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 2006-02-25

00:00:31BagderLGMS: use the normal cygwin installer and find "dev gdb" or something like that
00:00:40 Join ender` [0] (i=ychat@
00:03:02Paul_The_Nerdlinuxstb: Have you had scrollwheel troubles in the doom menu?
00:03:54Paul_The_NerdSorry, the pregame wad-selection menu?
00:06:00 Quit xmixahlx ("blah blah blah")
00:06:56linuxstbYes - not with the wheel though, but with other buttons.
00:07:16Paul_The_NerdI realized it's hitting select to choose something, causes it to immediately re-enter that menu.
00:07:23 Quit ender` (" Don't worry about the world coming to an end today. It's already tomorrow in Australia.")
00:07:29Paul_The_NerdLike it reads the button twice
00:07:40linuxstbYes, it seems to be reacting on both button press and button release events.
00:07:49midkayi've had the same, i think it scans for BUTTON_SELECT|BUTTON_REL or something.
00:08:57midkayyay, compiled alright. now the moment of truth.. *ejects ipod*
00:09:39midkayincompatible model..
00:09:52Paul_The_NerdHm. the doom_memory patch definitely has some issues
00:09:57*Paul_The_Nerd goes about compiling it without that patch.
00:11:42midkayah, the new doom.rock didn't copy over for some reason
00:13:44 Join ender` [0] (i=ychat@
00:13:45midkayawesome, it works! :D
00:14:04 Join damaki__ [0] (
00:14:25 Quit BHSPitMonkey (Remote closed the connection)
00:14:50 Quit xavierk ("CGI:IRC (EOF)")
00:15:16 Join midkay_ [0] (n=midkay@
00:16:33midkay_alright, so, it runs, and beautifully :D one thing is bugging me - the need for disk spinups (thus making the game lock up for a few seconds) to load stuff..
00:17:02 Quit moozooh (Read error: 104 (Connection reset by peer))
00:17:03midkay_i'd suggest an option to keep the disk spinning always so you don't need to wait like that.. i'd rather have less battery life than having to stop every 10 seconds :)
00:17:32*Paul_The_Nerd does the "My device is flash based" dance.
00:17:54*midkay_ does the "My device has 15-60x as much storage as yours" dance.
00:18:06 Quit quobl_ (Read error: 104 (Connection reset by peer))
00:18:20BagderPaul_The_Nerd: is that what "Flash Dance" was really about ?
00:18:40Paul_The_NerdMan, you mean you haven't seen it? Waaaay ahead of its time.
00:18:43midkay_is there currently any way to get back to the menu?
00:19:00 Join quobl_ [0] (n=quobl@tor/session/x-da21a8f944b4672c)
00:19:06BagderI guess I just didn't understand it
00:21:22 Quit Moos ("Glory to Rockbox!")
00:23:29 Join OneFix_Work [0] (n=onefix@
00:23:53OneFix_WorkI have a bad is the skipping on iPod Video builds?
00:24:42 Join XavierGr [0] (
00:25:37LinusNOneFix_Work: it's somewhere between "not too shabby" and "really awful"
00:26:25LinusNseriously, it all depends on what files you play
00:26:36 Join San [0] (
00:26:48 Quit perldiver (Read error: 110 (Connection timed out))
00:27:02OneFix_Workwell, I know AAC isn't that great on it right now...I'm just trying to figure if I should wait a little while longer
00:27:09midkay_with the EQ off, pretty much everything plays fine for me.. well, all mp3's and FLAC files.. some AAC..
00:27:13OneFix_WorkLinusN: What about battery life?
00:27:17Bagderwaiting is just for cowards ;-)
00:27:25LinusNOneFix_Work: i have no idea
00:28:07LinusNi guess everything will be fixed if we all just sat down and waited for something to happen
00:28:12 Join ashridah [0] (
00:28:14OneFix_WorkLinusN: Well, I used iPod Linux on my old Mini...and it was ok, but it wasn't as good as the iPod firmware
00:28:31OneFix_WorkI really just want to play OGG files on mine :)
00:28:45BagderOneFix_Work: I'd suggest you go ahead and try it
00:28:47Paul_The_NerdOgg is actually relatively fast on iPod, though not sure how fast on 5g.
00:29:01LinusNah, now i can finally control the hdd power on the x5
00:29:05OneFix_Workhave the Apple firmware problems been fixed now?
00:29:12 Join Moos [0] (
00:29:12Paul_The_Nerd"Apple firmware problems?"
00:29:20Bagderyes, we fixed their problems
00:29:35MoosLinusN: hi, Congratulations !!
00:29:42OneFix_WorkPaul_The_Nerd: Yea, the bootloader patch was causing problems with the Apple firmware
00:29:57Sananyone else have no audio on the latest optimized build?
00:30:09Paul_The_NerdOneFix_Work: What kind of problems? I wasn't aware of ones on 5g, but I may have missed them.
00:30:19BagderSan: I don't think most people here use that
00:30:34 Quit damaki_ (Read error: 110 (Connection timed out))
00:30:58Sansome might do though
00:31:12OneFix_Workit may have only been on the 4g
00:32:21OneFix_WorkOn the Linux4iPod bootloader there also used to be a way to make the apple firmware the default firmware...can you do that with rockbox?
00:33:07 Quit midkay (Read error: 110 (Connection timed out))
00:33:09Sanok, fixed it
00:33:23Sanif anyone asks, for future reference, it was a problem with the EQ
00:33:31Sanyou have to turn it off then it works again
00:33:52Bagderit seems it has is messed up often lately
00:34:00Bagderhm, typing
00:34:22BagderI'll stop this and go to sleep instead, night!
00:34:30Paul_The_NerdOneFix_Work: The apple firmware looking funny is only a problem with the grayscale ipods. Yes it's possible to make the apple firmware the default, but you have to compile your own bootloader to do that.
00:34:53OneFix_WorkPaul_The_Nerd: Oh, ok
00:35:29Paul_The_NerdThe primary goal of Rockbox is to replace the Apple firmware though. Ideally once you start using it, you won't really feel the desire to use their. (I know I haven't, at least.)
00:36:23Paul_The_NerdWell, by "Rockbox" I mean "rockbox on ipod" since when it was made for Archos' it probably wasn't intended to replace the Apple firmware.
00:36:42 Quit amiconn (Read error: 110 (Connection timed out))
00:37:04OneFix_WorkPaul_The_Nerd: Only a few problems I see right now...and then I would probably have no Album Video support that I know podcasting app that works with it...
00:37:17Sanlast question, is anyone working on fixing the scroll bug?
00:37:37 Join perldiver [0] (
00:37:43midkay_scroll bug?
00:37:55OneFix_WorkPaul_The_Nerd: Of course it needs to be a little faster, but that is gonna come
00:38:23 Quit Moos ("Glory to Rockbox!")
00:38:31Santhe scrolling is slower than it used to be and when you scroll with music playing the music stops untill you stop scrolling
00:39:06 Join doctor [0] (n=doc__@
00:39:37OneFix_WorkPaul_The_Nerd: For the iPod Video, video support is gonna be the killer app
00:40:09Paul_The_NerdOneFix_Work: Well, I have a Nano, and usually use my music player for music, so album art and videos don't affect me. Video support may be a little while in coming though.
00:40:10OneFix_WorkWehn I can play AVI, WMV, etc on the RockBox firmware, I'll probably switch
00:40:26*Paul_The_Nerd thinks it may be a while before you switch then.
00:40:43BagderI don't think Rockbox is for you
00:41:12OneFix_WorkWell, I don't do a lot of videos...most of the videos I do a re podcasts...
00:41:16 Join Mark_ [0] (
00:41:39 Quit ender` (" The idea is to die young as late as possible.")
00:41:42midkay_yet you'll wait for a wide range of video support to switch?
00:42:08Paul_The_Nerdmidkay_: There are people who won't switch for a wide variety of reasons, some of which baffle me to no end. :)
00:42:18midkay_Paul_The_Nerd, same.. :)
00:42:34Paul_The_Nerd"The menus don't look pretty" is one of my all time favorites though.
00:43:00*Paul_The_Nerd probably only looks at the screen of his DAP about 5% of the total time it's powered on, if not less.
00:43:27OneFix_WorkWell, I guess the main reason is that I now have most of my music in AAC format...and seeing as how RockBox's support for it is not the greatest, that's gonna keep a lot of ipod owners from switching
00:43:34midkay_yeah, that one's good.. this one may be about the best, though.. "i'll wait for months until video playback supports a wide range of formats.. oh, no, i don't really watch videos, but still.."
00:43:59Paul_The_NerdOneFix_Work: Well, we're always looking for someone to optimize the AAC codec. :)
00:44:10BagderOneFix_Work: join in and make it happen
00:44:13midkay_OneFix_Work, it's not like a permanent switch.. you can always switch between firmware right on the device, as well as completely uninstall rockbox..
00:45:25RotAtoRhmm, BUTTON_OFF seems to be broken in the sims
00:45:58 Join linuxstb_ [0] (
00:46:19OneFix_WorkHere's the big question...can the BroadCom chip on the Video do Audio or can it just do Video?
00:46:58 Quit Bagder ("Off to search for that connect-resetting peer guy!")
00:48:06 Quit lodesi (Read error: 110 (Connection timed out))
00:48:13OneFix_WorkActually, and I'll let you in on my secret...I switched to Linux4iPod back when I had my Mini...and something happened that skrewed up the database in the apple firmware...I had to reinstall the whole thing...that was 4GB...60GB would really p*ss me off if I had to reinstall it
00:48:52Paul_The_NerdYou may be best sticking with iPod's built in software.
00:48:56 Join lodesi [0] (
00:48:57Paul_The_NerdRockbox doesn't use the database.
00:49:07OneFix_WorkPaul_The_Nerd: I know that
00:49:44Paul_The_NerdSo, were you to want to use all your music, odds are you'd have to copy it all over again with a proper directory structure just to use rockbox.
00:49:46OneFix_WorkBut I think the problem was that iTunes no longer recognized the iPod when I tried using it to copy files to it
00:50:18OneFix_WorkPaul_The_Nerd: Yea, guess that's true
00:50:38OneFix_WorkDoes RockBox use PodZilla or is it a seperate app?
00:51:05Paul_The_NerdRockbox is its own app entirely, from the kernel up. Not linux based or anything.
00:51:33OneFix_WorkPaul_The_Nerd: Really...didn't know that...I thought it was Linux based...
00:51:48LinusNso many people think that, i wonder why
00:52:08 Join DJ_Dooms_Day [0] (
00:52:16Paul_The_NerdThere's a fair portion who also just think it's a modification on the existing firmware. It's strange.
00:52:35OneFix_WorkProbably because you give a "Thank You" to the Linux4iPod project...
00:53:18linuxstb_Probably about 2% of the code in Rockbox for the iPod is from ipodlinux.
00:53:36linuxstb_But it's a very important 2% - the low-level hardware drivers.
00:53:37OneFix_WorkPaul_The_Nerd: Well, I know better than that, but I thought it was really just a distro of the Linux4iPod project
00:54:23OneFix_Worklinuxstb: Yea, I figured that...but thought there was more code...
00:54:29LinusNthe ipl project deserves a great deal of respect for the remarkable reverse engineering
00:54:42Paul_The_NerdOneFix_Work: As linuxstb said, other than low level hardware code it's its own thing. Which is why the audio playback tends to be in the "superior" to "vastly superior" range, the boot time in the "wow, that's fast" range, and the application support in the "applications? It's a MUSIC player" range.
00:54:57OneFix_WorkI guess that's probably why RockBox can be better than the ipodlinux's more specialized
00:55:00imphasingRockbox boots way to fast
00:55:11imphasingIt never gives me a chance to choose other firmwares
00:55:41 Quit LGMS ("CGI:IRC (EOF)")
00:56:59OneFix_WorkI guess the real question there work on video support being done...and my guess is that the reason it will take so long it that the BroadCom chip it's based on is probably poorly documented...
00:57:50linuxstb_undocumented is the word.
00:57:58 Nick thegeek_ is now known as thegeek (
00:58:19imphasingAnd broadcom is a nasty company
00:58:25imphasingThey won't do jack
00:58:33OneFix_Worklinuxstb: So, it's something else that needs to be fully understood before something can be written for it...
00:58:36imphasingSe we will have to reverse engineer everything to get anything done
00:58:49OneFix_Workimphasing: Yea
00:58:56imphasingOneFix_Work: Hard to use a processor if you won't know what instruction set it uses ;)
00:58:58OneFix_Workimphasing: I understand...
00:59:12imphasingIt's possible that it's based on ARC though
00:59:27linuxstb_Personally, I'm not interested in the video processor. We still have a lot of work to do to properly support audio playback, without worrying about video.
00:59:32imphasingbut there's going to be lots of instructions that broadcom created for video processing purposes
00:59:55linuxstb_And the 5g is only one of the many ipods Rockbox is in the process of being ported to, and there are a tiny number of ipod developers.
01:00:06OneFix_Workimphasing: Well, since from what I read it only decodes video, it's probably different than most people have worked...from what I read, it's like a programmable hardware decoder
01:00:32ashridahimphasing: heh. you just need to 'promise' to order like 100,000 of them if they'll give you some engineering samples and a set of specs :)
01:00:53ashridah<broadcom> here you go, when can we expect the cheque?
01:00:59ashridah<imphasing> SUCKERS !
01:01:02ashridah* imphasing runs
01:01:36Papricamm the ipod color have rtc?
01:01:41OneFix_Workashridah: Apple is probably pulling an Atari on the market and making sure they horde all of the chips
01:01:49linuxstb_Paprica: All ipods do AFAIK.
01:02:00Papricain rockbox too?
01:02:34 Quit doctor ("Leaving")
01:02:42ashridahOneFix_Work: i'm not sure, if it's based on the portalplayer, and other manufacturers use portalplayer as well, such as iriver, don't they?
01:02:59linuxstb_Paprica: Yes.
01:03:28ashridahdont' know if the video is portalplayer's gear tho
01:03:57linuxstb_The 5g has one portalplayer chip, and one Broadcom chip. A match made in open source hell.
01:04:30ashridahah, didn't realise the broadcom was an addon by another party
01:04:42 Quit Thus0 ("Leaving")
01:04:58linuxstb_Broadcom _is_ the other party. They are famous for their refusal to help Linux driver developers.
01:05:38OneFix_Workashridah: Yea, there was some talk about what the broadcom chip was capable of doing...aparently the current Apple firmware only scratches the surface of what it can do...
01:06:20OneFix_Worklinuxstb: is visible by how well they have helped the WiFi community...
01:07:13OneFix_Worklinuxstb: Even though they could release a native driver anytime...because the Linksys WRT54G uses a broadcom wireless there's a module for it
01:09:07 Quit saratoga ("CGI:IRC (EOF)")
01:11:14 Join IcyStorM [0] (
01:14:26 Quit XavierGr ("Trillian (")
01:16:55 Join XavierGr [0] (
01:17:05 Quit XavierGr (Read error: 104 (Connection reset by peer))
01:17:16***Saving seen data "./dancer.seen"
01:17:53 Join XavierGr [0] (
01:19:54 Part LinusN
01:24:39 Quit imphasing ("Lost terminal")
01:24:40 Quit XavierGr (Read error: 104 (Connection reset by peer))
01:27:56lodesii'm trying to build a plugin
01:28:20lodesiand i get an "undefined reference to '_memset'" error ...
01:28:31lodesiwhile I don't use memset at all
01:28:34lodesiany idea ?
01:29:18markunlodesi: yes, it's an implicit memset
01:29:30markunyou can use rb->memset
01:29:43markuncan you show me the line where it goes wrong?
01:29:48lodesii don't have any memset in my plugin..
01:29:56 Quit midkay_ ("Leaving")
01:30:05lodesino line number its during linking
01:30:29 Nick Shadowarrior13 is now known as Shadow|ushering (
01:30:34linuxstb_gcc will be using it behind-the-scenes to initialise something.
01:30:37ashridahlodesi: it's done as part of rockbox's headers, attempting to attach memset to the rb struct
01:30:53linuxstb_You will need to define your own memset() function, which calls rb->memset()
01:31:28*ashridah shuts up
01:31:46 Quit slimx (Remote closed the connection)
01:32:27lodesiok thx, trying it
01:32:36linuxstb_lodesi: Look in apps/plugins/wav2wv.c for an example
01:33:22 Join moozooh [0] (n=moozooh@
01:33:34 Part IcyStorM
01:33:35lodesilinuxstb: works well, thx :)
01:38:06 Quit muesli- (Read error: 110 (Connection timed out))
01:44:35 Join webguest28 [0] (
01:44:46 Part Paul_The_Nerd
01:44:48webguest28anyone still awake?
01:46:22qwmi'm sitting here. doing nothing.
01:46:27webguest28oerr groovy
01:46:32webguest28its a bit like that
01:46:42 Quit Strath (Read error: 104 (Connection reset by peer))
01:46:50webguest28I dont spose you know anything about rockbox on ipods do you :P
01:46:51 Join Strath [0] (
01:47:26qwmcan't say i do.
01:47:56webguest28stoopid ipods nothings the same or as easy as it is with other players
01:48:08linuxstb_What's the problem?
01:50:20webguest28Eh trying to set backdrops.... I mean I can copy one into the backdrop directory, edit the theme .cfg file to make it bring it up and when I load the theme it works.. but a lot of bloody effort just to change the backdrop.. ive seen on other players you can "show all files" from the file browser and theres an option to set as backdrop from there.. but dunno how to get to either of those options with ipods and its pissing me off grrrr
01:50:48linuxstb_It's exactly the same on the ipod.
01:51:35webguest28so how do you get to the "show all files" menu
01:51:46webguest28hodl you obviously cant hold down "a+b"
01:52:07linuxstb_Err, press "menu"...
01:53:53 Quit linuxstb_ ("CGI:IRC")
01:55:16 Quit lodesi (Read error: 110 (Connection timed out))
01:55:20 Join Benacool [0] (
01:57:31 Quit Fuiou5 (Read error: 104 (Connection reset by peer))
01:57:47 Join virtualball2 [0] (
02:01:55webguest28bah didnt realise there was an option in the settings for filenames. yay
02:02:01 Quit webguest28 ("CGI:IRC")
02:03:45 Quit mikearthur (Remote closed the connection)
02:04:07 Join mikearthur [0] (
02:11:54 Join Sinbios [0] (
02:17:49 Join Jungti1234 [0] (n=jungti12@
02:45:22 Quit Mark_ (Read error: 104 (Connection reset by peer))
02:45:28 Join infamis [0] (
02:50:22 Quit infamis (Client Quit)
02:52:49 Join infamis [0] (
02:54:20infamisI'm trying to understand the rockbox there a main thread for UI?
02:56:07Jungti1234Rockbox UI?
02:57:14infamisyeah...what happens when Ipress this or that does it get routed
02:57:24infamisI'm trying to find that in the source
03:00:26Jungti1234Is it right?
03:00:33infamishm...I read that before but didn't really seem to show what I needed
03:01:15infamisfor ex: in apps/main.c that's where the firmware starts & where the while(1) loop is
03:01:27infamisbut where are the keyhandlers or whatever?
03:02:29infamisinterrupt handlers I guess?
03:02:49linuxstbAll the low-level hardware drivers, and the Rockbox kernel are in firmware/
03:03:01linuxstbThe button driver is firmware/drivers/button.c
03:06:36virtualball2I have a question, i know C and i was wondering, how do i compile .c files into .rock files?
03:07:07linuxstbPut your .c file in apps/plugins/ and add the name of your file to apps/plugins/SOURCES
03:07:20linuxstbHave a look at the helloworld.c example.
03:07:30linuxstb(which isn't in SOURCES, so isn't compiled)
03:07:50virtualball2ok thanks
03:08:17virtualball2and is there a wiki page for the lcd commands...i cant navigate your wiki :P
03:08:40virtualball2mkay thanks
03:11:32infamisis there a debug screen accessible from inside the sim anywhere? (in other words where can I get the DEBUGF outputs)?
03:11:37 Quit Daishi (Remote closed the connection)
03:12:02virtualball2isnt there a debug setting thing in the info menu?
03:12:32infamisI can't even get to the menu though
03:12:35infamison the x5 sim
03:12:47virtualball2sorry then i dont know what to do
03:14:03linuxstbAre you using Windows and SDL?
03:15:30infamisunder cygwin
03:16:03virtualball2Mac here
03:16:10*virtualball2 loves mac lol
03:16:19*infamis doesn't
03:16:26virtualball2... lol
03:16:54linuxstbDEBUGF should print to the console - but that doesn't work in the SDL sim under Windows.
03:17:06linuxstbHopefully someone is looking into it...
03:17:17***Saving seen data "./dancer.seen"
03:17:32 Quit matsl (Remote closed the connection)
03:17:59linuxstbHow are you running the sim? Are you typing "rockboxui" from cygwin, or double-clicking on it?
03:17:59infamisI might as well start using linux sometimes...I only use it for perl/php devel (even though php works in windows)
03:18:10infamis./rockboxui −−background
03:18:32linuxstbOK - I thought it might be working if you start it from the command-line. But obviously not.
03:18:56infamisdbl clicking always gives that msgbox err
03:18:59linuxstbYou'll find that Rockbox itself compiles a _lot_ faster under Linux.
03:19:38infamisyea, I'm aware of that...I wouldn't think about building a linux krnl under cygwin...even if it was possible!
03:20:39linuxstbAre you planning on helping with the x5 port?
03:20:58infamisyea...once I learn enough!
03:21:26linuxstbWorking with the sim - and making the higher-level parts of the port work is a good way to learn.
03:21:26infamisI used to be into computers heaviliy around ' a job as an ASP programmer & it seems I lost everything I once knew before I got the job. 5 years of ASP really dumbs yo u down
03:22:25infamis*5 years of nothing but ASP
03:23:42infamisI barely know C
03:25:12infamisI'm out
03:25:13 Quit infamis ("food")
03:46:24josh_18:16 < linuxstb> DEBUGF should print to the console - but that doesn't work in the SDL sim under Windows.
03:46:32josh_linuxstb: on Windows, SDL redirects stdout and stderr to files
03:46:39josh_try freopening them to "con" maybe
03:51:19 Part jaebird ("Leaving")
03:51:25 Join jaebird [0] (n=jaebird@
03:58:09 Quit Shadow|ushering ()
04:00:49 Quit ScoTTie_ (Read error: 110 (Connection timed out))
04:04:49 Quit virtualball2 ("virtualball2 has no reason")
04:38:19 Quit Benacool ()
04:48:52 Quit mikearthur (Remote closed the connection)
04:49:30 Join mikearthur [0] (
04:49:47 Quit mikearthur (Remote closed the connection)
05:03:18 Join Rob2222_ [0] (
05:06:39 Join mikearthur [0] (
05:17:20***Saving seen data "./dancer.seen"
05:19:06 Join mymomthelush [0] (
05:19:25 Quit mymomthelush (Client Quit)
05:20:13 Quit Rob2222 (Read error: 110 (Connection timed out))
05:45:02 Quit mikearthur ("Konversation terminated!")
05:47:17 Join mikearthur [0] (
06:08:07 Join imphasing [0] (
06:28:23 Quit RotAtoR ("zzzzz?")
06:37:20 Quit ts-x ("CGI:IRC (EOF)")
06:40:04 Join mymomthelush [0] (
06:48:10 Quit mymomthelush ("Trillian (")
07:04:28josh_anyone know why Rockbox can't use the helix decoding libraries?
07:05:27ashridahi don't recall the helix licenses being particularly GPL compatible
07:05:47 Join Arrogant [0] (
07:05:58josh_"What legal issues come up if I use GPL-incompatible libraries with GPL software?"
07:06:13josh_basically, it's OK iff you put a "special exception" in the GPL blurb for Rockbox
07:06:36ashridahthe rockbox team tend to prefer using the GPL on all of the code.
07:06:41josh_ah ok
07:07:13ashridahand not all of the code was contributed by current members, the authors of the mp3 codec may complain if we hook their GPL'ed code up with rockbox, and then use rockbox to hook it up to something less than free
07:08:42thegeekthe strength and backbone of rockbox is the gpl
07:11:16ashridahbut yes, i seem to recall that the helix license is kinda nonfree. the player itself is fairly opensource, but many of the codecs have click-wrap EULA's and whatnot before you can get to them
07:13:03ashridahbesides, which codecs do you expect to get out of them? are they fixed-point arithmetic based?
07:17:24***Saving seen data "./dancer.seen"
07:21:37ashridahhell, according to the site, to use the AAC fixed point implementation they've got you've still got to buy a license for some patents
07:21:59 Quit mikearthur (Remote closed the connection)
07:22:06 Nick Sinbios is now known as blah (
07:22:22 Nick blah is now known as Sinbios (
07:22:26ashridahas for realaudio, it's commercial, full stop
07:28:26 Quit Aditya ("Trillian (")
07:41:48 Join midkay [0] (n=midkay@
07:54:21 Join perplexity [0] (
08:03:23 Quit imphasing (Read error: 110 (Connection timed out))
08:11:44 Join damaki [0] (
08:12:22 Join Thus0 [0] (
08:13:36 Join amiconn [0] (
08:16:25 Join saratoga [0] (
08:17:29saratogaashridah: Regarding Helix AAC patents, that has nothing to do with Real's licenseing and everything to do with the fact that MS, fhg, and several other companies own patents required to implement AAC
08:18:49saratogaso technically, you can only distrubute AAC with their permission, as detailed on the licensing page:
08:19:17 Join Moos [0] (
08:19:42saratogathough the issue of software patents is not cut and dry in many countries like it is in the US
08:20:22 Quit saratoga (Client Quit)
08:24:40midkayamiconn, morning :) can you tell me how to get the effect lcd_clearrect used to achieve before your new-fangled drawing modes? :)
08:24:59 Quit damaki__ (Read error: 110 (Connection timed out))
08:25:26amiconnIt's rather simple and straightforward, and I had to convert all plugins to use it
08:25:46midkayi see DRMODE_INVERSEVID all over..
08:25:51midkaybut it seems to always work differently :)
08:25:53amiconnclock.c was a huge effort, since it has tons of linear code calling lcd funcs over and over
08:26:04midkayi never figured out how it works.. yeah.. i'm trying to do some simplification right now actually
08:26:09midkayyou should be pleasantly surprised :)
08:26:38 Join damaki_ [0] (
08:26:41midkaymay i quote.. 22.03.35 # <amiconn> Imho the clock plugin shows what happens if you don't care for code size (sorry midk). It is the second largest plugin code-wise, much larger than what I would expect concerning its functionality
08:26:47midkay:)) that gave me a laugh :)
08:27:02midkaythx for the link, i'll read up
08:27:04amiconnset drawmode to (DRMODE_SOLID|DRMODE_INVERSEVID), the use lcd_fillrect()
08:27:29midkayalso, by the way - is there a major difference in battery life/cpu usage between polling buttons at, say HZ/4 vs HZ/10 vs HZ/100?
08:28:17amiconnIt depends on the target, and on what you're doing in the loop itself
08:29:17amiconnAs long as you're waiting, the CPU is sleeping almost all the time, saving power
08:29:29midkayis HZ/100 almost unrealistic in most cases? for a game, say like.. brickmania..
08:30:00Mikachuit makes no sense to redraw the screen and not poll the buttons
08:30:00amiconnIt is unrealistic on a lot of targets
08:31:05amiconnWe can't do lcd refresh @ 100 fps
08:31:21Mikachuright, i got the factors backwards
08:31:31midkaygotcha - thanks
08:31:37amiconnHZ is one second
08:32:10midkaybtw, whoever's responsible for the graphics
08:32:10midkayAPI doc (amiconn?) - nice! well laid-out :)
08:32:12Mikachustrange name for it
08:32:23midkaywtf, why did that end up on two lines..
08:33:15amiconnMikachu: This is because HZ actually is a factor. It tells how many ticks are one second, i.e. how many Hz is the tick frequency
08:34:28amiconnThe timeout is almost everywhere given in ticks
08:35:05amiconnThis way the tick frequency could easily be changed, but all targets use HZ=100 (10 milliseconds) so far
08:35:55amiconnmidkay: (gfx api docs) thanks. Yes that was me
08:36:10midkaybeautiful :)
08:36:34midkaybtw, one more question.. does a function to calculate circle coordinates sound unreasonable?
08:37:10midkayi guess rather unuseful.. thinking about the clock drawing
08:37:19midkayi guess i could always include one in the plugin itself
08:37:34midkayuseful for scaling to different screen sizes, rather than a ton of different arrays
08:37:53amiconnYes. Some plugins already have a table-based fixed point sine function
08:38:13amiconnCheck cube.c, bounce.c ....
08:39:03midkayright, they have their circles set up in arrays..
08:39:42midkayi'd like a function to do that automatically based on screen size.. Linus sent me a script to generate circle tables like that when i first started working on the clock plugin, i wonder if i still have it.
08:40:49amiconnAh, yes, bounce.c uses a fixed table.
08:41:01amiconnThis needs to be changed anyway, maybe I'll do that sooon
08:41:24amiconncube.c has a table based sine function which can in turn be used to calculate circle coordinates
08:41:47 Quit damaki (Connection timed out)
08:41:59midkayamiconn, right, i noticed bounce is screwed up for other sizes - it uses LCD_WIDTH as the center of the clock hands, but the table remains 112x64-sized..
08:42:13midkayi'll check out cube..
08:42:19amiconnLooks funny on H300...
08:42:33amiconnH140 has no rtc, so the problem doesn't arise
08:42:33midkayyeah, funnier on ipod 5g :)
08:42:39midkayah, right..
08:42:48midkayondios don't either, right??
08:43:04amiconnyep, but they wouldn't have a problem if they did
08:43:13amiconnThey have the same lcd as the recorders
08:43:35midkayright - just checking for clock compatibility..
08:43:56amiconnclock should (ideally) run on:
08:44:16amiconn(already supported) recorder v1, recorder v2, fm recorder.
08:44:19midkayit should be compatible for every model with an RTC now (H300, iPods and archos recorders).. i actually haven't been able to find anyone to test the keymap i've got for the H300, maybe you're up to the task? :)
08:44:25amiconn(new) H300, all iPods
08:45:04Mikachudoes rb have timezone data and stuff?
08:45:37midkayi have it set up to "fake" the LCD width and height - the result is a small display on larger-LCD models.. temporarily :)
08:45:40amiconnIt just uses local time, as do most non- *nix OSes
08:45:49midkayafter cleaning up the code a bit more i'll commit it like this, and then get to scaling it up..
08:46:01 Join linuxstb_ [0] (
08:46:05 Quit linuxstb (Read error: 110 (Connection timed out))
08:46:58amiconnMikachu: One goal (even requirement) of rockbox is to be compact and simple.
08:47:22amiconnIt is running on an embedded device, wasting resources would be a bad idea
08:47:32Mikachui have no idea how much resources time zone data would take
08:47:52amiconnIt would definitely take code and table space
08:48:19amiconn..for functions which aren't needed at all
08:48:37amiconn- FAT32 time stamps are local time
08:48:53amiconn- some targets don't even have a real time clock
08:50:21*amiconn wonders why people asking for features which aren't really appropriate for embedded devices over and over
08:50:30*amiconn wonders about a lot of things recently
08:50:35Mikachui was just thinking about dst
08:51:55amiconnAutomatic DST could be supported w/o timezone data
08:52:05Mikachui could never get the time set correctly in apple os, either the status bar was right or the world clock was, never both at the same time
08:52:32Mikachuit would need the dst info for at least one country
08:53:01Mikachubut that is probably not so large
08:53:43amiconnIt could be a user setting; rockbox could even provide .cfg files with that info for various countries
08:53:48Mikachuand i guess if you wanted a world clock, that plugin could have the data
08:54:08*midkay sniffs out a new clock feature
08:54:12Mikachubut it's not as easy as just setting two dates, it's always between a saturday and sunday i think
08:54:27Mikachuat least in sweden :)
08:54:36amiconnYes, they aren't fixed dates, they include some calculation
08:54:57amiconnThere seems to be a quite common format to give dst data:
08:55:36 Join Mark_ [0] (
08:55:49 Join yeahx [0] (
08:56:02amiconn<month> (<day> / <weekday> <no-of-week>) HH:MM
08:56:23amiconnno-of-week can be either 1..4 or 'last'
08:56:34 Join SereR0kR [0] (
08:56:49amiconnIn germany it's March Sunday last 02:00 and October Sunday last 03:00
08:56:58 Join Paul_The_Nerd [0] (
08:58:43 Join damaki [0] (
09:05:11 Quit Arrogant ("Leaving")
09:08:28 Join Administrator [0] (
09:08:30Administratorit's so terrible the gerbil has to die
09:09:00midkayyes, it's just awful.
09:09:14 Nick Administrator is now known as orthogonal (
09:09:21Mark_novimon, the gerbil was a dick
09:09:25Mark_he had it coming
09:09:37Mark_noviwhat now?
09:09:49Mark_stupid xchat thinking i was refering to someone
09:10:02Mark_i said "no, the..."
09:10:19 Quit damaki_ (Read error: 110 (Connection timed out))
09:10:25midkayuncheck "automatic nick completion" in the Input Box section of Settings-Preferences.. :)
09:10:26orthogonalanybody know earHurts? Or earHertz?
09:10:32midkayi do! did.
09:10:45orthogonalmidkay: minight2k?
09:10:46midkayit's you.
09:10:56midkayyes, that was me :)
09:11:12orthogonalI'm flattered anyone recaslsl me
09:11:27orthogonalMy archos broke, you see
09:11:30midkayhaha. i could never forget! you're such a weirdo ;)
09:11:36orthogonalwell, yes ;)
09:12:20orthogonalthe guys I sent ithe archos to for repair screwed it up
09:12:30orthogonalBut now I have an ipod
09:12:36midkaygeniuses.. oh, very cool :)
09:12:38midkaywhat model?
09:12:53midkayawesome, same :)
09:13:11orthogonaloh really?
09:13:19midkayyep :)
09:13:26orthogonalhave you installed rockbox on your ipod?
09:13:34midkayyes :) hahaha
09:13:51orthogonalMark_: how much code have you contrbuted to the Rockbox project?
09:14:07midkaygive him a break. he's an AOL user.
09:14:22Mark_about 20 lines of wps code, and a few images which are used in cvs distributed WPSs :P
09:14:26Mark_so fuck all basically
09:14:51orthogonalMark_: I did a bit og the id3 stuff. Genre and composers support
09:15:09orthogonaland that poer-saving proff-of-concept;
09:15:39orthogonalwonder if anyone ever got any use from teh poeer save business?
09:15:59midkayorthogonal, of course not ;)
09:16:23orthogonalwell, once my archos broke down, it was ghjhard to keep testing, ahem.
09:17:25***Saving seen data "./dancer.seen"
09:17:29orthogonalSo midkay, how's rb on the ipod?
09:17:36midkayorthogonal, delicious :)
09:17:50orthogonalbut it doesn't do video, right?
09:18:00midkaynope, not yet..
09:18:17midkayyou can boot between the original OS and rockbox right on the device, though..
09:18:25orthogonalbut it does sound on the video ipod?
09:18:46midkayyes.. there's an EQ that kills MP3 playback as it is.. but FLAC and some AAC works wonderfully
09:18:47orthogonalfrom ipod to rb, or just rb to ipod, like onm teh archos?
09:18:49midkayas does MP3 with the EQ off..
09:19:10midkayum, you just hold menu while it boots to go to the original firmware.. it's not like ROLO
09:19:12orthogonalwait, if you use an eq zetting, mp3s don't play?
09:19:32midkayit seems to slow down buffering majorly or something - the buffer empties quick and playback stutters after that..
09:19:43orthogonalhow odd
09:20:00orthogonalis teh dev team mostly the same, or not?
09:20:05midkayyeah, flac works great though
09:20:24midkayhm, most of the same old, many new :)
09:20:27orthogonalwell, sure flac is simple
09:20:49midkayno decompressing to do - still, with the bitrate of the files i'd expect it to have a bit more trouble..
09:21:10orthogonalhow's power comsumption on rb compare to ipod firmware?
09:21:29midkaygeneral consensus seems to be "less".. battery meter isn't enabled yet, so i can't really tell..
09:22:20midkaycool to have and mess with but not a perfect alternative quite yet - some cool fullscreen, color games though :)
09:22:21Paul_The_NerdLess battery life he means
09:22:32midkayoh, yeah.
09:22:53Paul_The_NerdAlso, ogg tends to work better than MP3 at the moment, at the same bitrates.
09:23:02orthogonalMy main beef with teh ipod is teh wps screen
09:23:27midkayorthogonal, the apple OS one?
09:23:37yeahxI need to find a nice ogg setting and convert a bunch of stuff to test it out better
09:23:42orthogonalin fcat, the same thing that impelled me to contribnute to rockbox way back when: no genre or composer on teh wps
09:23:47midkaykind of a waste of space, yeah.. i don't mind it too much though
09:23:54orthogonalmidkay: yes, the apple firmware
09:24:03yeahxI thought I had a good setting but its telling me the bitrates are a little high in some
09:24:05midkayorthogonal, crossfading on rockbox though - that's very cool :)
09:24:16midkay+ gapless..
09:24:22orthogonalwait, rb can xfade?
09:24:28midkayyes :)
09:24:28amiconnhmm, no bagder...
09:24:44orthogonalwe couldn't do that on the archos
09:24:54Paul_The_NerdSince it's a software decoder, they can now.
09:24:55midkayright, only for software codec platforms i believe
09:24:59Paul_The_NerdOn the swcodec platforms
09:25:15orthogonalPaul_The_Nerd: mp3 is sw or hw on the ipod 5g?
09:25:23midkayorthogonal, there's also wav and ogg support - haven't tried those yet
09:25:36orthogonalwell, wav I don't want, for the obvious reason
09:25:43midkaymassive ;)
09:25:50orthogonaland ogg, I don'rt want to transcode my mp3s.
09:25:51yeahxthere is doom for rockbox now?
09:26:09Paul_The_Nerdorthogonal: Everything is software on everything non-archos so far.
09:26:16 Join aliask [0] (
09:26:41yeahxits handy if you find something only in ogg, there are a few things out there like that
09:26:55orthogonalPaul_The_Nerd: on teh apple firmware it's hardware though, right?
09:27:01midkayyeahx, it runs for me on my ipod 5g - seems to work on nanos and H300s too
09:27:23orthogonalmidkay: how long have you been riunning rb on the 5g?
09:27:35yeahxhow do you get it working on a nano? where do you get it?
09:27:36Paul_The_Nerdorthogonal: No, it's software.
09:27:51orthogonalPaul_The_Nerd: REALLY?
09:27:55orthogonalthat's interesting
09:27:58midkayorthogonal, a month perhaps..
09:28:06orthogonalmidkay: any probs?
09:28:08Paul_The_Nerdyeahx: It's in the patch tracker.
09:28:16midkayhaven't used it much until the past week or so
09:28:22yeahxId like to try it out just to show to friends
09:28:25midkaywhen audio support + faster LCD driver kicked in
09:28:27yeahxaah ok, nevermind
09:28:29orthogonalmidkay: arrgh! ;)
09:28:38midkayorthogonal, other than mp3 stutters with eq on - no, not really :)
09:28:52perplexityplease don't tell me rockbox is functional with audio on the ipod video.. coz then I just will have to go and get one..
09:28:52Paul_The_Nerdorthogonal: At least, I'm pretty sure it's software. If you use the debug mode, you can see the how much the load changes as the bitrate changes in a vbr file.
09:28:54orthogonaldidn't I help you write a .rock one time?
09:29:04midkayperplexity, i won't say a word ;)
09:29:08perplexitythanks :)
09:29:16midkayorthogonal, the clock? yes :)
09:29:18orthogonalPaul_The_Nerd: how much does the load change anyway?
09:29:21midkayyou're still in the credits :)
09:29:36orthogonalmidkay: are you doing any coding these days?
09:29:41orthogonalNo shit?
09:30:04midkayhaha. getting back into rockbox - almost done with a clock update adding more model support..
09:30:05Paul_The_Nerdorthogonal: On the iPod nano, with some vbr files I've got, where there's silence, and then sudden complexity, the load doesn't actually change fast enough and the Retail OS has a buffer underrun and the music skips. Rockbox doesn't have this problem. :)
09:30:09midkayand space invaders :)
09:30:30Paul_The_NerdI don't remember the exact change in load, but it was enough that they didn't handle it well.
09:30:45Paul_The_NerdApparently it's a common problem with lame-encoded VBR music.
09:30:50orthogonalPaul_The_Nerd: you're suggesting teh rockbox sw is superior??
09:31:13Paul_The_NerdWell, my music skipped in the apple software. It does not in rockbox.
09:31:37Paul_The_NerdTake that how you will, I have a Nano so I don't have the slow LCD issue on the 5g (which I believe affects things very little now, compared to before)
09:31:39orthogonalI know rb is far superior to archos's crap sw, but I'd thought apple would do better
09:31:54midkayPaul_The_Nerd, it's running quite well now (5g lcd)
09:32:39midkayorthogonal, other than the inability to use EQ with mp3's.. i'd say definitely rockbox, unless you play a lot of videos - crossfading and gapless is nice, as well as all the options + plugins + custom WPS stuff..
09:33:05orthogonalI only have a few videos
09:33:17orthogonalBut I have a lot of podcasts
09:33:18midkayyou can always quickly boot back to apple os for that
09:33:26Paul_The_NerdI think EQ with MP3s depends on your mp3's bitrate.
09:33:26midkayhm, i'm not sure how those work..
09:33:42midkayPaul_The_Nerd, i've done it on as low as 160kbit mp3s.. always seems to kill playback
09:33:46orthogonalmidkay, I don't know how much coding you've learned since we last spoke
09:33:53midkayorthogonal, hopefully a lot :)
09:34:02orthogonalbut UCBerkely hasd free podcasts of programming classes if you're interested
09:34:18Paul_The_NerdThough at the moment, on Nano, rockbox boosts something like 68% of the time with a 128kbps mp3, and 74% with a 320, which seems a bit small of a range.
09:34:20midkaynever tried podcasts, cool - maybe i shall soon :)
09:34:47midkayPaul_The_Nerd, what do you mean?
09:35:03midkaybattery life? boosts? i thought it was lower, was i wrong?
09:35:19Paul_The_Nerdmidkay: I've got a patch that makes the processor behave like it does on the iriver. It runs at a low speed (30mhz) and boosts to 75 when the PCM buffer gets low, to refill it.
09:35:43Paul_The_Nerdbattery life on rockbox is definitely lower than with retail OS at the moment, yes.
09:35:46midkayPaul_The_Nerd, awesome - why isn't it in cvs? :)
09:36:01Paul_The_Nerdmidkay: It's not my patch, for one thing.
09:36:07Paul_The_Nerdmidkay: It also has problems on 4Gs.
09:36:23Paul_The_NerdI can't remember which of linuxstb and preglow made it, at the moment.
09:36:24midkayhm. sounds pretty cool..
09:36:37Paul_The_NerdIt'll improve battery life, once it works.
09:37:17midkaygreat - i just can't wait for a faster mp3 codec or something that makes it more efficient
09:37:34midkayeq sounds great, but it's a shame i can't use it for now.. and bass/treble doesn't work either
09:37:34Paul_The_NerdI think it's less of needing a faster mp3 codec, as just optimizing the one we've got.
09:37:46midkayyeah, that :)
09:38:21Paul_The_NerdSince at the moment it's slower than the ogg vorbis codec on iPods, when it's faster on iRiver. But the ogg one had built in arm optimizations that the mp3 one doesn't, if I understand correctly.
09:38:40midkayah, i see
09:39:00orthogonalmidkay: did you do that foobar2000 itunesdb rebuild thingie?
09:39:23midkayorthogonal, no, i don't have much music so i just recopied it all to a seperate folder with artist -> album -> song structure
09:39:31midkaythat was before i heard of the rebuild thing though
09:39:35midkayi'm think of doing that
09:39:53orthogonalI have about 40gb and another 6 gb yet to add
09:40:26 Join Higgy_ [0] (
09:40:33 Quit Strath (Read error: 104 (Connection reset by peer))
09:40:38midkayhaha, nice
09:40:45orthogonalal;l legal, from emusic, btw
09:40:53midkayah, yeah, you're that emusic whore ;)
09:41:52 Join Strath [0] (
09:41:56orthogonalI was
09:41:59orthogonalI left emusic
09:42:14orthogonalbut I did write a freeware tool to allow emusic user to list what they'd downloaded
09:42:21midkaycool :)
09:43:14orthogonalyou use firefox?
09:43:37orthogonalheare, I wriote this too:
09:44:39midkayvery cool! *installs* :)
09:44:55midkayi usually use ctrl+w .. but definitely would be nice to have :)
09:45:05orthogonalit's useful if you have lots of tabs
09:45:13midkayi'm a tab user alright..
09:45:27orthogonalbe sure to check the optiosn dialog to set it up
09:45:27midkayctrl+w is usually quickest for me, i might have to change that habit though :)
09:45:47orthogonalOh, and you''ll have to set your mosue to send alt+left arrow
09:47:58midkayi'll have to install my mouse software for that..
09:48:19midkaythanks though! very useful :D
09:48:23midkaynice job!
09:48:25orthogonalFirebox doesn't giave a way to catch teh "Back" special key. ;(
09:48:31midkayyeah, damn
09:51:11midkayso what's up?
09:51:39orthogonalNot much
09:51:45 Join tvelocity [0] (
09:51:48orthogonalJust got the ipod for xmas
09:52:01midkaycool :)
09:52:04orthogonalDidn't realize rthe rb for ipod was worjking until tonight
09:52:17orthogonalI checked in jnuary, figured it would take longer
09:52:20midkayhaha, awesome :)
09:52:34orthogonaland now I'm going to install it.
09:52:49orthogonalthe lack of genre and composer is getting to me, see ;)
09:53:17midkayhahaha :)
09:53:47orthogonalDo you play much classical music?
09:53:59midkayno, none :)
09:54:23 Quit yeahx ()
09:54:30orthogonalmuch classical music has simuilar names. Indded teh "song" names are really directions how to play it: "fast", "slow", whatevcer
09:54:47orthogonalwithout a composer, it's a bit opaque what you're playing
09:54:57midkayhaha, ah, i see :)
09:55:00orthogonalbut I hate putting the composer in the album name
09:55:08orthogonalbecause that just sucks ass
09:55:23midkayWPS' are really fun now - tons more tags and much larger screen = yay :)
09:55:29midkayand bitmaps..
09:55:30orthogonal(although if tqo differnt albums have teh same name, itunes thinks thwey're th esame album)
09:55:52orthogonalheh. I wonder if they used my tag schme or not
09:55:52midkaydarn apple
09:55:59midkayhaha, what?
09:56:09orthogonalwhenm I added composer and genre (and a few othwers)
09:56:22orthogonalI made an estensible way to add more, that would also be efficuient.
09:56:40orthogonalbut it wasn't the most obvious code, because it was trying to be efficient
09:57:11orthogonalyou remember how we had that code size limitation on the archos?
09:57:31orthogonalSo all id3 tags are four characters long
09:57:51orthogonalI set it up so they weren't treated as stringsm, becasus ethat adds te h0 ternm,inatyor
09:58:13orthogonaland a double indirection
09:58:28midkayhaha. you and your technical mumbo-jumbo :)
09:59:33orthogonalan array of pointers to arrsays exacrly for char in length
09:59:52amiconnWe still need to keep more than one eye on code size, even though the hard 200KB limit on archos is lifted now
10:00:49midkayorthogonal, .. ok ;)
10:00:53orthogonalamiconn: yeah, the other problem wa sthat we (at that time) only allocated 300 bytes for all id3 tags, and that was statically allocated
10:01:15 Quit tvelocity ("Αποχώρησε")
10:01:16orthogonaland I tend to have long id3 tags
10:01:23amiconnWe would get into trouble again if the compressed size approaches 200KB, but there are other things which should us keep farther away from that limit
10:02:14orthogonalamiconn: we compress for the archios, then decompress into memorty?
10:02:49amiconnRockbox is self-decompressing for archoses were we are above the load-size limit of the original firmware
10:03:35amiconnThe load size limit is 200KB for player and recorder v1, and 400KB for recorder fm/v2 and ondios
10:03:59orthogonalyeah, I had a v2
10:04:03amiconnThe only target where we need to compress right now is recorder v1
10:04:11orthogonaluntil the piece of crap broke
10:04:21orthogonalI even installed a 60gb drive in it
10:05:10amiconnI have a recorder v1 upgraded to 80GB and white backlight mod. Almost 4 years old, still working ok
10:06:16amiconnThere is another size limit I wouldn't want to exceed: With a flashable box, it's possible to run rockbox directly from flash rom, which implies it needs to be uncompressed
10:06:37orthogonalright, the german guy's thing
10:06:43orthogonalyeah, I had that installe dtoo
10:07:19amiconnThe flash rom is 256KB, and the flash loader + bootbox (this combo is functionally equivalent to the bootloader we have on iriver) take 28KB
10:07:41amiconnSo I would prefer to stay below 224KB uncompressed size for all archos targets
10:08:01orthogonalmakes good sense
10:08:06amiconnAll my archoses are flashed and running rombox :)
10:08:24orthogonalok, I've installed rockbox on the 5g
10:08:28orthogonalI think
10:09:14orthogonallittle tiny fonts, I love it
10:09:37midkayi recommend xtal-14 - unless you want very tiny :)
10:09:40orthogonalOh god, i recall how much better rockbox was than teh archos firmware
10:09:52midkayxtal-14 is rather small, but .. considerably bigger than 'tiny' :)
10:09:57midkayjust right for me
10:11:39orthogonaljeeez, th efdaily build page is a lot more complicate dthan I recall
10:12:13midkayhaha. so many more builds now :)
10:14:31midkaybtw - directory cache is a nice setting (general settings - system - disk - directory cache) if you do a lot of browsing, reads your HD contents at boot and stores them in memory so you don't have to spinup to browse, just to load files
10:14:48amiconnorthogonal, midkay: Want more and larger fonts?">
10:15:20midkaywhoa! that's a _lot_ of fonts.. thanks :D
10:15:46orthogonalNah, I just need two fonts: a tiny and a big
10:16:15amiconnI still have to check these out; I want a really big font for my H340 for in-car use
10:16:30orthogonaler, good pint
10:19:02 Join safetydan [0] (
10:20:49orthogonalwhat's with all th ebmps?
10:24:51ghode|afkthey are for the WPS themes
10:24:57midkaybitmap support for the WPS..
10:25:10orthogonalneedless luxury
10:25:18midkayhaha, yes :)
10:25:41ghode|afkfor youmaybe ;)
10:25:57midkaysome very cool bitmap-oriented WPS' though :)
10:26:26orthogonaloh sweetness!!!!!!!
10:26:40orthogonalafter all these years, back on rockbox!!
10:27:27midkayoh yes, feels great, huh? :)
10:28:24 Join damaki_ [0] (
10:29:24orthogonalmidkay: my main screen is blank
10:29:38midkayorthogonal, what?
10:29:49midkayshow files = all?
10:30:23midkaydid you change anything? dircache?
10:30:58orthogonalcould be
10:31:22orthogonaldid a reset
10:31:27midkayyeah, see if that helps
10:31:34orthogonalstill nothibg
10:31:45midkayvery weird, hm
10:31:54midkaydid you try to restart rockbox?
10:31:58midkayreboot the ipod
10:32:10safetydanthe optimised build really triggers some odd problems
10:32:28*safetydan wonders if sanity checking of eq settings is needed
10:33:07orthogonalmidkay: thast works
10:33:26midkayorthogonal, weird but.. good :)
10:34:01*amiconn wonders why the optimised build is called optimised build
10:34:27orthogonalhow do I travel to the parent dir?
10:34:32Mikachupress left
10:34:35orthogonalit's not the mneu key as in teh ipod
10:34:36amiconnImho it should be called buggy custom build
10:34:42midkayamiconn, hahaha
10:34:47*Paul_The_Nerd agrees with amiconn.
10:34:50Mikachuamiconn: that doesn't have quite the same ring to it
10:35:12orthogonalit plays!
10:35:22Paul_The_NerdAnd then they made an "h100 optimized" build, and someone at misticriver suggested someone make an ipod one. =/
10:36:10orthogonalwhat's the best wps?
10:36:22ashridahcall it rockbox MAX
10:36:28ashridahmaximum taste, no function
10:36:34*ashridah nods
10:39:00Paul_The_NerdMy personal favorite is "Rockbox_default" with a small piece of art along the bottom half of the screen, and the peakmeter removed.
10:39:12Paul_The_NerdNot album art, mind you. Just a picture.
10:39:17orthogonalsomething's fucked up
10:39:32orthogonalit's showing id3 data not from what's being played
10:39:41midkaythat's the next track
10:39:49midkayin playlist
10:39:53orthogonalyou're kidding me
10:39:59midkayje ne kidding-you pas!!
10:40:02orthogonalholy crap
10:40:35 Quit novimon (Read error: 110 (Connection timed out))
10:41:11 Join ender` [0] (i=ychat@
10:41:46orthogonalIt';s doesn't do euro-accented characters?
10:41:57Mikachuit does, even japanese works fine
10:42:32safetydanIs there a clamp function/macro anywhere in Rockbox?
10:42:47orthogonalGötterdämmerung is showing as G tterd mmerung
10:42:53MikachuMAX(5,MIN(0,x)) ?
10:43:10Mikachuer, i always write those in the wrong order
10:43:31safetydanMikachu, yes I can write one, just wondering if there was one I could reuse
10:43:36Paul_The_Nerdorthogonal: Depends on what font you're using. Try Unifont.
10:43:40safetydanthe less code I write the better
10:44:01 Quit damaki (Read error: 110 (Connection timed out))
10:44:37orthogonalunifont is ugly
10:44:55orthogonalbut, yes, it does work.
10:45:25Paul_The_Nerdorthogonal: Yeah, the one mikachu mentioned also works. There's only a few unicode fonts so far.
10:45:57orthogonalsnap works, and isn't so ugly
10:46:20Mikachuif you want to draw a screen with two different fonts, would that be very slow?
10:46:34safetydanhrmm... clapming EQ settings on load isn't going to work as then they'll just get in-range crazy settings (like all bands +24 dB)
10:46:44orthogonalMikachu: under teh code a couple yeras ago, there was only mmeory for one font
10:46:52orthogonalbut I'm way out of date
10:47:11Mikachui was thinking theoretically about writing some japanese lookup progarm
10:47:42orthogonalhrist lots of .rocks
10:49:47orthogonalhow d o i escape from mandelbrot?
10:49:59Mikachumenu+select on ipod4g
10:50:03midkayselect+menu usually
10:50:52amiconnorthogonal: xtal-14 should work for accented chars
10:51:51 Join XavierGr [0] (
10:52:06orthogonalwrap arounf scrolloing is sweet'
10:52:42orthogonalztal is good looking
10:53:12midkayorthogonal, you bastard - i recommended it like an hour ago. :)
10:53:23 Join JdGordon [0] (
10:53:34orthogonalsame old midkay
10:53:40midkayfu man! ;)
10:55:15orthogonalhow do i get back to thre file system browser from the wps?
10:55:16 Join tvelocity [0] (
10:55:34Mikachuyou won this time
10:55:48midkayhaha. *narrows eyes and readies fingers for next question*
10:56:29orthogonalok, this is a hard one.
10:56:36*midkay == ready
10:56:52orthogonalhow's rb playback quality compare to ipod firmware playback?
10:57:11midkay*dances* IN YOUR FACE!
10:57:22Mikachui don't hear the difference between 128kbps mp3 and 320kbps ogg
10:57:42orthogonalanyone recorded putput to cpmpare it?
10:58:13MikachuAUDIO: 44100 Hz, 2 ch, s16le, 64.8 kbit/4.59% (ratio: 8100->176400)
10:58:19Mikachuwhat i'm listening to now
10:58:44midkayhaha. ew.
10:58:55Mikachunot my encode though, i usually ogg at 100 :)
10:58:55orthogonalrb plays unprotected aac?
10:59:01markunorthogonal: will you do a test?
10:59:15orthogonalmarkun: eventaully i will, yes
10:59:22orthogonalbut not for about a month
10:59:36orthogonalI need to refamilairize myself with teh rockbox codebase
11:00:04orthogonaland talk to whomever is doing teh rockbox ipod coding
11:03:08orthogonali hate to say it, but the apple wps looks "cleaner"
11:03:36Paul_The_Nerdorthogonal: How do you mean "cleaner?"
11:03:46orthogonalmore attarctive
11:03:59Paul_The_NerdBut... you can reproduce the apple WPS almost exactly in rockbox...
11:04:10orthogonalPaul_The_Nerd: is that apreset?
11:04:15 Quit San (Read error: 104 (Connection reset by peer))
11:04:21orthogonala preset, or do ai have to write teh wps myself?
11:04:28Paul_The_NerdNot really. The "Themes" that a re included are just those submitted by people.
11:04:47orthogonalyes, is there an existing applle-looking theme?
11:04:51midkayPaul_The_Nerd, well, also there's a lack of font anti-aliasing (even if it is fake)..
11:05:02Paul_The_NerdThere is that, I suppose.
11:05:11 Join webguest98 [0] (
11:05:19midkayorthogonal, it's easy to set one up..
11:05:20Paul_The_NerdBut then I use a tiny font anyway, so I wouldn't want it.
11:05:34midkayi could send you mine - it's pretty damn similar :)
11:05:41midkayusing xtal-14?
11:05:44orthogonalmidkay: would you?
11:06:27webguest98what is GPIO_FUNCTION in the source? for ex: in button_init() GPIO_FUNCTION |= 0x08000000; GPIO_ENABLE &= ~0x08000000; what are these doing?
11:07:21orthogonalwebguest98: they look like consrants, but tehy're beingused as variables
11:07:44webguest98??? to do what?
11:08:11perplexitythey are IO ports..
11:08:11 Nick webguest98 is now known as jsmitz (
11:09:35jsmitzisn't a port a 16bit?
11:09:40jsmitza port number
11:10:36perplexitywhich particular architecture are you refering to?
11:11:25amiconnAll our targets have memory-mapped i/o
11:11:47jsmitzaudio, but all of them mess with GPIO_FUNCTION..where is it defined
11:12:03amiconn...and are 32 bit archs (except the gmini target which is currently inactive)
11:12:12perplexityyes, I was more asking which architecture he got that particular snippet from.. I was going to point him in the direction of the relevant cpu manuals :)
11:12:23amiconnThe data width varies, can be 8, 16 or 32 bit
11:12:47amiconniaudio is coldfire, and coldfire gpio is 32 bit
11:12:56amiconnCheck mcf5250.h
11:13:16amiconn(or mcf5249.h for iriver)
11:13:27amiconnThey're almost identical
11:14:21jsmitzwhat path? didn't see in firmware/drivers
11:15:15jsmitzfound it;
11:17:27orthogonalhow muxch "excees" perocessor power do we have on teh ipod?
11:17:28***Saving seen data "./dancer.seen"
11:17:53jsmitzMBAR seems to be the base address? where exactly is MBAR at? on the coldfire chip itself?
11:18:51perplexityit's defined in crt0.S
11:19:16 Join muesli__ [0] (n=muesli_t@
11:20:28orthogonalamiconn: why's gmini inactive?
11:20:56jsmitzokay, I'm lost...what do I need to read to learn how memorymapped io works on coldfire?
11:21:13perplexitywhat are you trying to do?
11:21:21jsmitzcause I thought io messed with devices on the bus & not the chip (coldfire)
11:21:30perplexitythe IO registers are just mapped as memory..
11:21:33orthogonaldoesn't mmeory mapped just mean you cam read/write teh memory and let teh hardware wory about the rest?
11:21:54perplexityyes, but the coldfire has quite a lot of hadware built in, so you use internal registers to configure that hardware also
11:21:59jsmitzjust trying to understand rockbox code (and hardware programming in general)
11:22:00orthogonalno spceail INP OUTP commands
11:22:04perplexityyes orthogonal
11:22:10amiconnNo, it just means the i/o registers are mapped into the same address space as the memory
11:22:48orthogonaldecalte your variables volatile you access them like you would access memory addresses
11:23:14orthogonalyes, no spedial handling
11:23:27jsmitzso GPIO_FUNCTION is a register on the chip or adevice on the bus?
11:23:46orthogonaljsmitz: it's an abstraction so you don't have to ask that
11:24:13jsmitzok...but I hate being in bliss...
11:24:32orthogonalas Koenig says, "abstraction is selective ignorance"
11:25:35perplexitylook at the GPIO as hardware pins connected directly to a memory address.. or grab the coldfire manual for a pretty detailed schematic of the GPIO pin configuration
11:26:20jsmitzso GPIO is cpu specific & not device specific?
11:27:04perplexityGPIO in this context are a load of pins on the cpu.. so yes..
11:27:23perplexityGeneral Purpose Input Output .. GPIO
11:27:48amiconnWhat is connected to the various GPIO pins is of course device specific
11:31:01jsmitzahh...ok. I thought GPIO was just a 4byte memory location...but see it's 32bits each with a different function [pertaining to controlling devices]?
11:31:25perplexitypretty mich
11:31:37perplexitythough I'm sure mich is pretty also
11:32:29jsmitzok...that gets me one step closer to understanding hardware
11:33:50jsmitzok, so the iriver & archos targets have the GPIO pins pretty much defied (from the twiki) but theiaudio x5 only has a couple...does this mean it needs to be researched what pin does what?
11:34:22 Quit damaki_ (Remote closed the connection)
11:34:42jsmitzhow hard is it..or how would you identify each function?
11:34:59amiconnFollow the traces on the board...
11:35:01perplexitytrace the tracks on the board..
11:35:13perplexitydisassemble the original code..
11:35:29perplexitytoggle them in software and see where the smoke comes from..
11:35:40jsmitzso pretty hard I take it?
11:35:51perplexitymore time consuming
11:35:52jsmitzit's not something you can look up in a book
11:36:28XavierGrdoesn anyone has a link that shows how to open an H300?
11:37:07jsmitzso the only real way of knowing what each does is by having test code flashed to the ROM? cause you obiviously can't use the simulator (sdl) to find out what each does?
11:37:26perplexityor toggle them with the bdm..
11:38:38jsmitzso is this GPIO function definition process the current bottleneck for the iaudio x5 target?
11:38:44JdGordonhahah BHSPitLappy
11:39:33perplexityjsmitz: as with all targets, it's really a matter of figuring out what is connected where and then how to drive it.. so it's part of the process.. yes
11:40:33BHSPitLappyever get that print
11:40:43XavierGrBHSPitLappy: Thanks but no...
11:41:07jsmitzand only 1 person can do that, the one with the bdm..everyone else can only hypothesize what the pins do I take it
11:41:12 Join damaki [0] (
11:41:32perplexityjsmitz: no.. you can open your unit up and trace the tracks on the board :)
11:42:05JdGordonBHSPitLappy: no, not yet
11:42:07JdGordonstill waiting
11:42:24jsmitzbut I wouldn't really know what it does unless I could toggle it though?
11:42:31jsmitz...I thought
11:42:59perplexitywell, you know both ends of the trace.. and you have datasheets for both chips.. so you should know what it does :)
11:44:40jsmitzso just by knowing that pin X on the cpu connects to pin Y on say ATA controller, the ata controller (in documentation) will say that toggling this pin does Z?
11:45:05perplexitythat's the theory.. yes..
11:45:20jsmitzin that case...sounds like fun! :p
11:45:23ashridahdepending on the state of the ATA controller, and whether it's read on a rising edge of the clock or whatever
11:45:27perplexityif it connects to say pin 27 on the ATA controller and the manual says pin 27 is the enable pin.. you know what it does
11:45:30jsmitztrying to match 'em up! do you trace pins..when they disappear on the pcb?
11:49:06perplexitywell, they never really dissapear.. they just change layers or hide under components..
11:49:45ashridahjsmitz: in many cases, you take a dead version, and use a scanner on high resolution to get a clear image, and recreate a schematic from it
11:49:49amiconnOne option is what Linus did for the iriver H1x0:
11:49:58ashridahafter you strip the chips off it, i mean
11:50:14amiconnHe acquired a broken iriver and stripped all components from the board
11:50:49jsmitzhow many layers can there be? even if you have all the components stripped, how can you see "inside" the pcb?
11:50:52ashridahof course, this tactic is a bitch if the board has more than 2 layers
11:51:02Mikachui hear some pc motherboards have 5 or more?
11:51:03ashridahjsmitz: uh, some boards have 6 :)
11:51:22ashridahsome of which are often ground planes to reduce noise
11:51:27amiconnWell, every trace end must be on one of the outer layers
11:51:28perplexityI have an 8 layer board here
11:51:47jsmitzwhat's the point? lack of space or bad design ?
11:51:58ashridahjsmitz: good engineering
11:52:17amiconnFollowing a trace that runs thorugh inner layers is a bit difficult but doable. It has to reappear on an outer layer somewhere
11:52:31jsmitzmore layers = better?
11:52:40ashridahit can be used to reduce noise, save space, reduce cost
11:53:07Mikachuis it possible to use something like the connected tester of a multimeter?
11:53:10perplexityworst case when the board is stripped you use a fine point on one meter probe and a small brush on the other, then just sweep the board for beeps to narrow down the area of the other trace end
11:54:00jsmitzyou can use a open/short detector on a multi-meter & test all possibilities of pin A on everything else on the pcb...that would work?
11:54:23ashridahperplexity: i agree, after all, it can only connect to a component on the surface on either side.
11:55:13perplexitybut, it requires a sacrificial unit and lots of care removing the chips..
11:55:49jsmitzyou can't glue the chips back on or something?
11:56:20Mikachui doubt it...
11:56:28 Part Paul_The_Nerd
11:56:41Mikachuthe connections won't be as good as before, so the signals would need to be stronger
11:56:44perplexitywell, if you were super ultra careful and really good with smd rework I guess you could dismantle a working unit.. scope it out and reassemble it.. but man.. with the hours involved I'd rather buy a new unit and scrap it
11:58:25 Join Matze [0] (
11:59:15jsmitzI'm going to try doing some traces...sounds like fun...that is after I get my x5 back; had to rma it
11:59:36perplexityjust be careful or you might have to rma it again :)
12:00:17jsmitzif all I do is the "sweep" method, I shouldn't have anything to worry about?
12:00:54 Join BHSPitLappy_ [0] (
12:00:54perplexityoh yeah you will.. don't go jamming a couple of volts from your DMM continuity tester onto the board.. the odds are you *will* kill something..
12:01:06 Quit BHSPitLappy (Nick collision from services.)
12:01:12 Nick BHSPitLappy_ is now known as BHSPitLappy (
12:02:03jsmitzhaha..I just took my car receiver apart & ran continuity tests on random
12:03:05jsmitzperplexity, what if just measure resistance?
12:03:15jsmitzinstead of volts?
12:03:35jsmitzI don't know why you would measure volts
12:03:38perplexityno.. resistance is measured by measuring voltage drop by the meter..
12:04:13perplexityit requires injecting voltage across the terminals.. trust me.. don't go poking a live board or working board with a dmm doing anything other than measuring voltage just killed all the fun!
12:05:11amiconnThere's one situation where one has to strip components in order to trace connections: in case one of the chips is a BGA
12:05:18 Join IcyStorM [0] (
12:05:33amiconnThe iriver CPU is a BGA, but afaik the iAudio CPU is not
12:06:10jsmitzbga = buffer gate array?
12:06:18amiconnBall Grid Array
12:07:06amiconnA BGA package has its connections underneath the package, no way to reach the pads w/o unsoldering the chip
12:07:40jsmitzlike normal pc processors?
12:07:50 Join freqmod [0] (
12:07:50jsmitzwith all the pins underneath
12:08:17amiconnThe older PC processors are PGA and also have all pins underneath
12:08:55amiconn...but you won't find PGA packages soldered onto a small sized board, as PGA requires holes
12:09:40amiconnNewer pentium 4 are BGA
12:09:44jsmitzthe older processors? what's considered new? my athlon xp I guess is pga as it has pins on the bottom that fit into holes on the mobo
12:10:17amiconnYes, athlon is PGA
12:12:31jsmitzdamn I don't know jack about all this stuff! I'm far away from understanding rockbox since I don't understand hardware yet
12:12:55jsmitzall I really care is low-level stuff...I hate UI
12:12:59orthogonaljsmitz: you're going the wrong direction
12:13:10jsmitzhow so?
12:13:11*ashridah notes one usually brushes the surface with an electronics engineering degree.
12:13:18orthogonalthe hardware is abstracted for you, "hidden" behind a common interface
12:13:36orthogonallern the common interface
12:14:07orthogonaldepart from it only as you need to, to impement things the common interface doesn't already do
12:14:34jsmitzI'm more concerned about how the hardware is controlled...I really don't care about software programming;
12:14:57ashridahjsmitz: out of interest, what's your overall goal, besides learning how rockbox works?
12:14:59jsmitzI'm majoring in EE when I go to september; I just wanna get a head start on learning
12:15:04orthogonalwhat is it you're trying to do?
12:15:10jsmitzembedded systems programming
12:15:11ashridah(i might add reading design docs might be a better idea than reading source, btw) :)
12:16:33jsmitzI used to read the pentium system programming manuals but had no examples of source...
12:16:48jsmitzI know I have to read the pdfs & then study the code; I guess i"m just eager
12:17:18jsmitz& wanna hop right into to seeing code in action
12:18:40ashridahjsmitz: i'd pick up a 68k simulator from somewhere (used to use one a few years back, can't remember the name, and i didn't get it from a public location anyway) and poke at it.
12:18:53jsmitzand from all my comments thus can see I have no sense of direction in what I need to learn :)
12:18:57orthogonalor [eek at it
12:19:49jsmitzashridah, is that a physical device or a software program like the sdl simulator?
12:20:18ashridahbut the sdl simulator is a simulator of rockbox's lower level software, mostly
12:20:58ashridaha 68k simulator would actually simulate 68000 instructions, and provide a simple monitor to let you load code into memory, etc.
12:21:32jsmitzbut what could I do with a simulator when I won't have any "hardware" connected to it?
12:21:45*ashridah remembers his earlier years, working with a crappy 68HC11 cpu and an even crappier simulator written by the most annoying lecturer alive
12:21:51jsmitzor would it simulate hardware devices too?
12:21:56orthogonalit simulates teh hardware
12:22:07ashridahit's a simulator, it simulates the cpu and some connected hardware, depending
12:22:34ashridahnote that it's not to be confused with an emulator
12:22:55jsmitzwhat's the diff?
12:24:21JdGordonany1 know where the scrolling code is?
12:24:38ashridahsimply put, an emulator is like vmware. intended to pretend to be a complete, indistinguishable system, including timing and reactions to input, etc.
12:24:53ashridaha simulator is typically a lot less exact
12:25:03ashridahie, it's closer to an approximation in some cases.
12:26:15JdGordonany1..? scorrling text code??
12:26:53ashridahheh. rockbox needs an LXR
12:26:59ashridahwell, i suppose it'd be an RXR
12:27:43 Join Nico_P [0] (
12:28:01jsmitztoo many acronyms...acronymfinder has become my best friend as of recently
12:28:40ashridahah, LXR is basically 'linux cross-reference'
12:29:49 Join funkyhammond [0] (
12:30:06jsmitzok...does anyone use rb source in an ide or are most devs text-only people
12:30:48 Join Lost-ash [0] (
12:32:12funkyhammondhi all.. how do I email Zagor without it bouncing back?
12:32:25jsmitzand ashridah, there's a link to a coldfire emulator from the twiki...I would want an emulator over a simulator, correct?
12:32:26JdGordonfunkyhammond: cut the rubber cord off the email first
12:32:39funkyhammondexcuse me
12:33:32funkyhammondobviously I don't normally deal with "rubber cords"
12:35:41 Part funkyhammond
12:35:55Lost-ashjsmitz: i was thinking from a learning perspective, mostl
12:36:12orthogonalCool, my code is still in id3.c
12:36:15Lost-asha simulator may be an easier platform, but it'd be more for learning assembly than learning to deal with the hardware, i guess
12:36:21 Quit ashridah (Nick collision from services.)
12:36:24Lost-ashbloody dialup
12:37:03jsmitzI guess I need both...don't know 68k asm yet or how hardware works
12:37:16 Join mirak [0] (
12:37:50 Nick Lost-ash is now known as ashridah (
12:39:21jsmitzdamn cygwin is slow...the configure script takes 3 seconds for each 'checking for...' line
12:39:38ashridahyeah, it is
12:40:12IcyStorMHow do I patch rockbox so that I can play Doom on it?
12:40:33jsmitzrockbox can play doom?
12:40:49aliaskIcyStorM: What platform?
12:41:04amiconnashridah: The sim isn't suited for learning asm
12:41:06jsmitzpretty cool
12:41:24IcyStorMIpod 5G (video)
12:41:29ashridahamiconn: depends which simulator you're using, i guess
12:41:42amiconnThe sim simulates the rockbox ui, playback etc, but is compiled for the CPU it is running on
12:41:57ashridahamiconn: i wasn't referring to the SDL sim
12:42:06amiconnWhat else?
12:42:11 Part ashridah ("Leaving")
12:42:16 Join ashridah [0] (
12:42:23ashridahgod damn you xchat
12:42:34IcyStorMLlorean says that it works for iPod but its just a patch
12:42:39ashridahamiconn: i wasn't suggesting the *SDL* sim as a viable platform for learning asm
12:42:57ashridahinfact, i explicitly said that that was more for testing rockbox than hardware stuff
12:43:06jsmitzany general m68k/coldfire cpu simulator
12:43:10amiconnA simulator only simulates behaviour, afaik
12:43:19BHSPitLappyit's not an emulator
12:43:23amiconna cpu simulator is called emulator
12:43:34ashridahamiconn: well, i've used cpu simulators.
12:43:34BHSPitLappythough someone from rockbox did make an emulator of one of the platforms
12:43:35IcyStorMoh its an error :(
12:43:42aliaskIcyStorM: It won't work on the 5G yet, there are changes they have to make to the patch - keybindings at the very least
12:43:52midkaymy 5g runs it fine
12:43:56amiconnBHSPitLappy: Yes, there are emulators for gmini and iFP
12:43:56ashridahthe effective definition is basically an emulator attempts to be indistinguishable, whereas a simulator isn't
12:43:59midkaykeys aren't perfect, but it runs and is playable
12:44:06IcyStorMmidkay can you help me ?
12:44:09ashridah(so an emulator will attempt to give approximate timing and whatnot)
12:44:16midkayIcyStorM, do you have a dev environment?
12:44:24IcyStorMI dont think so
12:44:31IcyStorMthe page is just for group members it says.
12:44:35midkayIcyStorM, well, i guess i could send precompiled binaries..
12:44:39midkaywrong url
12:44:49jsmitzoops wrong console
12:46:17 Nick linuxstb_ is now known as linuxstb (
12:46:27 Join _StarScream [0] (n=am@
12:46:34linuxstbmorning all
12:46:38_StarScreamhi guys
12:47:13jsmitzmorning!??!...ah shit...I gotta sleep soon :)
12:47:20_StarScreami am just looking at the website and i was wondering if there was a page where i could check out which features rockbox has for a particular player
12:47:35BHSPitLappyalmost 6am...
12:47:35_StarScreami.e which functions of the H320 does it support etc
12:47:40linuxstborthogonal: You asked how much "excess" cpu power we have on the ipod. The answer is a whole CPU - the iPods have two arm7tdmi cores, and we currently only use one.
12:47:41 Quit Thus0 ("Leaving")
12:48:00BHSPitLappy_StarScream: what kinds of functions are you referring to?
12:48:34linuxstborthogonal: Plus, on the 5g we have a broadcom video dsp which is also completely unused.
12:48:40orthogonalwhy do we only use one?
12:49:52linuxstbIMO, we don't need it. To save power, we can just keep it sleeping and do everything Rockbox needs to do for audio playback on a single CPU.
12:50:12orthogonalis there any memeory we don't use?
12:50:33coobthe broadcom has 4mb on ram on it
12:50:39_StarScreamBHSPitLappy: well i was just curious if rockbox did everything that the iRvier firmware does..if not what i would be losing out on. I prefer FOSS where i can
12:50:50_StarScreambut i do use some features on the h320
12:50:50orthogonal4mb we don't use at all?
12:50:59_StarScreamlike usbotg
12:51:01orthogonaloh sweet
12:51:03ashridahorthogonal: some ipods have 64MB iirc
12:51:10ashridahwhich isn't being fully utilized
12:51:12coob60gb videos have 64mb ram afaik
12:51:22perplexity_StarScream rockbox does not do usbotg or play mpeg4 video.. thats about it.. Oh.. it does not record direct to mp3..
12:51:25orthogonaloh sweeet
12:51:33linuxstbThe Nano and 5g have 128KB of IRAM - currently Rockbox only uses 96KB of it, because that's how much the other targets have. So that needs fixing.
12:51:34ashridah_StarScream: the H320 can dual boot iriver and rockbox
12:51:44ashridahyou don't lose the original firmware if you've got rockbox installed
12:51:49orthogonalback in the archos days, I wanted to add a tag db to the archos. but i fugured there was too little capacity
12:51:50_StarScreamashridah: hmm that might be worth a look
12:52:03perplexityit's great _StarScream.. get's an A+ recommendation :)
12:52:03BHSPitLappy_StarScream: did you bother typing "iriver" into the search field on :P
12:52:04orthogonalwith an extra cpu and 4mb of memory, i can add a tag databse
12:52:20linuxstbAnd the 60GB 5g has 64MB RAM, compared to 32MB RAM on all other targets, and that extra memory is unused at the moment.
12:52:39ashridahorthogonal: look into tag cache (someone had a tag db going before, but it's been kinda deprecated and tagcache seems to be current, not sure)
12:52:41*perplexity imagines how many tracks can be buffered in that extra space :)
12:52:43linuxstbBut the 5g port has only really been working for about 2 weeks...
12:52:47_StarScreamBHSPitLappy: no i typed h320
12:52:48orthogonalah, that explains why my battery usage is so much higher than with teh apple firmware
12:52:58_StarScreamah got it
12:53:04linuxstborthogonal: Do you have a 60GB 5g?
12:53:10orthogonallinuxstb: yes, I do
12:53:10jsmitzquick q then I'm going to sleep: what is a wps?
12:53:13ashridahorthogonal: there's still plenty of work to do on the ipods
12:53:17orthogonalI just installed rb on it tonight
12:53:25ashridahjsmitz: WhilePlayingScreen
12:53:41orthogonalprior to this, I hadn't used rockbox for about two years, afyer my archios recorder v2 broke'
12:53:41 Quit jsmitz ("I'm out")
12:53:41linuxstbYou should at least create a custom build for yourself with the memory size set to 64MB.
12:53:44ashridahjsmitz: it's basically the status screen that shows while music is playing, and is customizable
12:54:11orthogonallinuxstb: indeed. Is there a wiki entry that explain what code changes rto make?
12:54:13linuxstbIt would be trivial to produce different Rockbox builds for the 32MB and 64MB versions, but I would prefer to just produce one.
12:54:29linuxstborthogonal: No. I think you only need to change the entry in the tools/configure script.
12:54:41linuxstbBut if you work it out, then add it to the wiki...
12:54:43_StarScreamBHSPitLappy: yeh the USBOTG is why i originally purchased the device, so i didn't have to buy lots of sd cards. Is this something that the develpers want to add when they get time?
12:55:10ashridah_StarScream: it's on the cards, like lots of things
12:55:15BHSPitLappy_StarScream: I'm neither a deveoper nor an iRiver owner :P
12:55:16orthogonallinuxstb: if it'sd trivial, we ought to do it, else peopel with 64 are going to think rb eats battery charge
12:55:24BHSPitLappydeveloper, even
12:55:39linuxstborthogonal: I would prefer not to implement temporary fixes.
12:55:50 Join Paul_The_Nerd [0] (
12:56:03orthogonalup to you.
12:56:31orthogonalI just wouldn't want to see peol;el saying rb doesn't work well if there's an easy fix
12:56:36_StarScreamashridah: ok .thank. i will check back at a later date. It looks like a great project and i'm extremely impressed that i at leat have the choice to run non proprietry firmware on the device.
12:56:52linuxstbThere are basically just two of us (myself and preglow) working on the ipod port, plus a few other people porting it to other ipods.
12:57:30orthogonalthen you're the man I want to talk to
12:57:40linuxstbNone of the other developers own ipods.
12:57:55ashridah_StarScream: some of the features rockbox provides are still worth testing now, like i say, you can use both, it just requires rebooting the unit and holding down a button to get to the iriver firmware
12:59:42orthogonallinuxstb: what needs to be done for the ipod port?
13:00:31linuxstbI guess the top priority is to finish implementing the hardware drivers - battery status is the main gap.
13:00:55orthogonalhow much of that requires reverse engineering?
13:01:00linuxstbThen it's really just bug fixes, and trying to reduce power consumption.
13:01:18 Join San [0] (
13:01:30linuxstborthogonal: None really - the source code of the ipodlinux project gives us all the information we need.
13:01:36_StarScreamashridah: ok thanks. I will do that. Also do you know if the H10's have a rockbox version available. I stupidly purchased one for my g/f only to realise it was a us one , and neither of us have a windows machine
13:01:57linuxstbAlso, the ipods have a PCF50605 power management unit, and a datasheet is available for that.
13:02:09orthogonallinuxstb: perhaps you could assign me some simple parts of it then, as I re-acclimate to rb?
13:02:11ashridahthe H10's based on the portalplayer chipset, not an motorola coldfire, so rockbox currently won't work on it
13:02:35_StarScreamashridah: yeh i figured they would be different. Thanks for your help.
13:04:29linuxstborthogonal: I can't really think of any specific jobs apart from the battery status, or making Rockbox detect and use the extra 32MB ram you have.
13:05:25orthogonalare you oe sometone else currently working on those? I don't want to duplicate or step on toes
13:05:43linuxstbNo, I'm not working on anything specific at the moment.
13:06:39orthogonaloh, here's a question: when I lasyt coded anything for rb, we were exclusvely using staticly allocated memory: no heap, no malloc. is that still true?
13:06:58linuxstbpreglow has implemented a CPU scaling patch which works fine on the 5g and Nano, but crashes my 4g ipod. I want to try and get that working.
13:07:13linuxstbBTW, I have a 30GB 5g and a 60GB 4G Color. preglow has a Nano.
13:07:28orthogonalI have th e60gb 5g
13:07:54linuxstbRockbox is basically just doing static allocation, but there is some allocation at startup from the main audio buffer.
13:09:27linuxstbAs you may know, Rockbox places itself at the start of RAM, and the plugin buffer (and now the codec plugin buffer) are at fixed addresses at the end of RAM.
13:09:58orthogonalanfd the dire cache too at the end, yes?
13:10:18linuxstbSo in order to support different RAM sizes detected at runtime, my idea is to put the plugin buffer and codec buffer at fixed addresses at the start of RAM, and Rockbox itself afterwards.
13:10:47linuxstbdircache is allocated from the audio buffer at startup - the size varies according to the number of files.
13:11:00orthogonalbut the plugins are at the end so they can grow down (effectively doing a malloc without teh associated bookeeping)
13:11:17linuxstbplugins can only use their fixed (512KB) buffer.
13:11:34linuxstbOr grab the entire audio buffer for their own use.
13:11:42linuxstb(but that stops playback)
13:11:43orthogonalah, right, that was it
13:12:09linuxstbPlugins are compiled and linked to run from a fixed address - they are just simple binary images.
13:14:55linuxstbamiconn: (and others) Can you think of any reason why we can't move plugins and codecs to the start of RAM for every target? It would seem messy to only do it for one build.
13:15:42amiconnMoving to the very start isn't possible, as this is where the vector tables are located
13:16:08amiconn...and leaving a hole between vector tables and code would blow up the binary
13:17:32***Saving seen data "./dancer.seen"
13:18:01linuxstbYes, I'm just realising it isn't a trivial change...
13:18:09amiconnThis is a problem on all targets, and it's a severe problem on archos
13:18:24 Join novimon [0] (
13:19:23orthogonalsomebody point me to waht I need to do to compile for the ipod?
13:19:31orthogonalfrom cygwin under windows?
13:20:00linuxstbDownload gcc-4.0.2 and binutils-2.16 (or later versions), and cross-compile them for arm-elf
13:20:18linuxstbGeneric cross-compiler instructions are here: We already have far too many ipod targets though, I would prefer not
13:20:25linuxstbOops, wrong paste....
13:20:44cooblinuxstb: does gcc-4 produce better arm code?
13:20:48coobthan 3.4
13:21:01linuxstbcoob: We had bugs on our code that disappeared when using gcc 4
13:21:13Paul_The_Nerdorthogonal: <−− Has instructions for a Cygwin arm GCC package.
13:21:18coobnotici was thinking more in terms of optimisation
13:21:28Paul_The_NerdSo you don't need to compile your own cross compiler under cygwin any more.
13:21:54linuxstbWell, we haven't really compared - working code was more important.
13:35:22Mikachuthat page says 4.0.x is not recommended in shiny orange text, and then you recommend it further down for ipods...
13:36:11linuxstbYes, it could do with being updated for the ipod.
13:36:37linuxstbCurrently, the recommended versions are the latest 3.3.x for SH, 3.4.x for Coldfire and 4.0.x for ARM.
13:41:45orthogonalonce I have the toolchainn, is there a .configure? There was targeting script if I recall?
13:42:21Paul_The_Nerd /tools/configure, yes.
13:43:03linuxstbmkdir build ; cd build ; ../tools/configure ; make ; make zip
13:44:19Paul_The_NerdCan configure be passed parameters, instead of having to manually input the target when it prompts for them?
13:45:00Mikachuecho 14\n\n|configure should work
13:45:20Mikachupossibly '' around \n
13:45:58linuxstbI didn't think echo supported \n - you could use printf instead.
13:46:18Mikachudepends on the shell i presume
13:47:11Paul_The_Nerdecho 14|../tools/configure seems to do the trick.
13:47:22Paul_The_NerdWhich is odd.
13:47:24linuxstbAh, you need the -e option to echo.
13:47:57orthogonalarrgh cygwin is bringing in all these weird packages
13:48:10orthogonalI alsways get ALL available packages fopr cygwin
13:48:36orthogonalit's helpful
13:48:49orthogonalone time I was on ethe phone on a job inteviiew
13:48:55Mikachui remember when i tried cygwin, it would redownload everything if you changed mirrors, even though the files are the same
13:49:07orthogonalguy akss if I known subversion arther than cvs
13:49:17orthogonalI go into cygwin, typwe svn, and there ir is
13:49:44orthogonalincidebntly, svn kicks cvs's ass
13:50:02Mikachuare you holding your fingers orthogonally to the keyboard?
13:50:24orthogonalMikachu: I have a disability, and I have to type with a stick I hold in my teeth
13:50:34Mikachuthat must suck
13:50:52orthogonalha ha
13:53:44Mikachuyou weren't joking?
13:55:56orthogonaldon't worry about it dude
13:57:19orthogonalnom, dude, I'm just fucking with you.
13:57:28orthogonalI'm just a terrible typist
13:57:51orthogonalI can't touvh typwe
14:01:36 Quit Rob2222_ ()
14:03:45Mikachuhm, my binaries got named sh-elf-* instead of arm-elf-*
14:03:52Mikachui'm stupid :)
14:04:29Mikachu"all instructions from here on are for sh-1"
14:07:00linuxstbMikachu: Can you remind me what patches you've submitted for the iPod?
14:07:13Mikachuisn't that what the tracker is for?
14:07:17linuxstbOK :)
14:07:23Mikachunot trying to be rude but i don't remember offhand
14:07:24 Quit orthogonal ("Chatzilla 0.9.70 [Firefox]")
14:09:36 Join gtkspert_ [0] (
14:10:21 Part IcyStorM
14:11:57 Quit ashridah ("Leaving")
14:12:23 Join orthogonal [0] (
14:13:10orthogonallinuxstb: what type of build, normal or bootloader?
14:13:46linuxstbUnless you want to build the bootloader...
14:13:47Mikachulinuxstb: and i think i have one bug open too
14:14:14Mikachuif you press MENU in the now playing context's View Playlist menu, you have to reboot
14:15:32linuxstbYes, I know about that. We still need to fix the button mappings in lots of places.
14:15:44Mikachuthat's sort of the only really annoying one :)
14:16:29linuxstbHave you tried preglow's CPU speed patch?
14:16:51orthogonalwhat does it do?
14:16:56 Join RoC_MM [0] (
14:17:05linuxstbDynamically changes the CPU speed depending on the load.
14:17:09freqmodwhere do i get the path?
14:17:12linuxstbSort of...
14:19:12 Quit tvelocity ("Ex-Chat")
14:20:53*JdGordon is bored and needs something to do...
14:20:58JdGordonany ideas?
14:20:59freqmodlinuxstb: i have added a fix to the text viewer on ipod ( )
14:21:29Mikachuyou can trim a patch
14:21:59linuxstbJdGordon: I have a pacman emulator I want to port to Rockbox...
14:22:00Nico_PJdGordon: maybe fox your color picker patch ?
14:22:07Nico_Pfox = fix...
14:22:16JdGordonNico_P: ye i prob should do that :p;
14:22:21Mikachuis it broken?
14:22:35orthogonala color picker???
14:22:38Nico_Pi think it's still malformed
14:22:49Mikachuah right, maybe i had to edit it a bit before i applied it
14:22:57 Join slimx [0] (
14:23:14Nico_Pwell if your edited version works, could you send it ?
14:23:35Nico_PIIRC i rtied to edit it but i must have made a mistake because it didn't build
14:24:20Nico_PMikachu: could you re-send ? my client has ignored the file...
14:24:25 Join tvelocity [0] (
14:24:33*Mikachu sighs :)
14:24:50Nico_Psorry... default setting :(
14:28:04 Quit aliask ("Chatzilla 0.9.70 [Firefox]")
14:31:03orthogonaljesus, this build is taking forever
14:31:31linuxstbYes, cygwin is slow, and Rockbox has grown very big.
14:31:37 Quit Xerion (Read error: 104 (Connection reset by peer))
14:31:54 Join Xerion [0] (
14:33:18amiconnorthogonal: make -j should speed it up somewhat
14:34:01orthogonalfhag's ghag ry?
14:34:36orthogonalwhat's that do?
14:35:40orthogonallinuxstb: where do I tell it I have 64MB memory?
14:36:28linuxstbTry changing tools/configure
14:37:16amiconnHmm, now I have a problem :/
14:37:22JdGordonlinuxstb: apart from the long lines and the translation, was there anything else that was needed to be fixed int he color patch?
14:37:45amiconnI have 2 functions that are *only* existing for the simulator, and I need to put them in the plugin api
14:38:09amiconnbut I can't include simulator .h files from apps/
14:38:24linuxstbYou can't just make them inline?
14:38:28Nico_PJdGordon: what about a few color presets ?
14:38:33amiconnlinuxstb: ??
14:38:50JdGordonNico_P: i thought about adding a preset text file and have a menu... saves extra code
14:38:54amiconnThey are simulator code, and need to be in the api
14:39:02midkayloadable preset 'themes' sound cool.. that plugins can access them.
14:39:39linuxstbThen I guess the code itself needs to be in either firmware or apps.
14:39:58amiconnThe code needs to be in uisimulator/*
14:40:11amiconnEither common or implementation specific
14:40:58Nico_PJdGordon: yea, that would be nice.
14:41:04amiconnThe code needs access to the lcd simulation (functions and data)
14:41:12Nico_Pand that way we can customise the text file
14:41:16Paul_The_Nerdmidkay: Loadable preset themes would just be config files stripped down to just have the colors.
14:41:21*amiconn is working on grayscale lib support for the sims
14:41:54amiconnI can just declare them in plugin.c (extern)
14:41:55orthogonallinuxstb: line 757
14:41:59amiconnDirty, but should work
14:42:00Paul_The_NerdI mean, you could have a folder called colors and under it, a foreground and background folder, and under it, several .cfgs that just have color entries, and not need to add any code at all to what's already there.
14:42:36Paul_The_NerdYou could probably even put it under the /themes/ folder, so that you can get to it with browse themes.
14:42:47Paul_The_NerdI don't know if browse themes will see subdirectories.
14:42:51midkayPaul_The_Nerd, right - that'd also be cool in that it could tie into other themes related to font, WPS, background, etc
14:43:07JdGordonPaul_The_Nerd: ye, but thats a hassle.. and not the nicest way to do it
14:43:15 Join spaceinvedersdev [0] (
14:43:30*Paul_The_Nerd shrugs.
14:43:46Paul_The_NerdIt seems like something you only really need to set up once, and then you're done, and it doesn't add any code size at all.
14:43:47spaceinvedersdevi've got a question - do how does using b&w bitmaps work?
14:44:15Paul_The_NerdYeah, I was gonna ask, what do you mean?
14:44:17spaceinvedersdevsorry - in plugins
14:44:26midkayyou need to use bmp2rb
14:44:45midkaywhich in turn requires a development environment
14:44:48linuxstbmidkay: No, the build system will do that for you.
14:44:48midkay(cygwin or linux)
14:44:54spaceinvedersdevon colour targets are bitmaps in /apps/plugins/bitmaps/native included?
14:44:59midkaylinuxstb, either way you need a dev environment
14:45:01 Quit RoC_MM ("Leaving")
14:45:21midkaylinuxstb, and that's if they're not contained within the plugin itself
14:45:23linuxstbmidkay: I sort of assumed spaceinvedersdev is developing a plugin, so he must have a dev environment
14:45:27 Quit tvelocity (Remote closed the connection)
14:45:28spaceinvedersdevi know what i'm doing with building and stuff - it's just i've got a b&w game and the bitmaps are 24 bit which is silly
14:45:33midkayhaha, ah, yeah :)
14:45:44linuxstbmidkay: The point is that they shouldn't be hard-coded in the plugin .c file any more.
14:46:04midkaylinuxstb, alright
14:46:10spaceinvedersdevso how do I get b&w bitmaps so I can also support b&w lcds
14:46:16linuxstbspaceinvedersdev: There are two types of bitmaps in Rockbox - mono and native. That determines which directory under plugins/bitmaps/ they go in.
14:46:22 Quit muesli__ (Read error: 110 (Connection timed out))
14:47:01spaceinvedersdevso I just save the pics in mono - but are they still included in colour builds?
14:47:08linuxstbmono bitmaps are always 1bpp, and are drawn (I think) using the current fg colour.
14:47:28linuxstbnative bitmaps are the native depth of the LCD - which can also be 1bpp on some LCDs.
14:47:47amiconnThey are drawn with the current drawmode, and the foreground and background colour depending on the drawmode
14:48:08spaceinvedersdevok thanks
14:50:24JdGordonlinuxstb: whats the diff between ID2P() and str() for the translation?
14:51:28linuxstbJdGordon: I don't know anything about the language system.
14:51:54linuxstbI just do what I see other code doing...
14:52:51spaceinvedersdevso sorry - i'm still confused; because my bitmaps are mono - they're going in the mono folder, but what do I name them at the end?
14:53:13spaceinvedersdevthe target's LCD size and bit depth or the bitmap's size and depth
14:53:27linuxstbThe convention so far is to name them based on the target lcd size.
14:54:22spaceinvedersdevand the bitdepth - so a monochrome bitmap called space_invaders_alien.200x176x16.bmp is ok
14:54:41spaceinvedersdevsorry 220 x176
14:55:04Paul_The_NerdIf that's the bitmap you want to use on 220x176x16 screens, then yeah
14:55:22linuxstbMaybe you don't need the x16 at the end - if your bitmaps will work equally well on LCDs with any depth.
14:55:39amiconnMono bitmaps should work equal on all target depths
14:56:29spaceinvedersdevi just thought bmp2rb might need that kind of stuff - it's remarkable how all I have to do is declare some const fb_data and the rest is all sorted out behind the scenes
14:56:58linuxstbNo, bmp2rb doesn't care about the filename - it just strips off everything after the first period.
14:59:37 Quit perplexity (Read error: 104 (Connection reset by peer))
15:00:35JdGordonany1 who matters: yes/no to the idea of having a text file under .rockbox/ with a list of colours for the chooser?
15:02:33 Join |Beowulf| [0] (
15:03:17linuxstbIMO it's an unneccessary complication.
15:03:52JdGordonit adds like 30lines of code... not really complicated
15:03:53safetydanJdGordon, couldn't you just use the cfg system?
15:05:04*JdGordon doesnt want it at all.. but im trying to make the wingers on MR happy :D
15:05:13linuxstbI may be wrong, but I imagine that most people won't care about setting colours manually themselves, but will just load a theme .cfg file.
15:05:28linuxstbThe real problem is a lack of themes
15:05:30safetydanJdGordon, that way lies madness
15:05:42JdGordoni know :'(
15:06:11linuxstbBut yes, a directory full of .cfg files would do the same thing with no code.
15:07:08linuxstbAlthough that would be a little wasteful on disk space...
15:07:16Paul_The_Nerdlinuxstb: People *really* want color presets. BADLY.
15:07:20Paul_The_NerdIt's strange.
15:07:38linuxstbPaul_The_Nerd: To me, "color presets" == "theme .cfg files"
15:07:47Paul_The_Nerdlinuxstb: That's *exactly* what I said.
15:08:11Paul_The_NerdI think though that a lot of the mysticriver people don't really see the flexibility config files allow.
15:08:52linuxstbIf we implemented colour presets, then it should be done properly via the language system. But that really needs langv2 so we don't waste a load of strings on colour names for targets without colour.
15:09:48JdGordonlinuxstb: thats 1 reason why i figured do it in a seperate txt file and load it if its asked for..
15:09:52spaceinvedersdevtalking of features people want - how about a make plugins: target in the root makefile that dows bmp2rb and builds the plugins?
15:09:52Paul_The_NerdApparently one of the big issues is "RGB colors are too confusing" for some of them.
15:10:12JdGordonsome ppl are stupid :p
15:10:26linuxstbHow can having a large rectangle showing you the colour be confusing? The RGB values are just an extra
15:11:04Paul_The_NerdI really don't know.
15:11:06JdGordonwhats the command to add a new file?? cvsdo something?
15:11:18safetydancvsdo add <file>
15:13:02midkaydoesn't optimization feel great?.. i just reduced 137 lines of code to under 20.
15:13:12JdGordon$ cvsdo add apps/gui/color_picker.c
15:13:12JdGordonUse of uninitialized value in concatenation (.) or string at /usr/bin/cvsdo line
15:13:12JdGordon 191, <ENTRIES> line 29. ??
15:13:51safetydannot a clue, probably a bug in the cvsdo script
15:14:57orthogonalhow so midkay ?
15:15:11JdGordonok, well im done... again... no lines longer than 85 chars (cant help a few), color not colour, translation done, compiles on b+w targets, no warnings on colour targets.. what have i forgotten?
15:15:21JdGordonapart from the new file not adding to the patch for whatevr reason...
15:15:25midkayorthogonal, my messy binary clock that relied on a series of if/elses i just switched to two for-loops :)
15:15:35safetydanJdGordon, are you doing cvs diff -uN ?
15:15:45safetydanyou need the N otherwise new files don't show
15:15:54 Join Bagder [0] (
15:16:39orthogonalmidkay: good work
15:16:40JdGordondone done done done done
15:16:57midkayorthogonal, :D thx.. unsure of why i didn't think of it before, haha.
15:17:04midkaydidn't know so much about optimization i guess
15:17:11orthogonalindeedd ;)
15:17:35***Saving seen data "./dancer.seen"
15:17:51JdGordonand ill just test to make sure the default colours load, not black on black
15:18:03linuxstbJdGordon: I've just read the misticriver thread about presets... It seems that one person is demanding presets, and everyone else is happy with the power of the sliders.
15:18:08 Quit freqmod (Remote closed the connection)
15:18:40JdGordonye, im calling it quits.. im sick of this code.. testing then uploading :p
15:19:01JdGordonyes! it works
15:19:32JdGordonand iv got it showing RGB:rrggbb for u...
15:19:37 Join perplexity [0] (i=heh16936@
15:20:41linuxstbJdGordon: :) Can you make it "RGB: #rrggbb" - that's how I normally see rgb colours written.
15:20:51linuxstbAlso, are the sliders still wrapping?
15:20:59Paul_The_NerdYeah, the # sign is vitally important. :-P
15:21:11*Paul_The_Nerd prefers it that way too, by the way.
15:21:16JdGordon# not 0x ??
15:21:35linuxstbYes - that seems to be the convention. e.g. in css/html files
15:21:35JdGordononly if u commit it tonight :p
15:22:06linuxstbonly if you stop the sliders wrapping...
15:22:28 Quit ender` (" When I was a kid I used to pray every night for a new bicycle. Then I realised that the Lord doesn't work that way so I stol)
15:22:36JdGordon? slidders not wrapping feels odd
15:22:49 Quit damaki (Read error: 104 (Connection reset by peer))
15:22:57Paul_The_NerdIt seems really odd to go from 31 to 0, though
15:23:03linuxstbIt just makes it hard to set the minimum and maximum values.
15:23:25 Join damaki [0] (
15:23:25JdGordon?? its eaiser... u dnont have to hit right 31 times to get max red
15:23:28Nico_PJdGordon: are colors still stored as 3 settings each in cfg files ?
15:23:46linuxstbYou just hold the right button down until it reaches the end.
15:24:09Nico_Pwouldn't it be easyer to store only one hex value rrggbb ?
15:24:10linuxstbMaybe it's different on the h300 - on the iPods, it's very quick to move to the start/end of the slider.
15:25:23Paul_The_NerdMaybe do what's done with the menus? Stop at the end, and then if someone release and presses again, wrap?
15:25:39linuxstbThat sounds perfect to me.
15:25:41Nico_Pi agree
15:25:47lostlogicanyone see my software codec playback wiki page? (amiconn in particular) is it helpful? whether it is or is not, what type of additional information would be wanted in it?
15:25:52*JdGordon doesnt.. too complicated :p
15:26:24linuxstbPaul_The_Nerd: But then again, that's hard with the ipod's scrollwheel. It doesn't give press/release events.
15:26:52 Join Rob2222 [0] (
15:27:50Paul_The_NerdHow 'bout just putting an ifdef, and only not wrapping with IPOD_4G_PAD, since it probably doesn't need to wrap.
15:28:05Paul_The_NerdI mean, yeah that breaks the "identical behaviour among targets" but then again, it's also a completely different input method.
15:28:22Paul_The_NerdOr rather, all ipod pads.
15:31:47linuxstbJdGordon: I have a clean source tree waiting for your patch when it's ready....
15:32:21JdGordonalmost ready
15:32:27JdGordontranslation is wrong for some reason
15:32:42JdGordonprobably a copy/paste error..s omewhere
15:35:01JdGordoni tihnk im wearing out my usb connection
15:37:41JdGordonok, something is very wierd... the text isnt being translated correctly... the wrong text is shown.. BUT i have just double checked the code and its all fine.. it should be loading the correct text..
15:38:25Nico_Phave you changed tu lang file on your player ?
15:42:04Jungti1234help me
15:42:49Jungti1234I modified source code.
15:43:05Jungti1234And is going to test it.
15:43:24Jungti1234Then, must I compile again?
15:43:40spaceinvedersdevJungti1234: yes
15:43:51linuxstbI thought the intention was to store the native colour (i.e. the 16-bit RGB565 value) in the config block?
15:44:11JdGordonlinuxstb: no, 8bit in config and native in ram
15:44:55linuxstbI thought 8-bit in external .cfg files, but native in the config block. changin the colour in the txt file was easier..?
15:45:24JdGordoni dunno.. change it if u cbf..
15:45:57linuxstbI can tell you're bored of this patch :) I'm happy to take it over.
15:46:13JdGordonhow could u tell??
15:46:20JdGordonalso its 1.45am here so im asleep
15:46:38linuxstbThe "cbf" is a clue (if I'm guessing correctly what it means)
15:46:40midkaythat's nothing.. 6:45am here! ;)
15:47:12JdGordonbut.. isnt it a pain to save to 8bit to the cfg file?
15:47:30linuxstbNot really - we have the macros to convert between the two formats now.
15:49:05JdGordonby pain i mean haveing to manually save/load it..
15:49:31JdGordonit will free up a whopping 24bit of RAM in the config block tho :p
15:49:59JdGordonanyway.. im going to bed
15:51:16 Quit JdGordon (Read error: 104 (Connection reset by peer))
15:51:39safetydan24 bits is a lot when you only have 1472 of them, it's nearly 2%
15:53:08Paul_The_NerdActually, wouldn't it only save 8-bits? Decreasing from 24 to 16?
15:53:10linuxstbI'm changing it now.
15:53:34*Paul_The_Nerd mutters about people on misticriver not even knowing if they're running optimized or not.
15:53:42linuxstbIt's 8 bits per colour. Currently we are only storing two colours, but I guess it's possible we will store more in the future.
15:53:49Paul_The_NerdOh, right
15:53:53Paul_The_NerdSo 16
15:54:00Paul_The_NerdI forgot there was BG and FG.
15:54:20linuxstbI'm thinking that at least the status bar should have different colours.
15:54:41Jungti1234Do you know where is eli?
15:54:42linuxstbBut that's all for the future...
15:54:52Papricahe is in the army
15:55:09Jungti1234May I modify Rockword?
15:55:19Paul_The_NerdIt seems like each of the status bar elements could have a token of its own, and you could essentially have a menu WPS with the menu viewport, and the status bar tokens scattered about it.
15:55:34Papricayou want to change the design?
15:55:53linuxstbPaul_The_Nerd: Yes, on second thoughts, a status bar WPS-like string is probably the way to go.
15:56:59Jungti1234Paprica: no
15:57:15Jungti1234I want to add function.
15:57:26 Join muesli__ [0] (n=muesli_t@
15:58:29Jungti1234open, save
15:58:51Papricagood luck ;]
15:59:18Jungti1234I don't know that I can do it. hehe
15:59:32Jungti1234Is going to try.
15:59:48 Quit Maxime (Read error: 110 (Connection timed out))
16:04:54spaceinvedersdevhmm after trying to use mono bitmaps and clean up code a bit space_invaders compiles without warning, but crashes both the win32 and sdl sims (from cygwin)
16:05:44safetydangdb back trace?
16:05:50spaceinvedersdevi get a "do you want to send an error report"
16:06:03spaceinvedersdevi wish i knew what gdb back trace was
16:06:14safetydanrun the sim like this
16:06:18safetydangdb ./rockboxui
16:06:26safetydanthen type "run"
16:06:46spaceinvedersdevi'll try
16:06:50safetydanthen make it crash again
16:06:55safetydangdb should tell you where it crashed
16:08:01spaceinvedersdevprogram recieved signal SIGSEGU, segmentation fault in memcpy.c 94
16:08:11safetydanokay, type
16:08:34safetydanthat should give you a back trace
16:09:21linuxstbAre you using rb->memcpy() ?
16:09:58spaceinvedersdevat the top is an lcd_bitmap_part (wich I never call) but below it is an lcd_bitmap which is my menu background
16:10:06spaceinvedersdevi don't use memcpy
16:10:13 Join doctor [0] (n=doc__@
16:10:40linuxstbCan you paste the line that calls lcd_bitmap?
16:11:59spaceinvedersdevok "rb->lcd_bitmap(space_invaders_menu_bg, 0, 0, LCD_WIDTH,LCD_HEIGHT);"
16:12:49safetydanAll I can think of is that space_invaders_menu_bg isn't a valid pointer
16:12:56 Join imphasing [0] (
16:16:03 Join Bger [0] (n=Bager@
16:17:41 Quit San (Read error: 110 (Connection timed out))
16:18:14spaceinvedersdevi am on windows - i bet it's just paint buggering me about
16:18:57spaceinvedersdevwouldn't the compiler pick up an invalid pointer?
16:19:23Paul_The_NerdNot if it's valid in the sense that "It points somewhere" but invalid in the sense "it doesn't point where you expect it"
16:20:21spaceinvedersdevright - but i've got "extern const fb_data space_invaders_menu_bg[];"
16:20:41spaceinvedersdevat top level
16:20:55 Quit gtkspert_ ("leaving")
16:21:38spaceinvedersdevand it all worked when I had 24bit bitmaps - now i've saved them as monochrome in /apps/plugins/bitmaps/mono.............
16:21:44spaceinvedersdevi've got an idea
16:22:03spaceinvedersdevi've got both a colour and a mono one in the same source tree
16:23:23spaceinvedersdevthat'll be it - so the extern const fb_data gets confused aecause there are 2 space_invaders_menu_bg's
16:23:24Paul_The_Nerdspaceinvadersdev: Are you sure you've got the line right in the makefile for the bitmaps?
16:23:44Paul_The_NerdErr, sources file
16:23:51 Quit SereR0kR ("XChat Aqua")
16:24:14 Join San [0] (
16:24:17 Join MarcoPolo [0] (
16:25:26spaceinvedersdevi've included both 1bit and 24bit bitmaps with the same name in the sources files - i've removed the 24bit ones and am recompiling
16:26:52Paul_The_NerdAh, you should #ifdef around various ones to make sure the right ones are included for the right targets
16:26:55Paul_The_NerdYou can't just lump them all in.
16:27:19 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
16:27:41Nico_P[15:53:41] * Paul_The_Nerd mutters about people on misticriver not even knowing if they're running optimized or not.
16:27:51Nico_PPaul_The_Nerd; are you llorean on MR ?
16:28:39Nico_Pjust read the thread you were talking about ;)
16:28:53Paul_The_NerdI just wanted to help, but it's really difficult if I can't even identify the problem.
16:29:15Nico_Pyea it's frustrating
16:29:18spaceinvedersdevit still crashes
16:30:26Paul_The_Nerdspaceinvadersdev: Want to paste that portion of your code, and maybe the SOURCES file at pastebin?
16:33:31spaceinvedersdevi've not changed my code since last time, and last time it worked - i'm just presuming windows is getting the bitmaps wrong
16:34:32linuxstbspaceinvedersdev: If your bitmap is in the apps/plugins/bitmaps/mono directory, you have to display it using rb->lcd_mono_bitmap()
16:34:59spaceinvedersdevOMg thanx linuxstb
16:35:06linuxstbThe lcd_bitmap() function is for native bitmaps.
16:35:06 Join mikearthur [0] (
16:35:38 Join SereR0kR [0] (
16:35:58 Join eli_sherer [0] (
16:36:10spaceinvedersdevi'm sooooo stoopid
16:36:40 Quit XavierGr ("Trillian (")
16:36:41 Quit doctor (Read error: 110 (Connection timed out))
16:41:31Nico_Plinuxstb: have you seen the latest album art update ?
16:41:59Nico_Pand would you have advice on how to put the bitmaps in the audio buffer, as you suggested ?
16:44:33linuxstbNo, I haven't had chance to look at it. I don't think it's easy to put the bitmaps in the audio buffer - I was hoping that someone else who knows the audio buffering code could help with that. Maybe lostlogic?
16:45:16Nico_Pand is the potential gain worth it ?
16:45:33linuxstbEspecially for people who don't use album art.
16:45:40Nico_Pwhat improvement would we see ? memory usage ? speed ? both ?
16:45:47linuxstbJust memory usage.
16:46:20linuxstbIt would also allow you to increase the limit on size - so a user could have full-screen album art if he/she wanted to.
16:47:01Nico_Pcurrently i've set the buffer size to 15250, to allow 125x125 max
16:48:13MarcoPolois album art support in cvs ?
16:48:26Nico_PMarcoPolo: no, it's only in a patch
16:48:31MarcoPoloNico_P: ok
16:48:43Nico_Pand also in needleboy's h300 optimised build
16:48:50MarcoPoloI tried rockbox on my iPod the whole week and it works rather well
16:49:16MarcoPoloit only isn't gapless on 500kbps OGG ;-)
16:49:36MarcoPolothe only problem is I can't load the battery AND play music
16:49:54Paul_The_NerdHold Menu while plugging in USB
16:50:06Paul_The_NerdIt'll charge instead of rebooting.
16:54:41 Quit spaceinvedersdev ("CGI:IRC")
16:57:22safetydanI'm really beginning to think we need to get the "optimised" build renamed.
16:57:56safetydanOr at least, a much stronger indication that it is potentially more unstable than builds
16:57:56Paul_The_NerdAnd like, big red bold letters that say "This is an unofficial build. Performance is not indicative of actual Rockbox performance."
16:59:10Paul_The_NerdAnd especially "Before reporting a bug, try a clean non-optimized build, and verify that this bug occurs in it."
17:01:36linuxstbHaven't we all said that a hundred times in the past?
17:01:57linuxstbBut that's the nature of the GPL - we have given away control.
17:02:09Paul_The_NerdBut has anyone asked needleboy to put a clearer disclaimer?
17:02:14Mikachudoesn't it require forks to change the name?
17:02:29safetydanDoesn't mean you can't try and get people to understand that it's not optimised as such, just more unstable
17:02:38Mikachuat least you didn't pick bsd license :)
17:03:20linuxstbsafetydan: I agree. My personal perference would be "experimental". If it's taken in that light, then I think unofficial builds are a good thing for testing and developing new patches
17:03:45Paul_The_Nerdlinuxstb: I was just going to suggest "experimental." My first thought was "unstable" but that sounds far too negative really.
17:04:41safetydanlinuxstb, experimental is probably the best
17:04:47safetydanit is a good testing ground for patches
17:06:13Paul_The_NerdI think it may be far too late to get the name changed though.
17:07:42linuxstbThe main problem is that none of the people with cvs write access have taken an interest in the UI side of Rockbox - which is the main difference in the optimised build, and the main complaint of MR users.
17:08:53Paul_The_NerdI dunno, other than JDGordon's patch, Optimized has what, scroll margins and album art?
17:09:22linuxstbThere are probably WPS-related patches there as well. The difference is smaller now than it was though.
17:09:34MarcoPolowhere is this experimental build ?
17:09:47Paul_The_NerdMarcoPolo: It's only for h300s
17:09:51 Join jlo [0] (
17:09:55MarcoPoloPaul_The_Nerd: doh :(
17:11:38 Join jlo_ [0] (
17:13:37 Join jlo__ [0] (
17:14:28 Quit jlo__ (Read error: 104 (Connection reset by peer))
17:16:26 Quit Moos ("Glory to Rockbox!")
17:16:44 Join jlo__ [0] (
17:17:39***Saving seen data "./dancer.seen"
17:18:56jlo__preglow : have you seen the lmods I did to fit your crossfeed needs ?
17:19:36 Quit aegray (Connection reset by peer)
17:21:44 Join aegray [0] (
17:22:31 Join jlo___ [0] (
17:22:54jlo___I get disconnected every minute !
17:25:41 Join Maxime [0] (
17:28:23 Quit jlo (Read error: 110 (Connection timed out))
17:28:29 Nick jlo___ is now known as jlo (
17:28:52 Join ts-x [0] (
17:29:01 Quit jlo ("ChatZilla 0.9.68 [Firefox 1.0.7/20050919]")
17:31:02 Quit jlo_ (Read error: 110 (Connection timed out))
17:37:32 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
17:37:42 Join jlo [0] (
17:39:35 Quit jlo__ (Read error: 110 (Connection timed out))
17:39:42 Join ender` [0] (i=ychat@
17:40:15 Join DT291 [0] (
17:41:56 Quit Matze ("Miranda IM! Smaller, Faster, Easier.")
17:47:12Jungti1234bye all
17:47:28 Quit Jungti1234 ()
17:48:04 Join RotAtoR [0] (
17:48:17 Join Hansmaulwurf [0] (
17:50:31 Quit jaebird ("Leaving")
17:52:03 Quit DreamTactix291 (Read error: 110 (Connection timed out))
17:57:17 Quit jlo ("ChatZilla 0.9.68 [Firefox 1.0.7/20050919]")
18:00:43linuxstbDoes anyone know the difference between str(), STR() and ID2P() in the language system?
18:11:02 Join Jungti1234 [0] (n=jungti12@
18:11:07*Bger tries to find their defines ...
18:11:34Jungti1234I have idea
18:11:37Mikachuyou guys use ctags don't you?
18:11:58Bgerapps/talk.h:55:#define STR(id) ID2P(id), id
18:12:45Jungti1234Can't I control size of font in WPS?
18:12:46pyroWhat does MMC stand for? ifdefs all over with MMC
18:12:56Bgerapps/settings.h:181:#define ID2P(id) (VIRT_PTR + id)
18:13:06Jungti1234If it's possible, can make more various WPS.
18:13:06MikachuJungti1234: the fonts are bitmapped
18:13:42Paul_The_Nerdpyro: Just a guess, since I have no clue: Maybe the media card in Ondios?
18:13:53MikachuJungti1234: you want to use many fonts at once?
18:13:56Paul_The_NerdJungti1234: So, each size must actually be a new font file, and there's no multiple-font support yet.
18:14:02pyroyeah - Mass Media Controller probably
18:14:03Jungti1234Or, specify various font in WPS
18:14:03linuxstbBger: Yes, I saw those defines, but still don't understand where I should str, where I should use STR, and where I should use ID2P.
18:14:28Jungti1234Mikachu: yes
18:14:46Paul_The_Nerdpyro: I was just thinking of the MMC flash cards.
18:14:51MikachuJungti1234: you can't
18:15:17Jungti1234Can't you make it?
18:15:22Paul_The_NerdJungti1234: There has been some discussion about this, but nobody has taken on this job yet. It requires a large amount of work.
18:15:23Jungti1234or we
18:15:54Jungti1234I want make U10 WPS
18:16:38orthogonal" Real" fonts would also use memeory. The more memory we use, the less room we have for teh mp3 buffer. The less buffer, the more we spin the disk. Teh more we spin teh disk, the less battery time we have.
18:17:03orthogonaland ultimattely, we want long battery time
18:17:30orthogonalJungti1234: there's only one font on that page
18:17:47Mikachuhm, the clock in the statusbar, how is that handled?
18:18:04orthogonalrest of the crap is graphics
18:18:06 Join SereRokR [0] (
18:18:24Mikachuyeah but i have the statusbar on and it shows the time with a smaller font
18:18:26linuxstbMikachu: That uses the built-in system font IIUC.
18:18:53Jungti1234orthogonal: So, I want to use various font.
18:19:47orthogonalJungti1234: can't
18:20:02linuxstbJungti1234: I don't think anyone objects to the idea, it's just that it hasn't been implemented yet.
18:20:04Paul_The_NerdJungti1234: Well, if you feel like creating the patch, I'm sure many other people would be happy.
18:20:22linuxstbIt's on that 5000-item Rockbox to-do list.
18:20:36orthogonalYeah, but it would take up memory
18:20:57Jungti1234Font buffer is problem yet.
18:21:25linuxstborthogonal: Yes, but not very much, and I think it's a worthwhile use of memory.
18:21:31Bgerlinuxstb where is the str() located?
18:21:33Paul_The_Nerdorthogonal: Yeah, but I have a flash based device, so I really don't care how much of *my* memory gets eaten up. :-P
18:22:00orthogonalPaul_The_Nerd: yes, i suppose you don't
18:22:50orthogonalBut teh additional font would only be used in teh wps?
18:22:52 Quit SereRokR (Client Quit)
18:23:21Paul_The_Nerdorthogonal: I think the idea is to have separate fonts between WPS, the menuing system, and the file browser. And within the WPS, the option of multiple fonts.
18:23:25orthogonalI mean, teh menus are only going to ever use one font, yes?
18:23:34 Join SereRokR [0] (
18:23:42linuxstbBger: I'm mainly confused by settings_menu.c - if you search in there, you can see str(), STR() and ID2P() being used. I can't find the definition of str itself.
18:23:42orthogonalThen add'l fonts should be embdeed in the wps
18:24:21Paul_The_NerdHow do you mean "embedded in the wps?"
18:24:47orthogonalwell, make it so that a wps could reserve more memory, and read teh font info with teh wps
18:25:12Bgerlinuxstb same with me ...
18:25:14orthogonalI mean, if I have a wos that only uses one fiont, I don't want to pay for reading 20 other fonts into memeory, yes?
18:25:16linuxstbWe have to have a fixed amount of memory for the wps - where would we get extra from?
18:25:27orthogonallinuxstb: from the top of memory
18:25:38linuxstbBut that is in use by the audio buffer
18:25:52orthogonalAt one point I looked at doing something like this: dyanamiclaly moving teh end of teh mpeg buffer
18:26:20Paul_The_NerdIt seems there should be a finite number of fonts for "dynamic" text in the WPS. Like, text using id3 info and stuff.
18:26:28orthogonalthere are several places whweree you'd change a constant to a function call
18:27:00Paul_The_Nerd"static" text, anything the user typed into the .wps file, can essentially be prerendered so the font only needs to be loaded on WPS loading, right. Then the necessary glyphs are stored somewhere, and the rest of the font doesn't need to be kept around.
18:27:41orthogonalPaul_The_Nerd: except that that text may move, as a result of being centered, scrolled, yes
18:27:56Paul_The_NerdBut I'm not talking about prerendering the screen image.
18:28:01orthogonalbut each static part of aline could be pre-rendere
18:28:06Paul_The_NerdJust only loading a subsection of fonts that don't need to be dynamic
18:28:07 Quit Jungti1234 ()
18:28:07linuxstborthogonal: Rockbox has also had unicode support added recently - which complicates the font handling by introducing a font cache.
18:28:21orthogonalwhat's in teh font cache?
18:28:24linuxstb(for fonts that contain a very large number of glyphs)
18:28:37orthogonalfor filthy furrinwers
18:28:58linuxstbYes. 26 characters should be enough for everyone...
18:29:12orthogonaland down south, ju' 19 or so
18:29:16Paul_The_NerdMan, two characters should be enough for everyone.
18:29:25 Quit SereR0kR (Success)
18:29:37Paul_The_NerdIf you can't just interpret long strings of binary into the appropriate ascii, ansi, or otherwise, I don't know what you're doing with a DAP, y'know?
18:30:17Mikachumaybe instead of rendering the fonts it could just echo the binary with the piezoclicker
18:30:41 Quit imphasing ("Lost terminal")
18:31:00Paul_The_NerdOr flash the HD Led or virtual HD led for those that don't have an iPod.
18:31:23orthogonalI dunno about a FONT CACHE
18:31:27orthogonali MEAN ONMAN IPOD, SURE
18:31:55Mikachuit seems you accidentally enabled caps lock
18:31:59orthogonalfor mpegs
18:32:05linuxstbOr you are very angry...
18:32:06orthogonalyeah, I have a hard time typing
18:32:16Paul_The_NerdOrthogonal: Well, there's the fact that a lot of the DAPs now supported have 32+ mb of ram.
18:32:19Paul_The_NerdOr, 16+
18:32:28*Paul_The_Nerd forgot about that one series.
18:32:36orthogonalbut the whole ptoject started to support the archos
18:32:41linuxstbThe purpose of the font cache is to reduce the amount of memory used by fonts...
18:32:46orthogonalit's kinda cruel to abandon them
18:33:42orthogonallinuxstb: oh, the build I did with 64? it can't find the codecs. Do I need to also compile the codeecs for 64?
18:33:48Paul_The_Nerdorthogonal: So, is it "don't add features that the archos can't support" or is it "force the archos to have less RAM available for mp3s by adding features that decrease it" or is it "add features, and only enable them in 8+ or 16+ mb targets" in your mind?
18:34:13linuxstborthogonal: You need to compile everything, do make zip and then unzip that to your ipod. But there may well be other places in Rockbox where 32MB is hard-coded.
18:34:18orthogonalPaul_The_Nerd: the later. alrthough I'd make iot more granular than that.
18:34:36orthogonallinuxstb: erg. Thanks
18:34:38linuxstbapps/plugins/ is one place to look.
18:34:53orthogonalthe Mormons???
18:35:40orthogonalPaul_The_Nerd: I guess in my haert I'm still thinking like a recorder owner
18:36:24orthogonallinuxstb: make zip?
18:36:25Paul_The_Nerdorthogonal: I'm all for any features that can run on Archos' to be supported in them, but at the same time you can't let that chain down the whole thing. 's all balance, 'eh?
18:36:56orthogonalI mean, I have a 64mb ipod
18:37:09orthogonalso eventually I'm gonna be all about features ;)
18:37:28Paul_The_NerdI have an H120, a missing AJBR (I suspect my brother has it somewhere) and an iPod Nano.
18:37:58orthogonaltell me about the h120
18:37:59linuxstborthogonal: Yes, make zip will create a file. This contains everything you need to install on your ipod - just unzip it to the root directory.
18:38:05 Join ender [0] (i=ychat@
18:38:12Paul_The_NerdThe H120 is my very favorite player.
18:38:19linuxstbi.e. fonts, plugins, codecs, themes...
18:38:26 Quit ender` (Nick collision from services.)
18:38:48 Nick ender is now known as ender` (i=ychat@
18:39:06orthogonalPaul_The_Nerd: specs?
18:39:09Paul_The_NerdIt has a nice, clean grayscale screen. It has an LCD remote, for easy use. It has spdif in and out, just 'cuz. It has FM Radio. And it's a swcodec platform. It's great for pure audio purposes, and has no visual flash.
18:39:45Paul_The_NerdThe H120 is 20 gigs.
18:39:50Paul_The_NerdPhysically it's fairly hefty.
18:40:02Paul_The_NerdI mean, not "big" but not small like iPods
18:40:02orthogonalah. My recorder was 20, until I toree out the drive and put in a 60
18:40:06linuxstbIMO, the only flaw in the h1x0/h3x0 is that the DAC can only work at 44.1KHz.
18:40:22Paul_The_NerdYeah, this one uses 1.8" drives, and can only fit a single platter, so I'm waiting for a 40 at least to be available.
18:40:37orthogonalPaul_The_Nerd: until I got the ipod, I'd have agreed with you. I have to admit I likes me a color screen
18:40:44BgerPaul_The_Nerd afaik it already is avail
18:41:16linuxstbBger: If you want to buy one, or want to buy a million?
18:41:26Paul_The_NerdBger: I couldn't find anywhere to purchase a 40gb single platter last time I looked.
18:41:30orthogonalcrap. he just tore a turtle into pieces and roasted it on a stick
18:41:40linuxstbYour ipod?
18:41:46Paul_The_Nerdorthogonal: My nano has a color screen, and besides the games in rockbox, I could care less. My WPS is black text on a white background.
18:42:02orthogonallinuxstb: crocodile hunter imittaion guy
18:42:17orthogonalPaul_The_Nerd: oh, screw teh games. I likes me teh videos
18:42:28 Join imphasing [0] (
18:42:30orthogonallinuxstb: no, the guy who has to survive in the wild for aweek'
18:42:36Paul_The_NerdSee, I don't want videos on something that small
18:42:42orthogonalSo I thought too
18:42:55*linuxstb doesn't have a clue what orthogonal is talking about....
18:42:59Paul_The_NerdIf I wanted video on a portable, I'd get a PSP so I had a proper screen, and a decent size memory stick.
18:43:00orthogonalBut it's great for webcasts of classes
18:43:14orthogonallinuxstb: on the science channel on tv
18:43:26orthogonalgood point
18:43:44 Quit paugh ("Leaving")
18:44:50orthogonallinuxstb: make zip isn't working fo rme
18:45:16linuxstbWhat does it say?
18:46:18 Join TCK [0] (
18:47:44orthogonallemme pm it to you?
18:48:02linuxstbYou need to be registered though.
18:48:38Mikachuyou can hax nickserv to allow msgs from unregged users
18:51:50lostlogics/hax/use a standard command as designed in/
18:51:56Mikachuhax is shorter
18:52:13Paul_The_NerdGah. I had a question for you lostlogic, and now I've forgotten it entirely.
18:52:40BgerPaul_The_Nerd, ?
18:53:09Paul_The_NerdBger: Wow, thanks.
18:54:14Bgerjust search for MK4006GAL or MK4007GAL (the newer version with perpendicular recording)
18:54:16linuxstblostlogic: What do you think about storing album art images in the audio buffer, just before the track itself? Will it complicate the playback system at all?
18:54:52lostlogiclinuxstb: we already store codecs there, if I understand it correctly, so it shouldn't be a problem
18:55:19 Join virtualball2 [0] (
18:55:27 Quit Bger (Read error: 104 (Connection reset by peer))
18:55:50virtualball2hey, whats the command to compile sources ?
18:56:26virtualball2ya i know make but whenever i type make, it gives me an error with
18:57:02Mikachuare you running it in your build dir where you ran configure?
18:57:04 Quit Rob2222 ()
18:57:30virtualball2well im not trying to compile rockbox, im trying to compile doom
18:57:35orthogonalI have a cygwin problem
18:57:37virtualball2and im using PATH
18:57:45orthogonalI thinK i hosed it installing teh prevbult toolchain
18:58:01Mikachudoom hooks into rb's build system
18:58:16virtualball2so i put it into the build directory?
18:58:35Mikachuyou want doom in apps/plugins, and apply the patch
18:58:55virtualball2ok thanks! then i just compile rockbox normally?
18:59:00Mikachushould work
18:59:08 Quit orthogonal ("Chatzilla 0.9.70 [Firefox]")
19:00:10linuxstbvirtualball2: Are you compiling for an ipod?
19:00:35virtualball2cuz doom now orks for the iPod
19:00:57linuxstbI think you will also need to increase the size of the PLUGIN_BUFFER_SIZE define in the config file for your iPod - firmware/export/config-ipod????.h
19:01:10linuxstbChange from 0x80000 to maybe 0xf0000
19:01:12Paul_The_Nerd0xA0000 is a good value for it.
19:01:19Paul_The_Nerdlinuxstb: Doesn't need to be NEAR that big.
19:01:22linuxstbIf 0xA0000 works, use that.
19:01:30linuxstbPaul_The_Nerd: I was just guessing....
19:01:41virtualball2wow you guys are really nice about these things :-P...if only the apple os loaded quicker, i wouldnt really need iPL lol
19:02:22Mikachuwhat do you do with ipl?
19:02:24Mikachuexcept play doom
19:02:33 Quit Nico_P ()
19:02:37virtualball2well they have more games and apps
19:02:51virtualball2\like a periodic table (which is very useful lol)
19:02:55 Part _StarScream ("Konversation terminated!")
19:03:03Mikachuyou should port all those useful apps to rb
19:03:17linuxstbvirtualball2: That can be your contribution back to Rockbox - port the IPL periodic table to Rockbox :)
19:05:16virtualball2i can try lol
19:05:40linuxstbIf you can compile Rockbox yourself, then you are half way there.
19:05:41virtualball2is it really hard too, cuz i know basic C and a little more
19:05:48virtualball2true lol
19:07:19 Quit TCK (Read error: 110 (Connection timed out))
19:07:38virtualball2so how hard would it be to port it? Are the wheel commands the same?
19:08:38linuxstbHow does the IPL periodic table use the wheel?
19:09:12linuxstbIn Rockbox, you call the "get_button()" function, and that returns things like BUTTON_MENU, BUTTON_PLAY or BUTTON_SCROLL_FWD
19:09:23virtualball2Thats like iPL
19:09:35virtualball2how does scrolling the wheel work?
19:09:54linuxstbAt the moment, it just sends button events - BUTTON_SCROLL_FWD and BUTTON_SCROLL_BACK.
19:10:03virtualball2oh ok
19:10:28linuxstbBut there is also a patch (not in CVS) that gives you the location on the wheel the user is currently touching.
19:10:53linuxstbBut Rockbox (and all the plugins) work on lots of different players - not just ipods.
19:11:40 Join Gibbed [0] (
19:11:42 Quit Rick (Nick collision from services.)
19:11:54 Nick Gibbed is now known as Rick (
19:12:13virtualball2ya that kinda bugs me, not that it works on other players but like how a game is up in the corner and like the calculator is in the lower left corner
19:12:57linuxstbThat's because the original plugin was designed to run on one specific player. We don't want to make the same mistake with new plugins.
19:13:12 Quit mikearthur (Remote closed the connection)
19:13:16Mikachumost of the buttons in calc don't do anything anyway
19:13:37linuxstbWe also need people to adjust the old plugins to work on all targets - it done for some of them, but there are still lots left to do.
19:16:45 Join Bger [0] (n=Bager@
19:16:51virtualball2i personally think that bejewled and bickMania are one of the best looking, if not THE best looking games on the iPod...eveer lol
19:16:52Papricamm is there some option to add another calendar to the cvs(color one)? (there is one for the archos...)
19:17:17Papricavirtualball2, thanks ;]
19:17:39MikachuPaprica: do you have an ipod?
19:17:41***Saving seen data "./dancer.seen"
19:17:43linuxstbvirtualball2: Doom isn't bad either....
19:17:49Papricano, h300
19:18:01Mikachui added a patch to the tracker for absolute positioning on the ipod
19:18:07virtualball2i didnt get it to work yet lol
19:18:21Papricayeah, i saw
19:18:50Papricai leave it to linuxstb or somthing else with ipod
19:18:54Mikachuthe login failed popup on the tracker is awesome
19:19:17 Join yeahx [0] (
19:20:13virtualball2Ok im compiling rockbox now, where do i find the .rock? or is that seperate from the built rb?
19:21:15Papricavirtualball2, are you compuling the simulator?
19:21:17linuxstbvirtualball2: You should do "make zip" and then unzip everything in the resulting to your ipod.
19:22:01virtualball2oh this is a lot more different the iP then i thaught lol
19:22:05linuxstbYou have to run the Doom .rock with the rockbox.ipod you are compiling now.
19:22:28linuxstbYes - you should think of Rockbox as a single entity.
19:22:31virtualball2ya i know
19:22:33Papricalinuxstb, what do you think on separate "browse plugins" to categories
19:22:55linuxstbPaprica: I think it's needed, but the problem is translating the categories to different languages.
19:23:08linuxstbSo the obvious solution of using subdirectories won't work.
19:23:36virtualball2wpsbuild notice: No remote LCD size, skipping rockbox_default.rwps
19:23:47virtualball2^^is that normal when i type in make zip ?
19:23:49linuxstbIgnore that - it's because ipods don't have a remote.
19:24:00virtualball2ok thanks
19:24:32 Join mikearthur [0] (
19:24:51Papricawe can separate the plugins folder to "apps/games/screen savers" etc.. and give them name is the .lang file
19:25:15safetydanlinuxstb, you don't have to do make zip then unzip
19:25:18safetydanjust make install will do
19:25:32safetydanignore that
19:25:37safetydanthought you were talking about the sim
19:25:52*linuxstb ignores safetydan
19:25:54Papricasafetydan, you cant do make zip on the sim
19:26:21Papricayou can
19:27:13 Quit perldiver (Read error: 110 (Connection timed out))
19:27:36linuxstbPaprica: The solution would probably be to make the "browse plugins" menu option call another menu, with the translated names "Games", "Screen Savers" etc. Those sub-menus would call the file browser with the name of the plugin subdirectory.
19:28:55 Join ender1 [0] (i=ychat@
19:28:56Papricathis will be very easy to do, my problem is the makefile file
19:29:33linuxstbIf you changed Rockbox, maybe someone else would look at the makefile stuff. That's assuming no-one would object to such a patch (I wouldn't).
19:29:46 Join MarcoPolo [0] (
19:30:38Paul_The_NerdI think the plugins are definitely ready for some organizational juju.
19:30:44 Quit ender` (Nick collision from services.)
19:30:46 Nick ender1 is now known as ender` (i=ychat@
19:31:47pyroDoes anyone know how the USB charge mode works? It appears to just change the status of usb_state=USB_POWERED in the USB thread. Does hardware automatically start charging it? I can't figure out if hardware controls this somewhere
19:32:55*Bger is with 2 hands up for the Paprica's suggestion
19:33:33 Quit imphasing ("Lost terminal")
19:34:28 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
19:34:30 Join imphasing [0] (
19:35:04 Join lodesi [0] (
19:35:29Papricabger, do you have experience with makefile stuff?
19:36:08lodesii have done a 'contact' plugin
19:36:23lodesiit's there, if you want to try it:
19:36:35 Part Paul_The_Nerd
19:36:45lodesibut don't expect any eyecandy ;)
19:36:46 Join mikearthur_ [0] (
19:37:07BgerPaprica hahaha :)
19:37:11Bgerask linuxstb ...
19:37:21 Quit mikearthur (Read error: 104 (Connection reset by peer))
19:37:26Papricaits all the work..
19:37:36Bgerabout the makefiles and about my all i know about them..
19:39:09linuxstbPaprica: Thinking about the plugins - it could also be nice to find a way to translate the plugin names.
19:39:58Papricai think too
19:40:05Bgerlinuxstb i think we should wait for the famous l10n v2
19:40:39Papricait looks ugly with the small letters
19:40:51Bgerlodesi put it on the patch tracker
19:41:12linuxstblodesi: Does your plugin work for all Rockbox targets?
19:41:30lodesiBger: it's already in the patch tracker :)
19:41:39Bgerme silly
19:42:02linuxstbI did read your comments, and wondered why you have hard-coded the maximum number of contacts. You should be able to detect this at run-time based on the available memory.
19:42:05lodesiworks on all target, but the buffer size is limited to the minimum supported by the recorder
19:42:22linuxstbI haven't looked at the code itself though.
19:43:05lodesii have an array of struct, can't be defined at runtime
19:43:14lodesi(i don't use the audio buffer)
19:43:37virtualball2hmm i seems like whenever i start up the iPod it says i have an error '-1' abd boots to the Apple OS, known causes?
19:43:54 Quit mikearthur_ (Read error: 104 (Connection reset by peer))
19:44:25 Join mikearthur [0] (
19:44:44linuxstbvirtualball2: You don't have a rockbox.ipod file on your ipod...
19:45:46virtualball2ya i do
19:45:56virtualball2oh wait its in a foldr
19:46:06*virtualball2 is embaressed
19:46:27*lodesi is away: Occupé
19:47:03linuxstblodesi: You can use the function rb->plugin_get_buffer() which will return a pointer to the remaining plugin buffer, and an integer containing the size of that buffer. Just use that memory for the array of structs.
19:48:41paughdircache unworkage! i vaguely remember a reason for it not working even tho it's enabled. can anyone help me out with a likely cause/fix ?
19:48:56paughdoes it depend on another setting?
19:49:57 Join damaki_ [0] (
19:52:11virtualball2linuxstb, i built it and moved everything to the iPod and it doesnt have a doom plugin in it, i dont get what is wrong
19:52:34linuxstbDid you download the AND apply the patch file?
19:52:56mirakwhat is the correspondance between the recording gain treeshold of rockbox and iriver record mode ?
19:53:17mirakthe rockbox have a lot more grain
19:53:20mirakin gain choosing
19:53:32virtualball2ya but it wouldnt find the files so when it said which file: i just dragged the files into it
19:53:59virtualball2should i have put the configure-videoipod.h instead of h300?
19:56:19linuxstbYou needed to do three things: 1) Download and unzip it into the apps/plugins/doom/ directory; 2) Download the patch file and (in your rockbox directory) type "patch -p0 < doom.diff" (or whatever it is called); 3) Change the PLUGIN_BUFFER_SIZE in firmware/export/config-ipodvideo.h
19:58:22virtualball2ok i did that and i get one problem
19:58:39virtualball2plugin.h fails the first 5 hunks
19:59:19virtualball2is that the problem?
19:59:46linuxstbDid you download "doomnew.diff" from the top of the page, or "doomcvs.diff" from the bottom of the page? It looks like doomcvs.diff is the newest one - you should use that.
20:00:22virtualball2Oh Ok i used Doomnew.diff
20:03:30linuxstbTry doing patch -p0 -R < doomnew.diff - to reverse those changes.
20:03:57linuxstbOr it may be safer to just delete that copy of the source and start again.
20:04:29 Quit damaki (Read error: 110 (Connection timed out))
20:05:04 Join Rob2222 [0] (
20:05:53virtualball2well all the hunks succedded
20:08:22 Quit tenzing (Read error: 104 (Connection reset by peer))
20:08:59 Join tenzing [0] (
20:09:41virtualball2no it didnt work
20:09:47 Quit SereRokR ("XChat Aqua")
20:10:24 Join SereR0kR [0] (
20:13:01 Join bluey [0] (
20:15:24 Quit lodesi (Read error: 110 (Connection timed out))
20:15:45 Join lodesi [0] (
20:16:58 Join perldiver [0] (
20:17:57|Beowulf|hi - if a TSR plugin is running, does that block other plugins from loading/running?
20:19:04 Join Moos [0] (
20:21:27Bger|Beowulf| blocks - no
20:21:48Bgerbut when you start a new plugin while the TSR one is running
20:22:18Bgerrb waits the TSR plugin to exit before really running the new one
20:22:53 Quit imphasing ("Lost terminal")
20:23:27|Beowulf|ok, thanks
20:28:28 Quit slimx ("Leaving")
20:31:17 Join petur [0] (
20:31:25 Join TCK [0] (
20:32:00 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
20:34:18 Join IcyStorM [0] (
20:38:47 Quit San (Read error: 110 (Connection timed out))
20:39:28 Join Thus0 [0] (
20:49:24 Quit virtualball2 ("virtualball2 has no reason")
20:49:58lodesidone, max contact number now computed at runtime :)
20:52:11lodesiIt's limited to 106 contacts on a recorder
20:53:12 Quit Hansmaulwurf ("( :: NoNameScript 3.81 :: )")
20:53:24linuxstbMy next question is whether you can increase that maximum number, or even get rid of the limitation completely?
20:54:06Bgerlodesi 106 when using the audio buffer ?
20:54:27lodesiwell.. the size of each field(name, phone...) is hardcoded ...
20:54:39lodesino, only plugin buffer
20:54:58Bgerlodesi at least for low-mem targets use the audio buffer
20:55:24Bgeror better, use the audio buffer if there's not enough mem in the plugin buff
20:55:39Bger(the last one on all targets)
20:55:59 Join Shadowarrior13 [0] (
20:56:34linuxstblodesi: You might like to read the Rockbox style guidelines -*checkout*/docs/CONTRIBUTING
20:56:38linuxstbYou seem to break every one :)
20:56:39lodesimmh.. no problem if the audio buffer is just after the plugin buffer.. otherwise would require much more modifications
20:56:42 Quit MarcoPolo (Remote closed the connection)
20:57:01*petur thinks about porting tombo to rb... then tosses the idea away: no time :(
20:57:08lodesilinuxstb: :'(
20:57:16Bgerpetur ? tombo ? what's that ?:)
20:57:18lodesionly checked this one :
20:57:22Shadowarrior13Sounds like tomba...
20:57:25Shadowarrior13I loved that game.
20:57:41linuxstblodesi: That's the same.
20:58:12lodesilinuxstb: ok, let me reread it then :)
20:58:33peturBger: nice plain text note organizer, I use it on my PDA all the time
20:58:50linuxstblodesi: The main problems are saw in your code were: 1) Tabs; 2) C++ style comments - // ; 3) lines longer than 80 chars
20:58:50Bgerah :)
20:58:59Bgerseems very interesting
20:59:15 Join quobl__ [0] (n=quobl@tor/session/x-b81ccc6486742d0a)
20:59:29lodesii corrected the tabs in the last file
20:59:36peturBger: even if it were just to view the notes only...
20:59:51lodesii'll do the 2 others right now
21:00:17 Quit BHSPitLappy ("Dropped my laptop.")
21:00:50 Join BHSPitLappy [0] (
21:04:59*petur tries to understand the peakmeter code right now and enjoys the comments 'this is proven to be inaccurate...' aaargh
21:05:32 Quit quobl_ (Remote closed the connection)
21:06:04peturit is proven, people complain about it ;)
21:09:21peturno really, the clip detection is awfull, it hardly detects clipping at all
21:09:42linuxstbAre you talking about playback or recording?
21:10:00peturboth I guess, the code is common
21:10:26linuxstbsoftware playback has peakmeter code in firmware/pcm_playback.c
21:10:44peturdoes it? oops
21:11:12 Join Paul_The_Nerd [0] (
21:11:52peturbut clipdetection is done in peak_meter_peek() I think (peakmeter.c)
21:12:48linuxstbI don't know.
21:12:49peturstill figuring out how it all hangs together there...
21:13:13peturwill find it, no problem there ;)
21:16:06peturlinuxstb: pcm_playback.c has really usefull code :D
21:16:13peturvoid pcm_calculate_peaks(int *left, int *right)
21:16:26petur{ (void)left; (void)right; }
21:17:14peturoh sorry, that's for an unimplemented platform
21:17:45***Saving seen data "./dancer.seen"
21:18:25Papricai need help with the make file
21:18:38Papricai dont know from where to start
21:19:23 Join MarcoPolo [0] (
21:21:45 Quit yeahx ()
21:22:33 Quit AliasCoffee ("Leaving")
21:24:01 Join jlo [0] (
21:26:13lodesican I assume the plugin buffer and the audio buffer are consecutive ?
21:36:17 Join infamis [0] (
21:36:41 Quit MarcoPolo (Remote closed the connection)
21:41:58 Join kyp [0] (
21:42:02kypman what a shitty channel
21:42:27peturwhat would you expect, it's saturday night
21:42:35kyplook, shut up
21:42:51peturonly freaks and loosers are at their pc right now...
21:43:01kypits 1pm and i just woke up
21:43:06 Quit eli_sherer ()
21:45:05 Part jlo
21:49:25 Join MarcoPolo [0] (
21:50:49infamiskyp, I beat you, I woke up at 2:15pm :p
21:51:06kyphey shut up
21:51:41infamisgo to sleep
21:51:44moozoohhello kyp.
21:51:51moozoohhow's life?
21:55:38infamisah, you know the usual, makin' some swings & tryin' to bring it own home, but you....oh....kyp.
21:57:29peturs/n ratio is getting worse every minute
22:01:00infamismaybe you should upgrade to a higher-quality device ;)
22:01:28*petur hugs his H340
22:07:20 Join solexx_ [0] (
22:09:11 Join qit [0] (
22:10:17 Part qit
22:11:00infamisso how bout that rockbox, 'eh?
22:11:05*infamis looks around
22:16:56 Join perl|wtf [0] (
22:18:55 Quit solexx (Read error: 110 (Connection timed out))
22:19:19IcyStorMWould it be possible to run iBoy on rockbox?
22:19:38*Mikachu points to rockboy
22:19:53BagderI believe rockboy is a more complete port
22:20:01IcyStorMIt is not
22:20:12Bagderbecause... ?
22:20:14IcyStorMCuz you cant even use key combinations or touch combinations for buttons
22:20:19 Join San [0] (
22:20:22IcyStorMthe iBoy keys are much better
22:20:24Bagderthat's just a key issue
22:20:32IcyStorMWith the Rockboy you cant play any game
22:20:33BagderI'm talking about the emulation
22:20:36Paul_The_NerdIcyStorM: That's not an issue with Rockboy. That's an issue with current limitations in the button driver.
22:21:05Paul_The_NerdIcyStorM: What doesn't work in Rockboy?
22:22:46Paul_The_NerdBesides the buttons/pad
22:22:46 Join Maxime` [0] (
22:22:46 Join StrathAFK [0] (
22:22:46 Quit Seed (Nick collision from services.)
22:22:46 Join Seed [0] (
22:22:54 Join Higgy__ [0] (
22:23:54linuxstbIcyStorM: What's stopping you installing ipodlinux and running iBoy there until somebody finishes adapting Rockboy for the iPods?
22:24:09IcyStorMBut why does it work so well with iBoy and linux then?
22:24:09 Join _Beowulf_ [0] (
22:24:09IcyStorMLook at the controls in iBoy and you should understand what I'm talkinb bout
22:24:10 Join Lynx [0] (
22:24:29Paul_The_NerdIcyStorM: As I said, the button drivers in Rockbox don't support that sort of function yet.
22:24:44Paul_The_NerdWell, button handler really.
22:24:48Mikachumy patch would maybe
22:24:52linuxstbIcyStorM: It's not that it can't be done, or we don't know what to do, it's just that nobody has done it yet.
22:24:57 Quit kyp (Read error: 110 (Connection timed out))
22:25:03Paul_The_NerdMikachu: Your patch would actually be exactly that, I believe
22:25:15 Join slasherx [0] (
22:25:43BagderIcyStorM: and do remember that Rockbox and Rockboy is multiplatform
22:25:50 Quit kernelsensei (Read error: 104 (Connection reset by peer))
22:26:15 Join kernelsensei [0] (n=boris@gentoo/developer/kernelsensei)
22:26:16Paul_The_NerdIcyStorM: Also, didn't you ask the same thing a day or two ago?
22:26:29IcyStorMThe problem is
22:26:45IcyStorMI have iPod 5G (video) and they are not supporting it yet..
22:26:49IcyStorMAn I really like rockbox
22:27:09IcyStorMThen Podzilla comes to 5G i will have all 3 firmwares on my iPod
22:27:11*Mikachu has a gba
22:27:18Paul_The_NerdIcyStorM: Well, if you used Mikachu's patch in the tracker, you could adapt Rockboy to use those controls. Why not give it a shot and submit a patch?
22:27:27*linuxstb has a collection of music to listen to...
22:27:38IcyStorMWhat patch?
22:27:49Paul_The_NerdThe absolute scrollwheel position one in the tracker.
22:28:16 Join lodesi_ [0] (
22:28:43lodesi_now my plugin can use the audio buffer
22:28:45 Quit perldiver (Read error: 110 (Connection timed out))
22:28:49lodesi_the limitation is of 5652 contacts on the recorder, if someone know that much people :)
22:29:20Paul_The_NerdMan, I don't even know 64 people that I'd store contacts for.
22:29:21safetydanSpeaking of plugin buffers... did we ever figure out if it was okay for the menu system to use the plugin buffer?
22:29:27*petur wonders how and when pcmrec_callback() gets called <−−- tip appreciated
22:29:28safetydanI'm talking about using it for file copying
22:29:35linuxstbThe problem is TSR plugins.
22:30:15safetydanah yes
22:30:29safetydanIs there any way to know how large they are?
22:30:31linuxstbWere you the person who did the file cut/copy/paste patch?
22:30:31Mikachuis there really any tsr plugin in addition to the battery test?
22:31:02safetydanlinuxstb, yeah that's me (and still working on it)
22:31:19safetydanrecursive directory merge has got me stuck at the moment
22:31:26linuxstbAt the very least you should use a 512 byte buffer - the size of one sector.
22:32:03linuxstbI think it should be OK to put that on the stack.
22:32:26safetydanon my local copy it's 512 bytes
22:32:48 Quit lodesi (Read error: 110 (Connection timed out))
22:32:53*petur discovers SYS_TIMEOUT and answers his own question
22:35:57 Join doctor [0] (n=doc__@
22:37:28 Quit Strath (Read error: 110 (Connection timed out))
22:39:16 Quit |Beowulf| (Read error: 110 (Connection timed out))
22:39:25 Quit Lynx_ (Connection timed out)
22:39:25 Nick Lynx is now known as Lynx_ (
22:39:27 Quit Maxime (Read error: 110 (Connection timed out))
22:39:30 Quit Higgy_ (Read error: 110 (Connection timed out))
22:41:00amiconnsafetydan: The menu system could use the plugin buffer, but keep in mind that the playlist viewer also uses it
22:41:44amiconnYou don't need to know yourself how large a tsr plugin is, plugin_get_buffer() will give you whatever is left
22:42:04Mikachuwhat to do if nothing available?
22:42:19amiconnThen it's just that - nothing available
22:42:51amiconnWould only happen with a tsr plugin that just fits the buffer
22:43:29Mikachutheoretically a tsr plugin could alloc all the buffer too
22:43:39Mikachuno? alright then
22:43:51amiconnTSR plugins must not allocate anything dynamically
22:43:51safetydanIf it fails, there's always the stack
22:44:12amiconnThis would e.g. make them crash when using the playlist viewer
22:44:38amiconnsafetydan: I don't know what you are planning to do, but keep in mind that plugins might also use menus
22:45:27safetydanamiconn, implementing file copy basically
22:45:41safetydantrying to find a large buffer to use for copying
22:45:43amiconnMikachu: There are 2 TSR plugins in CVS: the battery benchmark and alpine_cdc
22:45:55amiconnThe latter is archos only
22:46:30amiconnsafetydan: I keep thinking that more sophisticated file management could (should?) be a plugin
22:46:46safetydaninteresting, how would that work?
22:46:48Paul_The_Nerdsafetydan: It seems like it should be okay to terminate playback for that. I mean, it seems there's a decent chance playback could get pretty choppy anyway...
22:46:52safetydanCan a plugin hook the context menu?
22:47:19safetydanOr are you thinking more of a "file exporer" plugin?
22:47:36amiconnWhy would you want to 'hook' the context menu?
22:47:56safetydanBecause it's a clipboard, so you can cut/copy and paste files
22:48:11safetydanlike in explorer
22:48:17amiconnI have to say that I don't like the clipboard approach
22:48:38amiconnImho it's cumbersome to use
22:48:49safetydanPaul_The_Nerd, true, but I'd like to see how far we can go without stopping playback
22:49:01amiconn(judging from how it feels in the H3x0 stock firmware)
22:49:23peturwould it be ok if I replaced the current clip detection (two consecutive MAX samples) with a trigger on just one MAX sample?
22:49:25safetydanfor simple manipulation it works well enough
22:49:49Paul_The_Nerdsafetydan: A noble endeavor.
22:49:52amiconnMy idea would be a context menu entry for copy/move the same way as we have for delete right now
22:49:54Bagderpetur: I trust you on that, you're mr recording these days ;-)
22:50:11amiconnCould as well be implemented in the core provided it's not too much code
22:50:26peturthe way it is now is unreliable (misses quite some clipping)
22:50:37Moosamiconn: we already have one "create new dir" option, why not copy/paste/move ones?
22:50:38 Quit doctor ("Leaving")
22:50:49lostlogicamiconn: seen SoftwareCodecPlayback on the wiki? Is that the kind of doc on how it works you were looking for? :)
22:51:02safetydanamiconn, have you looked at the patch I've done so far?
22:51:14linuxstbamiconn: How would you specify the target directory for your copy/move ?
22:51:19peturBagder: and for recording, I think it's better to be on the sensitive side then to miss a clipped sample
22:51:25lostlogic4 in a row addressed to amiconn... popular guy.
22:51:29Bagderpetur: sounds logical, yes
22:51:43amiconnYou would navigate to the folder/file you want to copy/move, select (context menu)->copy, which would show another explorer view to select the target dir
22:51:54amiconnpetur: Just full scale is not clipping
22:52:26peturyes, there's no real detection possible is there?
22:52:44peturone can only guess it ent over the top
22:53:11peturbut two consecutive max samples really misses quite some clipping
22:53:19amiconnI'm not sure whether the UDA has clipping detection, I vaguely remember having read something about it
22:53:37peturI'll check the datasheets
22:53:45amiconnOn the archos, checking levels is the only way
22:54:14amiconnIt works reliable enough imho. Very slight clipping isn't really noticeable
22:54:39 Quit infamis ("food")
22:56:43peturheh... I can clap in my hands and get no clipping detection... and I'm processing all samples atm
23:00:36peturno clip detection on the uda (only overflow for output and silence for input
23:00:53peturI'll try something more intelligent like x max samples in s time... and see if it is sensitive enough
23:01:24peturbut with level monitoring, it will never be 100% reliable
23:01:51 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
23:03:35peturI'll keep the old way for playback and the sensitive way for recording
23:04:21Paul_The_NerdShould it even be bothering with clip detection during playback? Is there some behaviour when it clips?
23:05:10peturyou're right...
23:05:17amiconnIt should
23:05:33linuxstbI thought the playback peak meter displayed a clipping indicator.
23:05:38amiconnIt will tell you when it's clipping digitally, e.g. if your EQ gains are too high
23:05:47amiconnlinuxstb: It does
23:06:09peturdamn, you're right.... forgot all about EQ - never use it anyway
23:06:16BagderI noticed the tracker links from the forum were still the old ones
23:06:35Bagderso I added redirects for them
23:06:52 Part safetydan ("Leaving")
23:06:57Paul_The_NerdThe playback peak meter displays a clipping indicator?
23:07:03Paul_The_NerdWhat does it look like?
23:07:19linuxstbI think it's just solid vertical lines on the far right side.
23:07:26 Join miner49er [0] (
23:07:34linuxstbI saw them a lot when we had bugs in some of the audio codecs.
23:07:39peturyes, two-pixel wide lines
23:08:28amiconnpetur: Not only EQ - perhaps there's a badly replaygained (or even badly encoded) file
23:09:35peturso should I keep the old system for playback?
23:10:01miner49erhi there dudes
23:10:02amiconnI think so.
23:10:24miner49ercan I ask a question please?
23:10:47peturif a sample reaches max value, the chance is quite high that another one will clip.... I think
23:11:01miner49erI want to build both for archos and ipod nano.
23:11:13peturunless somebody normalized his audio to 100%
23:11:30miner49erbut the guide recommends different versions of GCC, so can I switch, if so how?
23:11:38amiconnThe current peakmeter behaviour has been like it is for a long time. The method was developed by kurzhaarrocker, and he tested it a lot
23:11:59Bagderminer49er: you on cygwin or linux?
23:12:06miner49eryunno, switch between versions to build for both?
23:12:07Paul_The_Nerdminer49er: Since you have to build the cross compilers, each version will have its own prefix anyway. As long as you set up the cross compilers as the instructions say, the configure script will handle using the right one
23:12:08amiconnHe plays in a band and uses his archos for recording
23:12:10miner49eroh, linux
23:12:26Bagderminer49er: then I suggest you download and install the recommended gcc versions
23:12:43miner49eroh right, ok cheers Paul
23:12:49peturamiconn: but it is unreliable for recording, I tested this after complaints (alright, one complaint) in the forum
23:12:52Bagderminer49er: since the ARM will be better with gcc4 while Archos is better with the gcc 3.3
23:13:16amiconnIn fact archos won't even build with 4.x
23:13:46peturm68k won't build either
23:14:39amiconnpetur: If you use every sample, it should be more reliable than on archos
23:15:11amiconnOn archos all we can use is the "quasi peak" value from the MAS, which we can read ~400 times per second at best
23:15:34peturok, then I'll do the sensitive thing only for SWCODEC recording
23:15:36amiconnClip detection is reliable enough with that...
23:15:56amiconnHmm. Perhaps it's actually a problem if we check every *single* sample?
23:16:39peturwe didn't, and I changed that, thinking the missed clipping detection was caused by it
23:17:10amiconnYes, but still we're checking single samples
23:17:11peturbut it was very lazy, only checking every 4th sample of only the first chunk
23:17:28amiconnMy guess is that the mas "quasi peak" is less fine-grained
23:17:48***Saving seen data "./dancer.seen"
23:17:53amiconn..and so it will be at max. if there's one max_value sample within a short window
23:18:07amiconnWhat about the following approach:
23:18:53amiconnDivide the pcm stream into "time slots" of HZ
23:19:09amiconn..and calculate the maximum across that
23:19:14 Quit Shadowarrior13 (Read error: 104 (Connection reset by peer))
23:19:31amiconnIf two subsequent time slots have a maximum of max_val, we have clipping
23:19:52 Quit slasherx ()
23:19:55 Quit San (Read error: 110 (Connection timed out))
23:20:31amiconnOf course we couldf experiment with the time slot length
23:21:33amiconnThat's certainly more sensitive than checking just 2 subsequent samples
23:21:51amiconn...but less sensitive to stray max_value samples
23:22:01 Quit lodesi_ ("Leaving")
23:22:21peturbut will it detect short clips? I doubt it
23:22:35amiconnIt should
23:22:48peturunless you keep the slot very small
23:23:01amiconnThe clipping indicator on archos does detect hand clapping as a clip
23:23:33peturI've seen it miss them on my h340
23:23:57 Join wizatcomputer [0] (
23:24:05peturI don't know...
23:25:11amiconnMy (wild) guess is that the MAs quasi peak uses the time resolution of the MP3 short blocks, which would be around 8,7 ms at 44.1kHz
23:25:57 Join Shadowarrior13 [0] (
23:25:57peturmaybe I could use the chunks of the dma as blocks
23:26:42peturbut clip detection isn't done there but in the peakmeter... bah
23:28:12pyroAre the forums or dev-mailing list better for asking technical development questions ?
23:28:19peturthe question is, how bad is it to indicate clipping if one sample reaches maximum... you're recorcing on the limit if that happens anyway
23:29:19peturpyro: you can ask here
23:30:54pyroWell, I'm trying to help look for how to configure the USB Charge Mode on the iriver h3x0 series.
23:30:55pyroLinusN wrote on the forums a while ago he went through all the GPIO pins looking for it with out any luck. I read all the i2c registers and compared before and after changing the iriver settings. No differences (other than obvious ones such as time changing)
23:31:11Paul_The_Nerdpetur: I think if any samples reach maximum while recording it's a good idea to indicate clipping, just because the user will know if that's a rare sharp noise, or indicative of the whole process.
23:31:30pyrohowever Linus wrote that a while ago, and was curious if there had been any updates since January 29th
23:32:17Bagdersounds like a question for Linus
23:32:26*petur points to lostlogic as well
23:32:37pyroyeah - lostlogic chimed in on the forum as well
23:32:58Bagderpyro: but to answer your prev q, I believe you reach more devs on the mailing list, even if quite a few also read the forums
23:33:42pyroI personally think it's kind of annoying having both communication channels - meaning forusm and mailing lists
23:34:04Bagderwell, the forums is the second place
23:34:19peturmore user oriented....
23:35:03 Join XavierGr [0] (
23:35:28Paul_The_NerdYeah, forums seem definitely more end-user oriented.
23:35:58pyrothat's true
23:40:02*lostlogic know nothing more than what's already been said WRT USB charging control
23:43:14peturwtf.. forget the previous descussion on whether to use two consecutive samples or not
23:44:31peturthere's something wrong with the way the value gets compared...
23:44:59peturtrying something to confirm this...
23:46:46amiconnBtw, I wonder why you've removed the local variables and use *peak_l and *peak_r directly
23:47:01amiconnAt lower -O levels, this might end up way slower...
23:47:31peturby not using registers you mean?
23:47:52amiconngcc is *sometimes* clever, but often rather stupid...
23:48:01 Quit Thus0 ("Leaving")
23:48:02peturI'm not that asm minded (yet)... sorry...
23:49:32lostlogicamiconn: on coldfire especially, even at -O2 gcc will tend to keep going back to memory for No Good Reason
23:49:47lostlogicseems to be much smarter on other arches for some reason.
23:50:00amiconnSame on SH1 (at least with -O)
23:50:37amiconnMaybe it's smart for x86...
23:51:07lostlogicx86 it'll optimize the world right out :-P
23:52:22amiconnI didn't check. x86 asm looks like gibberish...
23:54:07 Part IcyStorM
23:54:50 Join stereomike [0] (
23:58:04 Quit TCK (Read error: 110 (Connection timed out))

Previous day | Next day