Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2005-05-03

00:06:09*HCl hands amiconn a glass jar and a bug catching net.
00:06:14*HCl goes to sleep
00:09:10*HCl decides he's hungry and goes to make and eat food first :/
00:10:08ehntoofood is good.
00:13:07MoosCamarohi guys
00:13:33MoosCamaroamiconn: good luck for your soldering
00:13:58amiconnIt's already done−−−−−−−−, Ondio working again
00:14:22amiconnOops, I should pay attentin where to rest my finger
00:14:26MoosCamarocongratulations, resurection
00:17:03MoosCamaroamiconn: what's about your h140, have you receive it
00:17:26amiconnNope. Seems I have to try other sources :-/
00:17:54MoosCamaropostals problems?
00:18:36amiconnNo, shop simply doesn't ship it. Probably they went out of stock, but they should have informed me about that
00:20:24MoosCamarolot of people want it, and with the production's difficult to find one in perfect state
00:21:01 Quit lolo-laptop ("Client exiting")
00:22:23ehntoothere's always ebay:
00:22:47ehntoooh... it's even from misticaudio
00:23:22MoosCamarogood price
00:23:56MoosCamarobut his state isn't guaranty
00:24:51 Quit Harpy (Read error: 60 (Operation timed out))
00:25:08 Quit ashridah ("Leaving")
00:26:13amiconnHCl: Bug fixed. :)
00:27:11***Saving seen data "./dancer.seen"
00:27:17 Join TCK [0] (
00:27:38 Quit coob (Read error: 110 (Connection timed out))
00:27:59MoosCamaroehntoo: it's out of stock with Mistic Jeff
00:30:57 Quit Patr3ck ()
00:45:46 Join austriancoder [0] (~austrianc@
00:45:49austriancoderhi all
00:46:31MoosCamarohi austriancoder
00:47:17*austriancoder updates his gentoo laptop
00:49:17t0mashi austriancoder
00:49:37t0masweren't you working on the fm radio support for iriver?
00:52:19austriancoderi should do a commit.. but i haven't that much time the last week
00:52:20t0mashow's that going?
00:52:30austriancoderi2c stuff is done..
00:52:56t0masah.. and the TEA chip is the same as in archos?
00:53:22austriancoderso only driver porting.. that was a fine thing
00:54:36austriancoderthis weekend is very long 6 days free.. so i think i will commit my stuff at this time
00:55:10MoosCamarogood news, it's so cool :)
00:56:07austriancoderit will be cooler if the audio api could be tested ;)
00:56:22MoosCamarohihi :)
00:56:51BagderI've tested it
00:56:54Bagder15 seconds
00:56:56 Join lostlogic [0] (
00:57:36t0masthat's 3 times Linus's time ;)
00:57:44t0masBagder: saw my privmsg?
00:58:07Bagdert0mas: it was after Linus and I improved his initial code
00:58:17Bagdert0mas: yes, but I cannot comment since I don't know what it means
00:58:23t0masah ok
00:58:25Bagder"my own malloc"
00:58:39t0massimple function to know where in pluginbuffer I am
00:58:47t0masand return subsequent pointers in there
00:58:54t0masor return an error when using to much space
00:59:16Bagderin general, you can consider yourself boss over that plugin
00:59:23Bagderso you can decide those things
00:59:37t0masyes, but it was you complaining about to much stack ;)
00:59:51t0masso I wanted to know if this would be a good way of solving it
00:59:53BagderI was concerned you'd blow the stack, yes
01:00:06BagderI don't remember how big it is
01:00:36Bagderbut sure, it could be a fine way to avoid smashing the stack
01:00:37t0mashm... pretty big
01:01:09Bagder#define DEFAULT_STACK_SIZE 0x400 /* Bytes */
01:01:44t0mashm... then I would have blown it... but something saved me :P
01:02:04t0masmy new version uses 116 bytes
01:02:15 Join webguest70 [0] (
01:02:53Bagderthat define is only the default, some threads have larger
01:04:27t0masanybody with an archos around/awake?
01:05:24webguest70anybody with rockbox on iriver here?
01:05:43BagderI'm off for bed
01:05:44austriancoderwebguest70: yes
01:05:48t0masgood night Bagder
01:05:51austriancodergood night
01:06:02amiconnt0mas: me
01:06:11t0masok, have 5 minutes to test?
01:06:13amiconn(with archos(es))
01:06:38amiconnSure (although I found another bug, meh)
01:07:13t0masthere's the explanation on how to create a datafile..
01:07:32t0masand I'm committing a new version right now :)
01:07:32 Quit webguest70 ("CGI:IRC (EOF)")
01:08:20 Join webguest70 [0] (
01:09:37webguest70damn connection troubel..what i wanted to asked is how i can get rockbox on my iriver ,too? you got a link or something?
01:09:50t0mashmz... are you a developer?
01:10:00t0masor just wanting to play gameboy games?
01:10:12webguest70me−−nope..just a user wanting play a little and so on....
01:10:25webguest70one of these kiddies
01:10:29t0massee here:
01:10:39t0masand don't ignore the warnings...
01:11:00t0masmake sure the md5 of your update file is the same as one in the wiki..
01:11:18MoosCamarowebguest70: rockbox don't play music now
01:11:35webguest70i know it doesnt play music now
01:12:07t0masread that one..
01:12:14MoosCamaroif you have few hours free read
01:12:18t0masand this:
01:12:21webguest70thats what i where looking for
01:12:25t0masand then decide if you really want it
01:13:08t0masamiconn: saw my update?
01:13:17MoosCamarowarning rockbox is developing
01:13:52HClugh, hello.
01:14:04HClsome of my flatmates are such jerks, especially the gay one..
01:16:58amiconnt0mas: Grabbed it, building datafile...
01:17:05t0masok, thanks
01:18:01amiconnUgh, the .desc is 16 MB, the .index is 5 MB
01:19:49t0masthe 16 mb desc is just the data of the input files combined... and the 5 mb index is because of the fixed wordsize..
01:20:40amiconnHmz, that doesn't fit my Ondio as-is.
01:20:49amiconn(only 500 KB left)
01:22:00t0masand recoder?
01:22:11t0mas(as I haven't tested on any archos at all)
01:22:27amiconnYes, either I'll do it on my recorder, or throw out some music from the Ondio
01:22:37amiconnI found a bug in rdf2binary
01:22:57amiconnIt uses the wrong mode for the destination files; they end up write-only
01:23:21amiconn$ ls -l dict.*
01:23:21amiconn−−−−−−−−−− 1 Administrator Kein 17335684 May 3 01:16 dict.desc
01:23:21amiconn−−−−−−−−−− 1 Administrator Kein 5216184 May 3 01:16 dict.index
01:23:21DBUGEnqueued KICK amiconn
01:23:21amiconn-rw-r−−r−− 1 Administrator Kein 19142302 May 3 01:16 dict.preparsed
01:23:34amiconn...even no-rights
01:23:49CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
01:23:49*t0mas prods Slasheri
01:23:49 Quit webguest70 ("CGI:IRC (EOF)")
01:24:35*amiconn rebuilds current cvs for recorder
01:26:56amiconnBugs seem to recurse-multiply themselves :(
01:27:15HClit usually depends on coding style :/
01:30:50amiconnt0mas: dict.rock seems to work correctly. It's really quick, almost instantaneous on recorder
01:30:57amiconnOnly thing missing is scrolling...
01:31:14t0masthe scrolling thing is on my todo list
01:31:35t0masand it wastes 5 mb of diskspace to be quick :)
01:31:40t0masso it should be ;)
01:31:55amiconnWould it be possible to make this optional
01:32:09amiconn(so to be able to trade speed for space)
01:32:15t0mashm... it will be _really_ slow without the index
01:32:21amiconnThinking Ondio here...
01:32:32t0masyes, but isn't ondio cpu slower than iriver?
01:32:38amiconnMaybe it's slow there anyway; MMC access is sloww
01:32:57amiconnCPU is almost identical to recorder, even a little bit faster
01:33:01t0masyes, but we don't need much file access now
01:33:17amiconnIt just seeks to the necessary position?
01:33:25amiconnI'll check...
01:33:29t0mascheck debugf :)
01:33:39t0massearching for testy takes it 14 read's
01:33:51t0mas14 * 36 = 504 bytes
01:33:59t0masand then the description from the other file
01:34:08amiconnYes, but the seek() reads the fat
01:34:32amiconnI don't know exactly how much overhead that is, let's see
01:34:43t0masbut searching through it without the index would require reading the whole file... until we found the right word
01:35:03amiconnReading the whole 16 MB linear would take approx. one minute on Ondio this isn't an option imho
01:36:11amiconnLet's waste the 5 MB if it pays off
01:36:59t0masdo you know how struct-packing is working?
01:37:29amiconnUsually the members are aligned to member-sized units
01:37:56t0maswould it make any sense here?
01:37:58amiconnSo it's worth to group equally-sized members together
01:38:16t0masso: not usefull here?
01:38:50amiconnThe alignment must be done, otherwise the SH1 will crash with an CPUAdrErr exception
01:39:55amiconnIf you want to save space, group the members. 4 times a 'char' , then 4 times a 'long' gets no holes, but ordering them char/long/char/long... produces 4 3-byte holes
01:40:16t0masah, that's the idea
01:41:25amiconnAn even better idea is to put the shortest members at the top (char, bool (?? dunno if bool is 1-byte on coldfire too))), then short, the longs and pointers
01:41:54amiconnThis is not a space optimisation, but a speed optimisation with almost no tradeoffs on SH1
01:42:03austriancodergood night all..
01:42:13amiconnaustriancoder: night
01:42:32 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.12")
01:42:43t0masamiconn? your bug in the file permissions?
01:42:52t0masmight that be the open() and write() call's?
01:43:11t0masI changed them to fopen() and fwrite() and everything is working fine here..
01:45:25amiconnYOur open() calls are missing the mode (3rd argument; seems to be optional since you got no error)
01:45:42t0masit's not my tool
01:45:50t0masbut I changed it to fopen() and fwrite()
01:45:50amiconnIt seems the mode defaults to 0 (nothing) if left out, at least on cygwin
01:46:28t0masyou see any reasons for using write() and open() instead of fwrite() and fopen() ?
01:46:55MoosCamarogood night amiconn/t0mas
01:47:01amiconnopen() and write() are lower level, fopen() etc are higher level
01:47:14t0masMoosCamaro: night
01:47:26t0masamiconn: changing them here doesn't hurt right?
01:47:45 Part MoosCamaro
01:47:46amiconnNot on the PC
01:48:10amiconnrockbox doesn't provide the f* i/o functions
01:48:35t0masI know, but it's a pc tool :)
01:48:42t0masand I can't see why he used open() and write()
01:49:28amiconnI'd personally do the same; this way I wouldn't need to 'switch' thinking
01:49:52t0masfrom rockbox to pc?
01:51:03amiconnTried dict.rock on Ondio. Impressive speed :)
01:51:08t0masbut he hasn't coded anything in rockbox yet... afaik
01:51:30amiconnEven searching for a word coming late in the alphabet took ~2 sec
01:51:31t0masanyway.. just changed them to fopen() and fwrite()
01:51:41t0masyes, late isn't harder to find
01:51:49amiconnIt's not?
01:52:07 Quit rasher ("leaving")
01:52:08t0masit starts in the middle, and checks if the searched word is higher or lower
01:52:24amiconnAnyway, the point is that even on Ondio it's perfectly usable
01:52:24t0masand then splits that half again, and checks again
01:53:04t0maswhat's the slowest device in the rockbox-family?
01:53:40amiconnCPU-wise the recorders are slowest (11.0592 MHz). Player and Ondio are equal cpu-wise (12 MHz)
01:53:49t0masah ok
01:53:58t0masso ondio has a disadvantage here?
01:54:01amiconnThen Ondio has that sloow disk access
01:54:09t0masyes, that's what I ment
01:54:37amiconnThe MMC is hooked up to the SH1 SPI, which can do max. 3 Mbit/s at 12 MHz
01:54:39t0maswell... 2 seconds isn't a problem :)
01:55:00amiconn..and it needs bitswap because of the braindead spi implementation of the SH1
01:55:04t0mas5 or more would get annoying... but is still faster than a nokia phone :P
01:55:37amiconnReal-world read speed from an MMC is around 300 Kbyte/s
01:56:11amiconn(with all threads yield()ing sufficiently fast)
01:56:37t0mas5mb / 300 kb = 17 seconds for reading the index...
01:56:42t0masoh wait
01:56:54t0masfor lineair search it would need to read the full 17 mb
01:57:13amiconnAs I said, ~1 minute
01:57:38t0masthen I can't thing of a better way atm
01:57:56t0masthe index can be a little smaller... when changing the WORDLEN define..
01:58:07amiconnThe hd read speed of the disk-based archoses is ~4 MByte/s due to hgihly optimised asm pio transfer routines
01:58:32t0mashm... sounds not really fast for a HD
01:58:38t0masbut I guess that's the cpu limit>
01:58:54amiconnThis time no itswap is needed, but endian swap
01:59:05amiconnOtherwise we would use DMA of course
01:59:18t0masendian swap?
01:59:38t0masall read data needs to be swapped? :|
01:59:44amiconnYes. ATA is little endian (guess why), but the SH is big endian
02:00:01amiconnThe iriver does that in hardware
02:00:12amiconnAll read data and all write data needs to be swapped
02:00:19t0masah, iriver did something right in the design :)
02:01:18amiconnThe swap itself is not too costly; only it doesn't allow us to use DMA
02:02:02 Join DMJC [0] (
02:02:12amiconnThe SH1 does have special instructions for swapping 2 bytes (and 2 words) of a register
02:03:36*t0mas still has a lot to learn about ASM on the targets...
02:03:51t0masand bit/byte swapping and magic in general...
02:03:56t0masas I never did that before rockbox
02:04:17amiconnSee ata.c, lines 382ff and 690ff. I once wrote that stuff
02:05:25*t0mas notes in his rockbox info .txt
02:05:33t0massomething to read when I have some spare time
02:05:58amiconnThe spare time thing is a strange one...
02:06:09t0masyes, not happening a lot :)
02:06:56amiconnI could test dict.rock on the player as well, but that one really needs scrolling...
02:07:07t0masyes, just two lines?
02:07:10amiconn2 lines with 11 chars each are not too much
02:07:19t0mas(it doesn't display the searched word when you have < 4 lines)
02:07:48t0masbut still... 2x11 is to small :)
02:10:39 Join Quelsaruk [0] (~kvirc@
02:12:06 Quit DMJC ("Leaving")
02:13:22amiconnhi Quelsaruk
02:14:59Quelsarukhi amiconn!
02:15:08Quelsaruki've been away long time
02:15:20Quelsarukshouldn't you been sleeping?
02:18:36*t0mas should be sleeping...
02:18:45t0masshould've been for a few hours already...
02:18:50t0masas I have to get up early tomorrow...
02:18:52t0masso good night
02:18:54amiconnt0mas: It's working on player too (only that I can merely read the first 2 words or so)
02:19:06t0masamiconn: thanks for all the testing :)
02:19:10t0mashow fast is it on player?
02:19:15t0mas1-2 seconds too?
02:19:27amiconnMost of the time is the disk spinup
02:19:30t0masok, then the next thing is scrolling :)
02:19:38amiconnMy player has that dreaded Hitachi disk
02:19:38t0masghehe... raise the timeout... and type faster :P
02:20:03amiconnThe virtual keyboard is a bit slower to use on player too
02:20:10amiconnRemember: only 2 lines...
02:20:12t0massaw that in simulator
02:20:19amiconnAh, yes
02:20:22t0maswith the disk spinning on iriver it's < 1 sec
02:20:39amiconnI guess it's the same on the jukeboxes
02:20:50amiconnI have my disk timeout set to 5 sec
02:20:57t0masah, I have 10 to 20
02:21:10t0masdon't know for sure...
02:23:00t0masbut I'm away... sleeping :)
02:23:01amiconnA suggestion for better behaviour: check the existence of the .index and the .desc before opening the input screen
02:23:04t0massee you tomorrow
02:23:21t0masyes, to avoid typing for nothing? :)
02:23:30amiconnAnyway, g'night
02:23:36t0masgood night
02:24:20Quelsarukgood night t0mas
02:26:13 Quit Seed (Nick collision from services.)
02:26:20 Join Seed [0] (
02:27:13***Saving seen data "./dancer.seen"
02:28:37 Quit Seed (Nick collision from services.)
02:28:45 Join Seed [0] (
02:35:42 Quit Quelsaruk ("KVIrc 3.2.0 'Marmalade'")
02:36:10 Join DMJC [0] (
02:46:37 Quit thegeek (Read error: 54 (Connection reset by peer))
02:46:52 Join thegeek [0] (
03:00:01 Quit cYmen ("zZz")
03:08:31 Quit Aison (Read error: 104 (Connection reset by peer))
03:33:45 Join bagawk [0] (~Lee@bagawk.user)
03:58:49 Quit bagawk ("Leaving")
04:05:27 Join QT_ [0] (as@area51.users.madwifi)
04:10:20 Quit QT (Read error: 148 (No route to host))
04:27:17***Saving seen data "./dancer.seen"
04:54:50 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
05:16:23 Quit ehntoo ("Leaving")
05:26:25 Part DMJC ("Leaving")
05:44:06 Join ehntoo [0] (
06:27:21***Saving seen data "./dancer.seen"
06:48:35 Join noclue2 [0] (
07:05:21 Quit ehntoo (Read error: 110 (Connection timed out))
07:13:39 Join bobdbob [0] (
07:32:11 Part bobdbob
07:37:54 Join bobdbob [0] (
08:12:14 Quit bobdbob (Read error: 104 (Connection reset by peer))
08:27:22***Saving seen data "./dancer.seen"
08:32:10Rickrather, morning-evening
08:44:43 Quit edx (Read error: 110 (Connection timed out))
08:49:59 Join tor8008 [0] (
09:01:23 Quit tor8008 ("CGI:IRC")
09:18:26 Join Harpy [0] (
09:18:42 Join bobTHC [0] (
09:18:48bobTHChi folks !
09:18:52 Nick Lynx_awy is now known as Lynx_ (HydraIRC@
09:23:39 Quit Bagder ("Off to search for that connect-resetting peer guy!")
09:26:32 Join Bagder [0] (
09:27:51 Join ashridah [0] (
09:28:21Bagdert0mas: open() caused problems because you only used two arguments
09:28:29Bagderint open(const char *pathname, int flags, mode_t mode);
09:38:15Rickwhat is mode_t mode?
09:38:28Rickor flags
09:38:40Bagderflags is the O_ things
09:38:48Rickthat was there before
09:38:51Rickbut what's mode_t mode?
09:38:51Bagdermode is *nix for what permissions to set
09:39:23Rickcorrect me if i'm wrong (i'm tired), but arn't we using fat32?
09:39:39Bagderon target, yes
09:39:53Bagderbut not when the sim runs
09:40:01Rickoh, true that
09:40:30Rickalthough I don't see why you would add a mode to the file api of rockbox
09:40:33Rickto fix that
09:40:49Bagderbut t0mas problem was in a tool made for the host
09:40:56Rickok ;p
09:43:09 Join Zagor [0] (
09:43:17Zagorhey fellas
09:43:22bobTHChi mate !
09:44:45Zagorhave you heard the latest silly anti-GPL suit?
09:44:46*Bagder 's sound is gone today...!
09:45:38Zagor's a laugh
09:45:58ZagorBagder: been fiddling with your kernel?
09:46:56 Nick QT_ is now known as QT (as@area51.users.madwifi)
09:47:13Zagordwihno: yeah, someone's complaining "you're giving stuff away free! then people won't buy things from me! unfair! unfair!"
09:47:42Bagderamericans ;-)
09:47:47*Bagder ducks
09:49:50Bagderyday, a large virus batch was sent using my from email address
09:49:58BagderI say large
09:50:00Zagorooh, fun for the whole family
09:50:12Bagder~2000 bounces and counting
09:50:30Zagorany hate mail yet?
09:50:51Bagderprobably, I throw them all together and delete
09:54:10Bagderseen ?
09:54:29Bagderneat little thing
09:55:00Bagder(for those who understand swedish)
09:56:21bobTHChe which kind of test is it ?
09:56:23bobTHCsec ?
09:56:33Bagderit runs nesses on your machine
09:56:45Bagderand presents or emails the results back
09:57:22bobTHCgood servlet usage
09:57:59Zagor"Tyvärr var det inte möjligt att testa säkerheten på din dator. Det kan bero på följande saker: Din dator stängdes av under tiden testet genomfördes. Internetuppkopplingen mellan testverktyget och din dator avbröts."
10:00:39dwihnoThe obvious: You don't have any computer security whatsoever ;)
10:02:24ZagorDo I need a computer for this?
10:03:03*Zagor just got Hare Krishna spam
10:03:06bobTHC"Unfortunately each the not possible to test the security on your computer. The can depend on the following things: Your computer was shut down meanwhile the test was implemented. The the Internet connection the average physical tool and your computer was aborted"
10:03:41bobTHCraw translation of course
10:07:20Bagdermore coffee needed
10:10:25 Join Patr3ck [0] (
10:12:02Bagder"The remote host answers to an ICMP timestamp request. This allows an attacker
10:12:03Bagderto know the date which is set on your machine. "
10:12:19Bagderinformation leakage
10:12:22 Join edx [0] (
10:12:49Bagdersomehow I'd say they could estimate that pretty good anyway
10:13:31Rickwhat is nessus?
10:14:01bobTHCnessus is a famous sec audit tools
10:14:16Bagderit is Your Friend ;-)
10:14:27RickI don't need any auditing applications
10:14:35Rickbeing behind my router is enough
10:14:55Ricknot like I care if someone knows what time it is on my box
10:16:18bobTHClol, and u think your router avoid any flaw ?
10:16:33RickOf course not
10:16:36Bagderwhat, routers can have flaws?
10:16:43Ricki'm not that stupid
10:17:11Bagderit seems "Firewall ECE-bit bypass" is the most common high risk of the tests
10:17:15Bagdermade by that service
10:18:01Bagder(The remote host seems vulnerable to a bug wherein a remote attacker can circumvent the firewall by setting the ECE bit within the TCP flags field)
10:18:43Rick'the firewall'?
10:19:12ZagorBagder: interesting. it didn't say that for me.
10:19:29BagderZagor: and you allow it too?
10:20:03Zagorno idea :-)
10:20:24Zagori guess my router disallows it by default
10:20:32BagderI allow some icmp stuff
10:20:59bobTHCzagor > it's a brick or concrete wall ? ;)
10:21:14Zagorpaper, mostly :-)
10:22:07bobTHCit seems to be at least cardboard ;)
10:22:17Rickcardboard is paper
10:22:37bobTHCyes but a little bit solid
10:22:48Zagori get the icmp timestamp warning, but no other icmp issues
10:23:03Bagderme neither, only the timestamp thing
10:23:20Zagorit did find a freaking VIRUS on my wifes laptop though...
10:23:20Bagderrather harmless
10:24:03BagderI noticed it has a warning it can show that goes like "copyright protected material possibly found"... :-)
10:27:04Zagorstrange test. all mp3,asf,mpg,mpeg and ogg files trigger it.
10:27:24***Saving seen data "./dancer.seen"
10:28:24Zagorlovely bug: "UDP packets with source port of 53 bypass firewall rules"
10:28:33Bagdervery common thing
10:28:54Bagdersince it is for dns lookups
10:34:21amiconnt0mas: r u there?
10:36:43 Join bobdbob [0] (
10:36:59ashridahBagder: it doesn't have to be, however. you can force dns servers to not use that port as the source. i don't think i've run into any instances where i've had trouble with it.
10:37:16Bagderyes, and you can limit to only accept that from certain hosts
10:37:35Bagderthat's still the reason why so many allow it
10:53:38Zagorbah, the virus warning was a false alarm. it will trigger on any auth server.
10:58:52 Join amiconn_ [0] (
11:07:24 Quit amiconn (Read error: 60 (Operation timed out))
11:07:25 Nick amiconn_ is now known as amiconn (
11:14:50t0masamiconn: I am now :)
11:15:17amiconnI have two more suggestions for the dictionary plugin
11:15:39t0masok, just dump them in my privmsg next time ;)
11:16:11amiconn(1) The max. word length could be defined in the index header. This way the index is only as large as it needs to be for the current dictionary
11:17:00amiconn(2) Imho it would be better to combine the two files into one. This avoids the situation where the .index is there, but the .desc is missing
11:17:13t0masok, both was something I have tought of...
11:17:20t0maswas -> are
11:17:24amiconnThe file could then look something like this:
11:17:51t0masbut (1): the reading will be much more difficult... but I can try...
11:18:29t0mas(2) that's not a problem... but the rdf2binary tool should buffer and then calculate the offset... using a lot more memory
11:18:30amiconn(header (MAGIC, version, max_word_length, pointer_to_start_of_descs)) (index n*(index_entry)) (descs)
11:19:02t0masyes, and index_entry is max_word + 4 now
11:19:03amiconnWhy would reading be much more difficult if the max. word length is variable?
11:19:24t0maserr... no... wait.. I see :)
11:19:32amiconnOnly thing is that max_word is no longer constant
11:19:42t0masyes, so the struct can't be used...
11:20:09t0masthe max word len is in the struct
11:20:47t0masor can I define a struct at runtime?
11:21:00t0masafter reading the word_len?
11:24:10amiconnYou can't define the struct at runtime, but you can do the following:
11:24:32t0masread(word_len); read(4)
11:24:35t0masread separate?
11:24:46amiconn(1) Put the string at the end of your struct, and define it as a zero-length array (!)
11:25:10Bagderor a one-byte if you want it ansi clean
11:25:47amiconnThe only thing that does need changing is that sizeof(struct) doesn't give you the length of an index entry any more, you need to use sizeof(struct)+max_word instead
11:26:10amiconnZero-length arrays:
11:26:34amiconn"They are very useful as the last element of a structure which is really a header for a variable-length object:" </quote>
11:26:38BagderI don't think we should be using gcc extensions
11:26:39t0mashm... but when defining the struct with a 0 byte array... I still need to malloc() right?
11:26:59t0masotherwise I'll overwrite some other mem... after the struct
11:27:41amiconnBagder: Why not? We're already doing that is some other places
11:28:44BagderI don't like it anyway
11:28:52Bagdersince I don't want us locked in with gcc
11:29:27Bagderat least not more than we need to
11:30:14amiconnWe could also use a C99 flexible array member here, but that renders sizeof() non-working
11:31:03t0maswhy not use 1 byte/
11:31:25t0massizeof(struct) + max_word - 1
11:31:31amiconnYou could do that, but then you always have to take this 1 into accout
11:31:45t0masyeah, but it would be ansi compatible..
11:31:56t0maswich is a good thing if you ask me...
11:32:58t0masbut I'm away again...have to finish some things before lunch break
11:34:07amiconnPersonally I wouldn't care that much. (1) We're already using numerous gnu extensions (__atrribute__, asm() with operands etc) (2) Using it here would make the code easier to read
11:45:18 Join ferenczy [0] (
12:03:14 Join StrathAFK [0] (
12:04:32 Quit Seed (Nick collision from services.)
12:04:32 Join Seed [0] (
12:04:33 Join amiconn_ [0] (
12:07:51 Quit amiconn (Nick collision from services.)
12:07:51 Nick amiconn_ is now known as amiconn (
12:18:20 Join cYmen [0] (
12:21:17 Quit bobdbob (Read error: 110 (Connection timed out))
12:22:06 Quit Strath (Read error: 110 (Connection timed out))
12:27:28***Saving seen data "./dancer.seen"
12:27:34 Quit ashridah (Read error: 110 (Connection timed out))
12:28:22 Join ashridah [0] (
12:29:10 Join Aison [0] (
12:37:38 Join MoosCamaro [0] (
12:37:48MoosCamarohi all
12:43:17 Join asdsd____ [0] (
13:03:40 Join LinusN [0] (
13:06:30Bagderhi Linus
13:08:49*LinusN just had lunch with Björn
13:09:06amiconnhi LinusN
13:11:00MoosCamarohi Linus
13:15:40t0masBjörn == Brön Stenberg == Zagor?
13:15:52t0masor am I mixing some names now? :)
13:15:58Bagderthat's correct
13:16:12t0masand is he familiy of you Bagder?
13:16:16Bagderexcept that linus spoke iso8859-1 and you spoke utf8 ;-)
13:16:25Bagderhe's my brother
13:16:29t0masah ok
13:17:24*LinusN has to reboot
13:17:33LinusNcu soon
13:17:45 Part LinusN
13:21:03MoosCamarohi t0mas: the brothers Stenberg are cofondater with Linus... of rockbox
13:21:24t0masyes, I knew they were project admin and things like that... on sourceforge
13:21:32t0masbut I wasn't sure if they were familiy :)
13:21:50Bagder <= a crappy picture of us three
13:22:27t0masfrom left to right?
13:22:39Bagderme, Linus, Björn
13:22:42t0masmiddle one is Linus? (guessing)
13:23:01MoosCamaroleft Daniel
13:23:08MoosCamarooops Bagder
13:23:16MoosCamaroright Zagor
13:24:58t0masghehe... and Bagder wrote curl/c-ares :)
13:25:07t0masrecognized his name a few weeks ago somewhere :)
13:25:20Bagderkeeping myself busy
13:30:18 Join LinusN [0] (
13:35:30t0maswb LinusN
13:38:03MoosCamarohi LinusN,congratulations for your new car:), scuse me for this question, but have you any news about MP3 playback please?
13:39:26LinusNi've been on vacation, and also busy with other stuff
13:39:32LinusNso i have no news...
13:40:11Bagdergetting slightly longer than 15 seconds should not be too distant though
13:40:29LinusN16-17 maybe :-)
13:40:40Bagdershould be enough for everyone!
13:41:05Bagderwe can just ask everyone to chop up their songs in 16 second parts
13:41:16LinusNwe can develop a script
13:41:21MoosCamaroLinusN: a ok thanks and i wish you good vacation :)
13:41:49t0masnice idea Bagder :P
13:42:03t0mascut everything in 15 sec. parts... and put it in a playlist
13:42:14Bagderthen shuffle
13:42:15amiconn..and play it shuffled? ;)
13:42:21t0masthat's the best :P
13:42:28Bagderwe can call it Shuffle Extreme
13:42:37Bagderand add it as a feature
13:42:39t0masnew feature! :D
13:43:36t0masrasher should put it on the news ;) "New rockbox feature! Exteme Shuffel!" "ES is now availible on iriver, with the right tools you can even implement it on other mp3 players!"
13:47:51 Join rasher [0] (
13:47:59rasherspeaking of the news, it's still broken
13:48:02rasherI think
13:48:36rasheroh, no nevermind
13:48:46rasherjust the rss feed because of some illegal utf8
13:49:12Bagderthe script
13:49:21rasheror rather, something un-utf8
13:49:31rasheror.. something
13:49:56t0masthat MAGIC thing in the header...
13:49:57*LinusN is so happy, he gets lots and lots of emails :-)
13:50:00t0maswhat type should it have?
13:50:38t0masLinusN: interesting emails? or fanmail? ;)
13:50:50LinusNvirus bounces
13:51:01LinusNhundreds and hundreds
13:51:09t0masah, hate those...
13:51:21 Quit Bagder ("Off to search for that connect-resetting peer guy!")
13:51:21rasherYeah, I get a steady stream of about 50 of those per day
13:51:37t0masthat one LinusN/
13:51:49*t0mas has that one too :x
13:52:05LinusNyou're infected?
13:52:26t0masno, in my mail ;)
13:52:35t0maslinux can't be infected afaik
13:52:53ashridaht0mas: run it in wine
13:52:54amiconnt0mas: Usually the magic is a long, perhaps using only 3 bytes for the magic and one byte for the version
13:53:21t0masI did run some of those email things in vmware sometime...
13:53:33*rasher wanders off to send an old laptop he sold
13:53:33t0mascutted off the internet... sniffing what they did :)
13:54:47t0masyou have procmail?
13:54:53t0masor something alike?
13:55:03t0masthe messages always have an attachment named like this:,,,,,
13:55:14t0masand the size is always ~73 kb
13:55:21t0masso filtering isn't difficult :)
13:55:25LinusNyeah, i'm editing my procmail filter now
14:13:12 Join cYmen_ [0] (
14:13:35 Quit cYmen (Read error: 104 (Connection reset by peer))
14:21:49 Join Patr3ck_ [0] (
14:22:53 Quit thegeek (Read error: 54 (Connection reset by peer))
14:27:30***Saving seen data "./dancer.seen"
14:29:51t0masthat's how I understand your idea... please correct me if I see something wrong
14:33:20 Join hibou [0] (
14:39:08 Quit hibou (
14:39:08 Quit rasher (
14:39:08 Quit Aison (
14:39:08 Quit Harpy (
14:39:08 Quit TCK (
14:39:08 Quit DangerousDan (
14:39:08 Quit ferenczy (
14:39:08 Quit einhirn (
14:39:08 Quit LinusN (
14:39:08 Quit Hadaka (
14:39:08 Quit crash_ (
14:39:08 Quit asdsd____ (
14:39:08 Quit MoosCamaro (
14:39:08 Quit izzy (
14:39:08 Quit elinenbe (
14:39:08 Quit [zmaj] (
14:39:08 Quit Rick (
14:39:08 Quit mbr (
14:39:08 Quit HCl (
14:39:08 Quit Patr3ck_ (
14:39:08 Quit amiconn (
14:39:08 Quit Seed (
14:39:08 Quit t0mas (
14:39:08 Quit gromit` (
14:39:08 Quit Epoch`_ (
14:39:08 Quit Slasheri (
14:39:08 Quit CoCoLUS (
14:40:22 Quit Patr3ck (Read error: 110 (Connection timed out))
14:40:54 Join Bagder [0] (
14:40:54NJoinhibou [0] (
14:40:54NJoinPatr3ck_ [0] (
14:40:54NJoinrasher [0] (
14:40:54NJoinLinusN [0] (
14:40:54NJoinasdsd____ [0] (
14:40:54NJoinMoosCamaro [0] (
14:40:54NJoinAison [0] (
14:40:54NJoinamiconn [0] (
14:40:54NJoinSeed [0] (
14:40:54NJoinferenczy [0] (
14:40:54NJoinHarpy [0] (
14:40:54NJoinTCK [0] (
14:40:54NJoinDangerousDan [0] (
14:40:54NJoineinhirn [0] (
14:40:54NJoint0mas [0] (
14:40:54NJoingromit` [0] (~gromit`
14:40:54NJoinizzy [0] (laitinei@
14:40:54NJoinHadaka [0] (
14:40:54NJoincrash_ [0] (
14:40:54NJoinEpoch`_ [0] (
14:40:54NJoinSlasheri [0] (
14:40:54NJoinCoCoLUS [0] (
14:40:54NJoinelinenbe [0] (~elinenbe_@
14:40:54NJoin[zmaj] [0] (
14:40:54NJoinRick [0] (rick@Rick.user)
14:40:54NJoinmbr [0] (
14:40:54NJoinHCl [0] (
14:42:47t0maswb all :)
14:45:19 Join muesli- [0] (
14:46:03amiconnt0mas: Just two small remarks: (1) How do you know how many index entries exist? You can no longer deduce it from the file size with the new format
14:46:34amiconn(2) The global offset to the descriptions is probably unnecessary
14:47:06amiconnI'd suggest replacing the pointer to start of descriptions with a word count field
14:47:32amiconnThe indices to the descriptions (in the index part) would then be relative to the start of the whole file
14:48:56t0mashm... sounds good
14:50:42elinenbePorted cube.c from rockbox ( - a spinning 3D cube (now in CVS)
14:51:19Bagderthere's a video of that too somewhere
14:51:29t0masamiconn: changed :)
14:51:52rasheron the screenshots page
14:56:11amiconnelinenbe: I know... want to port that back to rockbox
14:56:19amiconn(the greyscale extension)
14:56:45muesli-btw: is there a greyscale patch?
14:56:51Bagdermuesli-: yes
14:56:59Bagdersee the patch tracker
14:57:09muesli-uhm, sorry?
14:57:16muesli-where do i find it
15:02:53muesli-there's file called gray.patch
15:03:04muesli-what do i do with it?
15:03:20Bagderread this
15:03:55t0mashm... when in usb mode... what can rockbox still do?
15:04:02t0maseverything except hd access?
15:04:07Bagdert0mas: everything
15:04:10rasherdepends on the device I guess
15:04:19t0masBagder: including hd access?
15:04:23Bagderno, excluding that
15:05:05 Nick Aison is now known as aison|away (
15:05:15Bagderthe iriver has the same usb chip as the newer archoses do
15:17:47 Join pfavr [0] (
15:24:31ashridahheh. you could load elevator music :)
15:25:09ashridahhave some crappy tune of the girl from ipanema or whatever it is driving people nuts
15:25:42 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
15:26:23 Quit muesli- (Read error: 145 (Connection timed out))
15:29:48 Quit lostlogic ("Going to the moon")
15:35:02 Join DMJC [0] (
15:42:58 Quit DMJC ("Leaving")
15:47:01 Part LinusN
15:52:06 Join DMJC [0] (
16:11:40 Quit DMJC ("Leaving")
16:12:24 Join thegeek [0] (
16:13:36 Join lolo-laptop [0] (~lostlogic@
16:17:15 Nick aison|away is now known as Aison (
16:23:24 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.7/20050420]")
16:23:56 Quit ashridah ("Leaving")
16:26:27 Quit Patr3ck_ ()
16:27:33***Saving seen data "./dancer.seen"
16:51:06t0masBagder? you wrote the old memory allocation system in rockbox?
16:51:56t0maswhy was it removed?
16:52:04t0masis there a mailinglist thread about it?
16:52:08Bagderbecase we don't need or want it
16:52:29Bagderwe've discussed adding one for pure plugin use though
16:53:26t0mashm.. using the pluginbuffer?
16:53:30 Join einhirn [0] (
16:53:48Bagderusing whatever buffer I guess
16:55:14BagderI'm off, see ya around
17:03:13 Join m8rix [0] (
17:05:30 Join Sucka [0] (
17:19:35 Nick m8rix is now known as m8rix|DND (
17:20:13 Join XShocK [0] (
17:21:39XShocKhi all
17:35:04 Quit XShocK ("CGI:IRC (EOF)")
17:39:53 Join _n2_ [0] (
17:41:29 Join muesli- [0] (
17:52:19 Part elinenbe
17:56:01 Quit noclue2 (Read error: 110 (Connection timed out))
18:00:35 Join DMJC [0] (
18:20:18 Join webguest64 [0] (
18:24:24 Quit webguest64 (Client Quit)
18:27:34***Saving seen data "./dancer.seen"
18:36:46*HCl blehs at stuff not working.
18:41:18 Quit muesli- (Read error: 110 (Connection timed out))
18:45:02 Quit bobTHC ("Smoke Weed Every Day !")
19:18:23 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
19:18:23 Quit MoosCamaro (Read error: 131 (Connection reset by peer))
19:18:40 Join MoosCamaro [0] (
19:37:30 Quit m8rix|DND ("There are only 10 different kinds of people: those who understand binary, and those who don't!")
20:01:31 Join DangerousDan [0] (
20:03:00 Quit DangerousDan (Read error: 54 (Connection reset by peer))
20:03:38 Join DangerousDan [0] (
20:05:29 Join cYmen [0] (
20:15:35 Join Strath [0] (
20:22:58 Quit cYmen_ ("zZz")
20:27:35***Saving seen data "./dancer.seen"
20:30:05 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
20:38:06 Quit StrathAFK (Read error: 110 (Connection timed out))
20:39:36 Nick Lynx_ is now known as Lynx_awy (HydraIRC@
20:58:31 Join _aLF [0] (
21:08:38 Join TCK- [0] (
21:14:51 Join xen` [0] (
21:17:10 Quit TCK (Read error: 60 (Operation timed out))
21:17:18 Quit asdsd____ (Read error: 145 (Connection timed out))
21:36:21 Quit lolo-laptop (Read error: 60 (Operation timed out))
21:45:22 Join lolo-laptop [0] (~lostlogic@
22:20:26 Nick _n2_ is now known as ehntoo (
22:26:13 Join austriancoder [0] (~austrianc@
22:26:18austriancoderhi all
22:27:38***Saving seen data "./dancer.seen"
22:28:13MoosCamarohi austriancoder
22:38:52 Quit TCK- (Read error: 110 (Connection timed out))
22:40:15MoosCamaroyou past a good vacations ;)
22:48:07 Join cYmen_ [0] (
22:49:18 Quit cYmen (Read error: 104 (Connection reset by peer))
23:00:49 Quit Zagor (Remote closed the connection)
23:04:17 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
23:05:48 Join DangerousDan [0] (
23:06:20 Join Patr3ck [0] (
23:14:09 Quit ehntoo ("Leaving")
23:16:31 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.12")
23:18:11 Join stripwax_ [0] (
23:18:23Bagderhi ho
23:18:55 Nick stripwax_ is now known as stripwax (
23:19:49stripwaxHow's things
23:20:01Patr3cki am wondering what the new dictionary plugin will do?
23:20:13 Join Harpy_ [0] (
23:20:26 Quit Harpy (Read error: 54 (Connection reset by peer))
23:21:33Patr3ckI have read this, but have still no clue what it will be used for ;-)
23:21:44CoCoLUSit's a dictionary?
23:21:58CoCoLUSlike in those portable translators
23:21:59Bagderor rather, for looking up words in a dictionary
23:22:01Patr3cklike english -> german
23:22:15Patr3ckah I see, a Duden (in german)
23:22:17Bagderif you want, yes
23:22:23CoCoLUSjo, duden ;)
23:22:29Bagderor "word => explanation of word"
23:22:57Patr3ckI had a different (wild) idea what it could be :-)
23:23:21Patr3ckwhen entering a search term to complete it
23:24:58stripwaxHas anyone built a bootloader for iRiver using the latest (13-day-old) cvs? if so, does it make any difference?
23:26:23Bagderdifference compared to what?
23:26:34stripwaxUm, the previous version?
23:26:55CoCoLUSpatr3ck, like t9 on mobiles?
23:27:14BagderI don't know of any versions in bootloaders
23:28:00Patr3ckI don't know t9, but it could be usefull if the dictionary would be artist names, song names, etc
23:28:02stripwaxI mean, does the commit made 13 days ago make any difference to compatibility / anything?
23:28:31Bagderah, that
23:28:42Bagderyes, it makes a difference if you use more than 16MB of ram
23:28:52Patr3ckbutton click, complete a partial entered word ;-)
23:28:56Bagderwhich the sound playback tests do
23:29:03Patr3ckso many ideas...
23:30:43stripwaxBagder .. ah- thanks. Makes sense. I'm guessing it doesn't affect/fix compatibility w 1.65, has anything more been found out about the ogg playback problem?
23:31:15Bagderstripwax: nothing's found, but Linus has an idea to test
23:32:19stripwaxBagder oh, what's the idea
23:32:46BagderI'm not sure, and I'd rather not repeat my skewed idea of it since it might be wrong
23:34:19 Join ashridah [0] (
23:36:10 Quit DMJC ("Leaving")
23:41:09 Join crashd [0] (
23:41:19crashdevening guys
23:51:09 Quit Harpy_ (Read error: 60 (Operation timed out))
23:58:43 Quit ashridah ("Leaving")
23:58:57 Join ashridah [0] (

Previous day | Next day