. in my case it's more a movem.l for 8 bytes. Is it still better than doing 8 byte moves ?
21:06:16amiconnyes, it should
21:06:43mirakdoes it change something to the problem if it's iram or sdram ?
21:06:52amiconnHowever, using movem.l for 8 bytes instead of two move.l isn't faster
21:07:22 Quit hcs (Read error: 110 (Connection timed out))
21:07:34mirakamiconn: doesn't it save one instruction cycle at least ?
21:07:44amiconnIt needs one cycle extra
21:07:56mirakone extra cycle per registry ?
21:08:02amiconnmovem.l needs 1+n cycles, where n is the number of registers
21:08:14mirakah I see
21:08:53mirakok, I can use only one register instead of two then.
21:09:23amiconnIn iram. it might even be that the byte stores are faster
21:10:22mirakthere is only burst read, not burst write ? One could think that a movem keep opened the "channel" to ram
21:10:25amiconnWith long stores, you need 3x lsl.l + 3x or.l + move.l = 7 cycles
21:11:07mirakwhy do you use or ?
21:11:08 Join aegray [0] (
21:11:10amiconn4x move.b is only 4 cycles - if you store to iram
21:11:25mirak"move.b %d0,%d2\n\t"
21:11:25mirak"lsl.l #8,%d2\n\t"
21:11:27amiconnReplace or.l by move.b, same timing
21:11:55preglowSlasheri: was that it?
21:12:21amiconnRegarding burst mode - the bus controller can only do certain burst lengths, namely 1, 2, 4 and 16 bytes. Not 8 bytes
21:12:33mirakhuh lol
21:12:36amiconnSo an 8 byte store is split in 2 4-byte-bursts
21:14:16mirakamiconn: when using the or however you need to have a clean destination registery before starting. you don't need that with a move and a lsl
21:14:33***Saving seen data "./dancer.seen"
21:14:55amiconnYes, but that wasn't the point
21:15:17amiconnYou still need 3x lsl.l + 3x move.b + move.l = 7 cycles
21:20:10 Join Lmcmann [0] (i=c0553201@gateway/web/cgi-irc/
21:20:10 Quit thegeek (Read error: 104 (Connection reset by peer))
21:20:14Slasheripreglow: yep
21:20:29Slasheripreglow: no longer crashes at least on my player
21:20:30 Join thegeek [0] (
21:22:21 Join pondlife [0] (
21:25:36BigMacanyone here use pandora
21:26:15preglowSlasheri: why did the scheduler provoke that?
21:26:19preglowSlasheri: or was it a coincidence+
21:26:30 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
21:27:07merbananlinuxstb_: the wma patch was a big mess
21:27:52linuxstb_I haven't looked at it in detail, but I'm not surprised...
21:28:06merbananI'll try to slim it down when I get time
21:28:31linuxstb_What do you mean by "slim it down" ?
21:28:57merbananit's lots of unneeded ffmpeg code in there
21:29:06linuxstb_Good - I was hoping you meant that.
21:29:13 Join CriamosAndy [0] (
21:30:56 Join davidc__ [0] (n=davidc@
21:36:32 Quit Rondom ("bb in 5 mins")
21:41:03 Part Lmcmann
21:41:28 Quit lightyear (Connection timed out)
21:41:51 Join lightyear [0] (
21:44:08 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
21:44:26Slasheripreglow: i think it was just coincidence
21:44:31 Join Lefungus [0] (
21:44:39 Quit XavierGr ()
21:44:41Slasheribut a good thing indeed
21:45:24 Quit Criamos (Read error: 110 (Connection timed out))
21:47:27 Quit lightyear (Remote closed the connection)
21:48:47Lefunguspreglow, need some feedback from you on mpc and seeking benchmarks
21:49:17 Join MadDog011 [0] (
21:50:14Lefunguslike how much granularity the seek table need ?
21:50:36Lefungusie, how much seek points per file approximately
21:54:48 Part mantono ("Don't worry, be happy")
21:56:53 Part MadDog011
21:56:53 Quit solarflare (Read error: 104 (Connection reset by peer))
21:58:34 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
22:04:04 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
22:04:15 Quit davidc__ ()
22:05:11 Quit rretzbach (Connection reset by peer)
22:05:24preglowLefungus: wouldn't know
22:05:45preglowLefungus: for pc use, you can get away with pretty many frames per slot, i'd guess
22:06:06Lefunguswe want it to work for both, as we may store a seek table in files directly
22:06:35 Quit aegray (Connection timed out)
22:06:47Lefungusmpc_uint32_t seek_table[10000];
22:07:09preglowwell, there's a tradeoff
22:07:13Lefungusfrom mpc.c in rockbox cvs, I interpret that as a 10000 entries seek table
22:07:14preglowyou don't want the seektable to bloat the file
22:07:34preglowLefungus: yeah, but that's per file, no matter how big it is
22:07:41tucozcongratulations Bagder
22:07:50preglowLefungus: if it's a long file, there'll be more frames per entry
22:08:07 Quit bytie ("Dana")
22:08:11preglowBagder: we were wondering if you were going to call him sansa :)
22:08:20tucozor e200
22:08:23Lefungusok, I just meant, how many entries per file, not how many samples per entry
22:08:26Bagderof course
22:08:31Bagdermy wife would love it! ;-)
22:08:39preglowa woman loves dedication in a man
22:08:46preglowespecially to hobbies
22:08:49tucozawww, what a pretty name. what does it mean?
22:09:08preglowLefungus: god knows, you pretty much have to see how much is practical
22:09:22preglowLefungus: i really don't have many musepack files yet so i haven't gotten to use it much
22:09:22Lefunguswell, is it working nicely enough currently ?
22:09:28preglowLefungus: sure, people say it is
22:09:35preglowi really haven't used rockbox myself for a pretty long while
22:09:40preglowdon't travel much these days
22:09:57Lefungusdo you happen to know whether the seek table is stored in internal or external memory ?
22:10:10Genre9mp3Bagder: Best wishes for the new family member! :)
22:10:14 Quit linuxstb__ (Read error: 110 (Connection timed out))
22:10:16preglowLefungus: you mean as in sram or dram?
22:10:25mirakis there on coldfire a function that reverse all bytes of a long word ?
22:10:41Lefungusdunno how you call it, internal=small&fast, external=big&slow
22:10:43preglowmirak: endian swap? sure
22:10:47mirakI remember something like that but can't find
22:10:49preglowdram, definitely, no reason to store anything like that in fast ram
22:10:52 Quit merbanan (Remote closed the connection)
22:10:54pondlifeBagder: Congratulations. I expect we'll see you on here at all sorts of odd hours now!
22:11:01mirakpreglow: yes, this would save me.
22:11:26Bagderyeah, they say two kids give you more spare time than just one ;-P
22:11:43preglow[04:34] * Bagder tears his hair out!
22:11:51pondlifeMaybe once one is old enough to look after the other...
22:11:56YouCeyEhello.. how can i change h10 firmware to rockbox?
22:12:03YouCeyE20gb model
22:12:41Paul_The_NerdDoes the H10 Rockbox manual have installation instructions yet?
22:13:00tucozno, unfortunatley not
22:13:17YouCeyEtucoz, is it fully functional now? i had iriver h120 and rockbox porting was easy
22:13:28preglowpondlife: or the other is old enough to be able to be safel looked after by the other...
22:13:43tucozYouCeyE: I think it's fairly usable
22:13:59preglowsomeday i'll learn myself english
22:14:18mirakpreglow: can't find it back
22:14:22pondlifeNah, safely is a subjective term
22:14:30Bagderwe should switch to norwegian as official Rockbox language!
22:14:34preglowmirak: system.h
22:14:35preglowmirak: swap32
22:14:51YouCeyEanyway the stock firmware sucks on h10.. i will just upgrade
22:15:02pondlifeIs there a nowegian port of C?
22:15:19mirakpreglow: in assembly ?
22:15:23preglowi've seen basic with norwegian keywords, heh
22:15:25preglowmirak: yes
22:15:51tucozIs C available in other "languages" than english?
22:16:14ArathisYouCeyE: radio and recodring isn't implemented yet in rockbox on h10
22:16:14 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
22:16:17preglowif you can't memorise what "if" means in other languages, you shouldn't program
22:16:19tucozwould that even be called C?
22:16:21preglowtucoz: never seen it
22:16:28preglowtucoz: but microsoft has almost certainly tried
22:16:39tucozfor sure
22:17:04mirakpreglow: can't find it in the manual. is that one cycle operation ?
22:17:07preglowpondlife: i've seen ancient footage of myself "taking care" of my little brother...
22:17:17preglowpondlife: nearly choked him a couple of times
22:17:26YouCeyEArathis, is mp3 playing fully functional? just like other
22:17:35 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
22:17:36Paul_The_Nerdhardeep: I *think* I know how the Insert situation can be improved a lot. Or made more complex. I'm not really sure. I responded in the forums again.
22:17:39preglowmirak: it's not an instruction, it's a routine
22:18:03preglowmirak: line 284
22:18:19pondlifeI threw a plank at my little brother and he ended up in hospital.
22:18:44Paul_The_NerdYouCeyE: On the H10 battery life is still not great, and music may skip if your WPS is too complex (primarily peakmeters).
22:18:48pondlifeYears later he threw a Rubik's cube at me and broke my front tooth
22:18:54pondlifeSo we're even
22:18:57mirakwell I want it to reverse bytes and get them one by one with a move.b and a right shift, since we can't rotate bits
22:19:04ArathisYouCeyE: if you use a wps without peakmeter and not an EQ it works very well. cpu booting isn't working fine.
22:19:23Paul_The_NerdIsn't working fine?
22:19:25Paul_The_NerdWhat's wrong with it?
22:20:36preglowi was all nice to my second little brother, though, probably because i wasn't too keen on strangling siblings i'm twelve years older than
22:20:53YouCeyEwould u suggest me to wait for sometime for full fledged h10 firmware?
22:20:55Paul_The_NerdArathis: I thought CPU boosting worked fine on the H10s now.
22:21:17YouCeyErockbox firmware works in ums mode?
22:21:20YouCeyEor ptp?
22:22:25Paul_The_NerdI believe
22:22:27ArathisUSB doesn't work, but if you have an MTP player UMS trick works, just you don't need to reset
22:22:56 Join petur [0] (n=petur@rockbox/developer/petur)
22:23:05Paul_The_NerdArathis: Is it posted somewhere about the CPU boosting thing?
22:23:18YouCeyEyes.. ums trick works on stock firmware.. but on rockbox firmware do i need to use the same trick?
22:23:27 Quit Genre9mp3 ("I don't suffer from Rockbox psychosis. I enjoy every minute of it.")
22:23:36ArathisPaul_The_Nerd: ask barrywardell
22:23:46mirakpreglow: byterev would be what I needed, damn 5249 ...
22:23:53ArathisYouCeyE: yes, because USB isn't supported yet
22:24:25ArathisPaul_The_Nerd: the last thing I know is it still causes freezes
22:26:02mirakpreglow: ok just 3 cycles. not bad
22:26:10preglowmirak: what?
22:28:22mirakpreglow: nothing I was reading the wrong swap function, as there is no swap.b on coldfire
22:31:05*tucoz doesn't understand how playlists work after reading
22:33:09tucozbut that might be because i seldom use playlists :)
22:34:04Paul_The_Nerdtucoz: That's okay, apparently I no longer understand how they work either.
22:34:48Paul_The_NerdI think though that removing the concept of "Insert Next" and "Insert Last" and just having "Insert" and "Insert, and move insertion pointer" (needs a shorter name) and then giving the user the option to manually choose where the pointer is set at any given time allows a lot more flexibility anyway
22:35:12Paul_The_NerdYou can then insert songs anywhere in a playlist easily, plus not get stuck with "You can't do that unless you insert them, then move them" situations.
22:35:35tucozHmm. simplifying is good, but wouldn't that force the user to take an extra step when making a playlist?
22:35:56Paul_The_NerdOnly when wanting to "Insert Last"
22:36:06Paul_The_NerdI think
22:36:11 Nick S0ap is now known as Soap (n=Soap@unaffiliated/s0ap)
22:36:18Paul_The_NerdA normal "Insert + increment" would work like "Insert" does now.
22:36:22tucozI think that i would only use insert next or insert last when making a playlist
22:36:56tucozhow does insert work now?
22:36:58Paul_The_NerdInsert Next does inverse order. If you have 1 2 3 4, playing 2, and do "Insert Next" on A, B, then C, you get 1 2 C B A 3 4
22:37:31tucozah, that should be the other way (imo)
22:37:42Paul_The_NerdDepending on how you created the playlist, the Insert Pointer in that list will either be at 2, or at 4. So if you "Insert" A B C, you get either 1 2 A B C 3 4, or 1 2 3 4 A B C
22:38:07tucozhmm. so that is the dynamic/static concept?
22:38:20LinusNtucoz: "insert next" means "insert after the currently playing track"
22:38:40Paul_The_NerdIt's actually a little more than that, since I believe if you just play a single song in a directory that still counts as "dynamic" but in that case the pointer will be the currently playing song.
22:39:08Paul_The_NerdThe problem comes up if you create a playlist by inserting several songs manually because you don't want to listen to whole folders, or want to mix and match.
22:39:24LinusNdynamic playlists are those that are created on the fly
22:39:28Paul_The_NerdOnce you've done *any* insertions during the process of making your list, it becomes impossible to later come back and add the A B C in.
22:39:41Paul_The_NerdUnless you insert them in inverse order.
22:39:49tucozLinusN: yes I know that. But lets say i listen to song 2, and want to hear 3 songs after that, i would want them to be played in the order i insert them
22:40:14LinusNthen don't use "insert next"
22:40:24tucozshould i use insert then?
22:40:25Paul_The_NerdThat's why I think it could be much better to scrap the Next/Last concepts and just have a user-locatable insertion point.
22:40:29Paul_The_Nerdtucoz: Usually.
22:41:25LinusN"SimpleBox", bah
22:41:37tucozLinusN: then I would first Insert next, and then Insert after that?
22:41:49Paul_The_NerdInsert Next doesn't currently move the insertion pointer.
22:42:15Paul_The_NerdSo, 1 2 3 4, Insert A, Insert Next B, Insert C gets you 1 2 B A C 3 4, if I understand correctly
22:42:45SoapI see the commit comment:
22:42:45SoapiCatcher/UniCatcher WPS updates by Ioannis Koutoulakis: code update to the new volume/battery enum system and new versions for H10 5GB, Gigabeat and X5 remote
22:43:08Soaphave the new volume/battery enum system changes been noted in the wiki or manuals?
22:43:41Paul_The_NerdI believe it's in the CustomWPS wiki page
22:43:52Paul_The_NerdSince isn't that just the arbitrary enum change?
22:44:12SoapI don't know, I was hoping to find out.
22:44:32tucozPaul_The_Nerd: your example is what I would expect to happen
22:44:42tucozor rather, want to happen
22:46:12Paul_The_Nerdtucoz: The only problem I have with that is that the C goes after A, rather than B. For me, that's not ideal. For others, it is.
22:47:06pixelmaSoap: yes, it's in the CustomWPS wiki page
22:47:12Soapthank you pixelma
22:48:40 Join TheMcPoon [0] (
22:49:05TheMcPoonHey guys, I was just wondering, with the new ipods, do people with 60gb videos get the firmware too?
22:49:11LinusNAhoy, remember e'eryone, tis' talk like a pirate day today Aye.
22:51:33BagderZagor: the manual needs the similar rename
22:51:34TheMcPoonAnyone know
22:51:43ZagorBagder: yes
22:52:24Paul_The_NerdTheMcPoon: There are *lots* of iPod specific places on the internet. This place is dedicated to users of various players who *don't* want to use the original firmware. Really think this is the best place to ask? But as far as I know, the new iPod got *a* firmware update. It may not be the same firmware though.
22:53:01Bagderunless TheMcPoon spoke about Rockbox as "firmware"
22:53:16TheMcPoonYes, I know it's people dedicated to using a different firmware, but I thought people would also know about the new ipod firmware.
22:53:40nlsArr.. Bagder congratulations to yarr matey..
22:53:56Paul_The_NerdBagder: Aye, I be congratulating ye' as well.
22:54:21Bagderaye aye
22:54:29SoapTheMcPoon - that would still be drifting off-topic. This channel isn't just for Rockbox users, but for Rockbox discussion.
22:55:21TheMcPoonAlright, I understand. So, I shall ask, does anyone know where I can go to ask/find firmware other than Rockbox?
22:55:45tucozAhoy, was it another game that was called Tetrox, or somethin' else that caused the rename? A pence for an old man o'de sea?
22:56:03Zagortucoz: I'm writing a mail explaining it
22:56:05Bagdertucoz: another game that uses a similar name
22:56:22tucozcool. I found a detergent with called Tetrox :)
22:56:23Bagder... pretty well known
22:56:56LinusNsomething like "Avast, ye demon copycats! Taste the wrath of my arse!" from a law firm
22:57:18 Join Mmmm [0] (
22:57:45tucozI think the rox part in terox made me think it was a rockboxed game
22:58:37nlswhatever happened to the jewels images buisness?
22:58:46Bagderthe never got back
22:59:15tucozwe should google next time a suspicious plugin gets committed
22:59:27nlsheh top searchresult for tetrox on google is a screenshot of rockbox tetrox :)
23:00:05BagderI don't think googling works for this kind of thing
23:00:11LinusNtucoz: wouldn't have helped in this case
23:00:23Bagder"find names that are trademarked that might look similar to this"
23:00:46LinusN"t*tris" is a pretty well-known trademark
23:00:47tucozaha. so Tetrocks or rocktris wouldn't work?
23:01:34 Quit Mmmm (Client Quit)
23:01:35Paul_The_NerdI didn't think Tetris *was* trademarked.
23:01:38*ze cluelessly jumps in without knowing what the coversations about:
23:01:39LinusNi like "rockblocks"
23:01:41zewhat about rockblocks?
23:01:51BagderPaul_The_Nerd: they claim it is
23:01:54Paul_The_NerdAnd anyway, Tetrox could easily have worked as a reference to Tetronimos, which is what they are.
23:02:08 Join Strath [0] (
23:02:13SoapIs this what the legal issue was you couldn't talk about Bagder?
23:02:26Bagderits a part of it, yes
23:03:03Paul_The_NerdHrm, well according to wikipedia someone does claim ownership of it at least.
23:03:08Paul_The_NerdAnd it's my arch-nemesis as well.
23:03:25nls"Familiar game with falling blocks." :-D
23:03:48 Join tucoz_ [0] (
23:03:53Paul_The_NerdI wonder if they could sue if you did call it "Falling Tetronimoes"
23:04:06Bagderprobably not
23:04:14 Quit tucoz (Nick collision from services.)
23:04:15ZagorPaul_The_Nerd: the thing is they can sue us whatever we do.
23:04:16 Nick tucoz_ is now known as tucoz (
23:04:20Paul_The_NerdRockblox works. :)
23:04:23Bagderin this case, the details are in the details
23:05:01Zagortesting trademark in court is way more expensive than it's worth
23:05:10Zagorat least for us in this case
23:05:41preglowand that's not even mentioning the amount of boredom
23:05:48preglowjust rename it and be done with it
23:06:02tucozthat is the easiest and best solution
23:06:02rigeldude, why the hell cant you get some EFF or OSDN associated lawyer to take it on pro bono?
23:06:15Zagorrigel: because there is no payoff
23:06:18Paul_The_Nerdrigel: Because it doesn't matter and isn't worth it?
23:06:45rigelchallenging intellectual property law is always worth it
23:07:01rigelbut im a bi of a zealot
23:07:03Zagorrigel: then become a lawyer and do it for us pro bono :)
23:07:04rigelbit, even
23:07:06Bagdereasy to say when you're not involved
23:07:07Paul_The_NerdSome intellectual property law is good. And trademarks are actually a very important thing, overall.
23:07:29Paul_The_NerdTrademark greatly increases the odds of you actually knowing who you're buying something from.
23:07:32rigeli didnt say IP law is bad, did i?
23:07:42Paul_The_NerdYou said it's always worth it to challenge it.
23:07:42rigeli just said challenging it is always good.
23:07:50rigelyes. they are completely different.
23:07:58Paul_The_NerdIf someone is actually *right* though, challenging it is good why, then?
23:08:10rigelsolidifying that precedent
23:08:21rigelor reexamining its basis, if that is necessary
23:08:28preglowideally, sure
23:08:31LinusNsounds like money well spent for rockbox
23:08:32preglowbut man, the boredom
23:08:39preglowthe stacks of paper
23:08:54rigelwho is this that owns the rockbox trademark?
23:09:10 Quit TheMcPoon ()
23:09:13Paul_The_Nerdrigel: I offer you a solution. Take our plugin code, adapt it to run on a PC, name it whatever you want, and then contact the EFF to protect *you* :-P
23:09:14Bagderthere is no generic such thing
23:09:34rigeli'm unclear on what the issue is then
23:09:42markunBagder: I guess I'm the last person here to tell you: congratulations!
23:09:49rigeloh yeah cngrats, badger
23:10:00markunok, so I was not the last :)
23:10:10Paul_The_Nerdrigel: The issue isn't in regard to the Rockbox name, but the name of one of our plugins.
23:10:18Bagderthanks. have a beer each for me ;-)
23:10:23rigelwhich one?
23:10:33markunBagder: will they leave us alone after we rename or is there more to it?
23:10:41rigeli thought someone was claiming rights to the rockbox name
23:10:44rigelmy bad
23:10:44preglowi have no beer!!1
23:10:47rigeli should have read the scroll
23:10:52markunI'll have a beer then
23:10:57preglowi have no beer!!1
23:10:58nlshmm is calling the doom plugin simply DOOM ok?
23:11:22 Part lowlight
23:11:34Bagdermarkun: there is more
23:11:57preglowthey've trademarked all our plugins? :>
23:12:13LinusNDeutsche Rockwool Mineralwoll-GmbH owns the Rockbox trademark in the EU :-)
23:12:35rigelwhich names are the subject of trademark disputes?
23:12:38ZagorLinusN: yeah, but not exactly in our product class :-)
23:12:38markunBagder: from the same guys?
23:12:41Paul_The_Nerdnls: Doom's source has been released under the GPL. We aren't just calling it doom. it IS Doom.
23:12:47LinusNZagor: :-)
23:13:14Bagdermarkun: yes, the name of the plugin is part of it but not the whole
23:13:22 Quit pondlife (""disconnected has pondlife"")
23:13:32 Quit lini ("lini has no reason")
23:13:56nlsPaul_The_Nerd, yeah but I was thinking that all the other modified doom variants outthere use different names.
23:14:09 Join lini [0] (i=pugsley@
23:14:16BagderGPL allows you to release clones with the same name
23:14:19Paul_The_Nerdnls: Like "Doom Legacy" or "zDoom" ?
23:14:36***Saving seen data "./dancer.seen"
23:14:43Bagderin fact, forbidding that makes it non-GPL compatible
23:14:44nlsdidn't know about those... ok
23:14:55Paul_The_NerdThose were, at one point, the most popular ones.
23:15:00rigelso uh, what is the trademark in dispute again?
23:15:01Paul_The_NerdIt's been a while since I've been in the Doom scene though
23:15:34tucozMaybe pacbox should be called rockman
23:15:54*preglow hoists the "please not rock in everything" banner
23:16:28tucozbut pacbox is still using the pac part
23:16:33Paul_The_Nerd"Little Yellow Guy"
23:16:38preglowthere you go!
23:16:48tucozyeah, LYGbox
23:17:42bluebrotherBagder, congrats
23:17:50*bluebrother just read logs ...
23:17:58nlswe should name everything with long strings of random letters that change on every compile ;-)
23:18:03tucozor xobox could be infringing to trademark
23:18:11bluebrotherwhy not use some locale names?
23:18:42bluebrotherI remember some linux application that used a german name as application name.
23:19:17nlsyeah name all the plugins from the credits
23:19:29bluebrothermaybe "egoshooter" for doom :)
23:19:39 Quit KN|stiff (Read error: 104 (Connection reset by peer))
23:20:09tucozand bollar_och_linjer for xobox
23:20:26rigelname them all in urdu
23:20:28*Paul_The_Nerd wants a plugin named after him.
23:20:30 Join rotator [0] (n=e@rockbox/developer/rotator)
23:20:31Paul_The_NerdWhich is most annoying?
23:20:34rigelthat should obviate any copyright issues
23:20:48bluebrotherPaulTheNerd-Plugin ;-)
23:20:49rigeler, trademark
23:20:53rigelheh, even
23:20:54rigeldamn kb
23:20:57*nls wanna be bubbles <3
23:21:18tucozsee you
23:21:20Zagorrigel: except in pakistan and india, then
23:21:20 Part tucoz
23:21:41rigelindia is notoriously lax about IP law
23:21:48Bagderand I bet there are trademarked urdu names already
23:22:04rigelat least w/r/t drug patents
23:22:15rigelwhich is closer to my actual area of expertise than coding
23:23:56nlswell gotta go to bed so I can get up...
23:24:03 Part nls
23:25:48preglowmight wanna duplicate that effort
23:30:01 Join EV|Toshiba [0] (
23:31:48 Join solarflare [0] (
23:32:10 Join webguest94 [0] (i=5211b106@gateway/web/cgi-irc/
23:32:42 Part LinusN
23:39:47 Quit Nibbier (Read error: 110 (Connection timed out))
23:40:29 Quit Pyromancer ("Ex-Chat")
23:41:05 Join Nibbier [0] (
23:43:24 Quit darkless (Read error: 104 (Connection reset by peer))
23:43:51 Join darkless [0] (n=darkless@
23:45:49 Join Angryman [0] (i=54389f07@gateway/web/cgi-irc/
23:46:19AngrymanDoes anyone know how to wipe sector 60 (Software advice recommend) ???
23:47:06hardeepPaul_The_Nerd: the only problem with your playlist proposal is that users would need to do an extra step to insert next/last. i'm betting there will be many complaints about that
23:47:19hardeepbut i'm all for simplifying the options
23:47:39rigeli cant figure out the controls of the h10 20gb
23:47:49rigeland the "manual" hasnt been updated since the controls were all changed
23:48:11AngrymanWhat do you wanna know
23:48:28rigelhow to get to these menus
23:48:39rigelto make playlists, etc
23:48:51AngrymanSettings are reached by long press either on right or left or a short press on power off
23:49:04 Join Pyromancer [0] (
23:49:29AngrymanDid that help?
23:49:44 Quit Lefungus ("Parti")
23:50:13rigelhow do i update the tag database
23:50:22AngrymanWhich version are you running
23:50:31 Quit EV|Toshiba ("Leaving")
23:50:32AngrymanWith or without Album Art
23:50:53 Join webguest38 [0] (i=46b37457@gateway/web/cgi-irc/
23:50:55 Quit Kohlrabi ("Quit")
23:51:29AngrymanYou#ll find it in file view
23:51:53webguest38Hey I use a mac os x and I just installed rockbox and is great anyway to make fonts bigger? also I have ipod video and I would like to know how to get a GUI bootloader I dont like the idea that rockbox is first on boot
23:52:08rigeli dont see it
23:53:01rigel"gui bootloader"?
23:53:05Angrymanshort press on Power off ->General Settings −−> File View −−>Tag cache
23:53:10Soapwebguest38 - The ipodlinux loader brings up a boot menu upon power-on
23:53:11midgey34webguest38: you need to install the font package
23:53:26Soapwebguest38 - you need to install the fonts package from the daily download page.
23:53:36webguest38ok hang on
23:53:43Angrymanyou need to play around more with rockbox ;-)
23:54:01webguest38I dont have ipodlinux installed
23:54:07midgey34webguest38: the rockbox bootloader isn't GUI based, you'll need the ipod linux bootloader like Soap said
23:54:13webguest38do I need to install it
23:54:35webguest38Through windows or mac do I install IPL
23:54:49Soapwebguest38 - rockbox is first on boot for many reasons, but an important one to consider is the fact that the Apple firmware boots from disk only if it has been shut off for 48(?) hours, and so most of the time once you boot into apple firmware you won't have to deal with the bootloader again until you want to boot rockbox. or atleast infrequently.
23:54:53AngrymanDoes anyone know what means this "*PANIC* Stkov tagcache ???
23:55:27Angrymanor better where the trouble is done in the code...?
23:55:44Angrymanor how to wipe sector 60 ;-)
23:55:46webguest38But i would like to have a GUI so how would i install IPL under windows or os x
23:56:40rigelok, i initialized and updated the tag cache
23:56:42Soapthat would be a question for #ipodlinux, but they might get rightfully pissy if you don't read their excellent website before asking basic questions.
23:56:44rigeland i still cant search
23:56:50rigelor look at tags at all
23:57:06webguest38would installing IPL erasse rockbox?
23:57:06midgey34webguest38: this might help -
23:57:40bluebrotherAngryman, if you want to wipe a specific sector on disk you can use dd and simply write from /dev/zero
23:57:44Angrymanyou can set it via file view settings
23:57:59AngrymanThx bluebrother
23:58:12webguest38thanks for the link I was about to gooogle it
23:58:26webguest38I hope i dont have to spend hours installing rockbox all over

Previous day | Next day