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 2009-09-09

00:00:40 Quit stripwax (Read error: 104 (Connection reset by peer))
00:00:51 Join truthtaco_ [0] (
00:01:15bughunter2well, I might wait a while to see if any more changes are made to that image
00:01:25bughunter2after that I might do the same thing I'm doing now, again
00:01:32bughunter2then re-ask if they want the images :)
00:02:13 Quit TruthTaco (Read error: 60 (Operation timed out))
00:03:04 Join stripwax [0] (
00:03:05 Quit stripwax__ (Read error: 104 (Connection reset by peer))
00:04:11 Quit stripwax (Read error: 104 (Connection reset by peer))
00:05:06bughunter2bertrik: with the OF in MSC mode, while copying multiple files simultaneously (i.e. with multiple explorer 'Copying' windows) the player sometimes spontaneously quits MSC mode (it's still visible in 'My Computer', but file transfers abort with errors)
00:05:14bughunter2any thoughts on that?
00:06:20bertrikno, but I think I haven't tried that on my ams players
00:06:22 Join stripwax [0] (
00:06:39bughunter2perhaps a native USB driver could fix that?
00:06:55bughunter2not sure if it has to do with USB though
00:07:02bughunter2it would seem odd
00:09:44 Quit BdN3504 ("CGI:IRC (EOF)")
00:11:15 Quit n17ikh|Server ("WeeChat")
00:13:27bughunter2rasher: after booting the OF again, suddenly the 'used space' went up again
00:13:33 Part domonoky1
00:14:11 Quit NAQOYQATSI (Read error: 145 (Connection timed out))
00:14:34 Quit jgarvey ("Leaving")
00:15:11 Join NAQOYQATSI [0] (n=yukayant@
00:16:18 Quit Eosphere46 ("Copywight 2007 Elmer Fudd. All wights wesewved.")
00:17:20 Quit bmbl ("Bye!")
00:21:38***Saving seen data "./dancer.seen"
00:26:09bughunter2rasher: fsck.vfat says 22883 clusters are in use, that's ~90MB, which equals what explorer reports
00:26:15bughunter2perhaps the OF marks clusters as BAD?
00:27:54rasherBut the OF is doing something. Let's leave it at that. Prod the disk all you like
00:28:14 Quit ender (" I used to believe in reincarnation, but that was in a previous life.")
00:28:16bughunter2I'm thinking about checking dosfstools sources, to see how they calculate used clusters
00:31:25bughunter2think I know, it marks clusters in use in the FAT, but there are no directory entries pointing to those FAT entries
00:31:43bughunter2probably just to reserve space for itself
00:31:49bughunter2sounds reasonable..
00:32:07gevaertssounds broken...
00:32:13bughunter2that too
00:33:25TheSevenchkdsk would convert unused clusterchains to "FILExxxx.CHK" when running on a volume with such a mess
00:33:34bughunter2good point
00:33:36bughunter2and it didn't
00:33:54bughunter2so then that would leave one thing: they're marked as bad clusters?
00:34:09bughunter2that would mean the OF tries to hide some clusters from the 'outside world'
00:36:30 Quit stripwax ("")
00:36:56TheSevenbughunter2: have you tried a chkdsk /f in a command window? that should print some detailed statistics when it's done...
00:37:05TheSeven(including bad block count)
00:37:08bughunter2it prints "no problems found"
00:37:11bughunter2ah, good point
00:38:52bughunter2it doesn't list the bad sectors :o
00:39:00bughunter2it tells me nothing about that at all
00:39:45bughunter2hmm? chkdsk /v does list files I don't see in explorer
00:40:09bughunter2a directory called ##MUSIC#
00:40:21TheSevenwell, that could explain some things
00:40:36bughunter2which contains dirs Music, Photo, Playlists, Video, Service, Service\Rhapsody
00:41:05bughunter2and also ##PORT#\DeviceIcon.ico and ##PORT#\Object.dat
00:41:06TheSevencan you dump that directory entry, to see why windows doesn't show it?
00:41:08bertrikyou'll see those ## directories in rockbox too
00:41:10bughunter2mostly MTP stuff, I'd think?
00:41:27bertrikIIRC, they have some kind of attribute to make them hidden on most OSs
00:41:45bertrikcombination of attributes rather
00:41:46bughunter2an undocumented attribute bit, then, or?
00:42:03bughunter2because normally I see all files, even system+hidden+readonly etc.
00:42:51bughunter2the chkdsk from windows XP x64 won't show any bad block info though
00:43:06bughunter2bertrik: indeed, I see them in RB :)
00:43:37bertrikI think funman know more about this, I can also vaguely remember reading something about this in the ams sansa forum thread
00:44:44bughunter2I don't know the file size of these files, but I don't think they'd be ~90 MB however
00:44:55bughunter2well, 90-11(rockbox' size)=79, then
00:45:20TheSevencan you cd into these dirs?
00:45:27bughunter2and Debian couldn't show the files either
00:45:33TheSeventhat's funny
00:45:39*TheSeven is wondering what they did to them
00:45:52TheSevencan you dump the raw directory entries?
00:46:05bughunter2I'll make a raw dump in Debian in a sec
00:48:20TheSevenbughunter2: btw, windows tools for such a job would be rohpod or winhex
00:48:43bughunter2debian is easier for me I think
00:49:01bughunter2I'm using VirtualBox for it
00:49:09TheSevenwell, i wouldn't know how to fish out that directory entry on linux
00:49:26bughunter2I'm making a dd if=/dev/sda dump of it
00:49:30bughunter2then analyzing it on windows later :)
00:49:55TheSeveni would have accessed the disk directly using the tools named above...
00:50:14bughunter2hmm, I may have winhex laying around
00:51:38 Quit bertrik ("sleep")
00:51:45 Part froggyman
00:51:51 Join decayedcell_ [0] (
00:52:49bughunter2TheSeven: found them
00:56:29bughunter2they're marked as both DIR | VOL_ID
00:56:50 Part decayedcell_
00:57:42 Join Rob2222 [0] (
00:57:54bughunter2let's see if I get in trouble for changing the attribute
00:59:44 Quit faemir ("Leaving")
01:02:46TheSevenso they did a similar hack to what microsoft themselves did when implementing long file names in FAT
01:02:52 Quit Hillshum (Read error: 54 (Connection reset by peer))
01:03:00bughunter2it isn't related to long filenames?
01:03:18TheSevenlong file names would have even some more bits set in that field
01:03:59TheSevenLFNs are typicall readonly,system,hidden,volumeid or even readonly,hidden,system,archive,directory,volumeid
01:04:54bughunter2this explains all,
01:05:00TheSeventhey probably relied on the fact that everything that is volumeid won't be shown, while it won't be treated because of some missing bits
01:05:24bughunter2not sure what's in there, but it's from the OF
01:05:36bughunter2no real data in there yet, most likely it just allocated it but didn't write anything
01:05:50bughunter2since it couldn't do that in just a few milliseconds before rebooting into RB
01:07:08TheSevenis it a bug or a feature that rockbox shows files flagged as VOLUME_ID? how does it deal with an actual volume id entry that isn't abused?
01:07:22bughunter2it hides the first volume ID entry
01:07:33bughunter2I think
01:07:45bughunter2I wonder what happens if I remove all those files, then pump the whole disk full, then reboot the OF, then reboot into RB
01:08:01bughunter2let's find out
01:09:09TheSevenif the OF is mean, it will just reformat it :-P
01:09:16 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
01:09:22bughunter2that'd be really mean
01:09:33TheSevenor it just crashes
01:10:57bughunter2HxDen can also open disks btw
01:11:04bughunter2not a bad hex editor imo
01:11:15bughunter2using it a while now, just found out it can also open disks
01:11:27TheSevencan it compare to winhex?
01:11:34bughunter2for me, I think it can
01:13:23 Quit DarkDefender ("Leaving")
01:27:00 Join barrywardell [0] (n=barrywar@
01:29:39bughunter2ok, 0 bytes free, let's see what happens if I reboot it
01:30:04bughunter2TheSeven: "Not enough space for Music Db, please free 90 MB"
01:30:16TheSevenwow, they even caught it :-)
01:30:31bughunter2cool :)
01:30:38 Join fdinel [0] (
01:30:46bughunter2well that suffices for me, time to enjoy RB
01:31:06bughunter2or first, a good night of sleep..
01:31:13bughunter2hard to choose, but sleep is important
01:31:15TheSevenwill it be possible to break it by putting thousands of smallish files on it, so that those 90MB will turn out to not be sufficient? :-P
01:31:27bughunter2haha, who knows
01:31:37bughunter2I'll give that a shot too
01:33:21TheSevenwho knows, maybe they calculated those 90MB (which seem to be quite a lot for 4GB of storage) for a 512 byte cluster size with one file per cluster :-)
01:34:13bughunter2they probably figured customers wouldn't notice
01:34:24bughunter2which in general is probably true
01:34:34bughunter2but still lame
01:35:19TheSevenare we talking about memory cards here or about internal flash? i don't see why they didn't solve this by hiding the space from the PC in the first place, if it's the latter
01:35:47bughunter2probably this was cheaper
01:36:19TheSeventhey would just have needed to reduce the MSC sector count and write that stuff to the sectors after the MSC visible range
01:36:48TheSevenin fact, this hack was probably more complex to accomplish...
01:37:03bughunter2probably a bad design combined with time pressure
01:37:33 Join Hillshum [0] (
01:37:51 Quit kkurbjun (Read error: 110 (Connection timed out))
01:38:25bughunter2lost my program to spam a drive with thousands of files
01:38:37 Join Strife89 [0] (
01:40:37bughunter2TheSeven: it'd be fun to spam the root directory so they can't create the directories
01:41:08TheSevenis it FAT16?
01:41:11bughunter2FAT 32
01:41:41TheSeventhis has no limit on root directory size
01:41:56bughunter2so if I spam it with files until it can't create files anymore...
01:42:28TheSeventhis will be when it runs out of space - or out of ram while parsing the directory, if they didn't write proper code :-)
01:42:42bughunter2would be fun to test
01:42:43 Quit Thundercloud (Remote closed the connection)
01:42:52TheSevensince FAT32, the root directory is just an ordinary cluster chain like every other file
01:42:57bughunter2just for the hell of it
01:48:21 Join Briantheman [0] (n=4aa12378@gateway/web/cgi-irc/
01:48:44Brianthemanhello I need some help please
01:49:02HillshumWith what?
01:50:51BrianthemanHOw can I get the artwork of an album to show up in a rockboxed sansa?
01:53:01 Join Eosphere46 [0] (i=Eosphere@
01:54:46 Join maraz_ [0] (
01:54:51 Quit maraz (Read error: 131 (Connection reset by peer))
01:56:11 Quit Briantheman ("CGI:IRC (Ping timeout)")
01:57:11 Quit DerPapst ("Leaving.")
02:01:35 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
02:02:11 Part toffe82
02:02:12 Quit Hillshum ("Ex-Chat")
02:03:09 Quit robin0800_ (Remote closed the connection)
02:06:27 Quit Rob2222 (Remote closed the connection)
02:09:40 Join Rob2222 [0] (
02:10:39 Join JdGordon [0] (i=d036044e@gateway/web/freenode/x-rylxehvinrsxeiow)
02:14:34 Quit JdGordon (Ping timeout: 180 seconds)
02:15:34 Join Briantheman [0] (
02:16:09 Join JdGordon [0] (i=d036044e@gateway/web/freenode/x-cagrretmpxseuxfc)
02:17:51Brianthemanhello All
02:17:58 Join AndyIL [0] (n=pasha_in@
02:18:39BrianthemanI need some help please .....
02:19:24 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
02:20:44 Join Eosphere46 [0] (i=Eosphere@
02:20:45Brianthemanhow can i place artwork in a sansa rockboxed with the arboxwidgets?
02:21:42***Saving seen data "./dancer.seen"
02:21:52JdGordondid you read the AlbumArt wiki page? and probably the manual?
02:21:52Brianthemani need some help
02:22:34Brianthemanwhere is that?
02:22:47*Eosphere46 plays dramatic music....
02:24:23JdGordonmeh, hint..... embedded AA doesnt work
02:26:13 Quit mt (Read error: 104 (Connection reset by peer))
02:26:36bughunter2TheSeven: I just made a little file system spammer :)
02:26:39bughunter2testing it on the player now ;)
02:27:24 Quit JdGordon (Ping timeout: 180 seconds)
02:29:05 Quit AndyI (Read error: 110 (Connection timed out))
02:35:27*Strife89 wonders what a file system spammer is good for.
02:36:29 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
02:37:07bughunter2Strife89: to stress test :)
02:37:17bughunter2it creates files until the fs driver gives up
02:37:24Strife89Ah. :)
02:37:32 Join Eosphere46 [0] (i=Eosphere@
02:37:55bughunter2I ocne made XP machines in uni go crazy by spamming my %TMP% dir
02:38:27Strife89Creepy stuff. :)
02:38:35rasheroff topic
02:38:43bughunter2there was no topic :P
02:39:09rasherThere is. Please read the guidelines if you have any questions. In fact, you should already have done so.
02:39:22Strife89rasher: Techincally, on topic, because he's stress testing the Rockbox code.
02:39:38bughunter2it was tongue-in-cheek-style anyway, hope it didn't hurt anyone
02:39:44bughunter2I'm not a troller, just FYI
02:46:37 Quit Briantheman ()
02:50:38bughunter2night all
03:00:44 Quit Eosphere46 ("Give a man a fish and he will eat for a day. Teach him how to fish, and he will sit in a boat and drink beer all day")
03:07:37 Join notben [0] (
03:11:36 Join Llorean [0] (
03:18:55 Join froggyman [0] (
03:20:47 Join Harry1337 [0] (n=ad42aa59@gateway/web/cgi-irc/
03:21:14 Nick fxb is now known as fxb__ (
03:23:24 Quit Harry1337 (Client Quit)
03:23:28 Join harrison [0] (n=ad42aa59@gateway/web/cgi-irc/
03:24:52 Quit harrison (Client Quit)
03:25:10 Part notben
03:26:14 Join webguest35 [0] (n=ad42aa59@gateway/web/cgi-irc/
03:26:55 Join webguest14 [0] (n=ad42aa59@gateway/web/cgi-irc/
03:27:46 Quit webguest35 (Client Quit)
03:27:54 Quit webguest14 (Client Quit)
03:37:14 Quit barrywardell ()
03:41:58 Quit BlakeJohnson86 ("Leaving.")
03:42:12 Join BlakeJohnson86 [0] (n=bjohnson@
03:43:11 Join webguest08 [0] (n=ad42aa59@gateway/web/cgi-irc/
03:43:25 Quit webguest08 (Client Quit)
03:55:29 Quit bughunter2 ("Leaving.")
04:06:27 Quit TheSeven (Nick collision from services.)
04:06:43 Join The_Seven [0] (
04:06:51 Nick The_Seven is now known as TheSeven (
04:21:47***Saving seen data "./dancer.seen"
04:29:27 Nick truthtaco_ is now known as TruthTaco (
04:42:51 Quit Rondom (Nick collision from services.)
04:43:01 Join Rondom [0] (n=Rondom@
04:48:42 Quit froggyman ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
04:54:04 Nick JdGordon1 is now known as JdGordon (
04:57:05 Quit Strife89 (Read error: 104 (Connection reset by peer))
05:05:06JdGordonumm..... anyone around with rather random skin to testa quick patch?
05:07:12JdGordonrasher: ping?
05:11:10JdGordonscorche: ping?
05:41:29 Quit Horscht ("Verlassend")
05:45:48 Join TFGBD [0] (
05:55:39 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:12:52 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:16:16 Quit Zarggg ()
06:18:09 Quit panni_ ("( :: NoNameScript 3.81 :: )")
06:19:52 Join Zarggg [0] (
06:21:48***Saving seen data "./dancer.seen"
06:54:39 Quit JdGordon ("Leaving.")
06:56:32 Join matsl [0] (
06:57:23 Join JdGordon [0] (
07:01:27 Join AsaelReiter [0] (n=d59730fa@gateway/web/cgi-irc/
07:02:30 Quit tarbo_ (Connection timed out)
07:04:38 Join tarbo [0] (n=me@unaffiliated/tarbo)
07:07:10 Quit AsaelReiter ("CGI:IRC (EOF)")
07:08:11 Part safetydan ("Leaving.")
07:08:58 Quit AlexP (Remote closed the connection)
07:09:08 Join AlexP [0] (n=alex@rockbox/staff/AlexP)
07:13:00 Join AsaelReiter [0] (n=d59730fa@gateway/web/cgi-irc/
07:27:19 Join merbanan [0] (
07:31:09 Quit JdGordon (Read error: 104 (Connection reset by peer))
07:35:05 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
07:35:59JdGordonwhich would be better/more understandable/consistant for the wps? I want to add an escape char for \n so you can break up new lines... should it be % or \ ?
07:37:56 Join stoffel [0] (
07:40:40*JdGordon is tempte to go % because thats the standard escape char for skins
07:40:45JdGordonbut \ is more universally inderstood
07:40:49JdGordonunderstood even
07:41:50 Quit AsaelReiter ("CGI:IRC (EOF)")
07:47:26 Join tarbo_ [0] (n=me@unaffiliated/tarbo)
07:50:11pixelmahuh, I thought there already is one
07:51:04pixelmaah, if you use # you can also "mask" a newline with it
07:52:13JdGordonthe whole thing is to fix 9640
07:52:29JdGordonits remarkably simple to do, just trying to decide if its wanted
07:54:42 Quit tarbo (Connection timed out)
07:56:29JdGordon... doing a forum thread for this
07:56:59pixelmaI don't know. I thought this was a feature of # - you can easily put the comment on another line but don't have a strong opinion
07:57:40 Join Grahack [0] (
07:58:45pixelmaas long as viewports didn't exist (or you don't use them) you would have to put the # on the same line then if it doesn't "eat" the newline, otherwise you'll get blank lines there
07:58:55JdGordonright, so what I've done is let you put % at the end of a line to break long ones... that would work the same as putting #comment at the end of a line
08:00:12pixelmathen quite a few people would have to rework there WPSs just for that and always put a % at the end
08:00:19pixelmaof some lines
08:01:55JdGordonthey wouldnt if they already split their lines with #comment
08:08:57pixelmaI don't understand your last statement but have to go to work now
08:21:40 Quit BHSPitMonkey (Remote closed the connection)
08:21:49***Saving seen data "./dancer.seen"
08:23:46 Join ender` [0] (
08:25:27 Join flydutch [0] (n=flydutch@
08:27:17 Join Rob2223 [0] (
08:31:37 Join bertrik [0] (
08:42:57 Quit NAQOYQATSI (Read error: 104 (Connection reset by peer))
08:44:32 Quit Rob2222 (Read error: 110 (Connection timed out))
08:52:14 Join barrywardell [0] (n=barrywar@
08:58:48 Quit scorche (" rawr...that is all...rawr")
09:00:22 Join petur [50] (n=petur@rockbox/developer/petur)
09:00:34 Quit Grahack ("Leaving.")
09:12:43 Join scorche [50] (n=scorche@rockbox/administrator/scorche)
09:16:03 Join GodEater__ [0] (n=9372e2b4@rockbox/staff/GodEater)
09:22:10 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
09:22:24 Join Thundercloud [0] (
09:26:55 Join NAQOYQATSI [0] (n=yukayant@
09:30:28 Quit bertrik (Read error: 113 (No route to host))
09:37:22 Join Grahack [0] (
09:41:06 Join DarkDefender [0] (
09:42:32GrahackHi, I'm trying to call pcfbuf_beep() defined in /apps/pcmbuf.c from the lua plugin (/apps/plugins/lua/rocklib.c). Is it even possible? I added #include "pcmbuf.h" in rocklib.c but this is not enough (I get a "failed to load /.rockbox/rocks/viewers/lua.rock, dlopen(simdisk/.rockbox/rocks/viewers/lua.rock): simdisk/.rockbox/rocks/viewers/lua.rock: undefined symbol: pcmbuf_beep"). Do I guess right with something like "only functions d
09:46:27 Quit GodEater__ ("CGI:IRC")
09:50:49 Quit Thundercloud (Remote closed the connection)
09:59:04 Quit Grahack ("Leaving.")
10:01:16 Quit NAQOYQATSI (Read error: 104 (Connection reset by peer))
10:02:04 Join NAQOYQATSI [0] (n=yukayant@
10:06:08***SPY: Authentication failed for Zagor
10:06:12***SPY: Authentication failed for Zagor
10:06:24***SPY: Authentication failed for Zagor
10:08:19***Server message 901: 'logbot_ logbot n=bjst rockbox/bot/logbot :You are now logged in. (id logbot, username n=bjst, hostname rockbox/bot/logbot)'
10:08:48***Server message 902: 'logbot_ 0 n=bjst rockbox/bot/logbot :You are now logged out. (id 0, username n=bjst, hostname rockbox/bot/logbot)'
10:09:22 Nick logbot_ is now known as logbot (n=bjst@rockbox/bot/logbot)
10:09:22Mode"logbot :+e" by services.
10:09:54 Quit Llorean (Read error: 54 (Connection reset by peer))
10:13:10 Join DerPapst [0] (n=DerPapst@
10:21:26 Quit DarkDefender (Remote closed the connection)
10:21:54***Saving seen data "./dancer.seen"
10:30:14 Quit jon-kha (Read error: 145 (Connection timed out))
10:41:40 Join DarkDefender [0] (
10:43:52 Quit stoffel (Remote closed the connection)
10:52:02 Join Llorean [0] (
10:53:44 Join BulldogPO [0] (
10:54:11BulldogPOcan anyone help me with rockbox
10:55:46BulldogPOas Alpine car player and other things that does offer iPod compatibility doesn´t work with my iPod 80GB Classic when Rockbox is in use
11:03:33 Join Grahack [0] (
11:14:49 Join robin0800 [0] (
11:15:27 Join stoffel [0] (
11:16:43robin0800JdGordon: yes your "sublinefix" works and does indeed fix the problem many thanks
11:18:47 Quit robin0800 (Remote closed the connection)
11:21:38 Join robin0800 [0] (
11:25:12 Join jocon [0] (n=jocon@
11:28:04joconI am interested in makeing a simple app for my ipod, where can i find some info
11:31:53Grahack and
11:37:42 Join sbhsu_ [0] (
11:43:55 Join webguest0 [0] (n=52d8dede@gateway/web/cgi-irc/
11:45:10 Quit webguest0 (Client Quit)
11:46:04 Quit sbhsu (Read error: 110 (Connection timed out))
11:47:40 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
11:53:27 Quit stoffel (Remote closed the connection)
11:55:42 Quit bughunter2 (Read error: 54 (Connection reset by peer))
12:02:08 Quit jocon ("This computer has gone to sleep")
12:03:11 Quit Lss (Read error: 104 (Connection reset by peer))
12:08:38 Quit robin0800 (Remote closed the connection)
12:18:36 Join stoffel [0] (
12:20:35 Quit stoffel (Remote closed the connection)
12:21:56***Saving seen data "./dancer.seen"
12:26:50 Join pamaury [0] (n=pamaury@
12:27:21 Join mt [0] (n=mt@
12:31:27 Quit barrywardell ()
12:38:39 Join jocon [0] (n=jocon@
12:39:24 Quit BulldogPO (
12:39:24 Quit Llorean (
12:39:24 Quit scorche (
12:39:24 Quit Rob2223 (
12:39:24 Quit merbanan (
12:39:24 Quit scorche|sh (
12:39:24 Quit Guest34014 (
12:39:24 Quit bittin (
12:39:24 Quit dionoea (
12:39:24 Quit jvd (
12:45:17 Join robin0800 [0] (
12:50:49NJoinBulldogPO [0] (
12:50:49 Join Llorean [0] (n=DarkkOne@rockbox/user/Llorean)
12:50:49NJoinscorche [50] (n=scorche@rockbox/administrator/scorche)
12:50:49NJoinRob2223 [0] (
12:50:49NJoinmerbanan [0] (
12:50:49NJoinjvd [0] (n=syscrash@poipu/developer/syscrash)
12:50:49NJoindionoea [0] (n=dionoea@videolan/developer/dionoea)
12:50:49NJoinbittin [0] (
12:50:49NJoinGuest34014 [0] (
12:50:49NJoinscorche|sh [50] (n=scorche@rockbox/administrator/scorche)
12:52:28 Quit jocon ("Leaving")
12:56:28 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:02:19 Quit pamaury ("exit(*(int *)0 / 0);")
13:09:06Sajber^hey guys, i just wanna ask if there is any news on the Vision:M port...or is it still idle?
13:09:55krazykitany information would be on the relevant wiki page and forum thread
13:10:24Sajber^noting new under then sun on that.. :/
13:10:33dionoeais the source code in svn somewhere? I was wondering if it would be possible to have stuff like sorting themes by publication date, download count, rating, ...
13:11:58AlexPIt is
13:13:05AlexPI assume it is up to date
13:14:38dionoeathanks, I'll have a look
13:23:32 Nick fxb__ is now known as fxb (
13:32:41 Join decayedcell_ [0] (
13:33:11 Join jon-kha [0] (i=jon-kha@
13:35:07 Join barrywardell [0] (
13:49:31 Quit FlynDice (Remote closed the connection)
13:58:38 Quit robin0800 (Remote closed the connection)
14:04:49 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
14:08:42 Quit Zagor ("Don't panic")
14:19:40 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
14:20:30bughunter2Hey, can some sansa hacker tell me how to recover my e200v2? I don't have any 'orig_image.bin' laying around (as described in )
14:21:21bughunter2also, if I connect the device in recovery mode, the contents of the disk don't seem like they should
14:21:28bughunter2e.g. no .rockbox FAT root dir entry
14:21:59***Saving seen data "./dancer.seen"
14:23:23 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
14:24:12bughunter2neither can I find the 'SANSA E200P' volume ID root dir entry
14:25:26 Join Omlet [0] (
14:34:58 Quit Zagor (Remote closed the connection)
14:55:39bughunter2think I found a trick, I hex edited the pattern of the disktidy app to match all the files I want to delete from the root dir
14:55:42bughunter2it's now cleaning :)
14:57:56 Join GodEater__ [0] (n=9372e2b4@rockbox/staff/GodEater)
15:00:46bughunter2it would actually be need if one could edit the patterns inside rockbox, instead of having to hexedit or recompile :)
15:02:15bughunter2or totally cleaning a partition, like, only leaving .rockbox in tact and delete the rest
15:05:06 Quit thegeek (Read error: 104 (Connection reset by peer))
15:05:07 Quit antil33t (Read error: 104 (Connection reset by peer))
15:05:10 Join thegeek_ [0] (
15:06:12 Join AsaelReiter [0] (n=5f565c51@gateway/web/cgi-irc/
15:07:06AsaelReiterCan somebody have a look at FS #10589?
15:07:45 Join Zagor [0] (n=bjorn@
15:13:00 Quit barrywardell ()
15:13:27 Join antil33t [0] (n=Mudkips@
15:13:30dionoeaAsaelReiter: looks good to me. btw, the two collision detection blocks seems to be duplicates of one another. They should be factorised.
15:13:40dionoeaAsaelReiter: and you might want to remove the now unused functions
15:15:19dionoeaAsaelReiter: it looks like you have copy/paste errors in that patch (food_size and argh_size mixups)
15:15:50AsaelReiterI will check
15:16:36AsaelReiterAnd which functions I need to remove? (food_collision and argh_collision are still used)
15:17:39dionoeaah ok. Those other uses don't need the same fix?
15:19:22AsaelReiterI think that no
15:19:48AsaelReiterThis two function are good when you want to check one point.
15:20:13 Quit AsaelReiter ("CGI:IRC")
15:20:24 Join AsaelReiter [0] (n=5f565c51@gateway/web/cgi-irc/
15:20:52AsaelReiterAnd I think that all the food_size and argh_size are correct.
15:22:21 Join robin0800 [0] (
15:22:49AsaelReiter(When checking between a food and an argh, we use both of them. when checking between two foods, we use twice food_size)
15:24:49robin0800JdGordon: just to repeat many thanks for patch it compiles and works fine
15:25:13dionoeaah, right.
15:34:23 Join panni_ [0] (
15:45:46 Quit AsaelReiter ("CGI:IRC")
15:46:49 Join pamaury [0] (n=pamaury@
15:49:16 Quit bughunter2 ("Leaving.")
15:56:04 Join AsaelReiter [0] (n=5f565c51@gateway/web/cgi-irc/
15:58:32 Quit AsaelReiter (Client Quit)
15:58:48 Join AsaelReiter [0] (n=5f565c51@gateway/web/cgi-irc/
16:08:42 Quit tmzt (Read error: 110 (Connection timed out))
16:11:08 Quit Omlet ()
16:21:03 Quit AsaelReiter ("CGI:IRC")
16:22:02***Saving seen data "./dancer.seen"
16:24:11 Quit Rob2223 ()
16:25:38 Quit robin0800 (Read error: 110 (Connection timed out))
16:27:09 Join webguest30 [0] (n=70c91b03@gateway/web/cgi-irc/
16:28:25 Join Lss [0] (
16:28:57 Quit webguest30 (Client Quit)
16:30:39 Join Rob2222 [0] (
16:39:30 Quit Grahack ("Leaving.")
16:42:52 Part decayedcell_
16:44:46 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-gjsutvewrrjhdbbi)
16:45:42 Quit mt (Read error: 113 (No route to host))
16:49:01 Quit evilnick (Client Quit)
16:52:47 Join tmzt [0] (
16:53:19 Join toffe82 [0] (n=chatzill@
17:02:12 Join PaulJam [0] (
17:03:06 Quit Zagor ("Don't panic")
17:05:46 Join mcuelenaere [0] (
17:05:58 Join Omlet [0] (
17:09:50 Join jgarvey [0] (
17:14:01 Quit liar (Read error: 113 (No route to host))
17:14:19 Join liar [0] (n=liar@
17:21:36 Join robin0800 [0] (
17:30:18 Join Blue_Dude [0] (
17:33:57 Quit robin0800 (Read error: 104 (Connection reset by peer))
17:34:59Blue_DudeI posted this the other day: I was wondering how the status bar "knows" exactly what's being played at the moment? I figure if the status bar "knows", then there might be a way to let the other display information know so that the WPS screen stays in sync.
17:38:02 Join barrywardell [0] (n=barrywar@
17:40:51saratogawhat gets out of sync in the WPS?
17:42:27CIA-43New commit by jdgordon (r22668): fix the first subline not displaying properly
17:43:18 Quit petur ("later")
17:45:59Blue_DudeThings get weird just before a track change. The playlist number changes early. The WPS seems to lock up if you pause just before the end of the track then try to change tracks. Basically I want to know how the GUI knows exactly which samples are being played by the hardware given that the samples were buffered several seconds before.
17:47:22Blue_DudeIf there's a way to track exactly what's going out the hardware, then there ought to be a way to lock the WPS more solidly so that these kinds of glitches stop happening.
17:48:51saratogai'm not familar with this code but it sounds like the playback engine is updating its state cleanly enough, not that we forget to update the WPS
17:50:04Blue_DudeYou meant "the playback engine is NOT updating cleanly enough..."?
17:51:04JdGordonthats a know issue
17:51:31JdGordonthe ~2s between when the track is technically finished being decoded and actually spat out the DAC is a wierd time
17:52:20Blue_DudeIt's a known issue, but is it fixable?
17:52:26saratogasorry "not updating"
17:52:34saratogabasically our playback engine is a mess
17:53:07Blue_DudeSo while we're at it, what else is busted?
17:53:25Blue_Dudewith playback?
17:53:36JdGordonyou dont really want to open that particular can...
17:53:42JdGordonlest the worms get out!!!
17:54:11Blue_DudeOh yes I do. If I'm going to wade in, I ought to know what I'm getting myself into.
17:54:36JdGordonI want to say search the logs... its been discussed plenty, but I couldnt tell you when or for what
17:55:21JdGordonthat particular problem is because of the way playback,buffering,playlist and the GUI all "work" together
17:55:25Blue_DudeI'll try but I suck at running searches. I never know what to search on.
17:55:35 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:57:01 Join Riku [0] (
17:58:25 Quit Lss (Read error: 54 (Connection reset by peer))
18:02:38saratogaBlue_Dude: theres been lots of discussion about playback, but mostly just vague stuff, I've never seen anyone lay out exactly what goes wrong, so I assume no one has figured it out entirely
18:02:58saratogai think its mostly just that a lot of cases on track changes aren't handled cleanly
18:03:25Blue_DudeWell, it vexes me. Must fix. :)
18:03:26saratogaincidently, if you happen to notice why playback locks up sometimes on low memory targets, please let us know :)
18:03:54Blue_DudeI don't have one, but if I come across something promising I'll mention it.
18:05:49saratogai think if you compile any target so that it has about 2-3MB total memory for rockbox + buffering it will lock up playback randomly
18:05:51 Join barrywardell_ [0] (n=barrywar@rockbox/developer/barrywardell)
18:05:57saratogaany software decoding target that is
18:06:01JdGordonvery quickly.... my beef with the system is that track changes (and where the "current track" is defined) all happen through playback.c... I'm thinking playlist shuld handle this
18:06:37 Quit barrywardell_ (Read error: 104 (Connection reset by peer))
18:06:49 Join barrywardell_ [0] (n=barrywar@
18:07:08JdGordonit shuold be... playlist knows track X is playing... playback/buffering will buffer X+N tracks untill its full, each time a track finishes actually being played it tells playlist that its finished which then lets the GUI know
18:07:33saratogawhat does playback.c handle then?
18:07:36JdGordonsame thing for forced track changes... the GUI tells playlist which then yells at playback
18:07:48Blue_DudeWell, how does playback know what's being played at the moment? It has to have some kind of feedback from the pcm handler.
18:08:35JdGordonplayback.c is the glue for the codecs and the actual audio stuff... it wouldnt change much... playlist would have the responsibility of keeping track of where in the playlist we actually are though
18:08:53JdGordonBlue_Dude: it does
18:09:10 Join AsaelReiter [0] (n=d44c6ebe@gateway/web/cgi-irc/
18:09:52JdGordoniirc it knows when a track is finished decoding, and when a track is finished through PCM... the problem you saw is because when it finishes decoding it tells playlist (wrongly) that the track changed
18:10:01JdGordonor something like that.. I havnt looked at it in a while
18:10:34Blue_DudeSo where's the problem? Playback? Or playlist?
18:11:19Blue_DudeThat narrows it down to about 4000 lines of code or so.
18:11:45 Join archivator [0] (n=Delyan@
18:17:42JdGordon... with all that said.. playlist.c should be split up into playlist generation, playlist manipulation and *current* playlist handling
18:20:24Blue_DudeFrak me. playlist.c is already 3500+ lines. I was kidding...
18:20:53 Join Strife89 [0] (n=nds@
18:21:17 Quit AsaelReiter ("CGI:IRC")
18:21:28 Quit barrywardell (Read error: 110 (Connection timed out))
18:21:29 Nick barrywardell_ is now known as barrywardell (n=barrywar@rockbox/developer/barrywardell)
18:22:05***Saving seen data "./dancer.seen"
18:25:23Blue_DudeIs playlist.c really the place you want for keeping track of which track is playing?
18:30:18 Join karltpb [0] (n=3b5eed7a@gateway/web/cgi-irc/
18:31:04 Join karltpb1 [0] (n=3b5eed7a@gateway/web/cgi-irc/
18:32:59 Join JdGordon_ [0] (i=441b6224@gateway/web/freenode/x-mixslabqivfpqoiq)
18:33:25karltpb1Hi I'm trying to compile checkwps on windows using cygwin but I get an error
18:33:37 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-vbkuorsmtrjpybcx)
18:33:41karltpb1LD checkwps.XXXXXXX.exe cc: no input files make: *** [checkwps.XXXXXXX.exe] Error 1
18:35:09JdGordon_do you only want it to for target?
18:35:42 Quit karltpb ("CGI:IRC (Ping timeout)")
18:36:48 Quit antil33t (Read error: 104 (Connection reset by peer))
18:37:02 Join antil33t [0] (n=Mudkips@
18:40:45karltpb1yeah but i'd love to compile it myself
18:42:18 Quit matsl (Read error: 110 (Connection timed out))
18:45:37 Quit jon-kha (Read error: 148 (No route to host))
18:49:18JdGordon_so use the configure option for that target
18:55:07karltpb1just tried that and it gives me errors too
18:55:16Blue_DudeHey, why does the GUI lock up at the end of a playlist? The screen returns to the main page, but the scroll wheel doesn't work. I have to touch the power button to wake it up.
18:56:13karltpb1$ make -j Generating dependencies CC apps/gui/skin_engine/wps_debug.c CC apps/gui/skin_engine/skin_parser.c CC apps/gui/skin_engine/skin_buffer.c CC apps/misc.c CC apps/recorder/bmp.c CC firmware/common/strlcpy.c CC tools/checkwps/checkwps.c In file included from /home/scarr/rockbox/apps/recorder/bmp.c:46: /home/scarr/rockbox/firmware/export/debug.h:27: error: parse error before "ATTRI BUTE_PRINTF"
18:56:15 Quit merbanan (Read error: 110 (Connection timed out))
18:56:37JdGordon_arg... it might be broken again
18:58:07karltpb1hang on let me try with the latest revision
18:58:50 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:59:22 Join bertrik [0] (
19:00:04karltpb1well I get the same error on revision 22668
19:00:32Blue_DudeWhich target are you trying to compile for? I'll try it here.
19:00:34JdGordon_it might be broken
19:00:42karltpb1ipod video 64mb
19:01:08JdGordon_use the sim with −−debugwps
19:02:01 Quit Strife89 ("To work.")
19:02:51Blue_DudeYeah, broke for me too.
19:03:16 Join Maltin [0] (
19:03:26Maltinhey guys
19:04:11mooslike that was proposed here, maybe we should have checkwps on the build table
19:06:04Blue_DudeThe checkWPS build really doesn't like the ATTRIBUTE_PRINTF tag. Or maybe it chokes on the "..." variable parameters.
19:07:52 Join stoffel [0] (
19:08:22karltpb1I'm unable to figure out this error:
19:08:47karltpb1Pasted the output from the simulator −−debugwps
19:08:58 Quit pamaury ("exit(*(int *)0 / 0);")
19:10:48JdGordon_there is no %xl|d|
19:10:56JdGordon_its breaking on %xdd
19:13:09karltpb1ah thanks
19:13:46karltpb1it's reporting the wrong line no though
19:13:50bertrikmy latest fix for the si4700 tuner chip in the sansa c200v2 didn't help much :(
19:13:53*domonoky wonders why the parser says line 41, and %xdd is on line 39 in this pastebin.
19:13:56 Quit JdGordon_ (Ping timeout: 180 seconds)
19:14:25domonokyand the error message could be better.. :-)
19:19:07 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
19:23:04 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-xeutmlsjunjxttse)
19:23:31 Quit karltpb1 ("CGI:IRC (Ping timeout)")
19:24:44JdGordon|yeah, I've been looking into making the debug output better
19:26:12JdGordon|I'm not sure why the line is 2 off though...
19:27:41JdGordon|it was worse 2 days ago when it didnt even say which image was being displayed... so at least there its easier to find the problem
19:28:03JdGordon|on that topic.... ping rasher
19:28:07JdGordon|or scorche
19:28:23JdGordon|or anyone who knows the theme site backend
19:28:44*domonoky thinks it would be very helpfull, if it would also output the token where it fails. ie Error on line 42, token %xdd.
19:31:15pixelmamoos: nice idea, but you compile checkwps for a certain target, would you add a column for each (and it can really differ in screen dimensions and depths and feature sets, e.g. virtual led etc.)?
19:31:27 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
19:31:47JdGordon|pixelma: checkwps is so fast I would suggest just adding it to the end of the sim build
19:31:54JdGordon|or the bootloader build
19:32:03JdGordon|it compiles llike 5 files only
19:32:09pixelmaa combined column?
19:33:02JdGordon|*or* get the mass checkwps going again and add a column for it, although its output will be very annoying because it wont be easy to see if its the same error for each target, or different ones
19:34:42pixelmahmm.. can you compile a sim *with* checkwps - does that even make sense?
19:34:57 Join zzy [0] (
19:38:53 Quit stoffel (Read error: 113 (No route to host))
19:39:27 Join Grahack [0] (
19:39:28 Join stoffel [0] (
19:39:40*domonoky thinks we should just add checkwps normally like the sims. And maybe change the buildtable orientation again.. (as the the table gets very wide)
19:40:26 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:41:13 Quit DataGhost (Nick collision from services.)
19:41:21 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
19:45:45zzyHi, I'm using the Rockbox 3.3 and and Boot v5.0 with a Sansa e260. When my e260 is off and I plug it in via USB it goes to the original Sansa firmware, is this expected?
19:47:21AB3JUzzy: there's a bootloader on flyspray that corrects that behavior
19:48:01AB3JUI'm not sure if anyone ever built it into sansapatcher though, so you may have to load it with... -a? can't remember, not at home right now
19:48:03JdGordon|domonoky: I dont think adding a new colum for every target is very helpful
19:50:02domonokyJdGordon|: but what are the alternatives ? a combined (either sim+checkwps or all checkwps builds) row/colum is also not really helpfull, as it will make it hard to find the errors.
19:50:35zzyAB3JU: thank you very much for the reply, know the FlySpray #?
19:51:00domonokyand if we change the orientation of the table again, it wont hurt so much to have another row.
19:51:16JdGordon|I think sim+checkwps does make sense... as long as the gap is obvious... remember checkwps is a tiny compile and wont break nearly as often as anything else (unless there is lot of skin changes like now)
19:51:17gevaertsAB3JU, zzy: please don't use the one from flyspray. It's horribly outdated
19:51:47gevaertsThere are candidate-official bootloaders at
19:52:14zzygevaerts: could you tell me, is this behavior expected and now being changed? or is what I'm experiencing a bug?
19:52:33AB3JUzzy: it's the old behavior from before the sansa had USB support enabled in rockbox
19:52:50gevaertszzy: AB3JU's answer is correct in every way, except that flyspray is not the right place to get a new bootloader
19:53:37AB3JUgevaerts: what's new in the new bootloader from fs?
19:53:42zzygevaerts & AB3JU: thanks... also if you guys have the FlySpry #, it would be cool, so I could follow the progress of the bug...
19:54:17AB3JUzzy: I want to say it's something along the lines of 7788 or 7878 or something like that
19:54:57gevaertsAB3JU: please stop referring to that flyspray bootloader. It is outdated, and the task will not be updated
19:56:55 Quit Grahack (Read error: 113 (No route to host))
19:56:55 Quit GodEater__ ("CGI:IRC (EOF)")
20:02:39 Join petur [0] (n=peter@rockbox/developer/petur)
20:02:57*GodEater had no idea the bug with ipod shutdowns happened so regularly for some people :(
20:04:02 Join jon-kha [0] (i=jon-kha@
20:04:44 Join Thundercloud [0] (
20:04:59 Join Grahack [0] (
20:06:45 Quit evilnick (Ping timeout: 180 seconds)
20:08:45 Join stoffel_ [0] (
20:12:17 Nick fxb is now known as fxb__ (
20:12:42 Join Strife89 [0] (n=michael@
20:15:35 Join vedos [0] (n=draft@
20:21:07 Quit Grahack ("Leaving.")
20:22:09***Saving seen data "./dancer.seen"
20:22:29zzygevaerts & AB3JU: I installed the new bootloader, and the Rockbox charging screen now appears when plugging in my e260 when off, cheers and thanks
20:25:25zzyis there a way to disconnect the USB through Rockbox, I read about a USB disconnect feature, but I don't know how to use it?
20:26:11gevaertsthere's no way to disconnect, no. The best you can do is plug in without a mass storage connection
20:28:40 Join stripwax [0] (
20:32:28 Join bluebrother [0] (
20:32:48 Join faemir [0] (n=faemir@
20:32:53zzygevaerts: oh, what does this flyspray bug mean, it discusses a usb disconnect feature:
20:33:51gevaertszzy: hm, I apparently forgot to close that. It explains why the disconnect handling was removed
20:35:44 Quit Strife89 ("Huzzah!")
20:36:15 Join Jaykay [0] (
20:37:57 Join gregzx [0] (
20:38:46zzygevaerts: thanks, you've been super helpful
20:47:58 Join FlynDice [0] (
20:49:16 Quit amiconn (Nick collision from services.)
20:49:19 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
20:49:22 Quit pixelma (Nick collision from services.)
20:49:24 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
20:49:39 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
20:49:43 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
20:50:07 Quit PaulJam (".")
20:53:01zzyCan playlists be created on a PC that include files on both the built-in storage and SD card?
20:53:23 Join Grahack [0] (
20:54:16saratogaIn theory yes, but I don't know if theres any software that can do it automatically without needing to find and replace the output
20:55:29 Join p3tur [50] (n=petur@rockbox/developer/petur)
21:01:40 Quit Omlet ()
21:03:10amiconnIirc it should work without search-and-replace
21:03:46 Quit flydutch ("/* empty */")
21:11:11JdGordon|drive mappings are different though...
21:11:31JdGordon|d:\ (or /sdcard) on the comp isnt even close to <microSD1>
21:12:05JdGordon|scorche|sh: how much does the themes backend rely on the checkwps output... I want to rearrange it but couldnt find where its used by the site
21:13:08 Quit BHSPitMonkey (Remote closed the connection)
21:14:13amiconnIirc the playlist code strips drive letters as well as top-level dirs if the file isn't found
21:14:45p3turbah... on h3x0, in the recording screen, the rec button no longer starts a recording... wtf?
21:15:00amiconnHmm, probably it won't add the <vol> dir though
21:15:18JdGordon|yeah, I didnt think it was that smart
21:16:39amiconnStripping is necessary for all targets, either the drive letter (playlists made on windows), or one or more top-level dirs (linux - the mount point)
21:17:12amiconnAdding one would only be necessary on multivolume - maybe that should be added
21:17:36 Quit stripwax ("")
21:19:27amiconnThe strip code once confused me a bit. I had the same album both on an MMC (i.e. /<MMC1>/genre/artist/album) as well as on the internal storage (just /album). I made a playlist from MMC contents only and started it. Then I stopped, pulled the MMC, and resumed. Guess what happened
21:20:15gevaertsstrange things?
21:20:44gevaertsit probably Just Worked
21:24:54 Quit stoffel (Remote closed the connection)
21:25:35Jaykaysaratoga: was there any progress in the supported vs. unsupported builds discussion since your last mail?
21:29:23p3turoy... why did MarcGuay remove the 'start new file if not recording' functionality from ACTION_REC_NEWFILE ???
21:29:44p3tur(revision 21034)
21:31:12p3turcan anybody remember why he did that?
21:31:33 Join AsaelReiter [0] (n=d59730fa@gateway/web/cgi-irc/
21:32:03domonokyJdGordon| take a look at private/themesire.class.php, checkwps is used there. And it looks like it only really uses the return value of checkwps (the text is presented to the user if uploading fails).
21:32:06AlexPThat doesn't say why, but was from the same time
21:32:40JdGordon|domonoky: ok thanks
21:32:49JdGordon|so I'm free to fiddle?
21:33:00AlexPpetur: And this is why apparently:
21:33:09AlexPIt caused issues on the nano or something
21:33:32domonokyJdGordon| i am not 100% sure, but it looks like its fine to fiddle with checkwps output.
21:34:02p3turAlexP: thanks
21:35:50*AlexP notes he doesn't have a little Rockbox logo in the irc log viewer
21:36:02*AlexP flicks zagor :)
21:36:54p3turanother fine misunderstanding.... I actually said I was OK with not having duplicate functionality on two keys, I never said I was OK with removing the REC key (which is later claimed to be what I said). argh
21:37:26 Join Lear [0] (
21:38:00*p3tur blames pixelma
21:42:33 Join froggyman [0] (
21:44:28 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
21:44:29 Join matsl [0] (
21:46:10saratogaJaykay: not that i know of
21:46:21saratogasomeone who can change the front page needs to do that
21:46:50Jaykaysuch a good idea
21:47:00Jaykayevery detail is in the mail
21:47:04gevaertssaratoga: the front page is in svn
21:47:05 Quit vedos ("Rockbox roxx!")
21:47:20Jaykayand nobody does it
21:50:49saratogaso we can edit it in SVN and then get bagder to upload to the webserver?
21:51:07AlexPas I understand it, yes
21:52:17 Quit Grahack ("Leaving.")
21:53:09saratogai guess this should be a patch on the FS then?
21:53:39gevaertsyou can just commit it. It's just that one of the Swedes has to run svn update on the webserver
21:57:27 Quit LambdaCalculus37 ()
21:57:57zzyamiconn: I don't think the drive letter stripping is occurring with my playlists
22:01:41 Quit p3tur (Remote closed the connection)
22:02:02 Join p3tur [50] (n=petur@rockbox/developer/petur)
22:04:41saratogashould the front page link to the actual wiki pages for unstable devices (e.g. SansaAMS) or just the TargetStatus page?
22:04:42 Quit moos (Read error: 145 (Connection timed out))
22:05:45 Quit BlakeJohnson86 (Read error: 145 (Connection timed out))
22:05:56 Quit AsaelReiter ("CGI:IRC (EOF)")
22:06:45gevaertsI think it should link either directly to the wiki pages, or to a restricted status page that only has information about the unstable devices (i.e. not the unsupported ones)
22:07:24gevaerts"the wiki pages" not necessarily being the standard port pages. Those tend to be a bit too technical and inconsistent for this I think
22:08:16saratogathe idea was that each wiki page would be cleaned up as requirement for being considered "unstable" so that it resembled the SansaAMS page right now
22:08:54gevaertssounds good
22:09:57gevaertsalthough the SansaAMS page is about what it says, i.e. the Sansa AMS devices, so not only the ones we would call supported
22:10:23gevaertsthat can be easily fixed though. An extra column in the table would do it I think
22:12:40 Join stripwax [0] (
22:13:16*pixelma whistles innocently
22:13:46pixelmaI wonder what petur blames on me
22:14:40p3turon May 22 you said I was OK with REC no longer starting a recording
22:15:04p3turI had only said I was OK when there were not two buttons to start recording
22:15:20p3turremoving REC is a bit strange
22:15:35p3turand people now start complainging of course...
22:16:15saratogagevaerts: first shot at updating index.t:
22:16:25saratogano idea how correct that is since i don't know HTML
22:17:17 Join Superandy [0] (n=dcf417db@gateway/web/cgi-irc/
22:17:17 Join robin0800 [0] (
22:17:34Superandydid you see linux4nano
22:17:39Superandythey got the dual boot for the 2g
22:17:51 Quit Superandy (Client Quit)
22:18:19gevaertssaratoga: I think that looks good
22:22:13***Saving seen data "./dancer.seen"
22:23:08 Join stripwax_ [0] (
22:26:25 Join BlakeJohnson86 [0] (
22:32:37 Quit robin0800 (Remote closed the connection)
22:39:41 Quit stoffel_ (Remote closed the connection)
22:40:15 Quit Lear ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
22:41:23 Quit barrywardell ()
22:41:26 Quit stripwax (Read error: 110 (Connection timed out))
22:42:21 Join Zagor [242] (
22:45:15 Quit stripwax_ (Read error: 54 (Connection reset by peer))
22:45:38bluebrotherREC doesn't start recording anymore? Did I miss something?
22:45:48 Quit Jaykay ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
22:47:49 Join stripwax [0] (
22:47:59 Join domonoky1 [0] (
22:49:08 Quit n1s (Read error: 110 (Connection timed out))
22:49:53pixelmaI said it was ok if Rec no longer starting recording - on an H100?
22:50:10pixelmaor H300?
22:50:43gevaertspixelma: read the bit of IRC log that AlexP pointed to. It's all a bit confused
22:51:57saratogaanyone else want to comment on this before I send it:
22:52:43pixelmajust read it - I did not speak about the use of the Rec button just about the "rec_newfile" action, I didn't know which of the two was mapped to the Rec button on these Irivers
22:53:03 Join madnick [0] (n=57a0ac81@gateway/web/cgi-irc/
22:53:16AlexPsaratoga: Looks good - do we want a link from the unstable devices line to somewhere?
22:53:21JdGordon|saratoga: you didnt put the mr500 and beast in supported?
22:53:29AlexPLike a device status page or something
22:53:44gevaertsAlexP: the individual lines link to such pages
22:54:00saratogaJdGordon|: the Beast needs cleaned up install directions
22:54:05Zagorhas the "devices" word been discussed? would "stable ports" be a better wording?
22:54:06saratogai don't know about the mr500
22:54:06AlexPgevaerts: For unstable?
22:54:25AlexPgevaerts: Er sorry, I meant unusable
22:54:36AlexP@ saratoga too
22:54:39pixelmagevaerts: the confusion starts with the fact that there are/were two ways to start a recording (except on the OndioFM IIRC because it doesn't have one of the two due to button shortage :P )
22:54:51gevaertsAlexP: there is a link on the next line
22:55:20AlexPgevaerts: Ah right, I didn't see that as it was after the Quick links heading
22:56:05saratogawe don't currnetly split up the devices between unstable and unusable on the wiki, so I didn't want to have seperate links
22:56:11saratogabut that could be fixed if people prefer it
22:57:01saratogaadding a link to ContributingToRockbox might be nice too
22:57:09AlexPsaratoga: I don't mind, I just didn't see the link at all
22:58:25 Quit madnick ("CGI:IRC")
22:58:27saratogashould "Stable Devices" have a description?
22:59:32 Quit bmbl ("Bye!")
23:01:27*Zagor predicts confused questions about why sansa fuzes are unstable
23:01:32*domonoky1 would prever "Port" instead of "device", its not the device which is unstable, its the code
23:01:41JdGordon|sometyhing like "The hardwrae manufacturers are embarrased by our amazing skill on these players...."
23:02:15Zagordomonoky1: me too
23:02:50 Quit domonoky (Read error: 110 (Connection timed out))
23:03:47 Join n1s [0] (n=n1s@rockbox/developer/n1s)
23:04:41domonoky1Zagor: i found a few problems in the buildmaster: bestfit_builds() will segfault if $totspeed over all clients is 0. And it never ends if there are builds which no client can build. But i dont know how to best fix that :-)
23:05:41Zagordomonoky1: ah, well hunted! I'll fix both bugs.
23:06:30domonoky1Zagor: i did not hunt them, i fell overthem when i tried to run a buildmaster locally :-)
23:06:41Zagorthat's not nearly as impressive :-)
23:08:08JdGordon|Zagor: dunno if you read the logs or not... we want to get checkwps going in the build table (its a configure oiption for targets now)....
23:08:28JdGordon|I feel the best way to do it is tack it onto the end of the targets sim build....
23:09:06gevaertsJdGordon|: if it's a configure target, it should be possible to just add them to builds
23:09:21JdGordon|thats another 40 odd columns though
23:09:24Zagorgevaerts: yeah but that adds ~50 new 1-second builds to the table
23:09:33domonoky1another not so nice thing in the buildsystem: the checks for tools like gcc,curl etc are done by checking the "which" output for a beginning slash. this will ofcourse fail on windows :-)
23:09:54rasherdomonoky1: not on cygwin..
23:09:57Zagordomonoky1: windows? what's that?
23:10:18gevaertsZagor: yes, but do we want independent builds to be listed together?
23:10:18*domonoky1 wants rbutil built automatically :-)
23:10:26*JdGordon| ditto's that
23:10:37domonoky1and for this i want the buildclients to run on win/lin/mac nativly :-)
23:10:45Zagorgevaerts: no, that's not terribly nice either. and requires changing the format of the builds file. I'm undecided.
23:10:47*rasher voices a vote for an "all checkwps targets" build, with one column
23:10:54*gevaerts agrees with rasher
23:11:13rasherIt wouldn't clutter the build table, and it should be pretty easy to read the log
23:11:21gevaertsor we can pick a few targets. That should catch most of the problems
23:11:30JdGordon|no, it will be a huge PITA to fix yellow/reds in it
23:11:38rasherJdGordon|: Why?
23:11:39JdGordon|gevaerts: 's idea is closer to workable
23:11:55domonoky1picking targets is dangerous, the selection is prone to get outdated...
23:11:56pixelmahow will that find out if WPS don't work on certain targets like with using backdrop and album art tags etc.?
23:11:58JdGordon|because it will spit out errors for every target... they may not all be identical
23:12:10Zagorgevaerts' idea works already. the only negative is the build table bloat
23:12:16gevaertspixelma: I think it's likely that the other builds will catch that
23:12:25JdGordon|we shold be able to work out which 5 builds need doing
23:12:33rasherJdGordon|: So what? You fix the errors it outputs.. And if you missed some, that's no different than what happens today
23:12:35JdGordon|to cover 90% of the options
23:12:41pixelmaor better: a part of checkwps that is "responsible" for checking that
23:13:11JdGordon|rasher: sure, but its the difference between 1 "fix red" ci and 5
23:13:17*gevaerts already found one
23:13:24gevaertscheckwps doesn't build for the Player
23:13:35JdGordon|it doesnt build at all again apparently
23:13:37rasherJdGordon|: I seriously doubt this is an issue, honestly.
23:13:59*domonoky1 thinks we should build checkwps for all targets. the presentation in the build-table is to be discussed :-)
23:14:01gevaertsJdGordon|: it builds for c200 here
23:14:07rasherBesides, picking only a few targets is bound to miss the targets where it breaks.
23:14:29*rasher is with domonoky1
23:14:47JdGordon|wps editor is actually the more interesting project imo
23:14:48rasherIt would be foolish to only build a selection on the hope that we picked the right ones
23:16:39 Join robin0800 [0] (
23:16:51 Quit freqmod (Remote closed the connection)
23:16:58 Join freqmod [0] (
23:17:40pixelmaJdGordon|: I thought your patch changes the behaviour of # always
23:19:30JdGordon|not intentioannly
23:19:39pixelmayour suggestion now that # will have a different meaning whether it is at the beginning of the line or at the end sounds even worse (and more complicated code-wise as well as understanding it)
23:20:13 Join Eosphere46 [0] (i=Eosphere@
23:20:30 Nick fxb__ is now known as fxb (
23:20:34rasherThe normal use of # is "everything after it is ignored". Problem is that in WPS code, a blank line is significant, whereas it usually isn't. For that reason, it requires special handling when a line is nothing but comment
23:20:47JdGordon|are you takling about my suggestion of using # instead of % to keep eat the \n?
23:21:08*gevaerts suspects that there is a subset of targets that covers all of checkwps, but he's starting to agree that it will be hard to get that subset right
23:21:08rasher"whereas it usually isn't" = in say, shell scripts or perl or config files
23:21:50rashergevaerts: That's likely true, but a silly task, if you ask me, when just building all of them is merely a problem of representation
23:22:07gevaertsrasher: that's what I'm starting to agree with
23:22:35Zagora problem I already fixed once, btw, by simpy filtering away all-green columns
23:22:39gevaertscollapsing greens on nozip builds sounds like a better solution
23:22:52JdGordon|player, recorder, h100, ipod video, clip.... that should be 99% of the posibilities
23:23:03rasherJdGordon|: What do you gain though by this bet?
23:23:15JdGordon|well nothing really
23:23:23JdGordon|easiier output to parse
23:23:25pixelmarasher: I don't think it needs special handling - and LaTex comments behave similarly
23:23:42rasherJdGordon|: You can just only look at the player, recorder, h100, ipod video, clip logs then..
23:23:59rasheror "parts of the log", if it turns into one log
23:24:06rasherJdGordon|: What?
23:24:30JdGordon|"roll eyes and groan" is longer to type
23:24:42pixelmaJdGordon|: I talked about "comments at the end of the lines *dont* eat the \n, comments which are the whoel line do eat the \"
23:24:53gevaertsThe only problem I see with the don't-show-greens system is that you lose the links to the downloads
23:25:27rasherJdGordon|: You'll have to explain. You just agreed that there is no good reason not to test them all. When something breaks, you go look at the logs that you think are needed to fix the problem
23:25:32rasherSame as any other build
23:26:10*JdGordon| is so bloody confused about the comment stuff
23:26:35rasherI just don't agree that only building some makes anything any easier at all
23:26:45 Quit freqmod (Remote closed the connection)
23:26:46rasherexcept perhaps for Zagor, but since when do we care about him?
23:26:51 Join freqmod [0] (
23:26:54pixelmaJdGordon|: which would mean that you have to know that # behaves differently whether you use it on a line on its own or not. By the way I wouldn't use # for inline comments after some other WPS code
23:26:55JdGordon|rasher: if everything is in one long log, a single warning is repeated X times... its very hard to figure out if its all the same one or multiple
23:27:16rasherJdGordon|: That's not different than the same warning repeated in X files.
23:27:54rasherEven then, you could have the checkwps build logs in separate files even though it shows up as one column, if you like, I guess.
23:28:00JdGordon|yes it is... right now if there is yellow 3 on every single column you can be sure (well almost) thats its all the same... if one has 5 though you can easily spot the difference
23:28:04gevaertsHas anyone ever considered just splitting the table?
23:28:20Zagorgevaerts: blasphemy!
23:28:50Zagoruh, no that's not the right word :)
23:29:27JdGordon|pixelma: like I keep saying... I dont write much wps code because i cant make any sense of it... so I'm trying to make it easier for those that do.... I would think that splitting long lines into mutliple shorter ones with comments would be nice
23:29:47Zagorgevaerts: that's more in line with my fuzz, yes
23:30:05 Quit archivator (Read error: 54 (Connection reset by peer))
23:30:12pixelmaJdGordon|: you can do that already
23:30:33 Quit Eosphere46 ("Not that there is anything wrong with that")
23:30:37 Join archivator [0] (n=Delyan@
23:31:06*JdGordon| gives up... he is way to confused
23:31:22rasherpixelma: I don't see it as "behaving differently", really. It's just how comments usually work everywhere else
23:31:43pixelmanow I'm confused
23:31:48gevaertsZagor: another way could be to split the table into four (or more...) tables (one per build type), and use css+js tabs
23:31:54 Quit archivator (Read error: 104 (Connection reset by peer))
23:31:58 Join Lynx_ [0] (
23:32:15 Join archivator [0] (n=Delyan@
23:32:31rasherpixelma: Text after # gets ignored. Lines that start with # are "comment lines" that might as well not exist
23:32:42Zagorgevaerts: sure, have fun! ;-)
23:32:46JdGordon|how about using css+js to hide various columns with single clicks... e.g "hide green", "hide sims", "hide bootloaders" etc?
23:32:51JdGordon|but all in one table
23:33:00rasherpixelma: Wait, let me phrase it better.
23:33:14pixelmaI think I understood now
23:33:15gevaertsZagor: I was afraid you'd say that
23:33:38rasherpixelma: # is a comment marker. Comments get ignored. If a comment is the only thing on a line, the entire line is the comment. If comment is only the latter, only the latter part is the comment, and hence the linefeed is not part of the comment
23:33:49 Quit Rob2222 (Remote closed the connection)
23:34:14rasherIt's just a matter of what you think of as "the comment"
23:34:23Zagorcollapsing greens is the best way imho. we can still make the full table too and have it available in a link.
23:36:45 Quit NAQOYQATSI (Read error: 145 (Connection timed out))
23:38:01 Join NAQOYQATSI [0] (n=yukayant@
23:38:29JdGordon|so wait a sec... isnt the bug then that comments mid line dont eat the \n?
23:38:37pixelmarasher: I think it's fine as it is - and as I said you can find other markup languages where the comment sign always eats the linefeed no matter if there is something before or not. E.g. the % in LaTeX marks the comment and I could (had to) use it to escape the linefeed when breaking too long \opt{blah,blub} lines as otherwise I got a space between the different things which broke the \opt for blah or blub
23:38:52pixelmaJdGordon|: for me a feature
23:39:08rasherIt means I have to litter my WPS with blank lines if I want to use inline comments
23:39:11pixelmaeh... no
23:39:11rasherLooks terrible
23:40:03pixelmathe "bug" is that it eats the newline which is just consistent to me
23:41:13rasherI admit to not knowing LaTeX well enough to know of this, but every other language I've used, inline comments don't require a blank line to be inserted below it
23:41:19pixelmarasher: I don't understand why you would want to use inline comments by the way
23:41:57rasherIt's a nice way to explain the structure of a wps. Or it would be if it didn't break down
23:42:04pixelmajust put the comment above or below
23:42:09 Quit bluebrother ("leaving")
23:42:24Zagoryeah that sounds like the simplest solution: only full-line comments allowed
23:42:34rasherLooks terrible and you can't do ascii art!
23:43:20gevaertsHow about committing a new builds file with a checkwps build for each target that has a simulator build? That should provide some real motivation to solve the table size problems
23:43:39Zagorgevaerts: sure, have fun! ;-)
23:43:41pixelmaif I add comments they are usually long enough on its own, putting them after a line of "random" WPS tags just make them unreadable to me
23:43:51rasher <−− these examples can't work.
23:44:13gevaertsZagor: I mean, I commit the new builds file, someone else gets motivated to work on the table :)
23:44:58Zagorgevaerts: sure, fire away
23:45:16CIA-43New commit by gevaerts (r22669): Add checkwps builds
23:45:26Zagorrasher: and is this really so terrible: ?
23:45:51 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
23:45:57rasherZagor: That doesn't carry the same amount of information
23:46:07rasherI don't like it one bit.
23:46:08CIA-43New commit by gevaerts (r22670): remove empty line at the end
23:46:11Zagorwhat information is lost?
23:46:20rasherWhich specific line is meant
23:46:26*JdGordon| prefers inline comments also
23:46:41rasherThe first ones also look terribly messy
23:47:05 Quit TheSeven (Read error: 54 (Connection reset by peer))
23:47:15JdGordon|pixelma: as for longer comments... wps lines arnt usually that long, and when they are, they could easily be split up with workign comment handling
23:47:18rasherAnyway, if it's such a big issue, I'd rather we ban inline comments completely than allow this unintuitive handling
23:47:50pixelmait's not unintuitive for me :\
23:48:54JdGordon|the difference between the 2 examples is that in rasher's you can sort of picture the wps by ignoring the comments... in Zagor's you cant do this.. the comments mess everything up
23:49:18 Join TheSeven [0] (
23:49:27JdGordon|which I tihnk is why the most comments you get for a viewport is a single line "aa viewport"
23:49:29rasherOnce you *know* this, you can adapt, but I doubt anyone fresh off the train expects a comment at the end of a line to join the two lines
23:50:04pixelmaJdGordon|: just look at some WPSs in SVN and you'll take that "they are usually not that long" back ;) My WPS never have a simple %ia for example as I use filename/directory fall backs etc.
23:50:14ZagorJdGordon|: any moderately complex wps loses any chance of being "visualized" in the source anyway
23:50:16 Part zzy
23:50:26JdGordon|pixelma: yes, but wouldnt it be nice to split those long lines up to make it readable?
23:50:47pixelmarashers example won't look that nice if you have lines that are differently long and longer overall
23:50:52rasherZagor: It worked pretty well for my widecabbie wps, if I may say so myself. At least before I realised I couldn't use inline comments..
23:50:52 Join zzy [0] (
23:50:56JdGordon|Zagor: yes and now... you at least can see the rough line placements correctly
23:51:00pixelmaJdGordon|: you can already do this...
23:51:02rasherpixelma: Just add some spaces before the # ..
23:51:42Zagoroh well, I don't really care either way. it was just a simple way out.
23:51:59pixelma"some"? - and how long will the long WPS line plus comment be?
23:52:37*JdGordon| goes to find a random wps line to use as an example
23:53:24gevaertsI don't really like the magic linefeed-eating, and I'd prefer an explicit continuation character. The main problem I'm afraid of is that I don't see a way to spot WPSes that break due to the change
23:53:25JdGordon|if that was mine, I'd split it on every ; with a short comment
23:53:29rasherpixelma: I managed most of widecabbie within 80 lines, including comments
23:53:37pixelmathose full comment lines can make it look messy too but to me they are still more readable (except rasher's too simple example)
23:54:08pixelmaJdGordon|: you can already do this now... someone just didn't do it
23:54:08rasherpixelma: Of course it's a simple example, I'm not going to write a full wps to show you how inline comments work...
23:54:45pixelmaI know how they work but won't look as nice in a real case
23:54:55rasherYes they will.
23:55:25rasherDo I really have to do this...
23:55:27pixelmato you... (and this getting stupid "religious" :\ )
23:55:28JdGordon|yep, once again, I dont know what I'm arguing.... inline comments definetly are more readable
23:55:43rasherpixelma: no, you're dismissing something of being "too simple", when it's quite close to reality
23:55:50rasherthis is not an opinion.
23:56:24 Quit DarkDefender ("Leaving")
23:56:31pixelmaI just compare to imagine mine
23:57:03rasherSo because you write wpses where every line is terribly long, everyone else suffers?
23:57:15rasherBecause you surely could never see a use for inline comments.
23:57:21JdGordon|you know what we *could* do.... add a control token to the start to define # behaviour for that wps
23:57:28 Quit jgarvey ("Leaving")
23:57:37rasherThe longest code line I have in widecabbie.wps is 78 characters.
23:57:43rasherMost are less than 50
23:57:43JdGordon|can we please ban the word "sigh" from irc?
23:57:51rasherPlenty of room for inline comments
23:58:06rasherThat is, if we were allowed to have them.

Previous day | Next day