**** BEGIN LOGGING AT Fri Apr 26 05:32:08 2002 --> adiamas (~adiamas@as5300-9.216-194-23-95.nyc.ny.metconnect.net) has joined #rockbox --- Topic for #rockbox is Open Source Jukebox Firmware - http://bjorn.haxx.se/rockbox/ --- Topic for #rockbox set by Zagor at Fri Apr 12 09:45:52 * adiamas heads off to bed now that logging works good night --- You are now known as adi|asleep * Bagder heads off for the lunch meeting --- Linus is now known as Linus|lunch is there any way to turn logging off on certain channels? in xchat grrr.. what did you guys break? tr: too many arguments Try `tr --help' for more information. cvs server: Pre-commit check failed cvs [server aborted]: correct above errors first! cvs commit: saving log message in /tmp/cvsoX6Lpv <-- wavey has quit (Read error: 104 (Connection reset by peer)) --> calpefrosch (~calpefros@62.52.174.30) has joined #rockbox --> wavey (~wavey@dlan1431.dircon.co.uk) has joined #rockbox too busy at work :( wanna code more for the rockster box :) me too i have code to manage settings and an outline of the playlist func but I just started to get used to C no better way to learn, cal :) hope so what do you do at work ß ? <-- calpefr|ipaq has quit ("Smartsoft PocketCHAT for PocketPCs - Http://www.sspocketchat.com") i'm a java enterprise consultant he, cool. I want to make the Java Certified Prorammer this year. cool. java's a lovely language bit c is sexy :) s/bit/but hmm, let's try to put JINI on the Jukebox ;-) I never did C before, so it's difficult for me to understand the code or JME and now to something completely different... is the code in the cvs ? my code? not yet can you send it to me ? it'll go in today, hopefully. not much point seeing it now unless you're desperate? ok what do you do at work? java too? yes. I make PL/SQL with Oracle8i and some Java-Tools for our product that is written in java cool. what company? look at www.empolis.com oh, you make rubics cubes? :) * wavey chuckles no :-( you're german then? i'm listening to a german band on my archos as we speak :) rammstein i don't understand any of the words, but it's fantastic music anyhow :) yes, and the words are not important, you can believe me you're a fan? which archos do you have ? recorder not a fan, but I can listen to some songs. I have to recorder, too. And the firmware is running on it ? ( a bit ) i haven't put the code on the recorder yet i don't trust it :) yeah, because of the hd-lock yus :) --- Linus|lunch is now known as Linus Linus: did you try the firmware on a recorder ? I haven't dared to... :-) Büch dich! (about the only Rammstein song I know) I think it's "Bück dich" I suppose. Ich kann nicht deutsch sprechen. oh, that was very good. I just read the lyrics, and it makes not much sense the song adi|asleep: you still awake? is he working at night, or living on the other side of the sea s/sea/world he's american ah --> alkorr (jbcoax@srs07v-8-205.n.club-internet.fr) has joined #rockbox * Bagder yawns is the recorder display really 64 height, not 80 lines ? what makes you think it is 80? the simulator-windows is not full filled window the controller can only work with 65 pixels height hmmm, then the scale is not correct that's possible I'm working on that problem, tnx --- ChanServ gives channel operator status to Zagor --- Zagor gives channel operator status to Bagder --- Zagor gives channel operator status to Linus * Zagor is bored :) is the simulator's job to simulate the ui alone, or can it pretend to be the entire device, enabling us to compile our code against it and test it without putting the mod on the device itself? my desire is to get display, keys and filesystem simulation to work, primarily that's what we need to develop the UI but those 3 parts are the entire device, save the music output.. which is good :) yup :) what I mean is that the simulator only has to simulate the highest-level APIs yes agreed what are you working on right now? doom.mod? ;) fat hehe, no i'm saving that for the weekend hehe is the fat hard to test? not so hard, actually you testing it on linux or the device? i'm writing a simulated ATA layer on linux aha. ok can the makefile 'clean' target deal with subdirectories in the sameway that the SRC define is done? yes, but it doesn't right now never seen: SRC := $(wildcard drivers/*.c before its a gnu extension gmake ok s'cool :) very useful indeed :) there are others there are indeed lots of others indeed there are indeed lots of others indeed just do under cygwin : "info -f /usr/info/make" and you will find a lot of things * wavey installs info ;) if you are not familiar with "info" : just place your cursor on a "*" to jump at the matching page i'm an emacs user :) to go back, press "u" (up) but thanks :) okay :) see you (must eat now) * wavey tries to work out which package info is in within cygwin * wavey remembers emacs * wavey slaps his forehead M-x info :-) --- Linus is now known as Linus|meeting does anybody know the simulator-code ? for win32 ? nope :-( , so I wait for edx he has duplicated the lcd code. i'll smack him for that :) hehe I extract all words from all my mp3 song titles now ah, how many? 3009 files guess the most common words! ;-) "The" "Mix" umm, in your collection? "hate" ;) nah bwahaha Of of You I mix Love In A To You Remix to Me in remix My I should lowercase them yes <-- alkorr has quit (Read error: 110 (Connection timed out)) the: 259 times funny enough, on 259 hours of music one "the" per hour ;-) "distention" 4 times hehe how many different words? how many >once 3106 good q so it's one unique word per song? seems a bit high Number of words unsed more than once: 1027 is that with lc ? lowercase yes I'll have a look in the list... ah, your filenames include artist+album, right? this is only id3 tags so only song name? supposedly interesting "technophobybitstreamdreammix": 1 time :-) I thought the "unique words/song" would be much lower it seems correct, juding from the look of the list now, running this list with wavey's script, I wonder how large percentage we'll cover Total Length: 932753 seconds: 259:05:53 Different artists: 302 Number of files: 3009 Number of files without id3 title: 183 cool I must go now ok bye <-- calpefrosch (~calpefros@62.52.174.30) has left #rockbox i'll check my perl script into the tools section :) * Bagder considers running a spell check on his id3 titles ;-) --> edx|school (edx@pD4B9E95C.dip.t-dialin.net) has joined #rockbox --- edx|school is now known as edx hi edx hi --- edx is now known as edx_ --- edx_ is now known as edx --- ChanServ gives channel operator status to edx edx: you shouldn't copy the lcd code into your win32 lcd.c why? shall i link both files with it? yes, like the X11 simulator does compile the drivers/lcd with the SIMULATOR flag set ok then in win32/lcd.c just include the code that is win32 specific got it. you should probably rename it lcd-win32.c to avoid collision ill do that later - *away* ok --- edx is now known as edx|away --> elinenbe (trilluser@bgp01029830bgs.sanarb01.mi.comcast.net) has joined #rockbox Hello. hi How is the status coming? coming along nicely. did you see the new status matrix on the web page? yeah -- that is nice. i'm working on the fat code right now --> alkorr (alkorr@srs07v-2-177.n.club-internet.fr) has joined #rockbox Fat32? I thought the fat16 code was finished? I have a question about the File System code... How does this work? The files system is implememnted in the firmware? But what does the USB chip detect? ? What I mean is could you have an open interface to implement more file systems on the drive? for the moment no what are you speaking about ? For isntace could a Mac user format it HFS+ and then the player could use it like that? about the filesystem or the chipset USB ? nope if you are speaking about arch*s firmware, nope --> jb1081 (~jb1081@lime.ircam.fr) has joined #rockbox the chipset USB only services generic scsi transfers between the computer and the harddisk i think he's asking about the possibility of using other filesystems with the device with the opensource firmware in the future or ata transfers if you like elinenbe: i think the only thing is you'd have to have a small fat32 partition at the beginning for the ROM firmware to load the new firmware from the drive just a sec if you plan to read mp3 files with your jukebox on a non-fat32 partition, it is impossible alkorr: impossible? alkorr: with why couldn't the firmware support other filesystems? but if you want to create a HFS+ on the harddisk knowing that your jukebox could not read it, yes you can but at your own risk --- Linus|meeting is now known as Linus if you are speaking about rockbox which is in fact a software, not a firmware there still a possibility to handle any kind of filesystem 's but it is not our priority I see. heh You can format it right now HFS+ and it will work fine as a portable hard drive for the Mac, but it will not work as a player. exactly unless some courageous people code a HFS+ filesystem for rockbox (when of course it would have features enough to compete against the firmware) for the moment it is too earlier :) can the USB-ATA access really handle other fses? Mine is formatted as a QNX 4.x partition heh the purpose of that chipset is not to handle filesystem bagd: its in the manual and everything that you can format it different things.. just without mp3 player functionality Bagder: i've run ext2 on my archos ok, I get it that is not controlled in the archos end but to give to the computer the access on the harddisk that's all but still, the firmware reads the software from a fat32 filesystem yeah so that would at least be required ah.... I see but that way there is nothing to have with the chipset USB right and yes we still need a first FAT32 partition <-- elinenbe has quit ("Leaving") just to download our software yep unless you have a flash instead a rom and you can change the firmware in the flash of still very dangerous ;) I'm curious. What are we talking about? someone was asking if other filesystems can be read we could replace the SH7034 with a PROM version and have our software PROMed! ;-) * Bagder ducks no i'm speaking about a software way to overwrite the flash without hardware mods isn't it a masked ROM version? i'm speaking about the external rom not the internal rom * Bagder is a bit out of his leage here Bagder: the firmware is located in an external (flash) ROM. so what's in the internal rom? I guess the code that sets up the oscillator stuff and some chip selects and stuff, to be able to execute from the flash. the internal rom would read the external rom and unscramble it if you used my digrom, the second file it produces comes from external rom and it is unreadable because it is indeed scrambled okay i explain my digrom creates two files the first is smaller and contains the internal rom the second is larger and contains the external rom the first is readable and has a vector table at the begining of the file <-- Bagder has quit ("http://daniel.haxx.se") c u <-- alkorr has quit () <-- Linus (~linus@labb.contactor.se) has left #rockbox --> Linus (~linus@labb.contactor.se) has joined #rockbox <-- jb1081 has quit ("[x]chat") how close is the simulator to faking the device's filesystem? has that work started at all? I'm pretty sure that the simulator only simulates the GUI. today, maybe :) we want to simulate the fs, the buttons and the lcd so that our runtime code can be tested on a pc yep feel free to write it :) i'll do so, once the playlist is complete cos i'll need it to test the playlist :) actually, it's not much to write at least not on unix that's the hope or windows since we want to use POSIX api windows might need some stub code yus i have no idea why i'm specialising this dev in windows.. i -hate- windows Actually, I think Windows has open(), close() and seek too. love, hate. there's a difference? it certainly does with cygwin installed :) Linus: opendir and readdir too? Getting philosophical here, eh? heh I guess so. Lemme check. i've never used opendir. why not just use FILE *? that's a silly q but i mean what operations does opendir give you opendir is for reading the directory entries, not a file it doesn't seem like Win32 has opendir() :-( i thought not and that's not available from a file handle? hmm a file handle lets you access a single file it's been years since i did any fs operations in C opendir returns a dir handle ok tnx then you do readdir(handle), which returns which files are available in the dir i find it odd that i've never had to do that in C java a million times.. weird I haven't done that many times i C either. see you later guys ok, bye bye --> alkorr (alkorr@srs06m-8-44.n.club-internet.fr) has joined #rockbox <-- alkorr has quit (Client Quit) --> Bagder (~daniel@as3-3-2.ras.s.bonet.se) has joined #rockbox boo hey ho is everything done now? ;-) Hacking FAT32 in 300mph nicers anything I can do to help on that? No. You are lame. :-) heheh Go awayt. you finally found out darned * Bagder hides in a corner *slap* if ( bpb->bpb_fatsz16 ) { printf("This is not FAT32. Go away!\n"); return -1; } this is all so far :*) :-) ;-) Gotta go away! :-) CU!!! quit <-- Linus (~linus@labb.contactor.se) has left #rockbox <-- wavey has quit (Read error: 110 (Connection timed out)) <-- Zagor (~bjst@labb.contactor.se) has left #rockbox --> calpefrosch (calpefrosc@p5082C7D7.dip0.t-ipconnect.de) has joined #rockbox hi again rehi edx is away for long? Did he said something ? I don't know, I just recently got here myself and he hasn't said anything the last ~30 mins tnx --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox welcome home ;-) hehe, thanks * Zagor just realized ata-sim.c can actually work on real devices ln -s /dev/hda1 disk.img *voila* and does it work? I think I want to comment out ata_write_sectors() first, though :) haven't tried yet. it should me needs food... you could dd the device to a file and then try I've already done that from a floppy ok worked fine (although the fat code isn't done yet) * Bagder looks around for food too so what worked? the dd? we could list a dir neat or part of, rather fat32's root dir isn't consecutive so we only got the first sector yet its a beginning <-- mid has quit (Remote closed the connection) it's a relief only having to support fat32 and not -16 or (horror) -12 fat12 is a mess the page at linux-usb.org is updated now, btw neato <-- calpefrosch has quit () * Zagor got a note from the post office today I think it's the donated Archos --> ironi (xircon@m213-101-132-36.swipnet.se) has joined #rockbox hello hi ironi hi Bagder nice with the double-space font that's a fun trick indeed yeah i wonder what it could be used for the "Rockbox" startup text! ;-) thats the only thing i could come to think of i would love to contribute to this project. too bad i dont know anything to be able to help many people have requested bigger fonts for use in the car Zagor, true, true Zagor/Bagder: off topic question: have you ever experienced that your bostream modem only has one lamp on? =) not yet :) i've only had it about two weeks nope oh ok i got mine in january I've had mine since... september last year used it for 4 months, and then i moved out but have been paying all the time until now and finally i got my own place again, and they connected it, but it wont work I have a friend who's bostream modem just stopped working one day didn't sync anymore took him weeks to convince the service guys to get there, check it and then it took some more weeks to replace it oh. really? might be same with mine, who knows they's *not* service minded "When you release the spec, six crazy Russians you never heard of will write an insanely fast encoder that implements it." they're (discussion about Vorbis not having proper specs) Ogg talk? yeha saw that one too i just enjoyed the "six crazy russians" hehe kinda funny hehe hehe, do you guys know what TWAIN stands for? no idea Toolkit Without An Interesting Name ! I love it! technology ok i tohught it was tehcnology and not toolkit thought :) i was away for a second i actually knew that one :) hehe http://base-linux.com/webcam.jpg it's a great name, nonetheless thats a crazy nic to have for private use Zagor, well yeah i guess a deeper meaning could be extracted twain -> mark twain -> riverboats -> carrier -> supporter -> driver -> twain ehm.....nah. =) kinda far-fetched very =) http://www.scannerplace.com.au/twain.htm http://dmoz.org/Computers/Hardware/Standards/TWAIN/desc.html no deeper meaning necessary hum, i'm reading the boot sector wrong... i want to develop something like anything =) what are your skills? i think i told you im more of a web stuff guy in addition i have very basic knowledge of c++ and java ok, sorry I have a pretty bad memory... you're very welcome to think up a prettier look for the web page thats what too much drugs do to you =) hehe uh, brain damage. how do I round up a division without +0.5 ? int (division) + 1? LOL (4/2)+1 != 2 ohh... too sad :/ hehe then i dont know actually... number/divisor + (number%divisor!=0) LOL hehe. creative, i'll give you that :) lol is there a more efficient way? (certainly..) yes there is, it's just fallen off my head at the moment... % is not a cheap operation well you could do assembler and then % and / is one operation and it doesnt cost anything :) someone once said: "Premature optimization is the root of all evil" :) i think i once even new the name of this someone - but in this case it is effective. as long as you dont know any other way :) Donald Knuth? yup hehe Zagor: what exact problem/math are you trying to solve with this? --- edx|away is now known as edx nothing special, i'm just bugged by having forgot how Zagor, is the webpage html lol ehm i mean, is it static html only ironi: yes --- edx is now known as edx|test --- edx|test is now known as edx ironi: 95% the mail archive is a cgi, for instance. and index.cgi is a load balancer but the rest is static built using a "make" template system that's why you only see .t files in cvs the webpage is in cvs too, so go there and look nick edx|testagain Zagor, okay... ./nick edx|testagain --- edx is now known as edx|testagain LOL --- edx|testagain is now known as edx Zagor, a make template system...well i have no clue how that works :/ Zagor, what's the point of that? the point is that all headers are in head.t instead of copied in every .html file so I just write the page-relevant data in the .t files and then run "make", which assembles head.t, page.t and foot.t into page.html that makes it very simple to change things ehm don't worry about that, just get us a better look :) ever heard of SSI SSI is the devil's work why? why do it dynamically when the data is static dynamic pages are for dynamic data, not for lazy web authors :) well you do have a point at least when the page is done but to run make while oyu develop is a bit time-consuming not really. it takes about 0.3 seconds it's mapped to Alt-C in emacs heh ok well i use include in php it is pretty common to do it that way, as far as i know SSI also has problems with parameters how do you mean never mind. i like static pages and i like this system :) i do some dynamic includes and so on just give me better colors :) * Zagor is on his knees but ok Zagor, is there php support btw nope Zagor, ok =) dynamic pages are 99% of the reason most web sites don't handle high traffic Zagor, hey do you like the show on tv2? what is it? its annika lantz ah she is hilarious =) ok well lemme see what I can do about the colors yeah, she's fun. I don't watch much tv anymore, though ok, good heh ok btw it doen't have to support lynx or somethingh like that right * Bagder accidentally mentions that he doesn't like that blue bgcolor ;-) Bagder, i'm on it too bad the jukebox doesn't have nice colors ironi: not really, but I'm not fond of too complex designs either Keep It Simple Zagor, i like simplicity KISS good i kinda like the logo see if i can do something around that concept ok, good --- edx is now known as edx|finaltest the colors of the activity table could use a looking-over too... :) --- edx|finaltest removes channel operator status from edx|finaltest --- edx|finaltest is now known as edx edx --- ChanServ gives channel operator status to edx haha *auto op script for edx on rockbox* lol amazing really. Zagor, hehe sure i think the logo should have a intropage =) maybe not. haha welcome to 1998 :) well a lot of pages have that in some kind of introduction phase to create a image it sucks we're a bare-bones to-the-metal project and I know users all over think so too our web page should reflect that but that doesn't mean it have to be this ugly :) Zagor, u got a point i have a thought can you get the layered image of the logo? i.e. do you have one I can get it I don't have it here ok i like it just the way it is, im just interested in changing the yellow a bit this is what i thought ok the jukebox has blue, gray and yellow the 6000 is in some orange color it fits very nice on the silver a little too strong for a background color, don't you think? well the purple-ish below the display too, but i'd like to disregard from that not for the logo well maybe, but giving it a shot would be cool also, we shouldn't borrow design elements from the archos gray is nice as bgcolor i'll see if I can get it layered so you can play with it well using it as inspiration couldnt hurt no *yawn* ironi: photoshop image coming soon <-- Bagder has quit ("http://daniel.haxx.se") Zagor, i was bored so i played way out :) heh ehm 4 mb im on gprs oh ok :) will take forever that the only one?` can u perhaps pack it i'll see what I can do that one's ~3500x1000 too, which may be a bit more than you need :) heh i'd say so, yes hmm, which file formats have layers? gimp can't save in psd format i use layered .png images in fireworks so try saving it as such "PNG can't handle layers" it says :( --- edx is now known as edx|away --- edx|away is now known as edx|sleeping cya guys good night oh ok well the fireworks is using some special png seems like it i guess you can't read XCF files? (gimp's format) i dont know i guess not probably not well ill try to do something tomorrow or so nlinks.org/dev/test.gif <-- just something i played around with, for fun .jpg http://nlinks.org/dev/test.jpg hey, cool! you're good! thanks :) well, I'm going to relax in front of the tv now...talk to you tomorrow... yeah, bye i think we can use the colors....well, talk to u later <-- ironi has quit ("zz ") wow.. that is cool... sure is * adi|asleep thinks we just found our web guy ;) --- You are now known as adi|awake --- You are now known as adiamas --> Moufle (~moof@195.101.176.125) has joined #rockbox hi hi I just bought a recorder 20g, and I run it on linux it's charging right now it looks exciting :) Is rockbox somewhat usable, or is it alpha right now? it's pre-alpha :) we're still writing parts of the code i'm working on the fat32 code right now hmm ok :) you've got previous experience in programming that kind of stuff? yes, i do embedded programming for a living I've got a very good general computer culture, but I don't speak C fluently ;) ok In fact, it's just a matter of changing bits on the chips and doing simple logic operation....? I know it's hard well, all programming is "just a matter of changing bits" :-) but you look well documented we have most of the documentation we need I'm gonna read your chat logs on the website ok :) But this project excites me a lot I hope i'll ba able to help you later you're welcome! BTW, what's needed to compile the firware, a special gcc (for the archos CPU?) yes. the sh-gcc. there's a page on the web site describing how to build it there's rpms available too Ok, I'm going to look at it and build it thanks a lot, good coding :) hehe, thanks <-- edx|sleeping has quit (Read error: 110 (Connection timed out)) <-- Zagor has quit (devlin.openprojects.net irc.openprojects.net) <-- Moufle has quit (devlin.openprojects.net irc.openprojects.net) <-- PsycoXul has quit (devlin.openprojects.net irc.openprojects.net) --> Moufle (~moof@195.101.176.125) has joined #rockbox --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> PsycoXul (psyco@adsl-63-205-41-159.dsl.lsan03.pacbell.net) has joined #rockbox --- ChanServ removes channel operator status from adi|work <-- Zagor has quit (devlin.openprojects.net irc.openprojects.net) <-- Moufle has quit (devlin.openprojects.net irc.openprojects.net) <-- PsycoXul has quit (devlin.openprojects.net irc.openprojects.net) --> Moufle (~moof@195.101.176.125) has joined #rockbox --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> PsycoXul (psyco@adsl-63-205-41-159.dsl.lsan03.pacbell.net) has joined #rockbox <-- Zagor has quit (devlin.openprojects.net irc.openprojects.net) <-- Moufle has quit (devlin.openprojects.net irc.openprojects.net) <-- PsycoXul has quit (devlin.openprojects.net irc.openprojects.net) --> Moufle (~moof@195.101.176.125) has joined #rockbox --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> PsycoXul (psyco@adsl-63-205-41-159.dsl.lsan03.pacbell.net) has joined #rockbox <-- Zagor has quit (devlin.openprojects.net irc.openprojects.net) <-- Moufle has quit (devlin.openprojects.net irc.openprojects.net) <-- PsycoXul has quit (devlin.openprojects.net irc.openprojects.net) --> Moufle (~moof@195.101.176.125) has joined #rockbox --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> PsycoXul (psyco@adsl-63-205-41-159.dsl.lsan03.pacbell.net) has joined #rockbox <-- Zagor has quit (devlin.openprojects.net irc.openprojects.net) <-- Moufle has quit (devlin.openprojects.net irc.openprojects.net) <-- PsycoXul has quit (devlin.openprojects.net irc.openprojects.net) --> Moufle (~moof@195.101.176.125) has joined #rockbox --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox --> PsycoXul (psyco@adsl-63-205-41-159.dsl.lsan03.pacbell.net) has joined #rockbox --- Disconnected (Remote host closed socket). **** ENDING LOGGING AT Fri Apr 26 20:11:53 2002 **** BEGIN LOGGING AT Fri Apr 26 20:12:05 2002 --> adiamas (~adiamas@as5300-9.216-194-23-95.nyc.ny.metconnect.net) has joined #rockbox --- Topic for #rockbox is Open Source Jukebox Firmware - http://bjorn.haxx.se/rockbox/ --- Topic for #rockbox set by Zagor at Fri Apr 12 09:45:52 --> adi|work (~adiamas@pool-151-204-140-140.ny325.east.verizon.net) has joined #rockbox --> coyote- (coyote@dreamhosted.borlange.se) has joined #rockbox <-- Moufle has quit (Read error: 113 (No route to host)) wooooohoooooo * Zagor just manage to parse a multisector directory :) phew sweet... * adiamas doesn't have a clue how difficult that is. it shouldn't be difficult, but when it's late and you have a silly bug that taunts you, it is... basically, it means the fat32 part of fat.c now works what's left is vfat (long filenames) fat.c just went YELLOW on the activity table :) time for bed. see you later <-- Zagor has quit ("bye") --> edx (edx@pD4B9E95C.dip.t-dialin.net) has joined #rockbox --- edx is now known as edx|good_morning --- edx|good_morning is now known as edx --- ChanServ gives channel operator status to edx --- edx is now known as edx|breakfast --> Linus (~linus@labb.contactor.se) has joined #rockbox <-- Linus (~linus@labb.contactor.se) has left #rockbox --> calpefrosch (calpefrosc@p5082C1BF.dip0.t-ipconnect.de) has joined #rockbox hi hi --- edx|breakfast is now known as edx --- edx is now known as edx|away --> alkorr (alkorr@srs01v-2-239.n.club-internet.fr) has joined #rockbox <-- alkorr has quit (Client Quit) <-- calpefrosch has quit (Read error: 110 (Connection timed out)) --> calpefrosch (calpefrosc@p5082C4C6.dip0.t-ipconnect.de) has joined #rockbox --- calpefrosch is now known as calpefrosch|away --> Linus (~linus@labb.contactor.se) has joined #rockbox <-- Linus (~linus@labb.contactor.se) has left #rockbox <-- edx|away has quit (Read error: 110 (Connection timed out)) --> Bagder (~daniel@as3-3-2.ras.s.bonet.se) has joined #rockbox <-- calpefrosch|away has quit (Read error: 110 (Connection timed out)) --> Zagor (~bjst@as9-5-6.k.s.bonet.se) has joined #rockbox hiya guys hey just got the donated archos cool very! new/old lcd? it's also very dead :-) but I knew that ah actually, it looks like an old i'll open up and take a look you gonna swap CPUs with this or what? well I can't if it's an old. I guess then I'll convince Peter to trade back with me :) * Bagder giggless he'll love me... * Bagder is engaged in a hair-tearing debug-per-email session hehe that mail on the curl list about 6.5.2 working but not 7.9.6 ok try and talk them into #curl :) yeah * Zagor notices the label over one the screws is labeled "Warrenty seal" what's warrenty? ;) Isn't that an actor? warrenbeaty? :-) haha fat.c went from status red to yellow yesterday wow so you can simulate a few things with it now? yeah, i can read a directory now short names only, so far still cool * Bagder watches a line formerly read now being yellow... s/read/red yup. it doesn't feel quite as obscure anymore, which is always good indeed getting a feel for the source and how things work is a huge part of things even before the functionality is there yes. i've simplified it a lot too. so now it's much more penetrable that's very good i can smell burnt electronics from the new archos :) hah whoa, it's almost as old as mine and linus a 20th century archos! gosh it's made 0052, just six weeks after ours 0046 --> edx|away (edx@pD950D222.dip.t-dialin.net) has joined #rockbox --- edx|away is now known as edx --- ChanServ gives channel operator status to edx hi edx lol just gave my private rockbox shirt in print ;) we wanna see pictures of you wearing it later hehe. you have to post photos! nlinks.org/dev/test.jpg <-- just something i played around with, for fun hey this is cool! wow i think we'll soon get a prettier web page :) he knows his stuff yup hehe... i will - i just had to put some names on the shirt hihi ppl of the list who develop stuff ;) hehe i put the internet site and the irc channel on it.. sounds like a cool shirt the site ironi is developing has one problem: the font he uses is Verdana i guess - it is not supported by linux i know, it's just a mockup hehe we'll use arial or helvetica or something zagor.. are you currently working on a linux machine? yes http://codeforce.d2g.com/ always does it look fucked up? lol no looks ok how is that? what font type is it? I dunno sans-serif arial, i think ok ... (is verdana on windows) i guess i put arial as alternative font then... well, web pages don't go blank if they can't find the font anyway yes you did :) font-family: Verdana, Tahoma, Arial; ok hehe you should att "sans-serif" to the end of that list add .. ok that's the fallback font whatever the browser can find otherwise most fonts will use Times when they can't find any of these three fonts most browsers *just added it* :) (heh it doesnt matter anyways because it is a msvc++ site so nobody from linux will ever go there haha) haha but maybe some mac people (unless i make them, like you) damn.. got a bunch of homework for this weekend... class test on monday :/ that all sucks.. me too: fat driver, x11 simulator... ;) edx: i received the donated archos today it's really cool how people want to help the project * Bagder agrees a donated archos.. great... you know who sent it? yes, a guy called Julien Labruyére he has too much monay?! lol no, he broke it and got a new recorder insteda so we have to fix the power circuit, but that's no problem be careful, one day archos is gonna "donate" you a jukebox with a free letter bomb hehe i played lotto today - if i get the 9 million Euro jackpot, ill donate something as well (haha) a t-shirt? ;-) a tshirt would be cool t-shirts for everyone.. in every color lol (do you knwo what lotto is btw...)? great! i'll never have to wash again! yes, it's the same word in swedish * Bagder has been offered a donated t-shirt in another project ok.. http://www.karelia.com/images/tshirt.jpg a 10% size-little-messed-up-version of the t-shirt: http://codeforce.d2g.com/rockbox/tshirt.jpg rock n' roll ok.. time to eat some thing... --- edx is now known as edx|eating what's "Watson"? it's a commerical Mac OS X program heavily based on libcurl ah gotta run an errend, back in while whats errend? lol --- edx|eating is now known as edx :) please, tell me... :D errand: [noun] A short trip taken to perform a specified task, usually for another. huh.. wasnt even in the online dictionary i use :/ ok thanks :) dictionary.com dict.leo.org isnt too bad either - it is German-English tho he spelled it wrong, maybe that's why didn't find it jup a short journey in order to do something for someone, for example delivering or getting something for them © Pearson Education Limited 2000 lol @ t-shirts.. if shipping wasn't too expensive we could have a t-shirt order service hehe hehe <-- coyote- has quit (carter.openprojects.net irc.openprojects.net) --> coyote- (coyote@dreamhosted.borlange.se) has joined #rockbox <-- edx has quit () --> edx (edx@pD950D222.dip.t-dialin.net) has joined #rockbox --- ChanServ gives channel operator status to edx hmmmmm... damn.. i put irc.openprojects.org on the t-shirt *noooooo!!* oops... well... nobody will be inspired to join the irc channel because of my shirt anyways --- edx removes channel operator status from edx --> calpefrosch (~hotrana@62.52.178.189) has joined #rockbox hi edx hi I have to go in 5 minutes but, can we meet later ? sure I would like to discuss some things about the simulator ill be here for at least 6 hours from now.. no problem ok, then cu l8r cu tnx bye <-- calpefrosch (~hotrana@62.52.178.189) has left #rockbox --> calpefrosch (~hotrana@p5082C472.dip0.t-ipconnect.de) has joined #rockbox --- calpefrosch is now known as calpefrosch|away I'll be away about half an hour --- Zagor is now known as Zagor|out * Bagder is back * edx greets Bagder and I nailed a bug! hehe.. what bug? unrelated to rockbox ah ok congrats anyways ;D heh, thanks curl => http://curl.haxx.se is my biggest open source project I'm involved in *looking at it* or, that's what takes my most time hmm i think i heard of curl before... not impossible it's getting quite widespread use these days there's also a commercial web language named curl, but that's another subject you are "haxx" (haxx.se) then... Haxx is our company, haxx.se is our domain jup Björn, me, Linus and a fourth friend that is cool.. you are four people together? yeps but we're employed by another company where we work full time Haxx is just a spare time company ;-) ah ok.. i was already wondering why bjorn told me he was at work yesterday - i guess you dont have you own office rooms for haxx hehe correct we work for Contactor => www.contactor.se (site in swedish only) as consultants hehe - dont understand a word :)_ how old are you? so Björn and Linus spend their days on Siemens I'm 31 then you are a little younger than bjorn, right? no, he's my kid brother. Björn's "only" 29 oh the int was linus who's 33 the int = then it right our "old man" ;-) hehe "programming since 85".. thats a lot (from you site).. yeah i was born in 86 hehe half my life wow more than my life I and Björn bought our first computer 1985 he was 12, I was 14 i started programming in 96 i think.. Quick Basic lol - i was 10 what computer was it? C64 my first as an ATARI ST something - it wasnt really mine but i used it a lot after that came a pc.. win95 stuff we made on the C64 => http://www.contactor.se/~kjer/horizon/ so i was nearly directly involved in windows business I took the route C64 => Amiga => Linux hehe --> edx|disconnected (edx@pD950D222.dip.t-dialin.net) has joined #rockbox <-- edx has quit (Read error: 104 (Connection reset by peer)) * edx|disconnected was disconnected --- edx|disconnected is now known as edx connection problems? i dont know just was disconencted the horizon stuff you did must have been really great and a lot of work for that time... oh yes we spent *many* hours on that and we did pretty good on the "demo competitions" i can imagine that.. :) Zagor is a really old nick ;) is indeed mine is only from 1987 ;-) old enough - mine is less than a year old - didnt have a nick before.. just didnt find one that fit my taste ;) you did a lot with this sinus stuff did you write the sin functions by your self (it is calculating the sinus values)? no, we usually generated sinus tables from a basic program ah :) "69248 chars in one single scroller." Wow. Where did you get that much text from? we had a machine on "copy party" and whoever wanted could write hehe remember that the machine only had 64K ram... what were those parties like? competitions..? the parties were like a couple of hundreds guys who brought their computers set up everything in a school or something LAN-Parties? hehe they call it that these days, yes but lanparties are like playing stuf.. we had no LANs back then and we didn't play hehe of course not we coded demos jup. night and day and had a competition in the end i could never find enough ppl in my age to make a "coding party" where the demos were usually screened using a bigscreen and people could vote hehe cool too bad, coding is that makes computers most fun ;-) I still think so how did you draw the pictures btw..? yea i think so too the pics were made like today, with paint programs only more basic programs and less colors available hmm they look so perfect hehe... did the programs support emboss effects or something like that or did you draw it yourself? manually pretty much everything was manual wow.. but since we became successful things really develop fast. :) we got lots of friends who wanted to help us with things like that so we got lots of pics and logos etc from other people yes, things were quite different back then not better, different i wish i would have started a little earlier. but 10 years is about as early as possible. it is you seem to have learned quite a lot anywaya heh i love my computer :P - i spent as much time with it as possible. --> alkorr (alkorr@srs06m-3-20.n.club-internet.fr) has joined #rockbox hey Alan hi hi everbody Björn got his donated Archos Björn is out ? yes good news and a working archos ? he should be back soon though no yea.. a few mins.. it doesn't work, but Björn thought he'd be able to fix it what it is the third archos with a broken serial !? we don't know yet on this one i'm reworking my ata stuff to have an interrupt-driven code cool we don't need a thread for working with ata just a queue of requests i think the same thing can be achieved for MAS it would probably be good, pretty efficiant i mean a queue to serialize operations should be enough and more efficient in performance terms yes unhopefully your devcon doesn't tell me why you want those three tasks nor what they really do * PsycoXul blinks if i could know what you exactly need, i can give the best way to implement since I know very the internals of SH1 hum... i'll be back. I hope Björn will be here when back <-- alkorr has quit () --- edx is now known as edx|away * Bagder drinks tea --- Zagor|out is now known as Zagor <-- calpefrosch|away has quit (Read error: 104 (Connection reset by peer)) --> calpefrosch (calpefrosc@p5082C472.dip0.t-ipconnect.de) has joined #rockbox --- calpefrosch is now known as calpefrosch|away --> alkorr (alkorr@srs05v-4-251.n.club-internet.fr) has joined #rockbox wakeup here ! :) * Zagor wakes up * Bagder yawns so you new toy ? yes, but it has a power problem so Linus is going to look at it (I knew it had this problem) what kind of problem ? I think he reversed the polarity of the charger and fried something well i suppose our generous donator was not so generous as to give you a working one ;P I still think it's very generous i mean i was atonished that a person will give a working archos i would be ok yes, I agree there's a difference but I'm very happy for this donation anyway because i thought it was a working archos :) but now i understand why --- edx|away is now known as edx sure we must gratulate this person for his donation yes, I have written it on the front page Björn, i need some details about you devcon, especially about the tasks Linus told me that you wanted a special thread to do ata stuff maybe, yes either that, or the interrupt handles everything because you thought we cannot use interrupts with the ata controller i explained him that if my ata.c had no interrupt code, it is because I was lazy quite now, i coding an ata driven-interrupt code where all what regards ata handling are done in the interrupt. yes but I think that's ok. we can start with polled and then do interrupts ok, nice it means that there is no thread but a queue for requests yes i'm committing my last ata changes now so I have no local modifications i'm still wondering if we need read-ahead not now i don't change your ata.c I want to do things very simple at first and add more complexity later please be cautious, i read your ata.c and see you oversimplify the code to an extent i fear that code generated for sh1 is worse. I know that's deliberate when the code works, then we make it small&fast I have gathered a little list of contributors in a CREDITS file, should I just commit it and have you fill in more/edit it? please do is the firmware dir fine? --> alan (alkorr@srs05v-4-251.n.club-internet.fr) has joined #rockbox <-- alkorr has quit (Read error: 104 (Connection reset by peer)) yup done btw.. could anyone write a file defining the file access functions (just the header) so i can implement them on the simulator? --- calpefrosch|away is now known as calpefrosch or shall i write the file (but i dont know the "devcon standard") hi all together imho, you can write it ok then we can just edit/discuss things we don't agree with ill try - you can still change them later exactly what name? and what dir? (the header file) edx, if you don't mind I would like to help you wuth the simulator. * Bagder leaves that decision to Björn sure, i dont mind * edx asks björn... ummm... cool, then I have some questions to you. fileio.h fs.h ? go ahead, calpefrosch tnx, hmmm... Zagor: I scanned the list archives quickly, checked the web and stuff to figure out the initial CREDITS but I'm sure I've missed a few names i'll take a look at it hm.. say a header name ;) I did no C before, so I have no idea how to change the struc with the colors. file.h ok the only thing you have to change are two lines in uisw32.h: my idea was to change the background-color in the struc #define UI_LCD_COLOR 46, 67, 49 // bkgnd color of LCD #define UI_LCD_BLACK 0, 0, 0 // black sure the first line is the background color but I want to change the color at runtime... oh... hm then you have to change the bmi structure. right but I had some problems with the location, where to change it. we would have to define an extra function for this... wait 1 minute, please Bjorn, is there a function to change backlight color in the lcd code (no) - could we define it (even if it is not yet implemented outside the simulator) edx: why would we want that? the color is green. there's no way to change that ahmm backlight ok, extern void lcd_backlight(bool on); and contrast... you can regulate backlight i think... with the menu <-- calpefrosch has quit (Read error: 104 (Connection reset by peer)) --> calpefrosch (calpefrosc@p5082C472.dip0.t-ipconnect.de) has joined #rockbox hey who change DISP_X to LCD_WIDTH constant ;) yeah, I saw this on Friday,too. it was LCD_WIDTH all the time arhg, I had to reboot. but it was DISP_X too edx: that was me. they both existed, I just removed the one with the worse name. allways, at the second time a switch on the archos with USB connected I get this damned BSOD ... :-( ok calpefrosch: ugha calpefrosch: hey... you use USB 2? no ok.. i have a similar problem but i think it is related to my usb2 adapter (it works with usb1.1) the first time after reboot everything is working. But at the 2. there comes this page fault at .. ok, but now to the bmi struc well.. changing background color only makes sense if we have a backlight function Zagor: can backlight be changed? I have to checkout the newest code... edx: just on and off wait im still editing calpefrosch: you just need to update at the recorder we can could the intensity, I think -can argh, ok again... we could change the ... yea thats what i thoguht too Zagor: what is this line and shall i use it in my files too? * $Id: rockbox-20020426.log,v 1.1 2002/04/27 21:57:22 adiamas Exp $ edx: yes is it automatically filled by cvs? yes (just removed unneeded lcd code) how do I delete the file lcd.c in the cvs (renamed it to lcd-win32.c) cvs rm file after it was deleted for real hmm, i have to fondle my girlfriend. sorry. edx: can we discuss this on monday ? when do you go to school ? is this usb2 really fast? yes :-) damn it rocks maybe I should by a controller. but my damn adapter is damaged or something. it works with win98/me/2k whatever but not with winxp hehe you have anything but windowsxp, calpefrosch? hehe, don't use xp ;-) I have w2k, xp, linux. what you want ! edx: sounds more like an XP problem and not an adapter problem sure it is an xp problem ok, cu. bye. bye calpefrosch cu edx: monday ? Zagor: do you have my 'cvscheck' tool? <-- calpefrosch has quit () no, what's that? a tiny script to display which files that are modified, by you or in the repository I find it handy to use haha ;) with windows xp i got a shell extension that displays modified files in red others in green and i can click-commit/add/remove etc :o) but what about "needs patch" ? you need to check with the server to know that ;-) what is "needs patch"? when the file has been modifed on the server hm now i just rightclick -> update then righ t but my script displays the status, it doesn't change it hm. damn. the tetris code is not working anymore :/ uh ok.. got the bug :) * Bagder salutes edx * edx smiles would it be useful to have getbacklight function returing a bool... ? edx: why? yes why ? hmm if we have a settings menu it need to determine whether bl is on or off *needs not necesserally settings should say how long the bl timeout should be, not permanent "on/off" ahhh that is what this setting does ok im working on the file functions now.. of course we *could* have an on/off setting too, but who wants backlight on all day? yea.. Zagor: possibly the guys using them in cars or similar file.h - into firmware or firmware/drivers/ or where? and during night Bagder: good point edx: common ok firmware/common i think we had a port pin that said when the charger is attached, don't we? PA0 according to my notes let me check are you speaking about DCIN ? yes yes int open (char *filename); is that ok, file.h? according to my notes too an int ? int open(const char *pathname, int flags); flags is fake ? edx: ssh into your sourceforge account and run "man open" etc, then you see the posix syntax no, flags for read or write etc only read supported for starters oh... yea.. *stupid me* * Zagor is #ifdef:ing out the write code in the first version :) i don't see the point to check if we can read or write on a file ? we might want to write a file it's not a check. it's wether to open it for reading or writing if you want that function for posix compliance like playlists? but we are not forced to handle all access rights yea.. *really important* a built-in playlist editor it's not about access it isn't worth extra code that's why there's only two arguments it's about if you want to read or write the file. truncate or not etc. it sucks that you have to write playlists on your pc... the access rights is in the *3rd* argument the third.. do we need that? no no k ah yes... but i was speaking read-only checking ok okay we can use that flag to determine if we open OR create a file exactly or append but we don't bother with checking for read-only file or stuff like it no open returns an int seems totaly useless what do you return in reatlity file handle that's an index in a table Does it matter we make it support N opened files yes because that's how posix open() works ... uhm.. guys.. take a look at that: int open (char *filename, int flags); int close (int handle); int read (int handle, void *buffer, unsigned int count); int write (int hnalde, void *buffer, unsigned int count); int seek (int handle, long offset, int origin); int unlink (char *filename); int rename (char *filename, char *newname); is that ok? lseek edx: yup it says seek @ devcon not lseek it's wrong it should be lseek it should be lseek(), right? i don't remember a posix seek ok lseek is posix then.. the opendir.. what is it supposed to do? * Zagor wonders what the 'l' is for * edx thinks it is for long (?) edx: return a dir handle aha... hmm man opendir ;) it's for doing readdir() on hmhm... it is for scanning a directory? yes you need some man pages edx ;-) zagor, it is for long seek with a 32-bit offset with 16-bit cpu (?) i don't remember exactly ok what does readdir do? (sorry.. win32 functions are different there..) ie. what parameters does it get? edx: it's like findfirst/findnext "returns a pointer to a dirent structure representing the next directory entry in the directory stream pointed to by dir." let me check int readdir(unsigned int fd, struct dirent *dirp, unsigned int count); no struct dirent *readdir(DIR *dir); Zagor: wrong readdir ;-) ok thanks. i will not include disk operations into file.h, right? oh edx: opendir and readdir goes into file.h opendir returns DIR then? DIR * just a minute, are you sure ? yes DIR *opendir (char *dirname); int closedir (DIR *dir); struct dirent *readdir(DIR *dir); is that ok? and readdir() takes a DIR * looks fine hm.. i gotta think up a way to do that in windows edx: cygwin ? well, we don't have any code using that function yet, so you have your time ;-) nah.. not cygwin linux ? clean windows code.. :) edx : you are under Windows or Linux ? if under Cygwin, try to find the file ".../cygwin/usr/include/sys/dirent.h" you will find the functions and structure which would likely be changed or maybe not some fields are junks DIR *opendir (const char *);struct dirent *readdir (DIR *);void rewinddir (DIR *);int closedir (DIR *); --> ironi (xircon@m213-101-132-24.swipnet.se) has joined #rockbox hey ironi I just love that web thing Bagder, that web thing? ??? * Bagder tries to find the URL ah ok Zagor pointed me to it Bagder, http://nlinks.org/dev/test.jpg * Zagor has leaked :) yeah i did it for fun while i was bored last night heh :-P that's the one, yes hey .. could anyone send me his dirent.h (the directory structure definitions) ironi: the page looks great :) who is the sexy girl ? Bagder, cool that you like it, maybe I could do something with it as inspiration edx, thx edx: i can edx: I think you'll reach results better if you just write up your own win32 version alan, some girl from photodisk.com :) i did not buy the image, though :) arf badger: yes, but i need to know what entries are in the DIR, dirent structires it's intersting how a photo can change the appearance of a very simple page are you okay for DDC SEND ? me? i guess so edx: actually, you don't. use whatever contents you like. it's just a pointer used as a handle Zagor: except for the one single one used for the file name posix defines a inode number, which you'll have a hard time emulating :) thanks, alan! Bagder: right, he needs d_name "char d_name[] of unspecified size, with at most NAME_MAX characters preceding the terminating null character." inode number is a junk for cygwin yes Zagor, the news entries, are they static? in fact we can use to have start cluster ironi: yes Zagor, ok alan: not a bad idea but edx won't need that in the simulator when you rename a file in ext2fs, you change its inode ? i don't think so Bagder: right, some fileds can be removed mv doesn't change inode, no inode number is what exactly ? it identifies a single file entry I'm not very good on fs internals though i wonder if an inode is a cluster equivalent no start-cluster equivalent I guess a file can change position and keep the same nodeid i'm pretty sure there's a difference between inode and nodeid ? hopefully i have explore2fs under windows, i can browse linuw source explore2fs is amazingly slow :) true hmm.. this might be enough: typedef HANDLE DIR; is it ok to allocate memory for DIR and return the allocated memory and free it in closedir? but with udma 100, it becomes less slower :) for you it is edx: in your emulator anything is ok hmm not necessarily ;) how do you do it on the jukebox device - where do you free the returned *DIR pointer> ? no edx: well have an array of structs static hmm ok and just return a pointer to the first unused isnt that a vaste of memory we want to avoid dynamic memory use and wouldnt it be nicer to have as many dirs open as memory is avaialbe? why we don't want many dirs open why not ;) it takes more memory oh well.. that is not that bad... those few bytes.. dynamic memory causes fragmentation okok.. fragmentation causes instability or at least non-determinism i'm not sure about that oh yes now i aggree dynamic memory will be used eventually anyway oh yes but we don't want to overuse it but we should stay out of it as much as possible in core stuff static array : space wasted when underused but simpler to start with yes dynamic array: if you use the same range for non pertinent structures there would be few fragmentation sorry i mean persistent indeed you must avoid to use persistent structures to avoid fragmentation. quite now use static array it is very easy to use malloc and free with windows but in our firmare it is another story :) heh hmm i actually do have a problem. the simulator needs to call the open function (from io.h) with the same parameters.. how can i make it understand which function is meant? for files you have open files, the same you have foe directories, open directory (DIR *) and how are you gonna do that in linux? i don't understand the question just a minute, what do you want to do ? edx: why simular that at all? simulate use the windows version the openfunction has a little different parameters.. file_open or something similar yea.. we could change names to file_open... * Bagder would prefer not to hm ok stop ! there are two sides the posix side and the specific side for file handling ? you mean there's an API i suppose if we want posix files it is for programming commodity and an implementation? yes #ifndef SIMULATOR define functions #else #include the windwos header file i mean an "open" is handling a table of open file and call the true open_file. to code a fat "open" is not a good idea I still don't get it yes, open will need to do stuff it could call open_file fat_open() fat functions are the real part, and posix functions wrappers with additional stuff yes the posix is an api okay so edx for its simulator just neet to code an "open" which call the true "FileOpen" (?) exactly sorry for mistakes but... yes ? windows already has an open(), doesn't it? you mean a posix compliant library ? hmm.. opendir.. does the first parameter contain wildcards??? no maybe... posix compliant open edx: no hm ok speaking of that, do we need wildcard support anywhere? what does opendir return if it fails btw..? -1? i don't see what for or rather NULL edx: you really need those man pages :) edx: NULL, right ok http://www.die.net/doc/linux/man/man3/opendir.3.html thx Bagder: for the moment, i don't think so jukebox has no keyboard i don't see how we can put wildcard * and ? :) it hasn't? B-] well we put code in it the code could use wildcards I think it *could* be done give me an example :) i cannot find one myself show only files that matchs "*.mp3" we don't write it until we need it wrong not a very good one anyway very true just a note about extension we don't need to include extension in the name, just add an extension type along with the filename ? the file name is stored via USB when will the extension be cut off? no, i'm speaking about the way rockbox can handle files you mean the dir browser? yea right once yo scan a directory that should be a lot smarter, indeed so "*.mp3" is a wrong example ;) no the files are still named .mp3 suppose author "Mr Doe" wants to write his own app continue he wants to store movies on his archos and ? and then scan for them recursively then we could use wildcards we=>he ??? never mind or he has to write his own wildcard function ;) exactly or do it another way i'm just speaking a way to disguish directory, playlists and mp3 files from others files I'm not usually you want only those three kinds of files usually yes using "*.[mp3|m3u]" is not the best way, without speaking about directories oh yes i see a problem can readdir see a directory ? yes all directory entries ok but how can you know if it is a directory ? there's a flag field too or something i hope c:\Programming\CVS Checkout\RockBox\uisimulator\win32\file-win32.cpp(33): error C2065: 'new' : undeclared identifier what the heck! * Zagor groans. C++ ! ;) ohhh just kidding edx, you compile as a C or a C++ ? C used malloc. works but you use a cpp file another thing.. _finddata_t is not defined yea a cpp file but i compile as c heh okay the trouble is here odd choice of name for a C file... that's confusing for C, operator new is not defined hmhmhm now if i use c it is also compatible to cpp files ;) no matter.. ill rename them and don't use cpp, but c extension i'm sure you create this file without having a look on the extension ;) yea i used c extension now well i had a look at the extension thought about it too late cpp is standard extension so i just didnt change it normally compiler, when it finds a .cpp extension switchs to c++ i disabled that. hmm it still has a very strange error. yes but just remember it for the next time typedef _finddata_t DIR; it does not recoginize _finddata_t what is _finddata_t DIR; ? i included io.h i define the type DIR nope it is supposed to be _finddata_t you need to commit some files if you want our help properly i mean the other structure it is a "system internal" strucutre definde in io.h yes i see by the way i see a filename of 260 chars instead of 256 you're right, Bagder, open is defined in io.h... "Long names are limited to 255 characters, not including the trailing NUL." -- Microsoft FAT32 document it is for a filename but 260 should be for a pathname path lengts are max 260 yes in fat? yes nope for windows :) right :) so we know there is no point to try to handle pathnames longer then 260 in rockbox if windows can do it * Bagder snickers devlishly.... yeah, they'll love us if we create files they can't read in windows... but otoh, who would create them? we :) fat_create_dir(), fat_create_file() "to be able to use this product, you must first put a file here => [261 letters]" ;-) hehe * Zagor hungre * Zagor needs food hey try to imagine that we have a longer pathname created with rockbox, how will windows react ? truncated files ? * Zagor go shop me too im going to order a pizza * Bagder considers pizza too... i had pizza yesterday edx: you use _findfirst/next/close ? i updated cvs files for ui simulator.. anyone who has a win32 compiler and can tell me what is wrong? alan: yea alan: only findfirst right now hum it is really rare i compile with VC alan: even typedef of DIR doesnt work :( i can try ok.. :) --- Zagor is now known as Zagor|shopping hwo come you can go shoppign on saturdays.. not in Germany :( lol Bagder, =) Bagder, a kebab-pizza in jönköping is soooo good. he's not in Germany... :-) they don't make them like that in 08 a bit far for me to order I guess ;-) so you can go shopping on saturdays in Sweden? yes hmm we should have that in Germany too :? on sundays too edx, dont you have malls? edx, but of course the grocery stores are open all day in weekens also in germany i would love to have ALDI here, itä's so cheap ironi, im afraid not oh sh*t it takes time to checkout :(( ironi, no when i wa sin st. louis i shopped a lot at aldi's, cause it was so good prices poor modem LOL @ Aldi LADI rocks ALDI prices are ok, indeed :) edx, what file i must open ? .sln ? alan, re-checkout... i added a few files that were missing - sorry the vcprj file my VC has no associate application with vcprj !? shouldn't be .prj ? ok you got VC++ 6? then you have to recreate a project file :/ i use vc7 yea vc6 edx: you might want a slightly different solution anyway hmm i gotta go eating... bagder: makefile? --- edx is now known as edx|eating edx|eating: the DIR struct myst have a d_name[] entry so you need to have your own DIR struct the dirent has that.. the DIR as well? ah silly me :) you're right cya in a few mins * Bagder hides heh edx: please generate a makefile not all people have vc7 you should find a way to allow other vc version to compile your program. * Bagder thinks alan is right edx: normally your vc7 has an option to generate this makefile hum, i wonder if i wouldn't use SDL... --- Zagor|shopping is now known as Zagor :-) <-- ironi has quit (Read error: 104 (Connection reset by peer)) just for info, here is the link for SDL : http://www.libsdl.org/intro/toc.html yes i know you wouldn't code with it :) heh but I don't need to if you do it ;-) I could do it too actually one day see you <-- alan has quit () * Zagor just realized fat_open() has no purpose linus just called to say the donated archos works fine * Bagder horrays it's a 4.53g version, so it's got a new lcd after all did he try the serial port? not yet. we want to be more careful this time, so we'll wait until monday when we can test with the remote control first good idea but that means they stopped using the old lcd very early i'm curious how many people actually have the old model can't be that many no we want the full donated-archos-story on the web/mailling list later yep didn't someone offer to produce irc digests? adiamas did that would be great, since so much discussion is off-list yep no commit mails so far today... i woke up late :) but things have been committed only not arrived by mail ah, ok problem with the sf mailserver then, I guess likely, yes --- edx|eating is now known as edx hey... cool i am about to get my first makefile working lol :-) welcome to the bright side :) always the same lame comment :P hihi how are resources compiled *unskilled* that's a windows question. we're all unix ponytails here :) but I think there's a Resource Compiler. maybe rc.exe ? I don't even know what "ressources" are ok lol we have complete faith in you :) edx is our man! lol the makefile is working *great* goodie ok.. cleanup code missing hmm.. maybe someone of you could look over the makefile again.. i am sure i did some stupid things in there is it committed? yup the last all: line looks a bit odd what should it look like? you should probably have this instead uisw32.exe: $(OBJS) $(CC) $(LDFLAGS) $(OBJS) i think hmmm aha but i'm not sure how to link on win32 i dont really know what it all means - i just guess ;) you want $(LIBS) there too ok, then you need a crash course in make! uisw32.exe: $(OBJS) $(CC) $(CFLAGS) $(SRCS) $(LIBS) $(LDFLAGS) make is a golden tool, you should learn it. if i have this it says it doesnt know how to make blabla.o the second line is the command line you want to use to link the appliction together which make are you using? nmake? yea i know what whas wrong its not .o it is .obj ah, change OBJS = $(SRCS:.c=.o) to OBJS = $(SRCS:.c=.obj) yup just did that works nice but it does much more than it did before it compiles everything twice (?) lol twice? like once for each object and then the command line that follows $(CC) $(CFLAGS) $(SRCS) $(LIBS) $(LDFLAGS) thsi one yes you should remove the SRCS from the link line i only need this one line actually ah ok ahhh yes missed that * Zagor realizes that fat_read() gets very complex without fat_open(), so adds it again :) LINK : fatal error LNK1181: cannot open input file '../../firmware/drivers/lcd.o bj' hm the output file is generated inside the current dir can i specify the source directories via a compiler parameter> ? yes but this should work i does not it produces the output file inside the wrong directory what is the parameter.. ? that's compiler specific bbl <-- Bagder has quit ("http://daniel.haxx.se") hmmmm... ahh i see how daniel did it.. *testing* yay. fat_read() now works. cool :) i fixed the makefile it works now nice shall it remove the obj files after linking? no ok the obj files are used to check which source files are modified i know.. just thought ;) so it doesn't have to recomile unmodified sources hehe make a "clean" target that removes obj files that's the normal way now alan can try to compile and help me fix the other error with the file-win32.h file :/ that's good when is this clean target processed? manually "make clean" $(RM) *.obj ? only if you have an RM variable defined yea i just defined it otherwise just "del $(OBJS)" you don't need a variable, really im about to understand the makefile thing :) yeah, it's not difficult. but very useful. yup. but a finished makefile looks damn freaky for a newbie like me :) hehe, yeah maybe. but then again all kinds of files look freaky when you don't understand them yea... do you get a mail for everything i commit *sorry for committing that much* normally, yes. but currently there seems to be something wrong with the sourceforge mail server don't worry about committing much. that's what cvs is here for maybe its down because i spammed it with commits (haha) :) * edx can't wait to wear his rockbox shirt * edx is angry that he misspelled irc server hehe irc.openprojects.net... why dont they set up an irc server there? aehmm. org irc.openprojects.org lol yeah, i was surprised their dns hasn't fixed that well.. maybe it will be someday... then the shirt is fixed too lol why not mail them and ask/suggest? hmm good idea (don't mention the tshirt :) Ladies and Gentlemen, due to the fact that i misspelled your server on a *very* important shirt, i want you to open an irc server on irc.openprojects.org ... LOL maybe not the best chance of success :) that was rather a joke ;) yes, i understood i need somebeody do fix that bug in file-win32.h i just cant do it i'll take a look hm you cant compile it... you want to see the error messages? ok c:\Programming\CVS Checkout\RockBox\uisimulator\win32\file-win32.h(27): error C2061: syntax error : identifier 'DIR' c:\Programming\CVS Checkout\RockBox\uisimulator\win32\file-win32.h(27): error C2059: syntax error : ';' well where is DIR defined? that means that _filedata_t is not defined for any reason. but i do include io.h dir is defined right in that line typedef _finddata_t DIR; ah strange indeed... you should probably free(p) before return NULL; that is a good idea i did nto code any further when the DIR thing didnt work i understand hmmmm that is so strange looking at io.hi find something really interesting: #ifndef _POSIX_ maybe io.h supports the posix functions :D yes i think it might ok.. one more thing i can try.. just rip the definition out of io.h an paste them before the DIR typedef lol (this is not afinal solution of course) haha whatever works hmmm if i include io.h AND define the struct it says redefinition something is *very* bad about this... #define DIR _finddata_t lol the problem is not the DIR thing. the problem is teh _finddata_t definition - it must be somewhat strange what if it's a macro? it is not a macro - i looked at it ok struct _finddata_t { unsigned attrib; time_t time_create; /* -1 for FAT file systems */ time_t time_access; /* -1 for FAT file systems */ time_t time_write; _fsize_t size; char name[260]; }; does typedef work with structs in C? lol yes it must be somethign really stuped stupid if I put dirent there it wont work either maybe it is the DIR - it is defined somewhere else??!!! hm no it is not --> elinenbe (trilluser@bgp01080511bgs.wanarb01.mi.comcast.net) has joined #rockbox hi hi elinenbe hey there. You sure have been busy. brb I looks like the fat32 code is coming along. <-- edx has quit () are you getting the mails? i dont --> edx (edx@pD950D222.dip.t-dialin.net) has joined #rockbox cvs mails, i mean oh finally.. it hadnt logged the irc sessions before --- edx is now known as edx| --- edx| is now known as edx but, yes, the fat32 code is coming along nicely. i just got read() working and am writing seek() right now In a Mr. Burns tone: exxxxcellent hehe --> Bagder (~daniel@as3-3-2.ras.s.bonet.se) has joined #rockbox hi hi again * Zagor grooves to Electric Skychurc is that a band? lol if that ain't a trip :) what is that? it's ambient dub electronic smooth beats heh silk for your soul :) * Zagor is a poet * edx grooves to emil bulls hard-rock (lol) you dont know emil bulls in Sweden, do you? well at least I don't. hard rock is not my taste. hehe :) they are German - they sing English but I guess they are not known anywhere but in Germany seek() done, time for long filenames kewlness great * Zagor is in The Zone :) * Bagder thinks Zagor listens to weird music what music do you listen to, Bagder ;) http://storebror.haxx.se:443 oh, he's very middle-of-the-road... try my stream just doing :) cool... what a quality i like the music :) what is this? *hey stream's gone* what does your player? say heh nothing look in the player window, the title shows there youst wont play anyithin.. press play again try again some tracks don't have an id3 tag my server runs all the time they should appear based on file name I think lol it says nothing for the title... but what is it * edx needs that as well heh I have only 3000 files on that stream Bagder: I found some tracks just don't show anything * edx laughs out loudly! which player do you have edx? *only* 3000 files??! windows media player ;) hm ok how many GB of mp3s do you have? I've mainly listened with winamp that's about 15-16 (wmp = the lames music listen tool there is haha) bjorn:~/music>find . -name "*.mp3" | wc -l 5588 :) heh 3189 here actually the best part: they all fit on my archos edx: try this http://storebror.haxx.se/mp3/ hehe * edx found 6783 mp3s on his hd that music repository is cool :) * elinenbe found 10095 MP3s <-- all Depeche Mode! hehe loool they sure are productive, those DMers what tool do you use to set up those streams? ample ampl.sf.net ample.sf.net * edx does not know that dead simple little thing as it should be is it for windows too? its full source but *nix style so you'd need to fiddle ;-) too sad :) get cygwin and it'll be less fiddling :-) is it command line or xwindow? neither its a server it says nothing, it just runs ;-) even better the way it should be, imho but the file accessing functions will have to be changed.. *too much work* not if you build with cygnus yea well... how much unix-specific code can it be? dir-reading socket stuff, yeah just about nah, most socket stuff works the same ok "most" like close needs to be replaced with closesocket() hmmm ... ill have a look at it.. hey.. only 10 files that looks like one could actually do it its only 6 source files hey, I bet that if you do it nicely and mail back a diff to the guy, he'll be grateful * Bagder did that already hmmm waht does config.h in linux? its is commonly generated by a script defining what features your particular machine/OS/CPU/environment has it is not included with the source though no, since it is generated you wanna see how mine looks like? ok vsnprintf is that the smae as sprintf almost http://storebror.haxx.se/ample-config.h whats the difference? thanks check vsprintf(), you must have that in windows too the 'v' means it takes variable arguments from your input parameters the extra n is only for the extra argument specifying the maximum output length but it does not take int as second parameter <-- elinenbe has quit ("ChatZilla 0.8.6 [Mozilla rv:0.9.9+/20020414]") * edx does not have the time to work through all that right now I fully understand that heh.. i guess it would need a week of free time to fully translate it there are tools for windows, certainly sure but finding a good free one is another issue probably windows is polluted with silly shareware and expensive software sure well - it is like free software most of the time - I dont spend too much money on software ;) i just found a mp3server.. im gonna try it :) 16kb/sek arent enough anyways lol check this out: http://codeforce.d2g.com:12345/file username: edx / pw: edxmp3 is a stream somewhere too? is there click on a file... should work.. well it just links to the mp3 hmm yes, then I download the mp3 well in windows you listen to it directly if you want to but only one file at a time *cool - accessing all my mp3s from school* jup ha not comparable to ample fun anyway ;-) yea hm i have to reconfigure.. right now it acccesses mp3s via LAN - ill store only a few mp3s on the server - then they are even accessable if my computer is not turned on well... gotta go... good n8 :)_ bye edx --- edx is now known as edx|sleeping bye good neight? ;-) sounds dutch LOL I wonder what it means? ;-) haha if somebody with windows ever gets here tell him to debug my code :o) cya yes sir! * Zagor just realised he has a strawberry/cheesecake ice cream in the freezer mmmm... yikes vfat is a mess fat32 is alright, but vfat is really bad each name is split into (up to) 13 entries each entry is split into 3 parts and to top it off the entries are stored backwards * Zagor is crying vfat, that's for the long file names only? yes you committed any working code? --> Linus (~linus@labb.contactor.se) has joined #rockbox not yet hey, linus! Yo! I'm curious hej Linus Hejsan! Whazzup? fat32 is working. open/read/seek is done. grinding away on vfat now. Ninja! Bagder: clarification: I have committed the working fat32 code, but not any vfat code so, theoreticly we should be closing in on reading a file now * Bagder updates yes. but we need the upper layer written right the fat code handles sectors, not bytes i figured that would be a fair tradeoff so the upper layer handles accesses for bytes within a sector agreed? Fair enough. indeed yesss... single-entry longnames working. I bet they hardly ever occur ;-) hehe are they fixed-size per entry? all files get longname entries, no matter the length, since longname entries have fewer restrictions on the name yes, 13 characters per entry wooooooooooooooooooooooooooo works C:\> dir ata: Reading sector 76 config-player.h (59) config-playerold.h (60) now on to cross-sector names... really neat prompt you made for your test program ;-) good old DOS-style * Zagor blames Linus yeah yeah we all do that :-) hehe * Linus denies everything bläääää, longnames can span THREE sectors :-( * Zagor needs a new strategy * Linus need som sleep to cope with the MAS music orgy tomorrow... :-) hehe. good night :-) nighty nighty Zagor. The charger is fried on your new archos. I will try to find out what needs to be replaced. ok Nighty. <-- Linus (~linus@labb.contactor.se) has left #rockbox