**** BEGIN LOGGING AT Sun May 19 13:32:37 2002 --> adi|home (~adiamas@as5300-9.216-194-23-128.nyc.ny.metconnect.net) has joined #rockbox --- Topic for #rockbox is Does your box rock? http://bjorn.haxx.se/rockbox/ --- Topic for #rockbox set by Bagder at Mon May 13 04:46:28 --> pippoblu (pippoblu@TVT-CaTV-dhcp-45-101.urbanet.ch) has joined #rockbox <-- pippoblu (pippoblu@TVT-CaTV-dhcp-45-101.urbanet.ch) has left #rockbox --> [TDM]Mr_B|away (~nx_2000@ip68-0-62-70.no.no.cox.net) has joined #rockbox --- [TDM]Mr_B|away is now known as [TDM]Mr_B <-- [TDM]Mr_B has quit () --> wavey (~wavey@dlan1431.dircon.co.uk) has joined #rockbox <-- wavey (~wavey@dlan1431.dircon.co.uk) has left #rockbox --> elinenbe (elinenbe@adder.engin.umich.edu) has joined #rockbox --- Tumm is now known as tumm hello there --> martin__ (~martin@pD9E7A37E.dip.t-dialin.net) has joined #rockbox <-- martin__ (~martin@pD9E7A37E.dip.t-dialin.net) has left #rockbox ("Client Exiting") <-- elinenbe has quit ("Leaving") --- tumm is now known as Tumm --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> fall3 (~francis@nas-cbv-9-62-147-161-190.dial.proxad.net) has joined #rockbox hi fall3 hi <-- fall3 (~francis@nas-cbv-9-62-147-161-190.dial.proxad.net) has left #rockbox ("Fermeture du client") ack!!! ppl... make them leave!!!! haha did the changes i made to the rockbox make sense? btw oh.. and none of them were reflected on the windows side... the credit stuff? or do you mean the archos dir? my !"#$%&/(/&%$# display seems broken :-( nothing i do works <-- Zagor has quit ("bye") * adi|home nods knowingly --> motiv01 (~trillian@sdn-ar-001ncraleP018.dialsprint.net) has joined #rockbox <-- motiv01 (~trillian@sdn-ar-001ncraleP018.dialsprint.net) has left #rockbox --- dw|gone is now known as dwihno --> Bagder (~daniel@as3-3-2.ras.s.bonet.se) has joined #rockbox HELLÖ! :D * dwihno vinkar --> calpefrosch|work (~calpefros@62.52.174.30) has joined #rockbox morning morning Höj höj I sold one of my computers last night So now I got to spend my money on archos stuff ,D cool yup * Bagder drowns in his email flow * dwihno sends some more mail to Bagder 10: print "hello" 20: goto 10 ack!!! <-- calpefrosch|work has quit (Read error: 104 (Connection reset by peer)) ppl here... no.. ppl arent allowed here... hey adiamas its been _my_ room for like 3 days... GO AWAY!!!!! hehe this is _my_ room!!!! heheh * Bagder boos have you seen the credits list on the uisim? I saw the code on the cvs list, that's all yet would work fine if i wern' for you sweds and your funny looking letters ah right Björn should change his name ;-) damn right... don't see me interjecting cyrillic letters in my name do ya? not yet ;-) as an exampel :) that and i fixed the splash sceen a bit you know.. really "useful" stuff.. --> Zagor_ (~bjst@labb.contactor.se) has joined #rockbox --- Zagor_ is now known as Zagor Dr. Zagmaster! woo there we should make the credits scroll hey Zagor The splash screen should be configurable - think about ppl wanting a quick startup nah.. ppl deal with the spash screen now.. while the firmware is "loading" spending precious CPU cycles with scrollers?! :-D and im sorry.. if you in a such a rush you can't wait 200 milliseconds.. Do you realize the insanity!? :) yes! 200 msec's is 195 too many! hehehe and fortanetly... 100 milliseconds faster then most of us think ;) http://www.hackles.org/cgi-bin/archives.pl?request=166 Zagor: did you run the app on target ? i tried, but I think my screen is broken :-( :-( asch * dwihno sends Zagor some karlssons klister You can fix just about everything with that goo! it works fine, the debug output tells me :-) hehe See, the glue fixed it! Never underestimate the power of the Karlsson side! Zagor: so we basicly need to add Linus' mpeg thread to have a v1? yup kewlness can the uisim play music yet? on linux, it should be able to I haven't tried cause i was just thinking.. it might be perversly cool to realase the uisim evenutally as another mp3 player front end :) haha hahaha * adi|home wonders if that was good or bad laughter perverse it is hell... least ppl could play tetris while listening ;) very very tiny tetris well.. we offer it to 3 years olds... their very very tiny ppl did we ever find out if there is a specific reason you can't use the remote control witht he recorders? you can i do i guess the earliest firmware versions simply didn't listen to the serial port. current versions do. cool remote control(?) there's a wire remote you plug into the headphone jack. it's an extra accessory cool does it come with the basic set or do you have to order some extra widgets? eww <-- smells like a dog all over the hands extra stupid dog, running around licking my hand it comes with the "travel kit", for some reason well.. first tme i saw this ver 1.21d "Implemented improved SCMS* (Serial Copyright Management System), Fixed bugs within SCMS, which caused errors when encoding from digital source." Should I get the travel kit? nah i never use the remote except for testing no you should get the travel kit instead adi|home: the SCMS can be disabled in software or what? and will we be implimenting it? * adi|home prepares to slap zagor "Serial Copyright Management System"? Sounds fishy adi|home: SCMS is software only, as far as I understand so we just won't have it * adi|home cheers wildly Why does the recorder store the streams in VBR btw? someday.. when i can afford to buy congressmen too.. im going to repeal the DMCA I guess you gurus might have the answer to that question ;D dwihno: the encoder chip only supports VBR Zagor: ah, okay... actually, VBR is in some ways a simpler MP3 than CBR. due to avoiding the use of "bit resevoirs" Simple answer to simple questions :) so.. seeing as i have more time this week... --> Linus (~linus@labb.contactor.se) has joined #rockbox how close to the jbr do we want the interface to be? wb Linus yo! and if you could have me work on any higher level stuff.. what would it be? adi|home: we'll make two separate interfaces, so make it as close as you like adiamas: we could use some code for dealing with playlists I mean, edit, add, delete etc nod have we implimented unlink() yet? we have no write support enabled at the moment as for the 999 playlist limit on the players... do we know if there is a specific reason for i? we can't think of any :) thats fine.. i can prob code around it.. ghost funcitons etc why do you need unlink() ? delete? so if someone were to build current CVS firmware and apps for -DARCHOS_PLAYER=1 on both, and threw it on their player, should they expect it to do more than turn the red LED on? heh how else do you delete a file from inside the player? yes.. it should wipe your player and send me your credit card numbers :) ok, no there's no other way. PsycoXul: well, zagor believed it worked ;-) I want to wait with introducing disk writes for some time. It's a major risk. PsycoXul: did you try it? well thats what i got heh it didn't even get rid of the ROM firmware's splash screen !? zagor.. you see any prob with getting an unlink() working? we could test that safely adi|home: no problems really, but we need to test it exhaustively before letting it out in public thats a given... do we want editing/deleting to be a feature in v1? no okay.. then it goes on the back burner.. i can code around it for now... ill just display text saying "your file would normally be " :) hehe PsycoXul: that's interesting. i thought it was just my frankenstein's archos that didin't work. i'll try on linus' archos next. ok i have a small recommendation... the lcd code... wouldnt it make sense to have a Recorder version and a player version? with 90% of the player version just empty functions? this way we don't need all the #ifdefs all over the place? and just localize the #ifdefs to inside the lcd.[ch]? lots of the lcd #ifdefs is due to how the player simulator is working ? becuase it uses the recorder simulator's LCD functions right... and? I think adi is talking about #ifdefs in the application code, such as tree.c yeah ah, ok, I'll shut up ;-) #ifdef HAVE_LCD_BITMAP lcd_update(); #endif in the app code all over is just hideous so if we did lcd_update lcd_update(){} for player it's a good point. should we have the same API for both? and the real function for recorder actually, I added a dummy player lcd_update() yesterday... :) if not the same.. then very similar * adi|home saw that.. its what gave me the idea... i could test out the theory and see how it works/looks before i toss it up.. pass it off to you guys to play with .. so the question is: should we have a dummy lcd_setfont() lcd_setmargins() etc for the player too? yeah.. but i woulnd't even say "dummy" i would do this... void lcd_update() { #ifdef..... #else #endif } keep it localized wherever possible the functions that have to differ, well those take diff. args. let the compiler optimize all the crap out actually, I was thinking of doing the player stubs as empty macros so they're not even called i assumed that if a function was empty, and called, then the compiler would just optimize it out anyway yeah, i guess no they're in different files ah, right. so it can't know ahh.. so it still has to link it in? yes so Zag.. what did you mean about the macros.. what were you thinking? #ifdef PLAYER #define lcd_update() #else void lcd_update(void) { dodum(); } #endif or, uh no. i mean the header file not the implementation do I make sense? so in the header youd have like #ifdef PLAYER #define lcd_update() #else void lcd_update(void); #endif ? Just a quick question... Should I order something extra than the 'basic' recorder package? yes, that was my thought nah you can get what you need off ebay dwihno: no, I don't think so ah, okay dwihno: better headphones Then I'll place my order hmmm.. well.. only down side .. what about functions taht take arguements? Zagor: Already got that ;D yeah. i hate the ones that come with the recorder adi|home: #define lcd_setfont(__font__) ? wtf does __font__ become? it eats the argument it's just a declaration that eats the argument ah.. so just like #define lcd_setfont(x) it is for #define moo(x) poo(x) nods yes hehe was the __ that through me... thought i was missing something :) ah, well you can skip those well.. might i sugest then #define lcd_setfont(IgnoreMe) or something? or whatever hmm.. let me go play for a min sure * dwihno gets the bucket and shovel and follows adi to the playground bagder, linus: what's your opinion about this kind of shadow functions? I'm still not sure what I think. personally.. i think as long as we state we're doing it.. (in the docs) its a good idea it could get confusing and bug ridden, since bad code will still "work" on one of the two I think they're fine, as it makes modules that are shared for both LCDs more readable I didn't follow want i should explain linus? Is it about accepting a setfont() call on the Player? yes sorta Then I think it's dirty, but acceptable we are still early enough on the road that if we don't like it.. we can always switch back right It's a more cosmetically nice way than 102401 #ifdef's indeed nods and once the api is set in stone. life gets easier for others that want to work on the project "oh just use the api, it takes care of itself" is easier and nicer to say then "remember to #ifdef around everything" right, but people will still do mistakes since the player won't require lcd_update() but the recorder does Well, I don't like to say "set in stone" but I see what you mean. thats true... As long as we are not assimilated by borg, we just have to deal with those problems as those arise. what we could do is have all the #defines call one common funciton that prints it was called on debug shrugs the application code works on linus' archos, but not mine :-( Zagor: Hmm, fishy! Has he some HW mods that yours don't have? well my archos is a monster. it's a new cpu and an old lcd. PsycoXul: did you make sure the Makefile for both apps and firmware specified the right player version? Zagor: Have you raped your archos? oh yes :) You mad sadist! we wouldn't be at this point without ripping them apart and poking around Next time, get some photos while you're at it ;) you mean http://bjorn.haxx.se/rockbox/internals/ isn't good enough? would it be "#defin lcd_update();" or #define lcd_update() ? semi-colon or not not Zagor: I want to see some pics of the user interface! :D <-- is a bit perverted, I know dwihno: there is no user interface to take pictures of yet. try the simulator yup... i have a really nice screen shot of my tic-tac-toe before my code disappeared :) Zagor: yes, i've got a studio 20 and i made sure they were both set to TARGET = -DARCHOS_PLAYER=1 hehe Zagor: Does it compile with mingw/32? :) dwihno: make it if not Yeah Sounds like a heap'o'fun Where do I start? get the code from CVS http://bjorn.haxx.se/rockbox/cvs.html get the firmware, apps and uisimulator directories to compile the player version of the ui, not the recorder... whats the make commandline? 'make' :-) adjust the Makefile oh damn you... bloody hell :) *checking out* cool... but ya don't need to tell us ;) of course hehehe I don't have anything better to atm So I'll just pester you, great huh :) ? sounds wonderful :) okay.. bedtime niht all.. night adi nighty --> wavey (~wavey@dlan1431.dircon.co.uk) has joined #rockbox hey wavey hi wav Is it possible to revert to older versions of the archos firmware without excessive fiddling? just copy the old file to the root dir ah I thought the newer versions would prevent this no Is the USB-PC communication code in the firmware? no it is beyond the firmware's control ah, okay... Might as well be :) * dwihno drools, today I'm gonna order myself the almighty archos :Õ~~~~~~~~~ dwihno: So we managed to convince you? the rockbox sales force :) We want commission from Archos Linus: Yeah, you did... I just hope I won't get disappointed when my unit arrives. I have extreme high hopes about this product well don't expect a miracle product. it does have flaws. it's just better than all the rest :) :) Do you think I will get satisified? The swedish magazine M3 recently had a review of a lame flash-based player with recording capapilities And can I trust the storage? I mean, I don't want the hardware to go bongo when I move all my music to it :) and they were ASTOUNDED that it actually cound record sound from the Line-In Bua! They are so lame! hm Journalists != techies I would rather say that Archos marketing has failed in Sweden and M3 are totally blindfolded -( )-( )- <-- blindness goggles don't trust the storage I always keep a mirror on my PC it doesn't fail easily, but never depend on something you keep in your pocket True. True. How many times has it failed you? never never okay Zagor: you had it for a year or so? Or is it just me being new in the scene? :) dwihno: about 18 months oh really strange I've never heard about the archos earlier. marketing failure, I presume exactly it's a small company, and Creative's weekly marketing budget dwarfs their yearly.. well, I do see the positive side - It's a non-hong kong product Never trust fishy hardware from hong kong well don't hope too much from archos. they're not "the good guys". You mean, the serial RIAA thingy? the build quality of the archos units are pretty poor but worth it in the long run no i just mean they are not a better company than anyone else wavey: in what way? does it feel way plastic or such? they are actually manufactured in China h hm Do they feel like they're gonna break apart any second? no no haha no! --- Linus is now known as Linus|lunch --- Zagor is now known as Zagor|lunch dw: my unit keeps resetting because of bad contacts needs resoldering happened after a couple of months geez :/ there's webpages showing where to resolder, because it's not uncommon this is on a recorder bought in january okay I just hope I will get a flawless unit i love the device, but i wish it was built by sony Heh +$200 on the price tag :) alas yes ;) Soun would never build a harddrive-based player Sony They love their silly memory sticks too much --- Linus|lunch is now known as Linus --- Zagor|lunch is now known as Zagor Memory-dick Yeah Expensive shit, just look at the price tags! It's probably in the same range as the recorder-20 ok in main.c i commented out everything after init(); in main() and put app_main(); there PsycoXul: when did you try the app? this weekend, or yesterday evening? and when i run that i get "Rockbox!" displayed on the screen and then when i press a button it goes away and the LED turns red ah nothing more? and the LCD still never clears properly on the player, i'm always left with the 03 on the 2nd line after the rom version 5.03 splashscreen Press the button and the LED turns reeed! :D (alternative infinite mass lyrics) "ROCKbox!" is all it says and then any button gets rid of that, red LED comes on, and thats it [and the 03's still there :p] if you keep the old browse_root() call, nothing comes up? I want to change the UI: the browser should be the first thing that comes up, not a "where do you want to go today" menu when you press MENU, *then* the menu comes up anyone against that? nope Sounds like a good idea ok i put a little sleep in Try to keep the button pressing to a minimum i also changed the DEBUGF's in the init() in main.c to lcd_prints's so i can see them, and just put a little sleep in after the init and saw it's if (disk_init()) { its failing there interesting those should be panic() calls the only thing that can fail in disk_init() is if you don't have a boot sector on your disk heh can you do a: dd if=/dev/sda of=disk.img bs=512 count=1 and send me the disk.img file heh ok and where's this panic()? firmware/panic.c panicf()? ah, yes your boot sector looks fine. sda was your archos, right? yes hmm ok i changed them to panicf()'s we'll see what i get ok its actualy a problem before that if(ata_init()) { its that ok Is is that stupid check_registers() thing? no, that's gone sorry, I meant check_hd() check_registers() is still there sorry, check_regs() is there, check_hd() is removed heh PsycoXul: make the panicf() call print the return code actually, ata_hard_reset() is probably a stupid thing, BTW Zagor: of ata_init? yes I can't see how PA2 is connected to the ATA interface at all But I might be wrong My guess is that it does nothing with the harddrive Sorry, it's PA1 Still, PA1 is the RAS signal So setting PA1=0 will do nothing whatsoever with the HD odd yup er why's %d for int, man.. i just booted it up to see it print 'i' :p %d is ancient. dunno why it came to be. probably for decimal decimal, probably. as opposed to %x hex yeah heh ata_init()'s returning -1 we could add a case 'i' in sprintf.c to avoid that mistake assuming %d and %i is the same heh Who did the ATA code? Zagor? i did based on alan's code ah, okay can the simulators do threading now? no do we need that at this point? that should be a priority, we'll need at least queues for keypresses soon why? this poll();sleep(1) isn't very good ah we'll just move that to the simulated keyget code it works on simulator, but gives one hell of a key repeat on target we can do that without threads right I just prefer to avoid threads in the simulator as far as possible ok. yes we can probably do without them for now so as far as i can tell wait_for_bsy() is timing out? OK. So the ATA code fails at ata_hard_reset()? apparently no, there's not wait_for_bsy() in ata_hard_reset() OK. wait_for_rdy() then big deal it times out *what* times out, is the question wait_for_rdy() he said wait_for_bsy() yeah but wait_for_rdy() calls wait_for_bsy doh. i'm a moron is ata_hard_reset() fails, it can only be wait_for_rdy() right which can only fail if wait_for_bsy times out try commenting it out which? the call to ata_hard_reset() ok in ata_init() then the next wait_for_bsy() call will fail... right * Zagor is a cotton head We should debugf() the contents of the ATA status register coffee time' yeah so then it just returns -4 heh add this line to ata_init(): panicf("stat: %02x", ATA_STATUS); where in there? just before ata_hard_reset() hmm says stat: 00 ok you have a "new" player, right? yeah hmm how big disk do you have? 20G ok is it a 6000 or a Studio? studio * Zagor hmms some more math 15360*1024 er heh hang on, testing on my unit stat: 50 hmm, hmm and hmm again maybe the rom firmware's doing something different? yes. or the hardware is different heh did any of alan's disk test code ever run on your archos? never tried it ok but you run the same archos.mod firmware updates as any 6000 user, right? yep 5.07a Damn. We need a serial-modded new player I wish the donated player had a working serial port... Silly me. It has. heh It's just that the charging unit was broken, so we decided to use Björns old display board If we fix the broken board, we have a "new" player to debug we==me, unfortunately... :-) heh I think it's time to do that. this "franchos" of ours isn't working too well anything else you want me to test? not right now ok menu.c is much too recorder specific not very surprising I have made a logic analyzer trace of the Old (tm) Player LCD init sequence ah! It doesn't seem to be the Solomon LCD that we thought it was uhhh, do I want to hear this? ;) Zagor: we should publish it and let people try to figure out what it is sure Damn You are evil hardware guru jultomtar dudes! Our LCD shifting code doesn't talk to the LCD in the same way that the original firmware does dwihno: is that good or bad? ;) Zagor: *jultomte* = excellent! :D It only sets the command/data bit in the last bit "1 bit should be enough for everyone" We do it for every bit I'm not sure that it matters at all, though the Solomon data sheet says that it is only sampled every 8th bit Linus: that's intentional and I assume that our LCD isn't that much different in that aspect we do it once, and then loop through all 8 bits it's less complex than stopping and handling one bit specially I assumed it would work the same I know, Zagor. Just pointng out a difference ok So now I need a similar trace of the New(tm) Player Zagor: If you don't mind running on batteries, you can just switch LCD board right away sure And we can trace your LCD while we're at it go go go 1x2 hmm, tree.c and menu.c have much in common. we should merge them There. My order has been placed. I just analyzed Björns New(tm) Player. The init sequence for new and old LCD are identical that's good at least really strange Well, almost identical. A single command differs But that might just be that the firmwares are of different revisions I gitta fly now. gotta <-- Linus (~linus@labb.contactor.se) has left #rockbox maybe the desc part of the credits isn't necessary in the software i know. i just made a quick port Bagder: I've ordered my archos now! :D when will you get it? hm dunno a couple of days, I presume Hm... 20 gigs over USB1 will take several hours yes get a usb2 card ;-) Bagder: the simulator has no tick, right? Zagor: tick? timer tick, for splash timeout for instance no, but sleep() should work ah. yes Bagder: Nah, I don't need it :) For initial sync it would be nice, but it will be nothing but overkill later on overkill speedwise is nice ;-) heh nah :) time to go home, see ya <-- Bagder (~daniel@as3-3-2.ras.s.bonet.se) has left #rockbox i'm off too. see you guys <-- Zagor (~bjst@labb.contactor.se) has left #rockbox --- dwihno is now known as dw|gone <-- wavey has quit (Read error: 110 (Connection timed out)) --> wavey (~wavey@host-54.valtech.co.uk) has joined #rockbox can i assume there is an off button on the player? --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox adi|home: you here? tänk på att om man slår yxan i båten, får man ro som fan :) * Tumm har tråkigt... väldigt tråkigt han pratar! :) ;) följer utvecklingen via mailinglistorna... ok tar studenten om 2 veckor, så jag har tyvvär inte hunnit skaffa mig så mycket kunskaper om inbyggda system... sitter mes och försöker knåpa ihop demos :) eller... vilken plattform? win32 för tillfället <-- wavey has quit (Read error: 110 (Connection timed out)) yeah Zagor whats up? the off key has to be #ifdef SIMULATOR nods the firmware must never exit ahhh, okay :) i'm fixing it now, so don't worry :) ah.. was about to commit :) then nm too late :) i was giving some thought to the whole 'not read a valid file' thing one concern was excessive spin up and down of the drive... but if you are planning to 'play' the file, then its not a concern is it? it might still be, if you take some time to select the next it will spin down and up agin we should filter out non-mp3 files you just read your first buffer in, then check and confirm that it is one of the format we want.. i dont evn think that... no reason you shouldn't be able to 'read' a text file right, but perhaps not in the mp3 player mode? i haven't thought about it... hmmm i don't think there should be a 'mp3' mode and a 'not mp3 mode' maybe not * adi|work shrugs ill be around later tonight.. we can try and think it through then... yup btw.. what time is it by you? 22:32 should be like midnight no? nods so ill see you in your morning ;) heh, ok <-- Zagor has quit ("bye") well the archos firmware does handle being turned off somehow, because custom firmware won't turn off with the power plugged in but archos firmware will --- Disconnected (Connection timed out). **** ENDING LOGGING AT Tue May 21 20:07:37 2002