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 2005-05-04

00:03:39MoosCamarogoodevening/afternoo/morning :) here in Paris midnight
00:05:11muesli-bonjour madame ;)
00:05:43MoosCamaro:) bonsoir mademoiselle
00:05:53 Quit Sucka ("a bird in the bush is worth two in your house")
00:07:51 Join Camilo [0] (
00:08:43 Quit Camilo (Client Quit)
00:09:23crashdhey guys
00:09:25crashdhas anyone here got a h10 ?
00:09:52MoosCamaronope h140 here sorry
00:10:18crashdahh, was just wondering if anyone had done much work on the h10 firmware yet is all : ]
00:10:43muesli-dream on dude :D
00:10:50crashdthat's what i thought man :\
00:11:00crashdapart from the codec chip, i dont evne know what hardware it's running
00:11:44dwihnoI recall something about the portalplayer hardware (same as ipod)
00:11:55dwihnoperhaps that's the codec ;)
00:12:06crashdthe codec chip is provided by wolfson electronic afaik
00:12:11crashdi had the spec sheet for it somewhere
00:12:23Bagder" I can confirm that the H10 does use a Portalplayer chipset."
00:12:34Bagderfrom misticriver
00:12:55Bagdercodec chip?
00:13:09 Join DMJC [0] (
00:13:26crashdaccording to some posts on the mistic river forums yah
00:17:34Bagderif it truly is portalplayer, then they surely have their codecs in the cpu
00:17:49crashdaccording to various posts it uses portalplayer and the wolfson micro chipset
00:18:25Bagderwhat would the wolfson do?
00:19:02crashdlol, im just passing on the information ive seen :| i assumed the portalplayer provided the infrastructure for the actual menus and stuff
00:19:07crashdand the wolfson did all the hardware decoding
00:19:16crashdmaybe it's just the dac and stuff
00:20:27Bagderclearly stories on this topic are confusing
00:20:33crashdquite clearly :\
00:22:09muesli-g'night mates
00:23:15t0masPatr3ck: The smart input was some other thing I've been thinking off
00:23:47crashdis the chipset spec i found on misticriver
00:23:52crashdbut i dont remember the actual thread url
00:24:50 Join XShocK [0] (
00:24:56t0mashi XShocK
00:27:41***Saving seen data "./dancer.seen"
00:28:03 Quit ashridah ("Leaving")
00:28:05 Join ehntoo [0] (
00:39:01 Part stripwax
00:42:35 Quit muesli- (Read error: 113 (No route to host))
00:47:20 Quit Aison ("( :: NoNameScript 3.72 :: )")
01:01:58 Quit t0mas ("goodnight")
01:27:57MoosCamarogood night all
01:28:19 Part MoosCamaro
01:34:28 Quit XShocK (Read error: 104 (Connection reset by peer))
01:34:56 Join XShocK [0] (
01:41:48 Join _ferenczy [0] (
01:56:22 Quit _ferenczy ()
02:24:39 Quit cYmen_ (Read error: 113 (No route to host))
02:27:43***Saving seen data "./dancer.seen"
02:33:01 Quit DMJC ("Leaving")
02:36:58 Join Howman [0] (
02:42:50 Quit Patr3ck (Read error: 110 (Connection timed out))
02:44:20 Join Guest3449434 [0] (
02:45:31 Quit thegeek (Read error: 60 (Operation timed out))
02:48:35 Join Biptoria [0] (
02:55:07 Quit XShocK (Read error: 110 (Connection timed out))
02:55:07 Nick Howman is now known as XShocK (
02:55:41 Quit Biptoria ("CGI:IRC")
03:10:48 Quit xen` ()
03:15:01 Join _aLEX [0] (
03:15:04 Quit _aLF (Read error: 54 (Connection reset by peer))
03:30:31 Quit lolo-laptop ("Client exiting")
04:00:09 Join lostlogic [0] (
04:05:27 Join QT_ [0] (as@area51.users.madwifi)
04:10:46 Quit XShocK (" HydraIRC -> <- IRC has never been so good")
04:13:13 Quit Strath ("Client closed")
04:18:04 Quit QT (Read error: 110 (Connection timed out))
04:27:46***Saving seen data "./dancer.seen"
04:47:43 Join _aLF [0] (
04:47:43 Quit _aLEX (Read error: 104 (Connection reset by peer))
04:52:01 Join asdsd____ [0] (
04:52:04 Part asdsd____
05:28:34 Join Strath [0] (
05:49:36 Join kergoth [0] (
05:56:31 Part kergoth ("Leaving")
06:21:37 Quit DangerousDan (
06:21:37 Quit Stryke` (
06:21:37 Quit einhirn (
06:21:37 Quit crash_ (
06:21:37 Quit rasher (
06:21:37 Quit hibou (
06:21:37 Quit Bagder (
06:21:37 Quit Hadaka (
06:21:37 Quit ferenczy (
06:21:37 Quit Guest3449434 (
06:21:37 Quit Rick (
06:21:37 Quit [zmaj] (
06:21:37 Quit izzy (
06:21:37 Quit HCl (
06:21:37 Quit mbr (
06:21:37 Quit crashd (
06:21:37 Quit CoCoLUS (
06:21:37 Quit Slasheri (
06:21:37 Quit Epoch`_ (
06:21:37 Quit amiconn (
06:21:37 Quit Seed (
06:21:37 Quit gromit` (
06:22:43NJoinGuest3449434 [0] (
06:22:43NJoincrashd [0] (
06:22:43NJoinDangerousDan [0] (
06:22:43NJoinStryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
06:22:43NJoineinhirn [0] (
06:22:43NJoinBagder [0] (
06:22:43NJoinhibou [0] (
06:22:43NJoinrasher [0] (
06:22:43NJoinamiconn [0] (
06:22:43NJoinSeed [0] (
06:22:43NJoinferenczy [0] (
06:22:43NJoingromit` [0] (~gromit`
06:22:43NJoinizzy [0] (laitinei@
06:22:43NJoinHadaka [0] (
06:22:43NJoincrash_ [0] (
06:22:43NJoinEpoch`_ [0] (
06:22:43NJoinSlasheri [0] (
06:22:43NJoinCoCoLUS [0] (
06:22:43NJoin[zmaj] [0] (
06:22:43NJoinRick [0] (rick@Rick.user)
06:22:43NJoinmbr [0] (
06:22:43NJoinHCl [0] (
06:26:37 Quit ehntoo ("Leaving")
06:27:50***Saving seen data "./dancer.seen"
06:31:22 Join ehntoo [0] (
06:54:32 Join DMJC [0] (
07:22:37 Join LinusN [0] (
07:40:26 Quit DMJC ("Leaving")
07:48:56 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
08:27:53***Saving seen data "./dancer.seen"
08:45:41 Quit edx (Read error: 110 (Connection timed out))
08:52:39 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
08:52:58 Join einhirn [0] (
09:02:46amiconnProbably I should switch to using ugt
09:03:47 Quit dwihno (Read error: 113 (No route to host))
09:10:13 Nick QT_ is now known as QT (as@area51.users.madwifi)
09:13:49 Join bobTHC [0] (
09:14:29 Join pfavr [0] (
09:15:01bobTHCmornin' folks !
09:16:38 Join austriancoder [0] (
09:16:44austriancodergood morning
09:23:45 Join webguest59 [0] (
09:35:52 Quit hibou ("CGI:IRC (EOF)")
09:45:02 Join Zagor [0] (
09:47:08 Join Harpy [0] (
09:52:36LinusNwee, binutils-2.16 is finally released!
09:55:52LinusNworks fine for m68k-elf, now building sh-elf
09:56:12 Join ashridah [0] (
10:04:45 Join DMJC [0] (
10:15:46 Quit DMJC ("Leaving")
10:18:13 Join t0mas [0] (
10:21:40 Join crash__ [0] (
10:22:03 Join rasher_ [0] (
10:22:06 Join Bagder_ [0] (
10:22:36 Part austriancoder
10:22:41 Quit Harpy (
10:22:41 Quit webguest59 (
10:22:41 Quit einhirn (
10:22:41 Quit LinusN (
10:22:41 Quit crash_ (
10:22:41 Quit rasher (
10:22:41 Quit Bagder (
10:22:41 Quit DangerousDan (
10:22:41 Quit Hadaka (
10:22:41 Quit ferenczy (
10:23:02 Join edx [0] (
10:27:54***Saving seen data "./dancer.seen"
10:28:14 Nick Strath is now known as StrathAFK (
10:29:06 Join Harpy [0] (
10:38:21 Join cYmen [0] (
10:40:22 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.7/20050420]")
10:58:59 Join amiconn_ [0] (
11:06:46NJoinLinusN [0] (
11:07:31 Quit amiconn (Read error: 60 (Operation timed out))
11:07:31 Nick amiconn_ is now known as amiconn (
11:08:09LinusNhmmm, major problems reaching here...
11:11:22amiconnLinusN: Maybe this issue is just a matter of waiting long enough while rockbox is rebuilding fsinfo?
11:14:08LinusNi replied
12:00:23 Join preglow [0] (
12:05:45 Join Aison [0] (
12:27:55***Saving seen data "./dancer.seen"
12:42:21 Join muesli- [0] (
13:07:12 Join MoosCamaro [0] (
13:07:24MoosCamarohey all
13:10:41 Quit muesli- (Read error: 104 (Connection reset by peer))
14:05:47 Quit lostlogic ("Going to the moon")
14:07:29 Quit Aison ("( :: NoNameScript 3.72 :: )")
14:27:09 Join Patr3ck [0] (~patr3ck@
14:27:59***Saving seen data "./dancer.seen"
14:39:10 Join lolo-laptop [0] (~lostlogic@
14:56:14preglowLinusN: finished the alternate asm version of imdct_l the other day, and it's heaps and heaps slower than the idiot version i previously made...
14:56:57LinusNcan i see?
14:58:25preglowjust the part to imdct_l_windowing:
14:58:34preglowthe windowing is unrolled, so i can understand it if that's slower
14:59:55LinusNi think i have an idea why it is so slow
15:00:03preglowpray tell
15:00:46LinusNsince the imdct36_long_array is in iram, the cpu interleaves opcode fetches in dram and operand fetches in iram
15:01:00LinusNtherefore it can't burst
15:01:47LinusNit would definitely benefit from having the code in iram
15:02:03 Quit Patr3ck (Read error: 104 (Connection reset by peer))
15:05:42*LinusN hasn't seen such an amount of assembler in one file since his amiga demo programming days
15:07:07bobTHCsome always avaible on the net ? ;)
15:07:46LinusNThe Real Horizon - Virtual Intelligence
15:08:05LinusNthe first and only amiga demo i made
15:08:19LinusNi == we
15:10:25preglowLinusN: well, yeah, having data in ram as well sure didn't help
15:10:42LinusNdata in dram?
15:11:07LinusNthat won't help a bit, it will only make it worse
15:11:26LinusNbursts can only do consecutive addresses
15:11:45preglowindeed, but if i'm to put code in iram as well, i'll need 64k for sure
15:11:52LinusNso i should have rephrased my first statement
15:12:30LinusNboth constants and data need to be in iram unless you find a way to burst read the constants
15:12:39LinusNs/and data/and code/
15:13:02preglowi do burst read the constant
15:13:07preglowin the .loop thing
15:15:32LinusNwhere is the data btw?
15:16:08preglowwhat data?
15:17:12LinusNthe samples
15:17:19preglowi think they're in iram
15:17:19preglowi hope so
15:17:58LinusNthey better be... :-)
15:18:30preglowyes, i hope so
15:18:39preglowor there's a lot of expensive accessesz
15:18:54LinusNi'm looking at the first loop
15:19:10LinusN mac.l %d0, %a2, (X2, %a0), %a2, %acc0
15:19:27LinusNthis will hurt badly
15:19:43preglowyeah, unless it's in iram, right?
15:19:56preglowi wish i could get sequantial access there
15:20:10LinusNyou could break out the loops and put them in iram
15:20:22preglowthere's always that
15:20:30preglowi should try it
15:20:41preglowbefore that there's very few data segment accesses
15:21:31preglowi'm actually contemplating just using those constants verbatim in the code
15:21:37preglowsince there's very few of them before the loop
15:23:59 Nick Bagder_ is now known as Bagder (
15:24:53*Zagor just got an offer for 24/8Mbit broadband
15:25:16Zagorthat's pretty fast
15:25:56bobTHCon FTTH ?
15:26:18Zagorno, cable
15:26:39Zagorthe fiber people are getting 100/100 already. they don't count. :-)
15:28:00bobTHCi've forgot that's swedish use the technology at is full potential directly, not like here where we are always bridle
15:29:04Zagorwe used to have that too, but for some reason the competition works pretty well in this field in sweden
15:29:54bobTHChere we have 1 city where FTTH is operational and it's just a 10/10 offer
15:30:02*ashridah considers moving
15:30:20Zagori guess it's because it was a company that wasn't Big Telco that started selling full speed access, so the others had to catch up. (but boy, did they try to resist!)
15:30:43Zagor"there is no need for full speed adsl, so we don't sell it"
15:31:04Bagder512k should be enough for everyone!
15:31:10ashridah"but we're trying to GIVE YOU MONEY FOR IT!"
15:31:47ashridah"we surveyed 1000 old grandmothers who can't see let alone use a computer, and none of them thought it was worth it"
15:32:02 Join dwihno [0] (~dw@
15:33:40bobTHCon cable here the best offer is 20M/512k it's a big scam
15:34:14bobTHCtechnicaly impossible to reach 20M with just 512 on up
15:35:47bobTHCisp are afraid with upload, the p2p syndrom
15:43:33 Quit dwihno (Read error: 60 (Operation timed out))
15:49:01 Join dwihno [0] (~dw@
15:50:55 Quit ashridah ("Leaving")
15:55:30 Part LinusN
16:07:16 Quit Lynx_awy (" bye")
16:18:59 Join Naked [0] (
16:19:10 Nick Naked is now known as Hadaka (
16:28:02***Saving seen data "./dancer.seen"
16:34:16 Join TCK [0] (
16:36:39 Join bobdbob [0] (
17:13:20 Quit bobdbob (Read error: 54 (Connection reset by peer))
17:15:39 Join bobdbob [0] (
17:19:45 Quit Rob- ("2.8.2-EPIC3.004 -- just do it.")
17:24:25t0masbobTHC: move to the netherlands
17:24:43t0mascable is 10/10 here
17:24:53t0masand adsl is up tp 2/8
17:24:57t0maserr. 8/2
17:25:04t0masah whatever... 8 down 2 up
17:25:33t0masand adsl2 is introduced in the bigger cities now...
17:25:34bobTHCif i move to netherland it's for weed and not for bandwith ;)
17:25:49t0masall french people say that..
17:25:56t0masand almost nobody smokes weed here...
17:26:05t0masjust old ladys and artists
17:26:12t0masand some junks...
17:26:13bobTHCi know, a lot of travel in ure country
17:26:16 Quit bobdbob ("Trillian (")
17:26:18amiconnt0mas: I have another suggestion for the dictionary plugin (after implementing the single-file format). Make it a viewer, and it will be very easy to use different dictionaries...
17:26:40t0masamiconn: someone else suggested making a selection before use...
17:26:46t0masbit maybe viewer would be better yes..
17:27:14 Join TCK- [0] (
17:27:16amiconnIt would be as simple as "playing" the dictionary file
17:27:30amiconnSuggested extension: .dict
17:27:39*t0mas emails Slasheri
17:28:53amiconnDoes the perl script sort the entries alphabetically, so that rdf2binary doesn't have to do that?
17:29:28amiconnIf so, the per-dictionary max_word handling should be simple
17:30:23t0masyes, the perl script should sort
17:30:36amiconn...because the perl script has to slurp in the whole source file before producing the .rdf. While doing that, it could check the maximum word length
17:30:45*t0mas prods rasher for the dict2rdf script :)
17:31:02t0masyes, that's simple
17:31:02rasher_I don't think that's happening
17:31:06amiconn...and perhaps put it as the very first line of the .rdf
17:31:35t0masamiconn: not a problem, but the C tool will have to read everything in for the new fileformat...
17:31:40t0masor use temp files, and combine later
17:32:42amiconnIf the first line of the .rdf tells it the maximum word length, it can work almost the same way as now
17:33:12amiconnAh, you mean because of combined index/desc?
17:33:19t0masI was just typing that
17:33:29t0masso.. yes :)
17:33:51amiconnI'd say just write 2 files, and combine them as the final step
17:34:29amiconn...and imho rdf2binary could be a perl script as well
17:34:29t0masyes, that would be the easy way...
17:34:48t0masnot if I have to do anything with it
17:36:38amiconnI'm not a perl expert, but I think I could write such a script. There is already a number of file conversion scripts in the rockbox tree
17:37:02t0masok, if you want to do it... go ahead
17:37:14t0masSlasheri did the file conversion thing for now... I just edited some things
17:44:03 Quit TCK (Read error: 110 (Connection timed out))
17:46:39 Join muesli- [0] (
17:51:07muesli-hi t0mas
17:51:14muesli-how's gauda goin? ;)
17:51:23t0masGouda ;)
17:51:54t0mas(I actually live right next to it...)
17:52:15muesli-living in a cheese *Wow* :D
17:52:34t0masI don't like cheese :P
17:52:41muesli-neither i do ;)
17:53:03muesli-we have cheese which smells like stinky socks
17:53:08t0masI know
17:53:08muesli-and i am not kidding
17:53:37muesli-dunno who is so pervert to eat..
17:53:44bobTHCmuesli ? which ?
17:54:09t0masschimmelkaas in dutch
17:54:19muesli-yeah kind of
17:54:22t0mascamembert (fromage bleu)
17:54:31muesli-its called harzer cheese
17:54:39muesli-harzer roll i.e.
17:54:48t0masfromage persillé :P
17:54:51bobTHCcamembert is not a a blue chease
17:55:00t0masor is my french to bad for this? ;)
17:55:04muesli-camembert ruelz
17:55:17bobTHChehe, i love it !!!
17:55:22*t0mas really hates cheese... in whatever form you get it :P
17:55:42t0masyou can't hide it in my food :P
17:55:53t0massome restaurant tried a few weeks ago..
17:56:00t0mas"No sir, there's no cheese in that."
17:57:36bobTHCroquefort, camembert, bleu de bresse, picodon .... it's so good
17:58:38bobTHChehe, stinky and strong taste but sooooooo good ;)
17:59:08muesli-like your pants? ^^
17:59:52muesli-try smoking your pants..will kick your ass more than thc ^^
18:00:46bobTHCif it's hemp pants perhaps, but i doubt...
18:01:15 Quit TCK- (Read error: 110 (Connection timed out))
18:04:46MoosCamarohi guys
18:04:53MoosCamarohi guys
18:05:28MoosCamaroit's a fromage discusion or thc discusion :)
18:05:33bobTHCeach country have this typical dishes, and they're seldoms likes by foreigners ;)
18:05:45MoosCamarohi bobTHC
18:05:57bobTHChi cheese mate ;)
18:05:59*muesli- is lost in cheeze....
18:06:02MoosCamaroyou're represant french ;)
18:06:14bobTHCalways and foreva ;)
18:06:39 Quit muesli- (Read error: 104 (Connection reset by peer))
18:07:50MoosCamarobobTHC: me, i prefer H to weed ;)
18:08:00MoosCamaroprobably cause I'm maroccan
18:08:28bobTHCfor sure
18:09:45MoosCamaroMorocco better than Netherlands for this ;)
18:10:24MoosCamarooops "best"
18:11:00bobTHCit's because ure maroccan imho , because for me weed is far better than H .
18:11:50MoosCamaro:) true
18:13:06 Join Domonoky [0] (
18:13:17MoosCamaroabsolument d'accord l'aspect chimique t est certainement pour beaucoup
18:13:28MoosCamarochimicals ;)
18:16:02 Join TCK [0] (
18:16:12Domonokyha.. iriver gets problems with :-)
18:17:03Domonokythey should deliver the source for the pmp-1xx player... because they use linux :-)
18:18:04Domonokyif this happens.. a chance for rockbox on the pmp-players :-))
18:18:46rasher_not necessarily so
18:18:53rasher_they could be using binary drivers
18:19:05rasher_which would make us none the wiser
18:20:18amiconnrasher_: Iiuc this would still be a gpl violation
18:20:27rasher_No more than the nvidia drivers
18:20:49 Quit preglow ("h")
18:21:39amiconnrasher_: Yes... the nvidia driver are violating the gpl too, iiuc
18:22:15rasher_Ah.. I'm not sure about that situation
18:22:41HClas far as i know they aren't.
18:22:47HClthey just have a binary driver for linux.
18:22:53amiconnI know
18:23:02HClthe loading bit is opensource.
18:23:11HClthey're not breaking gpl
18:23:46rasher_I think it's a grey area, but not sure
18:24:44amiconnAfaik, the gpl says everything that interacts with gpl code more closely than simply getting run by it (e.g. an application on Linux) needs to be gpl compatible
18:24:58amiconnA driver definitely isn't just a simple application
18:25:11HClamiconn: nafaik.
18:25:25HClanything BASED on gpl code must be gpl.
18:25:42rasher_No, what amiconn says is correct
18:25:53rasher_if you load a gpl library you need to be gpl
18:26:00HCli'd want to ask #fsf on that.
18:26:08HClbut we're not loading a gpl library.
18:26:12HClwe're doing the opposite.
18:26:41*HCl isn't interested in licensing too much as long as people don't steal stuff.
18:27:09rasher_well "stealling" involves not following the license
18:27:17Domonokylets hope irver havent done such things as binary driver loading... as they hide their code..
18:28:06***Saving seen data "./dancer.seen"
18:28:32 Quit StrathAFK ("Client closed")
18:29:04Domonokyand i dont think they change their software much.. if the have to release it..
18:29:07 Quit bobTHC ("Smoke Weed Every Day !")
18:31:19*HCl will go ask the fsf about it
18:36:19 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
18:39:07HCl18:32 < Goran__> nvidia driver is not "based" on the linux kernel. It is the
18:39:07HCl kernel that loads the module - the module becomes part of it.
18:39:07HCl18:32 < HCl> so its not against gpl, right
18:39:07HCl18:32 < HCl> ?
18:39:14HCl18:33 < Goran__> well, if it is, then you are breaking the GPL by loading it.
18:39:14HCl (since the module does not contain the kernel, they are not
18:39:14HCl responsible)
18:42:05HCl18:36 < Goran__> in this case all LKMs have to be violating it.
18:42:05HCl18:37 < araujo> Why all?
18:42:05HCl18:37 < Goran__> sorry, i mean all non-gpl
18:42:05HCl18:37 < araujo> Correct.
19:02:45 Quit Epoch`_ ("Leaving")
19:05:27 Nick Guest3449434 is now known as thegeek (
19:14:50*HCl guesses he should start on a searchengine editor plugin..
19:18:45 Join muesli- [0] (
19:34:51 Quit lolo-laptop ("Client exiting")
19:37:28 Quit muesli- (Read error: 104 (Connection reset by peer))
20:12:50 Join lostlogic [0] (
20:12:51 Quit crash__ (Read error: 104 (Connection reset by peer))
20:22:56 Join crash_ [0] (
20:28:08***Saving seen data "./dancer.seen"
20:41:42amiconnZagor: r u there?
20:47:30 Join austriancoder [0] (~austrianc@
20:47:39austriancoderhi all
20:54:31austriancoderis it ok that i reproduce the iriver orginal fm-to-mp3-switching button?
20:58:35 Join Sucka [0] (
20:59:57 Join DangerousDan [0] (
21:02:31*austriancoder plays with his fm at iriver
21:04:23 Quit Sucka ("a bird in the bush is worth two in your house")
21:05:26HClmmm, nice, got radio to work?
21:05:32rasher_austriancoder: how's fm doing?
21:05:34*HCl goes to shower, will be back.
21:07:14austriancoderrahser_: fm is in progess.. think i will have it fully in the next days
21:07:25austriancoderbut i dont know, how i can test it realy
21:10:23*amiconn slaps the bad bitfield design in settings.c
21:11:41amiconnI think I'll go change that, making the handling functions quite a bit simpler & smaller
21:22:26HClamiconn: you're bound to know this, putsxy is to put a string on the screen at an x,y, right? how can we know the length of a string on screen, and are there any wrapping printing functions yet?
21:23:42amiconnlcd_getstringsize() tells you the width and height of a string
21:24:22amiconnThere are no wrapping print functions, 'cause they're unnecessary in the core
21:24:48HCldo fonts have a fixed size per character
21:24:56HClor are they not?
21:24:58amiconnIf you want to output lines of text, imho it would be better to use lcd_puts()
21:25:05HCl(like i occupies less space than m)
21:25:31amiconnThis uses column, line instead of pixel coordinates
21:25:53HClwell, i'm gonna have to be able to underline a certain word as well...
21:25:54amiconnThere are both proportional and fixed width fonts
21:26:03HClick, and i can't choose, can i.
21:26:33HClis there any way to underline a certain word?
21:26:39amiconnYou can enforce using sysfont in your code, this one is fixed width
21:27:16amiconnNo underline, no emboldening etc
21:27:17HClokay.. how? :X or should i search for an example?
21:27:34HClwell, fixed with would allow me to calculate where to draw the line with drawline
21:28:04amiconnExample: apps/recorder/keyboard.c
21:28:35amiconnThis one uses the sysfont, and also has to know letter positions to draw the inverted box, the cursor bar etc
21:28:38*HCl prods his motivation to finish the frontend.
21:30:30amiconnHow do you intend to use underline?
21:30:51amiconn(semantically, not technically)
21:31:10HClto point out what token you're currently modifying.
21:32:52amiconnUnderline uses extra space... You could also invert the current token
21:33:06HClhow would i do that?
21:33:23t0masHCl: I have the same problem as you have now
21:33:26amiconnFirst, draw the text. The use invertrect()
21:33:32t0masmaybe we should work on some general output functions?
21:33:49t0masto just print text to the screen... and auto wrap it?
21:33:58amiconnHCl: keyboard.c does exactly that (for the "pick cursor").
21:34:09HCli'll look at it, thank.
21:34:13austriancoderso.. in Debugmenu under FM Radio, i get HW detected: yes Result: 000000000
21:34:18austriancoderon iriver
21:34:23austriancoderi think, this looks good
21:34:40amiconnt0mas: Nice idea. For general plugin use this should become part of the plugin library
21:35:01t0masok, I have some stuff now... but it actually just sucks :P
21:35:13t0masso maybe it would be best to make a separate function?
21:35:23t0masand make that auto-scroll and things like that?
21:35:36t0masso the plugin can just do something like printf() does on pc's?
21:35:50HCli'd like to get some more info than just printf though..
21:35:59HCllike i need to know where it has wrapped and stuff... :/
21:36:10HClcould it somehow return the x,y it printed to?
21:36:20t0mascheck my dict.c
21:36:23t0maslast part
21:36:31t0masthere's a very simple wordwrap thing
21:37:21t0masjust cutting a line on the last space... and continuing on the next line
21:45:30 Join markun [0] (
21:45:44HClmorning markun
21:45:56markunHi everyone (no, I have not finished grayscale.. :)
21:46:04markunHCl: hi!
21:47:45 Join TCK- [0] (
21:48:24 Join Strath [0] (
21:48:54austriancoderhi markun
21:49:07markunamiconn: I'm playing a bit with freetype to make a ttf2fnt converter.
21:49:12markunaustriancoder: hi!
21:49:53amiconnI expect most TTFs to look rather bad when converted to such small bitmap fonts
21:49:59markunthe problem is that most fonts have some very large glyphs in them. A possible solution could be to have variable line heights, but I don't know if that would look very good.
21:50:32markunamiconn: The look quite ok if they have proper hintin (like the MS fonts)
21:51:16markunthe -> they, hintin -> hinting
21:52:12markunHCl: did you think the output of the ttf program looked bad?
21:52:28HClnot really, why o.o
21:53:46markunamiconn: do you want to have a look at the source, or are you too busy?
21:57:05austriancoderi2c reading and fm should work, but i dont know who to test it - has anybody an idea?
22:04:54 Quit TCK (Read error: 110 (Connection timed out))
22:06:25*austriancoder thinks that he found a test methode for i2c_read
22:07:50ehntoodo tell
22:08:37 Quit DangerousDan (Read error: 110 (Connection timed out))
22:17:40 Join [IDC]Dragon [0] (
22:17:59amiconnGood evening to Hanover :)
22:18:12austriancoderdebug build for iriver is brocken
22:18:12[IDC]DragonHannover even
22:18:31[IDC]Dragonsound like Hangover
22:18:32amiconnIn english Hanover is correct spelling iirc
22:19:25[IDC]DragonToday I've tested my nasty bootbox
22:19:27 Join DangerousDan [0] (
22:19:40[IDC]Dragonit seems to do all exept rolo
22:20:10amiconnI fiddled a bit with your bitfield functions in settings.c (and mine in ata_mmc.c). Saving a bit on code size, but requiring a change in the order of fields that cross longword boundaries
22:20:28amiconnI'll bump the settings block version
22:20:41austriancoderguys.. i hope tomorrow i can ofer you a nice patch for i2c read and fm on irver :) Today i visit a nice girl ;)
22:20:45[IDC]Dragonis it woth it?
22:21:06amiconnI think so
22:21:26 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.12")
22:21:51amiconn112 bytes saved in settings.c, and another 48 in ata_mmc.c
22:22:34amiconnWell, not too much... but the functions are more straightforward now (at least they look like that to me)
22:23:06amiconnDid you check my modified mmc_usb patch?
22:23:15amiconn...and the comments on that?
22:23:38[IDC]DragonI hope for bootbox resolving the need for such stinginess
22:23:55[IDC]Dragonyes, saw that, but didn't update my code base yet
22:24:10amiconnAH, you mean the code size...
22:24:22[IDC]Dragonthe detection you mentioned would be nice indeed
22:24:45amiconnHmm, efficient coding is still A Good Thing (tm), imho
22:25:13[IDC]DragonI don't know if the bits were in correct order my way
22:25:15amiconnYour original handling of fields crossing a longword boundary is... strange
22:25:28*[IDC]Dragon loogs
22:25:44amiconnIt puts the bit from the second longword as lsbs...
22:26:53amiconnMy new code works much like using a (virtual) long long
22:27:22amiconnPerhaps using a real long long would be even better (in case gcc is better in handling that)
22:27:39[IDC]Dragonok, bring it in
22:27:58[IDC]Dragonnot the long long, but your code
22:28:12***Saving seen data "./dancer.seen"
22:28:35[IDC]DragonI'm not gonna dive in such tricky code tonight
22:31:15[IDC]Dragonbtw, the USB before mount patch is a crucial part of bootbox now
22:31:41amiconnNot necessarily
22:31:59[IDC]Dragonit has no event loop
22:32:06amiconnRockbox tries to mount first, but doesn't hang if it doesn't succeed
22:32:28[IDC]Dragoncan you guarantee that?
22:32:43[IDC]DragonI feel much better with USB first
22:33:02amiconnIt's certainly the clean solution
22:33:32*amiconn just had another idea for the bitfield handling
22:33:51[IDC]Dragondesirable for a fallback firmware, when the user has god-knows-how corrupted the disk
22:34:46oddwow that'd be awesome
22:34:59oddgod you guys rock so much
22:35:03oddi wanna hug you all
22:35:14*odd is the designated rockbox cheerleader
22:35:38ehntoothey rock because it's rockbox.
22:35:40[IDC]Dragononly if you are you female
22:35:46oddi'm not :(
22:35:57[IDC]Dragon(only one "you" intended)
22:36:09ehntoohow about an official title "Extreme Supporter" instead?
22:36:15oddand no, they rock because they're making a freakin' awesome piece of software that i and thousands of others will love so much!!
22:36:32*HCl should finish his search engine.
22:36:39oddsee, i'm inspiring
22:37:29ehntooI've gotta learn some C... I'd like to contribute to the project, but my current areas of programming don't help much
22:40:36 Quit Domonoky (Read error: 110 (Connection timed out))
22:42:26amiconn[IDC]Dragon: Btw, bootbox won't resolve the need to code efficiently. There's still that 200 KB limit for an .ajz loaded by archos, and at least the fm/v2 are quite close again...
22:42:45[IDC]Dragonoops, they are?
22:43:11[IDC]Dragonyou can use bootbox as a 2nd level loader
22:43:13amiconnAlmost 190 KB, only ~10 KB left...
22:43:42amiconnI don't want rockbox to become like windows... fat & bulky
22:51:41Bagderyeps, 190K is dangerously close to windows
22:52:05Bagderanyone ever did anything with AMR?
22:53:17*t0mas swallowes a remark on making big money with something eating ram...
22:53:19Bagderspeach codec for GSM/3G
22:53:22[IDC]Dragonthe voice codec?
22:53:43[IDC]Dragonnot personally, but in my group
22:53:55BagderI'm probably gonna get involved with it in a project very soon
22:54:07[IDC]Dragonnb or wb?
22:54:36Bagderwhatever is required ;-)
22:55:33Bagderit does use quite a lot of CPU, doesn't it?
22:55:44[IDC]Dragonwb, yes
22:58:49BagderI guess amr-wb can be in a mms
22:58:58Bagderso wb it is
23:01:57[IDC]Dragonwb is hifi, nb is voice
23:02:14[IDC]Dragonwb is expensive, like aac+
23:02:28HClwhats gsm/3g?
23:02:49[IDC]Dragoncan tell you more when at work
23:03:08[IDC]DragonHCl: 3rd generation GSM
23:03:53Bagderwe'll probably just have a risc ~100 MHz, so the amr decoder might go into a special-purpose FGPA
23:04:18[IDC]DragonBagder: I heared about an option to make the linker removing dead code on function level instead of module level
23:04:28[IDC]Dragondo you know of such?
23:05:00[IDC]Dragonwhat kind of risc? 100 MHz souds feasible
23:05:05Bagderbut ld has a zillion options
23:05:36[IDC]Dragonwhich are best documented where?
23:05:45amiconn[IDC]Dragon: I tried that once, think I have my modifications lying around somewhere
23:05:59[IDC]Dragonsounds great
23:06:01Bagder[IDC]Dragon: 'man ld' on a linux box ;-)
23:06:29[IDC]Dragonworks on cygwin, too
23:06:31amiconnBagder: 'man ld' also works in cygwin
23:06:36Bagderah, nice
23:07:26Bagder[IDC]Dragon: it isn't yet decided, but it might be a nios (Altera)
23:08:01[IDC]Dragonyou may want to check Emuzed for codecs, in general
23:08:31Bagderwill do, thanks
23:08:59[IDC]Dragontheir stuff is amoung the best
23:09:22[IDC]Dragondunno if they will recompile it for exotic cpu's, though
23:09:41amiconn[IDC]Dragon: I tried the function-level garbage collection on plugins back then.
23:10:01[IDC]Dragonwould be nice for bootbox
23:10:09amiconnMy files are really oldish stuff (old build system, dating 06/2004)
23:11:20[IDC]Dragoni just need the option
23:11:24amiconnThe basic idea is to compile with -ffunctions-sections -fdata-sections , then link with −−gc-sections
23:11:39amiconn-ffunction-sections even
23:11:57amiconnI remember that it required some trickery to make the linking work
23:12:37amiconnYes, there are .comment sections in the object files
23:13:03amiconnYou need to tell the linker to keep 'em, because otherwise they would end up empty
23:13:18amiconn..and objcopy doesn't cope with that (iirc)
23:13:56*[IDC]Dragon is a build system noob
23:14:19amiconn..and of course you need to explicitly keep a 'root' section, otherwise the garbage collector would clean out everything
23:14:43[IDC]Dragonit's not detecting main() ?
23:14:45amiconn(for a plugin this would be .entry)
23:15:19amiconnPerhaps it detects main(), but I didn't try that. Plugins don't have a literal main()">
23:17:11SlasheriI have used altera apex/cyclone fpga's with nios. Those are great powerful chips
23:18:17Bagderneat little buggers ;-)
23:18:52Bagderyou used uclinux on it?
23:19:10Slasherinope, own realtime os
23:19:22Slasheribut uclinux might be worth to try out
23:19:30Bagderthat's the route I'm heading
23:19:36Slasherigreat :)
23:20:21SlasheriI have just designed a "universal bluetooth controller" that has a cyclone core. Maybe I could try uClinux on that
23:21:31 Join muesli- [0] (
23:23:31 Quit muesli- (Read error: 104 (Connection reset by peer))
23:23:36 Join muesli- [0] (
23:23:46 Quit muesli- (Read error: 131 (Connection reset by peer))
23:23:52 Join muesli- [0] (
23:24:00[IDC]Dragonamiconn: my 1st attempt didn't make it smaller
23:24:20 Quit muesli- (Read error: 131 (Connection reset by peer))
23:24:21amiconnErgh, not quite thinking
23:24:29 Join muesli- [0] (
23:24:31[IDC]Dragonno substantial garbage collection
23:24:41muesli-sorry mates
23:24:58muesli-mirc is crashing the whole time
23:25:55[IDC]Dragonamiconn: in your makefile, whats "-Wl," used for?
23:26:10[IDC]Dragondoes it need a comma seperator?
23:26:33amiconnHmm... old stuff. Did you check the file sdates?
23:26:52amiconnIirc this is for passing options to the linker
23:27:19Bagder-Wl,option Pass option as an option to the linker. If option contains commas, it is split into multiple options at the commas.
23:27:56amiconnAh, yes. Of course this is for passing the −−gc-sections to the linker
23:28:34amiconn...and there is a -Wl already, for creating the .map files
23:30:31amiconn[IDC]Dragon: Hanover is indeed correct spelling in English, I checked my dictionary
23:31:07[IDC]Dragonok, ok
23:31:14 Join soxx [0] (
23:32:18[IDC]Dragonhmm, still no effect
23:32:36amiconnCheck the .map
23:33:11[IDC]Dragonfor what?
23:33:45amiconnWhether this is working at all
23:33:59[IDC]Dragonseems not
23:34:00amiconnIf it is, you should have individual sections per function
23:34:19amiconn..named like .text.myfunction (iirc)
23:34:40 Part soxx
23:34:55amiconnIf not, the compiler option isn't taking effect, so there is no garbage to collect for the linker
23:35:49[IDC]DragonI don't see such "decoration"
23:36:54[IDC]DragonI've put the compile options into debug
23:38:05[IDC]DragonBagder: I think the debug option is broken, in the makefile
23:38:39[IDC]Dragonbecause, first we add stuff to CFLAGS when debug build
23:39:05[IDC]Dragonlater in general, CFLAGS is set, not added
23:39:22[IDC]Dragonthus overwriting the debug option
23:39:42Bagder... right
23:39:54Bagderno one uses debug these days I guess ;-)
23:40:01[IDC]Dragonseems so
23:40:32amiconnTalking about the debug menu... imho we could throw out some stuff from it
23:40:57amiconn...not everything, of course. Some items are still useful
23:41:10[IDC]Dragonwhen did you last check your MAS registers?
23:41:27amiconnI did... when adding MAS3539F support
23:41:37[IDC]DragonI know, that aside
23:42:19rasher_define some out?
23:42:23rasher_or comment
23:42:57 Join Domonoky [0] (
23:43:17 Quit Harpy (Read error: 110 (Connection timed out))
23:44:55amiconnI think 'View/clr RTC RAM', the MAS stuff (total of 3 items when considering both player & recorder) and perhaps also 'View audio thread' could go
23:45:24[IDC]DragonI'm all for it
23:45:34[IDC]Dragonperhaps view runtime, too
23:45:52amiconnI'm using that one quite regularly
23:46:51[IDC]Dragonand we can remove the (keep out!) part then
23:47:07[IDC]Dragonif debug is read only stuff, no danger
23:47:37amiconnThe screendump feature might be irritating
23:47:51rasher_cpu frequency on iriver isn't read-only
23:48:16amiconnBtw, my sister would like that option for which there is a patch already: powering from usb without going into usb mode
23:48:47amiconnImho this would be useful at least for FM, V2, and Ondio
23:51:24amiconnHowever, I would like this to be activated by a holding a button instead of a settable menu option, for obvious reasons
23:51:43amiconnThen there is the problem with the shortness of buttons on Ondio...
23:54:42 Quit Zagor (Remote closed the connection)
23:55:46[IDC]Dragonyes, button query on boot would be useful for various things
23:56:17[IDC]Dragonincluding a settings reset (or rather, no settings restore)
23:56:41amiconnI don't mean on boot, but on plugging usb
23:56:55[IDC]Dragonmy .bin file just became 150 MB
23:57:08amiconnOn boot would be problematic, because it would collide with usb-before-mount
23:57:25Bagder[IDC]Dragon: that's usually because you get a section from address zero or similar

Previous day | Next day