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 2002-09-05

00:00:42alkorranyway, i'm in a true sleep mode and i didn't notice any skip or stop with 3s spindown
00:01:36Zagorsounds good
00:04:21alkorrand no disturbance in scrolling text too :)
00:06:38alkorrnow i try to test if cache is still valid even in sleep mode
00:07:59alkorrwell not very easy to test :/
00:08:49alkorrmy battery is still 100% :)
00:09:35alkorrdo you know how many hours to have 50% ?
00:10:07alkorrwell, Miran is out ?
00:10:13Zagor4 maybe
00:10:21alkorrskating ?
00:10:31Zagorwalking, i think
00:13:13alkorryou know what it could be interresting to have a menu to set the minimum percentage of power where we stop access on hd (playing back and so on) and diplays the time spent
00:14:24alkorrthat way, when fully charging the jukebox, we can play back, and have the time spent when power would reach a certain percentage. I dont know if it is a fair precision.
00:15:14 Quit Miran-Exercise (Read error: 104 (Connection reset by peer))
00:15:15Zagori think batteries in general are very inexact
00:15:17alkorrdo you sleep instruction in yield() ?
00:15:44alkorrwell Miran seems to be out for a long walk...
00:15:58alkorri mean do you USE sleep instruction...
00:16:07 Join Miran [0] (
00:16:54Miran3s, skips, 5s, warps to half way through next track then skips, 10s skips, the unit generally skipped faster then the 20.20 build :(
00:17:25alkorrrecorder seems to have problems :(
00:17:43MiranThe archos 1.27d doesn't skip
00:18:01Miranbut it sounds crap compared to rockbox :)
00:18:04alkorri was spoken about hardware
00:18:23alkorri wasn't speaking i mean
00:18:30Miranoh, my appologies
00:18:51alkorrwhat differs from player ?
00:19:38alkorrbtw it really sleeps down
00:19:51alkorri suppose so if it skipd
00:20:35 Quit sylvr (Remote closed the connection)
00:21:46alkorrzagor, ata_thread can sleep in middle of a ata_read_sectors, which is really bad (yield())
00:22:14alkorrso the ata_thread should avoid to sleep when mutex is taken no ?
00:23:19alkorri wonder if variable sleeping is volatile that would change something
00:23:19Zagorthat is already done, since perform_sleep takes the mutex. but it's not enough.
00:24:02Zagorit can still call perform_sleep between two read_sectors() call. so it should wait 500ms, using the timer fix I write
00:24:14Zagorcase Q_SLEEP: should probably be "last_disk_activity = current_tick - sleep_timeout + (HZ/2);" instead.
00:25:19alkorri was wondering if sleeping = 0 could be delayed in the code...
00:26:15***Saving seen data "./dancer.seen"
00:28:04 Join merwin [0] (
00:29:11merwinZagor: did the battery charge indicator :P
00:29:57alkorrmutex can yield even if aquired ?
00:32:57Zagormerwin: nice!
00:35:28alkorryou mean two ata_read_sectors ? but even so we run perform_soft_reset before reading...
00:35:57alkorra timer can switch threads ?
00:37:17Zagorno, only yield() can
00:37:45alkorryes i'm noticing that
00:38:07alkorrapparently you don't use the sleep instruction too yet
00:38:36Zagorwe need the more advanced scheduler for that
00:41:26alkorrin ata_thread one tests sleeping before mutex whereas in ata_read_sectors we first mutex then change sleeping variable. Is it not bad ?
00:41:48alkorrwe tests sleeping variable...
00:42:13Zagorwhy is the recorder so slow? this is very strange.
00:42:30Zagori've done some benchmarks. the player is almost twice as fast!
00:43:05alkorrwith my ata.c or in more general way ?
00:43:20Zagorcpu-wise. i will post to the mailing list
00:43:33merwinZagor: the player is almsot twice as fast as what?
00:43:38Zagoras the recorder
00:43:50alkorrcpu-wise well...
00:44:14alkorrit is weird because their frequency is not too different
00:44:19merwinZagor: twice as fast doing what though
00:44:53Zagorexactly. the frequency differs about 8% but the performance difference is MUCH larger.
00:44:58alkorrmaybe because DRAM in recorder is slower than player DRAM ?
00:45:03Zagormerwin: cpu work
00:45:06Zagorsimple busy loops
00:45:09alkorrsince most code is executed in DRAM
00:45:18Zagortried iram too. same thing.
00:45:41alkorrbut it is the same cpu !!!
00:45:53Zagori know. that's why i'm so puzzled.
00:46:09alkorrhow do you count the time ?
00:46:13merwinZagor: that's interesting... it's not physically very much faster
00:46:17Zagorthe tick
00:46:27Zagorwhich is interrupt based
00:46:30alkorrtick ? timer ?
00:47:07alkorrare you sure about frequency used to initialize timer ?
00:47:25Zagori'll check
00:47:46Zagorbut it's not just the numbers. the recorder *is* more than 8% slower
00:47:58 Join Lowfiler [0] (
00:48:03Zagormaybe it has slower ram
00:48:24alkorryes but if it runs in sram it should have same speed
00:48:38Lowfilerzagor: can u tell me where the games now are? (sorry im not up 2 date)
00:48:50Lowfilerdid u deleted them
00:49:00Zagorno, they are still in.
00:49:07Lowfilerhow can i play them
00:49:14Zagordid you try the latest build?
00:49:30Lowfileri checked out the latest files
00:49:47Lowfilerajbrec.ajz now hast ~80k
00:49:50Lowfilerand no games anymore
00:50:04Zagorrun ../tools/configure
00:50:13alkorrwell my player has still 100% power :)
00:51:01ZagorLowfiler: then you choose "yes" to have games
00:52:18Zagoralkorr: it's the RAM that slows things down
00:52:32alkorrDRAM you mean
00:52:37Zagori will post all numbers (and benchmark) in a mail
00:52:53alkorri prefer that answer
00:53:17alkorrwell a 4 KB SRAM is a little bit short :(
00:53:50Zagorthe player's dram is much much faster than the recorder's
00:53:53alkorrand what if a SH7032 with a 64 KB SRAM ?
00:54:30alkorrstill the problem to download the firmware from ROM...
00:55:19alkorrwell if we cannot change for a DRAM with more MB we can perhaps find a quicker one ?
00:56:00alkorrbut still you need skills to do so :(
00:56:07alkorrso we are definately stuck
00:56:46alkorrhey maybe we can speedup things !!!
00:56:58 Join GomGom [0] (
00:57:15alkorrthere are some mode to communicate with DRAM which are surely not used by your rockbox
00:57:41GomGomi would like to know what script you use to generate the Changelopg from tyhe cvs log
00:58:26 Nick Lowfiler is now known as Low[a] (
00:59:07merwinZagor: I just noticed something... Cingular stole the X out of your HAXX logo :P
01:08:30 Join merwin- [0] (
01:08:38 Quit merwin (Read error: 54 (Connection reset by peer))
01:08:46 Nick merwin- is now known as merwin (
01:12:16ZagorGomGom: we wrote it ourselves
01:12:18alkorrbye all
01:12:21merwinbye alkorr
01:12:22 Part alkorr
01:12:39Zagorjust a bit of perl parsing
01:13:08 Nick _seb_ is now known as seb-away (
01:13:30GomGomZagor: could you please send it to me, i like the output :-)
01:16:24 Quit Tsd ()
01:21:07GomGomZagor: anyway, if you want to send it to me, my email is <ed.gomez at>
01:21:33Zagorok, remind me tomorrow. i'm going to sleep now.
01:21:49GomGomi woud be usefull for XviD stable tree changelogs :)
01:22:38merwinGomGom: xvid no longer exists
01:22:58GomGommerwin: xvid still exists, i'm a xvid team member
01:23:03GomGomi know if it exists
01:23:25GomGomwe are just continuing dev discretly
01:23:37GomGombecause sigma is stalling us again
01:23:48merwindamn sigma
01:24:05merwintheir codec sucks too...
01:24:16GomGomanyway, i should not talk about that... sort of "confidential" information
01:24:47 Quit Zagor ("Client Exiting")
01:24:56GomGom"their codec sucks too" <−− hey what do you mean ? XviD sucks too ? ;-)
01:25:16merwinGomGom: no, i'm just saying that they can't even steal code well :)
01:25:39GomGomok :)) i prefer that last sentence :)
01:25:39merwinI personally love xvid
01:25:43GomGombye bye
01:25:56 Quit GomGom ("Going to bed")
01:29:03 Join Tsd [0] (Tsd@
01:35:13 Join Zocher02 [0] (
01:37:28 Quit Miran ("Its not that you do not see, it is mearly that your eyes do not perceive")
01:37:54Mode"#rockbox +o adi|home " by ChanServ (ChanServ@services.)
01:38:19Topic"Bagder wants YOUR vote!" by adi|home (
01:40:26merwinhey adiadi
01:48:31merwinI want to implement loadable eq files
01:58:57 Part Zocher02
02:02:46 Quit fragglet ("BitchX-1.0c19 -- just do it.")
02:12:14 Join mecraw [0] (~mecraw@
02:12:33 Nick mecraw is now known as mecraw|work (~mecraw@
02:12:47 Join mecraw [0] (~lmarlow@
02:14:29merwinhey mecraw
02:14:38*merwin is working on loadable eq's :)
02:15:21*mecraw is lacing up his running shoes
02:15:29 Nick mecraw is now known as mecraw|running (~lmarlow@
02:15:59merwinquestion quick
02:16:17merwinbuf_set[d++] ... if d is 0 before executing this, will this equal buf_set[0] or buf_set[1]
02:16:50merwinthanks :)
02:26:19***Saving seen data "./dancer.seen"
02:29:09adi|homed++ is post increment
02:29:14adi|home++d is pre
02:29:22adi|homejust think of it left to right
02:29:53pimlottcif you really wanna be l33t use both
02:30:26adi|homeyou mean
02:30:31adi|homei would hit you for that
02:31:23adi|homebesides.. compiler don't like that
02:31:45pimlottcoh ya?
02:31:53adi|hometry it
02:32:14adi|homeint main() {
02:32:15adi|homeint i=0;
02:32:15adi|homeprintf("%d\n", ++i++);
02:32:15DBUGEnqueued KICK adi|home
02:32:29pimlottcthat's too bad
02:33:00merwin++d++ :)
02:33:16CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
02:33:16*merwin is almost ready to put forth a loadable eq
02:33:31merwinin snprintf, what is integer?
02:35:55merwin%d always means decimal in my brain
02:36:16pimlottcit should
02:36:20pimlottcunless you want to output as hex
02:36:56elinenbemerwin: you have a loadable eq?
02:37:19merwinif(strcasecmp(buf_set,"volume")) means if they're equal, no?
02:37:22merwinelinenbe: almost
02:38:42elinenbemerwin: can't wait to try it out once it is committed
02:38:52merwinelinenbe: :-)
02:39:59pimlottcmerwin - you should install the devel manpages
02:42:51merwinpimlottc: ?
02:42:55merwinOK, another problem
02:42:56merwin snprintf(global_settings.volume, sizeof(global_settings.volume),
02:42:56merwin "%s", buf_val);
02:43:06pimlottcmerwin - are you on linux?
02:43:09merwinsettings.c: In function `settings_load_eq':
02:43:09merwinsettings.c:443: warning: passing arg 1 of `snprintf' makes pointer from integer
02:43:09merwinwithout a cast
02:43:16merwinpimlottc: windows
02:43:39merwinbuf_val characters
02:43:59merwinchar buf_val[8]
02:44:49pimlottcfirst arg should be the target buffer
02:45:09pimlottcyou don't have help files for these functions? (snprintf, strcasecmp)
02:45:37merwinpimlottc: I'm not really a C person.. i'm used to VB where you can put anything anywhere :)
02:46:03pimlottcyou are trying to print the volume I assume?
02:46:08merwini know what each of the fields is, i just don't know what i'm doing wrong
02:46:24merwinpimlottc: i'm trying to put what's in buf_val into global_settings.volume
02:47:41pimlottcwrong way around
02:47:46pimlottcsnprintf is from values to string
02:47:48pimlottcyou want ssanf
02:47:53pimlottcer sscanf
02:48:35merwindo we have sscanf?
02:48:38merwini don't believe we do
02:48:41pimlottcno we don't
02:49:10pimlottcyou could make your eq files binary
02:49:20pimlottcor do the conversion yourself
02:49:45merwindo the conversion myself?
02:49:52merwincan't be binary, since these are user editable
02:51:11merwinthis is the last hurdle in the loadable eq :P
02:51:15pimlottcwhat's the format like
02:51:31merwin[volume]70#(0 to 100)
02:51:32merwin[bass]15#(-15 to 15)
02:51:32merwin[treble]15#(-15 to 15)
02:51:32DBUGEnqueued KICK merwin
02:51:32merwin[balance]0#(-100 to 100)
02:51:45merwinthe (0 to 100) part isn't necessary
02:51:59pimlottcif you keep it as ints only it shouldn't be hard
02:52:08pimlottcfloating point wouldn't be too much harder
02:52:17merwinall of them are ints
02:53:19pimlottcwhile (more digits) number = number * 10 + (next_char - '0')
02:53:52pimlottcand look for the - at the start too
02:58:01merwinpimlottc: Volume 8030300 %
02:59:22pimlottcturn it down
02:59:25 Nick mecraw|running is now known as mecraw (~lmarlow@
02:59:26pimlottcyou'll damage your ears
02:59:44merwinOK, now it's Volume -729988952
02:59:46pimlottcprobably blow up your archos too :/
02:59:57merwinpimlottc: i think something is wrong with the calculation
03:00:07pimlottcmethinks you're right
03:00:22merwini'm going to check something
03:00:52merwinwhen all else fails, throw some debugging code in
03:01:42pimlottcthere was an episode of x-files where they had a 'bug' in a virtual reality game that ended up killing someone
03:02:03pimlottcso first thing they do is suit up and go into the game to look
03:02:33pimlottcI was like "wtf is wrong with you, at least put some printf's in it first"
03:02:58merwini remember that x-files
03:03:22pimlottcpretty laughable
03:03:38pimlottcah well still a good show overall
03:03:46pimlottcending was a bit weak
03:05:58pimlottcmulder didn't 'reveal' jack shit
03:07:19merwinso something is wrong with this:
03:07:19 Quit mecraw|work (Read error: 104 (Connection reset by peer))
03:07:19merwin if(strcasecmp(buf_set,"volume"))
03:07:19merwin for(j=0;j<sizeof(buf_val);j++)
03:07:19merwin global_settings.volume = global_settings.volume *
03:07:19merwin 10 + (buf_val[j] - '0');
03:07:43merwini had to make j an unsigned int, since the compiler was complaining :P
03:07:44pimlottcdid you initialize global_settings.volume to 0?
03:07:53 Join mecraw|work [0] (~mecraw@
03:07:59merwinoh, no i didn't
03:09:22merwinok, i set it to 0
03:09:27merwinbefore doing that
03:09:36merwinand i come out with -8302
03:09:45merwinit's SUPPOSED to be 30
03:10:24pimlottcbuf_val is all digits?
03:10:40merwinwhen i lcd_puts'd it it had 2 digits :P
03:10:55pimlottcmake sure
03:11:18merwinit seems to come out as -8302 no matter what buf_val is;
03:11:27merwindamn, now i'm even ending my sentences in ;
03:11:43pimlottctry strlen
03:11:46pimlottcnot sizeof
03:12:08pimlottcwell assuming it is a real str
03:12:18merwinis it ok that i'm using unsigned int for j?
03:13:17pimlottcyou sohuld
03:13:19pimlottcer should
03:13:20merwinnow it gets set to 0%
03:14:09CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
03:14:09*merwin is ignorant of the ways of c
03:17:18merwinpimlottc: you want to take a look see at the function in a whole?
03:20:41merwinpimlottc: it's in cvs now, ifdef'd as CUSTOM_EQ if you want to take a look at it. I'm going to take a break
03:21:03merwinfile format as:
03:21:03merwin[volume]70#(0 to 100)
03:21:03merwin[bass]15#(-15 to 15)
03:21:03merwin[treble]15#(-15 to 15)
03:21:03merwin[balance]0#(-100 to 100)
03:22:22merwinI may use 0 to 50 for the volume, 0 to 30 for the bass and treble, and 0 to 200 for the balance... to make it easier... dunno
03:24:01pimlottcya sure msg me the func
03:24:22merwinpimlottc: it's in cvs now, ifdef'd
03:24:36pimlottcI need to get me some of that
03:24:51merwinsome of what?
03:25:09pimlottccvs commit access
03:25:48merwinpimlottc: heh... i got it originally because i can't do any cvs unless it's ssh (firewalls)
03:31:18merwinpimlottc: the files it will load is .eq, in case you didn't notice :)
03:39:08merwinI was just thinking... with the correct boot disk, you could run linux on the jukebox
03:46:07 Join fragglet [0] (
03:51:35merwinpimlottc: bleah?
03:57:16pimlottcI would read in the full line first and then use str functions to grab individual parts
03:57:23pimlottc(assuming I had getline of course)
03:57:50pimlottcI guess this is more efficient but it's not pretty
03:58:12merwinpimlottc: but it doesn't work as of yet
03:59:26adi|homemerwin: define loadable EQ's?
04:00:08pimlottcbtw I'm sure you could run linux on the archos, I was thinking about that the other day
04:00:10merwinadi|home: .eq files
04:00:27merwinadi|home: volume, bass, treble, balance
04:02:20adi|homegot ya...
04:02:31adi|homei doubt you could run linux ont he player
04:02:41pimlottcwhy not
04:02:43merwinadi|home: yeah you could...
04:02:44adi|homearchos rather
04:02:48adi|homewell.. 1. memory
04:03:01adi|home2. processor speed would be intensly slow no?
04:03:07merwinadi|home: a boot disk that loads the usb driver and then loads linux off of the archos
04:03:08fraggletdoes the chip have a mmu
04:03:19pimlottcthere's already embedded linux
04:03:22merwinadi|home: i'm not talking standalone
04:03:28adi|homeOHHH okay..
04:03:29adi|homethats diff
04:03:55merwinWho's the one working on the recovery disk for the HD lock problem?
04:04:32pimlottcoh I was a bit
04:05:03pimlottcjust put a linux kernel on a floppy with loadlin running it from autoexec.bat
04:05:15pimlottcnever actually tested it though
04:05:33pimlottcwc64 made a self-extracting version
04:06:31*merwin REALLY wants to run linux on his laptop
04:06:43fraggleti run linux on my palmtop
04:06:43pimlottcwhat's stopping you
04:06:55pimlottcfragglet - zaurus?
04:07:03fraggletpsion revo
04:09:02pimlottcwhat can you actually run on it
04:09:11pimlottcnethack? :)
04:09:55fraggletbetter... picogui battleship!
04:10:28pimlottcso no nethack?
04:10:37fraggleti could compile nethack for it i gues
04:10:44pimlottcbattleship smattleship give me nethack and I'll be impressed
04:17:23pimlottcso does that thing have a pcmcia slot or anything
04:22:09fraggletunfortunately not
04:22:21fraggletthe 5mx has a compactflash slot
04:22:30fraggletwhich i think is basically cutdown pcmcia
04:23:50*merwin has decided he will drive from Chicago to Minneapolis instead of flying
04:24:19merwinI fly too much
04:24:23merwinI like driving :P
04:24:30adi|homedo your arms get tired?
04:24:51pimlottcfragglet - with a wireless card that'd be a neat toy
04:24:53merwinbesides, i've never been to wisconson
04:25:13fraggleti kind of regret getting a revo instead of a 5mx
04:25:19fraggletbut on the other hand, its soooo cute
04:26:14merwinIf I go through wisconson, i'll only have 11 states left to visit :)
04:26:17fraggletwith a compactflash ram card you can run debian actually
04:26:20***Saving seen data "./dancer.seen"
04:26:36pimlottcI found the page heh
04:27:35 Nick merwin is now known as merwin|zZzZz (
04:27:35DBUGEnqueued KICK merwin|zZzZz
04:28:34fraggletthe revo is basically a cutback 5mx
04:30:56 Quit Tsd ()
04:36:47pimlottclot cheaper I see
04:49:14 Quit edx|sleep (Read error: 110 (Connection timed out))
05:01:18 Join elinenbe_ [0] (
05:01:21 Quit elinenbe (Read error: 104 (Connection reset by peer))
05:26:09 Quit elinenbe_ (Read error: 104 (Connection reset by peer))
05:28:40 Join elinenbe [0] (
05:30:21 Nick elinenbe is now known as elinenbe|ZzZzZ (
05:41:27 Quit mecraw (
05:41:44NJoinmecraw [0] (~lmarlow@
06:00:53 Nick seb-away is now known as _seb_ (
06:15:06 Nick _seb_ is now known as seb-sleep (
06:26:24***Saving seen data "./dancer.seen"
06:29:09 Join edx|sleep [0] (
06:40:39 Quit edx|sleep ()
07:13:04 Join LinusN [0] (
07:14:05 Part LinusN
07:15:04 Join LinusN [0] (
07:25:46 Join Zagor [0] (
07:27:57LinusNZagor: already!!!???
07:28:25Zagoryeah, i've got an electrician coming soon...
07:28:58LinusNZagor: have you seen the new menus?
07:29:28LinusNthey're nice
07:35:00Zagordid elinenbe discuss it with you before committing?
07:36:07LinusNi was asleep
07:36:21pimlottcheh, didn't see that got commited
07:36:32adi|homehe didn't discuss with anyone...
07:36:48Zagori'm very unhappy about that kind of commits.
07:37:01LinusNso we can only commit when you're awake?
07:37:02pimlottcI didn't ask anyone to commit it; I sent it to the list
07:37:13adi|homepimlottc: its not your fault
07:37:31pimlottcjust making it clear :)
07:38:12ZagorLinusN: yes you can, but not random patches right off the mailing list
07:38:24Zagori want to be part of the decision process
07:38:27LinusNZagor: i know, i'm just being grumpy as usual :-)
07:38:51Zagorin this case, it's not a big deal and I approve of the changes. I just don't approve of how it was executed.
07:38:56adi|homewhy should we have you as part of the decision process..?
07:39:01adi|homeyour opinion don't count
07:39:11LinusNwe don't even like you
07:39:38CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
07:39:38*adi|home dubs zagor "prince of the untouchable mole ppl"
07:40:34adi|homei responded to a car accident today..
07:40:45adi|home2 girls w/o neck and back pain
07:41:08adi|homei felt bad.. we transported them.. and all i could think was "damn girl.. can i bend you over a barrel?"
07:41:10LinusNif course they wer in pain, they were without necks!
07:41:21adi|home18 year old nubile young latinas
07:41:23*adi|home drools
07:41:33adi|homesory.. WITH ...
07:41:40adi|homeoh.. they had necks alright ;)
07:42:18adi|homeit was sweet though.. both were really scared... they held hands the entire way to the hospital...
07:44:47LinusNZagor: the DRAM timings are scary!
07:45:10Zagorthey sure are
07:45:34pimlottcI keep thinking the battery charging is not working right
07:45:40pimlottcthen I keep finding my cable got unplugged
07:45:42LinusNi wonder if the 12MHz recorders have the same DRAM
07:46:39LinusNone of these days we'll have to adjust the timers on the 12MHz recorders
07:47:33LinusNdo we know what type of DRAM they have?
07:47:47LinusNthe recorders
07:48:19Zagoryeah, EDO
07:48:33Zagorplayers have FP
07:49:29 Quit pimlottc ("leaving")
07:49:32LinusNthat shouldn't make that much of a difference
07:49:54Zagorno. the data sheets also don't differ much in timing.
07:51:47LinusNwe have samsung dram sheets on the rockbox page, when the dram in (phil's) recorder is from ISSI
07:53:24LinusNwho said it was samsung?
07:53:29LinusNor was it a guess?
07:56:55LinusNZagor: do you have permission to publish the ISD300 data sheets?
07:56:56 Quit Zagor (Read error: 54 (Connection reset by peer))
07:57:19LinusNi guess the electrician cut the power :-)
08:02:03 Join Bagder [0] (
08:05:18LinusNyo Bagder!
08:05:33 Join rwood [0] (
08:11:52Bagderzagor's benchmarks sure are interesting
08:12:57 Join Zagor [0] (
08:13:03Bagderhey bro
08:13:38Zagormy electrician pulled the fuse... :-)
08:13:44Bagderinteresting benchmarks
08:13:58Zagoryeah, we must be doing something wrong
08:14:18LinusNi am investigating...
08:15:00rwoodi'm still hooked up to the DRAM and all of its control lines - anything you want me to look at?
08:15:17rwoodArchos vs Rockbox
08:15:23LinusNthe RAS/CAS timing in general
08:15:43LinusNZagor: Burst enable os 0 on Recorder
08:15:46LinusNis 0
08:15:59LinusNbut 1 on player
08:20:16LinusNZagor: set 0x5ffffa8 to 0x1e00 and try again
08:20:29LinusNi did, and my unit works
08:21:00Zagorwhere did you do it?
08:21:19LinusNin gdb
08:21:26Zagorah :-)
08:21:38LinusNloading my 4000 file playlist is much faster now
08:22:48LinusNat least it felt faster... :-)
08:23:06adi|homeoh.. btw...
08:23:11adi|homejust so everyone knows..
08:23:15adi|homewe have been consistant
08:23:20adi|homeit has been 24 hours.. and well...
08:25:06adi|homebagder how does the tshirt vote stand?
08:25:20Bagderyour design is in the lead
08:25:55BagderI think
08:26:11BagderI'm not 100% on how to weight the scores yet
08:26:27***Saving seen data "./dancer.seen"
08:26:35BagderI mean 1 - 2 - 3, how much is a first place worth compared to second or third
08:27:26adi|homeim not worried who's in first.. just who's top 3..
08:27:26LinusNi think 3 points for 1st place, 2 for second and 1 for third
08:27:44adi|homeid agree with linus
08:27:49Bagderthat's what I do now, but I'm not sure one first place equals three third places
08:27:53adi|homewhat i wold _love_ to see
08:27:57adi|homeis when we are down to 3
08:28:01adi|homehave on last vote
08:28:07adi|hometo choose between those 3
08:28:10adi|homebut *shrug*
08:28:19LinusNBagder: i know what you mean, do you have another suggestion?
08:28:30BagderI'm thinking perhaps 4, 3, 2
08:29:13LinusNyeah, why not?
08:29:47LinusNstill, i think the voters haven't really cared that much about 2nd and 3rd place
08:30:17LinusNthey have selected the 1st, and picked two more just because they felt they had to
08:30:27Bagderthat might be so, yes
08:30:38Zagorwhoa! burst mode made test2 go from 15s to 5s!
08:30:56LinusNthere must be a catch, though...
08:31:14Zagoryay, playlists now load at *2700* files/sec!
08:31:44adi|home Somehow your bug report to the Rockbox project must have been deleted. I can't find any bug report about crappy sound on 1.3 in our bug database.
08:31:44adi|home So if you don't file a new bug report, you will probably have to wait forever to get your sound problem fixed.
08:31:44adi|home /Linus
08:31:46adi|homeLOL i love it
08:32:00LinusNhehe, sarcasm feels great!
08:33:14LinusNthere MUST be a reason why the ROM firmware doesn't enable burst mode
08:34:05Zagori think iram vs dram test on player is interesting. it's only 3% faster with iram.
08:34:52Zagorsame on recorder with burst mode
08:35:41ZagorI say we enable burst and see if anything breaks
08:35:48Bagder39 people counted, 3 bad votes, 114 used
08:35:49BagderPos Design # Points 1st vote 2nd vote 3rd vote
08:35:52Bagder 1: 6 31 7 5 0
08:35:53Bagder 2: 24 26 7 1 3
08:35:53Bagder 3: 7 21 5 2 2
08:36:30Bagderthat's top-3
08:37:48adi|homeim just amazed at the variety of designs we got...
08:37:51adi|homesoooo cool
08:38:15LinusNZagor: i have a feeling that it might break older recorders, maybe the 12MHz ones
08:38:28LinusNbut let's find out, shall we? :-)
08:38:42Zagoryes, i think so. no points for being a coward.
08:40:39adi|homeoh.. in case anyone cares (and not that we need the press ;) ) Im starting to collect happy user comments in a file
08:40:43adi|homejust for future review :)
08:41:04Bagderego-boosts are never bad
08:41:56adi|home... the RockBox developing team, have (and very selflessly so) devoted your time and efforts to that of making the ownership & operation of our mechanical MP3 counterparts a much more pleasurable experience. That you do so, without expectation of remuneration, is truly a rare - and wondrous thing to me. It was earlier tonight that I had traveled back to the Archos F/W, because of its feature on "recording". But RockBox tra
08:41:56adi|homeature that Archos has...
08:42:01adi|homeis the only one i have as yet ;)
08:42:05LinusNhehe, look at this:
08:42:09LinusNIf you have an Archos Jukebox, trust me on this: Rockbox is nowhere near competing with Archos for the firmware on this thing. All I want is someone to fix the fucking shuffle on this thing, but no. I installed Rockbox and deleted it almost immediately. It sucked that much.
08:42:33LinusNhappy user comment
08:42:33Bagderbut that is old, isn't it?
08:42:38adi|homeheheh wher eyou get that?
08:42:53adi|homewhat you do, google Rockbox?
08:43:06LinusNa do that once in a while
08:47:35adi|homenice little linking job this kid did :)
08:52:19LinusN"I don't say that as the rockbox guys don't want to get
08:52:19LinusNdaily built's bugs posted"
08:52:31ZagorBagder: i'll answer
08:52:45Bagderbut why doesn't he get rolo working you think?
08:53:09LinusNbecause his mp3/m3u filter is on
08:53:32adi|homethe problem is between the keyboard and the chair
08:53:51rwoodhe has a nut loose on the console
08:53:55LinusNand he doesn't have his original firmware in a file, only in ROM
08:54:15LinusN"as I have the most recent ? player I don't have a firmware file."
08:54:43Bagderobviously Archos always lag to produce firmware files
08:54:52LinusN"it doesnt switch to USB mode while being on
08:54:52LinusN when plugged."
08:55:01LinusNnah, i guess he was in a menu
08:55:04LinusNor something
08:55:33adi|home"and mommy says its because stupid is as stupid does"
08:56:39LinusNrwood: what is RTS8891?
08:57:23rwoodLinusN: i don't know - where did it come from?
08:58:10ZagorLinusN: that was rob ward, not randy wood. close, but not the same :-)
08:58:44LinusNi am blind
08:58:50rwoodwhew - thought it was time to sign up for the home
08:59:03LinusNdoes anyone know what Rob Ward was asking about?
09:00:33*Zagor is cleaning out inactive people from cvs access list
09:02:03adi|homeid really like to know how many ppl are running rockbox on a daily basis now :(
09:02:07HesGood morning
09:02:30Bagdermorning hes
09:03:40adi|homelinus.. the guy that said we "sucked that much"
09:03:46adi|homewas it this url:
09:03:49adi|homelol sorry
09:03:52adi|homemis paste
09:04:04LinusNadi|home: hehe, your "READ THE FAQ" in the vorbis thread at funmp3players wasn't all that nice :-)
09:04:14LinusNespecially since he didn't ask about Rockbox
09:04:15adi|homeya think
09:04:20adi|homefuck em they deserve it...
09:04:45Bagderwe're losing rank on sourceforge! ;-)
09:04:57adi|homeim double checking to make sure i wasn't wrong/outta line
09:04:58adi|homehold on
09:05:02LinusNwooo, adi|home, are you trying to quit smoking? :-)
09:05:03Zagorback in a flash...
09:05:04 Quit Zagor ("Client Exiting")
09:05:39LinusNBagder: commit something, anything!
09:06:04LinusNwe should have an auto-commit script
09:06:12LinusNjust to keep the ranking position
09:06:25 Quit fragglet (
09:06:25 Quit Bagder (
09:06:25 Quit Hes (
09:06:25 Quit elinenbe|ZzZzZ (
09:06:25 Quit Low[a] (
09:06:25 Quit adi|home (
09:06:25 Quit Hadaka (
09:06:25 Quit miah (
09:06:25 Quit Tumm (
09:06:25 Quit mecraw (
09:06:25 Quit Synthe (
09:06:25 Quit datazone (
09:06:25 Quit dwihno|gone (
09:06:25 Quit laotan (
09:06:25 Quit rwood (
09:06:25 Quit merwin|zZzZz (
09:06:27 Quit PsycoXul (
09:06:27 Quit seb-sleep (
09:07:01NJoinadi|home [0] (
09:07:01NJoinrwood [0] (
09:07:01NJoinBagder [0] (
09:07:01NJoinmecraw [0] (~lmarlow@
09:07:01NJoinelinenbe|ZzZzZ [0] (
09:07:01NJoinfragglet [0] (
09:07:01NJoinmerwin|zZzZz [0] (
09:07:01NJoinLow[a] [0] (
09:07:01NJoinSynthe [0] (
09:07:01NJoindatazone [0] ([eMuNJtk1X@
09:07:01NJoindwihno|gone [0] (dwihno@Bald067.Baldakinen.Umea.SE)
09:07:01NJoinHadaka [0] (
09:07:01NJoinmiah [0] (
09:07:01NJoinlaotan [0] (
09:07:01NJoinHes [0] (
09:07:01NJoinPsycoXul [0] (
09:07:01NJoinTumm [0] (
09:07:01NJoinseb-sleep [0] (
09:07:01Mode"#rockbox +o adi|home " by
09:07:56adi|homeLinusN: appology posted
09:07:58adi|homesee if that helps
09:08:12rwoodLinusN: there's nothing remarkably different in the /RAS timing between the two - 270ns /RAS to /RAS
09:08:36LinusNrwood: i figured so
09:08:46adi|homei _wasn't_ out of line...
09:08:57LinusNi guess the burst mode is the key
09:09:09adi|homeread the subject "Ogg Vorbis and Archos, maybe Rockbox"
09:09:20*adi|home goes back to his bitching corner
09:09:38adi|homeand no.. no on the smoking.. just tired of stupid ppl in this world :)
09:10:39adi|homelinus.. the following post after mine.. wtf does that mean? " Chill. This concerns an fpu, or rather the lack of it. Tremor. Nex"
09:10:57 Join Zagor [0] (
09:11:03rwoodfloating point unit
09:12:41adi|homegot ya
09:18:43adi|homefear the off shoots
09:21:05LinusNhahaha, really cool
09:22:39adi|homelinus.. the FAQ i went off with..
09:22:40LinusNburst mode committed
09:22:48adi|homeread the subject of that message on the initial post ;)
09:22:51adi|homethats where it came from..
09:22:54adi|homebut i appologized
09:23:08 Join bobTHC [0] (
09:23:43Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
09:24:43bobTHCgood morning all!
09:24:53Zagorwhatever happened to wavey?
09:25:04Bagderwe don't know
09:25:27Bagdermaybe he got a life? ;-)
09:25:43LinusNpoor thing
09:26:41adi|homeis that a new feature i missed?
09:26:51*adi|home searches his rockbox for a 'life'
09:26:58Bagdernah, no life here ;-)
09:27:51Hes93 instances of Life on mine 8-)
09:28:56adi|homei have to say..i like the menu patch
09:29:03adi|homebut i still prefer [title]
09:29:07adi|homeon the menu titles
09:29:11adi|homelooks better to me...
09:29:17Zagorbjorn:~# man life
09:29:18ZagorNo manual entry for life
09:31:31adi|homewhat about man sarcasm?
09:31:37adi|home SARCASM - Its About Humor
09:31:37adi|home sarcasm [-v -d]
09:31:37***Alert Mode level 1
09:31:37adi|home Sarcasm Sar"casm, n. [F. sarcasme, L. sarcasmus, Gr.
09:31:48adi|home sarkasmo`s, from sarka`zein to tear flesh like dogs,
09:31:48adi|home to bite the lips in rage, to speak bitterly, to sneer, fr.
09:31:48adi|home sa`rx, sa`rkos, flesh.] A keen, reproachful expression; a
09:31:48***Alert Mode level 2
09:31:48adi|home satirical remark uttered with some degree of scorn or con-
09:31:48***Alert Mode level 3
09:31:48adi|home tempt; a taunt; a gibe; a cutting jest.
09:31:48***Alert Mode level 4
09:31:48adi|home sarcasm n : witty language used to convey insults or
09:31:50adi|home scorn; "he used sarcasm to upset his opponent"; "irony is
09:31:52adi|home wasted on the stupid"
09:31:56adi|home -v Verbose mode. Default is Silent.
09:31:58adi|home -d There is no option "d". See sarcasm.
09:32:00adi|homeSEE ALSO
09:32:02adi|home woman, man, irony ["sarcasm"], satire ["sarcasm"].
09:32:04Zagoruh, hurts my buffer
09:32:23adi|homeokay.. nap time
09:32:37adi|homedo ppl prefer menus with the [title]
09:32:40adi|homeor without?
09:33:07PsycoXulhow about sex.6
09:33:16 Join hardeep [0] (
09:33:36hardeephi all
09:33:54LinusNhi hardeep
09:35:13 Quit Synthe ("Connection Lost - Excess Blood")
09:37:01rwoodLinusN: i just download a clean copy of the CVS - failed with an undefined ref to ROLO_LOAD in tree.o
09:37:30Zagoryou run linux, right?
09:37:44 Join Synthe [0] (
09:38:04rwoodneed extern?
09:38:14hardeephmmm, i just did a clean compile on cygwin and it worked fine
09:38:38Zagordoes cygwin use the auto-find-sorce-files feature in the Makefiles?
09:38:58rwooddon't know - novice user
09:39:00hardeepzagor: yes
09:39:48Zagorno problem here either
09:41:49***Alert Mode OFF
09:42:23rwoodi screwed something up going from the computer with CVS to cygwin maching - sorry
09:50:55 Quit Low[a] (Read error: 60 (Operation timed out))
09:51:09 Join Low[a] [0] (
09:51:21 Join alkorr [0] (
09:52:45hardeephi alkorr
09:53:05LinusNhi alan
09:53:51alkorrLinus : it works now but I don't really understand the necessity of ata_thread
09:54:01LinusNalkorr: no?
09:54:42alkorrno, because having a sleep mode in a thread is not safe
09:55:06LinusNuhuh, so how should it be done then?
09:55:17LinusNand why is it not safe?
09:55:40alkorryou read or change variables without a mutex
09:56:41alkorri mean it is a little bit ackward especially if we need to have a better scheduler for future
09:56:44LinusNand the ata thread shared those variables with who?
09:57:06LinusNalkorr: we will probably never switch to a preemptive scheduler
09:57:18Bagderoh no, we won't go that way
09:57:29Bagderthat's give us much grief
09:57:40alkorrata_read_sectors for example (sleeping, sleep_timeout, etc...)
09:57:53LinusNalkorr: that is why it is a thread!
09:58:05Zagoralkorr: the better scheduler i talked about was about simply not scheduling sleeping/waiting threads
09:59:33rwoodLinusN: not that makes a difference - multiple CASs per RAS
09:59:44LinusNthe whole structure depends on and exploits the cooperative threading
10:00:04alkorryes it is why i didn't really like the ata_thread
10:00:11alkorrthere is no priority
10:00:40Bagderhow else should it be done?
10:03:21alkorrhave a timer like a timeout which performs sleep only if no disk operation occurs for a time lapsed, this timer you activate it exiting an operation, you disactivate it entering an operation par example.
10:03:59LinusNlike it is today
10:04:21LinusNand how is the timer handled? with an interrupt?
10:04:58alkorrdon't you have trigger timers
10:05:10LinusNtrigger timers?
10:05:59alkorrwell let's fall it
10:06:48LinusNi'm just curious, do you see a design problem in the current implementation?
10:07:01alkorrusing an ata thread that's way wastes cpu ressources.
10:07:20Bagderour thread system wastes cpu
10:07:39alkorrbecause you can wake this thread when it doesn't need (yield() doesn't remove it)
10:07:52Bagderbut that's our thread system
10:08:03alkorrit is why i don't like the use of ata_thread because of our thread system, Bagder
10:08:12LinusNand that's the improvement Zagor mentioned
10:08:28Bagderyes, but the fix is then *not* in the ata thread but in the thread system
10:08:40Bagderor am I wrong?
10:09:14alkorri don't say we need a preemptive threading btw. Just a good improvement of cooperative threading
10:09:35Bagderyes, and that's what we're saying as well ;-)
10:10:56alkorrBagder: ata_thread should not block with a mutex when trying a ata_perform_sleep but when other thread are USING ata functions
10:11:21alkorrlet me explain why
10:11:31alkorryou enter in ata_read_sectors
10:11:32Bagderyou're mixing things a lot now
10:11:48LinusNalkorr: yes, as it is today, the ata thread may try to SLEEP in the middle of a directory read
10:13:13LinusNfor example
10:17:32 Quit rwood ()
10:20:38 Join Blaster_Master [0] (
10:24:36 Join notch [0] (
10:26:31***Saving seen data "./dancer.seen"
10:26:56 Quit notch (Client Quit)
10:38:20Bagderis it Indian Rockbox Lunch Day today?
10:38:54BagderI'll be there
10:39:19Zagori'll try. depends if my electrician is finished demolishing my apartment :-)
10:39:48Bagderwhat is he supposed to do?
10:41:16 Quit Blaster_Master ()
10:41:47Zagorinstall power to my dish washing machine. he wound up having to pull a cable through some walls :-)
10:43:49Zagor"The classical Jukebox have an imput too, and he can probably record in PCM too"
10:43:58Zagorwhere do people get all their "facts"?
10:44:22Bagder45. Would it be possible to record from line in on the player?
10:50:06Zagori'll keep a little todo list there, in case people want to see what is up
10:50:13Bagderfix loadable fonts
10:50:18Zagorgive me a yell if you volunteer for a pont
10:50:27Bagdera pint?
10:51:08Zagora pint is always nice
10:58:46BagderI'm really not happy with the loadable font situation
10:59:22BagderI think I'll better try to understand the file format
10:59:35BagderI was thinking of rewriting the whole thing, but that might be more work
10:59:47Zagorplease do. i keep meaning to look at it, but there's always more pressing things.. :-(
11:08:12 Join idefx [0] (
11:09:13idefxHi, i have 2 questions: what's the better thing to do: make my own strcasestr function, or add the posix one to the firmware, and then, seconde question: where can i find posix source of strcasestr ?
11:09:31Bagderthere is no
11:09:38Bagderstrcasestr is not posix
11:09:43Bagderit's not standard at all
11:09:44idefxBagder: True ;), so, the gcc one ??
11:09:55idefxi may have to dl gcc source ?
11:09:57Bagderis there one in gcc?
11:10:04idefxBagder: yes.
11:10:14Bagderand you don't mean glibc?
11:10:14idefxto be true, my source compile for the simulator !!
11:10:28idefxBagder :yes, i mean glibc..
11:11:23idefxthat's no big deal to write my own one, i think i'm gonna do that. I have to test my search-function using database (and strcasestr is the only thing missing.)
11:12:57Bagderthe hard part would be to make it really fast on target
11:13:16BagderI bet a boyer moore would be a *lot* smarter
11:13:55 Join kargatron [0] (
11:18:51 Quit Zagor ("Client Exiting")
11:19:23BagderLinusN: ok, see ya in ten minutes then
11:20:55LinusNBagder: Zagor wants to wait an hour
11:21:19kargatrondoes one need to compile the source with a custom option to access the custom WPS functionality? I thought I saw that in the mailing list archives, wanted to make sure
11:23:05BagderLinusN: one hour it is
11:23:17Bagderkargatron: it is default
11:23:36kargatronah, so if the file exists, it looks at it?
11:24:04kargatronis the right name wps.config or default.wps? doc is confusing on that
11:24:28Bagderthe latter
11:25:31kargatronand oe more: each file line corresponds to a display line, and those display lines are #2-7 (not top or bottom)?
11:26:11Bagderthey should be 1-7 as the status bar can be removed
11:26:27BagderI would personally like to see the progress bar as a wps-config too
11:26:28kargatronbut the .wps file is 6 only?
11:26:46kargatron> * The file may be 2 lines long for the Player, and 6 lines for the Recorder.
11:27:03Bagder#ifdef HAVE_LCD_BITMAP
11:27:03Bagder#define MAX_LINES 10
11:28:30 Join Zagor [0] (
11:33:32kargatrondoes playlist name have a wps.config tag?
11:35:43kargatronlooking at it, just making sure it wasn't lacking any recent additions, since no pl name is noted
11:36:11BagderI have no clue
11:46:09 Quit bobTHC ("Trillian (")
11:50:27BagderI *think* I've found the problem with loadable fonts not being proportional
11:52:30 Join notch [0] (
11:52:59notchis burst DRAM mode in cvs ?
11:53:16notchwhooa... gotta check it out...
11:55:21 Quit hardeep ("Leaving")
12:00:19notchshould this make skipping a track faster on the recorder?
12:00:42LinusNeverything is faster
12:01:40 Quit idefx ("BitchX-1.0c19 -- just do it.")
12:14:05Zagorwhat is a good hexcalc in linux?
12:14:29Zagorcan it do hex?
12:14:56Bagderit wants uppercase letters though I think
12:16:05Bagderbdf2ajf bugs
12:23:21BagderZagor: you ready now then?
12:24:06LinusNi'm hungry!
12:24:10Bagderme too
12:24:17Zagoryup, leaving now
12:24:31Bagdersee ya in ~10
12:24:37 Nick Zagor is now known as Zagor|lunch (
12:26:35***Saving seen data "./dancer.seen"
12:30:53 Nick LinusN is now known as LinusN|lunch (
12:55:01 Join bobTHC [0] (
12:57:05*kargatron is afraid to reply when he knows he probably won't be able to answer any comments or questions
13:13:20 Nick seb-sleep is now known as _seb_ (
13:17:43notchthe burst mod DRAM rocks...
13:18:14*notch has to think faster to catch up with the recorder...
13:18:53bobTHCwhat is the advantage notch ?
13:19:28notcheverything is just faster... track skipping incl...
13:20:19notchlatest cvs...
13:20:31notchOnly on the recorder though...
13:24:21 Quit mecraw|work (Read error: 104 (Connection reset by peer))
13:24:51 Join mecraw|work [0] (~mecraw@
13:25:41bobTHCbadger, how many vote?
13:26:19alkorrthey are lunching
13:32:08 Quit kargatron (Read error: 113 (No route to host))
13:42:45 Nick _seb_ is now known as seb-school (
14:00:16 Join Blaster_Master [0] (
14:01:35Blaster_Mastera quick question, what is the best "Allocation Unit Size" to use when reformating the player6000 ?
14:02:39Blaster_Masterand is there anything to gain from formating to a smaler/larger "Allocation Unit Size" ?
14:04:01 Join coutta [0] (
14:20:43 Quit Blaster_Master ()
14:24:46 Join kargatron [0] (
14:26:38***Saving seen data "./dancer.seen"
14:29:20 Join idefx [0] (
14:29:54idefxHi, anyone want's to know how long my database-search function takes to scan 592 files ?
14:30:07kargatrondo we get to guess first?
14:30:16idefx(sorry bout it i'm happy;)
14:30:26kargatronthen i won't guess 591 secs
14:30:27idefxno, you don't need. less than 3 sec.
14:30:35idefxon the device !
14:30:45idefx(on a recorder)
14:31:32idefxi'll have to make things a bit cleaner, to have directory gestion and such, and an text input, but, it already have good result, so i'm happy ;)
14:34:34idefx(if someone want's to write a windows tool to make the database, that would be nice too)
14:35:42kargatrondoes the search scale? searching 4000 files would take 20 secs?
14:35:55bobTHCno devive database generation idefix?
14:37:13idefxkargatron: yes, takes the db 512 bytes per 512 bytes, so, just have to multiply!
14:37:30idefxbobTHC: if you wan't, but that will be incredibly slow !
14:38:13 Join Snorlax [0] (
14:38:41SnorlaxHEY HO!!
14:39:13idefxfirst try was on-device-without-database search string, and it took 45 sec to scan 45 files. So, imagine how long it would be if it had to write at the same time. I guess a device database generation could take 1 min to make 80-100 files, so, quite slow.
14:39:26idefxSnorlax: yes, we are here ;)
14:40:16idefxsorry: it tooked 45 sec to scan 64 files.
14:44:25bobTHCso with a good incremential management fof the db.... and a good optimization why not..
14:44:47idefxbobTHC: If you wan't to write it, that would be nice !
14:48:02bobTHCu generate the db only if u want to search... and it's a choice between browsing all dir to find it or generate and find it.
14:50:29idefxyes ;)
14:57:12alkorrwell i suppose with the improvement due to DRAM burst, you will reach a better speed
14:57:45alkorrdid you try it with DRAM burst improvement ?
14:57:59notchBurst rocks...
14:59:43idefxalkorr: don't know, i'm using yesterday's cvs.
15:00:15alkorrso take the new system.c
15:00:57idefxwhat's the way to get only one file from cvs ??
15:01:09alkorrupdate this file
15:01:27alkorrusing cvs or wincvs ?
15:01:52kargatronfor those of us lazy users that only grab daily build compiled files, when will burst show up, tomorrow?
15:02:00alkorrshould be : cvs u firmware/system.c (i'm not sure)
15:02:05Bagderkargatron: yes
15:02:40alkorrBagder must know :)
15:03:03Bagderto get one file?
15:03:13Bagdercvs update [filename]
15:03:20Bagderupdates that file only
15:03:25idefxBagder : nice ;
15:05:46 Nick LinusN|lunch is now known as LinusN (
15:05:47idefxdoesn't change much in fact.but it's hard to see difference in 5 sec.
15:07:34Bagderconfigure update still doesn't do right
15:12:34 Nick elinenbe|ZzZzZ is now known as elinenbe (
15:15:37 Nick Zagor|lunch is now known as Zagor (
15:15:50Zagoridefx: you should use sorted index files instead of using a single database
15:16:06Zagor20 sec for 4000 files isn't really acceptable IMHO
15:16:10ZagorI have 5800 files...
15:16:49elinenbeidefx: how is the searching coming?
15:17:10idefxZagor: what you mean by sorted index ??
15:17:51idefxelinenbe: fine. Still waiting a text input, but it only take 3 to 5 sec to "scan" 592 files...
15:18:40Bagderidefx: so this is made with a strcasestr() now?
15:18:48Zagoridefx: files like artist.idx, album.idx, song.idx that has all names sorted alphabetially, with an index to the position in the database
15:18:56idefxZagor: Bagder : yes, own-made strcasestr ;)
15:19:01Zagorthen you use binary search to locate a name
15:19:13elinenbeidexf: is that with some sort of database or just with files?
15:19:23idefxelinenbe: sort of database
15:19:25Zagor4000 files mean 12 string comparison == millisecond search time
15:19:28BagderZagor: well that doesn't work if he doesn't search on the beginning of the string
15:19:41ZagorBagder: right
15:19:42Bagderthat's why strstr() is wrong way to do it
15:19:49BagderBoyer Moore rocks compared to it
15:19:54Bagderfor free-text search
15:20:05Bagdereverything else is silly
15:20:18Zagori think most people will search start-string, so binary should be used aswell, imho
15:20:49idefxZagor: i nearly never search start-string as for myself.
15:21:34elinenbemy ~4000 song playlist only takesabout .5 seconds to load now (on the recorder)
15:21:46elinenbepretty impressive.
15:21:50kargatronfor the masses that have tag names First Last, but want to search on Last, start-string is no good, right?
15:21:50Zagorelinenbe: hehe
15:21:53Bagderthat is so neat
15:22:06Bagderkargatron: right
15:22:19LinusNthe original archos firmware probably runs with burst=OFF
15:22:19idefxZagor: the reason why i made this stuff, is that i couldn't find a version of "over the rainbow" in my jukebox during the holiday, and the title was: "what a wonderfull world-over the rainbow"
15:22:35Zagorkargatron: correct. but since binsearch only takes 12-13 operations it should be ran first
15:22:56Zagorit will complete in a fraction of a second
15:22:58kargatronsure, can't hurt
15:23:01elinenbeidefx: I think most people will not use the search often, but when they do, it will really help
15:23:23idefxelinenbe : i spend more than 1h30 searching this damned song !
15:23:36Bagderfind is your friend
15:24:00idefxBagder : i mean, i spent 1h30 searching it on my jukebox, 5000 km to my computer ;)
15:24:28idefxthat's why i wan't a nice search function on the jukebox, for when you are on holiday ;)
15:25:05elinenbeZagor: I really like the reorganized menu commit >-) wahahaha
15:26:04idefxi post a mail on the mailing list, calling for suggestion about the db format (and someone to make a text input :=))
15:26:48idefxZagor: your idea seems very good, but, this is not really the function i was looking. I want a search for everywhere in the string.
15:27:08Zagoridefx: use both
15:27:09idefxBagder : just a question: what is boyer moore
15:27:10idefx ?
15:27:16idefxZagor: so make it !
15:27:27kargatronidefx, that's what google is for, no?
15:27:34Bagder"The Boyer-Moore algorithm is considered as the most efficient string-matching algorithm in usual applications."
15:28:27idefxBagder: OK, i'm gonna use it instead of strcasestr
15:28:40Bagderit'll be much faster
15:28:42idefxheu, has it yet been implemented to rockbox ?
15:28:52elinenbeboyer moore does not just seach for string, move one letter down, search for string, move one letter down
15:29:21elinenbeboyer moore is much quicker
15:29:25idefxBagder: would it be possible/easy/efficient to do it ?
15:29:46Bagderidefx: the concept is really easy, it is quickly implemented
15:30:12 Quit coutta (Read error: 113 (No route to host))
15:30:32idefxBagder: OK, if someone do a text input, i do the boyer moore ;)
15:31:17Bagdersomeone: do it
15:31:21*Bagder grins
15:31:34Zagorwe have it already. search the archive.
15:32:35Bagderdo we?
15:32:40elinenbewe need t9 input... but with a custom dictionary
15:32:45idefxZagor: haven't tested it, but the author seemed to say he wasn't completely happy of it. I will try?
15:32:48elinenbewe use one with band and song names.
15:32:53Zagorsomeone wrote a text input demo, maybe he never posted it on the list
15:33:14BagderI read that nintendo explanation, but I never saw any code
15:33:21Zagorno, not that
15:33:36Zagora simpler version, with just a matrix of chars and a cursor to navigate them
15:34:26bobTHCelinenbe: T9 sounds good but it's hard for multilangual version...
15:34:33Zagorwavey says hi. he's buried in work and says he might come again some time.
15:35:02Bagdergood to hear he's still alive ;-)
15:36:15 Join sylvr [0] (
15:43:33 Join pimlottc [0] (
15:55:15LinusNZagor: have you tried the loopback device?
15:55:16elinenbeI am glad wavey is alive −− REALLY glad.
15:55:40elinenbeLinus & alkorr: what is the staus with the new ata.c?
15:55:56LinusNZagor is working on it
15:56:06LinusNi think
15:56:20LinusNif you mean the MULTI_SECTOR_READ
15:56:52elinenbethat is what I am talking about.
15:57:39LinusNZagor: said that he think he could see some cache hits...
15:58:36LinusNbut the performance boost wasn't all that noticeable
15:58:37Zagoryeah, i've got it
15:59:24LinusNgotta go now, bye all!
15:59:34Zagorit's a pretty marginal performance boost, but we should still use it
15:59:47 Part LinusN
16:19:31 Join edx [0] (
16:19:35 Quit notch ()
16:26:40***Saving seen data "./dancer.seen"
16:27:40 Part kargatron
16:38:28 Join quelsaruk [0] (
16:41:13 Quit edx (Read error: 104 (Connection reset by peer))
16:41:34 Join edx [0] (
16:47:41quelsarukBagder: nowadays we can just use continuous play on rockbox, no?
16:48:44 Quit mecraw|work (Read error: 104 (Connection reset by peer))
16:57:54 Part Bagder
16:58:09 Quit Snorlax ("gittar ny!")
17:06:47 Join Snorlax [0] (
17:07:00SnorlaxROLO doesn't seem to work for me!
17:07:03 Join merwin- [0] (
17:07:57SnorlaxI used the newest build, and turn the mp3/m3u filter off and serched up another .mod file and tried to press it, but it doesn't work!
17:08:20merwin-Snorlax: what did it do?
17:08:28*merwin- wonders what Warp mode is
17:09:10*Snorlax didn't get ROLO to work
17:09:52Zagormerwin-: warp mode means we can use iram and dram at the same time
17:10:00merwin-Zagor: much faster?
17:10:15Snorlaxare u supposed to turn ROLO on in some way?
17:10:16Zagorno, not really
17:10:23ZagorSnorlax: no, it just works.
17:10:36Snorlaxmmhhmm, it just works...
17:10:37ZagorSnorlax: you have a player, right?
17:10:58merwin-Zagor: how about dram burst mode... that much faster?
17:11:09Snorlaxso it doesn't work on players?
17:11:13Zagormerwin-: oh, only about 200% :-)
17:11:16Snorlaxyet.. =)
17:11:17merwin-Zagor: nice :)
17:11:25merwin-Snorlax: it works on players
17:12:08Snorlaxohh right, quote from zagor: "it just works"
17:12:19Zagor"works for me"
17:12:39merwin-Zagor: did you see the beginnnings to the loadable eq that I have? I can't get over one hurdle with it :p
17:12:56merwin-Zagor: data type conversions are why I hate C
17:13:46merwin-Zagor: i'm loading strings out of a file that are really supposed to be ints, and i can't seem to get them INTO ints
17:14:08Zagormerwin-: atoi()
17:14:41merwin-atoi() hmmm... so it's myint = atoi(mychar), right?
17:14:51Zagorno atoi(mystring)
17:15:02Zagori.e: atoi(mychar*)
17:15:02merwin-oh, it'll take a whole string?
17:15:20Zagorhum, i *think* we have it :-)
17:15:32merwin-so if the string is MyString[8], then i do atoi(MyString), right?
17:16:32merwin-the one nice thing about VB... variants :)
17:16:47Zagornah. there is no nice thing about VB :-)
17:17:18merwin-my boss just called me and says she's trying to get me a new laptop!
17:21:17merwin-Zagor: atoi turned the volume into 0% ...
17:21:34Zagorwhat did the string look like?
17:21:58merwin- if(strcasecmp(buf_set,"volume"))
17:21:58merwin- {
17:21:58merwin- global_settings.volume = atoi(buf_val);
17:21:58DBUGEnqueued KICK merwin-
17:21:58merwin- }
17:22:29Zagoryeah, but what was buf_val? exactly.
17:23:13merwin-yesterday it was "30"... i would assume it's still "30" :)
17:23:17Zagorthere must be no character before the number
17:25:09 Quit merwin|zZzZz (Read error: 113 (No route to host))
17:25:38merwin-what's this mean
17:25:38merwin-settings.c: In function `settings_load_eq':
17:25:38merwin-settings.c:448: warning: implicit declaration of function `atoi'
17:26:12pimlottcit means it wasn't declared anywhere
17:26:17Zagor#include "atoi.h"
17:28:49merwin-that is correct, right?
17:29:24Zagorbut please put a space after 'if'
17:29:43pimlottcI've never seen an atoi type function that goes backwards like that
17:30:26merwin-Zagor: space after if? never seen it doen that way
17:31:00pimlottcalways did val = val * 10 + (string[i++] - '0')
17:31:17Zagori prefer spaces around C keywords
17:32:18merwin-Zagor: if(strcasecmp(buf_set,"volume")) was incorrect... it needs to be if(!strcasecmp(buf_set,"volume"))
17:32:44Zagormerwin-: run it in the simulator with DEBUGF()s and breakpoints. makes debugging these things much simpler.
17:33:33merwin-Zagor: player sim doesn't really work on windows
17:33:35Zagormerwin-: strcmp returns 0 if the strings match.
17:34:12merwin-Zagor: The only one of the 4 settings lines that if(!strcasecmp(buf_set,"volume")) found to be true was the volume line
17:34:33Zagorthen it's another error
17:34:52merwin-Zagor: once i used the !, the function works beautifully
17:34:58Zagortry the parser code in a standalone C program. debugging without debugger is no fun...
17:36:23 Quit sylvr (Remote closed the connection)
17:36:25merwin-Zagor: if strcasecmp returns 0, then why is this written like:
17:36:25merwin- else if (!strcasecmp(&entry->d_name[len-3], ".eq"))
17:36:26merwin- dptr->attr |= TREE_ATTR_EQ;
17:36:48Zagorlooks good
17:45:18merwin-does atoi handle negatives?
17:45:39Zagorumm, i don't think so
17:45:51merwin-it didn't seem to work :)
17:46:19merwin-So i'd have to see if the first char is a - and then do atoi(buf_val++), right?
17:46:32merwin-and then multiply by -1
17:48:23Zagornot ++, +1
17:48:40Zagor++ would permanently move the pointer
17:48:41 Join sylvr [0] (
17:48:52Zagori have to go. see you later!
17:48:57 Nick Zagor is now known as Zagor|away (
17:59:46 Join mecraw12 [0] (~lmarlow@
17:59:46 Quit mecraw (Read error: 104 (Connection reset by peer))
18:07:55 Join Jojoba [0] (
18:08:17Jojobahi all, i just got my Jukebox 6000 :)
18:08:41Jojobaanyone know how long it has to charge before you can start it up?
18:16:08merwin-Jojoba: technically, about 30 minutes.... if you want good battery life, i'd charge it for a good 6-8 hours before using it the first time
18:16:15merwin-maybe even 10 hours
18:16:54merwin-Jojoba: in the long run though, personally i don't think it'll make much difference which one you choose to do :)
18:20:16 Quit pimlottc (Remote closed the connection)
18:24:37Jojobathanks, i've charged it for a good 2 or 3 hours but it still won't come on so far, teh manual said it only needed 1 hour before you could switch it on...
18:25:06quelsarukwell.. i think that is false
18:25:45quelsaruki mean, you can charge it for hours but you can't turn it on if it's charging
18:26:03quelsarukat least i can'ty
18:26:29Jojobaso it can't be on when plugged in the mains?
18:26:42***Saving seen data "./dancer.seen"
18:26:57merwin-Jojoba: I can turn mine on when it's charging :P try taking the charger out and turning it on
18:27:12Jojobajust switches off then :(
18:27:29Jojobaprobably needs more juice
18:27:36quelsarukmerwin-: can you turn it on while charging?
18:27:39 Quit Snorlax ("gittar ny!")
18:28:30merwin-quelsaruk: yup
18:28:39merwin-Jojoba: you need to hold the ON button for about 5 seconds to turn it on
18:29:01Jojobaah, ta, silly me ;)
18:29:48quelsarukmerwin-: as i said , i can't, maybe it's my rom firmware. I have a recorder with 1.17i at rom :(
18:30:12merwin-quelsaruk: hmm.. that's REALLY weird. i haven't heard of that before... it must be an old one or something?
18:30:26quelsarukit's old
18:30:26CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:30:26*merwin- needs to take a shower :P I stink
18:30:33 Nick merwin- is now known as merwin|shower (
18:30:33DBUGEnqueued KICK merwin|shower
18:30:45merwin|showerI'll be back in a bit... gotta take a shower before my 6 hour drive
18:31:36Jojobaeuh, it has a fujitsu drive :(
18:34:28 Join LinusN [0] (
18:34:32 Part LinusN
18:41:18merwin|showerJojoba: yup... my last on did too... my current one has a hitachi
18:41:39 Nick merwin|shower is now known as merwin (
18:41:39DBUGEnqueued KICK merwin
18:41:46 Nick merwin is now known as merwin|towel (
18:41:46DBUGEnqueued KICK merwin|towel
18:43:49Jojobanicer, i've had some bad experiences with fujitsu drives :(
18:44:47Jojobahmm, says can't support long filenames?
18:45:42 Quit Jojoba ()
18:56:55 Quit quelsaruk ("KVIrc 3.0.0-beta1 "Eve's Avatar"")
19:14:52 Join Bagder [0] (
19:19:07idefxno one here would like to implement boyer-moore algorithme. I'm tired, i don't fell like doing this ;'(
19:22:01merwin|towelwhat is boyer-moore?
19:22:09 Nick merwin|towel is now known as merwin|notowel (
19:22:09DBUGEnqueued KICK merwin|notowel
19:22:13Bagdera search algorithm
19:22:41Bagderclever one in fact
19:23:21idefxbut i haven't play the violin the whole day, this is bad ! so i stop programming for the day, Na ! ;-p
19:23:41merwin|notowelFor some reason, this is evaluating as being a true statement when "bass" = buf_set ... zagor says that's not correct
19:23:42merwin|notowel if (!strcasecmp(buf_set,"bass")) {
19:24:13Bagderit is correct
19:24:22Bagderstrcasecmp() returns 0 on a match
19:24:31Bagderso !0 is true
19:25:16merwin|notowelmaybe i just misunderstood what zagor said :)
19:25:56merwin|notowelBagder: i'm about 2 hours away from getting loadable eq done... well, actually about 8 hours since i have a 6 hour drive ahead of me
19:26:16merwin|notowelBasically, all i need to do is figure out mpeg_val2phys
19:26:47merwin|notowelThe config file is easy:
19:26:47merwin|notowel[volume]20#(0 to 100)
19:26:47merwin|notowel[bass]11#(-15 to 15)
19:26:47merwin|notowel[treble]14#(-15 to 15)
19:26:47***Alert Mode level 1
19:26:47merwin|notowel[balance]25#(-100 to 100)
19:27:39 Nick merwin|notowel is now known as merwin (
19:27:39DBUGEnqueued KICK merwin
19:27:39***Alert Mode level 2
19:30:47CtcpIgnored 5 channel CTCP requests in 57 minutes and 6 seconds at the last flood
19:30:47*merwin should stop coding and start packing
19:31:34merwinheh, deposits from my work show up as:
19:31:35merwinACH dep: TMOBILE-X
19:37:37merwingotta split... ttyl
19:37:40***Alert Mode OFF
19:37:43 Quit merwin ()
19:50:03 Join qulesaruk [0] (
19:52:39qulesarukbadger, just one thing
19:53:04qulesarukwhere's mpeg_queue defined?
19:53:10qulesaruki can't find it
19:54:12Bagderit's a private queue in the mpeg thread
19:56:10 Quit mecraw12 (Read error: 110 (Connection timed out))
19:57:56 Join lithermon [0] (
19:58:25 Join datazone-work [0] (
19:59:48 Quit qulesaruk (Read error: 54 (Connection reset by peer))
20:00:01 Join qulesaruk [0] (
20:01:03Bagdermuch nicer credits screen ;-)
20:01:39qulesarukkvirc crashed :)
20:12:54adi|homei think the scroll is a bit slow
20:13:27Bagderyes it might be
20:14:05adi|homebe nice if it was based off of a factor of the scoll setting ;)
20:14:33Bagderwell, its a completely different scroll so I don't think that would be very good
20:17:28adi|homeknow what... its MUCH faster scrolling onthe player..
20:17:34adi|homecompared to the sim
20:18:15BagderI noticed that too
20:20:47Bagdertime to go
20:20:49 Part Bagder
20:25:13 Quit qulesaruk (Read error: 54 (Connection reset by peer))
20:25:25 Join qulesaruk [0] (
20:26:44***Saving seen data "./dancer.seen"
20:42:43elinenbehello. back from skating today. it locked up on one song today. quite odd.
20:49:29 Quit Low[a] ("SpermaŁbertragung unterbrochen (broken pipe)")
20:55:35 Part alkorr
21:05:17 Join GomGom [0] (
21:05:45GomGomback to ask for the nice perl changelog generator :)
21:06:23GomGomwell zagor is not there
21:07:07 Quit idefx (Read error: 104 (Connection reset by peer))
21:09:13 Part elinenbe
21:15:01 Nick seb-school is now known as _seb_ (
21:16:01 Join LinusN [0] (
21:16:07 Part LinusN
21:26:35 Part qulesaruk (""Eve's Avatar - Avatar's Eve"")
21:35:04 Quit lithermon (Read error: 60 (Operation timed out))
21:44:44 Join lithermon [0] (
21:45:43 Join ironi [0] (
21:49:22 Part ironi
21:50:39 Nick Zagor|away is now known as Zagor (
21:50:47ZagorGomGom: where do I send it?
21:51:41GomGomi had no changelog generator in debian pools
21:52:01GomGomand as i'm using your firmware for more than 3months now
21:52:17GomGomi had noticed you were generating the logs :)
22:13:14 Quit GomGom (Read error: 104 (Connection reset by peer))
22:13:33 Join GomGom [0] (
22:13:55 Quit Zagor ("Client Exiting")
22:14:14 Quit lithermon ()
22:26:47***Saving seen data "./dancer.seen"
22:34:15 Nick edx is now known as edx|sleep (
22:47:08 Nick dwihno|gone is now known as dwihno (dwihno@Bald067.Baldakinen.Umea.SE)
22:50:43dwihnoI am dwihno
22:50:48dwihnoYou are you!
23:06:33 Join elinenbe [0] (
23:38:19 Quit sylvr (Remote closed the connection)
23:38:47 Join Jet8810 [0] (
23:41:35 Join hardeep [0] (

Previous day | Next day