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 2004-04-15

00:03:40[IDC]Dragonbut lcd_update_rect() saves you half a second, as expected
00:07:27[IDC]Dragontry this:
00:07:29[IDC]Dragonif ((y_pixel & 0x7) == 0x7)
00:07:30[IDC]Dragon rb->lcd_update_rect(0, y_pixel & ~7, LCD_WIDTH, 8);
00:07:30 Quit c0utta (" HydraIRC -> <- The dawn of a new age")
00:08:02[IDC]Dragoninstead of lcd_update()
00:12:48mattzzyup - thanks.
00:13:32mattzzI will have to change it because I am counting downto 0 now, right?
00:14:26[IDC]Dragoneven simpler then:
00:14:27[IDC]Dragon if ((y_pixel & 7) == 0)
00:14:27[IDC]Dragon rb->lcd_update_rect(0, y_pixel, LCD_WIDTH, 8);
00:15:43mattzz1.15secs @ 15 itarations ;-)
00:15:57mattzzlast update is missing
00:16:01[IDC]Dragonany reason for some variables to be unsigned char?
00:16:27mattzzthis is going to be a tough review ;-)
00:16:28[IDC]Dragon unsigned char n_iter;
00:16:29[IDC]Dragon unsigned char x_pixel, y_pixel;
00:17:00mattzzboy, my first code review...
00:17:18mattzzwhat do you propose instead of unsigned char?
00:17:42mattzzint for 32 bit performance optimum?
00:17:58[IDC]Dragonif you use smaller types than the native, extra code is needed after arithmetics to limit the value again.
00:18:26amiconnextu.b in this case...
00:18:33mattzzok, so I should use INT ?
00:18:39[IDC]Dragongeneral rule: use int or unsigned int if size doesn't matter
00:22:16[IDC]Dragonint is probably best for x/y_pixel, because you mix it with other signed values in the calculations
00:22:30[IDC]Dragonint/long is no difference
00:22:31amiconnmattzz: Use unsigned for x / y, shifting will be faster then.
00:22:37[IDC]Dragonon a 32 bit CPU
00:22:56amiconnGrr, nonsense.
00:23:34[IDC]Dragonthey are not always positive, aren't they?
00:24:05amiconnYes, of course. Nonsense, as I said.
00:29:24 Join Galik [0] (~galik@
00:35:22[IDC]Dragonstrange, a tail loop didn't help the iteration.
00:35:47[IDC]Dragonit's fastest with while (++n_iter < max_iter) at the top
00:37:18mattzzwhy not post increment?
00:37:46mattzzah... forget it
00:38:04[IDC]Dragonbecause then it can increment in the same register before the check
00:38:24mattzzI see
00:38:26[IDC]Dragonelse it would need another for the original value.
00:39:16[IDC]Dragonso this is cheating by one iteration ;-)
00:40:21mattzzit's worth it
00:40:28[IDC]Dragonstrange: unsigned char is faster for the iteration counter and limit
00:40:53[IDC]Dragonmaybe because of the shortermemory fetch?
00:40:58mattzzhow much? 3%?
00:41:10[IDC]Dragonnot that much
00:41:32[IDC]Dragonmaking up another nickname?
00:44:39mattzz1.12secs - that was another 0.03 secs @ 15 iterations. nearly 3%!
00:44:52[IDC]Dragondoing what?
00:45:10mattzzpre increment in while statement
00:45:25mattzzcould also be that one iteration less....
00:45:39[IDC]Dragondon't forget it dous one iteration less than "specified"
00:45:45mattzzbut the result matters, right?
00:46:32mattzzbut it _feels_ a lot faster now ;-)
00:48:26[IDC]Dragonare your screen lines OK now?
00:48:57mattzzYup, I put the b&w version on"> (Schwarzbrot release)
00:49:26[IDC]Dragonthat was a good one
00:50:01 Quit Galik (Remote closed the connection)
00:51:39 Quit matsl (Remote closed the connection)
00:52:18 Join Galik [0] (~galik@
00:53:04mattzz4<<28 ?
00:53:07[IDC]Dragonstrange magic number
00:53:25[IDC]Dragoncan't the compiler resolve that?
00:53:53mattzzdoes he have to?
00:55:03[IDC]Dragonyour coords are off by 1
00:55:19[IDC]Dragonstart at LCD_HEIGHT-1
00:55:28[IDC]Dragoncheck if >= 0
00:55:41[IDC]Dragon(so it has to be signed)
00:58:01mattzzhm, that would explain the strange simulator bahaviour
00:58:25[IDC]Dragonyou're writing off-screen
00:58:33[IDC]Dragonvery dangerous!
00:59:05mattzzback in file list screen the filenames where garbled....
00:59:12mattzzfixed now
00:59:33[IDC]DragonI just started your old plugin, what a breakthrough, this new one!
00:59:59[IDC]Dragonold took 36 secs for 10 depth
01:00:23mattzzresult of a good teamwork
01:00:52mattzzthere was an intermediate release with 3.6 secs
01:01:02mattzzin the patchtracker
01:01:46mattzzbut this one is much more fun
01:01:58[IDC]DragonI'm old-fashioned
01:02:04[IDC]Dragondo you if ((y_pixel & 0x7) == 0)
01:02:17[IDC]Dragonagain after fixing the coords?
01:05:02[IDC]Dragonand you may fix the true iterations, init to one more, display one less
01:07:17[IDC]Dragonwell for speed then, any more would need asm
01:07:30[IDC]Dragon(I guess)
01:09:23mattzznow I don't need the final lcd_update_rect anymore, due to the way of counting. nice.
01:11:28mattzzGentlemen, do we have a patchtracker candidate here, or what?
01:16:54 Quit AciD (Read error: 104 (Connection reset by peer))
01:17:21 Join AciD [0] (
01:25:53mattzzok, full patch committed to patchtracker
01:26:30[IDC]DragonI was just preparing it anyway
01:26:41mattzzah, I see
01:26:48scott666whats the total imrpovment speed?
01:27:18mattzzfrom 30 secs to 3.5 secs to 1.16 secs now
01:27:40mattzzwith much higher accuracy
01:29:05[IDC]Dragonwhy the lcd_clear_display() at the beginning of calc_mandelbrot_set()?
01:29:30mattzzjust a habit from old C64 times
01:29:32[IDC]Dragonyou don't update it, or qould you like to?
01:29:57[IDC]Dragonit was only happening in your old code
01:30:01mattzzhonestly, this line seems to be obsolete
01:30:45mattzzah, wait. it's not
01:31:04mattzzwe want a complete redraw.
01:32:08[IDC]DragonOK, I'll put an update behind it.
01:32:26[IDC]Dragon(you do a complete redraw anyway)
01:33:03[IDC]Dragondon't forget to take it out when you redraw a stripe ;-)
01:33:53mattzzgreyscale first
01:34:52amiconn[IDC]Dragon, mattzz: I just uploaded an update to grayscale.c (not my 3% tweak, since it looks too weird).
01:35:27[IDC]Dragonbut which?
01:37:42amiconnI've rewritten my safe replacement mechanism for lcd_update() to not cause flicker within the grayscale overlay area. And it contains a minor speed tweak.
01:38:20[IDC]Dragon"safe replacement"?
01:39:18amiconnAs you surely know, you cannot safely use lcd_xxx() functions that directly write to the lcd controller while having the interrupt running.
01:40:17amiconnSo I had defined a flag which is set by the replacement function and then read within then interrupt routine.
01:40:53amiconnThis used to do an lcd_update() *instead* of flipping to the next bitplane. So it caused flicker.
01:41:42amiconnThe new version first flips to the next bitplane *and* lcd_update_rect()'s everything around it afterwards.
01:44:54[IDC]Dragonmandelbrot wasn't in cvs yet?
01:46:05[IDC]Dragonamiconn: OK, understood, I wasn't aware of the "non-grayscale" side activities.
01:46:36mattzz[IDC]Dragon: nope
01:48:23amiconn[IDC]Dragon: This is useful if the grayscale overlay doesn't cover the whole screen, e.g. for a status line at the top/bottom in a game.
01:48:51[IDC]Dragonmattzz: now it is
01:49:21[IDC]DragonI took the liberty to brush it up a bit
01:49:39amiconnThis is useful because you cannot use fonts within grayscale unless you define some as bitmaps. (I'm thinking Minesweeper here).
01:49:47***Saving seen data "./dancer.seen"
01:49:57mattzzthanx a lot. could you please also close the patchtracker item?
01:50:20[IDC]Dragonif it already shows up
01:53:18amiconn[IDC]Dragon: There is still my patch for memcmp() which isn't that useful after all, I think. What should I do with that?
01:53:53[IDC]Dragonwe have no memcmp(), right?
01:55:07amiconnNo, we have memcmp() (in C), but it is used only in very few places, and my asm replacement is bigger, so I consider it a waste of space.
01:55:37[IDC]Dragonah, I remember
01:56:00[IDC]Dragonthe memcmp() places are not spped critical, iirc
01:57:55[IDC]Dragondoesn't cry for integration
01:58:16[IDC]Dragonbut I'd like your disk functions enabled
01:58:33amiconnNo it doesn't, thats why I ask what to do with it. Should (can) I close this myself?
01:58:47[IDC]Dragonhave you tried the write on somebody with a bad disk?
01:59:02amiconnNo, not yet.
01:59:08[IDC]Dragondunno if you can close it, try
01:59:55amiconnThe C writing is also sped up in the current builds, so far nobody did complain about it.
02:00:49[IDC]Dragonbedtime is long overdue
02:01:27amiconnPerhaps somebody (me?) should prepare 3 test builds (with read/write/both optimizations enabled) and post on the ml to test these (along with a warning).
02:02:14[IDC]Dragonsounds good :)
02:02:53amiconnI'm using both optimizations all the time without problems, but then there are these old Hitachi HDs...
02:02:54[IDC]DragonI'm off, nice talking to you guys.
02:03:08amiconnGoodnight, Jörg
02:03:13 Quit [IDC]Dragon ()
02:04:18 Quit pfavr ("ChatZilla 0.9.52B [Mozilla rv:1.6/1]")
02:08:01mattzzok, heading for some sleep, too - see you Jens
02:08:13amiconnGoodnight, Matthias.
02:08:20 Quit mattzz ("Client exiting")
02:29:15 Join TheGalik [0] (~galik@
02:29:15 Quit Galik (Read error: 104 (Connection reset by peer))
02:38:11 Part amiconn
02:40:40 Quit TheGalik ("Client exiting")
02:40:43 Join Galik [0] (~galik@
02:41:17 Quit Galik (Client Quit)
02:41:21 Join Galik [0] (~galik@
02:52:13 Quit Douche_Nutz (Read error: 110 (Connection timed out))
02:55:49 Join adi|home [0] (~adi|
02:57:54 Quit adi|home (Client Quit)
03:16:15 Quit BC|away (Read error: 104 (Connection reset by peer))
03:38:56 Quit AciD (Read error: 54 (Connection reset by peer))
03:47:05 Join BC [0] (
03:49:48***Saving seen data "./dancer.seen"
04:04:26 Join telliott [0] (telliott@
04:05:19telliottI'm confused about this
04:05:35telliottI'm not a programmer but I want to try this out.
04:06:17BCwhat is it?
04:07:08telliottRedefine Play while playing
04:07:15telliottA patch.
04:07:19BCoohh, dont know about that
04:07:29telliottRead the message.
04:07:36MTi still thinks its over the top
04:07:48BCyou mean you just wanna use it - or do other stuff To it?
04:08:01MTi dont want to redefine keys, i just want a dummy mode for when its not me using it
04:08:14MTBC: its the "party mode" from before
04:08:31telliottI just want to try it. I'm not a programmer. I'm hoping there's a custum build with the patch.
04:08:56MTnot very likely
04:09:33BCmt: thanks
04:09:43telliottI guess Lucas did a custom build for himself.
04:13:38BCI have a simple windoze devkit
04:15:38 Join Douche_Nutz [0] (apemanttt@
04:18:51BCI have had a number of requests for it and some positive feedback now - anyone wanna donate some web space?
04:20:51 Quit Galik ("Client exiting")
04:26:23 Part telliott
04:49:03BCdoes anyone know which C file prints "menu" on the file browser window?
04:49:04 Quit Nibbler (Read error: 104 (Connection reset by peer))
04:58:54 Quit Douche_Nutz (Read error: 60 (Operation timed out))
05:09:51 Join facted [0] (
05:11:35factedcan anyone point me in the right direction on how to install patches?
05:13:11scott666what OS?
05:14:30factedi installed cygwin
05:14:34factedbut i'm not sure where to go from there
05:14:48scott666did you find the docs on the rockbox page?
05:14:54factedthe patch ones?
05:14:57factedi read that...
05:15:03facteddidn't help very much
05:15:15factedany other good sites?
05:15:26scott666i tried a while ago, i never got it figured out
05:15:37factedyea...i gotta go, actually
05:15:40factedbut thanks for your help
05:15:44factedi'll try to figure it out
05:15:48scott666i suggest getting bluechips all-in-one dev kit
05:15:55scott666i can send it you, or BC can
05:16:20scott666It Works(c)
05:16:27scott666(except for the bugs)
05:16:47factedsure, send it to me please
05:16:52factedis e-mail ok?
05:16:56scott666its 7mb
05:17:16factedi'm not on a DCC client...using web irc
05:17:22factedftp? aim?
05:17:41 Quit facted ("Leaving")
05:17:48 Join facted [0] (
05:17:55factedsorry about that
05:19:01scott666i have aim
05:19:07factedok, i'm facted
05:19:14BCsorry - box didn't flash
05:19:36BCwas my presence desired?
05:19:58factedjust file transfer it to me?
05:20:00scott666he wants to install a patch
05:21:53factedok, thanks for the file...i'll install it and work on that
05:21:59factedunfortunately, i gotta go
05:22:03factedi'll try to come back tomorrow
05:22:16scott666you need a few other files too iirc
05:22:21scott666and instructions
05:22:55factedjust file transfer them to me?
05:22:58factedi have you set on automatic
05:23:04factedso i'll receive it all
05:23:09factedi thank you so much for your help...
05:24:03scott666thank BC, hes the one that created it
05:24:20factedthanks BC :)
05:25:07BCerr, ur welcome :)
05:25:12factedthanks :), i'm gone for real now
05:25:13factedlater guys
05:25:18 Quit facted ("Leaving")
05:25:24BCwhat was that?
05:25:52scott666i sent him your dev kit
05:26:01scott666and the .bat and profile
05:26:15scott666havnt told him what to do with it yet though
05:27:19scott666think you could type that up and put it in the same webspace as cards?
05:27:53BCI will sort out my webspace more sensibly later
05:27:55BCfor now...
05:28:40BC1. run the attached exe
05:28:40BC2. copy over the update patches
05:28:40BC ...minor fixes - mainly cosmetic
05:28:40DBUGEnqueued KICK BC
05:28:40BC3. run /DEVKIT/setup/setup.bat
05:28:40BC4. download a daily tarball
05:28:41***Alert Mode level 1
05:28:41BC5. put tarball in /DEVKIT/home/rockbox/
05:28:43BC6. unzip tarball "here"
05:28:45BC ...thus creating /DEVKIT/home/rockbox/tarball/apps
05:28:47BC If you have a folder called /DEVKIT/home/rockbox/apps you have screwed up!
05:28:49BC4. run rockbox.bat
05:28:51BC4a. if you are running win98 select for Simulator or AJZ build
05:28:53BC5. cd tarball
05:28:55BC mkdir mybuild
05:28:57BC cd mybuild
05:28:59BC ../tools/configure
05:29:01BC make
05:29:34BCthat's the text I'm sending out at the moment
05:32:36scott666im thinking im gonna have to completely start from scratch with OVERHAND
05:32:45scott666ill keep the random stuff and thats it
05:33:13BCi often do the same
05:33:21BCstill, it's only three lines (ish) of code
05:33:34BCI rewrote majot chunks of the card engine last night
05:34:00BCtechnically PIleCardCnt no longer exists - but it IS #defined ;)
05:34:04scott666i saw that
05:34:25BChe read my source :)
05:34:26scott666looks really nice
05:34:50scott666and the gfx!
05:35:43BCnot sure what the last u/l was
05:35:48BCtidying the gfx up no
05:36:03BCmy games mate came over tonight and critiqued
05:38:42***Alert Mode OFF
05:49:51***Saving seen data "./dancer.seen"
05:55:41BCyou got it yet?
05:56:21scott666no, i got disctracted by this guy i havnt talked to in like 2 years
06:26:52 Join Douche_Nutz [0] (apemanttt@
06:32:22 Join Nibbler [0] (
07:01:34 Quit Nibbler (Read error: 54 (Connection reset by peer))
07:11:00 Quit Douche_Nutz (Read error: 60 (Operation timed out))
07:16:04 Join adi|home [0] (~adi|
07:49:54***Saving seen data "./dancer.seen"
08:02:00 Quit scott666 ("i'll be back...eventually...")
08:21:15 Join amiconn [0] (
08:23:48 Join amiconn_ [0] (
08:29:17 Quit amiconn (Nick collision from services.)
08:29:17 Nick amiconn_ is now known as amiconn (
08:32:33 Join mattzz [0] (
08:32:48 Join Douche_Nutz [0] (apemanttt@
08:41:48 Join Nibbler [0] (
08:46:37mattzzs/morngin/morngin/ .... too tired
08:46:57mattzzanbody with CVS access here?
09:02:15 Quit Nibbler (Read error: 104 (Connection reset by peer))
09:10:26 Join monkey666 [0] (
09:10:38 Join [IDC]Dragon [0] (
09:11:45mattzz[IDC]Dragon: Good morning Joerg. Somehow mandelbrot does not show up in CVS
09:11:54[IDC]Dragonmattzz: what's the matter with cvs?
09:12:29[IDC]Dragonis it in the source tarball?
09:12:56mattzzhaven't checked, mompls
09:14:11[IDC]Dragonsourccforge brobably needs some time to digest it
09:14:13mattzzthere's a big time lag again
09:15:20mattzzok, I see
09:15:56[IDC]Dragongrayscale implemented? ;-)
09:16:28mattzzkind of.... ;-) I hacked in my dreams again. I hate that...
09:17:06[IDC]Dragonthen you shoudn't hack til 2am and be back in IRC at 8am
09:17:32mattzzthe good news is: I have the environment set up on my workstation @ work....
09:17:33[IDC]Dragon(neither me)
09:17:43[IDC]Dragonuh oh
09:19:01[IDC]DragonI'm not gonna do such
09:19:48mattzzit just happened
09:20:40BCdid anything happen with "plugin engine" code and directories yesterday after I left?
09:20:59BC(in terms of discussion, that is)
09:21:08[IDC]Dragonafaik, no
09:21:30mattzzthe question of using a directory for multiple plugin files?
09:21:51BCI thought /plugins/plugname and /plugins/frame/framename
09:24:03mattzzwhy not use /plugins/include for plugin framework files?
09:24:33BCbut to have a directory off /include for each framework
09:24:47BCthus allowing a framework to be in multiple files
09:25:23 Quit Douche_Nutz (Read error: 110 (Connection timed out))
09:40:45 Nick amiconn is now known as amiconn|away (
09:49:58***Saving seen data "./dancer.seen"
10:06:51 Quit [IDC]Dragon ()
10:21:06 Part BC
10:28:14 Join LinusN [200] (
10:45:53 Join lImbus [0] (
10:47:03 Join Nibbler [0] (
10:47:28 Join Douche_Nutz [0] (apemanttt@
10:54:17 Quit dwihno (Read error: 104 (Connection reset by peer))
10:54:18 Join dwihno_ [0] (~dw@
10:54:58 Join c0utta [0] (
11:08:29 Join [IDC]Dragon [0] (
11:10:49[IDC]DragonLinusN: while you're in that FAT area, how about the rmdir()? ;-)
11:12:05LinusNhehe, i just committed some very old changes i found on another computer :_)
11:12:42LinusNrmdir should not be different from the normal rm if the dir is empty. or?
11:13:20[IDC]Dragonno, delete() uses open(), which checks for dir attribute and fails if so
11:16:55LinusNi'm talking about fat_remove()
11:17:57LinusNhow far did you get with your rmdir patch?
11:18:42[IDC]Dragonit's all there, except the rmdir() itself working
11:19:01LinusNis it in the tracker?
11:19:06[IDC]DragonI made a rmdir() in dir.c
11:19:11[IDC]Dragonin cvs
11:19:39LinusNi don't see it
11:19:43[IDC]Dragononplay.c currently doesn't use rmdir(), this is commented out
11:19:57[IDC]Dragonoh, sorry.
11:20:21[IDC]Dragonthe rmdir() itself is not in cvs, I've sent you an email
11:20:41[IDC]Dragonfriday or so
11:20:47LinusNlemme check
11:22:12LinusNdamn i can't find it
11:26:55LinusNah, i looked in my private inbox, since you said "i sent you"
11:27:27[IDC]Dragonshould have gone there, I cc'd it to you and Björn.
11:27:48LinusNmy email filter removes duplicates
11:28:40[IDC]DragonOK, I won't try that again.
11:29:34 Join Misan [0] (
11:32:21[IDC]Dragononplay.c has all the code for the option, dirwalk and deletion.
11:32:37LinusNwhy do you call fat_truncate()?
11:32:57 Quit Douche_Nutz (Read error: 110 (Connection timed out))
11:33:11[IDC]Dragonthe remove_dir() helper function will need to call rmdir() then, see the comment there.
11:33:34[IDC]Dragonthat was my question as well, in the posting
11:33:56LinusNwell, we can't answer questions about your own code :-)
11:34:04[IDC]DragonI copied that from the remove() code, not knowing what it does
11:34:15LinusNremove() doesn't call fat_truncate()
11:34:49LinusNoh sorry, it does
11:34:50[IDC]Dragonline 257 of file.c
11:35:19LinusNvery unnecessary
11:35:37LinusNfat_truncate() frees unoccupied clusters
11:36:09[IDC]Dragonwhich have been reserved in advance to that file?
11:36:25LinusNthat is, if you (re)write to a long file with much shorter data
11:36:43[IDC]DragonI see.
11:37:01LinusNfat_remove() however, frees all clusters
11:37:15LinusNi'll look into it
11:37:27[IDC]Dragonmuchos gracias
11:47:10 Quit lImbus ("so long")
11:50:00***Saving seen data "./dancer.seen"
11:53:02 Join lImbus [0] (
12:12:05monkey666is it me or the CVS link removed from the website?
12:12:17monkey666*or is the
12:13:49Misanhi. i just found your website and found out, you won´t or can´t support gmini devices.
12:14:02Misanmaybe someone has heared of such a project?
12:14:42Misannevermind :)
12:15:01[IDC]Dragonsearch the list archive and the IRC logs.
12:15:17Misanok, i will
12:15:30[IDC]DragonThere's been some basic research, but iirc it didn't get as far as executing own code.
12:16:16[IDC]Dragonlook for "Telechip" (the CPU inside)
12:17:17LinusNmonkey666: it has been removed from the main menu
12:17:35LinusNyou find a link from the Download page
12:17:50[IDC]Dragon"Telechips", sorry
12:18:07monkey666LinusN: thank you
12:19:55[IDC]DragonMisan: the guy who was/is working on it is Laurent Giroud.
12:20:37[IDC]DragonAn IDE for that strange CPU was found on the web.
12:22:25monkey666i asked midk the other night if there was bookmark option for txt files... he said no is this (still) true?
12:23:31monkey666aw and i wanted to read my 1000 page book 12 letters at a time
12:24:16monkey666oh well i guess microsoft mary will have to read it to me instead
12:24:20 Join tboy [0] (~xxx@
12:24:48tboyidc dragon?
12:24:53tboyI have a question
12:24:58tboyabout the voice ui
12:25:10tboyI was hoping to port Icelandic
12:25:16tboyor make an icelandic voice file
12:25:25c0uttathanks for doing the cr/lf in cvs for me linus
12:25:27[IDC]Dragonnp, go ahead!
12:25:53tboybut I have the voice and the program that the voice comes along with says it uses sapi 4
12:25:59LinusNc0utta: thanks for pointing it out, we don't want cr/lf in the files, it was my mistake to commit that in the first place
12:26:04c0uttabut.. it made no difference!
12:26:20c0uttawhen i do a diff it is still showing every line
12:26:28[IDC]Dragonyou can use TextAloud for SAPI4 voices
12:26:39c0uttathen i patch −−dry-run and still get a HUNK error
12:26:45tboytextaloud ok I'll look into that
12:26:51 Quit monkey666 ()
12:27:10[IDC]Dragonmy tools have a converter to make an input file for it
12:27:21LinusNc0utta: delete action.h and update from cvs
12:27:27c0uttai have done that
12:27:41c0uttavery strange
12:28:20c0uttai have gone through with a hex editor too and even though some lines are the same they still get included in the diff
12:28:27LinusN c0utta: you do a diff against what?
12:28:40c0uttacvs diff -u action.h
12:28:54LinusNbut you deleted action.h?
12:29:23DBUGEnqueued KICK Misan
12:29:23Misan[IDC]Dragon: I found some spec sheets... i wil ltry to contact Laurent Giroud
12:30:34LinusNc0utta: make sure to remove all CR from your modified action.h
12:35:38 Quit Misan (Read error: 54 (Connection reset by peer))
12:36:32 Join Fragrag [0] (
12:36:41 Part Fragrag
12:40:51c0uttalinus: yes, done that too
12:41:15c0uttaused ultraedit to check the CR/LF and there's only LF
12:41:28 Quit [IDC]Dragon ("no fate but what we make (EOF)")
12:42:01 Join [IDC]Dragon [0] (
12:42:47 Nick dwihno_ is now known as dwihno (~dw@
12:55:23 Join Douche_Nutz [0] (apemanttt@
13:06:30c0uttain my diff it says:
13:06:37c0utta−−- action.h12 Mar 2004 10:17:19 -00001.1
13:06:37c0utta+++ action.h15 Apr 2004 09:29:03 -0000
13:07:03c0uttashouldn't the first line be 13 Apr 2004 since you updated it ?
13:08:41 Nick amiconn|away is now known as amiconn|work (
13:12:27 Quit mattzz ("CGI:IRC")
13:15:28 Quit mbr ("leaving")
13:20:58 Join mattzz [0] (
13:26:39tboy[IDC]Dragon: the voice came with this software called infovox
13:26:53tboythey say it uses sapi 4
13:27:16tboybut the voice seems to be fitted for their program only
13:30:42tboyI mean the voice file seems to be fitted
13:36:01 Quit mattzz ("CGI:IRC (Ping timeout)")
13:37:57 Join mattzz [0] (
13:41:58 Nick mattzz is now known as mattzz|afk (
13:50:02***Saving seen data "./dancer.seen"
14:03:43 Quit [IDC]Dragon ("no fate but what we make (EOF)")
14:26:28 Quit Douche_Nutz (Read error: 110 (Connection timed out))
14:31:53 Part tboy
14:39:28 Quit mattzz|afk ("CGI:IRC (EOF)")
14:43:54 Join AciD [0] (
14:45:03 Nick c0utta is now known as c0utta{zz} (
14:46:31c0utta{zz}linus: i have submitted my latest patch
14:47:52LinusNi see it, iäll look into it later
14:48:01LinusNgotta go now
14:48:03 Part LinusN
15:42:32 Join Douche_Nutz [0] (apemanttt@
15:50:04***Saving seen data "./dancer.seen"
15:53:48lImbushi, where can I find that lang.h that is needed to compile uisimulator@win32 ?
16:02:21 Join quelsaruk [0] (~fgjfg@
16:06:38amiconn|worklImbus: lang.h is dynamically generated while compiling
16:09:01lImbuscompilin of uisim or the archos firmware ?
16:14:03 Quit AciD (Read error: 104 (Connection reset by peer))
16:14:18lImbusrockbox itself doesn't compile either
16:14:32lImbus- at least on my machine
16:16:59quelsarukcompiling is building the source
16:17:03quelsaruki have to go
16:17:07 Quit quelsaruk ()
16:19:10lImbussorry, I know what building and compiling means and is, but it doesn't work
16:28:37lImbusok, I found that rockbox.dsp, but it cant be converted by my VS.NET2003 :-(
16:32:40CtcpIgnored 3 channel CTCP requests in 13 hours and 59 minutes at the last flood
16:32:40*lImbus is listening to 2. Lustmord - Heresy Part II
16:32:59lImbuserr. really not yet familiar with irc ore one of its clients
16:40:10amiconn|worklImbus: Don't know if may be able to compile the simulator, but it definitely isn't able to compile for the target, since doing that requires a cross compiler for SH1.
16:42:12amiconn|workI use cygwin for compiling both the sim and for the target. There is some howto on the rockbox docs page.
16:44:46lImbusI did all that. installed cygwin, cvs up, ... here what I get when I try to compile rockbox (the second time, so you don't see deps)
16:45:16lImbusmake[2]: *** [scramble] Error 255
16:45:24lImbusmake[1]: *** [../tools/convbdf] Error 2
16:45:30lImbusmake: *** [firmware] Error 2
16:46:24lImbusuisim seems to be compileable with was there are vcproj-files
16:47:34amiconn|workperhaps try the following: cd into <your_rockbox_source_dir>/tools, then do make. This should compile the tools (only). If it doesn't work correctly, try again after a "make clean".
16:49:05lImbusMDJ@LAPTOP-MDJ ~/tools
16:49:05lImbus$ make
16:49:05lImbusgcc -O -ansi scramble.c -o scramble
16:49:05DBUGEnqueued KICK lImbus
16:49:05lImbusmake: *** [scramble] Error 255
16:52:17amiconn|workQuote from the ml archive:
16:52:18amiconn|work> make[2]: Entering directory `/home/rockbox/tools`
16:52:18amiconn|work> gcc -0 -ansi scramble.c -o scramble
16:52:18DBUGEnqueued KICK amiconn|work
16:52:18amiconn|work> make[2]: *** [Scramble] error 255
16:52:18amiconn|work> . . .
16:52:19***Alert Mode level 1
16:52:19***Alert Mode level 2
16:52:19amiconn|work> So I tried running gcc to see if I could invoke it manually.
16:52:21amiconn|work> bash: gcc: command not found
16:52:23amiconn|workThis is because the "win32-sdk" doesn't include the plain vanilla gcc, only
16:52:25amiconn|workthe sh-elf one. That's why the page tells you to download and install that
16:52:27amiconn|worktools package separately (instead of building it yourself).
16:54:06amiconn|workOf course you can, instead of downloading the tools binaries, also install the native GCC in cygwin (with the cygwin installer).
16:54:23lImbusI downloaded it separately, then unzipped it and so on. I didn't want to make the tools myself, that was your idea. before, I had the same probs compilin the rockbox itself, that should work
16:54:28amiconn|workThis is what I did; it also enables you to compile the simulator with cygwin.
16:55:00lImbusthe only one real prob is I am running out of hd-space on my notebook
16:55:32amiconn|workThe problem is: if the tools' sources are newer than the compiled binaries, it tries to recompile them (that's what make is about).
16:56:15amiconn|workI could send you the current tools binaries.
16:57:05lImbusahh. for win32 ?
16:57:44lImbusAm I supposed to know how to run that DCC-thingy now ? ICQ ? FTP ? mail ?
16:58:06amiconn|workYes, for win32.
16:58:10 Join mecraw_ [0] (~mecraw@
16:59:16amiconn|workTell me the method that suits you best. The tools are 59 KB total (zipped).
16:59:38lImbushttp or ftp would be nice
16:59:43lImbusvery nice
17:00:16amiconn|workHttp then. Will take some minutes, stay tuned.
17:00:34lImbusnp, thanks in advance
17:02:20***Alert Mode OFF
17:03:59 Quit dwihno ("Ska bygga en rymdraket av cornflakespaket")
17:05:34lImbusthanks, got it
17:06:18lImbushydrairc doesn't let me copy&paste not click on that url over there. very annoying. feel back like in bbs-times.
17:07:40amiconn|workI'm also using HydraIRC. Just double-click the link, this works for me
17:11:39lImbusoh. thanks for the clue
17:11:54lImbusbtw: can't still compile rockbox...
17:12:06lImbusMDJ@LAPTOP-MDJ ~/build-dir
17:12:07lImbus$ make
17:12:07lImbusmake -C ../firmware TARGET=-DARCHOS_RECORDER NODEBUG=1 OBJDIR="/home/rockbox/build-dir" MEM=2 TOOLSDIR=../tools
17:12:07lImbusmake[1]: Entering directory `/home/rockbox/firmware'
17:12:07lImbusmake -C ../tools
17:12:08***Alert Mode level 1
17:12:08lImbusmake[2]: Entering directory `/home/rockbox/tools'
17:12:10lImbusgcc -O -ansi scramble.c -o scramble
17:12:12lImbusmake[2]: *** [scramble] Error 255
17:12:14lImbusmake[2]: Leaving directory `/home/rockbox/tools'
17:12:16lImbusmake[1]: *** [../tools/convbdf] Error 2
17:12:34lImbusI suppose he's still trying to update the tools. can't that be avoided ?
17:12:48amiconn|workGrr, it still tries to compile the tools.
17:13:47amiconn|workDid you cope the tools into the "tools" directory?
17:14:45lImbusI moved away what was in my /home/tools/ before, then unzipped your pack there.
17:14:49 Join pfavr [0] (
17:15:30amiconn|workIt needs to be in home/rockbox/tools, along with the sources and all.
17:16:21lImbussorry, I misunderstood the " ~/tools
17:16:33 Join methangas [0] (
17:16:35lImbus" on the page. I thought ~ is home
17:17:08lImbusnice, the compiler is glowing
17:17:48lImbusfuck, it crashes with a not found reentrypoint in cygwin.dll
17:18:34lImbusprobably we havn#t got the same version of cygwin, as it happens in the convbdf.exe
17:18:47lImbusim trying to revert to the original tools
17:19:47lImbusbuild resumed ;-)
17:19:51amiconn|workOpps, maybe this is the reason, I use a fairly recent cygwin installation. I could send you my current cygwin.dll as well, if reverting to the original tools doesn't help.
17:20:34lImbusit seems to work. I can't figure out if he restartet the whole build process or if he's able to resume
17:21:15lImbusnice, it did
17:21:29lImbusthanks alot
17:21:38amiconn|workmake is really clever: everything it already build doesn't get rebuilt until the source or some other file it depends on is changed.
17:21:52 Join elinenbe [0] (
17:22:08lImbusif make is so clever, what is autobuild for ?
17:22:09***Alert Mode OFF
17:22:37amiconn|workI dunno, haven't dealt with it yet.
17:24:03lImbusall I heard about is a silly. some told me nightmares that even autobuild is not able to build autobuild itself.
17:25:17lImbusuisim still doesn't compile, but I guess this is because I messed it up. I exe
17:25:32lImbusI executed a perl script by hand to get that lang.h...
17:25:42amiconn|workAs your sdk installation is now working, you can start developing for rockbox. Unfortunately you won't be able to compile the uisim with it unless you install native gcc.
17:26:40amiconn|workSorry, can't help with compiling the uisim on, since I don't have that.
17:28:10lImbusah well, ok, I'll make my first steps on the hardware directly
17:29:40 Join facted [0] (
17:41:52elinenbeamiconn|work: any work on the minesweeper clone yet?
17:42:26amiconn|workStill nothing done... I was tweaking grayscale.c instead.
17:43:21elinenbeamiconn|work: yeah, I saw that on the patch tracker
17:44:26amiconn|work..and giving tips for optimization to mattzz (mandelbrot.c)...
17:45:47elinenbeis the mandelbrot plugin heading to grayscale?
17:46:00elinenbeI just had to rebuild my cygwin enviroment yesterday....
17:46:47elinenbewhat a pain... recompiling everything.
17:47:40amiconn|work?? Did you compile cygwin yourself?
17:48:01elinenbeamiconn|work: no −− I did not recompile cygwin... I just downloaded it
17:48:14elinenbebut I did compile binutils, gcc, gdb, perl, etc...
17:48:22factedcan anyone help me apply a patch?
17:48:30elinenbeI have a nice clean enviroment again...
17:48:33 Join Galik [0] (~galik@
17:49:09factedI can compile the rockbox source, but I can't figure out how to apply the patch...I downloaded the "patch" utility
17:49:26elinenbefacted: what patch are you trying to apply?
17:50:06***Saving seen data "./dancer.seen"
17:50:13factedit's the play button patch
17:50:15factedit's a diff file
17:50:21factedhold one sec.
17:50:37amiconn|workI'm off for home. cu
17:50:38elinenbejust type something like "patch -po -i patch_filename
17:50:42elinenbeamiconn|work: later.
17:50:51 Nick amiconn|work is now known as amiconn|away (
17:50:51DBUGEnqueued KICK amiconn|away
17:51:49elinenbedownload the patch (the diff) file
17:51:56factedpatch_filename being the name of the patch_diff file?
17:52:16factedok, i have to set my "HOME" directory too for Patch to work right?
17:52:25elinenbeyou shouldn't need to.
17:52:35factedlet me try
17:53:08elinenbeyou can just copy the patch into your rockbox source dir.
17:53:24elinenbewhere the are subdirs apps, fimrware, flash, gdb, etc....
17:53:26factedi've done that
17:53:31factedi copied it to apps
17:53:46elinenbethen in that folder run "patch -p1 -i 00_patch_887081_play_button.diff"
17:53:48factedand then i run the patch command you gave me from that diretcory?
17:53:52elinenbedo not copy it to apps.
17:53:59elinenbethe parent directory of apps
17:54:32elinenbeto rockbox/ not rockbox/apps/
17:54:42elinenbethat should do it...
17:54:52 Quit pfavr ("ChatZilla 0.9.52B [Mozilla rv:1.6/1]")
17:54:54facteddidn't work
17:55:00factedpatch isn't a recognized command :(
17:55:06factedi guess I have to set my home directory...
17:55:14elinenbeyou just need patch in your path.
17:55:24factedwhat's the name of the patch utility?
17:55:35factedthere is no "patch" command file or anything in the patch set I downloaded
17:55:43factedyea, there is none though... :(
17:56:06factedis that ok?
17:57:27factedeven after copying patch.c and to the directory, the command still results in a "patch is not recognized, etc..."
17:59:02factedi did it again using my cygwin environment and it worked, but 1/6 of the patch was "rejected" :(
18:04:19 Quit Douche_Nutz (Read error: 110 (Connection timed out))
18:08:12elinenbeyou need to either download a compiled version of patch, or compile it.
18:08:21elinenbethat is just the source and the manual you have there.
18:08:36elinenbedoing it in cygwin seemed to work okay.
18:08:48elinenbeyou my have to grab the latest CVS and then do that patch on that.
18:10:03factedI got the latest cvs
18:10:12factedthanks, I already got it to work :)
18:10:33factedThank you for your help and your patience though
18:10:55factedoh, one question:
18:11:15factedonce I patched the files necessary, what files do I need to move into my archos recorder?
18:11:37factedthe rockbox.ucl and the main file (*.ajz)?
18:12:17elinenbeif you don't have scramble in your path then rockbox.ulc will be 5 bytes
18:12:31elinenbeyou will probably just want to compy *.ajz onto the recorder.
18:12:50elinenbethen "run it" by pressing play on the file.
18:12:52factedyea, rockbox.ucl is 5 bytes
18:12:58factedgot it
18:13:00factedthanks again!
18:14:15 Join AciD [0] (
18:16:22 Quit facted ("Leaving")
18:17:28 Quit c0utta{zz} (Read error: 54 (Connection reset by peer))
18:18:13 Join c0utta{zz} [0] (
18:24:35 Quit jkerman (Read error: 54 (Connection reset by peer))
18:40:53 Nick amiconn|away is now known as amiconn (
18:40:53DBUGEnqueued KICK amiconn
18:51:30 Quit mecraw_ (Read error: 54 (Connection reset by peer))
19:06:35 Quit lImbus (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
19:11:03 Join facted [0] (
19:14:10 Join mecraw_ [0] (~mecraw@
19:21:06 Quit facted ("Leaving")
19:21:16 Join Guest [0] (
19:26:12 Quit Guest (Client Quit)
19:30:07 Join Douche_Nutz [0] (apemanttt@
19:30:49 Quit mecraw_ ("Trillian (")
19:43:20 Join amiconn_ [0] (
19:43:30 Quit Galik ("Client exiting")
19:50:10***Saving seen data "./dancer.seen"
19:57:30 Join amiconn__ [0] (
19:57:56 Quit amiconn (Nick collision from services.)
19:58:10 Quit amiconn_ (Nick collision from services.)
19:58:20 Nick amiconn__ is now known as amiconn (
19:59:39 Join mecraw_ [0] (~mecraw@
20:06:36 Quit Douche_Nutz (Read error: 110 (Connection timed out))
20:09:09 Join facted [0] (
20:12:09factedif I get the error message: "can't find file to patch at input line 10", and I've tried -p0, -p1, and -p2, is the patch faulty or am I doing something incorrectly?
20:15:27 Join pvh [0] (PvH@
20:16:01pvhWhat do you guys think of a simplified play structure?
20:16:13pvhIt could be based on the F2/F3 menus.
20:16:22pvhPress Play, default play behaviour occurs.
20:16:38pvhHold down play, the pop-up menu appears (which I believe has pitch right now).
20:17:05pvhUp: Queue Next, Down: Queue Last, Left: Insert Next, Right: Insert Last.
20:17:25pvhFile-system stuff goes to On-Play for rename/delete.
20:17:42factedi like the customizable patches out right now
20:17:45pvhPlay on a directory, by default, plays it.
20:17:56pvhForward on a directory descends into it.
20:18:01factedyou can customize what play does, and what foward does
20:18:09pvhWhat's the patch number?
20:18:17factedi like that last suggestion...about the playing a directory
20:18:18factedhold on one sec.
20:18:20CtcpIgnored 4 channel CTCP requests in 1 hour and 51 minutes at the last flood
20:18:20*pvh is thinking more about what a good UI design would be.
20:19:01factedi do like your idea about holding the play button down for more options
20:19:05factedrather than for pitch
20:19:23pvhIt's intuitive, I figure.
20:19:31pvhButton-down, wait for a moment, then bring up the menu.
20:19:45pvhQuick down-up, and you get default behaviour.
20:19:54amiconnpvh: play + left/right/down is not possible at the same time
20:20:16pvhOkay, well, once the menu is open, just leave it up until a key gets hit.
20:20:28factedyea, that would make sense
20:20:39factedhold play on a song opens up the menu
20:20:40pvhIf play on a dir plays the directory, then play _in_ a dir could just play the song.
20:20:45amiconnThere are 2 groups of 4 keys each, where it is not possible to reliably detect pressing 2 or more of them at the same time.
20:21:28factedpvh, is this for a patch?
20:21:36pvhYeah, I'm not doing anything else today.
20:21:39amiconnGroup 1: f1/f2/f3/up, group 2: left/play/right/down. This is why there is no "up" in the quick screens.
20:21:51pvhAh soo.
20:21:57pvhI had wondered about that.
20:22:03pvhOff is a group of its own?
20:22:25amiconnOn and off are each connected to a port of its own.
20:22:50pvhWell, that shouldn't change too much.
20:23:11factedpvh, if you get it done today, will you pleast post the link on the newsgroup?
20:23:17factedi'd love to test this out for you
20:23:20pvhYes, of course I will.
20:24:21pvhWhere's the pitch menu moved to?
20:25:59factedsomewhere else :)
20:26:04factedi don't know
20:26:12pvhThat _was_ how you called it up before, right?
20:26:39factedbut I think this ability to more easily manipulate the playlist is much more important than the pitch menu...
20:26:58pvhI was going to look for it in the files to orient myself on the code.
20:27:12factedyou should also check elinenbe's patch for the WPS
20:27:21pvhWill do.
20:27:28factedpressing f3 would allow you to see the playlist...
20:27:39factedwhich isn't possible in any other way yet
20:27:41pvhAh yes, I think that's a good idea too.
20:27:43factedthat i knwo of
20:27:51pvhIt's like a two-liner patch, as well.
20:28:19factedi tried to apply it but I got some errors and I'm a newbie with the patches, but if you can incorporate that with your new patch...
20:28:34factedThat would do tons for the usability of the playlist
20:30:51pvhI'll probably keep my patch seperate.
20:31:19factedi sent you a private chat request...
20:33:59 Join monkey666 [0] (
20:37:04 Join quelsaruk [0] (~r00t@
20:40:45 Join LinusN [200] (
20:40:56pvhHi Linux.
20:41:02quelsarukhi LinusN
20:41:02pvhEr.. Damn finger reflexes.
20:41:02amiconnhi LinusN
20:41:06quelsarukthe man i was looking for
20:41:27LinusNpvh: the pitch menu is where it always have been, pressing the ON key in the wps
20:41:47pvhAh, I was suffering from a temporary lapse of sanity.
20:41:47LinusNpvh: i even write linux myself sometimes :-)
20:41:54amiconnquelsaruk: Nick completion is really a nice feature ;-)
20:42:01quelsarukyeah amiconn :)
20:42:02LinusNamiconn: yup
20:42:28pvhI use mIRC without any bells or whistles. I guess that makes me either hardcore or softcore, depending on how you look at it.
20:42:33quelsarukthe problem is when 2 people has a similar nick and you say something to the wrong man :)
20:42:47 Join jkerman [0] (
20:42:56pvhLinusN: Did you see my ideas about the play menu?
20:43:09LinusNxchat won't let you complete if there is any ambiguity
20:43:32quelsarukxchat show you all posibilities, no? afaik
20:43:54amiconnHydraIRC cycles through all possible matches.
20:44:21LinusNpvh: i don't like to hold a key for a menu to pop up
20:44:37LinusNthe extra delay is annoying
20:44:53pvhOkay, so make it momentary.
20:45:31LinusNthen it becomes more difficult to start the music
20:45:41LinusNi like to press Play to start
20:45:49pvhIt's simply a matter of finding the right timing on the button.
20:46:08LinusNthe concept we are working on is to use f3 for a context sensitive menu
20:46:25LinusNand f2 for a configurable menu
20:46:39pvhI think that's a good idea, but it's hard to explain those to people.
20:46:43LinusNand get rid of all 2-key combinations
20:47:03pvhI do think that's a good idea.
20:47:12LinusNwhy is "hold play for a second" easier than "press f3"?
20:47:49pvhBecause it looks like a play button.
20:47:52quelsaruki think, holding a button is not a good aproach
20:48:08*pvh shrugs.
20:48:18pvhI'll make a patch, and people can try it and tell me what they think.
20:48:31amiconnLinusN: I'm all for the idea of configurable & context sensitive menus, but some of the 2-key combinations are also _very_ handy...
20:48:49pvhOn-Up/On-Down for example
20:49:06LinusNwe have an idea of configurable "hot keys" as well
20:49:15amiconnpvh: I never use this (assume it is pitch).
20:49:17quelsaruk'cause some buttons are *slow*, i had that problem on my old recoerder, and sometimes using combos was awful, when pressing ON+PLAY, i pressed just PLAY, and so on...
20:49:43LinusNthe goal is to be able to operate rockbox with only one hand
20:49:48pvhAmi: No, it's paging through directories.
20:50:34LinusNpvh: in the dir browser, yes, but it is pitch in the wps
20:50:36amiconnAhh, in the dir browser it is paging. Didn't know that yet.
20:50:43pvhAmi: It's a good-un.
20:50:54pvhLinusN: Yeah, but I like the paging, and I don't like the pitch shift. ;)
20:51:04pvhLinusN: s/like/use/
20:51:25amiconnThe 2-key combo I'm using most is on+play.
20:51:26LinusNwe can still keep the paging, it has nothing to do with the context sensitive menus
20:51:37LinusNon+play will be moved to f3
20:52:20monkey666i think if i can play halo or mario sunshine i can press on+play
20:52:48LinusNwith one hand?
20:52:52amiconnLinusN: I think context sensitive menus will _require_ the button bar, otherwise the user may get very confused.
20:53:13LinusNamiconn: yes, that's why i implemented it
20:53:17quelsarukLinusN, i've been disconnected for 15 days... are there any changes/new things that i should have in mind for the splash menus? I had nearly 800 mails from rockbox and had to delete all them without reading or die in the try :(
20:53:48LinusNquelsaruk: no, most changes have been bug fixes or talkbox stuff
20:54:08pvhWhat do you guys think of the playlist F3 patch?
20:54:11pvhHas anyone else tried it?
20:54:22LinusNplaylist f3 patch?
20:54:32pvhJust replaces the F3 menu with the Playlist browser.
20:54:46LinusNi haven't tried it and i don't think i will
20:54:58monkey666i dont about your halo skills but i can play halo nearly on-handed
20:55:09amiconnLinusN: I noticed a minor bug with the button bar. For now it is supposed to work only in the dir browser.
20:55:22LinusNthere are lots of issues with it
20:55:38LinusNit is on in the plugin/font/lang browser for instance
20:56:02pvhYes, it's quite a preliminary patch.
20:56:20LinusNpvh: the playlist f3 patch is not interesting because we will have a context sensitive menu on f3
20:56:33amiconnYes, that's what I've noticed, and it is confusing, because the keys don't work then.
20:56:55LinusNamiconn: yup, it's on my (mile long) todo list
20:57:40LinusNgotta go away for a while, i'll be back
20:59:26quelsarukbtw, LinusN, i have 2 bugs/issues to report (i suppose you know them right now)
21:00:17quelsaruk1st: Inserting directories to a playlist is quite slow... reaaaaally slow, is it possible to make it be as fast as the create playlist function??
21:00:26 Join mecraw__ [0] (~mecraw@
21:01:08quelsaruk2nd: is it planned to kill the fade off/on/off when pressing STOP/PAUSE while the re-filling the buffer??
21:08:06 Nick AciD is now known as AcyD (
21:12:45 Nick AcyD is now known as AciD (
21:17:46 Join HenrikB [0] (
21:17:49 Quit mecraw_ (Read error: 110 (Connection timed out))
21:19:39quelsarukdinner time, cu!
21:21:14 Quit quelsaruk ()
21:25:30 Quit monkey666 ()
21:31:15 Join Douche_Nutz [0] (apemanttt@
21:34:00 Join Guest [0] (
21:35:04 Quit Guest (Client Quit)
21:46:05 Join scott666 [0] (
21:50:14***Saving seen data "./dancer.seen"
21:54:27 Quit methangas (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
21:55:48 Quit mecraw__ ("Trillian (")
21:56:16 Join cjnr11 [0] (
21:57:00 Part cjnr11
22:06:31 Quit facted ("Leaving")
22:07:14 Quit AciD (Client Quit)
22:32:11 Join Galik [0] (~galik@
22:52:17 Join solaris [0] (
22:54:20solarishello, I'm glad to be here
22:54:42 Quit Douche_Nutz (Read error: 110 (Connection timed out))
22:55:09solarisi need some help to revive my ajbr 20
22:56:35solarisI had rockbox 2.1 firmware flashed and wanted to upgrade the firmware to 2.2
22:57:25solarisand after using rockbox.url, it gave me some errors such as "imcompatible, rolo error, checksum error' and now....
22:57:38solariswhen i press ON it doesn't start
22:58:18solarisnothing happens 'except green led light
22:58:44amiconnWhat happens if you press F1+On to switch it on?
22:58:50solarisusing F1+ON will load up 1.40a
22:59:22amiconnSo the boot loader is ok, you just flashed an .ucl for the wrong Archos model.
22:59:43solarisi tried changing ajbrec.ajz
22:59:48solarisbut no go
23:00:04solarisi'm pretty sure i had the right version
23:00:23solarisbut it kept saying 'incompatible version at one point
23:00:56amiconnEasiest way to try: download Rockbox 2.2 (just double-check if it is for the right version), unzip it to the root of your Archos,..
23:01:31solarisi tried that... but no go, i'll try again
23:01:48amiconnthen after unplugging the USB restart the unit with F1+ON, This should boot into the Rockbox v2.2 .ajz _via_ the Archos v 1.28 ...
23:02:13solarishmmmm.... ok
23:02:15amiconnThen you can flash the correct .ucl
23:02:23solarisi have 1.40a version when i start
23:03:44amiconnThere are now pictures on the download page to spot the right model. If yours says v 1.40 then chances are it is either an FM Recorder or a Recorder V2.
23:04:12solarisit's a v2 without fm
23:04:36amiconnThen the V2 package definitely should work.
23:05:06amiconn(I own a Recorder v1 btw. That's why I mentioned v 1.28)
23:06:11solarisok, file copied and restarting
23:06:53solarisstill same thing..... no go
23:07:14solarisi'm in 1.40a, no rockbox
23:07:49amiconnHmmm. You said that while trying to flash it told you about checksum errors.
23:07:50solarisshould i try recorder? instead of v2?
23:08:47amiconnThis means that the .ucl was corrupted. Chances are that your .ajz is also corrupt. Did you re-download the package?
23:09:18solarisand directly copied from zip to the root
23:09:54solaristhat's what i said....
23:10:11 Join lImbus [0] (
23:10:53amiconnCould you send the .ajz _as read from your Archos' root_ to me? I could have a look.
23:11:05solarishold on
23:11:13solarislet me try the non-v2 version
23:11:43amiconnPerhaps the file corruption occurs while copying to your Archos. (Faulty USB port/cable/driver etc.)
23:12:42 Quit HenrikB ("Lämnar")
23:12:58amiconnOf course you may try v1, as well as FM. This only doesn't boot into Rockbox if you get it wrong.
23:13:38amiconnsolaris: Got it.
23:14:42solarissame problem with v1
23:15:11solarisso, the problem is that it won't run ajbrec.ajz
23:16:08solarisi can't see the .rockbox from 1.40a nor ajbrec.ajz, is that right?
23:16:25amiconnGood news: the ajbrec.ajz you sent is the genuine V2 version of Rockbox 2.2, and it is not corrupted.
23:16:25solariscan you see ajbrec.ajz from archos firmware?
23:16:52solarisnow, the bad news???
23:17:14amiconnNo, you can't. If it is there, it will be loaded on boot, if it is the correct version.
23:17:27amiconnDid you try the FM version?
23:17:32solarisnot yet
23:18:29solarisok, copied
23:19:46solarisnope, it stayed in archos
23:19:50amiconnHmm, just remembered that the Archos fw _sometimes_ has problems to find the correct ajbrec.ajz, if there are either old directory entries left over or there is some other file in the root whose name starts the same.
23:20:15amiconnYou can try the following (assuming you use windows)
23:20:55amiconn(1) Put back the V2 version of rockbox to archos root.
23:21:46amiconn(2) Delete all files from the root which are not absolutely necessary, especially those which start with ajb....
23:22:02amiconn*all _other_ files of course
23:22:23amiconn(3) Perform a scandisk, to make sure.
23:23:02amiconn(4) Perform a defrag (make sure you batteries are in a good shape, especially if you don't have USB 2.0).
23:23:16solarisi have over 15 gig of mp3 and audio books
23:23:21amiconn(5) Retry to reboot with F1+ON
23:24:03pvhsolaris: You don't have to delete all the files
23:24:11pvhsolaris: Just files in F:\
23:24:31amiconnStep (4) may take quite some time then, but it is one method to get rid of "dead" directory entries.
23:25:58amiconnThere is at least one other method that was mentioned some time ago, have to look up...
23:29:33amiconnsolaris: Look at the faq (, Question 64 contains some suggestions.
23:30:35 Join facted [0] (
23:30:56factedhow's the patch going pvh?
23:31:11pvhdoing some real work at the moment
23:31:24factedi guess that's acceptable :)
23:32:07solarisok, i'll try that, thx amiconn, brb
23:32:18pvhlinus doesn't seem to like the idea, but that's okay. if it works for me, that'll be good enough. ;)
23:32:27factedyea, i read some of the conversation
23:32:35factedi agree with you, to each his own...
23:32:45factedwho knows when the customizable menus will be ready anyway
23:32:49factedin the meantime, the patch is great
23:33:09amiconnLinusN: Are you around?
23:34:51 Join Quelsaruk [0] (~Swordmast@
23:35:07Quelsarukhi again
23:38:10Quelsarukamiconn, it was impossible to test your build
23:38:29amiconn? what? which? why?
23:40:04Quelsaruki mean, i left my old jukebox in my house, so as i went to the beach.. no testing :) and now.. i'm back in Portugal and the box is in Spain :(
23:40:20Quelsarukthe fast write/read build
23:40:34 Join Bagder [0] (
23:40:50amiconnAhh, I remember. This was quite some time ago.
23:41:02Quelsaruk15 days ago
23:41:04Quelsarukmore or less
23:41:14Quelsarukall the time i've been away from tech :D
23:41:17Quelsarukhi Bagder
23:41:39amiconnMeanwhile I did post a grayscale framework, within a demo plugin (well, the demo itself is not very amazing).
23:42:01 Join danhans224 [0] (
23:42:40 Quit facted (Read error: 104 (Connection reset by peer))
23:43:02 Join Douche_Nutz [0] (apemanttt@
23:43:52danhans224is there anyone here that worked on the rockbox?
23:45:41danhans224how did you all get started on this project, any skills you need to know?
23:46:04QuelsarukBagder, maybe you want this info: With 2100mAh batts, my box lasted for nearly 9 hours, doing a full disk (20GB playlist) random play (non-stop) with aprox. 50% CBR 192kbps, 20% CBR 128 and 30% VBR with an average bitrate of 190.
23:46:17danhans224it's so cool, just wondering how you all did it
23:46:45Bagderdanhans224: lots of hard work
23:46:56Bagderand many skilled people
23:47:25danhans224i know.. what type of stuff you have to learn to do this?
23:48:03Bagderyou mean to rewrite a firmware from scratch on an average consumer mp3?
23:48:12Bagdermp3 player
23:48:32Bagderelectronics, low level programming and higher level programming
23:48:57danhans224any specifics? anywhere to learn these skills?
23:48:59Quelsaruka techno priest supporting you is optional, but helps a bit ;)
23:49:22BagderI started learning on my c64
23:49:25Bagder20 years ago
23:50:15***Saving seen data "./dancer.seen"
23:50:36danhans224i'm new to programming and electronics.... the only ancient os i worked with was dos
23:50:46pvhYou mean CPM.
23:50:59pvhCPM was DOS before MS bought it.
23:51:25danhans224didn't know that... probably before i was able to walk... haha
23:52:17Bagdersleep is good
23:52:18 Quit Bagder ("")
23:58:35lImbusQuelsaruk: would rockbox have been possible if archos didn't build in a update-from-disk - method ?
23:58:47lImbusi mean without soldering out and burning the flash

Previous day | Next day