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 2004-06-14

00:01:02***Saving seen data "./dancer.seen"
00:05:06 Quit lImbus (" The IRC Client of the Gods! -> <- HydraIRC")
00:26:01 Part amiconn
00:48:57 Join NibbIer [0] (
01:15:17 Join pfavr [0] (
01:15:17 Quit NibbIer (Read error: 54 (Connection reset by peer))
01:16:55 Quit pfavr (Client Quit)
02:01:05***Saving seen data "./dancer.seen"
02:52:01 Join NibbIer [0] (
03:19:46 Quit NibbIer (Read error: 54 (Connection reset by peer))
03:35:56 Part xam
04:01:08***Saving seen data "./dancer.seen"
04:52:15 Join NibbIer [0] (
05:36:08 Quit NibbIer (Read error: 54 (Connection reset by peer))
06:01:12***Saving seen data "./dancer.seen"
06:31:07 Join LinusN [200] (
07:08:22 Quit midk ("sudden death, yo?")
07:17:15 Join midk [0] (
07:17:21 Join NibbIer [0] (
07:38:03 Quit NibbIer (Read error: 54 (Connection reset by peer))
07:51:21BC|codingYou're too damn efficient Linus ;)
07:54:04 Join oxygen77 [0] (
07:56:22 Part oxygen77 ("Leaving")
08:01:16***Saving seen data "./dancer.seen"
08:13:20 Join AciD [0] (
08:18:57LinusNhi BC|coding
08:23:21 Quit AciD (Read error: 104 (Connection reset by peer))
08:25:10 Join [IDC]Dragon [0] (
08:26:26[IDC]DragonHi LinusN, BC!
08:26:44LinusNhola [IDC]Dragon
08:27:11[IDC]DragonLinusN: breaking the VC sim was no fun :-(
08:28:10*[IDC]Dragon spent the avail hors on sunday getting this to work
08:28:24LinusNbut you managed to get it working?
08:28:31[IDC]Dragonby finally walking back in cvs
08:29:05[IDC]Dragonand got no settings code done, like I originally planned
08:29:37LinusNwhy are you forced to use vc++?
08:30:12[IDC]Dragonbecause that's the tool I'm living in, and I wanted the debugger
08:31:29[IDC]Dragondoes cygwin come with a debugger worth the name?
08:31:48midk*worth the name?
08:32:12[IDC]Dragongraphical, in the editor, variable contents popping up while hovering the mouse, etc. ?
08:32:58*[IDC]Dragon is spoiled
08:33:15LinusNyou are trapped in IDE hell, my friend
08:33:34[IDC]Dragonveryyyy comfy
08:35:14BC|codinghi [IDC]Dragon (back from making a cuppa)
08:36:05[IDC]DragonLinusN: you're trapped in command line hell ;-)
08:36:29*[IDC]Dragon not starting any flamewar, nono
08:44:34BC|codinglinus, I see you have been working on getting the sim back up to speed you know if we can expect libplugin support soon?
08:46:59LinusNi hope so
08:48:23BC|codingi have a real neat quick bodge in the mean time the bottom of lib.h have #ifdef SIM #include "lib.c"
08:49:29BC|codingsorry lib.c as in the counterpart to lib.h
08:49:49BC|codingeg. grey.h would include grey.c
08:49:54LinusNi get it
08:50:07BC|codingit a bodge, but it gets the sim up and running :)
08:50:14hardeep[IDC]Dragon: I've got a fix for getting vc++ working again
08:50:47hardeepLinusN: the changes add a couple of #ifdef's for vc++ in the common file code, is that okay?
08:50:59BC|codingonly problem is that plugins do not "update dependancies" so if you update a "lib" then you need to TOUCH all plugins that use it
08:51:05hardeep(just an include path and a typedef)
08:53:44[IDC]Dragonhardeep: great!
08:53:49LinusNhardeep: that's ok
08:54:02[IDC]DragonI had no time to look in the details of it
08:54:28[IDC]DragonSo I rolled back, hacked plugin.c prefixing
08:58:16LinusN[IDC]Dragon: if you want a graphical gdb in cygwin, try starting insight.exe instead of gdb.exe
08:59:45[IDC]Dragonah, that looks like the SH gdb
09:00:56[IDC]Dragonhave to remember that one
09:01:03LinusNunfortunately, the debug output ends up in silly dialogs
09:01:10hardeepVC++ is so much nicer to use. :)
09:01:21LinusNi agree that the vc++ debugger is nice
09:01:35LinusNbut i don't have vc++
09:01:35BC|codinganyone know about "zip: not found" error in "make zip" ?
09:01:38[IDC]Dragonlightyears ahead
09:01:51LinusNBC|coding: install zip
09:03:15LinusNit's under "Archive" in the cygwin installer
09:03:34BC|codingah "install" is a cygwin thing?
09:04:20hardeep[IDC]Dragon: changes committed. There's one problem: the vc++ compiler doesn't like empty structs and there's one in the settings_menu. I haven't figured out a nice fix so I just hack in something for now
09:04:54LinusNBC|coding: i guess your devkit should include the zip package as well
09:05:12[IDC]Dragonhardeep: I have a fix for that one, could commit it
09:05:36hardeepplease do, it's annoying. :)
09:06:06[IDC]Dragon#ifndef SIMULATOR
09:06:06[IDC]Dragon { STR(LANG_BATTERY_CAPACITY), battery_capacity },
09:06:06DBUGEnqueued KICK [IDC]Dragon
09:06:06[IDC]Dragon#ifndef HAVE_CHARGE_CTRL
09:06:06[IDC]Dragon { "Dummy", -1, NULL }, /* to have an entry at all */
09:06:06***Alert Mode level 1
09:06:40BC|codingLinusN: yes, that's precisely what I am checking right now ....i need to sort out what files I need ...problem is all this type of work was origianlly done by devZer0 ...and he seems to be doing other stuff now
09:07:06BC|codingall I did with the devkit was package it nicely
09:07:54LinusNmy guess is that all you need is the zip.exe file
09:08:29BC|codingI will try that - thanks :-)
09:08:37midkyo bcbc
09:08:44BC|codingmk, ur up l8
09:09:52midki'll be up a few more hours don't worry
09:10:04BC|codingno skool?
09:10:38midkno, it's still in effect
09:10:58midkwow. i have to get up in 5.5 hours if i go to bed right now
09:16:03BC|codinglinus - you're a guru ;) zip.exe :) Thanks
09:16:07***Alert Mode OFF
09:17:35 Join NibbIer [0] (
09:32:37[IDC]Dragoncu later
09:32:41 Quit [IDC]Dragon ()
09:36:59 Join Zagor_ [242] (
09:37:04 Nick Zagor_ is now known as Zagor (
09:53:11midkhi zagor
09:53:11 Quit NibbIer (Read error: 54 (Connection reset by peer))
09:53:40 Part LinusN
09:53:47midkbye linus
09:59:12BC|codingmk, have you scrapped the idea for breakout, or just put it on hold?
09:59:51midkhold, still need to do collision detection
09:59:58midkhaven't done any rockbox work lately
10:00:23BC|codingdid you put all your blocks in an array in the end - if so collision detection should be quite straight forward
10:01:18***Saving seen data "./dancer.seen"
10:01:26BC|codingrighty, the arrat will also allow special featur blocks for the future ;)
10:01:46midklet's do!
10:01:51 Join LinusN [200] (
10:08:09midkyay linus
10:12:34 Quit hardeep ("Killed by frash (Requested by panasync)")
10:19:30 Join [IDC]Dragon [0] (
10:20:03[IDC]DragonHi everybody! [IDC]Dragon is back! ;-)
10:20:19BC|coding<screams and cheers from the auditorium>
10:20:20midkhey idc
10:20:32midk*contributes a small squeak
10:20:46midkoh boy
10:20:52midki'll get about 4 hours of sleep
10:21:30BC|codingthe human sleep pattern is generally around 90mins
10:21:38midkBC|coding, i need your advice
10:21:45BC|codinguh huh
10:21:58midkit says here "essay: due by tuesday june 15, 2004"
10:22:08midkdoes that mean monday (BY tuesday) or ON tuesday morning
10:22:56BC|codingwe used to get a time as well at uni "by 16:00 monday" from that I would say "by tues" means "tues at the latest"
10:23:14midkby tuesday at 7:45
10:23:17BC|codingfor the teacher to mark tues eve
10:23:29midkso if i don't bring it today, monday, will i get in trouble
10:23:33 Join AciD [0] (
10:24:38BC|codingI dont know how "by the rules" your guys are or how much favour you carry with them ...but in my books if it says "by 07:45 tues" is means that if I hand it in at "07:44:59 on tuesday" everybody should be happy
10:25:09midklol mmkay thankies
10:25:40midkif he gives me trouble i will tell him that it's ok tu turn it in 7:44:59.99 tuesday
10:25:59midkalthough *i* can only go as precise as about "before 8am tuesday"
10:26:03midkand hope for the best
10:26:29BC|codingless inflamatory might be "I promise it will be with you by 07:45 tuesday as you requested"
10:26:55midk*doesn't get it in before then
10:27:01midk*gets f*
10:27:22BC|codingyeah, then be true to your word - that's how respect is earned ;)
10:28:00midki don't care if he respects me −− he hates me
10:28:22midk*eggs his car
10:28:36BC|codingit's all too easy to hate back, don't let him control you like that :)
10:28:50BC|codingahh, right, so you started it, so to speak , hmmmm
10:29:19midkstarted it?
10:29:27midki did not
10:30:23midkyay, a service tech comes tomorrow to supposedly fix midky's 30kb/s internet
10:30:46midk*waits for 34mb download to finish in.. 30 minutes
10:34:26 Quit [IDC]Dragon ("CGI:IRC (EOF)")
10:35:01 Join kurzhaarrocker [0] (
10:35:19midkonly kurz.
10:35:29midkhaven't seen you in a long time
10:35:32midk:) who were you again?
10:35:42kurzhaarrockerI was langhaarrocker :)
10:35:57midkwere you bothersome?
10:36:35midksmart? aggravating? terrifying For So Many Reasons?
10:36:43kurzhaarrockerI occasionally bothered some people...
10:37:29kurzhaarrockerZagor I assume you received my mail concerning the source of the split editor & co?
10:38:05Zagorstill going through the weekend mail, haven't seen it yet :)
10:38:20kurzhaarrockerok, keep wading
10:38:39midkOH YEAH
10:38:46midksplit editor.
10:38:58midk you did the oscillograph too right
10:39:10midkI KNEW IT
10:39:45kurzhaarrockerand the peak meter and wormlet and I was working on volume triggered recording.
10:40:22midkyou make peek meeter!
10:45:28kurzhaarrockerWhat's DH ?
10:45:34midkdeh deh!
10:45:56kurzhaarrocker<- doesn't know deh deh either.
10:46:09midkTEH TEH
10:52:44midkthe peak meter.
10:53:07 Nick midk is now known as midk|zzz (
10:53:15 Nick midk|zzz is now known as midk|zZz (
10:53:20CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
10:53:20*kurzhaarrocker is confused
10:53:20kurzhaarrockerWhat about the peak meter?
10:53:26midk|zZzyou made it?
10:53:30midk|zZzthe rockbox peak meter.
10:53:37midk|zZzomfgz muchly loves it!
10:53:59kurzhaarrocker:) me too
10:54:30midk|zZzi especially like the decay settings
10:54:35midk|zZzit is quite nice at ie 4 or 6
10:54:49kurzhaarrockerwell, I only wished it was more precise. Doesn't really help much when recording.
10:54:50midk|zZzsort of pointless suggestion: decay amount and decay speed
10:57:15kurzhaarrockerThat is related to each other as the decay speed is defined by the "display framerate" - which is fixed at 20 (?) fps.
10:57:42midk|zZzi once made a patch that let you update the peakmeter's framerate instead of a "performance" option
10:57:54midk|zZzi prefer framerate to performance anyday
10:58:50kurzhaarrockerThat doesn't help much because the lcd speed limits.
10:59:18kurzhaarrockerIncreasing the framerate doesn't make the thing more precise.
10:59:44kurzhaarrockerMore readouts / time makes the peak meter more precise and that is what the performance is about.
11:02:46midk|zZzyes, it just looks smoother at IE 30
11:03:09midk|zZzshould be 20fps save energy, 30fps high quality
11:03:33kurzhaarrocker<- must work again, see you
11:03:43 Quit kurzhaarrocker (Remote closed the connection)
11:13:05 Quit midk|zZz ("sudden death, yo?")
11:15:57 Join midk [0] (
11:16:16 Nick midk is now known as midk|zZz (
11:25:39 Join Bagder [241] (
11:25:56Bagderoh, food time coming up
11:27:36c0uttaguys, has there been much talk about "autoexec" in the last few days (i've been on holiday)
11:28:02BC|codingi think a more generic solution is looming post ml chats
11:29:00c0uttaok. my work related to f2/f3 enables an autoexec function as well
11:37:10 Join Nibbler [0] (
11:46:54BC|codinghandy - are you coordinating with pfavr?
11:47:37BC|codingzagor mentioned the idea of event-driven, such an idea would allow things like audio-3587 to redirect the sound controls to a plugin
11:54:06c0uttabc - no, my work is not that sophisticated. it simply enables actions to be executed depending upon different contexts
11:54:27c0uttamainly in response to user interaction
11:55:05BC|codingcould a user-interaction not constitute a context under which a plugin were run, much like autoexec?
11:55:51c0uttaprobably, but my brain is not capable of thinking that far outside my home-made square
11:56:41BC|codingeach time the user does a thing event(EVENT_ID) ...then let bool event(int id) do all the processing?
12:01:19***Saving seen data "./dancer.seen"
12:04:55c0uttayes, that's the general functionality. but how does this help "audio-3587 to redirect the sound controls to a plugin"
12:10:48 Join lImbus [0] (
12:13:52lImbusBagder, about the on HyperThreading-Processors: Are there any recommendations about the number of crunching threads ?
12:23:43Bagderno idea
12:24:08Zagori guess two would be optimal
12:24:32*Bagder spots a newcomer in the team!
12:25:28lImbusif I go for 1 cruncher, it looks like the processor it not utilized for 100%, If I got for 2 crunchers, the systray says I'm having 2,2 Mkeys/sec
12:27:03Bagderbtw, I guess you noticed that I added a subtitle on the docs page
12:31:10lImbusoff for lunch
12:31:14BC|codingc0utta: if the event handler is able to load a plugin, then it can call audio-3587?
12:31:42c0uttayes, i have plugins loading
12:32:55BC|codingthen that is how it solves the problem :)
12:33:12LinusNexactly, no need to "hijack" the sound settings
12:34:07BC|codingno, the originals can stay in tact for those who would rather simplicity over controlability
12:35:02 Join pfavr [0] (
12:36:15c0uttalinus, did you do the vbrfix plugin ?
12:54:08c0uttaok, got it in my cvs sync
12:56:35Bagderhey, the new grand challange is announced
12:56:41Bagdershould we make a rockbox-vehicle? ;-P
12:57:59Bagderits about making autonomous driving vehicles across the desert
13:58:27 Quit lImbus (Remote closed the connection)
14:01:22***Saving seen data "./dancer.seen"
14:04:14 Join lImbus [0] (
14:21:13 Quit pfavr ("ChatZilla 0.9.52B [Mozilla rv:1.6/1]")
14:58:09 Join [IDC]Dragon [0] (
14:58:35[IDC]DragonLinusN: do you read?
14:59:22*Bagder pokes LinusN with a stick
15:00:05[IDC]DragonI'm working on a new settings code, found that global_settings.rec_timesplit is not persisted, is that intentional?
15:01:45[IDC]DragonWith .cfg files, it's even more strange: it can get loaded, but is not saved.
15:02:25Zagori think there are more recording options that are not persistant
15:03:04[IDC]Dragonspeaking about the bits in the sector, this is the only one I found
15:03:39[IDC]Dragon(just made a sanity check for my code)
15:08:26[IDC]DragonI pack the bits now, we can fit almost all values in the RTC RAM
15:09:19[IDC]Dragonand it's table-driven, no more pages of code for the persisting
15:09:50Zagorwe have discussed for a long time to move settings out of rtc and into a config file
15:10:05Zagorall except the very frequently updated ones, such as resume position
15:10:13[IDC]Dragontable is also good for this
15:10:25LinusN[IDC]Dragon: here now
15:10:56*Bagder got all tired of that poking
15:11:02LinusNthe rec_timesplit isn't persistent, and it's intentional
15:11:57[IDC]Dragonbut you can load it from a .cfg file, also intentional, I guess?
15:12:26LinusNbut i can't remember why we didn't make it persistent...
15:13:40[IDC]Dragonsuch a special case is no good for a table...
15:14:03LinusNi think we can make it persistent
15:15:26[IDC]DragonWe will have 2 tables then, one for the RTC bits, one for the bits spilled into the HD sector.
15:15:51LinusNsounds good
15:16:33[IDC]DragonI can fit most in the first, was thinking if I place the once causing a HD spinup anyway while changed into the second
15:17:15LinusNalso remember that players don't have RTC
15:17:44[IDC]Dragonto minimize the effect of a setting being unchanged if the disk wasn't spinning before powering off
15:18:48[IDC]Dragonyes, sure.
15:22:34midk|zZz"rockblox" lol
15:23:36 Quit [IDC]Dragon ("no fate but what we make")
15:23:36BC|codingyes, I thought that was a funny choice (funny ha ha, not funny perculiar)
15:23:46 Join [IDC]Dragon [0] (
15:24:17BagderI think its funny in the proper nerdy way
15:24:27BC|codingyeah :)
15:26:08[IDC]Dragonfunny what?
15:26:33BC|coding"witty" is the English word, but I don't know if it translates
15:26:57midk|zZz*still votes for "magical falling blocks"
15:27:29BC|codingWitty: Quick to discern and express amusing insights or relationships.
15:27:42midk|zZzi know what witty means.
15:27:57BC|codingI didn't, I just knew when to use it
15:28:25BagderIt could also be named "falling rocks within a box" ;-)
15:28:44Zagorsounds by that like only a person can be witty, not a story/wording
15:29:52BC|codingit is also defined as: Possessing or demonstrating wit in speech or writing
15:30:12Zagorsounds like a person too :)
15:30:28Zagorbut I know, it's not only used for persons
15:30:34BC|codingargh - wrong one ....defintion 2: Characterized by or having the nature of wit
15:30:49BC|codingsorry pasted the wrong bit
15:36:30BC|codingI hate tautological definitions ...recursive: see recursive
15:37:59midk|zZzfalling rocks within a boix
15:38:01midk|zZzDO IT DO IT
15:38:23 Nick midk|zZz is now known as midk|gone (
15:45:51LinusNtime to go, cu guys
15:45:54 Part LinusN
16:01:27***Saving seen data "./dancer.seen"
16:17:35 Quit [IDC]Dragon ("no fate but what we make (EOF)")
16:32:00BC|codingBagder, quick Q ...does the viewers support multiple plugins per file extension?
16:39:29lImbuswhat about the Open with... ?
16:39:55Bagderyes, that's the way to allow other plugins to run the specific file
16:40:08Bagderbut there's only one default
16:40:53BC|codingnot that it is relevant for me now, would it be an option to goto open_with, if multiple rocks are registered to a filetype?
16:41:21Bagderhm, possibly
16:41:32Bagderbut I'm not sure it would be good
16:42:59BC|codingone example I can think of is if you had a jpg->rif (rockbox image format) then you might like to choose whether to view or convert, but the user could disable it easily enough by editing the config file
16:45:02BC|codingif you do find yourself in filetypes.c, you may like to consider that because the sim holds all of its rocks in one dir, the #define's for maxstr and maxfiles are too small
16:45:39 Join mecraw [0] (~mecraw@
16:46:55Bagderthe sim should make use of the viewers dir too
16:47:52BC|codingThere are a number of sim improvements that could be made ;)
16:48:06Bagderwe made some the other day
16:48:11Bagderin Linus' commit
16:48:17BC|codingYes, I saw :)
16:50:43Bagdera 'make install' for the sim would make sense
16:51:08*Bagder starts working on one
17:00:31Zagori think it would only be annoying to *always* have to select which plugin you want to use
17:00:56Zagor(always = if there are >1 for the extension)
17:01:36BC|codingbut with the config file, the power has been given to the user :)
17:01:46BC|codingnice work there btw
17:02:05Bagder(make zip && cd archos && unzip ../
17:02:14Bagderkinda handy
17:03:00ZagorBC|coding: yeah, but I'd first like to see 5 users actually want the power to be annoyed like that :)
17:03:41BC|codinglol - I'm not sure I can think of a better solution :(
17:05:10Zagorthere already is a solution: open with
17:06:09Bagdernow there's a "make install" for the simulator
17:06:28BC|codingdoes it set up the whole dir structure?
17:07:04BC|codingcool ...does "make all" put all the files in the right place, or is that only done at "make install" time
17:07:21Bagderonly with make install
17:07:42Bagderyou can make that 'make all install' to make it happen at once
17:07:54BC|codingbit of a problem that
17:08:14BC|codingmeans that during testing I need one compile and for a distro I need another
17:08:57Bagderwhy is that a problem?
17:09:12Bagderyou already needed a different make to build a package for target
17:09:53BC|codingdifferent code, not just a re-make
17:10:08Bagderwhy would the code be different?
17:10:25Bagderthat would be a bug in the sim, wouldn't it?
17:10:25BC|coding#ifsim use this dir #ifsef install use this dir #ifdef build use this dir
17:10:37Bagderthat's bad code
17:11:09Bagderthe sim should hide the sim-specific path-fiddling
17:11:20BC|codingdepends on your preferred coding style it makes it easier for me, as I do not need to touch core code that way
17:11:37Bagderyou shouldn't need to change path in the sim
17:11:56Bagdercan you show a specific example when you need to?
17:12:12BC|codingnot tried a compile since the new mods
17:12:23BC|codingbut I could show you enough before then
17:12:27Bagderthe win32 sim was all messed up before
17:12:39BC|codingand that's being nice
17:12:57BC|codinganythoughts of libplugin support?
17:13:12Bagdersupport for what?
17:13:28BC|codingMy libs - what I am about to release
17:14:07BC|codingI've got my rnd() one and my gfx() pretty much ready to roll
17:14:41BC|codinghopefully cards() won't be that far behind
17:15:29BC|codingcards() is mainly on hold because I needed to write gfx()
17:15:52BC|codingno rush, as I say the work around is uber-simple
17:16:09BC|codingjust one minor annoyance is that plugins do not have dependants
17:16:19BC|codingso if you update the lib, you need to touch the plugin
17:16:20Zagorwe have lib support in plugins already
17:16:35Bagderah, right
17:16:39BC|codingnot in sim, unless it is super-recent
17:16:47Bagderour depend-stuff is actually rather lame in general imho
17:17:05BC|codingI'd love to be able to have things like othelo.h
17:18:35BagderZagor: why isn't the plugins using the same dep stuff as the rest?
17:18:40Bagderyou remember?
17:18:51Zagorsimple laziness, i think
17:22:59Zagorplugins couldn't have dependencies in the beginning, so no more was needed
17:23:05Bagderme go home
17:23:08 Part Bagder
17:23:55ZagorBC|coding: you can. check out apps/plugins/lib
17:24:21*BC|coding looks a little embarrassed
17:24:31BC|codingI have no idea how to use cvs - tried a few times and gave up
17:24:50Zagorthat's because you are poisoned by windows ;)
17:25:03BC|codingactually, I'm a dos man
17:25:27Zagorit's a lot easier nowadays since we switched off of sourceforge though. you should have another go, methinks.
17:25:32BC|codingas far as dev goes, windows is just a dos multi-session tool for me
17:26:15BC|codingI wish I could get all my favourite apps for linux - but I've ust gotten tooo attached to things like trillian and ultra-edit
17:28:06Zagorgive emacs a chance and you'll be hooked in no time
17:28:15 Quit lImbus ()
17:28:25Zagoractually make that "a little time" ;)
17:28:59BC|codingmaybe one day when Im feeling masochistic again
17:29:43BC|codingi survived pine - only to see one guy pick his damn email up with emacs - LMAO
17:30:12Zagoryeah. there's not much emacs can't do
17:30:48BC|codingi think it's mainly the pain of having to learn it all, all over again that builds the most resistance
17:32:02BC|codingis eudora available for linux?
17:32:09BC|codingI know opera is :)
17:32:53Zagori don't know about eudora. opera is the only commercial program i use in linux.
17:33:46Zagoron the other hand I have never even *seen* a commercial mail reader that does not make me scream in pain after ten minutes...
18:01:29***Saving seen data "./dancer.seen"
18:06:22Zagorgoing home
18:06:25 Part Zagor
18:09:57BC|codingsorry I missed you -phone call
18:29:23BC|codingl8rz guyz
18:29:27 Part BC|coding
18:49:19 Join amiconn [0] (
18:49:19 Quit Nibbler (Read error: 54 (Connection reset by peer))
18:55:31 Quit AciD (
18:55:31 Quit Hadaka (
18:55:31 Quit mbr_ (
18:55:31 Quit ze (
18:55:31 Quit amiconn (
18:55:31 Quit Ka_ (
18:55:31 Quit dwihno (
18:55:31 Quit Hes (
18:55:31 Quit c0utta (
18:55:31 Quit silencer (
18:55:31 Quit julia (
18:57:10NJoinamiconn [0] (
18:57:10NJoinAciD [0] (
18:57:10NJoinKa_ [0] (
18:57:10NJoindwihno [0] (~dw@
18:57:10NJoinc0utta [0] (
18:57:10NJoinmbr_ [0] (
18:57:10NJoinze [20] (
18:57:10NJoinHadaka [0] (
18:57:10NJoinjulia [0] (
18:57:10NJoinsilencer [0] (
18:57:10NJoinHes [0] (
19:27:39 Quit midk|gone (Read error: 104 (Connection reset by peer))
19:27:44 Join midk [0] (
19:40:25 Quit midk (Read error: 104 (Connection reset by peer))
19:40:37 Join midk [0] (
19:44:39 Join Zagor [0] (
19:52:00 Join midknight2k3 [0] (
19:58:47 Quit midk (Read error: 60 (Operation timed out))
20:01:33***Saving seen data "./dancer.seen"
20:02:21 Join elinenbe [0] (
20:31:05 Quit AciD (Read error: 104 (Connection reset by peer))
20:33:20 Join Nibbler [0] (
21:11:21 Join AciD [0] (
21:44:51 Join rakaus [0] (Lmpbzktwn5@
21:45:16rakaushey all
21:46:11rakausi was just wondering if anyone knows if it's possible to change the LED's on the archos fm recorder
21:46:31rakauscuz i'm getting sick of the green ones, i want to know if there's anyway to remove them and replace them with red led's
21:52:22Zagoryes. it is possible
21:53:15rakausis it difficult to do so?
21:53:31Zagoryes, quite. you must be pretty good with the soldering pen.
21:54:05rakaushmm ok
21:54:10Zagor[idc]dragon posted some images to the mailing list a while back showing how he'd changed to white leds
21:54:30rakausok i'll check those out
21:54:33Zagorsomeone else changed to blue leds years ago, but I think white looked better
21:56:09rakausi'm lookin for it
22:01:37***Saving seen data "./dancer.seen"
22:04:15rakausyou wouldn't happen to know the rough date for it would you?
22:07:40rakausn/m i got it
22:36:32 Join Bagder [0] (
22:36:39 Join [IDC]Dragon [0] (
22:36:51Zagorhowdy there
22:37:02[IDC]Dragonhola Zagor
22:37:05amiconnfull house ;)
22:37:15[IDC]DragonHi Jens
22:37:26[IDC]DragonLinus is missing...
22:38:21amiconnMy box will soon get a larger hd (already ordered) :)
22:38:40[IDC]DragonI did the first part of the new settings code, a table for squeezing the bits to RTC
22:38:47[IDC]Dragonsaves about 1 KB
22:39:50amiconn[IDC]Dragon: I'm feeling a bit dumb just now (if somebody of you read the logs)
22:40:08[IDC]Dragonwe had a chat today
22:40:41amiconnMy idea of changing the bitswap algo and getting it faster this way is crap - as it seems I can't even read my own code correctly :(
22:40:45[IDC]Dragonthe RTC RAM content is currently "assembled" with a lot of code
22:41:15[IDC]DragonI've changed it into a table, and very few code
22:41:47[IDC]Dragonamiconn: I've missed the bitswap discussion
22:43:10amiconnSee log 09:38+, I explained it to Bluechip
22:43:38[IDC]Dragonbut you said it's "crap", should I bother then?
22:44:20amiconnAlthough my idea for a new bitswap algo will definitely work, and still save 100+ bytes of IRAM, it wouldn't be faster than the current implementation, but a bit slower (~16%)
22:44:54[IDC]Dragonwe're not short of IRAM currently
22:45:02[IDC]DragonI'd prefer the speed
22:45:07amiconnThe new algo would take 29 clock cycles for swapping 4 bytes, while the current one takes 25 cycles
22:45:14amiconn(in the main loop)
22:45:51amiconnThis is because I said it's crap, because yesterday I erred in that - thought the 25 cyles were for swapping 2 bytes
22:52:16[IDC]Dragondoes somebody know if and how the SH compiler pads structures?
22:52:46[IDC]Dragonwill a 32 bit value be 32 bit alined?
22:52:55BagderI'd guess so
22:53:11[IDC]Dragoncan it reorder struct members to minimize padding?
22:53:24BagderI don't think it does
22:53:33[IDC]Dragonso I have to do that...
22:53:38 Quit rakaus ()
22:55:24amiconn[IDC]Dragon: When you reorder struct members, cosider putting the most used members in the beginning (within reach of the direct offset addressing of the sh1) to speed up access
22:55:47elinenbeamiconn, [IDC]Dragon, Zagor, Bagder: anyone have anything exciting on the horizon for Rockbox?
22:56:08[IDC]Dragonif you call new settings code exciting...
22:56:23elinenbeI don't consider saving some IRAM or a quicker bitswap that exciting! :)
22:56:34elinenbe[IDC]Dragon: what is up with the settings code?
22:56:35Bagdera 'make install' for the sim then? ;-)
22:56:44*Bagder giggles
22:56:56elinenbeoh.. now that is exciting!
22:56:57[IDC]Dragonwe're really thrilling
22:57:53[IDC]Dragonelinenbe: the settings code is a pure internal change, saving some memory
22:58:22[IDC]Dragona user does't benefit from it
23:00:18elinenbeI was hoping to hear about the game boy emulator...
23:00:34[IDC]DragonZagor: I'm putting the .cfg strings in my table now, too
23:00:45elinenbeor an ID3 viewer plugin that supports jpegs in ID3v2 tags...
23:00:53[IDC]Dragonso we have them only once,and save another ton of code
23:01:06Bagder[IDC]Dragon: that sounds neat
23:01:16Bagderthat code has been ugly a long time
23:01:30[IDC]Dragonthe drawback is that .cfg files won't get internal comment blocks any more
23:01:46amiconnI had a look at the settings code once ...grr
23:02:24Zagorcomments shouldn't be hard to add, would they? just a few const char* in the table
23:03:06[IDC]Dragonbut then I'd need some strange escape table entries to control that
23:03:40Zagorok..? i guess I'll have to see your solution to understand that :)
23:05:52[IDC]Dragoneach value is represented by the following struct (excuse the flooding):
23:06:03[IDC]Dragon short bit_pos; /* bit position in RTC/sector */
23:06:03[IDC]Dragon unsigned char bit_size; /* how many bits within the bitfield */
23:06:03[IDC]Dragon bool is_signed; /* global_settings value is signed */
23:06:03DBUGEnqueued KICK [IDC]Dragon
23:06:03[IDC]Dragon short settings_offset; /* store position in global_settings struct */
23:06:03[IDC]Dragon unsigned char byte_size; /* how many bytes in the global_settings struct */
23:06:03***Alert Mode level 1
23:06:03[IDC]Dragon const char* cfg_name; /* variable name in a .cfg file */
23:06:05[IDC]Dragon const char* cfg_val; /* set of values, or NULL for a numerical value */
23:06:41Zagorso why not just add a const char* cfg_comment ?
23:06:45 Quit Hadaka (
23:06:45 Quit mbr_ (
23:06:45 Quit ze (
23:06:45 Quit Bagder (
23:06:45 Quit midknight2k3 (
23:06:53[IDC]Dragonthere is an array holding ~80 of those
23:07:22elinenbe[IDC]Dragon: how much code does the new settings stuff save?
23:07:46[IDC]DragonI'm careful not to introduce not too many "columns"
23:07:48NJoinHadaka [0] (
23:07:48NJoinze [20] (
23:07:48NJoinmbr_ [0] (
23:07:56NJoinmidknight2k3 [0] (
23:08:17[IDC]Dragona comment would be needed only by a few entries
23:08:38[IDC]Dragonactually, between the entries, to group them
23:08:56[IDC]Dragonthat's why I was thinking about "escape values"
23:09:02Zagorright. that's messy.
23:15:24NJoinBagder [0] (
23:15:39 Quit Bagder (Read error: 60 (Operation timed out))
23:16:04***Alert Mode OFF
23:17:56 Join Bagder [0] (
23:18:29 Join hardeep [0] (1098@
23:21:12 Quit [IDC]Dragon ()
23:21:19amiconnthat was fast
23:23:41hardeepmust've been really sleepy
23:24:30Zagorhas anyone tried creat()ing a file in the sims lately? it doesn't seem to work for me.
23:24:45CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:24:45*Bagder hasn't
23:26:35hardeepI'm assuming X11 sim?
23:26:44hardeeplet me try win32
23:30:24Zagorhmm, the O_RDONLY etc defines are taken from rockbox instead of the host OS. that breaks it royally.
23:31:07Zagorthis used to work...
23:31:38hardeephmmmm, it still works on the win32 sim
23:31:41Bagderits always good to have someone to blame
23:32:22ZagorO_CREAT is 0100 in linux. what is it in windows?
23:33:56hardeephmmm, the same
23:40:23hardeepoh, the code I was using was calling creat() directly.... let me try open with O_CREAT
23:41:04Zagorin the x11 sim at least, creat becomes sim_open. it never reaches sim_creat. i can't see where that happens.
23:41:28hardeepcreat becomes sim_creat
23:41:46hardeepwhen i stepped in it
23:42:21Zagortry a plugin
23:42:23Bagderhardeep: that's what linus commit introduced
23:42:55Bagdermaybe the plugins don't use the right function?
23:43:23Zagori checked that too. looks alright.
23:51:00Bagderit never includes any host include file with the right define
23:51:45Zagorthat's what it seems like
23:52:27Bagderits not that easy to add
23:57:56BagderI have fix pending now

Previous day | Next day