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 2003-07-10

00:00:14 Join paulheu [0] (
00:00:37 Quit paulheu (Client Quit)
00:20:24 Quit tpelliott (Read error: 54 (Connection reset by peer))
00:35:39 Join diddystar5 [0] (
00:35:55diddystar5Zagor: hey
00:36:35diddystar5Zagor: why is USE_GAMES in bounce.c never used?
00:37:05Zagorit's removed
00:37:31Zagorthe USE_GAMES macro is deprecated since all games are moved to plugins and thus don't take up extra space
00:37:46diddystar5Zagor: sorry i meant USE_CLOCK
00:38:15Zagorok. that's because it talks to the rtc, which is not supported by the plugin api
00:38:31diddystar5ok i see
00:39:33 Join tpelliott [0] (telliott@
00:39:58MTdiddystar5: you'll be pleased to know ive finished my wormlet patch, its on the patch tracker and Zagor is going to try and add it to main source sometime this week hopefully
00:40:30diddystar5oooh thats cool MT
00:40:50diddystar5i'm gonna give it a try now
00:41:58 Join BoD[] [0] (
00:42:23diddystar5hey there BoD[]
00:44:28diddystar5MT: have you tried to flash your box yet?
00:44:42diddystar5i mean the other flash ;)
00:46:26MTno, and i doubt i will
00:46:32MTboots fast enough for me atm
00:46:41*Stevie-O tries flashing with his digital camera flash
00:47:24diddystar5i gave it a try work very well and VERY fast
00:47:55 Quit mecraw ("Trillian (")
00:48:06 Part Dario
00:48:13MTnow zagor has fixed that bug i may give it a try
00:49:34diddystar5i got it to work without that bug fix
00:50:51 Quit diddystar5 ("Gotta Go!")
00:55:26 Join diddystar5 [0] (
00:57:23diddystar5MT: your patch does not work while running in flash, i had to do a normal boot
00:57:43MTsure it does
00:57:52MTyou just havent flashed the correct image
00:57:58BoD[]what is this flash thing
00:58:07MTas to accomodate those menus, the plugin api had to be changed
00:58:23MTso they wont work with default firmwares
00:58:38ZagorBoD[]: you can flash rockbox instead of the archos firmware
00:58:44diddystar5oooo i see
00:58:53BoD[]wow ... it works ?
00:59:03BoD[]:) does it mean it boots faster ?
00:59:06diddystar51 sec and ill change my flash ver
00:59:11Zagora lot faster
00:59:18diddystar54 secs for me bod
00:59:28BoD[]how do you flash it ?
00:59:41MTdiddystar5: it wont work unless you flash a compressed version of the firmware that has been compiled with my patch applied
00:59:49ZagorBoD[]: using a rockbox plugin
01:00:01Zagorcheck out Jörgs video to see the speed:
01:00:11diddystar5i know MT im making one right now
01:00:14BoD[]ok I'll watch that
01:00:39BoD[]I didn't try rockbox since there is the plugin system
01:01:16Zagoryou've missed a bit then :)
01:01:20BoD[]wow is it a divx
01:01:26BoD[]I need to install divx codecs
01:01:44Zagoryes it is
01:01:50BoD[]ok wait
01:02:46 Quit diddystar5 ("Gotta Go!")
01:02:55BoD[]while it's downloading ...
01:03:16BoD[]so is the "flash" plugin in the default rockbox download ?
01:03:37Zagorno, it's only experimental right now. it will be polished and released in the next few days
01:03:56BoD[]well maybe I can wait till then :)
01:05:00BoD[]wow :)))))))))
01:05:10BoD[]ok I've seen the video
01:05:16BoD[]and want to try it now :)))
01:05:28MTid wait tbh
01:05:50MTi didnt like the look of
01:06:24Zagorheh, that's how it goes when you become too eager :)
01:06:40BoD[]ok I'll wait then
01:06:54MTespecially fi you have an FM :)
01:07:00BoD[]I have a recorder
01:10:16BoD[]anyway that's REALLY great
01:10:39MTthe reasons i wont do it
01:10:49MTyou've only got 100 flashes in that beast
01:11:02Zagorno, you've got 100000 flashes
01:11:13Stevie-Oi thought it was 100k
01:11:29MTim sure Dragon said 100 on the mailing list
01:11:36MTthat changes my perspective
01:11:40MTflash me up
01:11:48Stevie-Odon't tell me you thought that that was a decimal separator
01:12:03Stevie-OI expect a comma there, but please, some common sense
01:12:06MTim british, shoot me now :/
01:12:21MTi shall endeavour to hide furtively now
01:12:39MT100 scared me a lot
01:12:48MTi make about that many a week
01:12:49*Stevie-O makes a new flash chip that can be flashed exactly 100.001 times (that's 1.00001x10^2)
01:13:21Stevie-Olol, I'm used to flash chips that are about a million write cycles
01:13:23Zagori still urge you all to be patient for another day or so until we make this more official
01:13:35BoD[]i will wait
01:15:31 Quit tpelliott ("Trillian (")
01:15:37BoD[]would it be feasible &/| easy to write a plugin that just make the light blink ... in rithm with the music ?
01:16:04Zagorwe can't extract the rythm, since we only see mp3 data
01:16:05MThave you been talking to tracktheripper?
01:16:39BoD[]well when I say "in rhythm" I mean just something like the vumeters
01:17:10Zagorwell yeah that would be possible. very strange-looking, but possible.
01:17:17 Join tracktheripper [0] (
01:17:24*Stevie-O tries some mozilla extensions
01:18:11BoD[]possible but easy ? :) I would like to write it but of course never wrote something for rockbox
01:18:35Stevie-Oif it was easy someone would have done it
01:19:02BoD[]stevie : maybe nobody has the need for something that would make the light blink ;)
01:19:11Zagorit would just blink apparently at random unless your music only contains very a loud bass drum
01:19:23BoD[]I assume it would discharge the battery fast?
01:19:23tracktheripperhi zagor stevie and bod
01:19:28BoD[]hi track!
01:19:35tracktheripperwhats up?
01:19:40tracktheripperdid I miss anything?
01:21:00tracktheripperwhat does ISO stand for? Ive seen ISO used a lot for computer and electronic terms
01:21:13BoD[]i guess international standard organism
01:21:19BoD[]or zation
01:21:46tracktheripperwell i thought ISO meant international standards organisation but I wasn't sure if that was correct
01:21:58BoD[]International Organization for Standardization
01:22:06BoD[]== IOS :) so that's stupid
01:22:20tracktheripperits ISO not IOS :)
01:22:43BoD[]I know but that's what they say on their site
01:22:47tracktheripperyea you are right
01:22:51tracktheripperive just seen it myself :-)
01:23:29BoD[]Because "International Organization for Standardization" would have different abbreviations in different languages ("IOS" in English, "OIN" in French for Organisation internationale de normalisation), it was decided at the outset to use a word derived from the Greek isos, meaning "equal". Therefore, whatever the country, whatever the language, the short form of the organization's name is always ISO.
01:24:15tracktheripperBoD What do u think of simulating the Player's LCD on the Recorder?
01:24:35BoD[]track : what for
01:24:53tracktheripperlarger icons, larger font, slightly less cluttered
01:25:29BoD[]you can just choose a larger font on the recorder !
01:25:48tracktheripper((never thought of that!))
01:25:56BoD[]I'm not sure about the icons though ? are they special characters in the font ?
01:26:12Zagorno, they are fixed
01:26:13tracktheripperthey are hard-wired into the Player's LCD
01:26:42BoD[]zagor : ok
01:26:52BoD[]track: I was talking about the recorder
01:27:06tracktheripperoh ok
01:27:26tracktheripperIve never used a Player so I cannot comment on its interface
01:27:32Zagori'm off to bed. night all.
01:27:40tracktheripperZagor how does the player's LCD compare to the Recorders?
01:27:48 Quit Zagor ("Client exiting")
01:28:07BoD[]well I'm sure they are crap :)
01:28:43tracktheripperthat feature request was cool "Invert the Player's LCD" :-)
01:29:02BoD[]did they answer it's not feasible ?
01:29:26Guest57hi all
01:29:34tracktheripperhi guest
01:29:40Guest57OIs there some news about the Falshale firmware
01:30:24BoD[]you mean flashable ?
01:30:48Stevie-Ono, the falshale
01:30:52Stevie-Ofrom Falshale, NJ
01:32:35Guest57no flashable , sorry
01:33:39BoD[]I don't know if it's news but you can look at this :
01:34:50tracktherippernight all
01:34:52 Quit tracktheripper ("Leaving")
01:39:01 Quit Schnueff (
01:39:01 Quit ricII|work (
01:41:47 Join Guest [0] (
01:42:54 Quit Guest (Client Quit)
01:48:04NJoinSchnueff [0] (
01:50:45Guest57has someone a .ucl file for the recorder that I could flash to aqvoid the ATA:-11 problem and get it to boot into rockbox in 3-4 secs .... like the others luky guys ... thanks
01:52:47NJoinricII|work [0] (
01:56:30***Saving seen data "./dancer.seen"
02:03:25 Join tpelliott [0] (telliott@
02:23:53 Quit tpelliott ("Trillian (")
02:27:17 Quit sobepmp (""Erection Reset by Beer"")
02:32:53 Join tpelliott [0] (telliott@
02:32:58*Stevie-O yawns
02:36:39tpelliottI just HAD to see if my Archos would flash Rombox. It did and I got the expected result "ATA error: -11" but I do get the safe ON+F1 boot so I'm happy it works.
02:37:31Stevie-Owhat's this ON+F1 boot thing?
02:38:08tpelliottIt's safe mode. Boots to the original Archos firmware.
02:38:48*Stevie-O would hardly consider that 'safe'... ;)
02:39:10tpelliottIt's better than a doorstop :)
02:39:57Stevie-Oonly by a hair :D
02:40:03Stevie-Oata error -11 eh?
02:40:52tpelliottYes. The flashed version of Rockbox still has a bug. Read thye mailing list.
02:43:44Stevie-Othis is strange
02:45:30Stevie-Ois the source tree that spawned the rockbox-flash code available?
03:00:42Stevie-Ohey tpelliott
03:00:44Stevie-Oyou here?
03:01:24tpelliottI'm here but I'm not a programmer.
03:01:33Stevie-Othat's quite alright
03:01:38Stevie-Odoes it say
03:02:00Stevie-OATA error: -11
03:02:06Stevie-OPress ON to debug
03:02:22Stevie-Osomething's not right
03:02:38Stevie-Osee, -11 isn't a valid value for us to be getting
03:02:54Stevie-Oi assume you have a plain old Recorder?
03:02:55tpelliottYou can still get it working by holding F1 when you turn it on.
03:03:16 Join earHurts [0] (
03:03:42tpelliottYes. I don't want the FM since they took of the digital I/O
03:03:46Stevie-Oone valid value IS -1
03:03:52earHurtsHow are you gentlemen !!
03:04:24earHurtsthat and the battery, I'm beginning to regret getting the fm
03:04:25BoD[]Somebody have set us up the bomb
03:04:30*Stevie-O thwaps td
03:04:37earHurtsbut it is smaller and lighter & better looking
03:04:38tpelliottI'm ok. Most of the others must be off programming.
03:04:55earHurtsMake your time
03:04:57Stevie-Ohas anybody here looked at the ata -11 issue?
03:05:20Stevie-Owhat I don't understand is that -11 isn't a valid value
03:05:22earHurtsthe issue with the rombox?
03:05:35tpelliottJoerg is off working on that now.
03:05:37Stevie-Ounless our snprintf is buggy
03:05:49Stevie-Ojoerg is probably asleep, it's 3am there
03:05:57earHurtsis (unsigned) -11?
03:06:06Stevie-OearHurts: no
03:06:26tpelliottok. I wasn't sure. It would be best to wait for the all-clear from him.
03:06:29earHurtsour printf rountines don't undeerstand shorts
03:06:40Stevie-Oit's not a short
03:06:42earHurtsor longs
03:06:48Stevie-Oit's not a long
03:06:53Stevie-Owheres the snprintf definition?
03:07:01BoD[]-11 == 245 right ?
03:07:18Stevie-Othat's the prototype
03:07:20earHurtsif it's a char
03:07:20Stevie-Owheres the soruce?
03:07:23Stevie-OBoD[]: only for bytes
03:07:26Stevie-Othis is an int
03:07:30earHurtsoh firmware/something
03:07:51earHurtsI'm not at my computer
03:08:09tpelliottActually, Joerg has some new files on his site. A couple of .ucl files yuou could burn as the secong copy.
03:08:15BoD[]then it's 65525 ? :)
03:08:40Stevie-Oour ints are 32 bits
03:08:54Stevie-Oso try 4294967285
03:08:59earHurtsthose .ucl are really britney spheres mpegs he's hiding from the riaa
03:09:01Stevie-Owhich is definitely not returned by ata_init
03:09:22Stevie-Oval = sign = -1
03:09:28Stevie-Oval = 1
03:09:45Stevie-O*−−str = '1'
03:09:51Stevie-Oval = 0
03:10:22tpelliottHave you guys done the rombox flashing yet?
03:10:30earHurtsnot me
03:10:35earHurtsI have an fm
03:10:55earHurtsit's tremedous work, i can't wait to try
03:11:35earHurtsdid anybody watch that stupid reality show last night?
03:11:44tpelliottI will probably try the .ucl files since my box is working in safe mode. Safe mode should still work.
03:11:48BoD[]what channel?
03:11:48earHurtsdid she pick the guy or the money?
03:12:01earHurtsuh, nbc?
03:12:07earHurtssafe mode??
03:12:39BoD[]anyway I have to go to bed:)
03:12:42BoD[]see you all!
03:12:45tpelliottturn on your box holding F1
03:13:01 Quit BoD[] ("dodozitude")
03:13:03earHurtsoh rombox safe mode
03:14:43tpelliottYes. Another reason I'm glad I have the older recorder, plus, if the batteries go dead, I can put in a spare set.
03:15:15earHurtsyeah. just taunt me ok? ;)
03:15:35tpelliottI can't believe they took out the digital I/O on the FM
03:15:54earHurtsthey didn't take it out
03:16:07earHurtsof the advertising & specs
03:16:07tpelliottSomeone told me they did.
03:16:33earHurtsi might not have bought it if they had
03:16:54earHurtsi remeber when i tried it
03:17:02earHurtsand got nothing
03:17:11tpelliottYou can still play your FM recorder through your reciever's Toslink connection? Cool.
03:18:22tpelliottI play mine through my Sound Blaster Extigy. It works even if the computer is turned off.
03:18:24Stevie-Otpelliott: you still have RB flashed?
03:18:56earHurtsi just play my mp3s on my pc
03:19:08Stevie-Omy laptop sound is broken :(
03:19:17tpelliottYes, but with the ATA error.
03:19:30Stevie-Otpelliott: but you can still boot the archos fw
03:20:00Stevie-Oif you have the rockbox ajbrec.ajz on there does it load that?
03:20:11tpelliottThey have some nice bookshelf systems with USB input for sound from your computer.
03:20:27tpelliottI can still boot Archos in safe mode.
03:20:40earHurtsbut that means no rolo
03:20:45Stevie-Othat's what I mean
03:20:50Stevie-Owhen you boot the archos in this 'safe mode'
03:21:03Stevie-Odoes the archos fw do its automatic ROLO thing?
03:21:14tpelliottYes. Safe mode roro's Rockbox like normal.
03:21:18earHurtsyeah, what stevie said
03:21:25Stevie-OI need you to do me a favor :)
03:21:31Stevie-Oin the debug menu
03:21:45Stevie-Othere should be an option to 'dump roms'
03:21:45earHurtsso it's still a quicker boot?
03:22:06tpelliottYes...I'm in the debug menu.
03:22:12Stevie-Oselect that option
03:22:18Stevie-Oit will create two files on your thing
03:22:26tpelliottI've done that, per instructions.
03:22:36Stevie-Oper what instructions?
03:22:48earHurtshe he, he said 'thing'
03:23:04Stevie-Oi'm about to make the rest of you hurt too
03:23:50*Stevie-O teaches joerg to wrap his goddamn txt lines
03:24:11earHurtsyeah, I had to d/l his instructions
03:24:30Stevie-Oso you have backups of those internal rom files on your pc?
03:24:38tpelliottI use Ultra Edit, which has a wrap icon.
03:24:51tpelliottThe large one.
03:24:57Stevie-Oi use Opera
03:25:08Stevie-Owhich doesn't have such a button
03:25:33Stevie-OI need you to use the debug menu to generate a *new* set of rom files
03:26:02Stevie-Othen I need you to zip both up and mail them to me
03:26:15tpelliottIt should be the same as in ""
03:26:38Stevie-Obecause those rom files are from your flash chip
03:27:02Stevie-Ounless you know something I don't
03:27:04earHurtshe means joerg's eelease
03:27:05tpelliottIsn't that what I just put on my flash chip?
03:27:29Stevie-Oisnt it a plugin and stuff?
03:27:53tpelliottThe plugin just loads the .bin.
03:27:53Stevie-Oi need the actual contents of your flash rom, as-is
03:27:53Stevie-OI don't have a Recorder
03:28:29earHurtsthey're probably the same. but give stevie what's there
03:28:46Stevie-OearHurts: they can't be the same, if it can boot the original archos fw
03:29:18tpelliottThat's built in to Joerg's .bin file (Safe mode).
03:29:46Stevie-Obut not everybody has the same archos fw
03:29:58tpelliottI'm going to try one of the .ucl files first.
03:30:10tpelliottSo you want my original?
03:30:19Stevie-OI need the current contents of your flash ROM
03:30:38Stevie-Oas the dump command would generate them right now
03:31:23Stevie-Obrb, i gotta find my laptop charger
03:31:55tpelliottI'll be back in a few minutes.
03:32:43 Quit Stevie-O (Read error: 54 (Connection reset by peer))
03:32:49 Join Stevie-O [0] (
03:34:26Stevie-Owho mythed me
03:34:59earHurtsyou're legendary, but I didn't myth you
03:36:17earHurtsstevie, what do you hope to find in tpe's dumps?
03:36:57Stevie-Osome information
03:37:27earHurtsthqank you, mr. obvious
03:38:16Stevie-Ohappy to help
03:38:58earHurtsi thought you might want to convert them to .bmps, print them out, and decorate your walls with them
03:39:07Stevie-Otheres an idea
03:39:56*Stevie-O wonders what tpelliott is doing
03:47:31Stevie-Othis would be a record amount of time for the dump_rom to complete?
03:48:22earHurtshe's hand verifying each bit is 1 or 0
03:48:29Stevie-Ouh huh
03:48:34Stevie-Ohow is he hand verifying it?
03:49:25earHurtsyou asked for it:
03:49:27Stevie-Ohe got a microscope?
03:49:28earHurtsby hand
03:49:44Stevie-Oan electron microscope
03:51:07MTwhats special about his rom?
03:51:39earHurtsit's flashed by rombox
03:53:31earHurtsit's handcrafted by Donald Knuth from pure free-ranged bits
03:54:25Stevie-Ono, it's not
03:54:30Stevie-Othe bits are kept in captivity
03:54:38Stevie-Oit's the veal of flash sectors
03:54:57earHurtsit's like free range chicken: they/were/ free ranging
03:55:46earHurtsnow they'ee on thrir way to the afterlife accompanied by an honor gaurd of l'orange
03:56:31***Saving seen data "./dancer.seen"
03:59:11tpelliottI'm back
04:00:22tpelliottI just tried "rockbox_master.ucl" and I naw have Rockbox working from ROM! I love it!
04:00:47earHurtson an fm?
04:01:02tpelliottNo, older recorder.
04:01:29tpelliottI'm sure a version for FM will be comming shortly.
04:01:57tpelliottThe newer .ucl file fixed the ATA errr.
04:03:03tpelliottYou can flash daily builds but you have to descramble the AJZ file, then compress it to a .ucl file.
04:06:24tpelliottDo you ust the digital i/o on your FM?
04:07:58tpelliottIt's great if you have a home theater setup.
04:11:10earHurtsdigital out doesn't work
04:11:20Stevie-Oi wanna know what this ucl file is
04:11:33earHurtsI just connect my pc to my sony mini sterep
04:13:55hardeepStevie-O: ucl is just the compression format that dragon is using for his roms
04:14:12Stevie-Oi wanna know more about this format]
04:14:42hardeepStevie-O: i believe there is a link in his text file
04:18:05Stevie-OI don't see any actual information on the algorithm
04:20:07hardeepStevie-O: yeah, it's quite sparse. there is some notes in the README in the tarball
04:20:16tpelliottI just tried the "rockbox_slave.ucl" file, just for the heck of it and it gave me another ATA error -41
04:21:25 Quit Stevie-O (Read error: 104 (Connection reset by peer))
04:21:27 Join OliverKlozoff [0] (
04:21:48tpelliottWorks fine, going back to the "rockbox_master.ucl" file.
04:38:17MTgood god
04:38:30MTi could ahve compiled a kernel in the time it takes ucl to ./configure
04:39:00tpelliottYou using Rombox?
04:39:15MTroot@linux.woodhall.barn [03:36 AM] '~/ucl-1.01'
04:39:15MT> # ./configure
04:39:22MTconfig.status: creating config.h
04:39:22MTroot@linux.woodhall.barn [03:40 AM] '~/ucl-1.01'
04:39:27MTok, so slight exaggeration
04:39:34MTi could probably have made bzimage
04:39:45OliverKlozofftry: time ./configure
04:39:57MTim not going through it again!
04:40:49tpelliottI'm not a programmer but I want a .ucl that dosen't check for firmware on the hard drive.
04:41:18tpelliottSPEED!!! :)
04:41:29OliverKlozoffits got to do the equivalent anyway
04:41:36OliverKlozoffbecause it has to spin up to read the root directory
04:42:01tpelliottok. I thaught Jorg said it added a second to boot.
04:42:20OliverKlozoffif it does then it needs to be fixed :P
04:42:47OliverKlozoffit should be able to do the initial rolo check when it's reading the root dir
04:43:37tpelliottI love the fact that we never have to see Archos firmware again!
04:43:52MTi find it incredibleyt scary actually
04:44:02MTthats what i really liked about rockbox
04:44:08MTi could fuck about with it
04:44:24MTand at the end of the day, i could fix all my mistakes just by plugging it into usb
04:44:41MTim confident about other peoples stuff, just my own changes that scare me
04:44:52OliverKlozoffthen don't flash it :D
04:45:06MTfm anyways :)
04:45:09tpelliottI used "rockbox_master.ucl" to get around the ATA error. Worked great!
04:45:33MTZagor fixed that in rockbox source
04:45:42MTyou need new images
04:45:51MTata-11 problem?
04:48:07tpelliottI got the fixed version anyway from his site.
04:48:27tpelliottYes. ATA Error: -11
04:50:29tpelliottI'm sure it will be released and we'll see a news flash on the main Rockbox page tomarrow.
04:51:08MTthere is one already ..
04:51:18MT2003-07-09 22:04 zagor firmware/drivers/ata.c
04:51:18MT Cold start fixes: master_slave_select() now checks for BSY as well as RDY (since disks are BSY during powerup). Also, wait_for_bsy() looks at ATA_STATUS instead of ATA_ALT_STATUS, since the address of ATA_ALT_STATUS is not determined until later.
04:52:12 Quit hardeep ("[BX] Silly wabbit, BitchX is for kids!")
04:52:40tpelliottI'm not a programmer but I want to eventually understand the source code enough to make personal changes.
05:03:08earHurtswhat changes?
05:05:12tpelliottMenu items, etc.
05:05:40tpelliottI just flashed with the current daily build and all is well!
05:30:27 Part tpelliott
05:42:33 Join tpelliott [0] (telliott@
05:44:33 Join Stevie-O [0] (
05:48:05tpelliottStevie-O, I read your latest message on the list but I no longer have the ATA error, in fact, I burned the latest (fixed) daily build.
05:53:18 Quit OliverKlozoff (Read error: 60 (Operation timed out))
05:56:34***Saving seen data "./dancer.seen"
05:58:03 Part tpelliott
06:12:25Stevie-Odoes anyone here know wtf a flyback transformer is?
06:13:45 Join f0xShot [0] (WinNT@
06:28:02 Nick thu_ is now known as thu (
07:02:15earHurtsa transformer that fits on the back of a fly
07:02:39earHurtsthey come in house, horse, and bluebottle sizes
07:16:51 Join [IDC]Dragon [0] (
07:17:11[IDC]DragonGood morning!
07:22:28[IDC]Dragon(everybody's sleeping)
07:48:10Stevie-Oisnt it like 6am there?
07:48:21Stevie-Oor hell, i bet its nearly 8am
07:48:58Stevie-Oseeing as how it's nearly 2am here
07:49:03Stevie-Obedtime phor me
07:52:12earHurtsIDC, are you the rombox writer?
07:52:57 Quit adi|home (Read error: 60 (Operation timed out))
07:56:35***Saving seen data "./dancer.seen"
08:06:15thuearHurts: ?
08:09:35 Quit Stevie-O (Read error: 110 (Connection timed out))
08:25:18 Quit thu ("reboot")
08:30:57[IDC]DragonSorry for not reading, was away. Gotta go now.
08:34:34earHurtsjust wanted to congratulate you on rombox
08:35:00[IDC]DragonJust made a new release, hope the ATA is fixed now, thanks th Björn.
08:35:09 Join adi|home [0] (~adi|
08:35:17earHurtsI'm waiting for fm
08:35:32earHurtslooks like it's a power issue
08:35:40[IDC]DragonOK, next week I'll get access to an FM from a friend.
08:35:53[IDC]DragonDont't try FM now!
08:36:04earHurtsyeah, I know
08:36:23[IDC]DragonThe bootloader isn't made for it, the firmware contains a recorder image.
08:36:52earHurtsbut according to someone on the list it acts like recorder.ajzs did until changes were made for the different power
08:37:20earHurtsI thought the bootloader /copied/ the existing rom image?
08:37:45[IDC]DragonYes, and it checks F1. That code has to be different.
08:38:11[IDC]DragonNot the existing image, the one in my big file.
08:38:17earHurtsah, I didn't realize the button hardware differed too
08:38:27[IDC]DragonSo this has to be authored for FM.
08:38:33earHurtsoh, wait, yes I recall the button thing
08:38:48earHurtsfirst time I loaded a recorder ajz. scary
08:39:18earHurtsI''m very anxious to try the rombox
08:39:42[IDC]DragonREally gotta go now.
08:39:44earHurtsit amazes me you wrote it without having a flashable rom in your archos
08:39:52 Quit [IDC]Dragon ()
09:25:26 Quit Guest57 ("Leaving")
09:47:02 Join thu [0] (
09:56:40***Saving seen data "./dancer.seen"
09:59:13 Join bobTHC [0] (
09:59:20bobTHCgood morning
10:00:29 Quit Nibbler (
10:00:29 Quit mbr (
10:00:29 Quit PsycoXul (
10:00:48NJoinNibbler [0] (
10:00:48NJoinmbr [0] (
10:00:48NJoinPsycoXul [20] (
11:08:26 Join Lynx0 [0] (user@
11:35:16 Join TotMacher [0] (
11:41:48 Join tracktheripper [0] (
11:44:42 Quit tracktheripper (Client Quit)
11:56:42***Saving seen data "./dancer.seen"
12:41:55 Join SillyFly [0] (
12:41:55 Quit bobTHC (Read error: 54 (Connection reset by peer))
12:42:43SillyFlyanybody available?
12:46:25SillyFlyguess not.....
13:56:43***Saving seen data "./dancer.seen"
13:56:47 Join choichoi [0] (
13:57:06choichoiy a des francais ici ?
13:58:38choichoican i change the battrey on my studio10
13:58:47choichoito have a long power
14:02:53 Join langhaarrocker [200] (
14:03:12langhaarrockerargh! A wormlet patch! :)
14:03:46Lynx0choichoi: yes, you can
14:04:01choichoiwith 1800mah
14:04:29Lynx0choichoi: i think so, even 2000mhA
14:04:58choichoiwhat is your archos ?
14:05:11choichoithee model
14:05:18Lynx0recorder 20
14:05:39 Quit SillyFly ("ChatZilla 0.8.31 [Mozilla rv:1.4/20030624]")
14:05:56choichoido you speak french ?
14:06:05choichoibecause i am
14:06:09Lynx0choichoi: look in your manual for description of how to replace batteries
14:06:27choichoia oui
14:06:45choichoiouai je sais comment on le demonte
14:06:54choichoic pas tre difficile !!!
14:07:20choichoii know the manuel
14:07:42Lynx0au revoir
14:08:04 Quit choichoi (Read error: 104 (Connection reset by peer))
14:08:34 Quit langhaarrocker (Remote closed the connection)
15:48:29 Join choichoi [0] (
15:50:03 Join joshn [0] (joshn@
15:50:27 Quit joshn (Client Quit)
15:53:20choichoihow do you change the memory cach
15:53:30Lynx0what do you mean?
15:53:32choichoiis it dificult ?
15:53:55choichoiof your juke box
15:53:58Lynx0memory cach??
15:54:03choichoi2mb a 8 mb
15:54:11choichoii am french
15:54:28Lynx0ah, i think it's difficult. have you seen pictures?
15:55:13choichoibut the man change the memory chip ?
15:55:31choichoior is it the same ?
15:56:47***Saving seen data "./dancer.seen"
15:56:48Lynx0i don't know
15:58:05choichoiand the sound mod with the resistance ?
15:58:34Lynx0haven't heard about that one
15:58:44Lynx0what does it do?
15:59:08choichoiil debloque l ampli non ?
15:59:26choichoithe volume is higher
16:01:43choichoiI see that on www.rockbox......
16:14:39 Join SillyFly [0] (
16:19:49*Stevie[FP] is back from [pewf] [gone 1day 17hrs 5mins 35secs] [KS]
16:20:46 Quit choichoi ()
16:27:14 Quit SillyFly ("ChatZilla 0.8.31 [Mozilla rv:1.4/20030624]")
16:35:46 Quit earHurts (Read error: 60 (Operation timed out))
16:50:35 Join hardeep [0] (1098@
17:23:55*Stevie[FP] is away [(Auto-Away after 1hr)] [KS-MsgLog Off]
17:26:59 Join Neo [0] (~Neo@
17:27:05Neohows rombox comming
17:32:13 Quit Neo ("Client exiting")
17:56:15 Join mecraw [0] (~mecraw@
17:56:22 Quit TotMacher ()
17:56:52***Saving seen data "./dancer.seen"
18:34:28 Quit Lynx0 ("Client Exiting")
19:30:31 Quit MT (Read error: 110 (Connection timed out))
19:31:27 Join MT [0] (
19:31:32 Quit MT (Remote closed the connection)
19:37:04 Join _aLF [0] (
19:50:00 Quit hardeep ("BitchX: it keeps going and going and going and going and...")
19:56:53***Saving seen data "./dancer.seen"
20:14:33 Join msoultan [0] (
20:17:22 Join Snorlax [0] (
20:36:19 Join Neo [0] (~Neo@
20:36:30Neohows rombox commin
20:36:59 Join MT [0] (
20:37:17 Nick MT is now known as MT` (
20:37:25Neoanyone here working on rombox?
20:37:31 Nick MT` is now known as MT (
20:39:32 Quit MT (Client Quit)
20:39:34Neoanyone here working on rombox?
20:40:17 Join MT [0] (
20:42:56 Join REBELinBLUE [0] (
20:45:09Neou workin on rombox or know how its goin?
20:45:51Neothe next rockbox
20:45:53Neoits a rom
20:45:58REBELinBLUEOh :-o
20:46:01Neoflashes the rom to put the os in
20:46:06REBELinBLUEah right
20:46:15REBELinBLUEsounds dangerous
20:46:25Neoits not
20:46:27Neoi odnt think
20:46:30REBELinBLUEdoes anyone know how easy it would be to modify the code so that playlists are shown first
20:46:43Neoplaylists the m3u?
20:46:58REBELinBLUEWell if the unit loses power/crashes etc whilst flashing the ROM you will have a blank ROM
20:47:16Neoi guess thats a chance ull have to take
20:47:19REBELinBLUEYes M3U files, as in so they show before MP3 files but after directories
20:47:39MTif your motherboard loses power during a flash, you also have a blank rom
20:47:40Neodo u want directories m3u then mp3
20:47:58Neoim not sure how u can do that
20:48:09MTthe safety net with rockbox flash is that its running on batteries
20:48:10REBELinBLUEcurrently the directories are shown first then MP3 and M3U are just mixed in together
20:48:17MTso unless you try to flash with zero charge
20:48:23Neoit does the files that u use more first then the other ones
20:48:30Neotry to sort them how u want on ur pc
20:48:32Neoor mac
20:48:39MTit does them in alphanumeric order
20:48:46REBELinBLUEyeh exactly
20:48:56MTbut directories appear first in the structure
20:49:01MTlet me look at the code
20:49:04NeoMT, u workin on rombox?
20:49:07REBELinBLUEI'll take a look at the rockbox source and try it, just wondered if anyone else had done it
20:49:15MTno, but im aware of the principles
20:49:25Neowhens it comming out
20:49:35REBELinBLUEYeh Folders before Files is a logical setup. I did that in my web file manager :D
20:50:00MTo~REBELinBLUE: if i get bored/distracted the file you should be looking at is apps/tree.c
20:50:30MToh sure
20:50:37MTit shouldnt be too hard
20:50:44MTwhat model do you have?
20:50:50MTi shall try and do it now
20:50:53REBELinBLUEFM Recorder
20:51:13REBELinBLUEthe reason I ask is because I have created a series of play lists
20:51:20MTi have a couple of patches applied to my source tree ..
20:51:39MTso i can either do you a patch for you to use to compile your own
20:51:45REBELinBLUEone in each album directory called "Play entire album" which basically plays the album in the correct order, but of course because it starts with P its half way down the list
20:51:57MTor use my firmware
20:52:39REBELinBLUEIf you wouldn't mind giving me a patch I would be very greatful
20:53:20MTif you ask me again in like 1hr
20:53:25MTi should ahve it done
20:53:33REBELinBLUESpent a few hours sorting out my Archos today, transferring everything to it again, creating playlists etc cause I had to get a replacement yesterday
20:53:39MTim packing to go on holiday tomorrow as well you see :)
20:53:48REBELinBLUEah right :D
20:53:59MTyeah, i know the feeling i spent virtually the entire weekend renaming, organising and tagging mp3s
20:54:31MTnow everything is arranged by artist name and album, and has correct tags and everything has the same structure :)
20:54:47REBELinBLUEouch! 17GB
20:56:40MTa subset of my mp3s :)
20:56:56MT94 gb
20:58:25REBELinBLUEDo you just rip every song you ever hear ;)
20:58:48MTi have a lot of cds :/
20:58:59MTabout 900
20:59:14REBELinBLUEexcuse me whilst I pass out
21:00:05 Quit thu (Connection timed out)
21:01:03 Join thu [0] (
21:14:09 Quit Snorlax (Read error: 104 (Connection reset by peer))
21:20:04REBELinBLUEam I doing something wrong cause I haven't been able to compile since plugin support was added
21:20:25REBELinBLUEmake[2]: Leaving directory `/home/rockbox/apps/plugins'
21:20:25REBELinBLUEmake[1]: *** [rocks] Error 2
21:20:25REBELinBLUEmake[1]: Leaving directory `/home/rockbox/apps'
21:20:25DBUGEnqueued KICK REBELinBLUE
21:20:25REBELinBLUEmake: *** [apps] Error 2
21:22:49Neowhat firmware u use to play per cd
21:22:53Neoor album
21:23:26REBELinBLUEI just downloaded the latest daily build but can't compile :(
21:25:05Neowats the prob?
21:25:40REBELinBLUEthats what I get
21:25:59Neou have an fm recorder?
21:26:19Neoi have recorder
21:26:27Neoi have the latest daily build
21:26:38Neothats werid though
21:26:38MTREBELinBLUE: its not the prettiest, but it works
21:26:50Neowhats not the prettiest
21:26:51REBELinBLUEhave I missed something out, have the build instructions changed
21:27:15 Quit Neo ("Client exiting")
21:27:25MTwhat it does
21:27:31MTis it sorts by filetype
21:30:54Stevie[FP]i need Z or Linus or somebody
21:34:34 Join [IDC]Dragon [0] (
21:35:01Stevie[FP]you can ignore that mail I sent you, I found the problem
21:35:09[IDC]DragonI just sent you email.
21:35:22[IDC]DragonWhat was it?
21:35:31Stevie[FP]I had an instruction implemented wrong
21:35:44Stevie[FP]one that the Archos fw obviously doesn't use to its full potential
21:36:07Stevie[FP]I also was missing the implementation for 3 instructions, one of which is used by Rombox startup, zero of which are used by Archos startup
21:36:13[IDC]DragonCompiles usually don't leave the beaten path.
21:36:56[IDC]DragonWhat exotic intructions are there?
21:37:16REBELinBLUEhas anyone heard of anyone else who hasn't been able to compile from source since plugins were added?
21:37:22REBELinBLUEam I doing something wrong?
21:37:28Stevie[FP]mov #imm:8, rn
21:37:42[IDC]Dragonwhere is that?
21:38:19Stevie[FP]loads a constant value into Rn (the value is the lower 8 bits of the instruction code)
21:38:41Stevie[FP]Archos actually used it, it just didn't use it to its full potential
21:38:51Stevie[FP]i failed to realize that the value is supposed to be sign-extended
21:38:58Stevie[FP]so when gcc did
21:39:03Stevie[FP]mov #0xFC, r5
21:39:18Stevie[FP]and used r5 (which should be 0xFFFFFFFC after that instruction) as a bitmask
21:39:34[IDC]DragonYes, I found that confusing in the disassemblies, too.
21:39:35Stevie[FP]i kinda lost the upper 24 bits
21:39:42REBELinBLUEOh I'm a bloody idiot. I haven't reinstalled Perl thats why I can't compile!
21:39:49Stevie[FP]one of the instructions I didn't implement was
21:39:53Stevie[FP]lds @rm+, PR
21:39:59[IDC]DragonI'ts quite often used for decrementing.
21:40:04Stevie[FP]which is used in thread switches
21:40:13Stevie[FP][IDC]Dragon: yah, add #0xff, r1
21:40:30[IDC]DragonI was using a Lauterbach emulator, it also had problems with it.
21:40:32Stevie[FP]I believe the archos fw doesn't use threads, at least not to the extent we do
21:41:04[IDC]DragonI had to patch a certain piece of code to fix that.
21:41:18Stevie[FP]sounds like fun =/
21:41:44[IDC]DragonPatching the SH code, with something equivalent, I mean.
21:41:58Stevie[FP]i thought you meant you patched the emulator
21:42:03Stevie[FP]which reminds me
21:42:09Stevie[FP]GCC is producing some VERY strange code
21:42:27[IDC]DragonBut much better than the one Archos was using.
21:42:28Stevie[FP]in sleep()
21:42:41Stevie[FP]I dunno, this one's fscked up
21:42:46Stevie[FP]in sleep()
21:42:59Stevie[FP]after executing the SLEEP instruction
21:43:01Stevie[FP]it does this:
21:43:11Stevie[FP]if (++next >= num_threads)
21:43:12Stevie[FP]next = 0;
21:43:59Stevie[FP]the assembly used to implement this goes like this:
21:44:21Stevie[FP]r7 = &next
21:44:28Stevie[FP]r1 = &num_threads
21:44:40Stevie[FP]r2 = next [r2 = *r7]
21:44:56Stevie[FP]r1 = num_threads [r1 = *r1]
21:45:14Stevie[FP]r3 = next+1 [r2+1]
21:45:38Stevie[FP]cmp/hi r3, r1 [next+1 >= num_threads?]
21:45:57Stevie[FP]at this point, if next+1 >= num_threads, T=1; otherwise T=0
21:46:08Stevie[FP]this is where things get strange
21:46:26Stevie[FP]movt r1 [r1=T]
21:46:51[IDC]DragonI'm too tired to follow this, if you don't mind.
21:47:00Stevie[FP]meh, ok
21:47:10Stevie[FP]it just generates some VERY bizarre code
21:47:36[IDC]DragonAs long as it works...
21:47:55 Quit REBELinBLUE ()
21:48:24[IDC]DragonExcuse me for a while
21:48:34 Nick [IDC]Dragon is now known as [IDC]Dragon|away (
21:55:50 Join matsl [0] (
21:56:56***Saving seen data "./dancer.seen"
22:00:02 Part f0xShot
22:00:07 Join groovingandi [0] (
22:02:05 Part groovingandi
22:08:00Stevie[FP]anyone here?
22:08:27CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
22:08:27*Stevie[FP] throws stuff at everyone
22:08:50 Join Guest57 [0] (
22:08:54*MT ducks
22:09:09Stevie[FP]do you have a Recorder?
22:09:14MTfm sorry :/
22:14:26Guest57hi all
22:14:48Guest57is there some news about the flashable firmware for the FM Jukebox
22:16:09Stevie[FP]the news is you can't do it yet
22:16:50 Join Zagor [242] (
22:17:09Zagorhi guys
22:17:40Guest57it seems that the "crazy" people who try it report the same effect that did appears with the first FM firmware : the shutdown problem cause of the diffence in the voltage level .. so I think it will be resolve .... :)
22:17:58Guest57isnt it ?
22:18:02Stevie[FP]sup Z
22:19:15Guest57see u later .
22:19:18 Quit Guest57 ("Leaving")
22:22:53Zagori'm pondering if we should rolo at all after flash start. i'm don't think we should.
22:23:53Stevie[FP]I'm in favor of it...
22:24:02_aLFand as a option ?
22:24:17Stevie[FP]mostly cuz I figure the best plan of attack is to have a simple boot-loader
22:24:44Stevie[FP]a known-working-version
22:25:10Stevie[FP]we have to spin the damn drive up anyway
22:25:25Stevie[FP]Hell, we even have to grab the contents of the root dir anyway
22:25:35Stevie[FP]we should be able to check for an ajz file for free
22:25:46Zagormy thinking is this: when you run flash, you most likely flash the version you want. thus rolo:ing is pointless and you tend to remove ajbrec.ajz to avoid it. but then you are out of luck if the flash version bugs, and you F1-boot into archos. no ajbrec.ajz means no rockbox and no chance to burn a new image without connecting to a pc.
22:26:55 Nick [IDC]Dragon|away is now known as [IDC]Dragon (
22:27:03Stevie[FP]I would recommend that the flash plugin auto-rename any /ajbrec.ajz
22:27:37Zagorthat still leaves us in the cold (as in archos firmware) on f1-boot
22:27:43Stevie[FP]to like ajbold-071003162600.ajz
22:27:58Stevie[FP]how about this?
22:28:12Stevie[FP]put some sort of unique signature in the file
22:28:21[IDC]DragonHi Zagor, yes, I'm close to take that auto-startup out again.
22:28:24Stevie[FP]like, say, the current date/time in some form when the image is built
22:29:02ZagorStevie[FP]: but my point is we only want to rolo-on-boot if we boot the archos firmware, and that firmware only looks for ajbrec.ajz
22:29:21[IDC]Dragontakes an extra second because it has to read the root dir or so
22:29:23Zagordragon: ok. now you know I agree with it :-)
22:29:59Stevie[FP]I suppose it wouldn't be unreasonable to expect the atypical user to manually ROLO the ajz file
22:30:13[IDC]DragonRenaming to ajbrec.ajz_ resolves that both would rolo.
22:30:15Stevie[FP][IDC]Dragon: we have to read the root dir anyway, for the directory tree
22:30:32[IDC]DragonEven when resuming straight away?
22:30:42Stevie[FP]I thought so
22:30:57[IDC]DragonI don't, why is is quicker without?
22:31:01Stevie[FP]yes we do
22:31:13Stevie[FP]because we need to scan the root dir for the damn playlist file
22:31:17[IDC]DragonZagor, please be our judge...
22:31:47Zagorwe resume before loading root
22:32:01Zagorthe resume info is stored in the config block
22:32:03Stevie[FP]but don't we need to scan the root dir for the playlist file?
22:32:06 Join Neo [0] (~Neo@
22:32:11Neohey all
22:32:25ZagorStevie[FP]: yes, to load the playlist file
22:32:34Stevie[FP]or even to find the mp3 we wish to play
22:32:42Neohows that rombox commin
22:32:46Zagorbut resuming still means one less dir scan
22:33:07Neostil workin on it?
22:33:16[IDC]DragonI think it doesn't store the filename and position, just a disk sector # or so.
22:33:17Zagorhowever the root dir scan is very fast. most of the rolo time is descrambling
22:33:34Zagordragon: no actually it's filename and position
22:33:40[IDC]DragonNo, I menat just the check for a non-present file.
22:33:44 Quit Neo (Client Quit)
22:34:01[IDC]DragonMaybe my root is too large, with hundreds of directories.
22:34:27Zagoryes, that will take some time. my root is ~30 entries
22:34:29Stevie[FP]which function scans a dir for a filename?
22:34:37ZagorStevie[FP]: open()
22:35:00Stevie[FP]wheres that defined?
22:35:40Stevie[FP]so actually opendir does...
22:35:44Stevie[FP]or readdir
22:36:16Zagorwell, readdir returns one entry per call. open is the one that looks for a particular entry.
22:36:43Stevie[FP]actually scans for filenames...
22:36:50*Stevie[FP] thinks of a more optimized version
22:37:01Stevie[FP]we could scan faster with a specialized fat_search function
22:37:14Zagornot without caching a lot
22:37:25Stevie[FP]no, this is what i'm thinking of
22:37:53Stevie[FP]err, i can't find fat_getnext
22:38:10Stevie[FP]fat.c is under drivers
22:38:46Stevie[FP]what we do is
22:38:50Stevie[FP]read the whole damn dir
22:38:57Stevie[FP]entry by entry
22:39:03Stevie[FP]looking for the name we desire, correct?
22:39:34Stevie[FP]fat_getnext needs to do this:
22:39:37Stevie[FP](1) fetch a directory entry
22:39:48 Join REBELinBLUE [0] (
22:40:10Stevie[FP](2) if this is an LFN entry, prepend this data to our LFN buffer
22:40:26Stevie[FP](3) if this is not the last LFN entry, goto step 1
22:41:18Zagorthis requires a huge cache. dirs can be Really Long
22:41:31Stevie[FP]i see the sectorcache thing
22:41:38Stevie[FP]that's what you're talkign about, right?
22:42:03Zagorno, i mean reading "the whole damn dir"
22:42:14Zagormaybe i'm misunderstanding?
22:42:18Stevie[FP]but we already do that
22:42:46 Join cst [0] (
22:43:20Stevie[FP]file.c line 120, for the 7/09 daily
22:43:26Zagorno, we only cache max three sectors
22:43:50Zagorok, apparently we are talking about different things :)
22:44:09Stevie[FP]ok, back to where I was
22:44:24_aLFwhen ajbrec.ajz is created ? only in usb mode ?
22:44:33 Join SillyFly [0] (
22:44:50Stevie[FP](4) if it is the last LFN entry, we fetch the next FAT direntry for the file attributes and stuff
22:44:54Zagor_aLF: it's never created, you copy it to the disk
22:45:14_aLFwhy not store in cmos after usb mode, if .ajz is here
22:45:40_aLFsorry, where config block is store
22:45:40Stevie[FP]when did we become a motherboard?
22:45:44Zagor_aLF: what are you trying to achieve?
22:46:00ZagorStevie[FP]: well, the rtc ram is sort of cmos ram...
22:46:05Stevie[FP](5) we stuff all the relevant data into a fat_direntry and return it to the user
22:46:08_aLFyes rtc ram
22:46:27 Quit msoultan ()
22:46:29Stevie[FP]I believe that's how fat_getnext operates
22:46:41SillyFlyhi. I'm having trouble compiling gcc with target=sh-elf on linux. anybody care to try and help me?
22:46:41Zagoryes it is
22:46:46_aLF-> when we boot, rombox check if .ajz is here with rtc
22:46:50Stevie[FP]that's perfectly fine for browsing dirs
22:46:51_aLFnot with fat
22:46:57Stevie[FP]cuz we need to get all the filenames anyway
22:46:58ZagorSillyFly: are you using the versions we suggest?
22:47:12Stevie[FP]but for things like the auto-ROLO
22:47:25Stevie[FP]and when we're playing from a playlist
22:47:28Zagor_aLF: as I said, checking the disk is not the slow part. actually roloing the file is.
22:47:39Stevie[FP]we don't care about any files other than the one we want
22:47:56_aLFZagor> ok
22:47:56SillyFlyZagor: no. newer versions. I suspect I'm missing some header files, but dunno which....
22:48:05Stevie[FP]in fact, we don't even care about the filename of the file we want, we just want to get the data for that file
22:48:16*Stevie[FP] ponders
22:48:49[IDC]DragonThats's how resume works, stores the disk position, right?
22:48:49ZagorSillyFly: there are problems with binutils in the newer versions. i recommend simply using the specified versions unless you feel like debugging the tools.
22:49:12Zagordragon: no, resume stores the filename and byte offset
22:49:15SillyFlyoh. ok. 10x.
22:49:40Stevie[FP][IDC]Dragon: it couldn't work −− what if the disk was defragged in usb mode?
22:49:49[IDC]DragonFilename? Isn't that a bit wasteful? what size can it have?
22:49:58Zagor260 bytes
22:50:08[IDC]DragonIn rtc ram?
22:50:09Zagorit's large, but it's The Right Way
22:50:17Zagorno, config block on disk
22:50:57[IDC]DragonThen we ave no space problenm.
22:53:49 Quit cst ("Wechsle Server")
22:54:23 Join cst [0] (
22:59:08[IDC]DragonZagor: thanks again for fixing ata::init(), that really did it for the people.
23:00:45[IDC]Dragon(not listening to take credits)
23:03:32 Join tpelliott [0] (telliott@
23:05:33 Join joshn [0] (joshn@
23:06:18Zagordragon: the pleasure was all mine. (it did it for me too... :-)
23:06:22Stevie[FP]what's this dir->entrycount thing?
23:06:51[IDC]DragonZagor: are you using Rockbox in flash now?
23:07:06Stevie[FP]i thought it was a cached copy of the # of files in the dir, but it gets reset to 0 when we get the end-of-dir
23:07:34Zagordragon: yes
23:07:48ZagorStevie[FP]: it's how many entries this file is using
23:07:59Stevie[FP]who uses that?
23:08:15[IDC]DragonZagor: Then it's not exactly a reference platform for Rockbox development any more.
23:08:19Stevie[FP]fat_open references it
23:08:23Stevie[FP]but I don't understand what it uses it for
23:08:54 Join hardeep [0] (1098@
23:08:58ZagorStevie[FP]: write_long_name uses it
23:09:08tpelliottI'm successfully running Rombox also. I wouldn't mind a version that dosen't check for firmware on hard drive.
23:09:30Zagordragon: true, but that's where the beauty of F1-ON comes in
23:09:44[IDC]Dragontp: I will do that.
23:10:16[IDC]DragonYOu can rename it for the time being, so that only Archos finds it.
23:10:46Stevie[FP]or is that the 'numentries' thing?
23:11:02ZagorStevie[FP]: a simpler example is fat_remove()
23:11:15Stevie[FP]oh i see
23:11:23 Quit cst ("Demo-Zeitüberschreitung")
23:11:42[IDC]Dragontpelliott, you read that?
23:12:09tpelliottWhen will Rombox make a splash on the Rockbox home page? It's one of the coolest features yet.
23:12:15 Join cst [0] (
23:12:30Stevie[FP]does this include the shortname? so a filename with 1 LFN entry (usually happens when changing the case of a name) has an entrycount of 2?
23:13:04ZagorStevie[FP]: yes, if my memory is correct
23:13:42[IDC]DragonZagor: how about F3-On starting gdb right away?
23:13:59[IDC]Dragon(Or F2-On)
23:14:32Zagordragon: nah, too few people have the serial mod to make it worthwile
23:15:10Zagorwe'd probably just get bug reports from confused users trying it :)
23:15:43[IDC]DragonMy autoring tool is prepared for n images, but not yet the bootloader and the update plugin.
23:16:52[IDC]DragonWhen I get my 512k flash chip, I wouldn't know how to fill it.
23:17:07[IDC]Dragon(Compression did work too well)
23:17:43Zagorwe can have a boot animation :)
23:18:03[IDC]DragonYeah, while waiting for the disk.
23:18:26[IDC]DragonAnd we could play a jingle.
23:18:42Zagorthat would probably get boring fast...
23:18:50[IDC]DragonHow about resuming from flash?
23:19:11Zagorstoring resume info in flash?
23:19:25[IDC]DragonNo, storing the MPEG buffer.
23:20:25[IDC]DragonOn another subjject: I'd like to check in my sources.
23:20:47[IDC]DragonHow do you like that organized?
23:20:52REBELinBLUEanyone know why I can't compile, ever since plugin support was added I get this output
23:21:32Zagordragon: could you send me a tarball so I can have a look?
23:22:03[IDC]DragonFor me Windows user, this would be a zip file.
23:22:05ZagorREBELinBLUE: we need more from the build log
23:22:12Zagordragon: sure, no problem
23:22:36REBELinBLUEOK Lets me try again and I'll post the whole thing
23:24:06[IDC]DragonGenerally peaking, I have the 2 plugins, the bootloader, an authoring tool, my "minimon" (getting injected for UART boot), RockFlash being the PC application that talkes to it (and so can flash, dump. upload, execute, etc.)
23:26:27[IDC]DragonProbably you don't want all that in the standard source tarball.
23:27:13MTwindows start sound :)
23:27:52[IDC]DragonFor the plugins, I don't know if both should be in the standard set, and then there's the question on where/how to keep that firmware image being used by the first.
23:27:54Zagordragon: actually, i probably do. i just need to come up with a good way to integrate it
23:28:20[IDC]DragonYou do what?
23:28:31Zagorwant it all in the standard source
23:28:54[IDC]DragonUART boot is rather exotic, like gdb.
23:29:07[IDC]DragonAuthoring probably too.
23:29:22Zagoryes, but gdb is in the standard source
23:29:34[IDC]DragonAh, i see.
23:29:42Zagorit's much simpler, maintenance wise, to have it all in one package
23:29:57[IDC]DragonDangerous, powerful toys.
23:30:38Zagoryes. we'll simply write big notes saying "don't play with things you don't understand"
23:31:06[IDC]DragonAs long as there's people out there who force recorder images into an FM...
23:31:18REBELinBLUE there you go, the complete log
23:31:42[IDC]DragonI was pretty upset about that.
23:31:57Zagordragon: yes, but we will never be able to stop those kinds of people.
23:32:20[IDC]DragonIs there a way to tell recorder/FM apart at runtime?
23:32:34[IDC]Dragon(Have asked that before, I know)
23:32:43Zagorwhat we can do is provide enough warnings to avoid mistakes. people who actually want to ruin their box cannot be prevented.
23:32:56Zagorum, we could easily add that
23:33:23*Stevie[FP] doesn't get this
23:33:28Zagorwell the binary is compiled for the specific target, so we could just do it with #ifdefs
23:33:36Stevie[FP]i have the root directory cluster in the fat 'superblock'
23:33:42REBELinBLUEanyone? it seems to build the firmware fine, its then when it gets on to the plugins
23:34:02Stevie[FP]after the first cluster
23:34:08[IDC]DragonI don't want 3 targets for the bootloader, maintanance-wise as you say
23:34:10Stevie[FP]how do I know what the next cluster is?
23:34:34Stevie[FP]the only page i could find on fat32 only talks about the directory entries
23:34:37Stevie[FP]and other structures
23:34:54Zagordragon: oh, you want to know in the boot loader
23:35:04MTbah, i disconnected just as dragon answered my question :/
23:35:16ZagorMT: check the log
23:35:25[IDC]Dragonwhat question?
23:35:56ZagorREBELinBLUE: i can't see what that error is about. everything seems to be going fine.
23:36:21Stevie[FP][IDC]Dragon: you want to prevent people from flashing the wrong thing, right?
23:36:24ZagorREBELinBLUE: you do get an ajbrec.ajz and all the .rocks, right?
23:36:30REBELinBLUE:| It doesn't really matter, because as I said the firmware compiles fine, its the plugins which don't
23:36:42REBELinBLUEhow many .rock files should there be/
23:36:55[IDC]DragonThat, and I want the bootloader to know how to check for F1.
23:37:01MToh, i actually disconnected before i asked the question, you just happened to be talking about the same subject when i reconnected :)
23:37:03REBELinBLUEI have 14, totaling 57.1k
23:37:11MT<MT> dragon: just so i understand, to develop/fix problems with rom flash for fm recorders, i would require serial mod and UART boot yes?
23:37:26Stevie[FP][IDC]Dragon: just have 3 diff versions
23:37:43ZagorREBELinBLUE: yup, that's all
23:37:51tpelliottI did notice one thing just now. If I turn on my Archos with the USB plugged in, it hangs at "loading". It starts USB fine if I turn it on before plugging in the USB.
23:37:52[IDC]DragonMT: No, you don't. I maybe will.
23:37:55Stevie[FP]we do it for rockbox, it won't be all that terrible for the bootloader
23:38:19REBELinBLUEweird :| It compiles fine I just get an error :D
23:38:24MTREBELinBLUE: i did your patch, but its crappy, and im not happy with it, if you check"> tomorrow, ill hopefully have done a version im happy with by then
23:38:27ZagorREBELinBLUE: yes, very strange
23:38:32MTotherwise i'll upload the crappy one
23:38:35REBELinBLUEah well as long as the firmware compiles :D
23:38:37ZagorREBELinBLUE: have you tried it? everything works?
23:38:44[IDC]DragonStevie: It's just an overkill, and when authoring the firmware file I have to be very careful to pick the right one.
23:38:54REBELinBLUEcool thanks :D I was just about to try and make it myself
23:39:04Stevie[FP][IDC]Dragon: what do you mean by 'authoring the firmware file'?
23:39:28MTah, actually
23:39:35MTi think if i change a -1 to 1
23:39:35Zagordragon: we need to make a new plugin that creates the initial firmware file from the existing rom
23:39:38REBELinBLUE<Zagor> REBELinBLUE: have you tried it? everything works? −− seems to work fine
23:40:05Stevie[FP][IDC]Dragon: that was something I didn't understand, why not use the image in ROM
23:40:09[IDC]DragonWell, I composed that large firmware_rec.bin file from the bootloader and the 2 images.
23:40:16Zagorthat's the only way I can think of to avoid voilating archos' copyright
23:40:27MTthats got it :)
23:40:28Stevie[FP]besides, we know for sure that the image in ROM works with the given HW
23:40:38[IDC]DragonStevie: I need to compress it to make room for Rockbox.
23:40:45Stevie[FP]otherwise it wouldn't have passed Archos's stringent QA...err
23:41:04[IDC]DragonZagor: well, we can do that.
23:41:17Zagorsounds like i need to fix that bigmem api... :-)
23:41:26Stevie[FP]hey Zagor...
23:41:27[IDC]DragonIs it broken?
23:41:47Zagorno, it just doesn't exist yet :)
23:42:03[IDC]DragonWe don't necessarily need to do all that on the box.
23:42:04 Join CS1 [0] (
23:42:09Stevie[FP]while you're at it, I think we oughta throw in a new plugin API routine to determine what we're running on - player, recorder, or fm
23:42:27[IDC]DragonI thought that's the first check.
23:42:37ZagorStevie[FP]: i though about that too, but... why?
23:42:55Stevie[FP]then the rombox plugin that flashes the board can make sure that the image they're flashing is correct
23:42:58[IDC]DragonBut then I wonder how that guy ran the plugin on an FM.
23:43:04Stevie[FP][IDC]Dragon: here it comes
23:43:09Stevie[FP]here it comes >:)
23:43:16Stevie[FP]when flashing the bootloader
23:43:17Zagordragon: good question indeed
23:43:23Stevie[FP]have a special reserved byte
23:43:25 Quit cst ("Demo-Zeitüberschreitung")
23:43:38hardeepi thought recorder and fm recorder were considered the same for plugins
23:43:39Stevie[FP]and set that byte value to indicate the firmware type
23:43:51hardeepso recorder plugins would work fine on fm and vice versa
23:43:52Stevie[FP]err hardawre type (fm/recorder)
23:44:13Stevie[FP]then your bootloader doesn't need to detect −− it knows
23:44:14Zagorhardeep: ah, yes they are
23:44:25[IDC]DragonStevie: OK, but this requires a correct run in the first place.
23:44:37 Quit _aLF ("bye")
23:44:45Stevie[FP][IDC]Dragon: you mean a correct copy of rockbox (fm or rec)?
23:45:15[IDC]DragonYes, or runnig the correct plugin.
23:45:22Stevie[FP]what do you mean, correct plugin?
23:45:28MTits very ugly, very inefficient, but it does what you wanted
23:45:38Stevie[FP]why do you need multiple plugins?
23:46:13REBELinBLUEah thanks :D
23:46:16[IDC]DragonOne for each platform.
23:46:29Stevie[FP]I figured that was what you meant by multiple plugins
23:46:32Stevie[FP]but that doesn't answer why
23:46:35[IDC]DragonI thought they don't ru cross-platform.
23:46:42[IDC]Dragondon't run
23:46:47Stevie[FP]I really don't see any reason why they couldn't
23:46:50Stevie[FP]they all have the same cpu
23:47:04[IDC]DragonThe "mandatory" first call is a check
23:47:27[IDC]DragonBut not the the same hardware.
23:47:43Stevie[FP]who makes this first call? [keep in mind: I know almost zilch about the plugin system]
23:47:59Stevie[FP]yes but I don't think Sokaban cares which ADC register your F1 key is on
23:48:11[IDC]DragonThe plugin calls TEST_PLUGIN_API(api);
23:48:26Stevie[FP]looks like a macro
23:48:37MTmm, excuse me for being dense, but surely just stick a is_fm() function into the api
23:48:47MTor define FM when compiling
23:49:12Stevie[FP]MT: the latter would not work for separately compiled .rocks
23:49:22[IDC]DragonCompiling so won't keep people from running it on another platform.
23:50:14Stevie[FP][IDC]Dragon: I would imagine most plugins to be games
23:50:36Stevie[FP]in which case the HW really is the same on both FM and recorder
23:50:49Stevie[FP]they have the same buttons and the same display
23:51:18Stevie[FP]likewise, they [usually] have the same ROM chip
23:51:30tpelliottBatteries and voltage is different.
23:51:30REBELinBLUEMT thanks it works :D
23:51:37[IDC]Dragonthe macro expands to a function which compares API version and model.
23:51:48Stevie[FP]tpelliott: why the hell would a game care about the battery type?
23:52:05Stevie[FP]where did Z go
23:52:13Stevie[FP]this doesn't sound right
23:52:20[IDC]Dragonmaybe there will be a radio plugin ;-)
23:52:23Zagori did it this way because the api for fm and recorder is identical
23:52:24tpelliottTrue. I thaught you were talking about Rombox.
23:52:33Stevie[FP][IDC]Dragon: so?
23:52:36Zagorthus I saw no reason to keep them separated
23:52:38Stevie[FP][IDC]Dragon: who cares?
23:53:14Stevie[FP]Zagor: is it true that a .rocks, once built, can only be run on a certain model (fm or recorder)?
23:53:48[IDC]DragonWell, at least I use the build config to make a suffix for the hard coded filename. The"_rec" of firmware_rec.bin".
23:56:01Stevie[FP]imho, the Right Way To Do It is this:
23:56:16ZagorStevie[FP]: there is no difference btwn fm and rec. only btwn recorders and players.
23:56:28Stevie[FP]Zagor: as IDC points out, FM has a radio
23:56:34Zagoryes, but the api doesn't
23:56:53Zagorand i'm not sure it ever will
23:56:58Stevie[FP]Zagor: here's the idea
23:57:00***Saving seen data "./dancer.seen"
23:57:12Stevie[FP]rombox bootloader needs a way to determine which model it's on
23:57:23Stevie[FP]so it knows which key is ON, and which key is F1
23:57:27Stevie[FP](they have different ADCs)
23:57:44Stevie[FP]now, afaik there's no trivial way to detect this
23:57:55Stevie[FP]other than by flashing the Recorder image and seeing if it doesn't boot
23:57:55tpelliottDoes the digital; i/o still work the same on the FM? Can it send digital to a reciever's Toslink connection? I've heard conflicting reports.
23:58:40Zagorthe proper way, imho, is to tell users "don't flash the wrong image". this works for all hardware makers, i don't see why it can't work for us.

Previous day | Next day