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 2008-03-08

00:00:07Burnninat0rwell, i like how the regular ipod firmware lets me know if i have already listened to something. and there is a menu specifically for podcasts.
00:00:59Burnninat0ri have been reading about the bookmark feature which is good, but i really don't want to go through my whole music database in order to find my podcasts.
00:01:47bluebrotheryou can setup a new search in the database
00:04:30 Quit bruges|afk (Read error: 110 (Connection timed out))
00:06:47 Join ctaylorr [0] (
00:07:14 Quit bruges|afk_ (Read error: 110 (Connection timed out))
00:07:52*petur remembers there was a patch for podcast searching, not sure if it was ever comitted
00:10:00 Quit mf0102 ("Verlassend")
00:11:25bluebrotherwe have almost 500 open feature requests in the tracker :(
00:12:14ali_asIf one of them isn't to extend the number of feature requests the tracker can handle, you'd better add it.
00:13:00 Join kushal_12_27_200 [0] (n=kushal@
00:13:22bluebrotherthere's a huge difference between "tracker can handle" and "tracker is unusable because it's flooded"
00:13:44n1sSSnake: depends on what you mean, regular music playback cannot run if you have claimed the buffer but your plugin can still play audio
00:14:48n1sbluebrother: I agree that feature requests are getting less useful...
00:15:17bluebrother162199 tasks opened during the last 12 month ... 217 bugs in the same time
00:15:30bluebrother*199 feature requests
00:16:06bluebrotherwe need another cleanup week.
00:16:24bluebrotherand maybe a bug-hunting-month?
00:16:31 Quit Burnninat0r (" ajax IRC Client")
00:16:34bluebrotherMORB −− Month of Rockbox Bugs?
00:17:09n1svery much so, I also think we should try harder to maintain the tasks in the tracker, I recently opened a duplicate bugreport myself because the original was marked as ipod video only...
00:17:19bluebrotherdoes someone know what "i-doser" files are?
00:17:27 Join csc` [0] (n=csc@archlinux/user/csc)
00:18:50 Quit petur ("Zzzzz")
00:19:26*linuxstb reads the site and still doesn't know what i-doser is
00:19:53 Quit japc (Read error: 104 (Connection reset by peer))
00:20:03 Quit kushal_12_27_200 ("This computer has gone to sleep")
00:21:29SSnakethanks n1s, that's all i need to know
00:24:13Nico_Plinuxstb: I read your post on the default theme thread. I'm sorry if I appeared to have forced the cabbie v2 change
00:24:42Nico_Pit seemed to me that a majority found it better, and I think it was discussed here at least once
00:25:53 Quit ctaylorr (Read error: 110 (Connection timed out))
00:28:05 Join Xorothal [0] (i=Joe@gateway/tor/x-b2ffe82f5a325977)
00:29:20 Join moos [0] (
00:30:44 Quit gevaerts ("falling asleep")
00:31:12 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
00:35:24faemirgeneral question: how many devs are there that regularly contribute to rockbox?
00:35:56n1sit varies a lot with time etc.
00:38:13n1sthere are about 70 committers some of which who are not active anymore
00:38:43 Join sarixe [0] (
00:39:59 Quit moos (Read error: 104 (Connection reset by peer))
00:40:12 Join moos [0] (
00:42:39 Quit w0rd54 (Client Quit)
00:43:06 Join w0rd54 [0] (
00:43:07 Quit faemir (Remote closed the connection)
00:43:21 Quit Nico_P (Remote closed the connection)
00:45:13 Quit sarixe ("(EE) Failed to load "quit" module")
00:45:55 Quit barrywardell ()
00:46:38 Quit moos (Read error: 113 (No route to host))
00:46:41 Quit Horscht ("")
00:46:50 Join moos [0] (
00:49:47 Quit jhMikeS (Nick collision from services.)
00:49:53 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
00:52:33amiconnSlasheri: RED
00:53:07amiconnBagder: The build system again didn't pick up a commit (Slasheri's this time)
00:54:07 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
00:54:54 Quit moos (Read error: 104 (Connection reset by peer))
00:55:08 Join moos [0] (
00:56:13 Part SSnake
00:56:21 Join TMM [0] (
00:56:55Slasheriamiconn: fixing
00:57:09 Quit n1s ()
01:05:14 Quit bluebrother ("leaving")
01:13:12 Quit amiconn (Nick collision from services.)
01:13:14***Saving seen data "./dancer.seen"
01:13:18 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
01:13:25 Nick m0f0x_ is now known as m0f0x (
01:13:52 Quit Xorothal (Remote closed the connection)
01:17:11 Quit moos ("Rockbox rules the DAP world")
01:25:34 Quit roolku ()
01:25:51 Quit corevette (Read error: 104 (Connection reset by peer))
01:27:40 Join corevette [0] (
01:28:11 Part toffe82
01:30:05 Join dr_kosmos [0] (i=dr_kosmo@
01:31:08dr_kosmoshey guys, i'm thinking about making my own wps... i've read the guides, and that seems all good, but how do i set the font (and/or change the font colour)?
01:31:42 Quit corevette (Remote closed the connection)
01:31:59linuxstbYou can't change that in the wps itself, they are set globally in Rockbox (via the theme .cfg file)
01:33:03dr_kosmosahh okay, that helps.. thanks
01:33:14 Join corevette [0] (
01:34:21 Join quaal [0] (
01:35:15 Quit XavierGr (Read error: 113 (No route to host))
01:35:18quaalcould someone please tell me why whenever i plug my ipod into the griffin usb car charger it goes on an endless loop through the Do not disconnect screen, Rockbox screen, Picture of a USB Plug screen, Apple screen
01:35:43quaalwhy is any of that shit necessary
01:35:43quaalcant we just make it charge in the background
01:37:32dr_kosmosit does charge in the background with my ipod fw charger
01:37:46dr_kosmosfor some reason your griffin is making it go into disk mode
01:41:36 Join kushal_12_27_200 [0] (n=kushal@
01:42:06 Quit kushal_12_27_200 (Client Quit)
01:45:28 Quit dr_kosmos ("Leaving")
01:51:46 Quit mirak ("Ex-Chat")
01:55:13 Join dandin1 [0] (
01:58:28 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
01:58:51 Join miepchen^schlaf [0] (
01:59:03quaalso uh.. any suggestions on what to do about it
02:00:51 Join kushal_12_27_200 [0] (n=kushal@
02:02:41 Quit DerDome ("Leaving.")
02:04:41 Quit ompaul (Client Quit)
02:08:40 Join kushal [0] (n=kushal@
02:09:21 Quit RoC_MasterMind ("Leaving")
02:09:29 Quit kushal_12_27_200 (No route to host)
02:23:03 Quit Rincewind ("bye")
02:30:40 Quit dandin1 ()
02:33:05 Part pixelma
02:33:30 Join jcollie [0] (
02:36:47 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
02:44:27 Join midgey [0] (
02:44:48 Quit midgey (Client Quit)
02:45:58 Join RoC_MasterMind [0] (
03:00:14 Join EspeonEefi [0] (i=espeonee@STRATTON-SIX-NINETY-SEVEN.MIT.EDU)
03:08:21 Quit quaal (Connection timed out)
03:11:48 Join yoyoned [0] (
03:13:16***Saving seen data "./dancer.seen"
03:16:38 Quit linuxstb (Read error: 110 (Connection timed out))
03:23:39 Quit ali_as ("zzzzzzzzzzzzzzzzzzzzzzzzzZZZZZZZZZZZZZZZZZZZZZZZZZZZZz")
03:24:02 Quit XavierGr (Nick collision from services.)
03:24:13 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
03:25:08 Join axionix_ [0] (
03:25:32 Quit XavierGr (Nick collision from services.)
03:25:43 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
03:27:31 Join midgey [0] (
03:28:05 Quit midgey (Client Quit)
04:02:32 Quit Thundercloud (Read error: 104 (Connection reset by peer))
04:22:06 Quit tessarakt ("Client exiting")
04:23:22 Join FOAD_ [0] (
04:23:28 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:23:39 Join miepchen^schlaf [0] (
04:24:17 Join quaal [0] (
04:35:49 Quit BlakeJohnson86 ("Leaving.")
04:36:03 Join BlakeJohnson86 [0] (
04:40:08 Quit FOAD (Read error: 110 (Connection timed out))
04:40:08 Nick FOAD_ is now known as FOAD (
04:53:50 Join esotericWISP [0] (n=62dcbd9e@gateway/web/cgi-irc/
04:55:54 Quit esotericWISP (Client Quit)
05:13:17***Saving seen data "./dancer.seen"
05:17:54 Join clangston [0] (n=44613593@gateway/web/cgi-irc/
05:22:05 Join midgey [0] (
05:23:05 Join BlakeJohnson87 [0] (
05:24:12 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
05:42:31 Join Davide-NYC [0] (
05:42:39 Quit Davide-NYC (Client Quit)
06:08:36 Join luckz [0] (
06:10:09 Part luckz
06:10:53 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
06:11:13 Join leox [0] (
06:13:51leoxhi, somebody can helpme in this channel?
06:17:09 Quit clangston ("CGI:IRC (EOF)")
06:19:14advcomp2019leox, we can if we know what it is
06:22:23 Join wisp_sh [0] (
06:23:12leoxwell, I have rockbox on my sansa e250, I have a 2gb micro sd card, I connect then I connect the player to the computer, its change the FW to the original, I can copy music to the internal memory but I dont see the M sd card folder
06:26:13advcomp2019are you able to see the mircosd card on the computer?
06:28:27 Quit quaal (Read error: 110 (Connection timed out))
06:29:30 Join TheBasha1 [0] (n=User@
06:29:52 Join Doomed_ [0] (
06:30:08 Part TheBasha1
06:30:20leoxno, only in rockbox when i go to files i see <MICRO_SD1>
06:32:22advcomp2019are you in MTP or MSC mode in the OF
06:34:50leoxmsc advcomp2019
06:35:46advcomp2019What operating system are you using?
06:36:51leoxwhen i connect it change to the original FW
06:37:18advcomp2019i am talking on the computer itself
06:37:59 Quit kies (Read error: 110 (Connection timed out))
06:38:06leoxoh, linux
06:38:08 Quit goffa_ (Read error: 104 (Connection reset by peer))
06:38:33advcomp2019have you manually mounting it
06:39:10 Join kies [0] (
06:40:39leoxthe msc mode is ok? in orig FW?
06:43:16advcomp2019yes that is fine because rockbox does not have usb support for the PP targets yet
06:47:29 Quit alienbiker99 (Read error: 110 (Connection timed out))
06:47:44 Quit csc` ("Powering Off")
06:50:10 Join csc` [0] (n=csc@archlinux/user/csc)
06:51:29 Quit csc` (Client Quit)
06:53:38 Quit XavierGr ()
06:56:47leoxnothing advcomp2019
06:58:18 Quit DaCapn (Read error: 110 (Connection timed out))
06:58:19advcomp2019have you tried reformatting the mircosd card or trying it into a different card reader?
06:59:04leoxidont have another reader
06:59:15HodappI've got a 3rd-gen iPod... you guys know any reason why the display repeatedly gets out of sync with what song is actually playing?
07:00:50advcomp2019leox, i will add one more thing.. have you tried it on a different computer?
07:05:13leoxadvcomp2019, no, but i think the problem thats i cant connect to usb running rockbox and this is why i dont see the msd card mounted in the player. what u say?
07:07:24 Join csc` [0] (n=csc@archlinux/user/csc)
07:09:12advcomp2019if it is a 2gb microsd card, OF should read it just fine but like i said before there is no USB support for the PP target while running rockbox and it is being worked on right now
07:13:03 Join cool_walking_ [0] (
07:13:19***Saving seen data "./dancer.seen"
07:13:58cool_walking_Is there a key mapped to "hold" on the ipod sim?
07:16:21cool_walking_ah... h
07:16:29*cool_walking_ feels stupid
07:23:48 Quit cool_walking_ (Read error: 104 (Connection reset by peer))
07:35:22leoxadvcomp2019, where is normaly the folder of the msd in the root dir of the player, and what is named?
07:42:50advcomp2019on the computer it should be another removable drive
07:44:04leoxbut its not inside de player file system?
07:47:25advcomp2019no, it mounts as two removable storage devices
07:51:23leoxoh! im a completly STUPID! now i found it, the player mount like 2 different devices, now i understand, thanks
08:03:33 Join Nimdae [0] (
08:04:20Nimdaei have to say, having usb in rockbox is kinda stable is it these days? so far i haven't hit any problems
08:15:59 Join Rob222241 [0] (
08:26:21 Quit midgey ()
08:28:47 Join daurnimator [0] (i=daurn@unaffiliated/daurnimator)
08:34:20 Quit Rob2222 (Read error: 110 (Connection timed out))
08:36:27 Quit spiorf (Remote closed the connection)
08:37:10 Join spiorf [0] (
08:38:10 Join davina [0] (
08:39:06 Quit daurn (Read error: 110 (Connection timed out))
08:42:55*amiconn thinks the proper fix would probably be to port the new api to the hwcodec playback engine
08:47:26Slasheriamiconn: indeed, or create a global event handling api for rockbox
08:47:47Slasherithat playback system and others could use
08:49:05amiconnWhere else do we need that kind of event handling?
08:49:50Slasherithat could be used in the ata callback for example that queues events to be ran when disk spins up
08:50:18amiconnHmm, true
08:50:29amiconnBut (swcodec) playback is apps, ata is firmware
08:51:04Slasherihmm, indeed. but maybe the the event handling could go to the firmware level
08:51:11Slasherithen it would work for hwcodec too
08:51:20amiconnYes, perhaps
08:52:15*amiconn thinks that needs a bit of discussion, but sounds like it might be a good solution
08:54:02amiconnOne question is how to define event ids
08:54:10Slasheritrue. then there could be just generic sysevent_add(DISK_SPINUP, handler);
08:54:14amiconnThey need to be globally unique
08:54:30Slasheritrue, those probably needs to be pre-defined in firmware level too
08:55:13amiconnMaybe they should be defined in events.h, but that separates them e.g. from playback.c
08:55:27amiconnOtoh, the ata callback is used from many places
08:56:53Slasheriif we would like to define the events in playback.[ch] then the code would need to register an event first with its own id
08:56:59amiconnPerhaps events.h could define the classes, and the .h file of the respective module would then define the individual events?
08:57:13Slasherii don't know if this would just add unnecessarily more complexity
08:57:31Slasherihmm, that would work
08:57:44amiconnLike #define EVENT_CLASS_DISK 0x00010000 / #define EVENT_CLASS_PLAYBACK 0x00020000 / ...
08:57:55Slasherithat sounds good
09:00:01amiconnSimilar thing in ata.h
09:00:40amiconnNot sure whether we need 32 bit
09:00:49Slasherihmm, that should work fine
09:01:15Slasherii don't think there will be that many events per class, or classes
09:01:23amiconnFor the current targets it doesn't matter, but in case someone wants to ever pick up calmrisc16 again, or even try rockchip, this might be nasty
09:02:05 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
09:02:47Slasherimaybe class could use 4 bits and the event definition the rest of 12. that should be more than enough
09:03:32amiconnOr even 8 bits + 8 bits
09:03:46amiconnDo you think one module will ever need more than 255 events?
09:03:55Slasheriunlikely :)
09:04:24amiconnOtoh, "640KByte ought to be enough for everyone" ;). Even the archos has more :P
09:04:56amiconnBut I guess we can simply extend the scheme in case it will become necessary
09:07:11Slasheribut maybe i will try how that kind of event handling would work
09:10:08amiconnThe ata callbacks are one-shot, so each registered handler would have to unregister itself
09:10:18pixelmaHodapp: I read your question about the 3rd gen Ipod in the logs. In the last month there were a few playback bugs (skipping tracks) which sometimes also looked like you describe (playback and WPS differed). What build are you running and have you tried the latest builds? The 3rd gen problems with some builds should also be fixed since yesterday evening...
09:10:30amiconnAre the playback events also one-shot?
09:11:27amiconnIn that case, the events should probably be made self-unregistering
09:11:35 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
09:11:39Slasherinope, playback events are permanent. but when we generate the event, it should be possible to specify whether keep or de-register the event
09:13:24***Saving seen data "./dancer.seen"
09:19:24 Join desowin [0] (n=desowin@atheme/developer/desowin)
09:23:21amiconnSlasheri: Does your simplification mean that buffered, but unplayed tracks will be counted as played now, e.g. in scrobbler.c?
09:24:26Slasheriin fact i don't know how scrobbler counts those tracks because it doesn't use unbuffer events at all.. that's weird
09:24:44Slasherihmm, in fact need to verify that
09:24:52 Join petur [50] (n=petur@rockbox/developer/petur)
09:25:45peturLlorean: ping
09:27:22Slasheribut the changes to unbuffer events (and changing it's name to track finished event) now makes sure those events are called correctly whenever a track has finished playing (skipped, end of playlist etc.)
09:28:04 Quit Nimdae ("leaving")
09:30:22Slasherihmm.. maybe a scrobbler dev should change the code to use track finished event and then remove all prev_track_elapsed code from playback.c
09:30:31amiconnpetur: I just had an idea. Did you try what happens if you send the "disable advanced power management" command to the problematic cards?
09:30:32Slasherii don't think the current changes should break scrobbler
09:30:54amiconnAnd if that works, whether sending the enable command after a disable command also works?
09:31:30amiconnMaybe the card doesn't like the enable command if apm is already enabled?
09:32:00amiconn(if that is the case, it still leaves the question how to guess the initial state)
09:32:26peturamiconn: I did, but I didn't write it down and I forgot. But that is an interesting idea, must try that later on
09:32:36*petur has serious memory issues :/
09:34:27amiconnIf the disable-enable sequence actually works, you could try whether a disable-disable-enable sequence also works
09:34:47 Quit martii (Read error: 104 (Connection reset by peer))
09:35:10amiconnThen we can be sure that disable always works, and always use the disable-enable(with desired parameters) sequence
09:37:01peturI'll try to find some time to test, today is pretty much stuffed with things to do...
09:38:07peturI actually popped in to ask Llorean to cool down a bit, certainly when handling long-term users. Telling them to RTFM surely offends them...
09:38:07 Quit petur ("later")
10:01:24 Join ch4os_ [0] (n=ch4os@unaffiliated/ch4os/x-059673)
10:01:55 Join barrywardell [0] (n=barrywar@rockbox/developer/barrywardell)
10:05:17 Join Siku [0] (
10:09:36 Join dr_kosmos [0] (i=dr_kosmo@
10:11:30 Join linuxstb [0] (
10:11:39dr_kosmoshey guys... i'm almost done making my first WPS, only i have one question: how can i get a progress bar (%pb) except for my volume level?
10:12:54linuxstbYou need to use conditional bitmaps - lots of WPSs do that if you want to see how it's done
10:13:46linuxstbe.g. cabbiev2
10:15:02dr_kosmoshmm... is there any limit to how detailed/big it can be? (i'd like to have it stretching across the entire screen, and reasonably detailed)
10:15:50linuxstbOnly the size of the WPS image buffer, and the limit of 52 images per WPS
10:16:27dr_kosmosok cool... 52 images should be plenty. cheers
10:17:08linuxstbAlthough it sounds like there should be a better way than this... (but it would need someone to code support for it in Rockbox0
10:18:50dr_kosmosmmm certainly would be handy if there was a single command to do it (like there is for %pb)
10:23:46 Quit dr_kosmos (Read error: 104 (Connection reset by peer))
10:32:00 Nick JdGordon|afk is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
10:35:38 Quit desowin ()
10:41:02 Join FOAD_ [0] (
10:42:47 Join einhirn [0] (
10:46:14pixelmalinuxstb: (because of your forum post) - what do you think about a NotSupported wiki page which explains what it means in Rockbox terms and maybe describes the problems for some popular models (similar to what we had for the 80GB Ipods last year but more generic). It could then be linked from the frontpage and we could point people to it...
10:46:43 Join waldo [0] (
10:48:48linuxstbpixelma: What about adapting the TargetStatus page?
10:51:50pixelmathat's already quite detailed, I just thought about making clear that "not supported" doesn't mean that "no one has tried" or "use at your own risk" first. Maybe that could be put on a glossary page which I think exists too..
10:52:11 Quit Horscht ("I got raided by the FBI and all i got is this lousy quit message")
10:52:21linuxstbI always try to use "does not work", rather than "not supported".
10:53:20pixelmamaybe it could be made more clear on the frontpage then which lists "supported" players and says "Nano 2nd/3rd gen is not supported"?
10:53:22 Join nplus [0] (
10:53:46pixelmaah no, that was already changed
10:55:18pixelmabut a lot of people still use "not supported" if someone asks for Rockbox on an Ipod Classic for example
10:56:35linuxstbYes, and IMO those comments are misleading - I don't think we should try and invent our own meaning of "supported", when lots of other projects/companies use it to mean "works, but don't ask us for help"
10:58:04pixelmaso a call to all who answer those questions to not use it anymore?
10:58:06 Quit FOAD (Read error: 110 (Connection timed out))
10:58:07 Nick FOAD_ is now known as FOAD (
10:58:41linuxstbI think so...
10:59:09linuxstbBut also, a clearer page with status of the major players (such as new ipods) would be helpful.
11:00:22pixelmasounds good, I just wonder where to put that...
11:00:52pixelma(the "call")
11:02:02linuxstbMaybe just put it in "General Discussion" ? It can't hurt for users to see it as well.
11:02:36pixelmaalright, going to start my first topic then :)
11:04:15linuxstbpixelma: I've just had another thought - we have an "unsupported builds" forum, so already use the term "supported" with a speciific meaning there...
11:05:52pixelmayes, realised that too now when I started thinking about posting something like "do not use 'not supported'"
11:06:34*linuxstb switches computers
11:06:39 Quit linuxstb ("Leaving")
11:08:23pixelma(for the logs) I'll try to cover that too, but I often need a bit to come up with a nice way of putting it...
11:08:52 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
11:09:40linuxstbpixelma: If only everyone took the time to do that... ;)
11:10:32*amiconn is undecided what to do next :/
11:10:53pixelmalinuxstb: wut do u meen? ;)
11:11:07linuxstbamiconn: What's on your list?
11:11:29 Join petur [50] (n=petur@rockbox/developer/petur)
11:11:51amiconn(1) greyscale gamma improvements (extensive test series, and coming up with a way for calibration by the user)
11:12:38amiconn(2) greyscale live gamma switching for 1st/2nd Gen (because the backlight inverts black & white, the lcd transfer curve must be changed live. Realted to (1))
11:12:50 Join replix_ [0] (
11:13:01amiconn(3) timer on cop, and putting the greyscale isr there
11:13:10amiconn(4) start the M3 port
11:13:25***Saving seen data "./dancer.seen"
11:13:56 Join mcuelenaere [0] (
11:14:39 Nick replix_ is now known as replix (
11:15:20 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
11:16:01amiconnOh, and (5) rework power management, and add suspend for 1st/2d Gen
11:16:26 Join stripwax [0] (
11:17:06linuxstbSounds like (5) would be the most useful in terms of core features...
11:17:32amiconnYes, but (4) is tempting too, and more targets == good
11:17:53amiconnAlso, weren't there a few requests for M3 support?
11:17:55*amiconn not sure
11:18:37amiconnM3 support also requires a bit of core cleanup (mostly reducing code duplication among lcd drivers)
11:19:14amiconnI am thinking about a similar approach as to what you used for the filters in libdemac
11:19:30amiconn(including a .c file)
11:20:43linuxstbYes, the duplication is a pain in the lcd drivers.
11:22:03amiconnMy idea would reduce them to a single driver for each pixel format
11:22:25amiconnlcd-remote-1bit-vert.c would set a few macros, then just include lcd-1bit-vert.c
11:22:57amiconnlcd-1bit-vert.c would one of those macros, and if it's not set, set it for main lcd
11:23:23amiconnThis way the driver can be compiled twice, and produce the 2 sets of functions
11:25:31JdGordon.. then there is majoy swcodec rework... and sw/hwcodec merging...
11:26:07amiconnWell, I don't really know how swcodec playback works, and in its current state, I'd rather not want to merge
11:27:09amiconnI'll probably have a try once swcodec playback got reasonably stable (although I'm currently not experiencing most of the reported problems)
11:27:54JdGordonI'm really not sure fixing all the bugs is the way to go
11:28:08*JdGordon still likes the idea of blosing away swcodec and startiong from scratch
11:28:28amiconnThere might be one other reason that could force me to try earlier though, and that would be adding pcm playback support for MAS
11:29:00amiconn(in the core I mean; WAV playback and recording is possible via plugin for quite some time now)
11:33:30amiconnWould it be a good idea to add an lcd/ sub-dir in firmware/drivers/ ?
11:33:31 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
11:35:25linuxstbamiconn: Yes, I think so.
11:35:26 Quit courtc (Read error: 104 (Connection reset by peer))
11:35:31 Join courtc [0] (
11:36:36 Join Lear [0] (i=chatzill@rockbox/developer/lear)
11:37:53*amiconn wonders whether we'll have targets with OLED displays soon
11:39:15linuxstbI was trying to think of a better name than "lcd" - as they are the code for rendering into a framebuffer
11:39:29stripwaxframebuffer? ;)
11:39:38linuxstbClever ;)
11:39:51*stripwax is still trying to get his phone to work..
11:39:52amiconnOr maybe just display_
11:40:06amiconnHmm, might be too generic
11:40:15 Join mcuelenaere [0] (
11:40:37linuxstb"framebuffer" is better than anything I've thought of...
11:41:27amiconnBut it's quite long
11:41:34linuxstbYes, it's not ideal...
11:42:39linuxstbBut I think it's accurate.
11:42:48JdGordonhow about..... fb
11:44:02linuxstbI think I would prefer to just use the whole word, rather than abbreviate
11:44:11amiconnThere's also the question how to call the low-level target specific part
11:44:55amiconnRight now, lcd is correct, but would be wrong on OLED targets. framebuffer would be completely wrong.
11:45:14linuxstbBut aren't they all in target-tree?
11:45:18amiconnThey are
11:45:21linuxstbOr are you talking about function names?
11:45:40linuxstbAh yes, I don't think we just rename functions to use "framebuffer_"
11:46:08 Quit einhirn (Read error: 104 (Connection reset by peer))
11:46:12amiconnAlso, framebuffer feels a bit strange. Then you would do e.g framebuffer_fillrect(....); lcd_update();
11:47:14amiconnframebuffer_update would be wrong because it doesn't update the framebuffer, but updates the display from the framebuffer
11:47:26linuxstbThose function names are more accurate though...
11:47:49linuxstbi.e. you draw in a framebuffer and then update the LCD - you don't draw directly on the LCD
11:49:51JdGordonwhats wrong with leaving it lcd_?
11:50:09pixelmaand would it be oled_update later?
11:50:20JdGordonwhy couldnt it still use lcd?
11:50:24JdGordonlcd is fairly generic
11:51:15linuxstbWhat if we ever get TV-Out working on a target? tv_update() ?
11:51:35 Join ender` [0] (
11:51:39linuxstbBut no, I think we can live with just calling everything an LCD...
11:51:54linuxstb(at least for now)
11:51:56JdGordonyeah, lets worry about that when that happens
11:52:23 Quit barrywardell ()
11:53:54 Quit replix ("Verlassend")
11:56:11JdGordonchanging topics a bit... apart from flicking hold and trying menu+select how else can I try getting a nano to turn on?
11:56:23JdGordonthe batt should be full.. its been cplugged in overnight
11:57:26*JdGordon 's sister is a drongo
11:58:21 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
12:01:32amiconnThe ipod hardware reset should always work, unless either the battery is totally flat, or the ipod is broken
12:01:41 Join replix [0] (
12:02:04JdGordonits chargiung with the comp so looks like the wall charger was just stuffed
12:02:07amiconnSometimes it takes 2 tries and/or holding the combo quite a while (> 15 seconds)
12:02:10JdGordonor rockbox charging isnt working
12:03:36amiconnRockbox only charges slowly, as we don't handle charging current setup. I'm not sure we know how to do this either
12:04:12JdGordonstill, it was plugged in overnight, so it should have had _Some_ charge
12:04:16amiconnFullsize ipods can even become discharged when used a lot while "charging" in rockbox. Minis and Nanos shouldn't, though
12:05:39 Join mcuelenaere [0] (
12:07:38 Join roolku [0] (
12:08:07 Quit stripwax (Read error: 104 (Connection reset by peer))
12:09:38 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
12:13:17 Join stripwax [0] (
12:13:55bluebrotherhmm, do we have recordbutton-to-rec-screen on iaudio?
12:15:48amiconnNo, as that's not possible
12:16:00amiconnThe record button is used for the menu on iaudio
12:16:44bluebrotherah, so long-rec won't work too?
12:16:47linuxstbstripwax: Any luck with getting those photos off your phone?
12:17:00amiconnLong rec is quickscreen
12:17:05*bluebrother is currently looking at FS #8566
12:17:58stripwaxlinuxstb - in the correct aspect ratio? no, not yet. still trying
12:19:13*stripwax curses Nokia PC Suite once more, and reboots once more
12:19:19 Quit stripwax (Read error: 104 (Connection reset by peer))
12:19:35 Join faemir [0] (
12:22:05 Join ali_as [0] (
12:22:47 Join dr_kosmos [0] (i=dr_kosmo@
12:23:38 Quit TMM (Read error: 110 (Connection timed out))
12:23:38 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
12:24:24dr_kosmoshey guys, i just finished my first WPS... can someone help me out with wiki write permissions so i can upload it
12:25:30 Join stripwax [0] (
12:29:16 Quit dr_kosmos ("Leaving")
12:30:18 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:30:49*stripwax hurrahs
12:31:00 Join mcuelenaere [0] ( and
12:31:57 Join Mathiasdm [0] (
12:32:53JdGordonnooooo!!!! you might have beaten the seattle tower for hieght!
12:35:05linuxstb\o/ - and that was without GodEater's contribution...
12:37:10 Join homielowe [0] (
12:38:48*pixelma wonders what other pub visitors said about this :)
12:40:04*homielowe didn't know rockbox was linux-based ;)
12:40:48pixelmawhy does the forum have "problems connecting to database" now that I wanted to post my first "start new thread" post. Is this a sign?
12:40:48 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
12:42:17 Quit Mathiasdm ("Yuuw!")
12:42:26 Quit replix ("Verlassend")
12:42:44*scorche submits a correction comment
12:43:21 Join seablue [0] (
12:44:14 Join Thundercloud [0] (
12:47:28*bluebrother wonders why he can't see the comment
12:48:16 Join mcuelenaere [0] (
12:48:20 Join roolku_ [0] (
12:48:43*linuxstb wonders what the point is in a wiki without an edit button
12:48:55scorchelinuxstb: i was thinking the same thing =/
12:48:56 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:51:53*bluebrother wonders why there is a description of the changes but no diffs.
12:51:58*gevaerts has newbie questions about ipod, but doesn't dare to ask
12:52:01*roolku_ found a view port bug
12:52:29*scorche directs gevaerts to his alternate forum account
12:52:49JdGordonroolku_: oh?
12:52:52 Join lee-qid [0] (
12:52:55roolku_JdGordon: the progress bar for deleting/copying is drawn on top of the text (tested gigabeat f and m:robe 100)
12:53:31JdGordondrawn ontop of which text?
12:53:59roolku_roolku: well, for deleting on top of what directories are deleted
12:54:46linuxstbgevaerts: Just ask - I think you're earned the right... ;)
12:54:52*gevaerts found the answer on apple's website
12:55:15roolku_JdGordon: and for copying on top of the directories
12:55:27gevaertsI was just wondering if there was a "cleaner" way to get out of EDM than menu-center reset
12:55:30JdGordonroolku_: well, its just supposed to show the progress thing somewhere on the screen... its not supposed to show the filenames as they go
12:55:38roolku_JdGordon: sorry, on top of the context menu
12:55:49JdGordonyeah, thats normal
12:55:57JdGordonwasnt it doing that pre-vp?
12:56:26roolku_JdGordon: it is showing the directories that are deleted (or at least it used to) - never tried with nested dirs
12:56:45linuxstbgevaerts: No... Unless you write to the first 32MB (I think) of the disk - that will cause EDM to reboot. Or you if you Rockbox's "reboot to EDM" option (as opposed to forcing EDM by holding SELECT+PLAY), disk mode will automatically reboot.
12:56:52JdGordonits not really a big deal is it? its only there to show it hasnt crashed
12:57:38roolku_for deleting it definately showed below the directories and it was a much nicer looking thick bar
12:58:04gevaertslinuxstb: thanks. I guess writing to the firmware partition doesn't count as "clean" though
12:59:05roolku_JdGordon: just tried with nested dirs and the updated dirs and the progress bar keep erasing each other - looks very ugly
12:59:58JdGordonAH, ok, ill have a look soonish
13:00:03 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
13:00:12 Quit BigBambi (Read error: 113 (No route to host))
13:02:19 Quit BobShield (Read error: 104 (Connection reset by peer))
13:03:50roolku_JdGordon: if there is an "overwrite" prompt when trying to copy, the bar appears on top of that - I think it should be erased as soon as the question has been answered
13:04:56 Quit roolku (Read error: 110 (Connection timed out))
13:05:35 Nick roolku_ is now known as roolku (
13:13:09 Quit ompaul (Read error: 113 (No route to host))
13:13:29***Saving seen data "./dancer.seen"
13:14:05 Join Buschel [0] (
13:14:40 Part yoyoned
13:19:04 Quit homielowe (Read error: 110 (Connection timed out))
13:24:52 Quit jhMikeS (Nick collision from services.)
13:24:58 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
13:28:52 Quit ch4os_ (Read error: 104 (Connection reset by peer))
13:30:32 Join bluebrot1er [0] (n=dom@
13:30:55 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
13:32:53pixelmaSlasheri: the custom WPS format you updated seems very old (for example it mentions specific play mode tags which don't exist anymore as there is now a conditional for it). I'm not sure if it makes sense to update just a few new things...
13:33:55pixelmathe previous update was August 21st 2005
13:34:20 Quit faemir (Remote closed the connection)
13:34:58 Join faemir [0] (
13:35:48pixelmaI meant the file...
13:36:49 Quit kushal ("This computer has gone to sleep")
13:47:59 Quit bluebrother (Read error: 110 (Connection timed out))
13:48:55Slasheripixelma: ah, that's why i wondered it didn't seem to be very up-to-date
13:49:41linuxstbLots of the docs/ files have been removed recently - maybe that's another candidate...
13:52:16Nico_PSlasheri: I'm liking your playback.c commits :)
13:53:48SlasheriNico_P: nice! :)
13:55:10Nico_Pit's good to see we're a few to touch it these days
13:57:50 Quit mcuelenaere (Read error: 110 (Connection timed out))
13:59:15pixelmawould be nice if I could check if the cabbiev2 files are now correctly included in the m:robe100 builds
13:59:25 Join PaulJam [0] (
13:59:52Nico_Ppixelma: can't you do that by running make zip on a mrobe100 build?
14:00:47linuxstbI think checking in the sim should be enough - i.e. a "make install" is just a "make zip" followed by "unzip"
14:03:14 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:03:59pixelmaI checked here (in a sim) but want to be sure if maybe a file was missing in the commit...
14:05:19pixelmabut the number of new files seems correct
14:08:18linuxstbIf you type "svn status" in the wps/ directory, unknown files will be shown with a ? (if that's what you want)
14:08:45 Join desowin [0] (n=desowin@atheme/developer/desowin)
14:08:59 Join moos [0] (
14:09:44pixelmaah, nice - yes that helped. There's really no difference. :)
14:12:04 Quit spiorf (Remote closed the connection)
14:13:39 Join jurrie__ [0] (
14:21:11 Quit jcollie (Read error: 113 (No route to host))
14:21:27roolkudoes anyone have an up-to-date (after 2/3/2008) rbutil binary for windows? The one on is to old.
14:21:41 Quit jurrie_ (Read error: 110 (Connection timed out))
14:22:07roolkuI would like to test an m:robe install now that we have the wps included
14:23:58Horschtuhm.... wow.
14:24:11Horschti just saw a 160GB 1.8" HD on ebay
14:24:14Horschtfor 33€
14:24:25Horschtsounds fishy :D
14:24:41linuxstbIs it from an ipod Classic?
14:25:32linuxstbThis one?
14:26:18Horschtno, i just noticed it's a 2.5" Drive
14:26:20 Join merbanan [0] (n=banan@
14:26:27Horschtbut it was listed in the 1.8" category
14:26:37*Nico_P fixed his rbutil build problem \o/
14:28:22roolkuNico_P: nice, does it mean you can build a windows binary for me :)
14:28:50Nico_Proolku: unfortunately no, I'm using linux and have no idea how to build a windows binary
14:29:31roolkuNico_P: no problem
14:34:07bluebrot1erroolku: give me a few minutes and I'll create you a current one
14:34:14 Nick bluebrot1er is now known as bluebrother (n=dom@
14:34:16 Part pixelma
14:34:33roolkubluebrother: thank you
14:37:21 Join spiorf [0] (
14:38:18gevaertsDoes RoLo work on ipod 5G ?
14:39:11linuxstbIt should do...
14:39:17gevaertsSeems it does
14:39:52linuxstbFor some reason the check-after-usb-connect is #defined out for ipods though... That needs to be fixed.
14:41:22bluebrothermaybe because usb connection previously always meant reboot?
14:41:45gevaertsThat's probably why I wasn't sure then. Not really a problem then, I need RoLo just for easy testing with safe fallback
14:42:04preglowamiconn: was _that_ the mysterious 5002 bug???
14:43:23amiconnlinuxstb: That 'some' reason is your invention
14:43:33amiconnIt's USB_IPOD_STYLE
14:43:34linuxstbAnd no-one stopped me? ;)
14:43:59amiconnSince ipods used to reboot for usb, there is no point in checking afterwards, as that could would never be reached
14:44:35amiconnThat reboot still happens in svn. When we enable ums, simply remove that #define from the config-*.h files
14:45:13preglowamiconn: how the hell did you find that?
14:45:18amiconnIn fact USB_IPOD_STYLE *should* be defined for all PP targets, but isn't defined for Sansa and H10
14:45:19preglowamiconn: that's got to be the most obscure bug i've seen
14:45:35amiconnpreglow: By lots of experimentation. It's quite esoteric, yes
14:46:28preglowamiconn: well, it's good to have it out of the way
14:46:31preglowamiconn: nice work, btw
14:46:45 Part Buschel
14:46:55bluebrotherbuild from current svn, so expect all new bugs to be present ;-)
14:47:36roolkubluebrother: thanks. downloading...
14:51:48 Quit nplus (Remote closed the connection)
14:52:30*gevaerts forgot USE_ROCKBOX_USB again...
14:53:23 Join Buschel [0] (
14:53:35 Join nplus [0] (
14:53:46 Join rasher_ [50] (n=rasher@rockbox/developer/rasher)
14:56:11 Join mcuelenaere [0] (
14:58:33 Join Horschti [0] (
14:58:41Tuplanollaon a 386 system, of course? :P
14:59:17 Quit Horscht (Nick collision from services.)
15:04:43roolkuthe full install with rbutil worked, but 1) it couldn't autodetect the player 2) it claimed the bootloader was already installed and uptodate
15:04:52 Quit faemir (Remote closed the connection)
15:05:32 Join faemir [0] (
15:05:35roolkubluebrother: is this something you can deal with or who do I report to?
15:06:44 Quit rasher (Read error: 113 (No route to host))
15:08:59roolkuit also claimed it would override the bootloader, but did the reame to OF.mi4 routine
15:09:01 Quit nplus (Remote closed the connection)
15:09:30bluebrotherroolku: do we have usb ids for the m:robe?
15:09:53roolkuyes, they should be in there
15:10:47*bluebrother found it
15:11:01bluebrotherhmm, in that case it should be possible to detect it. At least using the ids.
15:12:54roolkuhm, different to the one in the config file: #define USB_VENDOR_ID 0x07B4
15:13:32***Saving seen data "./dancer.seen"
15:14:02roolkucould you try with usbid=0x07B40280 ? Or I can detect again to be safe
15:14:32 Join nplus [0] (
15:15:19 Join dan_a [0] (n=user@
15:15:43roolkuyup VID:07B4 PID: 0280
15:16:19bluebrotherhmm, I guess olympus only uses on VID
15:16:26bluebrotherI'll create you an updated binary.
15:16:36 Quit spiorf (Remote closed the connection)
15:22:57bluebrotherroolku: updated the binary. Same URL as before ...
15:23:24 Join quaal [0] (
15:23:50 Quit petur ("switching....")
15:24:08 Join petur [50] (n=petur@rockbox/developer/petur)
15:24:14gevaertsDoes anyone expect problems if I broadcast SYS_USB_DISCONNECTED too often ? Will all threads still ack it ?
15:24:27 Join jcollie [0] (
15:24:35 Join Arathis [0] (
15:26:27roolkubluebrother: autodetection works now :)
15:28:02bluebrothernice. Looks like it was simply a typo
15:30:35 Quit Buschel ()
15:33:40*bluebrother already committed the change
15:36:58 Join mf0102 [0] (n=michi@
15:37:44 Quit jurrie__ (Read error: 104 (Connection reset by peer))
15:38:04 Join jurrie__ [0] (
15:40:08 Quit RoC_MasterMind (Read error: 110 (Connection timed out))
15:41:43 Join basti [0] (n=basti@
15:43:49 Quit mcuelenaere (Read error: 110 (Connection timed out))
15:43:53 Join maurus [0] (
15:44:01 Quit PaulJam (".")
15:44:31 Part maurus
15:44:44 Join mcuelenaere_ [0] (
15:54:22 Join Rondom [0] (
15:55:31 Quit Rondom (Read error: 104 (Connection reset by peer))
15:55:40 Join Rondom [0] (
15:57:13peturamiconn: tried setting features 9, A and 85 before feature 5, but feature 5 keeps failing
15:57:13 Quit mcuelenaere_ (Read error: 104 (Connection reset by peer))
15:58:20 Quit quaal ("Leaving")
15:58:31 Join quaal [0] (
16:05:01 Join mcuelenaere [0] (
16:15:06*gevaerts is struggling with using "removable" for two different purposes at the same time
16:15:13faemirIs there a good page on how to make themes for rockbox somewhere?
16:15:25peturthe wiki?
16:15:54peturfaemir: and peek at other themes ;)
16:16:55linuxstbfaemir: This is the only page I know about (apart from the CustomWPS page which lists all the tags) -
16:18:12gevaertsIf I use eject/load to simulate a removable drive for "normal" operation, allowing to "eject" the drive to go back to charging-only mode, and provide a rockbox menu item to "insert" the drive again, how do I handle the sansa microSD slot ?
16:19:18*petur gives up on the CF powermanagement issue
16:19:34 Join axionix__ [0] (
16:19:46peturgevaerts: that is a problem indeed
16:20:15linuxstbgevaerts: What's the problem case? Aren't they just two separate drives - so the eject-to-charge behaviour only applies to the built-in drive.
16:20:32 Join axionix_1 [0] (
16:20:32***Alert Mode level 1
16:20:32DBUGEnqueued KICK axionix
16:20:32DBUGEnqueued KICK axionix_
16:20:32***Alert Mode level 2
16:20:32DBUGEnqueued KICK axionix__
16:20:32DBUGEnqueued KICK axionix_1
16:20:32***Alert Mode level 3
16:20:57linuxstb(or more specifically, if both drives are ejected...)
16:21:29 Quit maddler ("connection reset by beer!")
16:23:21bluebrothercan't you just detect if the card has been re-inserted and then go back to usb mode?
16:23:37linuxstbThat shouldn't force usb mode (IMO).
16:23:59gevaertslinuxstb: maybe it's as easy as that, but I'm not entirely convinced yet. Suppose a user ejects the internal flash, because he doesn't need it. Then he sees the wrong sd card in the drive, so he software-ejects that as well (as he should), and then swaps the card. By now, rockbox has switched to charging mode, and if we're lucky, the user is not confused and chooses the "reconnect" option. If we're unlucky, someone has to tell him to RTFM on the
16:24:13gevaerts... forum yet again.
16:24:55linuxstbWould that be common? For a user to eject the internal flash whilst still wanting to use the SD card reader?
16:24:57gevaertsbluebrother: maybe he has swapped the card because he has finished copying his homework and now goes back to listening to music...
16:25:38CtcpIgnored 4 channel CTCP requests in 0 seconds at the last flood
16:25:38*bluebrother realized that this is a bit more complicated ...
16:26:00gevaertslinuxstb: maybe not. I'm not sure. Maybe we should just try it and wait what happens...
16:26:41*gevaerts expects all kinds of confused sansa users. Luckily the one-drive players aren't affected by this
16:27:04linuxstbBut I would hope that even the most confused users would try unplugging and reattaching...
16:27:37bluebrotherdon't forget users not following the standard procedure (i.e. ejecting the card before removing)
16:29:49gevaertsIf they don't choose "eject", they stay in UMS mode I guess. I don't mind users getting confused because they don't follow standard rules
16:30:33***Alert Mode OFF
16:31:09linuxstbAlso, won't the USB screen disappear as soon as they eject the SD card (in your example) ? That should give them a clue that something has happened, and they need to reattach somehow.
16:32:34*linuxstb will probably regret having such faith
16:33:21gevaertsThe problem as I fear it is that they won't have read the manual, so they won't know (a) that it's possible to reattach, and (b) how to do it. So they will think it's a bug...
16:34:32gevaertsOf course we could put the reattach option in the root menu, but then it would also be there when not connected, which is strange as well (unless we make it appear and disappear, but a number of people are opposed to that as well)
16:34:41linuxstbIf the "USB Screen" menu item goes in the main menu, and we make the USB screen always exit to the main menu, they shouldn't need to read it...
16:34:54linuxstbNo, the menu item would only appear when connected.
16:35:07 Quit axionix (Read error: 110 (Connection timed out))
16:35:11*gevaerts proposes just removing support for the micro-sd card from rockbox. That would neatly solve this problem
16:36:02linuxstbIn general, I agree with the people that don't like disappearing menu options, but I think in this case it makes sense.
16:36:22 Quit axionix_ (Read error: 110 (Connection timed out))
16:37:46gevaertsAnyway, the sansa sd driver has to be fixed first to properly send those inserted/removed events during a usb connection. I guess I'll work on that first (unless amiconn beats me to it, he talked about _maybe_ doing it this weekend)
16:39:24 Join midgey [0] (
16:39:33gevaertsCurrently, the reconnect command in my tree is in the debug menu. Maybe I'll leave it there and let someone else decide where it should properly go.
16:40:11bluebrotherwhy not add support for greying out menu items? That way such an entry won't disappear but make clear it's not working in all situations
16:43:23 Quit faemir (Remote closed the connection)
16:43:42 Quit csc` ("Powering Off")
16:44:15gevaertsThat would work as well, but it would clutter the menu unnecessarily in the most common case (when the player is not connected). On at least the c200, the main menu already doesn't fit on screen with the default font.
16:46:02*gevaerts wants to work on a new plugin that doesn't interact with all sorts of other subsystems in new and unexpected ways
16:49:38bluebrotherwell, or just leave the reconnect option in the debug menu ;)
16:51:28*gevaerts likes that idea
16:51:37mcuelenaerehi, I've got some questions about previous porting problems
16:51:45mcuelenaere(I'm the one working on the Creative ZVM port)
16:51:56mcuelenaerewere there any troubles with thread switching in the past?
16:52:02 Join faemir [0] (
16:55:21 Quit nplus (Remote closed the connection)
16:55:35 Quit axionix_1 ("Lost terminal")
16:55:35 Quit axionix__ ("Lost terminal")
16:55:47 Join nplus [0] (n=nplus@141.25.Globcom.Net)
16:56:09 Join axionix [0] (
16:56:15peturmcuelenaere: I think amiconn had some issues recently
16:56:31mcuelenaerepetur: which?
16:57:14petursomething with code alignment
16:57:54mcuelenaereah, that sounds familiar
16:58:06 Quit jurrie__ (Read error: 110 (Connection timed out))
16:58:14mcuelenaerebut weren't those related to PP5020 development?
16:58:21BigBambi_5002 I believe
16:58:22peturamiconn and jhMikeS are the ones to contact
16:58:35mcuelenaere5002 indeed
17:03:29 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:03:57 Join mirak [0] (
17:06:00 Quit mirak (SendQ exceeded)
17:06:59 Join mirak [0] (
17:09:46 Quit Rondom (Nick collision from services.)
17:10:04 Join Rondom [0] (
17:11:05 Join toffe82 [0] (
17:13:33***Saving seen data "./dancer.seen"
17:19:11 Join rasher [0] (
17:20:54 Join jurrie [0] (
17:22:08 Quit rasher_ (Read error: 145 (Connection timed out))
17:29:39 Join FOAD_ [0] (
17:38:49 Join krazykit [0] (
17:44:59 Quit midgey (Read error: 104 (Connection reset by peer))
17:45:30 Quit FOAD (Read error: 110 (Connection timed out))
17:45:31 Nick FOAD_ is now known as FOAD (
17:45:40 Join DaCapn [0] (
17:47:13 Quit Nico_P (Remote closed the connection)
17:48:41 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
17:51:35 Join Davide-NYC [0] (
17:52:26Davide-NYCdoes anyone here now have edit permissions for the tracker>
17:53:05Nico_PI have
17:53:44*petur too
17:54:04 Join DerDome [0] (
17:54:05 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
17:55:29Davide-NYCOK, this task should be a bug and not a feature request. Also please delete my inane comments while you're at it. ;-P
17:56:24Davide-NYCEdit: My first comment is valid. (the supportive one)
17:57:02*Lear just noticed that mono AAC files causes the simulator to crash.
17:57:06peturbut I can't undo the close request
17:57:45Davide-NYCThen we need to fix it quick! (hehe) I also feel that this is a serious bug. I'm not sure how to clear overload (from a UI perspective). Petur, what do you think?
17:58:28Davide-NYCPressing stop takes you back to the main menu (usually). Maybe that should be when the OL indicator gets cleared?
17:58:54 Join lesjohn [0] (n=johnl@
17:58:55peturyes, valid point, although you can see the value change in the peakmeters itself
17:59:21lesjohni'm trying to build rbutil and get the error "make[1]: *** No rule to make target `rbspeex', needed by `rbutilqt'. Stop."
17:59:41 Quit seablue ("life, death, life, death")
18:00:24Davide-NYCpetur: if the OL condition id occaisional and a new gain setting is set, and the taper is not staring at his/her meters and OL reset could be very useful
18:00:30Davide-NYCs id/is
18:00:53*petur has a quick peek at the code
18:01:33lesjohncan anyone help me compile?
18:01:38lesjohni ran svn co svn:// ~/Desktop/rockbox
18:01:43lesjohnthen switched to rbutil
18:02:02lesjohnran qmake, then make, and got the error "*** No rule to make target `rbspeex', needed by `rbutilqt'. Stop."
18:02:12Davide-NYCpetur: before you do that let me mention that the other nagging recording bug is lack of hold switch support.
18:02:17BigBambi_lesjohn: Do you have qt devel packages installed?
18:02:34peturDavide-NYC: ?
18:02:56lesjohnbigbambi: i'm not sure... how do i install them? (i'm a new ubuntu user)
18:03:11Davide-NYCpetur: apologies, my brain has a bad memory leak as of late.
18:03:40lesjohnwould it be qt4-dev-tools?
18:04:16 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
18:04:36*gevaerts finds all sorts of unexpected issues with this new eject handling
18:05:02 Quit lee-qid ("aufwiederbyebientotsayonara")
18:05:07BigBambi_lesjohn: I think so. Have you checked ?
18:05:19lesjohnbigbambi: no, i'll look at it
18:05:24lesjohnbigbambi: thanks a lot
18:06:03Davide-NYCfor the logs: the hold switch support is missing in recording for targets *without* hold switches. (FS #1055)
18:06:05gevaertslesjohn: try qmake-qt4 instead of qmake. On some setups "qmake" is for qt3, and rbutil requires qt4
18:07:01*petur remembers coding a virtual hold switch for recording as a special request... worked but did have issues
18:09:15 Nick BigBambi_ is now known as BigBambi (n=Alex@rockbox/staff/BigBambi)
18:10:31gevaertsThe new eject handling means that rockbox sometimes returns "not ready" when the OS tries to scan the partition table, especially on disk targets. This is not an error, and on polling again a bit later everything will work. The catch is of course that the linux kernel as such doesn't poll again. Windows polls, and I think gnome and macosx also do, so this is only a problem for people like me. It's possible to poll manually by accessing the drive
18:10:49gevaerts... somehow (which might fail the first time, but it tells rockbox we need it, and then
18:10:49gevaerts+everything works)
18:11:04*gevaerts doesn't like te irc line length limit
18:11:50 Quit jhulst ("Konversation terminated!")
18:12:22gevaertsSo the question is : is this a real problem, or can we ignore it ?
18:15:03 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
18:15:31 Quit Rondom (Nick collision from services.)
18:15:49 Join Rondom [0] (
18:15:52Davide-NYCFor the H1x0 modders: If "hold_on + usb_connect = not_diskmode" were implemented for the H1x0 it would only take a single wire mod to make the charging jack obsolete. If this feature were ever made available I would mod my player the same day and make a wiki page about it.
18:18:23 Quit Rondom (Nick collision from services.)
18:18:41 Join Rondom [0] (
18:19:20 Quit Lear ("ChatZilla 0.9.81 [Firefox 3.0b5pre/2008030607]")
18:20:17peturDavide-NYC: re 6147: you do mean clearing the clipping bar, not peak hold indication, right?
18:21:06Davide-NYCyes. Since I am fairly conservative with my levels generally I never noticed.
18:21:24 Join spiorf [0] (
18:21:31Davide-NYCIf you clip, it stays set until reboot!
18:21:32peturok, doing some last checks and then I'll commit
18:21:44peturno t doesn't
18:21:47Davide-NYChow do you clear it?
18:22:08peturyou set the clip hold time in display settings -> peakmeter
18:22:24peturmaybe the default is 'eternal'
18:22:31peturor you have it set that way
18:22:45peturmine is 5sec because cliplight uses it
18:23:42Davide-NYCOh boy, you are right (of course)
18:24:08Davide-NYCThe default is 60 sec. Which seems reasonable. So I guess that 6147 is moot?
18:24:29peturno, clearing is a good thing
18:24:33 Quit Arathis ("Bye, bye")
18:24:43Davide-NYChow would you clear it?
18:24:56Davide-NYCon any change?
18:25:05peturany gain change yes
18:25:14 Join Arathis [0] (
18:25:21Davide-NYCAwesome. I will test.
18:26:38peturI'm just puzzled by the function that is already there, it changes the 'eternal' flag too... must see that I don't break anything
18:27:30Davide-NYCWhat is LinusNielsenFeltzing 's IRC nick? I have a question about the ZIFToATAAdapter wiki page.
18:27:57peturit is in the irc nick wikipage, you know...
18:29:37 Quit lesjohn ("Leaving.")
18:31:23roolkuHas anybody got an opionion and/or suggestion what to do about this issue:
18:32:19roolkuare there other targets that use a superfloppy?
18:32:37peturhehe, I had a similar issue recently at work
18:32:54bluebrotherdetecting if the disc is superfloppy or not needs 4 seconds?
18:33:12*Davide-NYC wonders why all of the userIDs are not reconciled...
18:33:13peturwe could make rockbox support partitiontable-less disks...
18:33:13 Quit saratoga ("CGI:IRC (Ping timeout)")
18:33:43bluebrotherisn't superfloppy partitionless?
18:34:00roolkubluebrother: yes, on that horrible microdrive
18:34:18roolkuyes, superfloppy has no partitions
18:34:36bluebrotherwell, wouldn't it be sufficient to simply read sector 0?
18:34:41peturyes, the first sector is just the first sector of the partition
18:34:47bluebrotheror does reading that sector require 4 seconds?
18:35:29peturI guess rockbox goes hunting for partitiontables
18:35:46bluebrotherAFAIU sector 0 should hold the FAT BPB on a superfloppy. And a partitiontable on a partitioned drive
18:35:48roolkuI have not investigated why it is so slow - currently it is trying to mount all 4 possible partitions from what I read in the source
18:36:05peturbluebrother: correct
18:36:17bluebrotheryou posted that Rockbox already supports superfloppy ... or is this only for the MR port?
18:36:18 Quit quaal (Read error: 113 (No route to host))
18:37:04roolkurockbox tries to mount all 4 partions and if it fails it assumes it is a supperfloppy
18:37:16bluebrotherhmm. You need to read sector 0 anyway to get the partition table information. Mounting won't work for superfloppy as there is no valid partition table
18:37:22roolkuin the m:robe port the mount attempts are ifdefed out
18:37:38bluebrotherbut why takes finding this out that long? You just need to look at 4 * 16 bytes ...
18:37:51roolkubeats me
18:38:39*petur will have a peek after committing the peakfix
18:39:28bluebrotherroolku: which file does the mounting?
18:41:09 Join gregzx [0] (
18:41:15bluebrotherfound it
18:43:28roolkumaybe the logic could be reversed (m:robe only): try supperfloppy first and it fails try the partitions...
18:43:34peturDavide-NYC: committed...
18:45:04bluebrotherhmm, seems it checks every partition entry for a FAT
18:45:46bluebrothershouldn't it check the partition type before looking for a FAT?
18:45:53bluebrotheror am I overlooking something?
18:46:30 Join dandin1 [0] (
18:47:06bluebrotheri.e. check the value of pinfo->type against the FAT types list
18:48:18 Join goffa [0] (n=goffa@
18:48:19bluebrotheras far as I see for each entry fat_mount() is called. Which will try to read the BPB and thus read another sector
18:48:28roolkuis this reliable enough, or is there a risk that the partition is marked something else and still fat formatted
18:49:17goffahmmm.. updated rockbox on my gigabeat f60 last night.. getting a wierd error
18:49:28bluebrotherit's possible that this happens. But when the partition is marked using a different partition type than it really has the host OS most likely won't detect that drive
18:49:30goffait plays for about 30 mins.. then stops mid track
18:49:46goffai can restart the track and it plays through
18:49:57goffaand it goes another 30ish minutes
18:50:09bluebrothergoffa: is it possible the 30 min equal to the audio buffer?
18:51:28goffaanti skip is set to 5 seconds
18:51:43goffais there somewhere else i should be looking?
18:53:05goffaits not an exact 30 mins.. just approx
18:53:23goffacould be 20.. could be 40
18:54:19gevaertsgoffa: are your files encoded at approx. 128kbps ?
18:54:45goffaum... some could be... i'd say 99% no
18:54:47bluebrotherroolku: maybe a simple check like this?
18:55:18bluebrothercould be ifdeffed for m:robe only also.
18:55:26bluebrother(haven't checked the change)
18:56:17goffachecking tracks i know it locked on....
18:56:21gevaertsI asked because 30 mins at 128kbps is about the buffer size.
18:56:35roolkuI'll test what the improvement is...(still a little reluctant to render a targets with exotic partition types not working)
18:56:58goffayeah... this one was 240... one was 196 (these are vbr)
18:57:02goffastill looking
18:57:38roolkudid I say 4s - it's taking 14 now!
18:57:39bluebrotherroolku: I think we can safely assume the partition to hold the correct type −− at least windows won't recognize it otherwise.
18:57:40goffai'd say 128 isn't the issue
18:57:46linuxstbbluebrother: Could disk_init() check if sector 0 is a partition table, or if it's a "superfloppy" ?
18:57:52Davide-NYCpetur: what was wrong with CF6? I haven;t experienced and real issues. (though I have not done any real rigorous testing either)
18:57:58bluebrotherroolku: with or without the change?
18:58:07linuxstbbluebrother: I think it would be nice to remove #ifdefs if possible, not add them
18:58:15bluebrotherlinuxstb: I agree.
18:58:33roolkuwithout the change, wanted to confirm the measure
18:58:55goffausing r16550m-080308
18:58:56 Quit amigan_ (Read error: 104 (Connection reset by peer))
18:59:00goffai'm using 2 patches as well
18:59:21goffaclick insert... and an albumskip patch
18:59:45 Quit Horschti ("Snak 5.3.3 Unregistered copy. Evaluation period is over. Program will now quit. Thanks for using Snak.")
18:59:46bluebrotherwow, 14 seconds is huge
19:00:34goffabbl.. have to run to the grocery store...
19:00:39gevaertsgoffa: I'll leave it to the real playback experts.
19:01:01bluebrotherhmm, disk_mount calls disk_init, which reads sector 0
19:01:12goffai just figure'd i'd bring it up in case others were experiencing similar probs
19:01:27bluebrotherand disk_mount is called by disk_mount_all four times.
19:01:45gevaertsgoffa: what might be useful to know is the total size (in kb) of all the tracks up to the point where it stops
19:02:25*petur away for a while
19:02:30*linuxstb wonders what the albumskip patch is
19:02:46*gevaerts now shuts up. I don't want to get involved in playback debugging
19:02:47 Join japc [0] (
19:03:03linuxstbWith the recent changes to playback events, maybe that patch is breaking things
19:03:06bluebrotherfrom what I can see disc_mount is never called directly −− so disk_init could be called in prior and the pointer get passed directly.
19:03:31roolkubluebrother: with your patch it is not measurable, so this could be a solution if if doesn't break anything
19:03:48bluebrotherso mounting itself does work? ;-)
19:04:09roolkumounting as superfloppy yes
19:04:25bluebrotheranyway, as far as I can see there sector 0 gets read multiple times in case the fat partition isn't the first one
19:04:31*roolku compiles bootloader for gigabeat now...
19:04:47bluebrotherwhich could also be a speed hog on other players −− like Ipods with the data partition being the 2nd
19:05:09*bluebrother wonders why disk_mount() is exported
19:06:00*roolku has second thoughts - if this fails I will have 'bricked' my player :)
19:06:25*gevaerts has to go now
19:06:30 Quit gevaerts ("back later")
19:07:42bluebrotherroolku: why are you trying a bootloader?
19:08:06roolkuchanged my mind - rockbox binary uses the same function
19:08:28 Join Arathis2 [0] (
19:08:52 Quit PuppiesOnAcid (Nick collision from services.)
19:10:37 Quit Rondom (Nick collision from services.)
19:10:55 Join Rondom [0] (
19:11:15 Join Mathiasdm [0] (
19:11:38bluebrotherroolku: I'm trying to move the disk_init() call out of disk_mount(). Should give a speedup too.
19:11:44 Join Mathiasdm2 [0] (
19:11:45bluebrotherunless I'm doing it the wrong way ;-)
19:11:46 Quit Rondom (Read error: 104 (Connection reset by peer))
19:12:41 Join alienbiker99 [0] (
19:12:41 Quit Doomed_ (Read error: 104 (Connection reset by peer))
19:13:15 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
19:13:37***Saving seen data "./dancer.seen"
19:15:27roolkubluebrother: gigabeat works fine with the change
19:15:50bluebrotherlet me check my change too ...
19:16:29*bluebrother wonders if this gives a speedup on mini
19:17:58Davide-NYCis there a specifid amount of time that has to pass before another clip event is registered in the clip counter?
19:18:49 Quit mcuelenaere (Read error: 110 (Connection timed out))
19:20:01 Join mcuelenaere [0] (
19:20:01*bluebrother measured about half a second difference ... if measuring worked correctly
19:22:03 Quit Davide-NYC ("ChatZilla 0.9.81 [Firefox]")
19:22:18roolkuare you going to commit?
19:22:55bluebrothernow it's faster with the unchanged version
19:24:37 Quit Arathis (Read error: 110 (Connection timed out))
19:28:01 Join ali_as_ [0] (
19:28:01 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:28:01 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
19:29:37 Quit BigBambi (Read error: 110 (Connection timed out))
19:29:49 Quit Mathiasdm (Read error: 110 (Connection timed out))
19:30:24bluebrotherroolku: I tried some variations. Interestingly the first patch I posted was the fastest −− as far as I can measure.
19:30:43bluebrotherbut I'm somewhat unsure if this is really a good solution. I don't expect any issues from it though.
19:31:21bluebrotheranyway, I'd like to hear some other thoughts about this −− maybe petur ?
19:31:45 Nick Mathiasdm2 is now known as Mathiasdm (
19:31:53roolkumaybe would ask one of the oldtimers - they may think of potention pitfalls or remember obscure partition types
19:31:53linuxstbbluebrother: What does your patch look like?
19:32:17bluebrothermy "improved" version is here:
19:32:31 Nick BigBambi_ is now known as BigBambi (n=Alex@rockbox/staff/BigBambi)
19:32:43roolkus/would/we should/
19:33:58bluebrotherI should get some food now. Be back in a while (hopefully with positive remarks then ;-)
19:37:00*petur has visitors... not paying attention
19:37:38 Join codesquid_ [0] (
19:38:20 Join webguest76 [0] (n=48e601c5@gateway/web/cgi-irc/
19:38:37webguest76hi, does anyone know if its possible to use rockbox with an ipod that has a dead battery
19:39:05webguest76im trying to get it to work on mine but whenever I boot the screen just comes up blank and none of hte bottons work, I've heard it could be becaues the battery is dead ( i have it plugged in)
19:41:58 Quit webguest76 (Client Quit)
19:43:17 Quit ali_as (Read error: 110 (Connection timed out))
19:50:19roolkubluebrother: I am afraid the 'improved' version by itself is no improvement over the 14s
19:52:19 Quit nplus ("Leaving")
19:58:07bluebrotherroolku: I had the impression something was fishy with that version myself.
19:58:24bluebrotherwill check it later in detail.
19:58:32 Quit Zarggg (Read error: 104 (Connection reset by peer))
19:58:48 Join Zarggg [0] (
19:59:42 Quit moos ("bbl")
20:00:04 Join nplus [0] (
20:00:09 Join Crash91 [0] (n=Crash@
20:11:20 Join GrooveStix [0] (n=chatzill@
20:11:32 Quit krazykit (Read error: 110 (Connection timed out))
20:15:49 Quit Arathis2 ("Bye, bye")
20:20:29 Quit Nevtus ("Gone")
20:23:29 Quit spiorf (Remote closed the connection)
20:24:23GrooveStixhey people, anyone here?
20:24:47GrooveStixI am having a bit of a trouble with my Gigabeat
20:25:42GrooveStixand since I my hdd died few weeks ago, I got a newer one but when I loaded the dummy firmware and .rockbox, I get an error "00000002" everytime I reboot
20:25:48GrooveStixwhat am I doing wrong?
20:26:02GrooveStix(I believe the new hdd was used in an iPod)
20:27:30linuxstbHow did you format it?
20:29:19GrooveStixwith the winXp format
20:30:09linuxstbDid you repartition it? I don't know if it matters, but ipods have two partitions (first if for the firmware, second is FAT32) , whereas the gigabeat disks just have one.
20:30:47GrooveStixno I didn't do any of that, how do I know how many partitions I have
20:30:54GrooveStixI only see one drive in My COmputer
20:31:11linuxstbThe first partition is marked "empty", so won't appear as a drive.
20:31:46linuxstbI don't know what the Windows disk partitioning tool is, but I'm sure google will find some.
20:33:23GrooveStixI found it
20:33:25GrooveStixlemme see now
20:34:18GrooveStixyes, I see that there is 39MB unallocated on the drive
20:34:47GrooveStixcould that be the problem?
20:36:25GrooveStixmr. linuxstb ?
20:36:37 Quit desowin ()
20:36:39 Quit XavierGr (Nick collision from services.)
20:36:49 Quit jhulst (Read error: 110 (Connection timed out))
20:36:52 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:39:14 Quit Crash91 (Read error: 110 (Connection timed out))
20:40:47GrooveStixYES! the partitions were the problem! I just deleted the big one, and then I repartitioned the unallocated space and that did it! thanks for the hint linuxstb.
20:44:44SlasheriJdGordon|zzz: hi :) btw, does the gui list code support a multi-selection list (something like a checkbox list)?
20:45:59Slasherii was planning to add a category tag to the db, so you could put one song in a multiple user named categories (at maximum 32 different categories)
20:46:24 Join quaal [0] (
20:53:46 Join desowin [0] (n=desowin@atheme/developer/desowin)
20:56:24 Quit mf0102 ("Verlassend")
20:56:44 Quit GrooveStix ("ChatZilla 0.9.81 [Firefox]")
20:58:17 Quit waldo (Remote closed the connection)
20:58:41bluebrotherroolku: just thought about it −− my second version _can't_ be an improvement.
20:59:19linuxstbSlasheri: No, it doesn't...
20:59:46*bluebrother out for reboot, back in a bit
20:59:51 Quit bluebrother ("leaving")
21:00:10 Join waldo [0] (
21:02:33 Quit ol_schoola ()
21:04:13 Join tessarakt [0] (
21:05:43 Quit waldo (Remote closed the connection)
21:05:59 Join waldo [0] (
21:07:19 Join midkay_ [0] (
21:09:12 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
21:10:54 Quit waldo (Remote closed the connection)
21:11:09 Join waldo [0] (
21:13:39***Saving seen data "./dancer.seen"
21:14:59 Join Nevtus [0] (n=Nevtus@unaffiliated/nevtus)
21:15:09 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:21:39*petur returns
21:22:28 Join krazykit_ [0] (
21:23:49 Nick krazykit_ is now known as krazykit (
21:24:01 Join spiorf [0] (
21:24:20 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
21:28:58 Quit quaal (Read error: 110 (Connection timed out))
21:30:44 Quit Mathiasdm ("Yuuw!")
21:33:13 Quit midkay (Connection timed out)
21:39:02 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
21:40:54 Join Zagor [242] (
21:43:39 Quit jhulst (Connection timed out)
21:50:04 Join jurrie_ [0] (
21:51:04 Quit jurrie (Connection timed out)
21:52:59 Quit dan_a (Read error: 110 (Connection timed out))
21:55:12 Join gregrokz [0] (n=4c5d86e7@gateway/web/cgi-irc/
21:55:47gregrokzi have doom on my 5.5 gen ipod but the controlls for the game dont work...
21:57:22gregrokzdo you know what the problem might be?
21:57:41linuxstbWhat happens when you run Doom?
21:58:23gregrokzit loads up all the files that are required but once it gets to the actual gameplay i cant move around
21:59:09 Quit nplus ("Leaving")
22:01:16gregrokzdo you know what it is?
22:03:12linuxstbNo, I'm just trying it now.
22:04:48linuxstbI just started Doom, selected "Play game" from the first menu, then a few seconds later Doom started in demo mode. I brought up the Doom menu (toggling the hold switch), then selected new game, and it works...
22:05:45gregrokzwhere did you get your "doom" from?
22:05:57linuxstbThe Rockbox wik page
22:07:57gregrokzile try it again
22:11:06 Join stripwax [0] (
22:12:07 Join robin0800 [0] (
22:14:44peturbluebrother: disk_init() is indeed a bit simple in accepting the validity of the partitiontable, certainly because both MBR and Partition first sector end with 55AA....
22:17:36 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
22:18:05 Part hachi
22:20:51 Quit merbanan (Remote closed the connection)
22:23:28bluebrotheranother way of making it smarter would be to check if the calculated start sector of the partition does exist at all (and is != 0)
22:26:17 Join WatchDragon [0] (
22:26:44WatchDragonDoes Rockbox shuffle much like winamp does?
22:26:58bluebrotherwhat should "shuffle much" be?
22:27:09 Quit gregrokz ("CGI:IRC (EOF)")
22:27:27WatchDragonis it similar?
22:28:27Zagorhow does winamp shuffle?
22:28:48*bluebrother never understood this fuss about shuffling
22:29:26linuxstbRockbox shuffles like a deck of cards
22:29:31ZagorI can appreciate the difference btwn shuffle and random
22:30:06 Quit Xerion (Read error: 104 (Connection reset by peer))
22:30:34WatchDragonbluebrother, because of you have a radio like in my truck, its "Shuffle" feature is alaywas... 4, 10, 12, 85, 2, 1
22:30:45WatchDragonno matter what cd/mp cd you put in
22:30:53*bluebrother notices "Olympus" being mistyped in the according new ports thread :o
22:31:12ZagorWatchDragon: haha, that's a nice "shuffle"
22:31:15bluebrotherwell, then that's not shuffle.
22:31:17linuxstbWatchDragon: So you're asking if Rockbox's shuffle is broken like your truck's radio?
22:31:25WatchDragonnooo, lol
22:31:33*bluebrother still has no idea how winamp does and why this is related
22:32:01amiconnbluebrother: The dist_init() call in disk_mount() is necessary
22:32:54bluebrotherwhy is it necessary in disk_mount?
22:33:20peturamiconn: diskinit should check if that sector is indeed an MBR and not the start of a partition (superfloppy)
22:33:28bluebrotheras long as the disk itself doesn't change sector 0 shouldn't change too
22:33:39amiconndisk_init() reads the partitioning. There are at least 2 situations where it is important to re-read:
22:34:01amiconn(1) if the disk was repartitioned in usb mode (2) for all hotswap targets
22:34:43bluebrothersure, but it isn't necessary to do this multiple times when calling disk_mount_all
22:35:53bluebrotherso disk_mount_all could call disk_init() and make disk_mount() use the result, while in other cases disk_mount() will do the init itself
22:36:02amiconnIt isn't called multiple times
22:36:35amiconndisk_mount_all() calls disk_mount() once per drive, which in turn calls disk_init() once
22:37:17Zagoryikes, /me created a fork bomb on the server...
22:37:39Zagorload average: 272.69, 113.23, 47.43
22:38:05Bagder272 is a pretty high load
22:38:13bluebrotheryou're right. Somehow I mixed things up. My bad :(
22:38:13Zagorsort of...
22:38:20amiconnI can't see how any of these calls can be skipped safely
22:38:24bluebrotherZagor: a bit ;)
22:39:00bluebrotherwell, the simple solution would be to check the partition type and skip all partitions that aren't FAT
22:39:07amiconnpetur: disk_mount works by trial and error. Simple, and it works
22:39:24roolkuamiconn: would the check for partition type bluebrother suggested be acceptable
22:39:25amiconnIt tries each partition first, and if none works, finally tries superfloppy
22:39:42roolkuamicon: but this takes 14s on m:robe
22:39:54 Join webguest46 [0] (n=46a2cc3b@gateway/web/cgi-irc/
22:40:03webguest46hello everyone
22:40:20webguest46i had a question regarding the plugins
22:40:28amiconnroolku: huh? Then something is wrong with the driver on mrobe100
22:40:47amiconnAfaik it uses the same hdd as the h10, and that only needs a fraction of a second
22:41:01webguest46does anyone know if there is a website that has more plugins, i am looking for a checkbook plugin for rockbox that i can balance my checkbook with and have been searching but cannot find
22:41:15bluebrotherwebguest46: no.
22:41:19roolkuamiconn: is h10 formatted as superfloppy?
22:41:49amiconnThat is, I don't really know, need to check
22:41:50roolkuamiconn: maybe that is why. on m:robe it has to check all 4 partitions
22:42:03bluebrotherthing is that the current code tries to read a partition table from sector 0 and tries to mount each "partition" that is referenced there.
22:42:22linuxstbDoes the ATA driver check for valid sector values?
22:42:30amiconnWell, trying to mount a partition means reading a single sector. If it's as slow as you reported, something is clearly wrong
22:42:40bluebrotheras far as I understand not.
22:42:49roolkuthe m:robe has 4 co-incidental partitions as one can see in the debug menu, even though it's a superfloppy
22:43:02amiconnOn ipod video, we try each sector size, as we don't know what's in until we tried
22:43:14 Quit saratoga ("CGI:IRC")
22:43:20bluebrotherroolku: what's the start sector of those partitions?
22:43:30 Join axionix_ [0] (
22:43:39amiconnThat means that on a G5.5, it usually takes > 12 tries. And even that only takes an unnoticeable fraction of a second
22:43:50linuxstbamiconn: Yes, are probably still valid sectors, just wrong ones. I'm guessing on the mrobe the driver is seeking invalid sectors...
22:44:06linuxstb^insert "those" near the start
22:44:13amiconnI'm not sure whether the ata driver does range checking.
22:44:20amiconnIf it doesn't, that should be added
22:44:22linuxstbI've just checked, and I don't think it does.
22:44:30amiconnThe MMC driver definitely does
22:44:44webguest46thank you bluebrother
22:44:49roolkubluebrother: rather large numbers...smallest one is 2020534f
22:44:55 Quit webguest46 ("CGI:IRC")
22:45:18 Join disorganizer [0] (n=5b11da19@gateway/web/cgi-irc/
22:45:18amiconnIn the beginning of the Ondio port, it even used to panic on out-of-bound accesses, but Jrg changed that into an error later, because otherwise a similar effect (MMC formatted as superfloppy) made rockbox panic...
22:45:24bluebrotherdoesn't sound like a valid sector ...
22:45:51amiconnSo, we need range checking in the ata driver.
22:46:21amiconnThe partition type check might or might not be a good idea. We need to make sure to cover all types
22:46:32 Join Kqkq [0] (
22:46:46linuxstbIs the number of sectors in the identify_info ?
22:46:57amiconnAnd the posted patch does unnecessarily many checks
22:47:04roolkumaybe the disk doesn't like out-of bound access or is that checked
22:47:42*roolku is slow
22:47:56amiconnyes :P
22:48:57bluebrotheryep, that's what I disliked about checking all known partition types
22:50:22amiconntypes 0xb..0xf are contiguous
22:50:37amiconnSo the check can easily be shortened from 7 to 4 comparisons
22:52:18roolkuwhile the partition type check is probably good enough, there is still the risks of false positives
22:52:25amiconnHmm, type 0xd doesn't exist
22:52:57roolkuI wonder how other OS solve this - must be a common problem
22:52:57amiconnAnd checking partition types isn't a good idea for those 160GB archos users
22:53:24bluebrotherwhy that?
22:53:27linuxstbWhat about checking for superfloppy first?
22:53:31roolkucould we inverse the logic
22:53:40amiconnThese users want to hide the second partiton by setting an unusual partition type, in order to avoid overwriting innocent data
22:53:52amiconn(the usb-ata bridge in the archos doesn't support lba48)
22:54:16bluebrotherthere are "hidden FATx" types around
22:54:18amiconnroolku: Imo the only proper fix would be to add range checking to the ata driver
22:55:06roolkuamiconn: yes, but it is an annoying overhead for every sector access
22:55:37 Join mcuelenaere [0] (
22:55:57amiconnImo it's a necessity... and the overhead is neglectible
22:56:39amiconnIt's not to be checked for every sector, but one per call to ata_read_sectors() or ata_write_sectors()
22:57:03 Part Kqkq
22:57:05roolkuhow would an out-of-bounds be handled - empty read? error code?
22:57:36amiconnReturn an error
22:58:42amiconnActually, in the mmc driver it's only ata_read_sectors() that returns an error on out-of-bounds access. ata_write_setcors() would panic
22:59:02 Quit axionix (Read error: 110 (Connection timed out))
22:59:14amiconnSee ata_mmc.c lines 698ff and 816ff
23:00:45 Join Mathiasdm [0] (
23:00:52 Join maddler [0] (
23:02:00 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
23:03:13 Quit corevette (Remote closed the connection)
23:03:17 Nick midkay_ is now known as midkay (
23:03:34 Quit mirak ("Ex-Chat")
23:05:15roolkuso one would just need to obtain the disk size/last available sector in ata_read_sectors() somehow...
23:05:24 Join corevette [0] (
23:05:32linuxstbThat can be set in ata_init - from the identify_info IIUC
23:05:33amiconnThat "somehow" is even half done already
23:06:34amiconnIt's in the identify info, and the LBA48 code checks for the standard 28 bit sector count to be set to its maximum value, and the lba48 feature set supported, in order to switch to lba48
23:06:51 Join daywalker_ [0] (
23:07:17daywalker_wow, i didn't expect such a crowded channel :)
23:07:41amiconnThe lba48 sector count is also in the identify info, at a different position. The rockbox lba48 code only supports "lba32" - still enough for up to 2TB
23:12:52amiconnlba28 sector count is in words 60..61, lba48 sector count is in words 100..103
23:13:16amiconn(yes, they wasted a word)
23:13:40***Saving seen data "./dancer.seen"
23:14:06amiconnOne single question remains - what sector count does the MK8011GAH report?
23:14:35amiconnAn identify info dump from a G5.5/80GB should tell
23:16:09 Quit davina (Remote closed the connection)
23:32:53 Quit maddler ("connection reset by beer!")
23:33:34 Join maddler [0] (
23:34:07amiconnNico_P: around?
23:35:06amiconnCould you provide an identify info dump from the G5.5?
23:35:22Nico_Psure. what do I need to do?
23:36:37Galoisso turn off iptables
23:36:39amiconnGo to the debug menu, and execute "Dump ATA identify info"
23:36:54amiconnThen grab the .bin file from the root
23:38:04Nico_Pamiconn: is DCC ok by you?
23:38:26*petur thought amiconn said that dump ata function would not be used :p
23:38:43amiconnYes. You could also attach it to the cf mod wiki page though, as we're collecting identify dumps there
23:39:29Nico_Pamiconn: can I let you do that?
23:39:45amiconnThanks for the file
23:40:21Nico_Pyou're welcome :)
23:40:40roolkuamiconn: does this mean you are implementing the check? don't want to duplicate work
23:41:19amiconnNice, the MK8011GAH simply reports its size in terms of standard 512 byte sectors
23:43:29peturamiconn: could you live with my latest go at the CF problem? I know it is not elegant but I don't want to waste more time on it
23:45:37peturand I still think disk_init shoul check more than 55AA to conclude it is a bootsector, because the first sector of a partition also ends that way
23:46:10peturanyway, gotta run...
23:46:21 Quit petur ("real life....")
23:48:13 Join stripwax [0] (
23:48:59 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
23:49:21amiconnThe MK8011GAH doesn't state apm support
23:51:48 Quit maddler ("connection reset by beer!")
23:52:01 Join maddler [0] (
23:56:24 Join mcuelenaere [0] (

Previous day | Next day