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

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

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

#rockbox log for 2005-07-03

00:00:05mufftiso i tried to install rockbox
00:00:21mufftibut now it shows everything on the display
00:00:31ansivirusby everything what do you mena?
00:00:35ansiviruserr mean
00:01:09mufftii need some time to write in english, sorry
00:01:10ansiviruslike dir listing etc.?
00:01:30mufftino, everything the display is able to display
00:01:42mufftiblack boxes instead of letters
00:02:02mufftievery pixel on the display is activatet...
00:02:04ansivirussounds like a flash update gone wrong to me but don't know
00:02:05Bagdermuffti: you _never_ see the jukebox text when it starts?
00:02:16Bagdermuffti: did you flash it?
00:02:40Bagderthen you _should_ see that text first
00:02:43mufftijust copied the .rockbox and the other file to the harddisk
00:03:02Bagdersince the text is written there by the built-in firmware before rockbox starts
00:03:22ansivirusyeah you're not ever getting ot the point of loading rockboxc
00:03:30Bagder(unless the player version functions differently than the recorder)
00:04:02Bagdermuffti: can you still use the USB when connected to your PC?
00:04:18mufftiwell, as i told you, i got the jukebox because my friend thought its dead
00:04:30mufftiyes, but only at win xp
00:04:44mufftii normaly use a apple powerbook
00:04:56Bagderand you have the proper drivers installed on the mac?
00:05:00mufftimac os x cannot find the juke box
00:05:13mufftino, i tried it on the win box
00:05:29Bagderthe studio 10 is not usb-storage compliant
00:05:33Bagderyou need special drivers
00:05:36mufftioh, ok
00:06:05Bagderpossibly win xp comes with them included
00:07:00ColdtoastLuthe rVandross died!
00:07:19ColdtoastLuther Vandros
00:07:20Moosnooooooo :(
00:07:30ansiviruswhen did that happen?
00:07:32mufftihm... there are no drivers for os x tiger
00:07:53Coldtoastsame age as my Dad
00:08:10ansiviruswhat did he die of? natural causes?
00:08:26Coldtoastafter suffering a major stroke
00:08:38Coldtoast2 years ago
00:08:48Coldtoastthen yesterday died in his sleep
00:09:10Coldtoastsays he'd never fully recovered from teh stroke
00:09:52ansivirusthat's too bad
00:10:05Coldtoastit is
00:10:05mufftihm... i think i'll tell my friend to buy an ipod
00:10:06ansivirustalented man.. will be missed by many
00:10:11Coldtoastso young too
00:10:20ColdtoastI mean, 54 isn't really too old
00:10:24ansivirusno it's not..
00:10:54Moosansivirus:respect of the talent man
00:15:14 Quit ansivirus (Read error: 131 (Connection reset by peer))
00:23:00 Quit muffti ("CGI:IRC (EOF)")
00:35:24 Quit webguest91 ("CGI:IRC")
00:45:01 Quit Coldtoast (Read error: 104 (Connection reset by peer))
00:56:22Moosgoodnight :)
00:56:38 Part Moos
00:57:10 Join tvelocity [0] (
01:04:27 Join TCK [0] (
01:13:20 Quit Bagder (Read error: 145 (Connection timed out))
01:22:57 Part Musicmad
01:29:59 Join ansivirus [0] (
01:38:53 Quit hicks ("Too lazy to change my quit message")
01:48:05***Saving seen data "./dancer.seen"
02:11:13 Join Omar [0] (
02:11:38 Nick Omar is now known as Odie (
02:27:04 Quit kenshin (Read error: 110 (Connection timed out))
02:54:13 Quit tvelocity ("Leaving")
03:06:13 Join ashridah [0] (
03:32:40 Join StrathAFK [0] (
03:38:38 Quit ghostiger (Remote closed the connection)
03:48:09***Saving seen data "./dancer.seen"
03:51:13 Quit Strath (Read error: 110 (Connection timed out))
03:57:54 Quit CheeseBurgerMan (Read error: 104 (Connection reset by peer))
03:59:45 Join CheeseBurgerMan [0] (~Dan@
04:05:05 Quit cYmen_ ("zZz")
04:05:43 Join QT_ [0] (as@area51.users.madwifi)
04:06:03 Quit QT (Read error: 131 (Connection reset by peer))
05:21:51 Quit pill (Read error: 145 (Connection timed out))
05:48:10***Saving seen data "./dancer.seen"
05:51:17 Nick StrathAFK is now known as Strath (
06:06:15 Join pill [0] (
06:19:05 Quit CheeseBurgerMan (Read error: 110 (Connection timed out))
06:39:08 Join Coldtoast [0] (
07:30:04 Quit pike (Read error: 110 (Connection timed out))
07:48:11***Saving seen data "./dancer.seen"
08:18:19 Join StrathAFK [0] (
08:36:38 Quit Strath (Read error: 110 (Connection timed out))
08:49:27 Join pike [0] (
08:56:27 Join matsl [0] (
09:12:11 Join Lost-ash [0] (
09:18:08 Nick QT_ is now known as QT (as@area51.users.madwifi)
09:18:16 Part stoffel
09:28:39 Quit ashridah (Read error: 113 (No route to host))
09:33:40 Nick Lost-ash is now known as ashridah (
09:48:14***Saving seen data "./dancer.seen"
10:18:10 Quit pike (Read error: 110 (Connection timed out))
10:18:23 Join Chamois [0] (
10:20:17 Quit Chamois (Client Quit)
10:33:13 Quit `nert|Uni (Read error: 104 (Connection reset by peer))
10:49:31 Join belgarath [0] (
10:49:49belgarathwould ir
10:49:53belgarathwould ot
10:51:03 Quit belgarath (Client Quit)
10:58:32*HCl wants to do something useful :/
10:58:40*HCl prods Slasheri and amiconn
11:04:02Coldtoastyaaay! got my PDA back
11:07:03 Join tvelocity [0] (
11:07:22ashridahport rockbox to it :)
11:07:42 Join einhirn [0] (
11:07:43 Quit einhirn (Read error: 131 (Connection reset by peer))
11:07:49 Join einhirn [0] (
11:08:21Coldtoastrockbox on an x50v
11:08:25Coldtoastimagine that
11:14:12 Join PaulJ [0] (
11:22:15*HCl has linux on his pda...
11:22:20HCli could probably just run the x11 sim on it
11:41:24 Join pike [0] (
11:41:33Coldtoastare they good? what are the specs?
11:42:05Coldtoastmy x50v would be SO great to run Linux on. 624MHz, VGA display
11:42:19Coldtoastintell accelerated gfx with 16MB
11:42:37 Join Moos [0] (
11:42:48MoosMorning all
11:42:51Coldtoastwifi and bluetooth, SD and CF slot
11:42:55Coldtoasthey Moos
11:48:15***Saving seen data "./dancer.seen"
11:55:25Slasheriiriver battery test (128k mp3, crossfade on) 14h with standard old battery and still going :)
11:56:06ColdtoastI'll do a test with my 2200mAH when I get it
11:56:08tvelocitySlasheri, volume?
11:56:20crwlwasn't that something around 15-16 hours with the original firmware?
11:56:40Coldtoastwhat data would be handy? maybe what voltage every 15mins?
11:56:45Slasherivolume 62, no headphones connected. I think this should not affect very much to the battery life
11:56:59ColdtoastI get just under 16hrs with my standard battery
11:56:59Slashericrwl: yep
11:57:41tvelocityhow many mAH is irivers standard battery?
11:58:38tvelocitycool, i think i should get one of those 2200mAH ones too...
11:58:54Coldtoastthey're really cheap
11:59:03Coldtoastget it off of ebay
11:59:38tvelocityyeah, but i'm not sure i'm able to replace it
11:59:53Coldtoaston ebay, search for "2200mah" then look for Ipod 1st gen ones
12:00:01crwli've also got one of those 2200 mAh batteries coming
12:00:09Coldtoastwell, 1st and 2nd gen
12:00:26Coldtoastsame as that?
12:01:35crwlColdtoast, nope, hmm...
12:01:49Coldtoastoff ebay?
12:02:30crwlthey're the same though, it seems
12:02:30Coldtoastsame battery
12:02:33Coldtoastdifferent seller
12:02:48Coldtoastyou only get 6months warranty tho
12:03:18crwlheh, well
12:04:25tvelocity2200mAH is the maximum available?
12:08:58Coldtoastlooks like it, yeah
12:10:33Coldtoastah cool! a guy I know from Weta is online. not seen him on in ages
12:10:42ColdtoastI imagine they keep him damn busy tho
12:11:31 Join ghostiger [0] (tor@07486633c5c23074.session.tor)
12:13:32tvelocityand how is it? heavier?
12:13:42Coldtoastdon't have mine yet
12:13:58ColdtoastHCl and markun have them tho
12:19:54 Quit tvelocity ("Leaving")
12:44:15 Join cYmen [0] (
12:47:56 Join hicks [0] (
12:49:42Slasheriiriver with rockbox 15h and still playing!
12:51:05Slasheriat least the battery performance is no worse than with iriver firmware :)
12:59:25Coldtoastyou recording the battery info so the battery remain time can be configured properly?
13:01:55Slasheriunfortunately no :/
13:02:20Slasheribut i think the battery meter should be quite accurate now. Only the remaining time has to be configured
13:02:21Coldtoastbummer. heh
13:02:41Coldtoastyeah. my remain time is way out. never seen it go >10hrs
13:02:55Coldtoastwhen it's fully charged it says something like "8hrs 53mins"
13:03:06Slasheriyep, mine too
13:03:19 Join belgarath [0] (
13:03:26Slasherii think we can get close to the 16 hours
13:03:37Slasheriand with a fresh battery maybe even more
13:03:56Coldtoastif I change the battery to 2000mAH tho, it says 16hrs or so I think
13:04:29Slasherihehe, yes. i think it should be easy to correct that reamining time meter
13:04:42belgarathis the estimated battery time still for archos?
13:04:53belgarathit sometimes seems quite innacurate
13:05:03Slasheriyes, it's not calibrated for iriver yet
13:05:40 Join Febs [0] (
13:05:54belgarathif it goes down to 0 minutes does the unit shut down even if there is several hours playback time left?
13:05:54ColdtoastSlasheri: so do you think pretty much all bugs are gone from crossfade?
13:06:12SlasheriColdtoast: should be, currently i don't know any bugs with crossfade :)
13:06:30Coldtoastcool. Mine's not crashed inteh past 24hrs or so
13:06:39Slasheriand it doesn't affect almost at all to the battery life
13:06:47Slasherigreat :)
13:06:51Coldtoastgood work!
13:07:10Coldtoastnice to see mono MP3s are sorted
13:07:41belgarathhmm i had a problem yesterday when i played a buggy m3u file, and my unit was on resume, so whenever i booted the player up it froze
13:07:45Coldtoasttho I get clicking with 2200KHz mono. that'd just be a resamplingthing tho wouldn't it?
13:07:51belgarathi was forced to boot into iriver firmware
13:08:13 Join stripwax [0] (
13:08:16belgarathis there a way to turn resume off without going into the settings (like when starting player up)
13:08:23Slasheriresampling is still a little buggy and that may crash the player
13:08:44Coldtoastahoy hoy
13:09:24belgarathi don't believe it - i am about to ship my h120 off to iriver america to fix a broken reset button, and i will be without it and rockbox for 3 weeks!
13:09:52Coldtoastdid you flash an unmodded firmware?
13:10:04ColdtoastI think I would
13:10:05belgarathi'm going to lol
13:10:11 Part stripwax
13:10:13belgarathi haven't sent it yet
13:10:21 Join stripwax [0] (
13:10:26belgarathi also formatted it so they can't steal my music...
13:10:26stripwaxhello again. ;-)
13:10:32Coldtoasthow'd reset get brken?
13:11:13ColdtoastI use a wooden toothpick on mine
13:11:24stripwaxi'm more a paperclip kinda guy
13:11:35Coldtoastwell, a paperclip is metal tho
13:11:45stripwax.. ?
13:11:50Coldtoastmetal and plastic don't get along that well
13:11:59Coldtoastit's easy to wreck plastic with metal
13:12:20Coldtoastdn't want to bust my reset button all to bits with a metal clip
13:13:03belgarathyeah i used a pin
13:13:12belgarathpaper clips work good tho
13:13:30stripwaxi put a hole thru my reset button with a pin :-( so that's why i now use a big fat paperclip instead
13:13:47belgarathi'd prefer it if they just fixed the reset button but it looks like i will get a replacement
13:14:03*belgarath prays i won't get an old refurbished model
13:14:19ColdtoastI used to use this pointy metal thing I have but one time it felt like I pierced the reset button
13:14:30Coldtoastso now I prefer the toothpick
13:14:46belgarath i wouldn't mind if they ran out of h120s and sent a new h320 but would probably prefer to have optical in/out
13:14:48Coldtoastman. what are the chances they still have a H120 lying around?
13:14:55stripwaxhere's a thought - could we have a rockbox option to autoreset on an illegal instruction/etc? rather than always need to reach for a toothpick/pin/clip which may not be available?
13:15:04belgarathooh good idea
13:15:24belgarathif you can implement that i won't send my player away lol
13:15:52Coldtoastyou could dismantle your player and fix it yourself belgarath :)]
13:15:59Coldtoastbut you'd void the warranty
13:16:18ColdtoastI think I still have 5 months left on mine
13:16:27stripwaxmine's long gone
13:16:43Coldtoastmaybe only 4 actually
13:16:48belgarathi have like a year and 5 monthes
13:16:55belgarathbecause europe has 2 years
13:16:59Coldtoastoh? nice!
13:17:08belgaratheven tho technically rockbox has voided it
13:17:09stripwaxhmm, mine didn't?!
13:17:09Coldtoastmaybe I have 2 years here in Australia
13:17:42belgarathreally? mine came with a warranty card saying 1 year for america and 2 for europe
13:17:52crwleurope has warranty for 2 years?
13:18:07belgarathapparently yes
13:18:09crwli thought i have one year, and it's gone now, bought it in june 2004, i think
13:18:21belgarathdunno if they changed it tho
13:20:16 Part stripwax
13:20:24 Join stripwax [0] (
13:20:34stripwaxsomething keeps killing my irc window..
13:20:51belgarathwould this be hard to implement:
13:21:20Coldtoastthat'd be very nice
13:22:41Coldtoastwho here has dismantled their iriver?
13:25:31 Quit Moos (" HydraIRC -> <- IRC with a difference")
13:26:54belgarathi would, so that i could fit a new battery
13:27:07belgarathhowever i can't find anywhere that sells them
13:27:19stripwaxebay dude!
13:27:27stripwaxthey're about 7 ...
13:27:53belgarathwhat do you search for
13:28:01belgarathi kept getting 1600mAH batterys
13:28:11stripwax2200 mah ipod ...
13:28:11belgarathwhich isn't much of an improvement
13:28:53belgarathi stand corrected thanks
13:29:14stripwaxHCI reckons it could be 2300 mah
13:29:26stripwaxbut i never found any of those
13:29:45Coldtoastit's 2200mAH stripwax
13:29:49belgaraththey're all 4mm wide
13:29:54belgarathi mean deep
13:29:55ColdtoastIordered the same battery as him
13:30:02belgarathwill that fit in an h120?
13:30:15Coldtoastsnug fit
13:30:28belgarathbut that's not bad tho?
13:30:44belgarathbecause i'd really love a new battery
13:30:56Coldtoastwhy bother tho? youonly get another ~10hrs battery :)
13:31:02stripwaxi'll do this just as soon as my battery performance really starts to die
13:31:25stripwaxi'd rather improve battery performance of rockbox than buy a new battery ;-)
13:31:35ColdtoastI'm doing it now, even tho my battery's fine and only 7months old
13:32:19ColdtoastI'm one of only 2 ppl I know who has an iriver
13:32:36Coldtoastmy friend has a h340 tho. mine's a h140
13:32:46stripwaxi know three others (two h120, two h340)
13:32:57belgarathdo you have to buy a torx screwdriver to fit the battery?
13:33:24stripwaxyeah, the cheapest set i found was at
13:33:26Coldtoastsome ppl sell them with a torx to fit the ipod
13:33:46Coldtoastlike the guy I bought from. comes with a torx
13:34:22ColdtoastI'll need it for my phone too, if i fits. replacing the cracked LCD next week and need a torx
13:34:27belgaraththis the kind of thing?
13:34:55 Join jipi [0] (
13:35:02 Join jipi_ [0] (
13:35:25Coldtoastwhat do you need for the iriver? T6?
13:36:02belgarathis it difficult to switch the polarity?
13:36:02Coldtoastthought that might be the case. And how about the ipod?
13:36:07 Quit jipi_ (Client Quit)
13:36:07 Quit jipi (Client Quit)
13:36:14Coldtoastit's VEY easy belgarath
13:36:17 Join jipi [0] (
13:36:17 Join jipi_ [0] (
13:36:37Coldtoastdone it a few times with various things
13:36:40stripwaxbelgrath - i got this
13:37:08belgarathcool thanks
13:37:30belgarathi will buy one of these batterys and install it when my h120 comes back from america
13:37:30 Quit jipi_ (Client Quit)
13:37:37ColdtoastSlasheri: are the only torx screws the 4 on the top and 4 on the bottom?
13:38:45stripwaxisn't there one on the side/
13:39:01Coldtoastthink I read there's no need to undo tha one tho
13:39:10stripwaxah cool
13:39:20Coldtoasthope that's the case
13:39:26 Quit jipi (Client Quit)
13:39:36 Join jipi [0] (
13:39:42 Quit jipi (Read error: 131 (Connection reset by peer))
13:41:05SlasheriColdtoast: yes, there are two screws on the sides too
13:41:19Coldtoastah damn
13:41:24Slasherithose are the same T5 though
13:41:39Coldtoastwas hoping there was no need to take those out
13:42:10Slasherihmm, it's easy to take those out
13:42:35Coldtoastand no little fiddly bits fall out when youdo?
13:42:37Slasherijust don't tighten any screws too much when you put them back in
13:45:59belgarathhmm ionity batterys look good as well
13:47:11Coldtoastyeah... at 4 times the prce
13:47:17Coldtoastand smaller capacity
13:47:47ColdtoastIonity is rated at 1900mAH (tho they sometimes list them as 1900-2000mAH)
13:48:17***Saving seen data "./dancer.seen"
13:49:09Coldtoastdamn! ipod uses T6
13:51:01belgarathi suppose
13:51:16belgarathbut i don't know how i would switch the molex connectors
13:51:28belgarathon the 2100mah battery
13:52:24Coldtoastit's liek this
13:52:45Coldtoastthe wires have a metal bit crimped to them which fits into the plastic connector
13:54:08Coldtoasterr, this is hard to explain
13:55:05Coldtoastbasically, you get something small and pointy, slip it into the top of the connector to lift the lplastic that holds the metal crimped bit into the connector
13:55:29Slasherinow the iriver has passed 16h and still playing! better battery performance with rb than iriver fw :)
13:55:45Coldtoastpull the wires out and switch them and put them intot he connector again
13:56:27 Join bipak_ [0] (
13:56:36belgaraththats great news slasheri - is this with 1300mah battery?
13:56:41Slasheribelgarath: yes
13:56:55ashridahSlasheri: nice work.
13:57:05belgarathit must be the extra codec optimisations and smarter hd buffering
13:57:22belgarathwhat bitrate btw?
13:57:28ashridahis that coming down to the caching policy rockbox has vs iriver's, or something else?
13:57:37Slasheribelgarath: 128k cbr mp3
13:58:11*ashridah notes he came across an ogg that claimed to be 60000kbps@32kHz last night
13:58:25belgarathhow did you get that?
13:59:09ashridahi wouldn't be surprised if it was an issue in the ogg file, but they were ripped out of gta:Sa
13:59:28ashridahactually, no, it wasn't the ogg, now that i think about it.
13:59:36ashridahdefinently rockbox's bitrate detection for vorbis :/
14:02:20belgaraththats strange i was quite pleased with the way rockbox detected lame vbr bitrates
14:02:33belgarathbetter than iriver just copping out and displaying "vbr"
14:02:51ashridahyeah, but that's not vorbis, that's mp3.
14:02:59ashridahapparently vorbis is a bit harder
14:03:20crwlwith ogg files, iriver firmwares show the nominal bitrate field in the vorbis header
14:03:29crwlwhich is quite lame too
14:12:31belgarathoh dear my sister's ipod mini crashed...
14:12:39belgarathand it doesn't have a reset button
14:13:53 Quit stripwax (Read error: 110 (Connection timed out))
14:14:42 Join stripwax [0] (
14:14:50stripwaxgrr. hi again! keep dying
14:14:58 Part stripwax
14:15:06 Join stripwax [0] (
14:15:18crwlit's quite irresponsible to create a device with a fixed battery that doesn't have a reset button
14:15:49belgarathyes especially since the procedure described by apple doesn't seem to work
14:16:12Slasherihi HCl :)
14:16:18stripwaxwe bitching about ipods? brilliant!
14:16:18HClhey slasheri
14:16:23HClany progress on those events? :)
14:16:31SlasheriHCl: there are already done :)
14:16:38Slasheriyou may start using them ;)
14:16:52Slasherijust look at the playback.h
14:18:13HClwhats the disk spinning bool for?
14:19:01Slasheriyou can use that variable to determine if the disk is currently spinning (there are some special cases when it's not)
14:19:07 Quit bipak (Read error: 110 (Connection timed out))
14:20:54 Join Chamois [0] (
14:21:21HClwell if it isn't i'd have to spin it up anyways, dunno if it matters
14:21:34SlasheriHmm, ok
14:22:16Slasherimaybe i could fix even that, just ignore the variable
14:24:15belgarathwould it be possible for, when an error occurs causing you to manually reset the player, it automatically resets itself?
14:24:29belgarath*without you having to forage for a paperclip...
14:25:10belgaraththis was suggested by stripwax
14:26:01*HCl scratches his head
14:26:16HClmmm, i think the last track bool would be more useful at the unbuffer bulk event o.o
14:26:34Slasherireally? :D
14:26:42*Rick bounces
14:26:44HClyea cause when you're buffering you're just reading, not writing
14:26:48Rickso what's new?
14:26:51stripwaxshould be a piece of cake.. just change the infinite while loop in system.c to do a reset... based on a config option maybe
14:26:54RickI havn't looked at rockbox in a couple months
14:26:55HCland when unbuffering it would allow me to fsync only at the end
14:26:56Slasheriah, yes. I could add it to both
14:27:01HClplease :)
14:27:19Rickno college means not much use of my iriver :<
14:27:40stripwax(line 424)
14:30:20 Join amiconn [0] (
14:31:29*HCl goes to rewrite his code
14:36:53Slasheriiriver battery test finished, playback time: 16 hours 40 minutes
14:38:44stripwaxwhat was the test setup?
14:39:04ashridahyou know, there's something decidedly odd about listening to the pastoral symphony while gunning down gangs in gta :)
14:40:23Slasheri128 cbr mp3 (one folder with repeat on), crossfade on, no remote, volume 62 (no headphones)
14:40:41stripwaxstock battery or upgraded?
14:40:52Slasheristandard old stock battery (1300 mAh)
14:40:58stripwaxvery nice
14:41:11Slasherihi amiconn
14:41:14Rickhello there amiconn
14:41:17HClmmm, my new headphones rock compared to my old ones
14:41:19amiconnSlasheri: Was that iriver fw or rockbox?
14:41:28Rickpretty sure rockbox
14:41:31Slasheriamiconn: rockbox of course (iriver has no crossfade ;)
14:41:37Rickmax volume of iriver is 40 iirc
14:41:38HClnice :)
14:41:49HCli'll try to do a battery test later on, i have an 2200mah battery..
14:41:50amiconnAnd btw, I would recommnd to use mp3 ~192 kbps for runtime tests
14:42:07amiconn(That's what the rockbox estimations on archos are based on as well)
14:42:36Slasheriah, i just wanted to make the test with about same setup than iriver has based the 16h estimation
14:43:01amiconnAh ok
14:43:09amiconnSo we get a nice comparison :)
14:43:35amiconnStill someone should probably do a 192 kbps test (perhaps with iriver fw as well)
14:44:18stripwaxthe battery upgrade forums on misticriver have detailed battery tests on iriver fw already
14:45:06amiconnstripwax: Are there results for the stock battery as well?
14:45:24stripwaxpretty sure there are..
14:45:51SlasheriHmm, but those tests was made with the old rockbox code?
14:46:53stripwaxSlasheri- how old?
14:47:51Slasheristripwax: at least the mp3 code has got some performance enhancements due to the dsp system
14:55:59 Quit belgarath ("CGI:IRC")
15:03:03stripwaxam i right in thinking cpu boost is only enabled while filling the pcm buffer? or is it always enabled all the time while playing vorbis tracks?
15:06:07 Join Jleagle [0] (
15:06:31Slasheristripwax: it's always enabled while filling the buffer and after that only when it's necessary
15:06:41stripwaxSlasheri - cool
15:08:54stripwaxah yes, just found the pcm_watermark_callback stuff. neat-o!
15:09:39HCla short on an 16bits cpu is still 16bits, right?
15:09:49stripwaxHCl yeah usually
15:10:59Slasheriyes, but coldfire is 32 bits cpu
15:11:18Slasheriso short = 16 bits, int 32
15:14:01amiconnHCl: char, short and long are always 8, 16, and 32 bits on both 16 and 32 bit cpus. Only int changes size
15:15:22stripwaxamiconn - not sure that's guaranteed - only that char <= short <= int <= long.
15:15:55amiconnI said on 16 and 32 bit cpus - not everywhere
15:16:13amiconnThings get very tricky when considering 64 bit archs
15:20:21 Quit Jleagle ("Chatzilla [Firefox 1.0.4/20050511]")
15:29:49 Quit ashridah (Read error: 110 (Connection timed out))
15:32:03 Join ashridah [0] (
15:35:08stripwaxam i right in thinking DEBUGF output can go to the iriver remote? i'm wondering if i can get useful debug output to find out why my hard drive is sometimes squeeling...?
15:35:53HClamiconn: i just saved about 700 bytes, there's a slight chance we might have rombox after i finished implementing the 2 event approach thing... dunno how much you need..
15:36:19amiconnNeato :)
15:36:35amiconnCurrently I would need 720 bytes less to make rombox fit
15:36:46amiconn(with sh-elf-gcc 3.3.5)
15:37:13amiconnI'm currently working on centralising the IDATA/ICODE #defines
15:37:29amiconnI need that for greyscale - otherwise readability would suffer
15:38:17stripwaxanyone know why auto bookmark writing on poweroff doesn't work if Harddrive Power Off is enabled?
15:38:24amiconnHCl: What are you working on? More efficient database code?
15:38:25stripwaxis that expected?
15:38:36amiconnstripwax: Nope
15:39:07amiconnI don't use autobookmarking, so I didn't experience it myself
15:39:50amiconnHarddrive poweroff just saves a little bit more power by phyisically switching off HD power instead of using standby mode only
15:40:06amiconnStandby current saved, but at the cost of a little bit longer spinup time
15:40:07stripwaxi'm wondering if attempting to write an automatic bookmark on power off is knackering my harddrive and causing that squeeling sound
15:40:32amiconnIt should never do that on sw poweroff platforms, as the iriver is
15:41:21stripwaxwell something's not right.. can i get debugf output on the remote to see what it says when my hd squeals?
15:41:43amiconnYou would have to add that yourself, I think
15:41:46Slasheristripwax: you should use logf for that, i don't know if DEBUGF works on iriver
15:41:55amiconnNo it doesn't
15:42:14amiconnDEBUGF() is for debug builds, and there is still no gdb stub for iriver
15:42:15stripwaxI wouldn't know where to add it .. given that i don't know what the cause is. could debugf be madded to logf somehow?
15:42:36amiconnPossibly, it's a macro after all
15:44:05stripwaxhm, the ata_hard_reset code should probably be using the atomic bitfield macros
15:44:34amiconnThere are very probably many places that should use them
15:44:47amiconnThat's why I added that ToDo: to my commit msg
15:47:29HClamiconn: converting the runtime database code to store its data in the id3 structure
15:47:59HClnoticed a buffer that was being used for filenames that was 1kb, rather than MAX_PATH
15:48:12HCli do have to remember to implement that limit in the database generator though
15:48:18***Saving seen data "./dancer.seen"
15:48:32HClit should print out severe warnings / errors if it goes above the limit set in the code
15:48:46HClor better, just skip files that have a filename longer than that
15:48:48amiconnMAX_PATH is the limit in rockbox almost everywhere
15:48:52HClyea, i know
15:49:31amiconnThat buffer size is bss space though, doesn't save binary size
15:49:49amiconnStill nice to save it; archos RAM isn't exactly large..
15:50:15HCloh, okay :/
16:07:19HClSlasheri: did you have time to fix the last track thing yet ?
16:07:43 Part amiconn
16:08:57stripwaxhmm. ata_read_sectors has a 400ns wait implemented as 5 nops - is that correct for iriver? seems dod
16:21:38*HCl fetches his iriver
16:21:41*HCl prods Slasheri
16:21:50HCli got questions...
16:30:54SlasheriHCl: yes, just a moment..
16:30:57Slasherii try to do it soon :)
16:32:34HClk :)
16:32:41HCli had questions
16:32:51HCldo the unbuffer events get called on shutdown?
16:32:58HClcause they need to be called then for the remaining tracks
16:33:04HClthat was the most important question, i think
16:33:13HCland when you press stop
16:34:42SlasheriHCl: yes they do
16:35:12Slasheri(if the shutdown process calls audio_stop..)
16:35:29Slasheriand i think it should always do that
16:36:21HClthen i'm ready to test my new implementation, i think
16:37:22Slasherithat's good, i hope the database will be someday compatible with amaroK
16:38:29Slasheria great music player for kde, it has at least some ipod db sync functionality
16:38:39Slasheri(scoring, how often tracks are played etc.)
16:38:50Slasheriit would be cool to be able to sync it against iriver too :)
16:39:00HClthegeek was wanting support for playcount tags too
16:39:18HClit shouldn't be *too* hard to implement, in the java version at least
16:39:26Slasherigood :)
16:39:38HClits not of a high priority at the moment..
16:40:32RoriY HELO THAR
16:47:19Coldtoasthow's the ol' wps thing going now? :)
16:49:09RoriI got a workaround
16:49:32Roribut it means you cannot have a permanent line like 'Next track...' in the line
16:49:52Roriit works
16:50:46Roriif you FF b4 the buffer kicks in though it won't work until you FF again but it's no big deal
16:52:15Coldtoasttime to free up 9GB of HDD space I think
16:56:24Rorioh the iriver?
16:56:26 Quit ashridah ("Leaving")
16:56:50DBUGEnqueued KICK Rori
16:56:50RoriHard Disks: (Total/Free: 308.28/113.06GB Total/Free space on: C: 74.53/10.85GB D: 231.74/102.19GB J: 2.01/0.01GB)
16:56:50RoriNetwork Drives: (Total/Free: 867.21/142.15GB Total/Free space on: P: 4.38/0GB T: 74.52/12.81GB U: 279.47/36.23GB V: 189.92/34.47GB W: 76.32/39.43GB X: 19.08/19.02GB Y: 111.76/0.03GB Z:
16:57:09stripwaxRori :-)
16:57:15DBUGEnqueued KICK Coldtoast
16:57:15ColdtoastHard Disks: (Total/Free: 409.88/315.18GB Total/Free space on: C: 20/14.67GB D: 92.7/86.59GB E: 92.7/45.19GB F: 92.7/77.1GB G: 111.79/91.63GB)
16:57:27ColdtoastRAID goodness
16:57:33RoriI need to raid
16:57:49RoriNext time I get a raid and a few hd's
16:58:00Coldtoastonly 300GB of that is RAID tho. Then I have 120GB of regular IDE and another 80GB for Linux
16:58:19RoriI always have too much data on the drives to make a raid
16:58:20ColdtoastRAID0 too......
16:58:24RoriI'd have to backup first
16:59:11Rori2 drives in this pc and 8 upstairs
16:59:25Coldtoastcan't be for anything legal. tee hee
16:59:40Rorisure it's all my pr0n ;)
16:59:48Coldtoastno doubt. heh
16:59:52Coldtoastall jpg too
17:00:00Roriactually it's mostly anime and MP3
17:00:12Coldtoastsee? nothing legal then
17:00:26Coldtoasttho ppl DO argue about fansubs
17:00:31Roriwell anime can be a bit of a grey area for non-licensed shows
17:00:56RoriI am sure the Japanese would say it's not grey at all though :)
17:01:16Coldtoastwho gives a crap anyway? I have close to 500GB of HDD in this machine and only using, like, 75
17:01:30RoriMost don't chase people because they know it's how they get a market outside of Japan
17:01:50 Join amiconn [0] (
17:01:54Roriotherwise most shows would go unheard of and people would not bother
17:07:53stripwaxhow fast does the iriver cpu run when boosted?
17:09:44stripwaxhmm, so that would mean we'd need 50 nops to implement a 400ns wait?
17:10:55stripwaxrather than the 5 currently in ata.c ? am i missing some info?
17:11:18HCli dunno
17:11:23HCli'm not a hardware kind of person
17:12:20stripwaxi'm wondering if the screaming sound coming from my harddrive is a result of missing wait_for_start_of_transfer and doing a soft_reset. it definitely sounds like the drive does some reset seeking when it's screaming.
17:12:39stripwaxok right, i'm going to patch my local build with a bunch more nops and see how things go
17:15:09HCldoes it happen with iriver firmware?
17:16:36stripwaxreckon a for(i=0;i<15;i++) nop should be about 400ns.. i'll see if it makes a difference
17:16:48amiconnstripwax: The additional delay for higher clock frequencies is achieved by setting more waitstates for the IDE port
17:17:03amiconnLook at system.c
17:18:53stripwaxamiconn - hmm, so any other ideas what might cause my problem?
17:19:07amiconnUnfortunately not
17:32:17stripwaxamiconn - not sure I understand how 0x106000 | (5<<10) sets CS2 pre and CS2 post. Is that comment just wrong? looks like it sets CS2 pre and CS1 post (not CS2 post).
17:32:53amiconnHmm, I didn't look that close. Linus wrote this stuff
17:34:41stripwaxok. also looks like we only ever set one wait state for CS2, irrespective of cpu freq. i could be wrong
17:35:10stripwaxyeah, i am wrong ;-) we use zws unless it's boosted, in which case we use 1 ws.
17:48:20***Saving seen data "./dancer.seen"
17:48:52 Quit PaulJ (Remote closed the connection)
17:50:25 Join bagawk [0] (1000@bagawk.user)
17:57:37 Quit Odie ()
18:20:33 Join Sucka [0] (
18:22:57 Join CheeseBurgerMan [0] (~Dan@
18:32:11 Part amiconn
18:45:45 Join nphase [0] (nphase@nphase.user)
18:47:11nphasedoes anyone in here know evilwalrus?
19:03:59 Part nphase
19:22:26 Join yyz [0] (
19:26:21 Join Stryke` [0] (
19:35:08 Join gravygoat [0] (
19:37:16 Part gravygoat ("Leaving")
19:42:31 Join Musicmad [0] (
19:43:34 Quit Stryke` (Read error: 145 (Connection timed out))
19:43:40 Quit bagawk ("Leaving")
19:48:25***Saving seen data "./dancer.seen"
19:49:53 Quit yyz (Read error: 131 (Connection reset by peer))
20:21:15stripwaxHmm... I reckon that DCR manipulation in set_cpu_frequency is wrong
20:21:36stripwaxunless it's the intention to write 28 decimal into the refresh counter? sounds weird..
20:27:58 Join TCK- [0] (
20:30:32stripwaxignore above. i'm sure i'm just an idiot..
20:36:39SlasheriHCl: fixed the event callbacks :)
20:36:54Slasheriand now you don't need to worry about the disk spinup
20:39:04stripwaxHmm.. what's the difference between CPUFREQ_NORMAL and 'default'?
20:39:59SlasheriHmm, i think one of them might be 11 MHz and the other 48 MHz
20:40:24stripwaxoh so default isn't normal. weird, but gotcha. thx
20:45:30 Join belgarath [0] (
20:46:27 Quit TCK (Read error: 110 (Connection timed out))
20:48:01belgarathis it true that the latest commit means that you no longer have to wait 3 minutes into a track to get the next song info?
20:48:09Slasheribelgarath: yes :)
20:48:23Slasherinow you get that info instantly
20:48:26belgarathi would test it but my player is all packaged up
20:48:32Coldtoastno you don't Slasheri....
20:48:42Coldtoastnot instantly :)
20:48:44SlasheriColdtoast: Hmm?
20:48:48Coldtoastonce it buffers
20:48:53Slasheriok, there is a few second delay ;)
20:48:58Coldtoasthehe. yeah
20:49:06Coldtoastworks great, btw
20:50:03Coldtoaststill need that workaround Rori mentioned tho
20:50:17Coldtoastthus: %s%t0;%?Ia< %Ia| Awaiting Buffer...>
20:50:44Coldtoastwithout the %t0 it doesn't update the wps when i has the info
20:51:03Coldtoastbut if you ff a teeny bit, it updates
20:52:52CtcpIgnored 2 channel CTCP requests in 25 seconds at the last flood
20:52:52*belgarath prays that his h120 will come back from america soon after being posted tomorrow
20:53:16Coldtoastand oddly, it only updates the first line of next track info right away
20:53:41Coldtoast%s%t0;%?Ia< %Ia| Awaiting Buffer...>
20:53:41Coldtoast%s%t0; %It
20:54:08Coldtoastwhen it has the info, it updates the next artist but there's a few secs before it updates teh wps with the next track info
20:54:42Coldtoastneed %s%t0; %It
20:54:49 Join Lear [0] (
20:55:29Coldtoastneed %s%t0;%?It< %It| %Fn> so it looks like it updates the next track info til it actually does
20:56:02Coldtoastjust on the first track tho.
20:56:50Coldtoastafter the first track, it works properly, even when skipping back/next
21:00:53belgarathhow difficult is it to install a new hard drive into a h120?
21:01:08belgarathi'm thinking of getting a single platter 30gb
21:01:22stripwaxbelgrath - pretty easy apparently
21:01:55belgaratheasier than switching polarities on the ipod battery?
21:02:07thegeekI would say so yes
21:02:14thegeekit's just basically plugging it in
21:02:31stripwax- after partitioning it, of course..
21:02:57thegeekthe iriver firmware can format it;)
21:03:34belgarathi thought you would only need to partition drives above 40gb
21:03:50belgarathbecause that is the maximum iriver firmware can read?
21:03:59thegeekyou sure?
21:04:06thegeekhow do you know that?
21:04:19belgarathi think i read it on misticriver
21:04:24thegeekI think I saw a thread on misticriver with a 80gb?
21:04:26stripwaxcheck again!! unmodified iriver firmware can use 80GB drives.. check misticriver for details
21:04:34stripwax(thegeek - jinx! :-))
21:04:37belgarathok stand corrected
21:07:06ColdtoastI don't even visit misticriver any more
21:08:35belgarathreally, who come?
21:08:48 Join Aison [0] (
21:09:05Coldtoastno need to
21:09:20belgarathi have to admit there has been a huge influx of h300 and h10 owners who keep asking the same questions again and again
21:09:23Coldtoastthere's never any news relevent to the h140
21:09:33ColdtoastI have zero issues with my player
21:09:38Coldtoastthere's nothing there for me
21:11:15Coldtoastthere's yellow builds, btw
21:15:34belgarathwonder what errors were introduced...
21:19:39Slasherioh, fixing that
21:23:57 Join Leperkawn [0] (
21:24:43 Quit Leperkawn (Client Quit)
21:24:44stripwaxdoes sleep(1) sleep for 1 ms? or some other duration?
21:25:02 Quit StrathAFK ("Client closed")
21:25:03Slasherii think it's slightly more than 1ms
21:25:09stripwaxless than 2ms?
21:25:18Slasherithose sleeps are not very accurate at small intervalls
21:25:41stripwaxcan we guarantee that sleep(1) is > 2ms?
21:25:55stripwaxcos if not then ata_hard_reset should be fixed for iriver
21:26:32Slasherijust a moment
21:27:19SlasheriHZ = 100, kernel tick timer interval: 1000/HZ, so one tick = 10ms
21:27:28Slasherithat is the shortest sleep period
21:27:50stripwaxSlasheri - thx!
21:37:24*HCl goes to test his new code
21:37:28HClits hot..
21:39:48stripwaxHCl the dynamic db stuff?
21:43:16 Join webguest48 [0] (
21:43:23webguest48hello all
21:43:42webguest48anyone work in voice ui?
21:43:49webguest48it's very usuful
21:43:52webguest48please do
21:43:56 Quit webguest48 (Client Quit)
21:44:59HClstripwax: yea
21:45:12HClmore production quality, more battery life etc
21:45:47HCljust checking whether its correct that i'm not getting any messages
21:45:50HCland that seems to be the case...
21:45:53HCli need to add more debug messages
21:48:25LearJust noticed that after stopping playback, cpu boost is still active (at 1)...
21:48:27***Saving seen data "./dancer.seen"
21:50:21HClthats not good
21:50:25HClSlasheri: ?
21:52:44stripwaxdoes cpu clock ever fall back to 11Mhz after boot, or is at always switched between to 48/120?
21:53:10HCl11mhz isn't used iirc
21:53:14HClmaybe during usb mode
21:53:24stripwaxduring system reset afaict
21:53:41stripwaxso that would include bootloader usb mode? (i guess?_
21:53:43LearA call to pcm_set_boost_mode in audio_stop_playback looks right...
21:54:01SlasheriLear: Hmm, i can't reproduce that
21:54:42LearI just start playback, and then stop pretty soon after that (possibly while still buffering, thus in boosted state).
21:55:06 Join webguest52 [0] (
21:55:30SlasheriHmm, doesn't work for me..
21:55:34Slasheriwhat version are you using?
21:55:44webguest52can anyone tell me how the playlist funtion works for rockbox on h1x's
21:56:04webguest52or if it does work at all
21:56:10LearBuild from today, about six o'clock...
21:56:11Coldtoastit sure does work
21:56:23webguest52im having trouble creating a list
21:56:49Coldtoastwhen you're in the fle view, hold down the joystick button (don't move it in teh down direction, depress it)
21:57:08Coldtoasthold it for a more-than-split-second
21:57:21webguest52doesnt that make it play the file?
21:57:22Coldtoastthen go into the Playlist menu
21:57:37ColdtoastI said HOLD it down. not just press it
21:58:02Coldtoastsee he Playlist menu?
21:58:33Coldtoastcool. go into there
21:58:40SlasheriLear: Hmm, i committed a fix for that
21:58:49webguest52just created new playlist
21:58:53Coldtoastyou have a couple of ways to add tracks to the playlist
21:58:54SlasheriNow it should always call the pcm_boost(false) while in audio_stop_playback
21:59:36webguest52when i held down
21:59:47webguest52it gave me some options
21:59:56webguest52playlist rename etc
22:01:01*HCl stares at his iriver
22:01:06HClif i keep playing the same album
22:01:11HClwill it not rebuffer at all
22:01:20HCli'm scared to press play cause it activates a lot of my code ;x
22:02:05webguest52aiight i got it
22:02:11webguest52thanks toast
22:02:17Coldtoastno probs
22:02:52webguest52one last thing what is the recursively insert directories
22:03:50webguest52its under the playlist options
22:05:50stripwaxif enabled it will insert files below the selected directory too, not just the files IN the selected directory...
22:05:54*HCl stares at stuff..
22:06:00*HCl immediately realizes he has a bug
22:06:34 Join tourist08 [0] (
22:07:42tourist08hi all
22:08:01tourist08first time online
22:08:14Coldtoastfirst tiem EVER?
22:08:24Coldtoastlike, you've never even checked email before?
22:08:42webguest52an internet virgin
22:08:44tourist08first time using irc
22:09:15CheeseBurgerManWelcome to my world.
22:09:29tourist08rockbox progress is really great!
22:09:31*HCl broke the entire runtime database, and doesn't even understand how.
22:09:36HCland on top of that i'm lagging
22:09:42HClwhere is my beeping wire
22:12:45tourist08would it be difficult to add a -clear playlist- entry in the playlist, i think it would make sense
22:13:39tourist08so you could create a new playlist without stopping the actually playing song
22:14:57Coldtoastyou can clear the playlist in stopped mode....
22:15:08HCli'm an idiot
22:15:15Coldtoastbut that's no good to you, huh?
22:17:59HClbut the debug messages i added are good either way
22:18:02SlasheriHCl: Hmm, currently the playback code doesn't recognize if a requested track is already buffered (that will be fixed)
22:18:09*HCl completely forgot about amiconn's disable feature
22:18:16HClSlasheri: wha?
22:18:34HCloh yea
22:18:40HCli was planning to use the track change event
22:18:40SlasheriHCl: so it will rebuffer if you repeat the same album
22:18:43HClto tell whether a track has finished
22:18:48HClso you won't have to rebuffer
22:18:53HCloh right
22:19:01stripwaxSlasheri - it will rebuffer even if you repeat the same track?
22:19:03HCl :p
22:19:09HCli thought it was already not rebuffering
22:19:11Slasheristripwax: yes it will
22:19:16HClbut if it is still rebuffering then it won't matter
22:19:22stripwaxok, that fix'll improve battery tests
22:19:40SlasheriHCl: it should never matter how you handle the stuff in your code
22:19:55Slasheriplayback system will always generate the events, whether it's buffering or not
22:20:03stripwaxactually, it'll only improve flawed battery tests :-p what i am thinking
22:20:05HClwell the buffer events would be illogical to give when its not actually buffering
22:20:47SlasheriHCl: hmm, yes.. maybe some better callback name should be invented :)
22:21:19Slasherisomething like track_finish, track_end
22:21:23Slasherithat's how it works now
22:21:26 Quit CheeseBurgerMan (Read error: 145 (Connection timed out))
22:21:35Slasherioh, track_begin i mean :D
22:21:49HClthere's a typo in playback.h
22:21:52HClbut don't bother fixing it
22:21:56HClmy commit will fix it
22:21:56Slasheriwhere? :)
22:22:03HClunbufer -> unbuffer
22:22:07HClbut don't fix it!
22:22:10Slasheriah :D
22:22:10HClor i'll just get a conflict
22:22:13HCland get more trouble
22:22:13HCl :P
22:22:37Slasheriare you sure there is unbufer?
22:22:45Slasherii could'nt find that..
22:22:48HCljust don't fix it
22:22:51HClforget i said it
22:22:54Slasherii wont
22:23:00Slasheribecause i don't even have it :D
22:23:45Slasherioh, it was in playback.h =)
22:23:54Slasheriok, i don't touch it
22:25:55SlasheriHCl: maybe you should change the callback names also: track_buffer -> track_begin, track_unbuffer -> track_finish
22:26:37Slasherijust replace all occurences
22:31:20 Join amiconn [0] (
22:35:23HClmmm, i prefer you to do that after i committed, okay?
22:35:29HCli'm not far away, i hope
22:35:41HClamiconns default-disabled runtime database option bit me without me noticing :P
22:35:51HCland i was like argh why doesn't it work
22:35:52HCl :P
22:36:21amiconnHCl: I thought that would be a good thing ... you pointed out yourself that it is experimental
22:36:26HClofcourse it was :)
22:36:31HClits just something i forgot
22:36:41HCli'm happy to know what caused it though, i hate debugging
22:36:48HCljust making a build with more messages so i can check more
22:37:06SlasheriHCl: please also disable the event handlers (with NULL pointer) if the code is disabled
22:37:22HClmk, i'll add that before committing, good point
22:37:38 Part Musicmad
22:38:37HClthats not good... it should've printed my message..
22:39:49*HCl slaps himself awake
22:39:57HClstupid runtime database option biting me again xD
22:40:16HClamiconn: where is the option located?
22:40:20amiconntourist08: You can't clear the current playlist while playing music. The whole rockbox playback system is playlist based
22:40:25HCl :)
22:41:00amiconnGeneral Settings -> Playback -> Gather runtime data
22:41:05HClyea, got it :)
22:41:12HClcode does stuff now
22:41:17crwlmaybe at least something like "remove all but the currently playing song" would be nice
22:41:18Coldtoastjust so you know guys. I've had no crashes for ~3 days now
22:41:29 Quit webguest52 ("CGI:IRC")
22:41:39 Join Bagder [0] (
22:41:47HClrundb: no entry and not found
22:41:49HClthats correct
22:42:44SlasheriColdtoast: that's really good to know. I think the fixes i made might be working :)
22:43:01Coldtoastyeah. I ALWAYS have crossfade enabled too
22:43:06Slasherii have neither had crashes
22:43:15Slasherime too ;)
22:43:25belgarathwould it again be possible to implement an option where if the unit crashes it automatically resets?
22:43:39ColdtoastI'm using mine around 3hrs a day
22:43:43tourist08amiconn: :( thought it should be no problem.
22:43:44 Join muesli- [0] (
22:43:49belgarathor would that be impossible?
22:43:53Slasheribelgarath: for some type of crashes i think the answer is yes
22:44:21belgarathhi muesli
22:44:37stripwaxbelgrath/Slasheri - just put a system_reset call in the UIE handler?
22:44:39muesli-hi belgarath
22:44:57 Quit TCK- (Read error: 145 (Connection timed out))
22:45:05stripwaxbelgarath - i'll let you know if that just works
22:45:20belgarathok thanks
22:45:47Coldtoastlistening to a podcast here and they're talking about MP3s and the RIAA. Wonder if it'd ever get SO bad the authorities would stop you on the street if you have a DAP of any kind
22:45:47amiconnstripwax: There is a slight proplem to solve
22:45:47belgarathalthough its getting more an more difficult to purposefully crash the player now
22:46:13amiconnstripwax: You can do that, but it should not reset immediately, it should wait for a button press
22:46:23stripwaxColdtoast - that'll never happen
22:46:26stripwaxamiconn - why?
22:46:36 Quit Infirit (Remote closed the connection)
22:46:39amiconnErm, in order to read the error msg?
22:46:43Coldtoastso you cna see the error?
22:46:55amiconnHowever, you can't use the button driver, you have to access the hardware directly
22:47:13amiconnThat also means you can only use buttons which are connected to port pings
22:47:24amiconn...i.e. Play
22:47:49stripwaxamiconn - what if I don't want to see the error? :-p
22:47:58Coldtoastyou should have something liek an event viewer for Rockbox
22:48:13Coldtoastor some sort of error log
22:48:27Bagderwhen it crashes, how should it live to save the log?
22:48:48stripwaxbagder - eprom?
22:48:51Coldtoastleave that one up to you smart boffins to sort out? heh
22:48:58Bagdera log in eprom?
22:48:58stripwaxerr.. you know what i mean
22:49:36stripwaxany mileage in keeping > 1 error log anyway>
22:50:14tourist08amiconn: it is even possible to delete playlist entries manually while playing, so why not automaticly
22:50:26BagderSlasheri and HCl: I'm hoping for some email replies from you
22:50:27Coldtoastas fixes are made and erros change, you'd be able to see the pattern of errors?
22:51:03stripwaxColdtoast - could you use a piece of paper and a pen for that? :-)
22:51:09SlasheriBagder: ah, i will look at that simulator code soon
22:51:13belgarathmuesli: hope you understand what i meant to say...
22:51:24Coldtoaststripwax: not if I'm ou and about or at teh gym or something :)
22:51:26BagderSlasheri: ok!
22:51:37Coldtoastwhich is pretty much where I use my player
22:52:00stripwaxColdtoast mm good point - are you expecting more than one crash during a single gym session?
22:52:06BagderSlasheri: I'll then work on getting sound on linux
22:52:12muesli-belgarath: we will see, check out my answer ;)
22:52:20Coldtoaststripwax: until about 3 days ago? absolutely :)
22:52:45Coldtoastwas getting crashes with crossfade, the pause crash, etc
22:53:04SlasheriBagder: sounds great :)
22:53:11Coldtoastanyway. not very useful for me of course
22:53:21SlasheriBagder: i believe it should not be very hard to fix the code
22:53:24stripwaxColdtoast - ah, ok. I just generally press play and leave it alone after that. It still crashed once though, without crossfade, while just playing a playlist of vorbis :-(
22:53:46BagderSlasheri: nah, the existing plugin method should be a good point of inspiration
22:53:47Coldtoaststripwax: yeah. ppl keep talking to me at teh gym tho so I pause the music while they dribble crap at me
22:54:23Coldtoastthink I'll gather all teh Tool albums in the house and re-rip them as ogg
22:54:46*HCl gets bitten by the 21 character limit as well
22:54:51HClbut i think my code worked, i hope at least
22:54:51Coldtoastactually, I'll do that with all the Pink Floyd albums as their tracks often run into eachother
22:56:58HClSlasheri: am i guaranteed to get a last_track true quickly?
22:57:04HClat the end of a bulk?
22:58:18Coldtoasthey. so iare the %i options in WPS classed as dynamic?
22:58:28Coldtoast%ia, %it, etc?
22:58:58LearColdtoast: nope, only stuff that changes during a track (play time, etc.).
22:59:19Coldtoastthat's what I thought. THere's still the wps issues then?
22:59:25Coldtoastupdate issues
22:59:33LearThere are? I haven't noticed...
22:59:59Coldtoast%s%?Ia< %Ia| Filename:>
23:00:00Coldtoast%s%?It< %It| %Fn>
23:00:04SlasheriHCl: hmm, you will get all of the track events in one single chunk
23:00:05Coldtoastshould be fine, right?
23:00:13Slasheriso there should be no delays between the events
23:00:17muesli-who is coding the wps generaly asking!?
23:00:35Bagdermuesli-: nobody and everyone
23:01:18muesli-i would love to get codec and khz in the bottom bar beside the battery/volume etc.. to save space on the screen
23:01:19LearColdtoast: "la"? Or is that upper case "i"?
23:01:28Coldtoastas Rori was saying yesterday, and I've been testign today, even with the next track detail updates from today, where the info it available very very quickly, the %s causes the wps to not update with the track info
23:01:45Coldtoastuppercase i
23:01:53BagderColdtoast: that is an old bug
23:02:01BagderI say this over and over again
23:02:02Coldtoastif you ff a teeny bit tho, the info gets filled in
23:02:04Bagderevery day
23:02:08HCli'm almost done with the new rundb implementation
23:02:19Bagderit is even in the bug tracker
23:02:23muesli-cheers HCl
23:02:46ColdtoastBagder: (Coldtoast): that's what I thought. THere's still the wps issues then?
23:02:58belgarathwhen will you be able to add databased songs to a playlist?
23:03:01Bagder"wps issues" ?
23:03:14LearBut what is the problem anyway? :)
23:03:26ColdtoastI know there have been fixes over the last day to the .wps stuff. which is why I asked
23:03:34BagderLear: updating info in scrolling lines
23:03:36Coldtoaster wps=.wps
23:04:08Coldtoastthat's why I asked Bagder. on the off chance that was one of the fixes over the last day
23:04:20Bagdersure, but I don't know what "wps issues" means
23:04:23*HCl bites the anti social person being extremely rude with the shared bandwidth of the wireless router
23:04:29Bagderit will always have issues
23:04:39 Quit ghode|afk (Connection timed out)
23:05:07Coldtoastalso, you'll notice I did ask if things like %it were classed as dynamic
23:05:22Coldtoastthe url you just posted is inr egards to dynamic content
23:05:23LearBagder: But wasn't Coldtoast having problems with e.g. %Ia? That's not what I call dynamic...
23:05:31belgarathissues aside, its still loads better than the iriver wps
23:05:39Bagderthe problem is updating a scrolling line
23:05:41Coldtoastit is, yes
23:05:48Bagderno matter what info it is
23:05:57ColdtoastBagder: here's teh thing tho
23:06:08HClwhy is that limitation there anyways?
23:06:15Coldtoastwith %s there, when it has the info it doesn't fill it in til track change
23:06:22Coldtoastafte that, it all works perfectly fine
23:06:22 Quit belgarath ("CGI:IRC (EOF)")
23:06:23Bagderwhy is there a bug?
23:06:46Bagderor why don't people fix the flaw?
23:06:57Coldtoastwith %s it only fills the info in (and works with scrolling and everything) when you either ff a bit or track change
23:07:04LearColdtoast: Ah, so %I is semi-dynamic. :) But Slasheris recent fix should handle that, because next track info always will be available.
23:07:56Coldtoastyeah. the info is definitely there. But for some reason, when it's available it doesn't get updated til you either ff/rw or change tracks. After that, it works great
23:08:10Coldtoastscrolls, updates next track change, scrolls, etc. it's good
23:08:35Coldtoastit's just the first track you play that has the problem
23:08:36SlasheriHmm, adding track_changed = true to the end of read_next_metadata() function could fix that
23:08:40HClSlasheri: could you change the track change event to give the mp3entry struct rather than the track info?
23:08:46Slasherifeel free to do that, i am little away now
23:09:04SlasheriHCl: you still need that function??
23:09:06Coldtoastso am I making sense? or just dribbling?
23:09:12HClyea, at the moment i d
23:09:14Slasheriyes i can.. tomorrow
23:09:18HClor otherwise playcount gets increased
23:09:23HClfor songs that have been prebuffered
23:09:26HCleven if they were never played
23:09:47Slasheriif they were never played, you should determine that when they are unbuffered
23:09:55HCldunno how
23:09:56Slasheriyou will not need the other event for that
23:10:02Slashericheck the elapsed time
23:10:03HCli'm planning to move the playcount increase anywho
23:10:13HCli'm tired
23:10:21Bagderweren't you getting an event at track change?
23:10:23Slasheriincrease the playcount _only_ when you get the unbuffer event
23:10:39HCli was doing that, and it added playcount for all the songs in the buffer
23:10:43HCleven though i hadn't played any of them
23:10:51HClwell, just the first one
23:10:55LearColdtoast: I understand the problem; next track info isn't always available when the WPS is updated, so scrolling that causes problems...
23:11:09Coldtoastah ok
23:11:13*HCl goes to find his kitty
23:11:29Coldtoastso it's just a matter of there being no wps update event once the info is available?
23:11:47 Quit tourist08 ()
23:11:56Coldtoastbut when you ff or track change, there's an update?
23:12:58LearSeems like it yes; I'm not that familiar with the ff/rewind code though.
23:13:14Coldtoastthat makes sense
23:15:02LearHe, I just noticed that it is flagged as dynamic in the WPS code; removing that seems like a good idea actually. Any scrolling should then re-start, but as that doesn't happen very often, that shouldn't be a problem...
23:17:24muesli-cya mates
23:18:59LearColdtoast: No, scratch that, that wouldn't work too well. I've haven't really looked at the wps code for too long to remember all the details... :)
23:19:24Bagderits quite... complicated (to say it in a friendly way)
23:19:39amiconnBagder: See the discussion on when to get and process events in the runtime database code in yesterday's irc log
23:19:42Coldtoastheh. ok :)
23:19:53LearIt wasn't too compilated when I worked on it, but that was a couple of years ago...
23:20:14BagderI remember the alternating line thing added lots of complexity
23:20:21amiconnBagder: Processing a track change event at the exact time of track change is bad. It causes a disk spinup at the track change
23:20:37Bagderonly if you use the disk
23:20:48amiconnYes, but the runtime database has to
23:20:55Bagdernot really
23:21:24Bagderwell, you manage to design a way that avoids it, right?
23:21:26stripwaxcould it not buffer changes in memory and only write them out to the runtime db when the disk next spins up?
23:21:32amiconnBtw, I don't repeat everything from y'day now
23:21:38amiconnBagder: Yes.
23:21:54Bagderstripwax: yes, that's my thinking
23:21:59stripwaxah ok
23:22:06amiconnThe event handling should be split in 2; a track buffer event and a track unbuffer event
23:22:16stripwaxis there a 'disk spinup' event that apps can register to?
23:22:31Bagderstripwax: no, but it would make sense to have one
23:22:35stripwaxcertainly would
23:23:02amiconnI can't imagine where that would be needed
23:23:17stripwaxamiconn - lazy updates to files, no?
23:23:20Slasheriamiconn: that is already done
23:23:42Bagderamiconn: any more specific position in the 2nd july log?
23:23:55amiconnSlasheri: I know. However I don't know whether HCl did the full hookup change yet
23:24:00Bagderah, found it
23:24:21Bagderhm, no
23:24:27Slasheriah, ok :)
23:25:26Bagderanyway, having events on buffer/unbuffer is simply adding a different type of requirement to the rundb code
23:25:51BagderI guess it makes it work anyway
23:27:33 Quit muesli- (Read error: 145 (Connection timed out))
23:27:42amiconnHCl: Yellow and red builds...
23:27:56HCli'll look at it.
23:28:02HCltoo many people are asking me things
23:29:54HClthats not mine
23:29:58HCli did not touch drivers/button.c
23:30:28amiconn database.c: In function `loadruntimeinfo':
23:30:28amiconndatabase.c:486: warning: empty body in an if-statement
23:30:30HCland the track buffer and unbuffer event has to be implemented on archos
23:30:38HCllet me look at that
23:30:47amiconn build-recordersim/apps/database.o(.text+0x510): In function `rundb_shutdown':
23:30:47amiconnapps/database.c:392: undefined reference to `audio_set_track_buffer_event'
23:31:05HCloh what kind of beep is that.
23:31:08HClits not an empty statement
23:31:09amiconnThe former happens for all builds, the latter for all sims
23:31:17HClits only empty because that logf define gets set to 0
23:31:21HClsomeone fix the logf define to not be empty
23:31:34HCli'm really uber busy and hardly have time
23:31:46Bagderit is supposed to be empty
23:32:13HClthen what am i supposed to do?
23:32:26Bagderfix it in another way
23:32:36Bagder1) #ifdef the if()
23:32:43Bagder2) add more code in the block
23:32:44amiconnHCl: Then that if statement is a waste for non-logf, and should be surrounded with #ifdef ROCKBOX_HAS_LOGF
23:33:44HCltoo late
23:33:48HCli already committed a "fix"
23:34:15BagderPLEASE do not do that
23:34:42Bagdermay I fix?
23:35:08HCli told you i'm uber busy and really stressed and don't have time x.x;
23:35:20Bagdersure, but then don't commit
23:35:48BagderI don't mean to be rude, but commiting means some responsibility
23:35:58HCli don't work well under stress x.x
23:36:16 Quit hicks (Remote closed the connection)
23:36:22HClespecially when i have someone asking me for help on how to learn c++
23:36:32HCland they don't even have the computer skills to browse a website
23:36:35Bagderbtw, try making functions static if they're only used within the same file
23:36:44HCli tried to do that a bit
23:37:55 Quit Lear ("Chatzilla [Firefox 1.0.4/20050512]")
23:38:01stripwaxwhy can't logf define to do{}while( 0 ) or something?
23:38:17stripwax(which looks like a statement but compiles to nothing)
23:38:34Bagderyes it could
23:38:52Bagderbut that would've hidden this
23:39:14BagderI'm not sure what the best alternative is
23:39:18HClthe normal way to do an empty logf define is to do (void)<format string here>;
23:39:34HCli gotta go x.x;
23:39:35stripwaxhidden? I guess I thought the problem was an "if( blah ) logf( foo )" or something?
23:39:38HClsorry about the crappy fix
23:39:59Bagderstripwax: yes, the fact that unused code was added would not have been noticed
23:40:20Bagderthe 'blah' in this case is a function call
23:40:33stripwaxah sorry. gotcha
23:40:42 Join hicks [0] (
23:41:15stripwaxhmm.. my hard drive still squeals sometimes. i've tried a bunch of things now and no closer to working out what's up O_o?
23:43:30stripwaxmost often happens on bootup .. anyone point me to things I could try?
23:48:30***Saving seen data "./dancer.seen"
23:54:17*Bagder has no idea

Previous day | Next day