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 2005-01-18

00:08:37[IDC]Dragonbuilding the db from cygwin places a /cygdrive/e/ in front of all
00:08:49Zagor[IDC]Dragon: use −−strip
00:09:21***Saving seen data "./dancer.seen"
00:09:51amiconn"No playing implemented yet", hmm.
00:09:58[IDC]DragonI see
00:17:22jypI'm currently replacing ints by longs in fat.c ...
00:17:39jypHave you changed alot in there [IDC]Dragon ?
00:18:39[IDC]Dragonnot after cvs
00:18:58[IDC]Dragonor, not outside of cvs
00:19:17jypIs it in a snapshot ?
00:19:43Bagderjyp: you're not using cvs ?
00:20:05jypI'm using my own repo.
00:20:20amiconnZagor: The perl from the wiki *is* ActiveState perl
00:20:49Zagorreally? ok
00:20:56Bagderjyp: sounds awkward to me
00:21:00[IDC]Dragon−−strip didn't work for me - am I not supposed to name the common part there?
00:21:09Zagor[IDC]Dragon: yes
00:21:11jypBagder: ... working on a local copy is not practicable
00:21:56[IDC]DragonI still have those /cygdrive/e/ in there...
00:22:15[IDC]Dragonperl −−path /cygdrive/e −−strip /cygdrive/e
00:22:23jypI already lost my files trying to do that last week
00:22:45Bagderjyp: you could have a private one to work in plus a common one
00:23:03Bagderthen you'd easier stay up to date and easier do good patches
00:23:20amiconn[IDC]Dragon: Looks correct, only that I didn't use "perl" in front.
00:24:13 Quit Zagor ("Client exiting")
00:24:36[IDC]Dragontrying again...
00:24:41jypThere's the cvs import vendor sources ...
00:24:51[IDC]DragonI should run it on a subdir
00:25:33jyp... but it whines when trying to import directly from another repository
00:27:16BagderI didn't mean 'cvs import'
00:27:23Bagderbut I won't tell you what to do
00:27:28Bagderit's your call
00:28:16jypI'm actually intersted in advices...
00:28:25[IDC]Dragonjyp: much easier with a "moving target"
00:28:59[IDC]Dragonyou can automatically update your sources with our changes, so stay up to date
00:29:51[IDC]Dragonand then make a patch from it which diffs against the current version automagically
00:29:57jypYes, but I need to have a working repository
00:30:18[IDC]Dragononce check out our source tree, yes
00:30:21jypfor saving work & sharing it
00:30:40BagderI think jyp means that he has too many too big changes to just do 'cvs up'
00:30:55jypWhat I really need is a distributed source control system
00:30:58Bagder... which in itself is the problem
00:31:07[IDC]Dragonah, and they want to share that branch
00:31:08jypa la dracs
00:31:32Bagderor just use a branch
00:32:01Bagderfor experimenting, then merging to HEAD for deploy
00:32:06jypYup, that would solve the problem
00:32:23amiconnID3 browsing also works on the player; only the files get the wrong icon (they also get the folder icon)
00:32:53[IDC]Dragonhey, my first working db :-)
00:34:07[IDC]Dragonthe track display is nicer, with no numbers in front of it
00:35:08 Quit elinenbe (Read error: 110 (Connection timed out))
00:35:20[IDC]Dragontoo bad Zagor jumped off
00:35:41amiconnI wonder how to build a db for the Ondio. Getting this right would need more than 1 db file...
00:36:33[IDC]Dragonhaha, a db for a flash player
00:36:59[IDC]Dragoninsert card #35, hit play
00:38:07amiconnWell, with upcoming 2 GB cards...
00:40:00jypre. the db format, etc...
00:40:33jypArchos has its own format for gminis
00:41:25jyp(I'm not using it, just mentioning)
00:41:38jyp... and there's this project ...
00:41:59jypto generate compatible dbs
00:42:11jyp(never tryied it)
00:42:37Bagderyou'll soon be able to use the Rockbox one! ;-)
00:42:58jypHopefully ,)
00:43:15jypThere's still alot to do though ;)
00:43:31Bagderhow's the ATA going? does it work?
00:43:42jypATA seems to work perfect
00:44:11jypthe int size in fat is now the problem
00:44:25jypso disk is not so accessible
00:46:01[IDC]Dragonnow I got a full db
00:46:33[IDC]Dragonscrolling is (way) slower than the key repeat, causing a lag
00:48:14[IDC]Dragonnow I see all the double interprets, cause by different spelling
00:48:46BagderI can see how that will be a FAQ very quickly
00:50:16[IDC]Dragonyes, I "smell" lots of user issues
00:52:08 Join elinenbe [0] (
00:53:57*[IDC]Dragon says goodnight
00:54:04 Quit [IDC]Dragon ()
01:08:38 Quit ze ("damn hardware its needed to be powered off to work on")
01:10:52*jyp waves
01:10:53 Quit jyp ("poof!")
01:25:45 Join Y2e8n7 [0] (
01:27:50 Part Y2e8n7
01:28:59 Quit elinenbe (Read error: 104 (Connection reset by peer))
01:43:29 Nick Spida_ is now known as Spida (
01:57:50 Part amiconn
02:09:22***Saving seen data "./dancer.seen"
02:14:06 Quit edx (Read error: 60 (Operation timed out))
02:41:42 Join ze [0] (
02:44:09 Quit mecraw_ ("Trillian (")
02:44:12 Quit Stryke` (Read error: 110 (Connection timed out))
03:44:13 Quit lImbus (Remote closed the connection)
03:44:28 Join ashridah [0] (
03:57:58 Join edx [0] (
04:09:25***Saving seen data "./dancer.seen"
05:58:52 Join bagawk [0] (~lee@bagawk.user)
06:09:29***No seen item changed, no save performed.
06:15:50 Quit bagawk ("Leaving")
06:35:41 Join midk [0] (
07:15:29 Quit midk ("Leaving")
08:09:32***Saving seen data "./dancer.seen"
08:42:22 Join Zagor [242] (
09:45:32 Quit Nibbler (Connection timed out)
10:08:33 Join [IDC]Dragon [0] (
10:09:36***Saving seen data "./dancer.seen"
10:10:32 Join LinusN [0] (
10:18:49[IDC]Dragongood morning!
10:24:49 Join Nibbler [0] (
10:27:24 Join R3nTiL [0] (~zorroz@
10:35:21BagderI mailed jyp and offered him CVS access
10:36:11*LinusN should commit hit iriver ATA patch before jyp
10:36:21LinusNit is major
10:36:31Bagderthen you really should
10:36:39Bagderbtw, check it for int/long issues too
10:36:45 Join amiconn [0] (
10:36:58LinusNwell, it really doesn't matter who commits first
10:37:01Bagderwe should also get a CalmRISC16 gcc installed
10:37:13Bagderso that we can do proper gmini-builds for the cvs table
10:37:27LinusNone of us will have a helluva lot of work to resolve the conflicts
10:37:45amiconnraining ;) (at least in Berlin)
10:37:51[IDC]Dragonthat's why you want to go first ;-)
10:38:01Zagorthe #gmemu topic talks about a 3.4.3 gcc for calmrisc
10:38:04[IDC]Dragonhere too
10:38:19Zagorsounds a lot nicer than that 2.97
10:38:28Bagder2.97 sounds scary to me
10:40:56amiconnZagor: Did you notice - db browsing broke resume :(
10:44:02Zagoryeah, i'll fix that today
10:44:50amiconnThe correct icons for the player was an easy addition
10:45:00[IDC]Dragonamiconn: did you get a _norom player dump?
10:46:28amiconnNot yet; that's why I posted the request in the forum
10:46:46[IDC]Dragonyes, saw that, so no response yet.
10:47:25[IDC]Dragoncan you forward it to me then?
10:49:18amiconnOf course.
10:51:09[IDC]Dragonbefore releasing, I'd like to check it a bit with the simulator
10:53:33 Join amiconn_ [0] (
10:57:16 Quit amiconn (Nick collision from services.)
10:57:17 Nick amiconn_ is now known as amiconn (
10:57:39amiconn[IDC]Dragon: You have an archos simulator?
10:57:56LinusNlauterbach sh simulator, is suppose
11:01:04 Quit R3nTiL ()
11:02:12[IDC]Dragonyes, that one
11:17:29[IDC]Dragonony trace of Stevie-O ?
11:17:51[IDC]Dragonhe seemed to have vanished from this planet
11:18:11Bagderhe vanished
11:18:22Bagderwithout passing on what he had done so far
11:18:24[IDC]Dragonwith his promising simulator :-(
11:24:18[IDC]Dragonemail bounces...
11:25:08 Join quelsaruk [0] (~kvirc@
11:25:12quelsarukgood morning
11:25:33[IDC]DragonI vaguely remember he gave me a different one, but that fell down the 3-month ledge of my freemail account
11:26:15quelsarukhmmm... why deutsch.lang has different strings than english.lang?
11:26:38quelsaruksouldn't all languages have the same strings in the same position than english.lang??
11:27:02Bagderthe order is only important in english.lang
11:28:00[IDC]Dragonwhich doesn't incline you should make a mess from it ;-)
11:28:20quelsarukcan't i?
11:28:35[IDC]Dragongerman was horrible when I picked it up
11:28:45Bagderin fact, uplang _could_ be fixed to output all the translated strings in the same order the english one has them
11:28:51quelsarukand now is even worse?? ;)
11:29:07Bagdertime to go, see ya around
11:29:13[IDC]Dragonhehe, now Jens was at it
11:29:18quelsaruki can say that spanish lang has the same order than english
11:29:37[IDC]DragonI needed some order again, to give it a voice
11:35:46amiconnI didn't touch the order...
11:37:11[IDC]Dragonno, I was joking
11:37:42[IDC]Dragonwhen I first introduced german voice, I had to sort out the file big time
11:52:53 Join ze__ [0] (
12:01:51 Quit ze (Read error: 60 (Operation timed out))
12:01:51 Nick ze__ is now known as ze (
12:09:39***Saving seen data "./dancer.seen"
12:26:46 Join R3nTiL [0] (~zorroz@
12:35:15 Quit R3nTiL ()
12:38:16 Nick quelsaruk is now known as Quel|away (~kvirc@
12:49:25Zagorhmm, i can't repeat the resume problem anymore. can you amiconn?
13:07:23 Join R3nTiL [0] (~zorroz@
13:20:53amiconnZagor: I can't try it atm. Though I could compile, I don't have my A-A cable with me..
13:26:50amiconnYou didn't change anything... so I have current cvs running on my box.
13:27:14amiconnThe resume problem is definitely there, I tried it more than once.
13:28:10amiconnIt stores the resume position on stop. When you hit On, then Play, it tries to resume, but then doesn't start playing. It says "End of playlist", then stops.
13:35:50Zagorworks for me :(
13:37:01amiconnDo you have uncommitted changes?
13:50:46 Join jyp [0] (
14:03:12 Quit R3nTiL ()
14:09:42***Saving seen data "./dancer.seen"
14:26:58Quel|awayby the way
14:27:21Quel|awayshouldn't rockbox spell especial chars?
14:27:35Quel|awaylike spanish **
14:28:02Zagoryou mean in voice mode?
14:29:49Quel|awaya lang_char_ :D
14:29:52Quel|awayand so on
14:30:01Quel|awayif it's going to spell a word
14:30:10Quel|awaymust know how to spell it properly
14:30:12Zagorit's difficult to fit all chars in ram
14:30:18Quel|awayi know
14:30:23Quel|awaytoo many chars
14:30:58Zagori think it should rather fall back and spell the closest ascii char
14:31:00ashridahgod, i really need to fix up my gateway, all i saw then was �
14:31:24ashridahbloody thing has broken unicode
14:31:58*ashridah wanders off muttering vile things about the ineffectiveness of unicode_start
14:34:56dwihnoWhen replacing a heatsink, what do you recommend to get rid of the old paste?
14:35:17LinusNa hammer
14:35:59LinusNor nail polish
14:38:13Quel|awayoh... and goat's blood
14:43:16 Quit jyp ("poof!")
14:45:21ashridahdwihno: light buffing with sandpaper.
14:54:51Zagordwihno: "Thermal pads can be scraped off with a plastic tool that will not scratch the bottom then the remnants can be removed with a xylene based cleaner, (Goof Off and some carburetor cleaners) acetone, mineral spirits, or high-purity isopropyl alcohol." −−
14:57:41dwihnothermal goo!
14:57:44dwihnofree goo!
15:00:13dwihnothanks Zagor
15:00:21dwihnoI owe you kebab, starka ssen p.
15:08:49 Join midk [0] (
15:13:42 Join webguest51 [0] (
15:14:34 Quit webguest51 (Client Quit)
15:19:35 Quit [IDC]Dragon ("CGI:IRC")
15:32:27 Quit ashridah ("sleep")
15:48:24 Join Bruno [0] (
15:48:35BrunoHi all
15:49:46Brunohow is the developement of the boot for iriver ? Progresing?
15:53:28Zagorno news at this moment
15:54:26Zagorwhee, now I can browse infinitely long tables
15:56:18Lynx_Zagor: getting rockbox ready for future storage media? ;)
15:56:38amiconnZagor: I'm already thinking of a db creation plugin...
15:56:52Zagori guessed you would ;)
15:56:59Lynx_amiconn: i'd rather have someone port perl to the box ;)
15:57:12Bagdera perl-plugin!
15:58:30LinusNBruno: everyone is waiting for me to finish the boot loader
15:58:37amiconnI need to find an algorithm (I know that it's possible) to sort a file larger than available memory
15:58:55Bruno^^ Linus
15:58:59LinusNamiconn: google for "merge sort"
15:59:49BrunoHope you'll find out how to make it work :)
16:00:08LinusNthat's not a problem, the problem is finding the time to do it
16:00:41Lynx_LinusN: what, you mean you are still wasting time in bed sleeping? ;)
16:01:01LinusNall night (3hours) :-)
16:01:22dwihnoNo pressure on that guy :)
16:02:13BrunoLol i know that you don't have much time to do it.... real life +sleep time = no time left
16:03:02BrunoBut no preasure i guess you'll be faster than iriver devs ^^
16:06:51crash_LinusN: have you already begon dreaming of the bootloader ? ;)
16:08:44Brunosee you
16:09:06 Quit Bruno ("Chatzilla 0.9.66c [Mozilla rv:1.7.5/20041108]")
16:09:09dwihnoLinusN: This is the moment you say you haven't even started writing the boot loader, or know where to start, for that matter :)
16:09:32LinusNit is started
16:09:43***Saving seen data "./dancer.seen"
16:14:03dwihnothe question that remains, is, who loaded the boot?
16:22:19 Join methangas [0] (
16:29:31LinusNwho put the bomp?
16:31:40Quel|awayBagder: btw, there's also another thing that should be changed in uplang script :)
16:31:43 Nick Quel|away is now known as quelsaruk (~kvirc@
16:33:34quelsarukvoice line is changed with english one if it exists, but if it does not exist (or a new lang_id is added, then the voice line is not added, only desc, eng and new lines are added)
16:33:41 Part LinusN
16:57:16quelsarukZagor, Bagder any of you awake?
16:57:44quelsaruki have one of my fantastic bugs :)
16:58:32quelsaruki completed espanol.lang, tried to binlang it and worked, tested the espanol.lng on the box and... still in english!
16:58:51quelsarukcan i send you the espanol.lang file so you can test?
16:59:00quelsarukmaybe is the size?
16:59:48Zagorno time right now im afraid. i can test later.
16:59:52 Join mecraw_ [0] (~lmarlow@
17:00:07quelsarukok then
17:00:17quelsarukone day more or less is not a proble
17:00:41quelsarukspanish has been outdated for months.. maybe nearly 1 year right now.. or more...
17:07:51Zagorgotta go
17:07:52 Part Zagor
17:09:06amiconnquelsaruk: How many bytes is your new espanol.lng (not .lang!)?
17:10:16quelsarukis quite big compared to english.lng
17:17:07amiconnSo that's the problem then. MAX_LANGUAGE_SIZE is 6500 bytes. It needs to be increased and rockbox recompiled for your new espanol.lng to work
17:17:35quelsaruklet me change that
17:17:37amiconnI once had that with deutsch.lang too
17:17:44quelsarukcheck if works and commit the change then
17:18:00quelsarukmy first commit in... thousand years
17:18:46quelsarukdo you like 7000? or should i just increase it to 6700?
17:24:40amiconnI'd say increase it to 7000
17:35:25quelsarukhmm.. i forgot i have to rebuild the crosscompiler :D
17:37:54amiconnWindows/cygwin or Linux? If cygwin, you can install the updated sh-elf-gcc package provided by Eric Lassauge
17:41:39quelsarukhm.. i always built it, so is ok
17:41:43 Join Spida_ [0] (
17:41:58quelsarukjust have to download binutils and gcc :)
17:42:46 Quit Spida (Nick collision from services.)
17:42:58 Nick Spida_ is now known as Spida (
17:43:15crash_someone there who can tell me if there is a tool out there for getting a password from a rar file?
17:43:19crash_for linux
17:44:45Lynx_crash_: dozens...for windows
17:44:53crash_yeah for windows :/
17:45:17Lynx_crash_: for linux bash and a lot of time would do ;)
17:50:51crash_ok dont need it anymore, there was a hint in the description of the rar _;)
17:53:17crash_'blabla' has no website, blabla was pwd, funny guy who made this
18:07:51 Nick quelsaruk is now known as Quel|away (~kvirc@
18:09:44***Saving seen data "./dancer.seen"
19:32:37 Join preglow [0] (
19:35:31 Join Qrutpomp [0] (
19:54:06 Join Yokalosh [0] (
19:54:29YokaloshHi all
19:55:26YokaloshHmm i need an answer from someone who knows absolutely everything about the recorder 10's hardware and about programming rockbox
19:57:20amiconnI think no one know everything about it, but maybe I can help...
19:58:12Yokalosher the question
19:59:20YokaloshJorg says it is possible to fit a colour lcd onto my recorder 10
19:59:47YokaloshJorg says it is possible to fit a colour lcd to my box
20:00:05Yokaloshyou there?
20:00:29YokaloshI was wondering what the specs of a screen i would have to find would have to be
20:01:26YokaloshCan you help?
20:02:23amiconnWell, you would need a display with the right controller & supply voltage.
20:02:42Yokaloshcould you send me a spec sheet?
20:02:57YokaloshI can give you my email address if u know the full specs
20:03:05preglowmonochrome and colour lcds have the same kind of interface?
20:03:22amiconnI don't know such beasts. You'll have to find out yourself
20:03:45amiconnThe controller must have accessible documentation, because you'll have to rewrite the lcd driver.
20:04:30YokaloshDo you know when Jorg or stenberg get on here?
20:04:36amiconnThe interface must be serial (SPI), because that's the only one that will fit the current interface. All other interfaces would need more control lines than are available
20:05:09YokaloshThat is what Jorg said
20:06:19YokaloshAny other info?
20:06:22amiconnOf course the display needs to be the right size, and you will need white backlight with substantially more brightness as well.
20:06:35Yokaloshi know
20:07:59amiconnNothing more, I'd say. Of course the rockbox adaptions are quite some work, especially if you want to actually profit from having colour.
20:08:28Yokaloshi need to work on that stuff
20:08:34YokaloshI need a good teacher
20:08:59YokaloshI can easily be tought
20:09:04Yokaloshi'm 14
20:09:18Yokaloshits just finding someone who knows what they are talking about
20:09:46***No seen item changed, no save performed.
20:09:49YokaloshLike you..........
20:10:43YokaloshCould you teach me?
20:14:10amiconnWell, I can give tips.. here.. like now,
20:14:20amiconnbut not much more I'm afraid.
20:14:51YokaloshBut i need to be taught to program rockbox
20:15:27preglowdo you know how to program at all?
20:15:47YokaloshEr no
20:15:50preglowthen you need to learn that first, i'm afraid
20:16:04YokaloshWho by? and how?
20:16:18preglowpick up a book on C or something
20:16:36preglowthere's rather a lot to be learnt before you can start programming hardware stuff
20:17:17YokaloshThis sucks.........
20:18:22YokaloshYou know how to program?
20:18:25preglowsure does, but that's life
20:18:46preglowi could probably do what you're describing, but i've also been programming for years and know a thing or two about electronics
20:19:50preglowwithout knowing any of those two, you are pretty much stumped
20:20:25YokaloshI can do the electronics part
20:20:25 Join Domonoky [0] (
20:20:36YokaloshJust not the programming
20:21:05preglowwell, by electronics i really do mean knowing how to program the lcd controller
20:21:27preglowyou've have to be able to read a data sheet
20:21:40Yokaloshi'm learning how to do it at school now
20:22:37YokaloshAny idea of a fitting screen?
20:23:08preglowno idea, no, i've written a lcd driver in my life, but that's all i've had to do with them
20:23:46preglowbut if it uses an spi interface, it shouldn't be that hard
20:25:37Yokaloshi'm currently getting a spec sheet for a gameboy colour screen
20:25:37 Join Tang [0] (
20:25:42Yokaloshi have plenty of them
20:25:52Yokaloshi have plent of them
20:27:21YokaloshHi :D
20:27:30TangLinus or any Rbx memebers is here? :)
20:27:38TangI had a question
20:27:52Tang(about the API wiki discussion)
20:28:00YokaloshAmiconn is helpful
20:28:11Tang(the idea semms wonderful)
20:28:18TangAmiconn is here? :)
20:28:23TangWould be nice
20:28:34amiconnYes I'm here
20:28:41TangHello AMiconn :)
20:28:51TangGoood new year first
20:29:00amiconnHi... you don't need the shift key, it's just amiconn
20:29:06TangDo we want to attempt cross-fading between a 44.1KHz file and a 48KHz file? Is a "gapless" change in playback frequency possible on the iRiver?
20:29:26TangI read this and i didnt' really understood the last question
20:29:37preglowhmm, that could be strange, yes
20:29:41Tangdoes it concerns gapless or another things
20:29:56preglowit concerns gapless, but only between files of different sampling rates
20:30:00preglowwhich you shouldn't see too often
20:30:04preglowmost files are 44.1
20:30:07TangAh okay
20:30:15amiconnWell, that simply means whether changing the playback sample rate will produce a short gap, caused by reinitializing/resynchronizing the dac
20:30:16preglowordinary gapless playback should be no problem
20:30:40Tangokay thanks wthat was just what i was wondering in fact
20:30:52TangAre you Rbx memeber too?
20:30:54amiconnpreglow: I wouldn't count on that. CD uses 44.1 kHz. E.g. DAB uses 48 kHz (and mp2)
20:31:15preglowamiconn: yes, i encode all my vinyls as 48khz as well
20:31:28preglowamiconn: but most music is ripped from cds, so 44.1 is by far most common
20:32:03amiconnBoth crossfading and gapless would be possible if we never change the sample frequency that goes to the dac, but that means a lot of resampling...
20:32:04TangIndeed also 48kHz is not supported for Ogg by iHP i think
20:32:10Tang(not sure however)
20:32:33preglowi'm almost certain it is
20:33:12preglowamiconn: but i really see no point in gapless playback between files of different sampling rates, they clearly don't belong together anyway
20:33:23preglowamiconn: since material that does belong together really should have the same sampling rate
20:33:55amiconnI also see no sense in crossfading, but there are different opinions on that one.
20:34:42preglowagreed there as well
20:34:49preglowi would never use crossfading
20:35:10Quel|away(building the crosscomplier is being a really slow task....)
20:35:15preglowbut ofcourse, if you listen to alot of mixed artist material, you could use it
20:36:24amiconnCrossfading would only make sense if you would also be able to mix for real, i.e. match the beat. I can't imagine how this could be done with only a few buttons...
20:36:45preglowbeat matching on the h120 would be SO cool
20:37:36TangAnother thing...
20:37:37amiconnOf course it might be possible to do this in software, at least with some kinds of music
20:37:59preglowit could be doable, stick up/down for speed up track 1 momentarily, stick right/left for speed up track 2 momentarily
20:38:09preglowpress button, and then same controls for constant pitch change
20:38:14amiconnGrr, I need Zagor! The resume bug is still there, 100% reproducable on all 3 units I have.
20:38:28preglowhell yes, i'm really into drum 'n bass and beat matching is used extensively by djs there
20:38:41TangIn the iRiver port RAM loader section i read someone had his iHP refusing firmware upgrade due to false code
20:39:14TangWho is cocnerned? The matter is still here? :s
20:40:53YokaloshDRUM 'N' BASS KICKS ASS!
20:42:33TangAnyway i've to go
20:42:35preglowbut will the h1x0 units have enough cpu to run two mp3 decoders in parallel ?
20:42:40preglowbye, tang
20:42:46TangBye preglow
20:42:55Yokaloshc ya
20:42:55preglowit'd be neat to make this crossfading thing work, hehe
20:43:15Tang(i stay a little bit this point is important to me (gapless)
20:43:33preglowi don't think we're going to say much more about it ;)
20:43:36Tanghum ok it's just for corssfading
20:43:44Tangsorry i mistaken
20:43:54Tangyou were talking about crossfasing
20:44:03preglowgapless playback is the reason i got interested in rockbox, so sure, i understand your concern, hehe
20:44:08Tangi thought you were talking about gapless
20:44:23preglowi cry every time my h120 makes a gap
20:44:26TangOki i understand
20:44:28preglowmy soul cringes
20:44:36TangCheers preglow
20:44:41TangI've to go
20:44:47preglowsure, see you around
20:44:48Tangbest regards
20:44:58amiconnpreglow: (2 decoders in parallel) Yes I think so, given that it's a 140 MHz coldfire. A 50 MHz 68060 is able to decode mp3 at full quality, with ~50% cpu load (integer arithmetics only).
20:45:18 Quit Tang ("Chatzilla 0.9.66 [Mozilla rv:1.7.5/20041108]")
20:45:42preglowamiconn: yes, and the coldfire should have the dsp caps to decode mp3 more efficiently than a 68k
20:46:08preglowamiconn: i think i read somewhere that motorola has an mp3 decoder for the coldfire that uses about 10% cpu
20:46:14preglowthat's probably a 64kbps file, but...
20:47:20YokaloshSo amiconn what is the current 'must have' box ..and please don't say iPod
20:47:28 Join LinusN [0] (
20:47:32preglowhi, linus
20:47:38LinusNhey ho
20:48:01amiconnhi again
20:48:22amiconnZagor's id3db code has many problems... :(
20:48:35LinusNregarding the ram loader, paul tried to patch the original firmware into loading his code from disk
20:49:27LinusNit failed, and he managed to disable the code that does the normal firmware upgrades, so he can't restore the flash
20:49:34LinusNamiconn: of course it has
20:49:42LinusNit's alpha status
20:50:07amiconnYes, but it also broke other places that worked before.
20:50:21LinusNas the resume, i guess
20:50:45amiconn(1) Resume doesn't work.
20:51:00amiconn(2) After the failed resume, you are stuck in the dir level the failed-to-resume files were in.
20:51:16amiconnYou can go to a lower dir level, but not higher.
20:51:17LinusNhow nice :-)
20:51:44amiconn(3) It is impossible to leave sub-browsers (fonts, plugins etc.)
20:52:32amiconn..hence it is impossible to change font, wps etc. Because you cannot leave the sub-browser, settings are not saved...
20:52:44LinusNhe should return his badge and gun, and leave the code police force
20:52:49amiconn(4) If browsing mode is set to id3 browsing, the sub-browsers use it as well
20:53:18amiconnHe promised he'll fix it today, but then said he cannot reproduce it anymore
20:53:26 Quit Yokalosh ("CGI:IRC")
20:54:05amiconnI tried it on the recorder, the player and the Ondio, recompiled after make clean from current cvs.
20:54:40LinusNbtw, do you know a good place to do the uart boot mod on the player?
20:54:48amiconnI tried the official bleeding edge recorder build as well, to rule out compiler problems.
20:55:00amiconnHow do you mean "place"?
20:55:36LinusNsome good vias or pads to cut and solder
20:55:38amiconnThere is a nice pad area (looking like meant to take a chip) that has all necessary signals.
20:55:52 Quit Domonoky (Read error: 60 (Operation timed out))
20:55:53amiconnHowever, your board revision may be totally different.
20:56:13amiconnI intend to take pictures and write a wiki article when I undo the mod.
20:56:20amiconn(this week)
20:57:28LinusNis it on the cpu board or the display board?
20:57:35amiconnMain board
20:57:48LinusNnext to the isd200?
20:57:52LinusNaha display board
20:58:12amiconnThere is no nice resistor array to cut and pull down instead of up.
20:58:37amiconnYou need to pull down against the pull-up resistors (which are scattered across the board I think)
20:59:04amiconnPull-ups are 10 kOhm, so I pulled down with 1 kOhm
20:59:45amiconnIn , you can clearly see the pad area.
21:00:35amiconnIt's the same the serial wires go to.
21:01:29LinusNi see
21:01:43LinusNthose pics are my player btw
21:01:52amiconn...where left column, 5th pin from top and right column, 4th pin from top are Tx and Rx, the 3 topmost pins in the left column are PB1..PB3
21:02:00amiconn(which you need to pull down)
21:03:00amiconnYou'll need the pull down switchable, releasing it after the uart boot started.
21:03:36amiconnOtherwise the CPU may not be able to drive the pins high across the strong pulldown, preventing the LCD from working
21:04:51amiconnI connected a 5-wire piece of ribbon cable to the pad area. This conveniently fits, you only have to solder one wire to the other side.
21:05:05 Join Stryke` [0] (
21:05:11amiconnFirst 3 go to left column, 4th goes right and 5th left again
21:05:59LinusNwell, the serial mod is already there, since year 2000
21:07:30LinusNok, will do that later tonight
21:07:51amiconnBe prepared that the lcd may not correctly initialize if you have rockbox flashed and start the box by inserting batteries.
21:07:59amiconnThis isn't rockbox' fault; it also happens with pure archos flash contents if you have the ON button pressed while inserting them.
21:08:31LinusNold players don't start when inserting the batteries
21:09:06amiconnNew players also don't do this with archos firmware if you don't hold ON.
21:09:36amiconnThe archos firmware checks button status on boot, and immediately shuts down if ON is not pressed
21:09:43LinusNoops, i was wrong
21:10:04LinusNi meant that it doesn't start when inserting the charger
21:11:26LinusNgotta go, cu around
21:11:38 Part LinusN
21:20:34 Join Zagor [242] (
21:21:27amiconnhi Zagor
21:22:16amiconnDid you check the logs? Many problems with id3db code :(
21:22:19Zagor"Since 22 December, mail servers at have been configured not to accept connections from Europe by default." <−−- explains all the bounces we get
21:22:52Quel|awayi wonder why
21:22:58 Nick Quel|away is now known as quelsaruk (~kvirc@
21:23:03Zagor"to stop spam"
21:23:17ZagorVerizon three million DSL customers waiting for emails from Europe were advised to use alternative forms of communication. "If it's really important you might want to make a phone call," he said.
21:23:44Zagorgotta love friendly ISPs :)
21:24:14amiconn...especially those that advise you not to uses it's services, haha
21:24:27preglowverizon are assholes, no surprises there
21:24:56 Quit methangas (" HydraIRC -> <- *I* use it, so it must be good!")
21:26:50preglowthe fact that they actually went through with such a braindead approach, speaks volumes about the management
21:32:16Zagoramiconn: i still can't repeat it. :-( just tried a clean build and install. can you help me?
21:33:10amiconnDid you try a build from the site?
21:33:23Zagorno. i'll do that.
21:33:29amiconnI can only try to describe what happens as detailed as possible.
21:33:41amiconnActually there are a number of problems.
21:33:50Zagoryeah, but only the resume is important
21:34:07Zagorthe rest only occurs when you use id3 mode, which isn't ready for use anyway
21:34:16amiconnI think the unability to leave any sub-browser is also important
21:34:31amiconn(This happens in any mode, not just id3 browsing!)
21:34:52Zagorok, i don't get that either... weird, this
21:36:03amiconnI tried with fresh builds from current cvs, with only 2 little changes in (totally unrelated to browsing - my alternate recording routines on the recorder, and enabled ata poweroff for the player)
21:36:24amiconnI tried official builds as well. The behaviour is identical on all 3 units.
21:36:28Zagorwhich gcc are you using?
21:36:33amiconngcc 3.3.5
21:37:04amiconn(Eric's cygwin build)
21:37:13Zagori've tried builds with 3.0.4 and 3.3.4
21:37:29amiconnIs your box flashed? If yes: rombox?
21:37:47Zagoryes, no
21:37:58amiconn(But then rombox can't be the problem, because this also happens with roloed builds)
21:38:24amiconnWhat I tried and what happended, in detail:
21:38:38Zagorcan you make it happen in the simulator?
21:39:03amiconn- Resume = ask, Shuffle = no, Repeat = off
21:39:31amiconn- Go to a sub-dir and play a file
21:39:41amiconn- Hit stop.
21:40:28amiconnI'm now back in the dir where the tracks are.
21:40:45amiconn- Hit ON, then PLAY to resume.
21:41:21amiconnIt immediately displays "End of playlist". I'm back in the dir the tracks are in
21:41:46amiconnIt's now *impossible* to go up even one dir level.
21:41:54Zagorbleeding works for me as well..
21:42:25amiconnHuh? How can this be?
21:42:35Zagori don't know. very strange.
21:42:47Zagorcan you make it happen in the simulator?
21:42:56*amiconn tries
21:43:30amiconnbuilding simulator...
21:47:10 Join HCl [0] (
21:47:52quelsarukZagor: i get the same that amiconn
21:47:56amiconnHmm, really strange. It works in the sim
21:47:57quelsarukand my brother was going to kill me
21:48:14HCli don't suppose there's anything i can do to help speed up iriver development?
21:48:16quelsaruk'cause i was testing the new spanish.lang
21:48:23ZagorHCl: cheer? ;)
21:48:26HCl xD
21:48:29HClyea, i'll do that :P
21:49:16HCljust wait and pray then :X
21:49:21Zagorare you a programmer, hcl?
21:49:28HCli think its safe to say i am o.o
21:49:56HCla bored one, i might say
21:51:55quelsarukand he also got another bug using bleeding edge
21:52:09quelsarukchanging font *hanged* the browser
21:52:27quelsarukcould not move back to the menu, or to another screen
21:53:20amiconnquelsaruk: It's the same as I get. The browser is not hung, only it is impossible to leave the browser
21:54:33amiconnResume on the sim has problems too.
21:54:59amiconnWhile the sim is running, it works, but when I "power off" and then start again, it seems to resume, but doesn't display any tag in the wps.
21:55:34amiconnIt says it plays a track from the root dir, which is also wrong.
21:55:51quelsarukthat's what i have just realized
21:56:15quelsaruki've changed the max_lang_size... to 7000 so i can load espanol.lng
21:56:22quelsarukZagor: should i commit that change?
21:56:53Zagorquelsaruk: write shorter :)
21:56:56amiconnTrying X11 sim to double-check...
21:57:29quelsarukZagor: i'd do, but spanish guys wont understand
21:57:54quelsarukamiconn: i have just built a cvs version and happens the same
21:58:18Zagoramiconn: great, i can repeat that too. i'll dig into it.
21:58:26quelsarukbut, resuming after poweroff works
21:59:10amiconnNow _that_ is strange. 3 guys compile from the same source, and all get different behaviour :/
21:59:29Zagorsounds like unitialized data
21:59:41quelsarukamiconn: oh.. you don't know the legend of the digital goats...
22:00:49amiconnZagor: Sa0e behav564r w5th the X11 s50,
22:00:58amiconnOoops, still numlocked :)
22:01:07quelsaruka new language
22:01:12amiconnI mean, same behaviour with the X11 sim
22:07:38amiconnZagor: Leaving sub-browsers is also impossible in both sims
22:08:45quelsarukthen, someone changed something in the browsing
22:09:47quelsarukdinner time
22:09:50***Saving seen data "./dancer.seen"
22:20:57amiconnI just received a _norom player flash dump :)
22:57:07 Nick quelsaruk is now known as quel|nwn (~kvirc@
23:13:36Zagorgah, i hate bugs that makes you ask "how did this ever work?"
23:14:13quel|nwndid you repair the browser back button bug?
23:14:32Zagorfound the resume issue, and the same thing is causing the back bug
23:14:53quel|nwntsk.. tsk..
23:15:10quel|nwnyou are not the great code police i thought
23:15:12amiconnZagor: do you have an idea why this worked for you?
23:15:18quel|nwnyou disapointed me :D
23:15:22Zagoramiconn: no
23:15:43Zagorquel|nwn: :)
23:16:53amiconnZagor: Do you remember the language array issue? I have an idea how to solve this, and save almost 1500 bytes of binary size.
23:17:08Zagorremind me
23:17:49amiconnThe language is internally stored as an array of pointers to strings, which initially point to the builtin strings.
23:18:07amiconnSo these pointers are initialized data, and take up space in the binary.
23:18:45amiconnWhen you load a complete language file, these pointers do now all point to the loaded strings.
23:19:20amiconnIf you then load another language which is *not* complete, the last few pointers still point to the old positions.
23:20:13amiconnIf the newly loaded language also has longer strings than the old one, the last few may pointers end up pointing mid-string anywhere.
23:20:55amiconnMy initial idea was to make the builtin pointer array read-only, and copy it every time before loading a language.
23:21:36amiconnHowever, this would not decrease binary size, and would even need more ram (unless you use rombox, where ram usage would be identical)
23:22:34amiconnMy new idea is to leave out the initial pointer array from the binary completely, only leaving a pointer to the start of the strings.
23:23:23amiconnThe strings would then be separated by *single* zero-bytes, and the pointer array gets calculated by "walking that long, \0-separated string.
23:23:26 Join jyp [0] (
23:23:49Bagderamiconn: just make sure you can handle gaps
23:23:49amiconnThis would not need more ram, and the left-out pointers make the binary smaller.
23:24:12Bagderbut I like the idea
23:24:17amiconnBagder: Well, if I put a *singe* zero byte at each string end, this is easy.
23:24:46BagderI just wanted to mention it
23:24:51amiconnI would actually output this as one string (from the binlang script):
23:25:37amiconnInstead of writing "foo", "bar", "bas" simply output "foo\0" "bar\0" "bas\0" and let the compiler concatenate
23:25:47Bagderno need for the \0
23:26:39amiconnIf I leave it as an array, the compiler may decide to pad the strings, and I may end up with more than exactly one zero byte...
23:27:21amiconnBtw: The same trick can be applied to the credits list.
23:27:53amiconnThere the pointer array is even not needed at all, because the list is read in sequence
23:28:02Bagderjyp now officially has cvs commit access
23:28:22amiconnOk, good :)
23:29:06*jyp bounces
23:29:24Bagderjyp: is there any docs anywhere on what gcc/binutils needed to build?
23:29:30*jyp shakes hands w/ everyone
23:29:38Bagderwould be neat to get it added to the cvs build table
23:29:56jypThat's a pretty complicated story
23:30:09amiconnWe need to define at least the basic buttons if we want a simulator soon...
23:30:21amiconn(gmini simulator that is)
23:31:01jypre gcc: Because right now we use an old version of the binutils
23:31:32jypwhich is included in a full development suite
23:31:43jypbut the gcc version in there does not work correctly
23:32:02Bagder"full development suite" ?
23:32:20jypIt resides somewhere on
23:32:24Bagderisn't it just gcc and binutils?
23:32:44jypThere's a bunch of other utilites there
23:32:58jypgdb, sim...
23:33:11jypthere's even a version of tcl/tk/etc for calmrisc!
23:33:28jypI was planning to release a stripped down version
23:33:42Bagdersee this:
23:34:15Bagdercan't that be done for CalmRISC too?
23:34:26BagderI don't want a development kit
23:34:49Bagderor rather gcc and bintuils is that
23:35:03jypI can be done more or less
23:35:53Bagderto me, that is the simplest approach
23:36:31jypthe problem is that you'll need custom versions of each package
23:36:39Bagdercustom how?
23:37:18jypWell, fsf gcc/gas do not support calmrisc16...
23:37:30 Join [IDC]Dragon [0] (
23:37:44jypYou need to "special" version I'm maintaining
23:37:45Bagderisn't the patches required available then?
23:38:09jypThey are available but sort of unstable
23:38:16amiconnhi [IDC]Dragon
23:38:32[IDC]Dragonhi guys
23:38:35Bagderjyp: well still, if that is the only compiler we have, that is the path we walk
23:39:07amiconn[IDC]Dragon: I got a player _norom dump (rom 5.08, archos 5.08). Interested?
23:39:24jypAre you ok with checking out gcc code from sourceforge cvs ?
23:39:57Bagderyou have the whole gcc code there?
23:40:01amiconn[IDC]Dragon:i I also built a firmware_player_norom.bin, but I'd rather not distribute it with current cvs rockbox due to the browsing bugs.
23:40:07Bagderbut sure
23:40:14jypBagder: all we need
23:40:31BagderI wouldn't have done it that way, that's just it
23:40:38jyp(but not yet binutils)
23:40:40[IDC]Dragonamiconn: how about using 2.4?
23:40:54amiconnHmm, possible.
23:41:02[IDC]Dragon(dump) sure, bring it on!
23:41:14jypBagder: I know this is awkward... But we' on the "bleeding edge" here
23:41:19*[IDC]Dragon reads a lot of log
23:41:34Bagderjyp: I know about that, I would still not do it like that
23:41:40BagderI would provide patches against official releases
23:42:08Bagderso then we'd download release XXX and your patch YYY
23:42:23ZagorBagder: well there's nothing stopping us from doing that now
23:42:26Bagderbut I'm not complaining
23:42:31BagderZagor: very true
23:42:36BagderI can check out and make a patch
23:42:50Bagderif I know from what release it is branched
23:44:03jypWell you can do a patch with a diff from the vendor tag
23:44:30amiconnBagder: How can I see which cvs commits took place between the 2.4 release snapshot and 2004-12-25, 09:00 ?
23:44:40amiconn(if there were any)
23:44:45Bagderuse cvs log -d
23:44:56Bagder-d "fromdate<todate" I believe
23:45:03Bagderor cvs diff even
23:46:51amiconnHow do I do this between a version and a date? -d takes 2 dates iiuc
23:47:25Bagderget the date for the version first
23:49:21*[IDC]Dragon is done reading logs, a lot of fun here
23:49:52[IDC]DragonLinus did solder a new flash in his player?
23:50:13amiconnNot yet, but he intended to do it later this evening
23:50:20[IDC]DragonYokalosh learns everything... ;-)
23:50:36amiconn(or at least he wanted to uart boot the player)
23:50:57[IDC]DragonI regret telling him about the theoretical chance to change the LCD
23:51:10quel|nwnZagor: did you repair that bug already?
23:51:22 Nick quel|nwn is now known as quelsaruk (~kvirc@
23:51:52Zagoryes, resume is fixed. the back-in-plugin-browser remains still.
23:52:43amiconnI don't get this cvs command right :(
23:53:14Bagderwhat dates is it?
23:54:17amiconnAs I told above... between v2_4 and 2004-12-25, 09:00
23:54:30Bagderso when was 2.4?
23:54:35amiconn(I don't think there are many)
23:56:54Bagdercvs diff -D "2004-12-23" -D "2004-12-25 09:00"
23:58:12amiconnThanks, that worked :) Does that work the same with cvs log ?
23:58:37Bagdercvs log -d "2004-12-23<2004-12-25"
23:59:01Bagderwasn't that obvious? ;-)

Previous day | Next day