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

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

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

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

#rockbox log for 2005-08-23

00:00:28Shebbto get it to work you will require a .lrc file in the same directory as the file you are playing
00:00:47Shebbnamed the same as the file you are playing just wil .lrc as its extension
00:01:22stripwax_Hm, I like the idea (on the mailing list) about 'unboosting' to 11Mhz if we've been stuck at 48Mhz for a while.
00:01:32 Join webguest77 [0] (
00:01:38stripwax_So I guess my games plugins didn't make it into the feature freeze :-(
00:01:45stripwax_^plugins^patches (sorry)
00:02:56 Quit Shebb ("CGI:IRC (EOF)")
00:03:00 Join randname [0] (
00:03:09 Nick randname is now known as Shbb (
00:03:40 Nick Shbb is now known as Shebb (
00:04:18ShebbI used: which is correct
00:04:57LinusNstripwax_: which ones?
00:06:14stripwax_ (which I think amiconn was looking at) and (which is pretty recent)
00:06:44stripwax_Although - is iriver not part of 2.5? if not I guess it's not relevant anyway
00:08:56Shebbpatches like those are nice for the cvs users though
00:09:47ShebbAnd as they almost like optimisions or fixes couldn't they get through the freeze?
00:10:24stripwax_If iriver isn't part of 2.5 it's not important
00:10:46stripwax_(I'm not sure if it is or isn't - it's not very clear)
00:12:26LinusNShebb: i created my own elapsed display plugin, and it is accurate
00:12:41ShebbNo its not, the iriver port will be released when all the completion things are done
00:12:58ShebbLinusN: What did you do different from me?
00:13:04LinusNShebb: your time display is not based on the real elapsed time
00:13:17LinusNint t = mint;
00:14:08LinusNfirst of all, don't add the latency
00:14:39LinusNhmm, you display a lot of different times :-)
00:14:46Shebbthat was just a test thing I should have removed that
00:14:58Shebbthe important one is the top one by the while(1)
00:15:09Shebbthat is based off file->elapsed
00:15:25Shebbif you ignor the latency which should not be there
00:16:50LinusNi now pasted your code into my plugin, and i don't see any difference from the wps
00:19:01ShebbI will get my build right up to date... I need to start using cvs to make this easier
00:21:37 Part stripwax_
00:25:41hardeepis the elapsed time on the wps accurate though? I know there's a fudge factor added to the resume offset in the playback code
00:27:35MoosHi hardeep, what's about your idea to add the patch about a warn if dynamic playlist modified ?
00:28:02hardeepMoos: Yeah, I liked that patch... just didn't get around to committing it
00:28:12hardeepwill have to wait until after 2.5 is released now
00:28:17Moosah ok :)
00:29:53LinusNhardeep: yes, the pcm buffer latency
00:30:39LinusNbut the "elapsed" time in the id3 struct is the same whether it is read by the wps or a plugin
00:32:12ShebbOk I can be more specific now
00:32:35ShebbIf I stop playback then play my song the laggin will occur
00:33:12Shebbhowever, if I rewind the audio file whilst it is playing
00:33:20Shebbthen restart my plugin
00:33:28ShebbMy plugin will show the correct time
00:34:00LinusNweird, i start a file, then start the plugin
00:34:07LinusNshows the correct time
00:35:38Shebbno, each time I stop and restart playback I have the wrong time until I rewind the audio a little
00:35:52Shebbthis is the latest cvs
00:35:58Shebbwith only my pluggin added
00:40:56LinusNrunning your plugin now
00:41:44LinusNthe .lrc file says "we're talking away" is at 0:36:0, but both the wps and your plugin time display say it happens on 0:34:0
00:42:10LinusNbut i don't see a difference between the wps and your plugin
00:43:38LinusNfunny, the lyrics catch up later in the song
00:47:09ShebbIf you play that, then rewind the audio
00:47:24ShebbDo the starting lyrics match up as they should?
00:47:26ShebbThey do for me
00:47:45Shebb"we're talking away" happens on 36 for me if I do that
00:47:58 Quit Strath ("Client closed")
00:48:27LinusNhappens at the very same time even if i rewind
00:48:45 Quit thegeek (Read error: 110 (Connection timed out))
00:48:58LinusNand the lyrics have the correct time later in the song
00:49:27ShebbStrange, I do not like different behaviour
00:49:50LinusNis your file by any chance vbr?
00:50:39Shebbstraight 256
00:51:56LinusNmine is straight 128
00:52:08 Quit markun ()
00:52:32ShebbShall we exchange? Since we both own them I guess that would not be illegal...
00:57:12 Quit Moos (" HydraIRC -> <- IRC has never been so good")
01:04:01 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
01:05:17LinusNShebb: i /msg:d you
01:07:18ShebbThanks! I will have to email I think
01:07:26ShebbWhat is your addres?
01:08:36 Quit ender` (Read error: 113 (No route to host))
01:09:42LinusNlinus at haxx dot se
01:10:36***Saving seen data "./dancer.seen"
01:11:05ShebbI have tried your file and I get exactly your problems
01:14:59ShebbStrange, my start will realign when I rewind, your start never aligns
01:18:47ShebbI am going to test out some different audio files...
01:19:20 Quit ashridah ("Leaving")
01:27:05ShebbSmells like teen spirit, mp3 straight 224 works fine
01:30:25 Join TCK [0] (
01:33:28ShebbKashmir, mp3 vbr 237 avg works fine too
01:38:16LinusNgotta sleep, cu tomorrow
01:38:22 Part LinusN
01:40:06ShebbWell there seems to be some slight bug in the elapsed time...
01:51:45 Quit pike ()
02:18:58 Join Strath [0] (
02:19:21 Quit hicks (Remote closed the connection)
02:21:22 Quit Shebb ("CGI:IRC (EOF)")
02:21:43 Join pike [0] (
02:24:54 Part arp
02:40:51 Join Febs [0] (
03:03:35 Join webguest68 [0] (
03:04:01webguest68how do i turn radio off on ihp 120 i'm getting radio and ogg playback at same time!
03:04:14FebsPress stop while on the radio screen.
03:04:34webguest68cool tdhat was weird
03:05:07webguest68how can i reshuffle playlist?
03:05:33FebsDo you mean in the middle of the playlist, or have it reshuffle when it is done?
03:05:41webguest68in the middle
03:05:54FebsTry re-loading it.
03:06:33webguest68does rockbox use same ogg codec as orig firmware?
03:09:32FebsAs far as I know, no. Rockbox was not based on the original firmware at all. I don't know if anyone even knows what the original firmware used.
03:10:40***Saving seen data "./dancer.seen"
03:12:29 Join phaedrus96 [0] (
03:15:05Paul_The_NerdI thought it was assumed it used Tremor?
03:17:32FebsCould be. Like I said, I don't know one way or another.
03:19:18 Join billytwowilly [0] (
03:19:24webguest68sounds cleaner to me
03:20:02billytwowillyhey, how do I delete files on the handheld without hooking it up via usb?
03:21:01webguest68is there a quick way to scroll to bottom of dir list?
03:22:15Febswebguest68, just press the up button at the top of the directory.
03:23:00webguest68ah i see
03:23:06webguest68if you're holding it it doesn't do it
03:23:31 Quit Xetrov ("random chances operate in my favour")
03:23:37billytwowillymy rockbox thinks it has a read only filesystem when i mount it. even if I chmod it.
03:23:41billytwowillyhow do I fix this?
03:26:00Febsbillytwowilly, not sure about your 2nd question, but as to the first, I believe delete is in the file menu. Navigate to the file, press and hold select, and when the file menu appears, choose "delete."
03:28:34billytwowillyI don't have select. I have a archos jukebox 6000
03:28:41 Quit phaedrus961 (Read error: 110 (Connection timed out))
03:30:05webguest68is there a quick way to turn shuffling off
03:31:18webguest68so everything's based off a playlist?
03:32:12Febsbillytwowilly, press ON+PLAY or press and hold PLAY.
03:32:21Febswebguest68, yes.
03:32:22Paul_The_NerdWebguest68: General Settings -> Playback, top of the list.
03:33:40webguest68record button does nothing right now?
03:33:47Febswebguest68, you can also toggle shuffle from the quick menu. Press and hold the A-B button.
03:35:04webguest68rockbox..well rocks
03:36:10FebsMuch of this stuff is in the manual. The manual is for the Archos models, but much of it applies to iriver, and you can use this wiki page to figure out the correct buttons:
03:36:46FebsOr you can look in the wiki manual, although that is very much a work in progress:
03:37:27*Paul_The_Nerd always forgets about the quick menus.
03:37:47*Febs contemplates writing some more of the wiki manual.
03:40:31webguest68what do you guys develop on linux?
03:43:00 Quit hardeep ("[BX] Have you huggled your BitchX today?")
03:43:42 Quit Strath ("Client closed")
04:03:41 Nick phaedrus96 is now known as phaedrus961 (
04:05:43 Quit webguest68 ("CGI:IRC")
04:58:13 Join bagawk [0] (
05:10:42***Saving seen data "./dancer.seen"
05:55:03 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.6/20050716]")
06:32:25 Quit bagawk ("Leaving")
06:34:17 Join ReKleSS [0] (
06:41:33 Quit dionoea (Remote closed the connection)
06:48:56 Join dionoea [0] (
06:55:28 Join tvelocity [0] (
07:10:44***Saving seen data "./dancer.seen"
07:15:40 Join LinusN [0] (
07:29:50Bgermorning :)
07:39:18t0mastime to go
07:40:27 Join Gibbed [0] (
07:40:29 Quit Rick (Nick collision from services.)
07:40:41 Nick Gibbed is now known as Rick (
07:47:25 Join thegeek [0] (
08:14:03 Join ender` [0] (
08:16:40 Join webguest88 [0] (
08:16:54CoCoLUSthese new creative x-fi soundcards are getting quite good reviews...
08:17:26 Quit webguest88 (Client Quit)
08:19:49tvelocitycreative soundcards always get good reviews...
08:19:54tvelocitybut they still suck
08:23:32CoCoLUSsounds like a prejudice :P
08:23:55tvelocityit is:P
08:24:36CoCoLUSthey are a bit pricey, though
08:27:49amiconnLinusN: You idea to simply cut the HD power on a crash to avoid it spinning all the time should be good enough, although it will probably cause the HD to perform an emergency park
08:28:00amiconnCrashes shouldn't happen too often
08:28:43amiconnBtw, we have 2 places where dead-end crashes are handled: panic() and UIE()
08:29:39amiconnThe handling of button check for rebbooting, powering down HD, setting low CPU frequency etc for these 2 places should be unified
08:32:05CoCoLUSreminds me of the worst case scenario... happend to me yesterday, while i was in the train, rb hang, disk kept spinning
08:32:22CoCoLUSand i had nothing small enough with me to press the reset button :)
08:33:46CoCoLUSlet me tell you, that thing gets -hot- aber running continously for an hour
08:34:42amiconnI know
08:35:49amiconnElevated temperatures will shorten the lifetime of a LiIon battery... bad
08:36:40CoCoLUSbut what could i've done...
08:37:35CoCoLUSi tried fidgeting a "pin" out of some paper tissue i had with me... then i tried to bend the keychain... nothing was small enough to fit in the reset hole
08:37:58CoCoLUSsure was funny as hell for the seat neighbour...
08:39:30CoCoLUSyou think that one-hour-long experience will noticable shorten the liftetime of the battery?
08:44:36amiconnProbably not
08:56:39ReKleSSanyone know anything about the simulator FPE thing yet?
08:59:09amiconnLinusN: I see someone upgraded sh-elf-gcc on the site... :-)
09:00, rockboxui dies before it even reaches _init()
09:05:30 Join einhirn [0] (
09:10:48***Saving seen data "./dancer.seen"
09:14:38BgerReKleSS: gentoo ?
09:20:58ReKleSSBger: yep
09:21:08Bgerknown problem
09:21:27Bgerit gives you FP exception, yeah ?
09:21:40 Join webguest45 [0] (
09:22:00Bgersearch in the irc logs starting from last week ...
09:22:27webguest45hi, i was wondering if it would be possible to have a feature in the playlist to empty the whole playlist ?
09:22:59webguest45or am I missing something, there is no other way than to manually remove one song after the other right ?
09:26:38ReKleSSBger: ok
09:27:02BgerReKleSS: starting from last week backward ...
09:28:42 Join B4gder [0] (
09:31:14webguest45RTFM ?
09:32:07ReKleSSBger: I've found something about it, about glibc 2.3.4 working and 2.3.5 not
09:32:24Bgersearch in
09:32:33BgerReKleSS: read more ...
09:32:38ReKleSSbah, just upgrade glibc
09:32:49Bgeri think it was working at last :)
09:33:31ReKleSSwould that be worth a todo list entry?
09:37:00webguest45hmm, sorry for being thick, but I cant find anything in the manual.
09:37:10LinusNReKleSS: todo for what?
09:37:35LinusNwebguest45: no, there is no way of clearing the playlist
09:38:13BgerReKleSS: maybe it's good idea:)
09:42:36 Join ashridah [0] (
09:44:14 Quit tvelocity ("Leaving")
09:49:39webguest45wouldn't that be really handy, and easy to implement ?
09:49:56webguest45have I missed the featue freeze ?
09:50:46B4gderyou're right in it!
09:53:41B4gdergood to see the fm build fit within 200K now
09:54:56CoCoLUSif i would want to write an entry into a textfile for every file thats played longer than 40 percent, where would such code fit best? :)
09:55:31B4gderit should use the same logic the runtimedb already does, imho
09:55:56CoCoLUShcl wrote that, right?
09:56:00B4gderbut it would requires some new thinking if we want to be able to choose what runtime code to run
09:56:22CoCoLUSi'm still avid about that audioscrobbler
09:56:28B4gderhcl wrote the runtimedb parts, Slasher added the callbacks and amiconn did them for the archos parts
09:57:42CoCoLUSi know hcl did already implement "times played" into his db
09:57:53 Join Lynx_ [0] (
09:58:16CoCoLUShowever, audioscrobbler plugins are supposed to only count tracks played longer than 40%/240 seconds
09:58:43ReKleSSwrong one
09:59:11B4gderCoCoLUS: yes, but either that code should be corrected to do the same, or we should work out a way to allow a user to chose what runtime code/counter to use
10:00:47CoCoLUSasking hcl to implement it into his code seems a lot easier
10:01:08LinusNCoCoLUS: don't hold your breath, hcl isn't very active nowadays...
10:01:16CoCoLUSi noticed
10:01:25LinusNthat bothers me
10:01:33B4gderthe whole runtimedb is in limbo
10:01:42LinusNwe might have to remove the runtimedb from the release
10:01:54LinusNand the database plugins
10:02:52CoCoLUSwould be a shame; maybe some other dev would be willing to look after those parts?
10:03:08LinusNlet's hope so
10:03:36B4gderunfortunately all those are very undocumented
10:03:48B4gderso its a lot of reading up
10:04:06B4gderactually, I don't think it is so much of a shame
10:04:13B4gdersince this is the situation right now
10:04:22B4gderand we don't see many people asking for this
10:04:25B4gderthus no one uses it
10:04:29B4gderand no one will miss it
10:04:52CoCoLUSmaybe it works so flawlessly nobody complains? :)
10:04:55LinusNhow could they use it? afaik, nobody knows how to...
10:05:07B4gderCoCoLUS: you ever used it?
10:05:59CoCoLUSnope, and i'm not going to, however, as i said, i'm avid about that audioscrobbler idea, and if hcls already existing code could provide support with only some small changes...
10:06:18B4gderI agree
10:06:27B4gderI would personally like to extract stats from it
10:06:45Bgerimho the idea behind the runtimedb is good, but it needs more work on it ...
10:07:02Bgerat least for documentation/bugfixes
10:07:24CoCoLUSthere's no way for rockbox to actually "look at" the files transfered during usb mode, is there?
10:08:08BgerCoCoLUS: rockbox doesn't have controll over the hard disk once the player is attached to computer
10:08:12CoCoLUSbecause if rb would build the database transparently during/after transfer, the whole db-thing would be a lot easier to handle... no perl/java scripts, etc.
10:08:30BgerCoCoLUS: no way
10:08:39B4gderwell, there's a C implementation partly done
10:09:35LinusNanother one of those hit-and-run commits... :-)
10:09:45B4gderbut he said that before he committed
10:09:49LinusNyes i know
10:09:50B4gderI asked him to commit
10:10:45LinusNi played with a pre-alpha of the lyrics plugin yesterday, quite amusing
10:11:00B4gderdid you sort out the time problem?
10:11:06CoCoLUSjeez, sounds like the original firmware
10:11:07LinusNi can't make it happen
10:11:21LinusNnot with my mp3 and not with his
10:12:05CoCoLUSis lyrics support actually wanted by users?
10:12:16LinusNi think so
10:12:22LinusNand it's a fun toy
10:12:28B4gderits fun
10:12:31B4gderkaroake time!
10:12:56CoCoLUSi'd rather see the darth vader vocoder :P
10:13:20B4gderwe are many, we can do them all
10:13:28LinusNnow *that* would be much more useful than lyrics support, wouldn't it? :-)
10:13:52CoCoLUSwell if we are on the topic of practically useless additions... ;)
10:13:55B4gder"finally a music player that can make me sound like darth vader" ;-)
10:14:24LinusNin the feature comparison table as well
10:14:36B4gderyes, that's be very good in the table
10:14:37LinusNDarth Vader mode: Yes No No
10:14:49*B4gder giggles
10:14:57CoCoLUSalthough i would not look away from a decent equalizer as well
10:15:17B4gderthen you better start hacking!
10:15:20LinusNthat will definitely come
10:15:30LinusNi have faith in preglow
10:15:55CoCoLUSb4gder, you should have noticed by now, i'd rather complain then code :P
10:16:03LinusNthat was his initial goal, before he got trapped in codec optimizing hell
10:16:26B4gderand I don't think we should let him out of that cage just yet
10:16:32LinusNoh no
10:16:49CoCoLUSi guess he already got nightmares about that... trapped in the infinite darkness between assembler and emac, or whatever
10:17:38 Join kurzhaarrocker [0] (
10:17:45eplHas anyone ever talked about creating a integer encoder for OGG?
10:18:13B4gderit wouldn't surprise me
10:18:14kurzhaarrockerThank you LinusN for committing the a-b repeat patch. Finally another _useful_ feature for rockbox.
10:18:45LinusNkurzhaarrocker: long time between the useful additions nowadays, eh? :-)
10:19:37webguest45after the Darth Vader mode, do you think you could implement "Playlist reset" ?
10:19:43CoCoLUSseems reasonable though, during a feature freeze period ;)
10:19:43kurzhaarrockerWell - obviously its more fun to investigate new hardware architectures. (which is a good thing, too)
10:19:43B4gderI would actually go for any good lossy format integer-encoder
10:20:16webguest45hmm, just thinking, as a workaround, could I not just load an empty playlist ?
10:20:46LinusNwebguest45: try it
10:21:33CoCoLUSbe sure to have something small at hand, though, something that fits into the reset hole ;)
10:23:02webguest45hmm, just realized, just off/on empties the current playlist
10:23:18webguest45not too elegant, but working
10:24:16 Quit kurzhaarrocker (Remote closed the connection)
10:27:57CoCoLUSso, i guess i have to delay the coding of my scrobblerbox-audioscrobbler-plugin :(
10:32:14CoCoLUSbecause of the obviously imminent removal of hcls code?
10:32:53B4gderwell, if the situation would be changed, the situation would be... eh, changed
10:33:12B4gderin its current state it is useless
10:33:16CoCoLUSi really have no problem coding the helper/submitting tool for the pc, but that whole playback code/runtime db thing is over my head
10:35:19CoCoLUSnot everyone is an embedded device professional ;)
10:35:36B4gderyou sure? all my friends are! ;-P
10:36:30CoCoLUSthat's... fine ;)
10:40:52 Join BBub_ [0] (
10:44:11amiconnLinusN: I don't think the runtimedb code should be removed for release
10:44:44amiconnRemoving it might introduce other bugs, and the runtimedb code seems to run quite stable
10:45:10amiconnThere is some documentation as well, and using the database isn't too difficult either
10:45:18B4gderyou used it?
10:45:35amiconnB4gder: Obviously I had to, since I wrote the arhos hooks
10:46:23B4gderbut saying it isn't too difficult is an understatement imho
10:46:37B4gderits hard as hell
10:46:48B4gderenabling the db is easy
10:46:53B4gderplaying music is easy
10:46:56B4gderbut then...
10:47:02amiconnJust enable it, and it will collect runtime data if you have a suitable tagdb
10:47:06B4gderhow the heck do you search or extract data from it?
10:47:16amiconn(suitable == containing hashes)
10:47:21B4gderI know that
10:47:23B4gderI've done that
10:47:49B4gderI gave up the last part though
10:48:13amiconnYou can then make search request files with databox
10:48:31B4gderif you understand it
10:48:45B4gderI guess I wasn't motivated enough
10:49:03amiconnRunning them will start searchengine, which will run the request against the tagdb and rundb, and generate search.m3u containing the result
10:49:07B4gderbut I don't think many users will be able to penetrate that
10:49:26amiconndatabox is quite simple, although a bit cumbersome in some points
10:49:46B4gderI felt like I was tricked into a maze
10:50:14B4gdertrying things blindly
10:50:14amiconnDatabox itself is quite logical, just some details need to be made more user friendly
10:50:29B4gderhoping it would be what I thought it is
10:50:56amiconnLike, we need a simple number entry screen. Currently databox uses the vkeyboard, which is sub-optimal for numbers
10:51:19amiconnAnd, our vkeyboard should have a caption saying what to enter
10:51:29Bgermaybe something with up-down buttons (inc/dec)
10:51:44Bgeramiconn: and maybe a "charset"
10:51:44amiconnYes, and left-right to choose the position
10:51:54Bgeryep :)
10:52:45 Join [IDC]Dragon [0] (
10:52:49B4gderwell, feel free to document what you know about this then because you seem to understand and know a lot more than me about these things
10:53:01B4gderI ran away screaming
10:53:10BgerB4gder: don't run too far :)
10:53:16amiconnThere is just one problem unsolved with databox, and that is you can't really enter requests that don't fit the screen, because you won't see what you select
10:53:24amiconnThis is a real problem on the player
10:53:35[IDC]Dragonhi guys
10:53:55B4gderhi [IDC]Dragon
10:54:20amiconnB4gder: Just ask me what you don't understand, because I wouldn't know what needs more documentation.
10:54:49CoCoLUSamiconn, since you seem to have a deeper understanding, what do you think, how much work would this "only count when 40% played" idea be?
10:54:58Bgerbtw, i must agree with B4gder that on first sight the things look very complicated ...
10:55:06B4gderamiconn: 1) a link from tha runtimedatabase wiki page to a page describing how to use the runtimedatabase to start with
10:55:07amiconnI can only repeat myself. Databox is quite logical, you basically 'click together' an if() condition
10:55:13 Join ghode [0] (
10:55:21 Quit webguest45 ("CGI:IRC")
10:55:26B4gderdatabox has no docs what so ever afaik
10:55:45*[IDC]Dragon is completely disconnected, no idea what you're talking about :-/
10:55:46Bgeramiconn: maybe the best is to write a user-end friendly doc...
10:55:52ghodeHmm i use the runtime db all the time...
10:55:56Bger[IDC]Dragon: about databox ... runtime db
10:56:04ghodeits pretty much explained on the wiki
10:56:25B4gderthe concepts are described, yes
10:56:28B4gdernot how to use it
10:56:30*[IDC]Dragon vaguely remembers something he never tried
10:57:05[IDC]Dragonthis is not ID3, but a dictionary or so, right?
10:57:10amiconn[IDC]Dragon: You can try now, even without an iriver :)
10:57:27B4gder[IDC]Dragon: it stores info about run-time info, what songs that have been played etc
10:57:54[IDC]Dragonah, so no "unrelated" db
10:58:08 Quit BBub (Read error: 110 (Connection timed out))
10:58:08 Nick BBub_ is now known as BBub (
10:58:33[IDC]DragonI came back for a minute to help Bootbox into v2.5
10:58:49CoCoLUSis there a way for plugins to read this runtime db, ie, is it in the plugin api?
10:58:56[IDC]Dragonit's all ready now
10:59:21[IDC]Dragontoday I've built images for all the Archos targets
11:00:02[IDC]Dragonand fixed the ROM start addresses to Bootbox requirements (not yet committed)
11:00:04amiconnDid you use latest cvs for that, i.e. containing yesterday's standby fix for USB mode?
11:00:26[IDC]Dragonyes, I've built it this morning
11:01:19amiconn[IDC]Dragon: I want to do 2 more things for full bootbox support, and work against our current code size tightness
11:01:31[IDC]Dragonwhich is?
11:01:41amiconn(1) Integrate ROM saving into firmware_flash, and remove it from the debug menu
11:01:57amiconn(2) Add the start address check for rombox images to rockbox_flash
11:02:09[IDC]Dragon(2) is done
11:02:16amiconnOkay, nice :)
11:02:39[IDC]DragonI needed to link that address into the image
11:02:47 Quit ghode ("CGI:IRC (EOF)")
11:03:44amiconnLooks like a clean way, but would not have been necessary
11:03:59[IDC]DragonI hope I didn't violate the feature freeze
11:04:14amiconnWe could just have used the start address from the vector table
11:04:40[IDC]Dragonthat is the reset vector, not the start address
11:04:46amiconnWe know the offset from start of image is constant as long as we don't change the linking convention
11:05:43[IDC]Dragonisn't the reset vector pointing anywhere?
11:05:59amiconnI'm not talking about the reset vector
11:06:13[IDC]Dragonrelying on that would rely on code ordering, optimization, etc.
11:07:21[IDC]Dragonafaik there is no such thing as a start address in the vector table
11:08:21[IDC]Dragonabout (1): there is also a "proposal" to unite the 2 flash plugins
11:08:49[IDC]Dragonand perhaps make a really user-frienly "suite" from it
11:10:50***Saving seen data "./dancer.seen"
11:12:36 Join [-AIR-] [0] (
11:12:48amiconnYes, it relies on the code ordering, but only on the fact that .init.text will stay directly behind .vectors
11:12:56[-AIR-]hey. the the irvier flash plugin for updating the firmware in rockbox ?
11:13:09[-AIR-]this one: apps/plugins/firmware_flash.c
11:13:42amiconn[IDC]Dragon: ..and the very first entry in the vector table is the start address
11:13:54amiconnI checked with the RAM version (easier to check)
11:14:17amiconnImage starts at 0x09000000, .init.text section is at 0x09000200
11:14:30amiconnFirst vector table entry is - 0x09000200
11:15:07[IDC]Dragonand the reset vector is the second entry?
11:16:19[IDC]Dragondoesn't seem so
11:17:22amiconnThis first entry is the power-on reset vector, and it does point to our start() function
11:17:27[IDC]DragonI'll check later
11:18:13[IDC]Dragonperhaps the vector just looked odd, since rombox has an extra 0x10 offset to be behind my flash header
11:19:34amiconnif (image[0x10] != image_load_addr + 0x210) ...
11:19:39 Join webguest89 [0] (
11:19:47webguest89Good Morning
11:20:17LinusN[IDC]Dragon: would it be possible for you to upload the uart boot files to the wiki
11:20:27[IDC]Dragonperhaps I'll revert and save 4 bytes
11:20:28LinusNyour site seems to be down
11:20:41[IDC]DragonI changed the provider
11:20:47webguest89I know the iriver is not the reason for the feature freeze, but....
11:20:47webguest89while you're in bug fixing mode...
11:20:55[IDC]Dragonand thought they are in wiki
11:20:56webguest89On the iriver a number of the confirmation message screens are wrong...
11:20:56webguest89For example recursivly inserting a directory in a play list, you are prompted
11:20:56webguest89and asked to press PLAY to confirm, it actually wants you to press SELECT
11:20:56DBUGEnqueued KICK webguest89
11:20:56webguest89there are other similar prompts that I can't remember.
11:20:57amiconnLinusN: I gave a current uart_boox package to rooomish
11:21:12amiconnI'll wait for confirmation that it works, the I can upload that
11:21:33B4gderwebguest89: that's a know flaw but no easy/quick fix
11:21:34amiconnwebguest89: That will be solved with localisation v2
11:21:51amiconn..which will hopefully implemented soon after the 2.5 release
11:22:13webguest89OK - ta... I guessed it would be known
11:22:24 Quit webguest89 (Client Quit)
11:27:08LinusN[IDC]Dragon: the link to the .zip with uart_boot.exe pointed to
11:27:50 Join hicks [0] (
11:28:13amiconnLinusN: Yes, and does no longer exists, already for some months
11:28:27amiconnT-Online dropped the old homepage server
11:28:30LinusNso it would be nice to have it in the wiki instead
11:28:40amiconn[11:20:57] <amiconn> LinusN: I gave a current uart_boox package to rooomish
11:28:41amiconn[11:21:12] <amiconn> I'll wait for confirmation that it works, the I can upload that
11:30:24amiconnHe didn't get it to work yesterday, but that was almost sure because he got the wrong pulldown resistors
11:30:48amiconnPulling down with 10kOhm against 10kOhm pullups won't give low level...
11:33:10LinusNyeah, saw that
11:34:40LinusNi have seen quite a few jukeboxes where the Rx input is broken in the cpu
11:34:53LinusNlet's hope his works
11:35:41amiconnHmm. Broken Rx means the remote won't work on such boxes
11:36:47LinusNhmmm wait a sec
11:37:06LinusNwhen i come to think of it, it has often been the Tx, not Rx
11:37:47LinusNi've seen it on two Players and on my FM
11:37:56LinusNand one Recorder
11:46:51 Quit phaedrus961 (Read error: 113 (No route to host))
11:49:26ReKleSSyay, ui simulator working I just need to find the controls
11:50:26LinusNReKleSS: good news
11:50:45LinusNso a glibc update solved it?
11:51:45ReKleSSis it just me or is a numpad absolutely necessary?
11:53:18 Join phaedrus961 [0] (
11:56:04amiconnReKleSS: The x11 sim has alternate key assignments which will work without numpad
11:56:23amiconnJust look at the console from where you started it
11:56:39ReKleSSI was looking through the source
11:59:52 Join midk [0] (
12:31:32midkHAHA, i just remembered the ollitrop "incident"... :) where's adi these days, anyhow?
12:33:15LinusNi wish i knew
12:34:10midkhm, logbot says he was on last a month ago..
12:44:51B4gderlogbot is the allknowning oracle
12:46:07midklogbot is the fast-tempered, kick-happy whore of the channel. :)
12:47:03B4gder"Thou Shalt Not Upset logbot" - old #rockbox saying
12:47:20midkfortunately, he doesn't have op.. such a loser... :)
12:48:33 Quit dionoea (Read error: 110 (Connection timed out))
12:55:31LinusNB4gder: does the script attempt to preserve the runtime info?
12:56:53B4gderyou mean v3? ;-)
12:57:10B4gderthe runtime info is not present in the songdb
12:58:20LinusN" Basic tagdb generation done, doesn't hash files and doesn't support keeping the runtime db in sync. (yet)"
12:58:30B4gderit hashes
12:58:38B4gderI don't know what "keeping the runtime db in sync" means
12:58:45B4gderand it isn't v2, it is v3
12:59:18B4gderbut I guess the v number doesn't matter in this case
12:59:36B4gderthat's just another problem with this db stuff
13:01:37B4gderI've updated
13:01:45LinusNi wonder where amiconn found the documentation that describes how to use the database plugins?
13:02:08B4gderI think he figured it out without docs
13:02:47LinusN"B4gder: Just ask me what you don't understand, because I wouldn't know what needs more documentation."
13:02:55B4gderah, right
13:03:02LinusNthere is obviously some docs that we don't know about
13:03:22B4gderthat's the only one I know of
13:03:54B4gderand it doesn't explain much of how to use things
13:05:34LinusNthe database plugin has no docs whatsoeever
13:06:26B4gder've never understood why they are two and not one
13:07:12B4gderis it because it isn't easy to make it both a normal and a viewer perhaps?
13:07:26LinusNi think it's so you can save queries as files and then run the search engine as a viewer
13:08:04LinusNit's a pity that the query files have a binary format
13:08:20LinusNso you can't make queries on your pc
13:08:26B4gderthat is silly
13:08:30LinusNoh yes
13:08:50B4gderand it would make sense to ship a few good default searches in the release build
13:10:53***Saving seen data "./dancer.seen"
13:11:43amiconnLinusN: Some of the usage-howto of the databox plugin can be derived from
13:11:56LinusNthe key is "derived"
13:11:57amiconnI figured out the rest
13:12:27 Join preglow [0] (
13:12:30amiconnI'd say it was fairly easy, but the I did some work on databox as well
13:12:35preglowhellos, all
13:12:38amiconn(The input handling for player)
13:12:46LinusNi can imagine that very few end users would figure out how to use it from those docs
13:13:09B4gderI'd be very surprised if the amount is more than 5
13:13:13amiconnI don't see why the binary search definition format prevents it from being generated on the PC
13:13:21B4gderit doesn't prevent it
13:13:28B4gderit just makes it hard
13:13:40B4gderwithout justification imho
13:13:41LinusNit just forces you to develop a tool for it
13:13:53LinusNi would prefer a text editor
13:13:59amiconnAh, you mean it should be plain ASCII?
13:14:04LinusNof course
13:14:05B4gderplain text is goodness
13:14:16amiconnShould be possible...
13:14:53LinusNmy wife will probably have a hard time with this: "QUERY := MEXPR (SORTBY SORTTYPE (AND SORTTYPE)*)? (CLIPBY CLIPPING)? (SHUFFLE)? EOF"
13:15:04LinusNclear as mud
13:15:31LinusNand it doesn't even say that i need the databox plugin to create the query
13:15:53LinusNyou can fogure it out by reading the suggestion at the bottom of the page
13:16:18preglowdoes codec opts count as bugfixes? :-)
13:16:33LinusNand i thought the wps language would be intimidating
13:16:45LinusNpreglow: if we are generous
13:17:11preglowi need something to do, don't i
13:17:20preglowmight as well start working on the eq or something
13:17:23HClwell, linus, i actually did write it down in ebnf
13:17:24B4gderhehe, there he is
13:17:38HClwhich is the standard language for writing down languages
13:17:43HClcomputer languages, that is
13:17:52B4gdertell that to my wife
13:18:22LinusNHCl: out problem is that we will have problems releasing 2.5 with the database as it is now
13:18:57B4gderone word: docs
13:18:58LinusNit suffers from severe lack of docs
13:19:19HCli could try to write some, but i'm not too good at writing documentation
13:19:24LinusNand the database viewer context menu doesn't work
13:19:38HClwha? o.o
13:19:43B4gderyou could start by describing how a human actually would use it
13:19:46HCldatabase viewer context menu?
13:20:00preglowdo we need the manual ready for the release?
13:20:03LinusNpush and hold the joystick
13:20:17LinusNpreglow: that is definitely preferred
13:20:21HClB4gder: its pretty simple, really. you start the databox plugin, enter a filename to edit, then make a search query with the joystick o.o
13:20:35B4gderI tried
13:20:36HClthen when you're done you click the <end> thing and it writes it to disk
13:20:39B4gderI gave up
13:20:43HCland then you click the file it made
13:20:45preglowi wonder if cassandra started the latex conversion
13:20:46HCland the searchengine runs
13:20:54HCland executes the search
13:21:00HCland writes the search results to search.m3u
13:21:03B4gderHCl: there's no docs saying all this
13:21:16B4gdernot easily accessable at least
13:21:37HCli can write it down in the databox section, i guess
13:21:49LinusNthere is no databox section :-)
13:21:56HClmhm, well i'll make one
13:22:08HClthere deserves to be a PluginDatabox, at least
13:22:16LinusNi'd like the two plugins to be one
13:22:23HClthe searchengine plugin is pretty much low level
13:22:39HClLinusN: that would eliminate all possibility of having predefined searches, or searches in a list that you can select easily.
13:22:40B4gderLinusN: that would probably not count as bugfix
13:22:50B4gderHCl: not at all
13:22:51LinusNHCl: why?
13:23:11HClbecause the plugin would have no way of knowing whether it has to execute the file or edit it
13:23:34B4gderwell, you can't click on one to edit it now either, can you?
13:23:47HClno, ofcourse not, that opens it with the searchengine
13:23:49HClif you want to edit it
13:23:55HClyou simply have to put in the filename in databox
13:24:11B4gderthen there will be no difference
13:24:25HClyes there would be?
13:24:36B4gderyou'd start it without a file and then load one to edit
13:24:45HClin the end, i'd like to have something like .rockbox/searches/
13:24:46B4gderor you'd click one and execute it
13:24:56HClwith searches a user has made, and can execute immediately
13:25:03B4gderbut we want them searches to be plain yext
13:25:06HCland they end up in the current playlist
13:25:11B4gdereditable on your pc
13:25:14HClwhy do you want searches to be plain text?
13:25:18HClthat is so not idiot-proof
13:25:22LinusNyes, plain text .rsp files would be nice
13:25:23HCldatabox *enforces* at the moment
13:25:26B4gderbut user-friendly
13:25:32HClthat the search engine querys are incorrect
13:25:41HCli so do not want to deal with retards who cant form a proper search query
13:25:52HClso databox enforces it to be correct or tells them to go to hell
13:26:04preglowwell, those people will still probably use databox
13:26:04LinusNHCl: databox can still enforce it
13:26:06HClit simply refuses to save a query
13:26:07preglowso it can still enforce
13:26:08HClif its not correct
13:26:18B4gderit could still be a picky parser
13:26:21preglowpeople who have no clue, wont try writing queries by hand
13:26:32LinusNas it is now, i can't generate the searches on my pc
13:26:49HClLinusN: i'd have to write a lexical scanner for databox, which isn't *that* much of a problem
13:26:50B4gderwe could even have a search query generator web pge
13:26:57HClbut users would get things like
13:27:11preglowB4gder: neat idea
13:27:16HCl"number found where string compare expected"
13:27:18HClstuff like that
13:27:24LinusNmy first try was actually to write a query in a text editor, because i couldn't imagine that the query files were not ascii :-)
13:27:38preglowi agree they need to be ascii anyway
13:27:40preglowbinary is badness
13:27:58HCli absolutely refuse to deal with retard users who mess it up cause of that..
13:28:04preglowthen dont
13:28:05HClso you'd have to deal with that
13:28:07B4gderwelcome to rockbox
13:28:17B4gderwe have end users
13:28:22LinusNwe deal with retard users every day
13:28:24HClB4gder: i specifically made databox in such a way that they *can't* screw it up
13:28:30preglowyou still can
13:28:36B4gderthey can't use it at all :-)
13:28:39preglowpeople can STILL edit your binary files
13:28:41CoCoLUSlet the mr-users vote :)
13:28:46LinusNHCl: no, you made it in a way that nobody would be able to use it :-)
13:28:52HClpreglow: yea, but thats their own fault if they edit the binary files
13:28:58HClLinusN: its really not *that* hard to use
13:28:59preglowas it is if they edit the ascii files
13:29:04HCldatabox is pretty straightforward
13:29:04B4gderHCl: it really is
13:29:25preglowi'm moving to faster internet land, so gotta get packing
13:29:28B4gderbut it is only due to the lack of ocs
13:29:32LinusNusing the databox plugin is pure guesswork today, as there are no docs at all
13:29:35 Quit preglow ("CGI:IRC")
13:29:36HClyes, so i'll write some documentation
13:29:49CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
13:29:49*amiconn feels like a nobody as I am able to use it...
13:29:57LinusNi am too
13:30:16LinusNi know how to use it, and i have even made searches with it
13:31:31B4gderI fired it up once and got so scared I've never tried it again
13:32:10B4gderI felt like writing a compiled regex
13:32:30amiconnNah, regex is certainly harder to understand than databox
13:32:41B4gdernot to me
13:33:19B4gderit is possibly because I had a different idea of what I'd meet
13:36:37LinusNHCl: what's the status on read/write of runtime db in rockbox? (the wiki says 90%)
13:38:52HClLinusN: it pretty much works, doesn't work on archos yet because archos needs the audio hooks that the runtime database uses implemented
13:38:58HClaside from that it should work fine
13:39:09LinusNHCl: those are there now, iirc
13:39:10B4gderamiconn fixed those hooks, right?
13:39:19HCloh, then it should work fine :)
13:39:44LinusNand what is this? " Save info to runtime db on play (fairly decent code in cvs)"
13:40:10HClwe changed the "fairly decent code" to a buffered design
13:40:14HClso that can be changed to done
13:40:21HClthough i could use some more bugtesting on it
13:40:38B4gder... more people will use it once they learn how to
13:41:31amiconnHCl: Don't you check the commits? ;) I added the hooks on Sunday, they seem to work okay
13:41:36B4gderand with plain text searches, it will be a killer
13:41:43amiconn...although I didn't test all cases
13:41:47HClamiconn: i haven't been able to look at rockbox for quite a while
13:42:50amiconnLinusN: Seen my 2 additions for the 2.5 todo list?
13:43:45B4gderI don't think the code size is a problem as long as we manage to keep below 200K until release
13:43:54LinusNamiconn: null pointer and code size?
13:43:56B4gderand then I suggest we go 2-step boot after 2.5
13:45:03*HCl has to charge his iriver..
13:46:27amiconnB4gder: I don't like this idea
13:46:39B4gdertwo step?
13:46:42LinusNamiconn: a 2-step boot loader?
13:46:50B4gderwhat other solutions is there+
13:47:01B4gderstop adding features?
13:47:01amiconnReduce code bloat...
13:47:15LinusNamiconn: where is this huge bloat?
13:47:50LinusNi can imagine that we can trim a few kbyte here and there
13:47:53B4gderI don't see how we can continue expanding features while remain at the same code size
13:48:09amiconnIt's not exactly huge, but there are quite some areas where we can save space without removing features
13:48:20[IDC]Dragonfrom the off: stop expanding sounds good ;-)
13:48:42amiconnOne point is localisation v2, but certainly not the only one
13:48:55[IDC]Dragonif we continue growing, there will be no mp3 buffer left
13:49:16[IDC]Dragonmissing the point of an mp3 player
13:49:17B4gderby the year 2043
13:49:36*HCl made a crude databox usage thing
13:49:47[IDC]Dragonhey, we have so many features, we had to drop the music playing thing
13:49:47amiconnLinusN: One thing I would like to do is to tell gcc to use .rodata.srt1.1 instead of .rodata.str1.4 on SH1
13:49:51HCli have to go in a bit, i know its pretty horrid and such, but i tried to explain the basics
13:49:59amiconnI didn't find a way to do this yet...
13:50:20[IDC]Dragonwhat is this 1.1 / 1.4 stuff?
13:51:16amiconn.rodata.str1 means simple (char) string, as opposed to e.g. .rodata.str2 would mean wchar
13:51:23amiconnthe .1 / .4 is the alignment
13:51:46[IDC]Dragonwewaste space in front of text strings?
13:51:51[IDC]Dragonwe waste
13:52:11amiconnactually between and after them, 1.5 bytes per string on average
13:52:25B4gderI don't see how we can manage to stay below 200K long-term
13:52:47[IDC]Dragon200k sounds reasonable for such a box
13:52:51B4gderwe'll have to remove features
13:53:03B4gderand make things iriver-only
13:53:06[IDC]Dragonor, out-source them
13:53:43amiconnSome things could indeed be removed from the core
13:53:52[IDC]Dragonif you guys keep insisting on a large firmware, you can use bootbox to start it
13:54:00B4gdernot all can flash
13:54:11[IDC]Dragonno, as a 1st leve loader
13:54:16[IDC]Dragon level
13:54:17amiconnE.g, if we bundle some default searches, the built-in database search can and should be removed
13:54:17B4gderyes, right
13:54:43LinusN[IDC]Dragon: who are "you guys"?
13:55:00[IDC]Dragonthose who like >200k ;-)
13:55:07*amiconn doesn't
13:55:12B4gderbuilt-in database search?
13:55:19B4gderisn't that a plugin?
13:55:32amiconnI mean the one Zagor coded
13:55:41LinusNah, the freetext search?
13:56:07B4gderthat could indeed be a plugin
13:56:11*[IDC]Dragon would welcome an option-decluttering
13:56:13B4gderI guess
13:56:24LinusN[IDC]Dragon: agree
13:56:26B4gder[IDC]Dragon: good luck managing that ;-)
13:56:42[IDC]Dragonworks only by authority
13:57:00[IDC]Dragondon't try demokracy on that
13:57:10LinusNone example is the line selector
13:57:29LinusNit could just as well be an inverse bar
13:57:46[IDC]DragonI always use the inverse bar
13:57:50LinusNme too
13:57:52[IDC]Dragonsaves a char
13:57:54B4gderI agree
13:58:23LinusNsaves a bit of code space
13:58:41*LinusN is tempted to remove it right away
13:58:52[IDC]Dragonso the FM build went >200K?
13:59:03[IDC]Dragonhow come it's green again?
13:59:10LinusNi changed the compiler
13:59:19[IDC]Dragonto which?
13:59:22amiconnThey are 80 bytes less than 200K now with gcc 3.3.6
13:59:39[IDC]Dragonah, I use 3.3.5
14:00:25amiconnMaybe 3.3.5 will manage that too, but 3.3.4 didn't
14:00:40[IDC]Dragonhow about the credits in a file?
14:00:46LinusNa plugin
14:00:52[IDC]Dragonor that
14:00:53B4gderwe make it a direct-used plugin
14:01:09 Quit webguest77 ("CGI:IRC (Ping timeout)")
14:01:28B4gderof the kind of previously removed ;-)
14:01:39B4gder"we previously"
14:02:32*amiconn contemplates applying a similar method to the credits as he did for the language strings
14:02:43amiconn...saving the need to store the pointers
14:03:01amiconn...and circumventing the string alignment problem at the same time
14:03:02LinusNwon't be necessary if we move it to a plugin
14:03:25amiconnWill save ~5.5 bytes per credit
14:03:53LinusNthe plugin will fit in memory without those optimizations
14:04:12[IDC]Dragonwhat size are the credits?
14:04:33B4gderthere are some 130 names now
14:05:22amiconn~2 KB
14:05:33amiconn(with all the alignment waste)
14:05:36amiconnPlus pointers
14:06:03LinusN30k left then
14:06:04amiconnAnother ~500 bytes
14:06:34[IDC]Dragonok, let's add pictures of everybody ;-)
14:07:17LinusN"the rockbox developer calendar 2005"
14:07:18B4gder1860 bytes for names + zero bytes, + 496 bytes for the pointers
14:09:21Bgerhq: at least 40k/picture ;)
14:09:31HCl :P
14:10:15 Join Shebb [0] (
14:11:09[IDC]Dragonabout 2.5K, not bad, I've had more effort in the past for similar savings
14:11:38[IDC]Dragon(menu system, config bits)
14:12:19B4gderwe can in fact make other parts to use that method too
14:12:29B4gderif we want to minimize core
14:12:37B4gderlike FM screen, recording screen etc
14:12:46LinusNno, no, no
14:13:39LinusNi can imagine moving some debugging stuff to a plugin though
14:13:58 Join muesli- [0] (
14:14:18[IDC]Dragondefinitely, I thought you must have already discussed that
14:14:26LinusNwe have
14:14:34B4gdera positive side effect would be that we'd get an even better plugin API
14:14:56[IDC]Dragonat points where it doesn't require adding very dirty access functions
14:15:08 Join stamppot [0] (
14:15:26LinusN[IDC]Dragon: i don't think we need to add access functions
14:15:41LinusNthe debugging plugin can take any shortcut it wants to
14:16:04LinusNread and write directly to memory etc
14:16:27[IDC]Dragonbut for I2C the can opens
14:16:52LinusNbut that os no different from having it in-core, is it?
14:17:34[IDC]Dragonno, but you have to expose it
14:18:53[IDC]Dragonand people will abuse it to write very dirty plugins
14:19:00ShebbLinusN: Have you looked into the elapsed time problems any more at all?
14:19:22LinusNno, as i can't make it happen on my player, not even with your mp3 file
14:20:20ShebbBut with your original file there is a clear problem: time elapsed reads 34 when it should be 36
14:20:56 Quit billytwowilly (Read error: 104 (Connection reset by peer))
14:21:02LinusNand what says that the .lrc file is more correct than the time counter?
14:21:17ShebbBecause I checked playing back the files on the computer
14:24:51[IDC]Dragonohh, no hello.rock any more
14:25:24LinusNShebb: i just played my mp3 file in Media Player, he still starts singing at 0:34
14:25:46 Join dionoea [0] (
14:26:10 Join tvelocity [0] (
14:26:54ShebbHmm, I think there must be something strange in my mp3 file, because I just tested yours without a problem
14:27:33ShebbMy mp3 file still changes its elapsed time when I rewind it but this is the only file I have managed to get this behaviour with
14:28:01ShebbI guess I have been wasting your time with this...
14:29:23LinusNShebb: how do i make it happen with your file?
14:30:05ShebbHow it works on mine is if I start it playing then later rewind
14:30:39LinusNi try rewinding, but the lyrics always start at 0:34, no matter what i do
14:32:33ShebbThen I really do not know why... Each time I rewind the lyrics start at 36 instead of 34 with my file
14:33:53LinusNah, now i see it
14:33:58ShebbIf I fast forward my file the lyrics will also start at 36
14:34:10LinusNi misunderstood and always rewound to the beginning
14:34:36LinusNok, the ff/rew seems to forget about the pcm latency
14:34:46ShebbHowever, this seems to be the only file I have that does this
14:34:56LinusNi doubt it
14:35:04 Join DangerousDan [0] (
14:35:34ShebbIt might have been something I encoded long back with mindows media player which would have used the Fraun whatever codec
14:38:20HCltemporarily, anyways
14:40:30 Join noC|andY`fRa [0] (
14:41:50ShebbIs it possible to retrieve the decoded information sent by the codecs from a plugin?
14:43:10ShebbSo that I might be able to play an mp3 and either save it to WAV or fx it up a little?
14:45:44LinusNShebb: i think i know what the problem is
14:46:30 Quit tvelocity ("Leaving")
14:46:33LinusNthe id3 tag is 77kbytes
14:46:43Bgerwow :)
14:47:03LinusNcover image
14:47:21LinusNand the mpa plugin doesn't account for the id3 tag when it seeks
14:47:41CoCoLUSHCl? :)
14:49:25 Part Shebb
14:49:36 Join Shebb [0] (
14:50:17ShebbRight, well its good that that problem was found
14:53:14LinusNShebb: your persistence paid off
14:53:22LinusNi have now committed a fix
14:53:31Bgerdoes "Mono Left" mean that u take the samples from the left channel and "write" them out to both of them?
14:53:54Bgerok, 10x ;)
14:53:54CoCoLUSmust sound horrible :)
14:54:12LinusNor rather, it will work that way when it is implemented on the iriver
14:55:08Bgeron archos it's different ?
14:55:44LinusNit is done in hardware on archos
14:56:00ShebbShouldnt sound bad. Is just the same as listening through one headphone
14:56:15BgerShebb: maybe a little better in fact :)
14:56:49Shebbyeah, I hope!
14:59:01ShebbSomething I would like to have in rockbox would be some study mode where you could skip by a specified time preferably by a time given in milliseconds
14:59:21ShebbHowever this kind of thing would really be code bloat in the core firmware
14:59:44ShebbWould it be possible to play with the architecture so I could implant this in a plugin?
14:59:45LinusNwhat should be in the core if not playback features
14:59:50CoCoLUSthe darth vader mode would be code bloat :)
15:00:06stamppotdarth vader mode?
15:00:09LinusNShebb: no need for a plugin imho, it should be in the core
15:00:20CoCoLUSyou know, dark, evil guy? ;)
15:00:24 Join muesli__ [0] (
15:00:36stamppotOh, you mean Gates mode...
15:00:45ShebbIt would be cool in the core but the way I would like to use it would be rather specialist
15:00:54stamppotDo we intend to play Windows Media then... :-)
15:01:01LinusNShebb: how would you use it?
15:01:20LinusN"we-we-we-welcome!" :-)
15:01:21ShebbI would take a song with a constant tempo
15:01:42ShebbI would calculate how many milliseconds make up a bar or something
15:02:02ShebbThen I would use the study mode to remix in a rahter simple way
15:02:49LinusNso what you really want is a playback speed setting?
15:03:01LinusNwithout changing the pitch
15:03:17ShebbNo, that is a seperate feature that would be nice
15:03:35ShebbHave you ever used Traktion or something similar?
15:04:10LinusNanyway, i gotta go now
15:04:13LinusNcu later
15:04:19Bgercu :)
15:04:22 Part LinusN
15:04:35Shebbright, I will post it in the forums anyway. There is a topic on it.
15:05:45Bgerthese "LANG_ON" and "LANG_OFF" are used in something like "blala: str(LANG_ON)" ??
15:07:09t0massomething like that
15:07:18t0masto make rockbox display things in the native language
15:07:28Bgeri mean "<setting>:str()"
15:07:33Bgernot the button?
15:10:55***Saving seen data "./dancer.seen"
15:17:39 Join zezayer [0] (n=jake@
15:17:40 Quit muesli__ (Read error: 104 (Connection reset by peer))
15:18:44 Quit muesli- (Read error: 110 (Connection timed out))
15:18:51 Quit ashridah ("sleep")
15:18:55 Quit Febs (Read error: 110 (Connection timed out))
15:19:32 Join guyg [0] (
15:20:20guygHi there. I've been on holiday for a fortnight, and want to see what's been going on with rockbox. Is there somewhere I can find a CVS log for the past two weeks? Thanks
15:22:39 Join Domonoky [0] (
15:28:15B4gderin CVS ;-)
15:28:38B4gderpossibly we should make such a little log
15:28:41Bgerguyg: use dayly builds
15:28:47Bgeri mean
15:28:56B4gderright, there's a daily changelog for each daily
15:29:33 Join webguest49 [0] (
15:30:14Bgerand change the day number in this URL
15:30:51guygthanks a lot. I'll check them out
15:30:58Bger <= better way :)
15:31:11guyg(not in the CVS sense, obviously)
15:35:00 Join billytwowilly [0] (
15:35:32 Quit guyg ("CGI:IRC")
15:36:20 Quit webguest49 ("CGI:IRC")
15:37:46BgerB4gder:<= the anchors here are invalid (point to or it's just my firefox ?
15:38:33B4gderah, right
15:39:51[IDC]Dragonamiconn, do you read?
15:40:35Bgergood B4gder
15:41:02B4gderI should make that page get auto-updated
15:41:10B4gderbut not right now
15:43:17BgerB4gder: any idea whether the "player" works with cyrillic chars ?
15:43:26B4gderit doesn't
15:43:41Bgerok, it doesn't have them in its charset, but ...
15:44:09B4gderthe player only has 8 (or 4) reprogrammable letters
15:44:53Bgeri think amiconn said something about remapping cyr. chars to their like-looking latin ones
15:45:12B4gderah, yes that might be a way
15:45:26BgerB4gder: i know for the 4 and 8 chars ...
15:45:53 Join ep0ch [0] (n=ep0ch@
15:47:14B4gdercurrently the player code strives hard to support iso8859-1
15:47:21ep0chhey, does anyone know if there's is a BPM plugin for rockbox? i.e. calculates the tempo of button tapping?
15:47:32B4gderbut it could of course be changed to do the same for cyrillic
15:47:57Bgeri can help with that .. :)
15:48:14B4gderand it uses a replacement table for the cases where it can't reprogram the letters
15:49:22B4gderep0ch: I believe the metronome plugin have something like that
15:50:20ep0chdamn and i wanted to write my first rockbox plugin :)
15:59:59amiconn[IDC]Dragon: Here now
16:00:25amiconnB4gder: I still plan to implement cyrillic for the player
16:00:40B4gder"still" ?
16:00:42amiconnI have a complete replacement table
16:00:56amiconnI already plan to do that for some months
16:01:02B4gderdidn't know
16:01:31B4gderisn't it mostly to make a setting and use alternative tables?
16:01:40amiconnI'm a bit undecided when to start; this will for sure need adjustment for unicode support
16:01:51B4gderoh right
16:02:12amiconnYes, it's "just" using alternate tables, however, the current tables have fixed sizes
16:02:25Shebbep0ch, you could always do a nicer bpm calc, one that averages it all out rather than relying on the interval between each pair of clicks
16:02:53amiconnShebb: The metronome plugin does average
16:03:23amiconnB4gder: My first attempt wil probably be hackish, i.e. support cyrillic *instead* of extended latin
16:03:41B4gdermake sense
16:05:15Bgeramiconn? i don't think this is really good
16:05:28Bgerif you have both latin & cyrillic tags for example ?
16:06:19amiconnI said extended latin, not latin
16:06:50Bgeraaaa, yeah
16:06:57amiconnPlain latin, i.e. ascii, is part of all codepages
16:12:21 Join muesli- [0] (
16:17:36 Join rooomish [0] (
16:17:36 Quit muesli- (Read error: 104 (Connection reset by peer))
16:19:22rooomishamiconn: Parcial success - I tried your exe - downloading monitor done that is OK, but f.e. dumping file - nothing happened (how long it takes)?
16:19:56amiconnTried to dump the rom to a file via uart_boot?
16:20:23amiconnShould take around 2.5 minutes
16:20:36rooomishOKI try to wait
16:21:39amiconnRS232 is sloooow, especially with 14400 bps...
16:21:56amiconn[IDC]Dragon: ping...
16:23:38rooomish[IDC]Dragon: I waited here for uart_boot exes more than one month :-(, please have you link for downloading?
16:23:43 Join Paul_The_Nerd [0] (
16:24:59amiconnrooomish: If you manage to get it working, I can upload my compilation
16:25:08amiconn...because I then know it works
16:27:05 Nick t0mas is now known as t0mas_ (
16:28:04 Nick t0mas_ is now known as t0mas (
16:28:31 Nick t0mas is now known as t0mas_ (
16:29:29rooomishamiconn: In that case yes - but now it is about 10 minutes :-( and cursor is still only blinking. Do you think, that it could be "thanks" to your exes?
16:30:20 Nick t0mas_ is now known as t0mas (
16:30:48t0maswhat's the rockbox policy on hostname cloaks?
16:30:49 Join muesli- [0] (
16:30:54t0masyou can request a cloak on freenode
16:31:50B4gderwhat's that?
16:31:55 Join webguest05 [0] (
16:32:09t0masthis: /whois t0mas
16:32:11t0mas* [t0mas] (n=Tomas@unaffiliated/t0mas): Tomas
16:32:18t0masyou can change unaffilated to rockbox
16:32:28amiconnrooomish: Sometimes I got similar behaviour. It seems that a reconnect doesn't always work, i.e. downloading minimon in one call of uart_boot, then try to do another action in a second call often fails
16:32:40amiconnDoing it all in one call worked for me
16:32:52t0masB4gder: are you registered with the rockbox group on freenode?
16:33:01t0masas in.. is linus the only owner? or are you it too?
16:33:03B4gderI once was
16:33:07B4gderlinus is not owner
16:33:12t0masah ok
16:33:19B4gderadiamas is the owner
16:33:22B4gderand he is gone
16:33:25t0masoh lol
16:33:31rooomishamiconn: how to manage it on one step?
16:35:15amiconnuart_boot -p <Your_COM> -dump <filename>
16:35:35amiconnDoing it in a second call actually requires an additional option -nodownlaod
16:35:40amiconn* -nodownload prevent the tool from trying to download minimon again (which won't work)
16:36:30 Join b3u [0] (i=beu@gentoo/developer/beu)
16:36:32rooomishamiconn: I used: uart_boot -p COM1 -f flashup.bin, than there was a message Downloading minimon done and then second line "dumping file 1.bin" with cursor blinking
16:36:56rooomishsoryy flashing file 1.bin
16:37:09rooomishthe same with dumping
16:37:14amiconnHmm, I have no idea how long *flashing* will take
16:37:53amiconnI never flashed via uart_boot. I just used it for rom dumps (-dump) and executing code (-exec)
16:37:55b3uloha :)
16:38:04rooomishamiconn: aha
16:38:23amiconnHmm, perhaps something is wrong with the minimon in my package...
16:38:47rooomishamiconn: ? you fount something,
16:39:11amiconnNo, it's just a suspicion
16:39:32amiconn[IDC]Dragon: ping...
16:39:44rooomishamiconn: it it teriible, that IDC Dragon is not responding waaaaau
16:40:34rooomish[IDC]Dragon: I waited here for uart_boot exes more than one month :-(, please have you link for downloading?
16:41:07[IDC]DragonLinusN also said it's not in the wiki, to my surprise
16:41:25[IDC]DragonI'll add it tonight, if missing
16:42:03rooomishYes it is missing me too much ans so looong
16:42:19rooomishOKU will wait
16:43:11rooomishamiconn: do you thing that I can interrupt the "flashing"?
16:44:11[IDC]Dragonwhy would you want to?
16:45:02rooomishI have some exes from amiconn, but maybe it is not working - see my few messages above
16:45:18rooomishnow it is about 15 minutes
16:45:25 Join webguest24 [0] (
16:46:29rooomish[IDC]Dragon: and amiconn is not shure if they are correct
16:46:33webguest24with the replaygain stuff does it set the levels according to some baseline? Or all files relative to each other?
16:46:55amiconn[IDC]Dragon: You wanted to ask me sth?
16:48:01[IDC]DragonArchos is "bugging" me how we progress with the wav codec
16:48:16amiconnDid we get the okay?
16:48:44amiconnI would like to puplish the docs in the wiki, that would help development a lot
16:48:45[IDC]Dragonwe can start coding
16:48:46Paul_The_NerdWebgues24: Replaygain is described at
16:49:14[IDC]Dragonno publishing OK yet, that's my standard return question
16:49:17amiconn[IDC]Dragon: *we* == you and me only as long as we can't publish the docs
16:49:32 Part ep0ch
16:49:55amiconn..and unfortunately I don't have much time for hacking on it as there are a couple of other things that need work
16:50:16[IDC]Dragonthis is a running gag, since I wa bugging them for so long
16:50:39amiconnYes I know. Someone pls invent a time machine :-/
16:50:51[IDC]Dragonnow they bug me, to "take revenge"
16:53:19[IDC]DragonI had the impression you're deeply interested in this
16:53:26amiconnYes I am
16:53:36 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
16:54:00amiconnHowever, I have ideas and rockbox things enough for months of coding time :-(
16:54:04 Quit muesli- (Read error: 110 (Connection timed out))
16:55:17 Join Lear [0] (
16:55:57rooomish[IDC]Dragon: so PLS do not forget to update wiki - thank you so much!
16:58:19webguest24i looked at i couldn't really figure it out. I understand that you can set the gains on an individual track basis, or as an album, but how does it make songs from different albums the same loudness? is there a baseline loudness it uses or does it scan all the files and set them relative to each other? What happens when you buy a new album? Does that have the same relative loudness as the already replaygained albums? I'm guessing it
16:58:46Paul_The_NerdThere's a baseline
16:59:38 Part b3u ("freightlined")
16:59:38Paul_The_Nerd89dB, if I recall what I read somewhere.
17:00:04*Paul_The_Nerd knows nothing about audio, so does not know what this number really represents.
17:02:53webguest24cool pretty smart
17:03:07webguest24works really well so nice to not have to mess with the volume every song in the car
17:03:52Paul_The_NerdHonestly, if all rockbox did was offer basic music playback and replaygain, that'd be enough to get me to switch. :)
17:04:20*amiconn doesn't care about replaygain
17:04:36Paul_The_NerdI didn't used to.
17:04:37Learwell, it does that (on the iriver at least), and a lot more...
17:04:49Paul_The_NerdThen I downloaded the OCRemix archive, and... ouch.
17:05:11Paul_The_NerdYeah, I'm loving everything else it does too Lear. I switched before replaygain support was even in.
17:05:51LearAh, I read it as if you didn't use rb... :)
17:06:23Paul_The_NerdNo, I've been using it for a while. I keep trying to come up with some ways I can be more supportive of it.
17:06:40webguest24it's pretty sweet i'm impressed
17:06:50webguest24gapless is huge
17:06:59webguest24and being able to reshuffle playlists
17:07:16webguest24seems to sound better also
17:10:19 Quit Lear ("Chatzilla [Firefox 1.0+/undefined]")
17:10:58***Saving seen data "./dancer.seen"
17:12:07 Quit webguest05 ("CGI:IRC (EOF)")
17:13:33 Quit webguest24 ("CGI:IRC")
17:17:13 Join webguest07 [0] (
17:23:50 Part zezayer
17:24:18 Quit Shebb ("CGI:IRC (EOF)")
17:26:39 Join webguest56 [0] (
17:27:21webguest56is the UDA1380 the best DAC available in its class ?
17:28:07amiconnI don't think it is
17:28:40amiconnI'm not sure whether it is DAC related, but the crosstalk suppression on iriver is rather poor
17:32:40webguest56hmm so there is room for ipmrovement
17:34:58 Join Lear [0] (
17:35:17webguest56is the UDA1380 a 8bit or 10 bit ADC ?
17:37:14 Join tvelocity [0] (
17:39:21phaedrus961hey all, I've found a problem where convbdf writes incorrect offsets if the glyphs in the bdf are out of order
17:39:32phaedrus961see here for patch:
17:41:19 Part webguest56
17:46:50 Join amiconn_ [0] (
17:48:14 Quit pike (Read error: 104 (Connection reset by peer))
17:54:15 Quit amiconn (Nick collision from services.)
17:54:15 Nick amiconn_ is now known as amiconn (
17:59:08 Join crwl [0] (
18:00:14 Join pike [0] (
18:01:46 Nick Lynx_ is now known as Lynx_awy (
18:04:59 Join bagawk [0] (
18:05:38 Nick Lear is now known as Lear_away (
18:11:41 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
18:12:47 Quit TCK (Read error: 104 (Connection reset by peer))
18:16:03 Join DangerousDan [0] (
18:31:32 Join Sucka [0] (
18:47:09 Join dpassen1 [0] (
19:00:56 Quit Sucka ("a bird in the bush is worth two in your house")
19:02:36 Join godzirra [0] (
19:02:46*godzirra waves
19:03:46 Join markun [0] (
19:06:50[IDC]DragonI'm off
19:06:56 Quit [IDC]Dragon ("CGI:IRC")
19:11:02***Saving seen data "./dancer.seen"
19:14:38 Join Moos [0] (
19:19:23 Quit dpassen1 (Read error: 110 (Connection timed out))
19:21:22 Quit tvelocity ("Leaving")
19:21:50 Join dpassen1 [0] (
19:28:02 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.6/20050716]")
19:29:51 Quit bagawk ("Leaving")
19:32:14 Join solex [0] (
19:33:22 Join matsl [0] (
19:44:20 Quit solex_ (Read error: 110 (Connection timed out))
19:45:40 Quit dpassen1 (Read error: 110 (Connection timed out))
20:06:54 Join Sucka [0] (
20:07:00 Quit Sucka (Remote closed the connection)
20:08:20 Part godzirra
20:15:08 Join Shebb [0] (
20:21:00 Quit Lear_away ("Chatzilla [Firefox 1.0+/undefined]")
20:23:36 Quit rooomish (Read error: 104 (Connection reset by peer))
20:23:43 Join rooomish [0] (
20:29:24 Quit stamppot (Ping timeout: 14400 seconds)
20:37:56 Join wrecker [0] (
20:40:16 Quit wrecker (Client Quit)
20:40:29 Join wrecker [0] (
20:56:10 Part Domonoky
21:11:05***Saving seen data "./dancer.seen"
21:14:41 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
21:19:49 Quit thegeek (Read error: 104 (Connection reset by peer))
21:21:01 Join lost|X40 [0] (
21:21:13 Join thegeek [0] (
21:23:29lost|X40not complaining, just wanted to mention that a recent rockbox build at least partially kills usb on my linux box
21:24:41lost|X40by that I mean, using kernel 2.6.12 and udev, loading iHP w/ iHP firmware = listing in /dev and mount = success; loading iHP w/ rockbox = no listing in /dev, of course no mount; going back to iHP firmware = no listing in /dev
21:25:22lost|X40it only took me like 2 hours to figure out this was not a suspend problem..
21:27:34amiconnrecent meaning what here?
21:27:53lost|X40downloaded day or two ago, daily builds
21:28:19amiconnPls check the exact version.
21:28:53lost|X40is there a way to check that from w/i the iHP?
21:29:21lost|X40i see CVS <> flash at the bottom of the screen when booting, but too quickly for me to read
21:29:36lost|X40take that back 20050821
21:29:48amiconnCheck Menu->Info->version
21:29:57amiconnThere you have 2 seconds to read the version
21:30:21lost|X40ok, yea 050821
21:30:29amiconn20050821 had a bug with USB. 20050822 fixed that
21:31:22lost|X40that's quick turnaround, thanks alot
21:32:04lost|X40wasn't even looking for a fix, just wanted to let you know
21:32:04amiconnBtw, I wonder why Linux doesn't recover from a failed USB connect attempt
21:32:15lost|X40yea, that's my next chore
21:32:15amiconnOthers reported that too
21:32:51lost|X40or it might be too great of a chore for me, but i'll try
21:33:08lost|X40*shrug* i blame everything /dev related on udev
21:33:12amiconnWindows doesn't have problems with that. Of course connecting an iriver 20050821 daily doesn't work, but subsequent attempts (new version, bootloader USB etc) work fine
21:33:45lost|X40yea, i noticed that too
21:35:53lost|X40ok, time to reboot to fix usb, thanks for the help
21:35:59 Quit lost|X40 ("$##@ usb")
21:44:17 Join lost|X40 [0] (
21:44:30phaedrus961amiconn: what did you mean when you said the unicode patch needs proper bidi support?
21:45:21phaedrus961(current bidi code is now working with unicode)
21:47:40amiconnI'm not an unicode expert, but I think there are at least 2 points:
21:48:29amiconn(1) The implementation should know which languages need which direction, and e.g. print arabic and hebrew parts of a string rtl, and latin etc ltr
21:49:10amiconn(2) Iirc there are some 'bidi-mark' codes, usable to manually switch direction
21:49:20phaedrus961(1) is done
21:50:04amiconnIf (1) is done, the option for enabling bidi support can go away. It is only needed for the current codepage-based implementation
21:50:57phaedrus961good point, I'll change that
21:51:10wreckerI'm just updating the WikiManual. Can someone explain to me in easy terms the difference between Track and Album gain on replaygain??
21:51:55phaedrus961also, I dont know if the entire string should be reversed or just blocks at a time, the current code does it in blocks, but has support for both ways
21:51:56paughhrm. usb unplug before unmounting is a bad things i take it?
21:52:42amiconnphaedrus961: The string can contain parts in different languages, so reversing the entire string is usually wrong
21:53:43amiconnUnicode has some more features that are difficult to implement in a compact way, e.g. combining marks
21:53:51amiconn- imho -
21:55:07phaedrus961yes, what I mean is this: the current implementation will change this "hebrew1 english1 hebrew2" to this "1werbeh english1 2werbeh"
21:55:51phaedrus961I think maybe it should be this "2werbeh english1 1werbeh"
21:56:39phaedrus961combining marks are also a problem, yes
21:57:22amiconn'Fun' to read...
21:57:44phaedrus961maybe I should ask someone who uses the language how it should look?
21:58:52paughwrecker, <- replay gain info
22:00:43 Join stamppot [0] (
22:01:46 Part stamppot ("Kopete 0.10.2 :")
22:03:20phaedrus961anyways, I will post the patch to the tracker soon, so that others can test it and give feedback on what could be done differently/better
22:11:06amiconnphaedrus961: There is some reference code for bidi linked in the web page I gave
22:11:40amiconn...even a C implementation exists, although not explicitly mentioned
22:12:54wreckerpaugh, I've just looked up some useful descriptions for this on the foobar website. Thanks anyway. I'll check out your link too.
22:19:17paughwrecker, okies
22:20:09 Join muesli- [0] (
22:23:30 Quit wrecker ("CGI:IRC (EOF)")
22:27:17phaedrus961hmm, it seems a bit involved to implement (just skimmed through it), unfortunately I don't have the time to do it
22:58:15 Join Strath [0] (
22:59:58 Quit DangerousDan (Read error: 110 (Connection timed out))
23:02:35 Quit matsl (Remote closed the connection)
23:06:39 Join DangerousDan [0] (
23:11:06***Saving seen data "./dancer.seen"
23:22:39 Quit muesli- (Read error: 113 (No route to host))
23:32:52Bagderphaedrus961: here?
23:37:09 Quit paugh ("Leaving")
23:38:28t0masgood night
23:38:47Bagderif you tell me your last name, I'll add your name to the credits
23:39:03Bagderjust committed your convbdf patch
23:40:19phaedrus961maybe you could look at this one too?
23:40:35phaedrus961it's not really a bug fix, but not a feature either
23:41:54Bagderstill, I'd rather wait with that
23:42:11phaedrus961np, just thought I'd ask
23:42:47 Join mongey [0] (
23:46:18 Quit thegeek (Read error: 104 (Connection reset by peer))
23:46:18 Quit mongey (Client Quit)
23:48:59 Join muesli- [0] (
23:49:03amiconnBagder: What do you think about disabling functions which do exist for completeness, but aren't used by current code
23:49:27Bagderyou have an example?
23:49:40amiconnI mean functions like checkbox(), or the menu functions for moving entries
23:50:00amiconnWe could disable them with #if 0 and leave a comment
23:50:03BagderI'm for disabling them
23:51:39amiconnOn an unrelated note, I already saved >100 bytes in apps/recorder/widgets.c (from a total of ~620) without disabling checkbox()
23:52:14amiconnIt is no longer necessary to check coordinate boundaries with the current lcd code
23:52:20 Join thegeek [0] (
23:52:24amiconn...because current lcd code handles clipping
23:52:56Bagdermultiple checking for the same things is probably present all over
23:55:18amiconn..and iirc the rockbox philosophy is to not check parameters to just protect from programmer errors
23:55:44 Join muesli__ [0] (
23:55:48amiconnParameter checking to protect from user error, corrupt file data etc is of course a different story
23:55:56 Quit muesli__ (Client Quit)
23:57:48 Join einhirn [0] (
23:57:54 Quit ender` (Read error: 113 (No route to host))

Previous day | Next day