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 2007-11-04

00:00:11saratogathis just powers off the disk's ATA controller when its done?
00:00:28amiconnIt cuts power to the hdd, yes
00:00:55 Join random_desu_is_s [0] (
00:01:04amiconnIt looks like we could even switch off the clock of the PP's ata controller, but this means we need to learn how to init it
00:01:10linuxstbBTW, how did you identify it in the ROM?
00:01:26saratogaamiconn: do we switch it off for the PP5024 ?
00:01:54amiconnJust via the 'HD Power Test' string, and then disassembling a bit above it
00:02:07amiconnThe function looks exactly the same as on mini
00:02:24 Quit lazka ("I'm off now")
00:02:25amiconnG5 has a slightly different test (and it's a different bit)
00:02:34 Quit cooz ()
00:02:35 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
00:02:57Soapsaratoga: considering the Nano which uses the same PP as the Video shares a similar runtime percentage-of-stock...
00:04:12amiconnlinuxstb: Plus following the call chain of course
00:04:36 Join Spiritsoulx [0] (n=eyes_of_@
00:04:49SpiritsoulxHow do you turn a nano to hard drive mode?
00:04:50 Join random_desu_is_s [0] (
00:05:12saratogaactually i guess the 5024 doesn't have an ATA controller, so there'd be no gain for the sansa from disableing it
00:05:38amiconnAre you sure that it doesn't have an ata controller?
00:05:39 Quit JRoT (Read error: 104 (Connection reset by peer))
00:05:48SpiritsoulxHow do you turn a nano to hard drive mode?
00:05:56amiconnBut I would think that the rom already switches it off...
00:06:49 Join JRoT [0] (
00:06:50 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
00:06:50SpiritsoulxHow do you turn a nano to hard drive mode?
00:07:04amiconnIt's a bit in DEV_EN, and DEV_EN is shown in the debug menu
00:07:15saratogaamiconn: the product brief doesn't show one where the PP5022 does
00:07:33amiconnDEV_EN bit 25
00:07:55 Join random_desu_is_s [0] (
00:08:09saratogathough it would be interesting to see if battery benching the sansa on with the retail bootloader is any different then when flashed with ours
00:08:26amiconnJust check that bit...
00:08:51SpiritsoulxHow do you turn a nano to hard drive mode?
00:08:52saratogaoh didn't realize you were speaking to me
00:08:57saratogai'll check right now
00:10:36saratogaC440597F, so its 0
00:10:37 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
00:11:14SpiritsoulxHow do you turn a nano to hard drive mode?
00:11:16saratogausing barry's bootloader if that matters
00:11:40amiconnYes, it's disabled (if it's there at all)
00:11:59 Join random_desu_is_s [0] (
00:12:29saratogagood luck with your hacking
00:13:09 Quit saratoga ("CGI:IRC (EOF)")
00:13:48SpiritsoulxHow do you turn a nano to hard drive mode?
00:14:31BigBambiSpiritsoulx: Pleae stop repeating that
00:14:32 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
00:14:51SpiritsoulxDoes anyone know >___>
00:14:51linuxstbamiconn: Yes, that works ;)
00:15:04amiconnlinuxstb: Nice :)
00:15:34BigBambiSpiritsoulx: People wil answer if they have one, continually repeating yourself is just annoying
00:15:56 Join random_desu_is_s [0] (
00:16:41linuxstbSpiritsoulx: If you mean "disk mode", then see here
00:17:36 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
00:20:06 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
00:21:00amiconnInteresting.... the H10 has 2 bits in GPO32 set, but the mini G2 doesn't
00:23:15 Join random_desu_is_s [0] (
00:24:46 Quit random_desu_is_s (Client Quit)
00:27:00 Nick parafin is now known as parafin|away (
00:27:36 Join linuxstb_ [0] (
00:27:53 Quit linuxstb (Nick collision from services.)
00:27:59 Nick linuxstb_ is now known as linuxstb (
00:29:34 Join colin__ [0] (
00:30:36amiconnHdd poweroff is working like a charm on my Mini G2 and the Video :)
00:31:52 Join bwhiteford [0] (
00:31:56bwhitefordhey everyone
00:32:14 Quit Spiritsoulx ()
00:32:20bwhitefordanybody around?
00:32:48linuxstb146 people and a bot
00:33:18bwhitefordi've got a question about rockbox on a 60gb ipod video
00:33:45bwhitefordit installs just fine, but whenever i try to build the database, it hangs
00:34:43bwhitefordany ideas?
00:35:46 Join lespea [0] (
00:36:18lespeaI was listening to my music and I just had a brilliant (i think) idea
00:36:30bwhitefordman lame room
00:37:02Mouser_Xbwhiteford: Did you read the IrcGuidlines?
00:37:04lespeait would be very cool if one of the "now playing" menus would be a similar tab which would let you quickly view similar songs by album/artist or whatever
00:37:14lespeaala amarok
00:37:26 Quit jhMikeS (Nick collision from services.)
00:37:28 Join JdGordon [0] (
00:37:32 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
00:37:35 Join linuxstb_ [0] (
00:37:49Mouser_Xbwhiteford: I ask, because there's specific rules in this channel, for a reason.
00:37:54 Quit z35 (Read error: 110 (Connection timed out))
00:38:08bwhitefordI understand them Mouser, but it appears that no one is interested in helping.
00:38:09 Quit colin_ (Read error: 110 (Connection timed out))
00:38:12 Quit linuxstb (Nick collision from services.)
00:38:13lespeadoes anybody else think that would be a good idea?
00:38:16 Nick linuxstb_ is now known as linuxstb (
00:38:52scorchelespea: feel free to either work on it, or put in a feature request if one doesnt currently exist for it :)
00:39:51 Part bwhiteford
00:40:04scorchepatient one...
00:41:06lespeascorche: okay cool
00:41:16amiconnI think it would be a good idea to wait a tiny bit after powering on the hdd on all targets
00:42:02amiconnRight now only the H10 does this, to prevent lockups, but it seems cleaner to me in general
00:42:03 Join perrikwp__ [0] (n=chatzill@
00:42:04 Nick perrikwp__ is now known as perrikwp (n=chatzill@
00:42:05amiconnJust 50ms
00:44:48 Join Spiritsoulx [0] (n=eyes_of_@
00:45:09SpiritsoulxAnyone know how the rockboy keys are mapped on a nano?
00:45:47bertrikamiconn: I don't have a HDD player, but 50 ms seems reasonable and imperceptible to the user I think
00:46:07amiconnYeah, compared to the spinup time it's short
00:46:18amiconnAnd I'll rather go 20ms (the H10 currently uses 10ms)
00:47:06lostlogicbertrik: how's the latest build?
00:47:12lostlogicbertrik: better? worse?
00:47:24bertriklostlogic: about the same, i can still trigger it
00:47:33Mouser_XSpiritsoulx: I assume this is the case for all targets, but you can remap the keys in Rockboy, from the menu. Just press buttons until you either quit Rockboy, or find the menu.
00:47:58 Join PaulJam [0] (
00:48:07lostlogicbertrik: I'm feeling defeated by this bug
00:48:21bertrikI hope I'm not overlooking something that's unique to my situation or anything like that, I'll try an official binary
00:48:22Mouser_X:( @ lostlogic
00:48:44 Quit colin__ ()
00:48:52bertrikI'll wipe my entire .rockbox
00:48:59lostlogicbertrik: Oh, I did have one thought −− what kind of tags do you put on your mp3s?
00:49:14lostlogicIIRC, id3v2 and id3v1 write to the beginning of the file
00:49:17lostlogicand that could be a difference
00:49:19PaulJamlostlogic: you closed FS #8040. for me it still doesn't work (H300, r15442)
00:49:28bertriktags were made with tag&rename 3.0.1
00:49:44lostlogicPaulJam: hmm, reopen then :(
00:50:28bertrikthe mp3 on the album I used for testing have both v1 and v2 tags I think
00:50:34 Quit linuxstb ("ChatZilla [Firefox]")
00:50:44lostlogicbertrik: ok, I _think_ that's related to the breakage.
00:50:50 Quit mf0102 ("Verlassend")
00:51:56lostlogicPaulJam: reopened... I thought that my changes of how tracks were unbuffered would fix it but I guess we still have the problem of data not getting copied back to the buffer.
00:52:07SpiritsoulxI can't find the menu. That's the problem.
00:52:22w1ll14m-join #rockbox-community
00:52:27w1ll14mwhoops... ;)
00:52:40Mouser_XSpiritsoulx: Like I said, push buttons until you find it.
00:53:35SpiritsoulxLol, the hold button is the menu. >_<
00:53:36 Quit webguest89 ("CGI:IRC (EOF)")
00:54:01kugelIs there a variable or something in rockbox which contains the actual screen?
00:54:29lostlogicbertrik: is 8077 currently happening mid-song or close-to the end always or what?
00:55:07lostlogicbertrik: hmm, then not tag related, nvm
00:55:37kugeli.e.if I was in the main menu, the variable would return main_menu or something
00:55:51 Quit PaulJam (".")
00:56:21JdGordonkugel: what do you want to do?
00:56:33bertrikI can also reproduce it with an official build
00:57:16lostlogicbertrik: figured
00:58:02bertriklostlogic: do you think it is sansa specific?
00:58:16kugelJdGordon: I want to save the current screen in a variable, and then go to any other screen, and on pressing back go properly to the screen saved in the vairable
00:58:21bertrikwhen the problem occurs, shutdown also takes very long, alsmost like a forced shutdown
00:58:35kugelI've been trying with GO_TO_PREVIOUS but it didn't work
00:58:44JdGordonkugel: it does that already.... but global_status.last_screen is about as close as youll get
00:58:44bertrikcould it be that there's a thread that hangs?
00:58:45lostlogicbertrik: I actually think I was able to trigger it by some trickity splickty skipping on my ipv. So yes, it is sansa specific, but it's really a race condition that is triggered consistently on sansa and not on other targets
00:59:06lostlogicbertrik: there is most likely a thread that hangs, but I don't know which one (codec or buffer or both) or why
00:59:48lostlogicbertrik: my only thought is that the codec gets stuck in a bufgetdata operation that is now blocking while the buffering thread for some reason cannot fill the buffer but I can't seem to find any potential livelocks or races between them
01:00:02lostlogic(I've eliminated several potential ones in trying to kill this bug)
01:01:40Mouser_XYay for bug killing. Too bad it was the wrong one.
01:02:05 Quit Spiritsoulx ()
01:02:08 Quit ompaul (Client Quit)
01:08:15bertriklostlogic: sansa always returns 0 (false) for ata_disk_is_active
01:08:42lostlogicbertrik: yeah, but that should be only an optimization, should not be needed for functionality
01:08:45bertrikif i change this to true, the buffer is refilled as soon as there is room
01:08:54amiconnThat bug in buffering still isn't fixed?
01:08:58bertrikso the buffer never runs out and the problem doesn't occur
01:09:13lostlogicbertrik: yeah, but it also would kill battery life ona disk based player
01:09:16amiconnata_disk_is_active was oroginally never meant to be used outside the ata driver
01:09:22moosamiconn: the saga isn't finished yet :)
01:09:36lostlogicbertrik: I'll try disabling that block of code in my build and see if I can then reproduce but I doubt that's the source
01:09:43bertrikare there any other flash players that return 0 (false) there?
01:10:27 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
01:10:28bertrikthe problem's still there undoubtly, but ata_disk_is_active returning true may hide it
01:10:31amiconnThe other flash targets return the true status
01:11:00amiconnI.e. true while the flash is accessed, false otherwise
01:11:12 Quit Ebert ()
01:13:25 Join colin_ [0] (
01:14:21 Join Ger [0] (
01:14:26***Saving seen data "./dancer.seen"
01:14:59bertrikI'm no buffering.c expert (yet), but on line 1250 (or thereabouts) there is a line "else if (ata_disk_is_active() && queue_empty(&buffering_queue))"
01:15:15bertrikthis line will never evaluate to true for sansa
01:15:24lostlogicbertrik: yeah, it's just an optimization though
01:15:30lostlogicnothing should break if that is always false
01:15:37lostlogic(I'm testing on my video with false inserted there)
01:15:45JdGordonata_disk_is_active() should be renamed or never called on HAVE_FLLASH_STORAGE targes imo
01:16:29amiconnata_disk_is_active should be *removed* completely, or if needed, stay as a static function in ata.c
01:16:56JdGordonsomethin to let apps know the disk is spinning is needed though
01:16:57lostlogicbertrik: no change in behavior by removing that block entirely on ipod
01:17:07amiconnThe modules needing to hook into a spinup should use the callback, that's what it's meant for
01:17:46GerHi, i have sansa e2X0. Can I install Rockbox under Linux? And when I installed Rockbox can I copy photos, videos etc. without the software?
01:18:07JdGordonand they do, but (I said this already..) callbacks which then post an event on another htreads queue might.. the disk might be spun down before that queue reacts so a disk spinup would occur
01:18:07lostlogicamiconn: I'm considering just removing that block for now and we can re-add the optimization using the callback.
01:18:09 Join faemir [0] (i=55d31ec1@gateway/web/cgi-irc/
01:18:39faemirDoes anyone know a good place to get gigabeat f40s?
01:18:53JdGordonlostlogic: just ifdef thet is_active() call out for either the sansa or HAVE_FLASH_STORAGE
01:19:00amiconnUsing the callback both stops those modlues from getting in the way of the access that originally triggered the spinup, and ensures the ata thread (or whatever flash thread) knows that there'll be some extra access
01:19:06toffe82faemir: ebay
01:19:22lostlogicJdGordon: probably better to remove it for simplicity until we do it right, imo
01:19:58faemirtoffe82 heh, I tried gigabeat instead of gigbeat f and it brought out loads of f40s, just what i need! :D
01:20:00 Quit faemir (Client Quit)
01:20:08 Join z35 [0] (
01:20:24amiconnlostlogic: If the callback causes problems with thread synchonisation, there is another idea of mine that I need to implement for proper suspend as well
01:21:03amiconn(registering for sys messages which then need to be replied by the registrant)
01:21:27JdGordonits not so much thread syncro problems.. its more makeing sure the disk is still spininng by the time the other thread is active again
01:21:43 Join FOAD_ [0] (
01:21:54amiconnYeah, and that *is* a synchronisation problem
01:22:00bertrikthread switching isn't that slow, is it?
01:22:11amiconnNo, why?
01:22:24JdGordonno.. but the disk is shutdown as soon as the callbacks are all called..
01:22:56amiconnThe problem is that currently the callback *must* be processed synchronously. The ata thread might shut down the disk as soon as the callback returns
01:23:56lostlogicamiconn: that seems problematic, since buffering could take a fairly long time
01:23:56JdGordonone thig that might work is that if ata_spin() is called in the callback, then the disk will wait the usual time after the callbacks are done before shutttind donw (but not trigge them again untill next spinup)
01:24:28BigBambiGer: Yes and yes
01:24:29bertrikI'm new to this, but perhaps we can do something like the CPU boost counter
01:24:47lostlogicwell regardless, disabled that optimization with a #if 0, reducing a difference between sansa and ipv
01:24:55amiconnJdGordon: That would be nasty...
01:25:03amiconnbertrik, lostlogic: I already mentioned my idea...
01:26:09*JdGordon thinks amiconn's idea might be a good one :)
01:26:12bertrikwhat idea?
01:26:19GerAh okay thx. But the warranty is void?
01:26:22amiconnRegistering for sys messages
01:26:32JdGordonbut not need to register.. just send it out like the usb message
01:26:49 Join faemir [0] (i=55d31ec1@gateway/web/cgi-irc/
01:26:50amiconnJdGordon: That would be bad
01:27:07lostlogicyeah, registering for a message that requires a reply would be fine for buffering's needs.
01:27:08BigBambiGer: Probably technically, but reports seem to say they don't normally check. However, no guarantees.
01:27:09faemirum, can someone tell me where to find the page with battery life on all the players is? I can't find it again >=(
01:27:17amiconnMost threads don't care about such a message, but with broadcast, all threads with a public queue would have to reply
01:27:43GerThanks and good night.
01:27:46 Quit Ger (Remote closed the connection)
01:27:51JdGordonit would be handled in the default handler, so unless a thread needed it there would be no extra code
01:27:51amiconnRight now it's usb only (which btw isn't handled in a 100% clean way too)
01:28:09amiconnThe default handler is for the main thread *only*
01:28:25JdGordonoh... ok
01:28:49 Quit ender` (" A ‘good’ landing is one from which you can walk away. A ‘great’ landing is one after which they can use the plane again.")
01:28:52amiconnWhen I am going to implement this, there will be some more such system messages, for clean handling of other events like shutdown or suspend
01:29:06 Quit Thundercloud (Remote closed the connection)
01:29:12 Join bistouri [0] (i=58a10615@gateway/web/cgi-irc/
01:29:15JdGordonbeore or after viewports :p
01:29:19amiconnThe current shutdown handling is quite messy
01:29:30amiconnMost likely before
01:29:53amiconnLower level first
01:31:33faemiranyone? =[
01:32:00bertrikwhat about a ATA "boost" level that is increased immediately by the callback and prevents the ATA from shutdown, and which is later decreased as each registered thread is completely done?
01:33:10*amiconn fails to see how this is different from registering for sys messages
01:33:28JdGordonslightly simpler to hack in quickly...
01:33:38bertrikoh, sorry, I'm not familiar with sys messages yet
01:33:40amiconnI doubt it
01:34:56 Quit dandin1 ()
01:36:50 Quit faemir ("CGI:IRC (Ping timeout)")
01:37:00bertrikI'm off to bed, have a nice day
01:37:12 Quit bertrik ("bye")
01:37:49 Quit karashata ("Leaving.")
01:38:17 Join HellDragon_ [0] (n=Nocebo@unaffiliated/helldragon)
01:38:18 Join karashata [0] (
01:38:45 Quit FOAD (Read error: 110 (Connection timed out))
01:38:46 Nick FOAD_ is now known as FOAD (
01:38:53 Quit colin_ ()
01:38:59JdGordon:( queue_send() isnt on hwcodec
01:40:28 Quit HellDragon (Read error: 104 (Connection reset by peer))
01:41:04crashdhow will the power control update affect battery life?
01:41:24 Quit HellDragon_ (Read error: 104 (Connection reset by peer))
01:42:13Mouser_Xcrashd: It allows Rockbox to turn off the HDD.
01:42:30jhMikeSJdGordon: so add the #define or remove the requirement for it if you need it :)
01:42:37crashdaye, i know, but has anyone run any benchs on it yet :)
01:42:41Mouser_Xamiconn is testing with it right now, but if it's implemented well, it should save on battery life.
01:42:53Mouser_XNo, not that I'm aware of.
01:42:59crashdfair deal :) that's all i wanted to know
01:43:02crashdcheers Mouser
01:43:10 Join colin_ [0] (
01:43:29amiconnMouser_X: It's committed...
01:44:02Mouser_XSo, it's already turning off the drive? Or, the debug code is commited?
01:44:26Soapa brave soul with an inline ammeter could save a lot of hassle by testing power draw before and after HDD shutdown. ;)
01:44:32Mouser_X(I thought the debug code had been commited. I didn't realize it had already been setup to turn off the HDD and all that.)
01:45:38 Quit moos ("Glory to Rockbox")
01:45:56 Quit colin_ (Client Quit)
01:46:09Mouser_XAh, I didn't see the most recent commits.
01:46:42jhMikeSJdGordon: btw, any queue using that is to be taken as owned by a single thread.
01:46:47 Join colin_ [0] (
01:47:20 Quit bistouri ("CGI:IRC (EOF)")
01:49:17 Quit J3TC- (".•«UPP»•.")
01:51:30kugelwhy does the sim crash when I try to DEBUGF(global_status.last_screen)?
01:51:48JdGordonbecause DEBUGF() is printf
01:52:05JdGordonthats not how printf works
01:52:44kugelis there a function which can put that out?
01:52:53JdGordonyes... printf
01:52:57*kugel is a bit new to C, that's why the dumb questions
01:52:59jhMikeSor DEBUGF :)
01:53:20kugelI thought printf cannot
01:53:41JdGordonit would be useless if it couldntnt... you jst doing it wrong
01:53:42jhMikeSsure it can, if used correctly. should be plenty of references online about printf
01:53:44kugelif DEBUGF is printf, and DEBUGF crashes...
01:54:06*JdGordon remembers printf being in his very first c lesson :p
01:54:26jhMikeSbesides, use examples elsewhere in the code that print variables
01:55:30 Nick fxb__ is now known as fxb (
01:55:43 Join J3TC- [0] (
01:56:51kugelI was here yesterday (not that you think I wouldn't look at references) but I didn't look through there:
01:56:59homielowe Does anyone have this bug when viewing "rockbox info" and the time will only update if there is user input ( i.e. scroll etc.) ?, i'll file a bug report probably though
01:57:12kugelI just notice it's c++ reference
01:57:25JdGordonkugel: printf is the same in c and c++
01:58:22JdGordonhomielowe: dont bother.. ill fix that now
01:58:36TMMBagder: hey man, there?
01:58:41homieloweJdGordon: sounds good :)
01:58:46*jhMikeS preferrs cout << this << that << otherthing in C++ :)
02:00:20 Quit mirak (Remote closed the connection)
02:00:53 Quit karashata ("Leaving.")
02:01:31 Join karashata [0] (
02:01:41*lostlogic hates cout
02:01:54*jhMikeS wonders what happend to C+ (C-plus) and (perhaps wrongly) recalls such a thing existed
02:01:59*JdGordon not so fond of the new order in the info screen
02:02:44 Nick fxb is now known as fxb__ (
02:02:46kugelCan you now tell me how to make it not crash?
02:02:57lostlogicok, so now that it's definitely not caused by the misuse of ata_disk_is_active, what the heck _is_ the problem with playback on sansa pausing at buffer-refill time?
02:03:44jhMikeSlostlogic: some pernicious thing I added to the ata driver in preparation for MoB :)
02:03:48amiconnRead speed? Differences in yielding behaviour?
02:04:27 Join colin__ [0] (
02:04:35jhMikeSthat ata is definitely different for yielding behavior
02:05:28lostlogicjhMikeS: :-P
02:05:53 Quit Seed (Nick collision from services.)
02:06:00 Join Seed [0] (
02:06:58lostlogicamiconn: read speed is where I've been looking but I doon't see why all flash targets wouldn't then be hit
02:08:03amiconn'all' are just 3 for MoB (swcodec): the Sansas and the Nano
02:08:27amiconnWell, and the iFP, but I don't know whether MoB was tested at all on it so far
02:09:00 Join tdtooke [0] (
02:09:32lostlogicfor some reason I was thinking we had more flash targets.
02:09:39lostlogicok, so flash targets broken by MoB.
02:09:58lostlogicwho has a non?
02:10:14tdtookeforgive the intrusion... but will those last 2 amiconn commits effect battery life? Looked like it was something about IO stuff and HDD control, I'm completely ignorant of hardwareish things so I gotta ask
02:10:18amiconnAnd the Nano uses the standard ata driver (with some hackaround for the standby issue)
02:10:44tdtookeaffect even
02:10:54*JdGordon can try and find his sisters nano...
02:10:59lostlogicdo you by chance know what the behavioral differences between the two ata drivers would be?
02:13:19 Quit colin_ (Read error: 110 (Connection timed out))
02:13:25 Quit colin__ ()
02:13:55 Quit lespea (Remote closed the connection)
02:14:17jhMikeSamiconn: nice find on the atomic GPIO stuff btw. it's good for all GPIO related regs?
02:14:44amiconnjhMikeS: That stuff was already known, just only used in a very few places
02:15:05amiconnIt works for all standard GPIO ports (A thru L) on PP502x
02:15:14jhMikeSwhere was it? I never came across it that I recall.
02:15:33amiconnG5 backlight, some other places
02:16:04amiconnThe 0x800 offset might even be unnecessary, but that's what the neneric gpio bit function does on the ipods
02:16:51*tdtooke will endeavor to say really brainey things worthy of response next time
02:17:10 Quit tdtooke ("thanks for you time guys.....seriously")
02:17:38*amiconn greps
02:17:47jhMikeSI don't know of any way at the usual addresses to enable or disable the interrupts atomically.
02:18:54jhMikeSor is the majic actually the use of val << 8 ?
02:18:56amiconnI don't know whether this method also works for the GPIO_INT_* registers, but in fact it might
02:19:39lostlogicI think I see th e bug and I hate it still
02:19:46lostlogictoo bad bertrik left for the night already
02:19:52amiconnIt's not just (val << 8). That's the mask which defines the bits to touch. The actual values are still in bits 0..7
02:20:00 Quit lee-qid (Read error: 110 (Connection timed out))
02:20:07lostlogicjhMikeS: take a look at yield_codec() in buffering.c −− I see a perfectly possible infinite loop in there, do you?
02:20:14jhMikeSamiconn: right
02:20:28amiconnSo with ((vall << 8) | val) you set all bits in the mask, and with just (val << 8) you reset them
02:21:08amiconnYou can even combine that, e.g. 0x4404 would set bit 2, reset bit 6, and keep all others
02:22:14amiconnVarious ipod backlight setting functions already use that (using the old nasty outl() with literal addresses)
02:22:43jhMikeSlostlogic: hmmm...
02:24:02lostlogicpcmbuf goes low because it can't read data from the filebuf
02:24:18lostlogicfilebuf is not critical because it has _allocated_ data on it
02:24:25lostlogicand so it stays in their... forever
02:24:51jhMikeSuntil a message is queued. would that prevent it?
02:25:23lostlogicwho's going to queue a message? but yes, a message being queued would break the loop
02:25:58 Quit w1ll14m ("CGI:IRC (EOF)")
02:26:09jhMikeSsome playback thread would eventually I suppose
02:26:26lostlogicnope, playback is just sitting there waiting for commands
02:26:56lostlogicthat didn't _used to_ infinite loop (premob) because there wasn't a concept of allocated but not written data so much
02:27:06 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
02:28:19jhMikeSso MoB reserves space in advance but doesn't buffer, correct?
02:28:48lostlogichmm but then why didn't 15440 fix it? THat should make the audio thread active and kickstart it out of the loop.
02:30:13 Quit Robin0800 (" HydraIRC -> <- Now with extra fish!")
02:31:10lostlogiccombination slow cpu and fast disk on sansa is what makes this appear there.
02:31:47jhMikeSmore race conditions eh?
02:32:04JdGordonand you thought slow disk!!
02:32:37jhMikeSwhy even bother with yield codec anyway?
02:33:01lostlogicjhMikeS: to prevent pcmbuf underruns during filling is the theory behind the function
02:33:21jhMikeSis there any actuality behind it? :)
02:33:40lostlogicthere was pre-MoB
02:33:42jhMikeSnow that priority threading exists
02:33:58*lostlogic doesn't run priority scheduling
02:34:00*lostlogic doesn't like it
02:34:28*jhMikeS doesn't share that sentiment
02:35:26jhMikeSIt is there, so I guess take advantage
02:35:57lostlogicblech, O(n) scheduling
02:39:49jhMikeSn = ?
02:40:55lostlogicok, I'm inclined to make yield_codec not a loop bu tI don't know how it will impact slow codecs on slow targets
02:41:00jhMikeSthreads ready to run actually which isn't usually more than one.
02:42:21jhMikeSusually that loop won't even execute since the current thread will probably be the one that matches the level
02:42:40lostlogichmm, maybe I'm too hard on it
02:42:58lostlogic*shrug* I don't have a problem with the flat scheduling ;)
02:43:17jhMikeSthe only thread in the running list are the ones actually running. others take no cpu (blocked) or barely any (timeout)
02:44:22lostlogic <−− what do you think of this instead of the yield loop?
02:44:35lostlogicwhat's the target most likely to skip during buffering of a 'working' codec so we can test it?
02:45:00jhMikeSthe timeout operation is extremely lazy and it only executes the checks when a thread is actually due to wake. heck they aren't even taken off the tmo list when explicitly woken but are garbage collected later when the next check is due
02:45:42lostlogicso you're saying that priority scheduling is a good thing and I'm being stupid not to run it?
02:46:53jhMikeSit's a good thing when used carefully. I think just blasting the codec thread to realtime suddedly should be more gradual
02:47:28 Quit Mouser_X ("CGI:IRC (Ping timeout)")
02:47:29jhMikeSA non-looping solution is better IMO too...sleep(2) is certainly a really long time
02:47:44lostlogicshrug, just basing it on the existing loop
02:47:56lostlogicall codecs tha tI have handy are so fast on ipod now that it doesn't even need more than the yield
02:48:33 Quit obo ("bye")
02:49:16jhMikeSthe codec thread should always be ready to run unless it's sleeping waiting for pcm space. it will get executed on the next buffer yield.
02:50:35 Join emeraldd [0] (
02:50:40lostlogicright, but is that single yield enough to recover from a pcmbuf_low situation on a slow codec? no.
02:50:55lostlogicthat's what led to the creation of the yield_codec function in the first place
02:51:09lostlogicI think a sleep(2) to let it gain some breathing room if it's low is reasonable
02:51:21lostlogic2/100 of a second is long in pcmbuf time but short in buffering time
02:51:24jhMikeSwe also don't run every thread on every task switch either
02:52:24jhMikeSthe old threading woke evey thread every time no matter what it was doing. now you'll only have buffering and codec going most times.
02:52:56*lostlogic still running 'old threading' ;)
02:52:59jhMikeSheck test_codec could only work with genuine blocking
02:53:14lostlogicor... am I not?
02:53:23lostlogicugh, I'm so out of date still
02:54:01jhMikeSonly ready to run stuff gets any CPU cycles
02:54:28emeralddI take it from the gest of this conversation that there is still a buffering issue in the current build?
02:54:33lostlogicso, you think I should commit this and see what all the targets have to say about it?
02:54:36lostlogicemeraldd: only on sansa
02:54:52 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
02:54:54lostlogicemeraldd: have you experienced it? If so I'd love for you to test a patch
02:55:09emeralddbeen getting really annoying
02:55:13emeralddwhere can I get the patch
02:55:20lostlogicemeraldd: can you build or do you need a build?
02:55:30emeralddI can build quite readily
02:55:37lostlogicget the altest svn and add that patch
02:55:46lostlogicand I think you'll get no more buffering dath
02:55:50jhMikeSlostlogic: sure. I think it will yield alot of time away from buffering. Don't forget the ATA yields quite often on any targets I've the point recording makes no explicit yields in flushing
02:56:27lostlogicjhMikeS: it only does an explicit yield ever 32k, which it would do with the loop any way
02:56:48emeralddlostlogic: buidling now
02:56:51lostlogiclet's see what a sansa does with it before I commit :)
02:56:53lostlogicthanks emeraldd
02:57:16jhMikeSlostlogic: I think that should be more than enough down time for buffering
02:57:56lostlogicjhMikeS: *nod* well if pcmbuf goes low it's going to sleep for now, maybe that can change further in a future edition, for now I'm basically keeping the semantics of yield_codec without the death loop
02:58:16 Join webguest77 [0] (i=4ca81f01@gateway/web/cgi-irc/
02:58:18jhMikeSwhat about making sure the buffer is also sufficiently full?
02:59:25lostlogicjhMikeS: we don't have an effective way to do that currently without O(num_handles) to find out how much data is on buffer
02:59:27emeralddlostlogic: I get this warning while building:
02:59:32jhMikeSthere is a point where both pcm buffer and the buffer are nearly empty
02:59:36emeralddunused variable breakable
02:59:48emeralddbuffering.c:554: warning: unused variable 'breakable'
02:59:54jhMikeSblah, ignore that warning
02:59:58emeralddI am
03:00:00lostlogicemeraldd: go ahead and run the test any way, I'll fix before commit :)
03:00:14lostlogicwebguest77: don't ask to ask, just ask
03:00:24*jhMikeS votes for more yellow
03:00:57emeralddstupid database refresh
03:01:05jhMikeSto ask about asking is metaassking
03:01:38lostlogicthere are metagies available in not metaasking
03:01:40webguest77hello. i have a 5.5 g ipod. i downloaded rockbox and installed it. but then i switched the .rockbox reg folder to the .rockbox in this folder EvilG-Fusion-60-80G-20070717. it loades and everthing but it doesent work. like if i press my center button nothing happens. it doesent open anything.. what did i do wrong?
03:03:02lostlogicwebguest77: so you are trying to use a build named EvilG-Fusion?
03:03:17lostlogicif so then we do not support that here, we only support rockbox official builds, downloaded via
03:03:37emeralddlostlogic: I'm seeing a solid read icon in the upper right hand corner, That's new
03:03:54webguest77oh ok. but i downloaded it from a link from this website...
03:04:00emeralddwell, nearly solid
03:04:06lostlogicemeraldd: check out the buffering thread screen and see waht's going on
03:04:24lostlogicwebguest77: where'd you find the link?
03:04:25 Join RaRe` [0] (
03:05:16emeralddlostlogic: what are you looking for specifically, I'm seeind 70-80% boost
03:05:29emeralddpcmbufdesc 10-15/21
03:05:33 Join colin_ [0] (
03:05:56lostlogicemeraldd: I'm curious what the usefl slider is doing
03:06:02lostlogicand whether the disk is still staying active
03:06:22emeralddthat's the bar at the top of the buffering screen?
03:06:31*jhMikeS wants a 5-thread mpegplayer but MAXTHREADS is also expanding. Though there's no important thread code that relies on thread entries to be in an array and plugins could allocate their own.
03:07:06lostlogicemeraldd: usefl is the 4th bar down my screen but I'm not sure how it looks on sansa
03:07:36webguest77from here : i went down until i found a theme called flow. i read this :This is my first WPS for the iPod. Inspired by the iPhone and its cover flow feature, I came up with this. Enjoy! Tested with EvilG's Fusion build, but should work on any build with AlbumArt, Bitmap Resize, Customline, Multifont & Multifont-userfonts patches. i clicked the link to the is not
03:08:01emeralddlet's see here
03:08:19lostlogicwebguest77: sorry but we can't support that build here, it's heavily patched to support the album art and stuff
03:08:37webguest77oh ok thank you
03:08:39emeralddit seems to be counting down
03:08:41jhMikeSlostlogic: I don't get what's different about some people's players. I had been running e200 and gigabeat and never had a hangup since the first problem was fixed.
03:09:03emeralddSlowly but surely dropping
03:09:06 Quit zicho (Remote closed the connection)
03:09:23emeralddWell, looks like I hit the end of the song
03:09:34lostlogicjhMikeS: yeah, probably depends on bitrates and stuff too...
03:09:36webguest77are there any themes i can downloaded from the real website?
03:09:49lostlogicemeraldd: and how are things going?
03:09:52jhMikeSlostlogic: always high bitrates here
03:10:00lostlogicwebguest77: sure, visit our wiki
03:10:12emeralddMade it through 5:50seconds at 128k with no problems
03:10:23webguest77oh ok ^_^ thank you
03:10:30lostlogicemeraldd: I assume you were normally seeing freezes after a full buffer ran out?
03:10:42emeralddAll through the songs
03:11:10emeralddIt would maek it about half way through and start reading the disk at 1to 2 second intervals and then pause for a
03:11:12lostlogicwebguest77: look for ones that say "build standard"
03:11:19emeralddfull 2 to 3 seconds,
03:11:24emeralddnothing like that so far
03:11:27lostlogicemeraldd: ah, good!
03:11:52emeralddI am bothered that I haven't seen a disk read on the second song though
03:12:08lostlogicemeraldd: you shouldn't, it's buffered
03:12:16lostlogicemeraldd: you shouldn't see more disk reading until about 20 minutes from now
03:12:36emeralddThat's more like it ;)
03:12:39lostlogicbut make sure that it keeps playing for at least 20 minutes before we say "it works"
03:13:00emeralddhow much does data rate and track count matter
03:13:07emeralddwould a 1:30 minute track work?
03:13:11emeralddat 64k
03:13:35lostlogicwhat do you mean?
03:13:50lostlogicwhat I'm looking for is a full playlist that fills the buffer completley to see what happens when the buffer then runs low
03:14:29***Saving seen data "./dancer.seen"
03:14:37webguest77um after i click on wiki where do i go to get the themes?
03:15:02Mouser_Xwebguest77: lostlogic gave you a link to follow.
03:15:05lostlogicwebguest77: see the link I posted 4 minutes ago fo rthe ipod 5g themes and then look for ones that say "build: standard"
03:15:46webguest77oh ok, but i have a 5.5 g. is that ok?
03:15:58lostlogicwebguest77: yeah, the screen is the same so the wps are the same
03:16:15lostlogicgood lookin' out though.
03:16:26webguest77oh ok thank you very much
03:16:54lostlogicwebguest77: you're welcome, hopefully we'll have album art in the official build sometime so you can get some of the fancier themes ;)
03:18:12lostlogic(or you'll get more knowledgeable and be able to deal with the bugs that crop up in the unofficial builds better, either way :))
03:18:43*Mouser_X probably wouldn't use album art, even if it was supported in official builds.
03:18:52*lostlogic definitely wouldn't
03:18:58lostlogicbut lots of people would apparently
03:19:08*karashata definitely wouldn't, screen's not big enough on the H10's
03:19:10Mouser_XI might by accident though, if the folder I copied over has stuff in it.
03:19:15*jhMikeS is one of those that like that pretty stuff
03:19:53emeralddso, usefl identifies what about buffering?
03:20:17karashataMouser_X: unless you were using a theme that supported it, you wouldn't
03:20:19lostlogicemeraldd: usefl is how much data is left that is _likely_ playable
03:21:02jhMikeSMoB = Maybe on Buffer?
03:21:11lostlogicjhMikeS: you're mean :( :)
03:21:15Mouser_XThe theme I'm using now is *supposed* to support it, but I removed it.
03:21:39jhMikeSam not...just a tottaly sacastic bastard. I'm as bad about myself.
03:21:43 Quit RaRe (Connection timed out)
03:21:56emeralddwhat unit is usefl in I'm seeing six figures here
03:22:12lostlogicso that's kilobytes, and it's getting low-ish
03:22:18emeralddsorry, five figures
03:22:22lostlogicwhen it gets just below 6 figures expect it to fill
03:22:37lostlogicif it doesn't fill then that would be considered a bad thing
03:23:02lostlogic(what bitrate? 64?
03:23:06emeraldd(Note: it started at five figures, I mis counted at first)
03:23:11emeralddbit rate is 128 right now
03:23:17lostlogicah, has it started filling yet?
03:23:20jhMikeSBad Thing (TM) or just generic bad thing?
03:23:23lostlogicie counting up instead?
03:23:31emeralddnot yet
03:23:37lostlogic::blink blink::
03:23:43emeralddIt's about 1/3 what it started at
03:23:59*lostlogic gets ready to cheer/cry
03:24:05emeralddIf the water marks are still at 25% and 75% it should be just a little longer
03:24:08 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:24:18lostlogicthe watermark should be at 5 seconds of audio left
03:24:29 Join miepchen^schlaf [0] (
03:24:38emeralddI remember comments talking about 25/75
03:24:41lostlogicso at 128kbps that would be 32kB
03:24:46lostlogicer 48kB
03:24:48emeralddgot it
03:24:55*jhMikeS predicts it will drop to 25%, refill up to 26% :p. (more sarcasm)
03:25:10lostlogicthe 75% was the high_watermark
03:25:30lostlogicwhich only applies when the disk is spinning (I guess should apply always on flash targets) and which I've disabled completely for the time being
03:25:31jhMikeSmaybe it's dyslexic and will go to 57%
03:25:47*lostlogic hunts for his blow-dart gun
03:26:58 Quit Mouser_X ("CGI:IRC (Ping timeout)")
03:27:19*jhMikeS could use the relaxation time
03:28:45lostlogicemeraldd: ?
03:29:05emeralddnot refilling yet :(
03:29:10lostlogicemeraldd: what's it at?
03:29:36lostlogicseems to be counting awful slow, wondering if the digits are rounded on sansa
03:29:46lostlogicwait, that's still 5.5 megabytes
03:29:50lostlogicwe've still got 5 minutes till it fills
03:30:06emeralddso the unit is not bytes
03:30:11lostlogicyes it is
03:30:19emeralddwait my brain is fried
03:30:19lostlogic5,500,000 <−− 5 million bytes
03:30:30lostlogicwhen it gets down closer to 100 thousand bytes, it'll fill
03:30:32lostlogic(i hope)
03:32:00emeralddcrap I think I just told it to skip a song
03:32:09lostlogichaha, I do that all the time :-P
03:32:10emeralddand it jus timed out
03:32:16lostlogictimed out?
03:32:21emeralddI hear silence
03:32:34lostlogicanything going on on the screen?
03:32:34emeralddat 1:02 of 7:11
03:32:59emeralddstill responsive though
03:33:03lostlogicI'm so confused, so you exited the buffering screen, then skipped and the skip didn't work?
03:33:08*jhMikeS thinks that queue hack in playback should be fixed first
03:33:16emeralddI didn't exit the buffering screen
03:33:25lostlogicthen how do you know the time?
03:33:37jhMikeS a clock?
03:33:40lostlogicjhMikeS: queue hack? oh the queue_peak thing?
03:33:43emeralddi hit the down button to keep the back light on
03:33:46emeralddit skipped a track
03:33:50emeralddeverything stopped
03:33:53jhMikeSlostlogic: yeah, not safe at all
03:33:58emeralddI then left the screen
03:34:04lostlogicemeraldd: ah.
03:34:06emeralddchecked the time index on now playing
03:34:19emeraldd1:02 of 7:12
03:34:21lostlogicw.t.f. sansa.
03:34:29emeralddmy thoughts exactly
03:34:42lostlogicI mean I'm sure ther is a bug in the code but ... I don't get it.
03:35:14emeralddI think there might be a bug in th buffering screen :(
03:35:24emeralddIs there a nice way to artifically shrink the buffer?
03:35:34lostlogicthe buffering screen is designed to allow skips for basically that purpose
03:35:49emeralddthen that's what I just caused :(
03:35:59lostlogicso what I do when testing this is to watch the screen, skip forward (which is a clockwise touch on the ipod) until the last track and then let that run out
03:36:21emeralddI wish I had known that twenty minutes ago ;)
03:36:49lostlogicsigh and it still works dandy fine on the ipod
03:36:54emeralddno problem
03:37:02emeralddso what should it do at five seconds?
03:37:20lostlogicalloc should start dropping rapidly, as should real
03:37:28lostlogicand usefl should start counting up at a medium pace
03:38:00 Join eigma [0] (n=cat@
03:38:06eigmakkurbjun: I think i finally broke my m:robe
03:38:14emeralddI dont' see anything labled that
03:38:36lostlogicemeraldd: might be dropped from your screen
03:38:49lostlogicI know some things had to drop because sansa real estate is limited
03:39:24emeralddthat time it worked
03:39:39emeralddjust fine in fact
03:39:50emeralddI had to reboot the player though to get it play again
03:40:42lostlogicemeraldd: let me knwo if you figure out what broked it
03:41:03emeralddworked that time as well
03:41:29 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:41:44lostlogicsounds like it's better, where the hell else could it get stuck
03:41:48*jhMikeS just might do a speedy state-machinish rework of the audio queue_peek.
03:42:00emeraldddeadlock between playback and buffering thread?
03:42:15lostlogicemeraldd: no potential deadlocks present
03:42:20lostlogicit's a livelock if anything
03:42:36 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
03:42:38lostlogic(and yes I know I"m mincing words)
03:43:14lostlogiclive lock means that the threads are actively looping on something that's not changing as opposed to a deadlock where they are in a mutex_wait or other wait operation and not executing any code
03:43:48jhMikeSstuff like if (!playing || playlist_end || ci.stop_codec) has to go too...hmmm....
03:44:12lostlogicjhMikeS: you have fun beating on playback.c, I'll confine my toys to buffering.c for now :)
03:44:54emeralddjust locked again
03:45:03jhMikeSI'm worried buffering requests could build up in the queue
03:45:06lostlogicon a skip or a run down?
03:45:19emeralddskip to last buffered track
03:45:21emeralddand run down
03:45:27emeralddran out at 43 seconds
03:45:32emeralddof 3:25
03:45:36lostlogicjhMikeS: it bounces out to service them −− that's why it uses the peak... if it's _not_ a buffering message, send a buffering message
03:45:40emeralddtrack 38 out of 51
03:45:46lostlogicemeraldd: when it stops playing what is usefl at?
03:45:56 Quit Mouser_X ("CGI:IRC 0.5.9 (2006/06/06)")
03:46:08lostlogicjhMikeS: one thing there is not is a risk of buildup
03:46:09jhMikeSbut it could do buffering, not buffering, buffering
03:46:12emeralddright now 16398
03:46:20emeralddlast time it was 14xxx something
03:46:28lostlogicalways seems to be about 16k left when it fails
03:46:32lostlogicwtf is special about 16k?
03:46:36emeralddIf I remember one file chunk is 32k
03:46:40jhMikeSthen it won't remove it if it is in fact a buffering message while in the loop
03:46:47emeralddso something less than one file chunk maybe?
03:47:19lostlogicemeraldd: something like that but for it to stop it also must be less than the amount the codec is requesting... I think
03:47:30 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
03:47:32lostlogicjhMikeS: it breaks out and lets the main thread loop run to remove it
03:47:54jhMikeSah, woops, reading the break in the wrong spot :P
03:49:24 Quit colin_ (Read error: 110 (Connection timed out))
03:49:46 Quit kugel ("ChatZilla [Firefox]")
03:50:05jhMikeSdoes that really count as a postbuffer event?
03:50:22lostlogicjhMikeS: I wondered the same when fiddling in there yesterday
03:50:47emeralddWhere can I find the meaning of the thread states on thread screen?
03:50:57 Join DarkDaemon [0] (
03:51:15DarkDaemonjust wondering, if i get flac songs
03:51:18DarkDaemonwhat can i expect?
03:51:21DarkDaemonbetter sound?
03:51:41jhMikeSemeraldd: B=blocked, T=Blocke W Timeout, S=Sleeping, R = Running (there's a couple others you won't see)
03:51:47Mouser_XDarkDaemon: Enter does not = spacebar.
03:52:00DogBoydepends DarkDaemon
03:52:11lostlogicDarkDaemon: in theeory yes... especially on transients and other hard to encode sections of audio
03:52:33DarkDaemondepends on what DogBoy?
03:52:35lostlogicDarkDaemon: a flac will sound exactly like a cd, an mp3 or ogg _may_ sound different, losing detail or range
03:52:52DogBoyon the gear you're using and your ears etc
03:52:54emeralddbuffering is T, codec is S, audio is T
03:53:17lostlogicemeraldd: are they changing?
03:53:18DogBoythe source audio
03:53:25DogBoydepends on a lot of things
03:53:34DarkDaemonlostlogic so is it worth the download? since they size is fairly big. im not worried abt space cause i have a 80gb ipod but more if i should wait for the download :P
03:54:02lostlogicDarkDaemon: to _me_ it would be, but I have a 3k stereo system and $300 ear phones to go with my digital music...
03:54:11Mouser_XI'd be probably be more concerned with battery life.
03:54:22DogBoyyea and for me as well, I would prefer the flac files
03:54:28Mouser_XFLACs will decrease your battery life because there will be more disk reads.
03:54:38lostlogicand what Mouser_X said
03:54:50DarkDaemonwell i have around 80$ dj headphones (the big ones) and ipod 5.5g 80gb
03:54:50DogBoygood point
03:55:02lostlogicbut you can keep the flac on your computer for hi-fi use and transcode to anything you want with no additional loss of quality that you would get from transcode betwen lossy formats
03:55:07lostlogic(that is what I do)
03:55:15lostlogic(I put ogg q7 on ipod and flac on pc)
03:55:28Mouser_XI'd go with lostlogic on that one.
03:55:30DogBoythat's what I do too but I use musepack 7
03:55:34DarkDaemonMouser_X, thats not a big problem for me. i dont listen to music during school and sutff but when i do listen it, i like higher quality
03:55:47Mouser_XWell then, go for it.
03:55:59Mouser_XOther than bandwidth and time, what can it hurt?
03:56:08DarkDaemongood point
03:56:18DarkDaemonok thanks alot for the help
03:56:22DarkDaemonhave a good day
03:56:34 Quit DarkDaemon ("If You Can't Be Famous, Be Infamous...")
03:56:46lostlogicstrange feller... who doesn't just leave IRC running?
03:56:58*karashata doesn't
03:57:13Mouser_XI can't. I'm not on any of my stuff (I'm using someone else's PC).
03:57:21DogBoywe know karashata, we're watching your every move
03:57:37Mouser_XAlso, using CGI:IRC can cause problems if the back buffer gets too big.
03:57:42emeralddI'm running off a laptop, can't leave it in one place too long
03:57:44karashataDogBoy: like I'm worried, it's not like I do anything illegal...
03:57:51lostlogicmy irc client runs on a leased server in a data center... which I just log into when I want to pay attention it from anywhere
03:58:10emeralddhow much does that cost you per month>
03:58:12Mouser_XHeh. Nice.
03:58:15lostlogictoo much.
03:58:23DogBoyit's generally a waste to use flac files on a portable though
03:58:30lostlogicDogBoy: yep
03:58:46karashataI dunno, I kinda like the FLAC files I have
03:58:49DogBoyyou're not gonna hear the difference most of the time
03:58:57lostlogicSoap: bite me
03:59:01Mouser_XI had some stuff in FLAC on my Gigabeat, but they took up too much space.
03:59:25karashataI don't need the space yet, so I'm leaving mine there
03:59:29lostlogicSoap: if there was signal for the noise to disrupt I'd agree, but there isn't
03:59:31 Join toffe82_ [0] (
03:59:41emeralddlostlogic: Interesting, I just saw it start to fill, and lock on the second block
03:59:47DogBoyit eats up people's log files though
03:59:52emeralddfollowed by the pcm buffer emptying
03:59:53lostlogicemeraldd: what do you mean second block?
03:59:59DogBoymakes them harder to parse
04:00:05emeralddsecond update
04:00:14emeralddor so,
04:00:14scorchelostlogic: the noise is to those who read logs
04:00:30emeralddbut the buffering deffinately stopped before the pcm thread died
04:00:30lostlogicfire me.
04:00:54lostlogicemeraldd: was usefl still down to below a file chunk when it stopped?
04:01:16emeralddit's at 17679
04:01:21emeralddI think that is about half a file chunk
04:01:26emeralddlittle over
04:01:49lostlogicyeah ok
04:02:02lostlogicwhat codec? mp3 128k?
04:02:24emeralddnote it started filling before it locked
04:02:40lostlogicyeah, I'm trying to figure out what new information that provides.
04:04:02jhMikeSlostlogic: how bout dumping that call to postbuffer when only responding to a message? it looks like it will just rifle through them real fast otherwise before buffering as actually done.
04:04:27lostlogicjhMikeS: makes sense to me, I considered doing it, but ... I've been changing a lot lately :)
04:04:36webguest77excuse me, how can i get to the ipodlinux irc?
04:04:54DogBoywith a real irc client
04:04:55scorchewebguest77: you cant using the rockbox web client
04:04:55lostlogicwebguest77: you'd need to use a different irc client... check out (I think)
04:04:56 Quit DerPapst (Nick collision from services.)
04:05:06scorchelostlogic: you pay attention :
04:05:15lostlogicscorche: no, I'm an illusion
04:05:23*scorche sighs
04:07:13lostlogicjhMikeS: at least most of my deltas are green, that makes me happy −− I alway say that I try to gauge my success by negative lines of code
04:07:42lostlogicemeraldd: you used logf at all? know hwo to turn it on?
04:08:35emeralddnever used it
04:08:41emeralddbut willing to give it a shot
04:09:01 Quit webguest77 ("CGI:IRC (EOF)")
04:09:10lostlogicemeraldd: ok, great −− reconfigure with your target, advanced, logf
04:09:27lostlogicthen modify playback.c and buffering.c, in each one uncomment the LOGF_ENABLE near the top
04:09:51lostlogicthen in both files, copy the LOGF_QUEUES_FILENAME define up outside of the #ifdef SIMULATOR block
04:10:00lostlogic(if any of those instructions are not clear or specific enough, let me know)
04:10:40emeralddone moment
04:13:19 Quit karashata ("Leaving.")
04:13:25emeralddI'm not seeing the LOGF_QUEUES_FILENAME define in playback.c
04:13:42 Join Llorean [0] (
04:14:00 Join karashata [0] (
04:14:10 Quit toffe82 (Read error: 110 (Connection timed out))
04:14:20lostlogicemeraldd: lemme check
04:14:49lostlogicBUFFERING_LOGQUEUES in buffering.c
04:17:02emeralddyou had already done that in bufferinf with your patch
04:17:14emeralddIs there anything I need to do before rebuilding?
04:19:24 Quit eigma ()
04:20:29emeralddthis might take a few on my pIII 800
04:20:58Mouser_Xemeraldd: It takes me an hour+ to build on my craptop.
04:21:54emeralddmouser_X: I've forgotten what those days were like, building a linux kernel on a 486 with 16/ 32 meg
04:34:19 Quit japc (Read error: 110 (Connection timed out))
04:34:25emeralddlostlogic: I've got a build, transfering it now
04:36:03emeralddIs there a way to kill the database refresh in the OF on sansa c200 ?
04:38:35 Join RoC_MasterMind [0] (
04:38:58emeralddlostlogic: dead on the first try,
04:39:11emeralddlostlogic: now how do i get the needed data out of this thing
04:40:04 Join jmspeex [0] (
04:42:38emeralddlostlogic: never mind that issue, I have the logf file, where would you like me to send it?
04:44:48lostlogicemeraldd: can post to pastebin for me?
04:44:58emeralddlet's see
04:46:49emeralddIf I did that right
04:49:09 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
04:49:50lostlogicemeraldd: looks good
04:50:33emeralddany clues?
04:50:53lostlogiceating, might be slow to figure that part out
04:51:05emeralddthat's where you went
04:51:16hcslostlogic: greetings, I've determined that the old mp3 freeze does in fact happen at the same time, but it looks like you have new commits
04:51:17Mouser_XHowdy HCS.
04:51:17emeralddIt's 10:51 my time, well past dinner ;)
04:51:50lostlogichcs: yeah, seems to still happen, possibly less with the new stuff
04:53:03 Quit spiorf (Connection timed out)
04:53:03lostlogicemeraldd: whereabouts in central time are you?
04:53:17emeralddnorth east texas ;)
04:53:55lostlogicgotcha, I used to live on the opposite side of central time in Chicago
04:54:45hcslostlogic: ok, I can get to it pretty fast now, I'll see if the one I experienced is fixed
04:54:58lostlogichcs: cool, thanks
04:55:38hcsand I'm eager to try out the ipod disk powerdown
04:55:47emeralddI've never made it there, but I've told many a machine that it lived there to get the right time zone ;)
04:55:52lostlogicok, so why does it hop back out after just one track of buffering?
04:56:02lostlogicemeraldd: yeah, I tell machins I live in LA ;)
04:56:30lostlogicI bought a car from a guy in Dallas once... and drove it back to Chicago :)
04:57:10lostlogicI feel like I'm chasing a goddamned gremlin with this bug
04:57:19emeralddthat's one heck of a drive
04:57:22lostlogicit's not the same problem as it used to be, it just moved
04:57:32lostlogicemeraldd: I did it straight other than a short roadside nap ;)
04:57:50lostlogicemeraldd: then again I also rode my motorcycle from Chicago to Seattle to move here... in 3 days
04:58:43emeralddI stick to enclosed travel
04:58:53emeralddso my wife doesn't decide to end me ;)
04:59:05 Join BigMac [0] (
04:59:29Mouser_XFonts in a theme (or WPS for that matter) can only be 1 color throughout, correct?
04:59:39emeralddis that a register dump at line 300-307?
04:59:40 Quit BigMac (Client Quit)
04:59:57 Join BigMac [0] (
05:00:14Mouser_X(Line 1 of text = red; Line 2 of text = blue; Line 3 of text = yellow etc; is not possible, correct?)
05:00:21lostlogicemeraldd: that's just it telling us the addresses where the various buffers are allocated
05:00:57karashataMouser_X: unless you have the font colouring patch for the WPS, no
05:01:18Mouser_XThat's what I thought. Thanks.
05:01:27emeralddlostlogic: why there and not after line 102 as well?
05:01:28lostlogicemeraldd: malloc buf and mallocbuf end, filebuf and filebuf end, guard buf and buardbufend and pcmbuf and pcmbuf end
05:01:54lostlogicemeraldd: only happens when the buffers are first used or they cahnge sizes
05:05:25 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:05:38 Join miepchen^schlaf [0] (
05:07:18lostlogicemeraldd: do dyou happen to remember what the value of usefl was when this logf stopped playing?
05:07:33emeraldd17xxx something
05:07:46lostlogicemeraldd: ok, thanks
05:07:55emeralddIt's ranged between 14xxx and 17xxx or so everytime
05:08:28lostlogicso the good news is that the audio thread stays alive now and manages to supply the buffering thread with all of the information it needs to fill the buffer
05:08:40lostlogicthe bad news is that the buffering thread just sits there on its ass and doesn't do anything with it.
05:09:06lostlogicemeraldd: can you add (yet another) define to buffering.c for me and do it again?
05:09:19lostlogic#define BUFFERING_LOGQUEUES_SYS_TIMEOUT
05:09:21emeralddsure :)
05:10:51lostlogichcs: how's yours behaving?
05:11:16emeralddI'm assuming outside the ifdef
05:11:27hcslostlogic: still building, ran into some patch trouble
05:11:36lostlogichcs: ah, ok
05:11:42lostlogicemeraldd: yeah, right next to the other one
05:11:46lostlogicthat you moved out
05:12:25lostlogicchasing gremlins *mutter*
05:13:10lostlogicemeraldd: actually before you build, add another line to that file
05:13:12lostlogiclemme find where.
05:13:22Mouser_Xlostlogic: I think it's a collection of gremlins, working together, to devise the perfect means to making your life miserable.
05:14:02Mouser_XSlowly, you've been able to kill some of them, at least.
05:14:13hcsbut they breed like rabbits
05:14:27Mouser_XHmmm. Good point.
05:14:33***Saving seen data "./dancer.seen"
05:14:35emeralddcould be worse
05:14:39lostlogicemeraldd: in the function buffer_handle near line 616 add: logf("buffer_handle exit: %d", ret);
05:14:49lostlogicthey're breeding!???!
05:14:57lostlogicgah, I'm going to hafta get out my flame thrower
05:15:03lostlogicjust don't get them wet... right?
05:15:08*karashata laughs a little
05:15:25*Mouser_X would never dare get his Gigabeat wet...
05:15:26lostlogicsomeone already fed them after midnight.
05:15:39 Quit bb (Nick collision from services.)
05:15:44 Join bb_ [0] (n=bb@unaffiliated/bb)
05:15:49lostlogicemeraldd: that logf should go right abouve return ret
05:15:59lostlogic(in case that wasn't somewhat obvious)
05:16:52emeralddlostlogic: got it,
05:17:13emeralddI am somewhat familiar with c, but I really hate large numbers of ifdef lines
05:17:26lostlogicemeraldd: gotcha, I'll keep that in mind as we debug :)
05:17:41lostlogicor degremlin
05:18:11emeralddIf you want to see something scary, try to trace through the code in the r128 driver in Mesa
05:18:24emeralddIt's #define macro upon macro upon macroe
05:18:39lostlogic*shudder* I stick to linux kernel and rockbox C personally.
05:19:05 Join webguest57 [0] (i=4ca81f01@gateway/web/cgi-irc/
05:19:28webguest57excuse me, but every gameboy game is supposed to run in fullscreen right?
05:19:38emeralddunfortunately, I have a r128 and someone mixed up bits and bytes in the depths of that monstrosity
05:19:49lostlogicemeraldd: ouch!
05:20:13hcslostlogic: still MP3 freezing, here's what I see: on this track, the metadata for the next track hasn't loaded yet, just before the freeze the metadata loads.
05:20:16lostlogicwebguest57: kkurbjun is your man for rockboy stuff, not sure if he's around
05:20:41lostlogichcs: yeah, I think the metadata loading is the result of my most recent commit... but it's still effing broken :)
05:20:46webguest57oh thats what the emu is called? rockboy?
05:20:56lostlogicwebguest57: afaik yeah
05:21:08webguest57can i run diff emu using rockbox?
05:21:39lostlogicwebguest57: rockboy is a port of gnuboy to rockbox, in theory another emu could be ported but it took a _lot_ of work to get this one even
05:21:53emeralddhere we go again
05:21:58lostlogicemeraldd: wheee
05:22:16webguest57wat is gnuboy
05:22:28emeralddI need to make another long playlist, I'm going to get tired of queen
05:22:28webguest57oh nvr mind
05:22:29lostlogicwebguest57: the gnu gameboy emulator or some such
05:22:46lostlogicemeraldd: haha, I always just play on random, hard to get bored of 3600 songs
05:22:51webguest57lol my bad
05:23:18Mouser_XThe max playlist size is too small for the number of songs I have on my Gigabeat.
05:23:28 Quit karashata ("Leaving.")
05:23:28lostlogicMouser_X: you play a lot of sid and nsf and stuff right?
05:23:32webguest57ok thanks
05:23:32lostlogicmore than 30k on the player?
05:23:50emeralddlostlogic: dead again, 17854 this time
05:23:56Mouser_X(I think I have 2x the amount of the max playlist size. Possibly more.)
05:23:57lostlogicemeraldd: yay, logf! :)
05:24:40emeralddNow, if only for the usb stack ;)
05:24:54lostlogicemeraldd: hrm?
05:25:14emeraldddifferent issue ;) (c200 has to use the of for usb
05:25:26lostlogicemeraldd: ah, so does ipod
05:25:47lostlogicemeraldd: will be nice when someday we have a working USB stack in rockbox and can access it on these SOCs
05:25:52lostlogicmight even allow USB hosting
05:26:21Mouser_Xlostlogic: I made a playlist that supposedly has all of my currently present songs in it. It's over 3 MB in size.
05:26:33lostlogicthe playlist is larger than most of the songs, laugh
05:26:38 Join tihoc4n [0] (
05:26:48emeralddlostlogic: update or new post?
05:27:02lostlogicemeraldd: don't care, jsut link me either way :)
05:28:30tihoc4nHi, Just noticed one of the updates on the HomePage, about the powerdown HDD. - How much is the batterylife expected to improve because of this?
05:28:51Mouser_Xtihoc4n: No idea.
05:29:08Mouser_XNo one has done enough battery benches to give any kind of indication.
05:29:09lostlogictihoc4n: should now be better than the apple firmware, according to amiconn
05:29:25lostlogictihoc4n: but like Mouser_X says, we really don't know, why don't you test it :):)
05:29:57lostlogicemeraldd: why? why is it zero?
05:30:19emeralddwhich line?
05:30:30tihoc4nIf I had the time for setting up, I would. But I'll defnatly be giving it a wirl!
05:30:34lostlogicoh, I know why it's zero
05:30:46lostlogicok... now why... why doesn't it start a 'normal' buffer fill right after that?
05:32:32 Quit tihoc4n ("Leaving.")
05:32:39lostlogicemeraldd: add a yield() after call_buffer_low_callbacks() near line 1178
05:32:50lostlogicemeraldd: I'm pretty sure it will then work, but that's not the right solution
05:33:06lostlogicemeraldd: also, you can take out the LOGFQUEUE_SYS_TIMEOUT
05:33:09lostlogicto stop flooding the logf
05:33:43emeralddwould the SYS_TIMEOUTs keep it from doing a fill?
05:34:03 Part toffe82_
05:34:25emeralddyou have a queue_empty call wrapping your fill
05:34:29 Quit BigMac ("Leaving")
05:34:29 Quit RoC_MasterMind (Connection timed out)
05:35:07lostlogicemeraldd: yeah, thats why I wanted the sys_timeouts log
05:35:12lostlogicit told me that the queue was indeed empty
05:35:21lostlogicnot getting flooded by some messages
05:35:34 Join SoulinEther [0] (
05:35:52lostlogicand it tells us that it's entering that lower block in there
05:36:02lostlogicat least it should... if the event is SYS_TIMEOUT
05:36:09lostlogicso why aren't those two conditions right?
05:36:10lostlogiclet's logf them
05:36:26lostlogicactually hmm, do you mind not puttin gin that yield() and instead putting in this logf?
05:36:37emeralddnot a probelm
05:36:39emeralddnot a problem
05:36:45emeralddif I can spell ;)
05:37:24emeralddlogf inside the queue_empty if,
05:37:32emeralddand then in each of the two conditions?
05:37:37lostlogic logf("TMO bufcheck: %d, %d", data_counters.remaining, data_counters.useful);
05:37:51lostlogicyeah, basically
05:37:57lostlogicreally just need it inside the sys_timeout one
05:37:59lostlogicthe filling one works
05:38:18lostlogicI just developed a new theory though
05:38:38lostlogiccould have to do with the updates to data counters
05:38:59 Quit homielowe (Read error: 104 (Connection reset by peer))
05:40:38emeralddhere is what mine looks like right now, just so we are on the same page
05:41:30lostlogicemeraldd: cool −− weird that it's showing the diffs against svn that I committed?
05:41:42lostlogicah, you haven't svn up'd since using the aptch
05:42:56emeralddwould that help?
05:43:27lostlogicno, but probably better to sync up −− I'd revert the yield patch and update to svn if I were you
05:44:00 Quit JdGordon ("Konversation terminated!")
05:45:13emeralddI think I sent the wrong diff
05:45:21emeralddI did pull that yield out
05:46:29lostlogicemeraldd: k
05:49:09emeralddinteresting, my working copy seems to have other differences that what was comitted,
05:49:19 Join homielowe [0] (
05:49:29lostlogicemeraldd: I think I tweaked and commented before committing
05:51:23lostlogicyeah, the svn version is 'right' other than the logfs that we've been adding
05:53:58 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
05:54:47 Join JdGordon [0] (
05:55:01emeralddok, I should be working off the most recent svn version now,
05:55:58emeralddsize_t is an int right?
05:56:16lostlogicit's generally an unsigned long int (32 bit int)
05:56:37lostlogicyou won't find any floats in rockbox
05:57:37emeralddI get a warning about the %d in that last logf
05:57:45emeralddshould it be a %u or something similar?
05:57:45lostlogicoh, shoot, should be %ld
05:58:02lostlogicbut in practice it will work fine either way ;)
05:58:17lostlogic(and I was whining about others using %d for %ld earlier today too) *smacks self*
05:58:36emeralddbeen there and done that ;)
05:58:43emeralddslightly different context
06:03:07lostlogic is the next thing I want to test (depending on the result of the current logf)
06:03:19 Join eigma [0] (n=cat@
06:03:21eigmajhMikeS: ping
06:03:24emeraldddead with 17854
06:03:58lostlogicdamn, you're getting tragically good at breaking it, logf me :)
06:03:58emeralddand I can't spell buffering
06:04:43emeralddbetter to be good at killing it than to not be able to duplicate the probelm except on a whim
06:05:05lostlogicyes, quite.
06:07:41lostlogic24 megs remaining... 17854 on the buffer
06:07:49lostlogicit should be shrinking and filling
06:08:28 Join ashridah [0] (
06:10:18lostlogic logf("conf_watermark=%d",conf_watermark);
06:10:29emeralddis that %d or %ld?
06:10:29lostlogicemeraldd: stick that in Q_SET_WATERMARK block in buffering_thread
06:10:37lostlogiclol not sure, %ld probably
06:10:44lostlogicright before the break;
06:11:59lostlogicbecause if that's < 17k, I'm going to punch something :-P
06:12:16lostlogicit should be 512k
06:15:16 Part ashridah ("Leaving")
06:16:05SoulinEther...should I try repartitioning my nano when i have installed rockbox onto it properly (im not asinine) and the bootloader spits at me an "unable to find rockbox.ipod" error?
06:16:25SoulinEtheror... something? i've exhausted both search means and myself in trying to find a solution.
06:16:45lostlogicSoulinEther: where is rockbox.ipod located on the drive?
06:16:57 Quit JdGordon ("Konversation terminated!")
06:17:04SoulinEtherallow me to check.
06:17:17emeralddI took the misspelled log line out as well
06:18:03 Quit Mouser_X ("CGI:IRC (EOF)")
06:18:07SoulinEtherha, apparently i am asinine
06:18:13 Join alienbiker99 [0] (
06:18:28lostlogicemeraldd: :)
06:18:42lostlogicSoulinEther: laugh, or perhaps just made an error ;)
06:18:43SoulinEtherOk, well, fsck.vfat wasnt too friendly to my installation of rockbox
06:18:55lostlogicSoulinEther: you have fs corruption?
06:19:12emeralddlostlogic: dead at 17854 again
06:19:17SoulinEtherhum... what are the red flags of said issue?
06:19:24 Join Calcipher [0] (
06:19:59 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:20:00jhMikeSeigma: gnip?
06:20:02 Quit Mouser_X (Client Quit)
06:20:35SoulinEtherhm, your specific wording has given me guidance on google, thank you lostlogic
06:21:24CalcipherAnyone recall any kind of feature request for a "speak whats playing now screen" button on rb?
06:21:50 Join webguest51 [0] (i=d2d59c0b@gateway/web/cgi-irc/
06:21:55 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:22:16emeralddlostlogic: guess what the water mark is
06:22:32emeralddfind something soft before reading that
06:22:43CalcipherI have a blind friend who I've just installed rb for, on a Sansa e260R
06:22:47 Join midkay [0] (n=midkay@rockbox/developer/midkay)
06:23:00lostlogicwhat in th eliving name of all that is holy
06:23:01lostlogichow the fuck
06:23:37lostlogicemeraldd: thank you. that sure is odd, I wonder how it gets to be zero, at least it lays plain how it gets broken
06:23:41*Mouser_X hears screaming coming from the rough direction of Washington state...
06:24:22emeralddcheck playback?
06:26:35Calcipherwe were thinking that using the rec button on the e200 series players as a "speak current item, or wps " function would be a great accesability improvement
06:27:06emeralddlostlogic: check the code in set_filebuf_watermark
06:27:37Calcipherthats only an example of how the feature could be mapped, since I don't think the rec button does anything on my rb'ed Sansa
06:28:22emeralddand line 2390 in playback.c
06:28:35emeralddbuffer_margin is set to 0 at line 259
06:29:09emeralddso when you get to line 2390 in audio_load_track
06:29:15 Join JdGordon [0] (
06:29:39emeralddyou have set_filebuf_watermark(buffer_margin, 0) which is (0,0)
06:29:59emeralddset_filebuf_watermark takes the max of those (0)
06:30:04lostlogicline 3279
06:30:11lostlogic#ifndef flash
06:30:15lostlogicsets the margin
06:30:46emeralddc200 has flash storage alright
06:31:08lostlogicCalcipher: that does sound like a nice feature, but most often all buttons are already mapped... still, worth submitting to flyspray as a feature to track... most of the 'major' devs don't seem to be in here atm
06:31:39 Quit JdGordon (Client Quit)
06:32:36Calcipherah ok, thanks lostlogic, I'll do that, maybe it should be an optional setting?... don't know how it would be approached for other targets, since I myself only have a e200 series player
06:33:36lostlogicCalcipher: yeah, I hear ya −− could even be a build-time option to map remap another function elsewhere for an accessible build.
06:34:33 Part SoulinEther
06:34:34 Quit Mouser_X ("CGI:IRC 0.5.9 (2006/06/06)")
06:35:08Calcipheryeah, do you want to go ahead and add the item yourself, you have a better understanding of rockbox, so you can suggest better ways to implement that right off the bat
06:36:45lostlogicemeraldd: committed
06:36:48emeralddlostlogic: perhaps this?
06:36:53lostlogicCalcipher: no, I don't
06:37:22lostlogicemeraldd: see what I committed instead (took me two commits, because I'm special)
06:38:26 Join JdGordon [0] (
06:38:57Calcipher•lostlogic• ok, so I will make the request, and wait and see what the other devs think up, on what the best way to implement that kind of function
06:39:26Calcipherwill be my first request made on flyspray
06:39:48emeralddthat will return the same result, i think
06:40:19lostlogicCalcipher: yeah, I'm not sure what you should expect in the way of a response, you'll probably want to come back here and poke at some of the other guys who are more UI focussed once the request is in FS to solicit them to comment on it and to make sure it's not forgotten
06:40:26emeralddAhh, neverminf
06:40:57CalcipherI will
06:41:17lostlogicCalcipher: I personally stay as far away from the UI code as I can ;)
06:41:37Calcipherhaha, why is that?
06:41:50lostlogicemeraldd: my theory is that this way it's defaulted to 512k unless a codec says otherwise... I think...
06:42:12lostlogicCalcipher: dont' like it, don't use it... I turn my player on on shuffle over a large playlist and ignore it ;)
06:44:16 Join hannesd_ [0] (
06:45:25emeralddlostlogic: that looks like what it should do . . . . building
06:45:37lostlogicemeraldd: sighugh
06:45:53lostlogicall that chasing of other race conditions that _should_ have been recovering
06:45:56lostlogicI guess it's good
06:46:00lostlogicbetter not to hafta be 'recovering'
06:46:03lostlogicbut still what a couple of days
06:46:13hcslostlogic: fwiw I'm testing it too, just gonna let it run on the foul MP3s whilst I shower
06:46:21lostlogichcs: thanks man
06:46:33lostlogicI feel like I've exhausted 3 testers on this bug now
06:47:01hcsso the root of the issue was that someone thought flash access was instantaneous?
06:47:05emeralddit's been fun
06:47:19lostlogicfwiw, I'm also going to commit some other fixage that should make it not need that fixage... in a bit
06:47:20hcsprecognitive, even?
06:47:36lostlogichcs: loosely yeah basically didn't fully look at the side effects of setting the watermark to zero
06:47:40lostlogicie it will never fill
06:48:10lostlogicI'm considering adding an explicit minimum watermark of CONF_FILECHUNK into the buffering side so that it protects itself from idiot clients
06:48:23emeralddthat would be a good idea
06:48:29emeraldddeals with broken codec's as well
06:48:33lostlogicemeraldd: good point.
06:48:44lostlogicalways good to protect oneself from those.
06:49:35emeraldda logf for the bad watermark would probably be a good idea,
06:49:56lostlogicemeraldd: mmhmm
06:50:53emeralddhere we gooooo . . . . .
06:50:58*lostlogic ducks
06:51:37lostlogicemeraldd: I get the feeling that you will be developing with us soon... you're a pretty quick study.
06:52:11emeralddIt filled!
06:52:32hcssame here
06:52:55lostlogicclosing damn gremlin filled bug from hell
06:53:43emeralddMy day job is hunting bugs in an archiac c/perl/informix 4gl/spl system
06:53:49emeralddthat thing is a nightmare
06:53:59hcsin retrospect it seems obvious that the places to check would have been HAVE_FLASH_STORAGE
06:53:59lostlogicemeraldd: good god... large perl applications are ... ouch
06:54:20lostlogichcs: yeah, I didn't realize there were any inside playback code :(
06:54:26eigmaJdGordon: ping
06:54:32lostlogicthat's what i get for having not been active for so many months leading up to MoB
06:54:37JdGordonhey eigma
06:54:47eigmaJdGordon: I broke'd my m:robe :(
06:55:03JdGordonoh no
06:55:11emeralddlostlogic: and it did it again :) cool it's working
06:55:14eigmathere's something seriously wrong with the screen
06:55:28JdGordonbroken? or software problem?
06:55:29eigmathe colours are completely screwed.. it doesn't show black at all
06:55:33eigmaI don't know
06:55:39eigmaI mean, I didn't crack it or anything like that
06:55:58JdGordonits open yeah? maybe something got shorted?
06:56:03eigmaI'm hoping it's a sw problem, but leaving it with the battery unplugged for a wihle didn't help
06:56:12JdGordon:'( that sucks
06:56:36eigmaI was sending some strange commands over SPI with the backlight chip selected, but I don't see how that could have persistent bad effects
06:57:51eigmayou think they'd take it back in warranty? ;) −−
06:58:05eigma(note wires sticking out of back)
06:58:32 Join RoC_MasterMind [0] (
06:58:44emeralddlostlogic:So how long till the commit ?
07:00:36 Quit hannesd (Read error: 110 (Connection timed out))
07:00:36 Nick hannesd_ is now known as hannesd (
07:02:03emeralddlostlogic: crap, I'm brain dead
07:02:04lostlogicemeraldd: just did a couple more commits
07:02:37lostlogiceigma: jtag doesn't void warranty does it???
07:02:52lostlogicI sure have been giving the build servers a workout lately
07:04:33Calcipheroh damn, daylight savings kicked in.
07:04:56lostlogiccouple more hours here, hehe
07:04:56Calcipherfuckin government messing with time.
07:05:20eigmalostlogic: I'm pretty sure soldering on the board and drilling a hole in the case does..
07:05:31eigma...unfortunately ;)
07:06:44 Join amin [0] (
07:07:02eigmazomg is down.. must have crashed because of new DST
07:07:28aminthere's no un-tagged music in my itunes, that Ive transferred to my ipod, but after accessing that music from .rockbox it still is showing some part of the music as untagged :|
07:08:12emeralddtag version?
07:08:42aminemeraldd didnt get you ..
07:08:55lostlogic'night all, hope no gremlins are biting
07:09:36emeralddI'm thinking that I've seen that if I have music that is either badly tagged or conflictingly tagged . . . .
07:09:48emeralddI could be thinking of something completely different though
07:10:59aminrockbox is showing [Title] = No info, [Artist] = Kylie Minouge, [Album] = Fever
07:11:17aminof that particular music that it's showing untagged
07:11:25aminbut same track shows up fine from itunes
07:14:34***Saving seen data "./dancer.seen"
07:15:00emeralddno clue :(
07:22:28webguest51how do i get ubuntu to recognize my ipod? it was autodetected until i put rockbox onto it
07:22:40 Quit JdGordon ("Konversation terminated!")
07:24:13 Join JdGordon [0] (
07:28:36emeralddamin: have you tried retagging the file? ( clearing all tags and resetting them?)
07:31:59aminemeraldd no I havent how to do that
07:33:54Calcipheraww crap, added a feature request on the first, and of course I had to misspell the title
07:34:51CalcipherIf anyone cares enough to edit the title, and also consider the request hehe, that would be appreciated
07:34:58emeralddfoo, unfortunately, I know next to nothing about itunes :(
07:35:33CalcipherFS #8083
07:36:15aminemeraldd dont tell me abt itunes.. just tell me the procedure to tag every track by resetting them, Ill delete everything from ituens and update it again
07:38:03 Join Llorea1 [0] (n=llorean@
07:38:21 Quit Llorean (Nick collision from services.)
07:38:23 Nick Llorea1 is now known as Llorean (n=llorean@
07:38:59emeralddamin:how I would do it probably won't work for you, I'd use amarok/mp3info on linux
07:40:14CalcipherI like amarok, what linux are you using emeraldd?
07:40:48Calcipheroh, I've never used gentoo
07:41:15emeralddyou need a good chip, every install is a compile
07:41:26CalcipherI have only used Mandrake and Ubuntu
07:42:00emeralddamin:look for an id3 tag editor online
07:42:11CalcipherI know next to nothing about Linux, and what I learned I forgot most of
07:42:30emeralddcalcipher: the way you keep it is to use it all the time :)
07:42:50emeralddI don't have machine that I use regularly with anything but linux/unix on it
07:43:44CalcipherI see, I have a Mac and a frankenstein PC with windows XP
07:44:23CalcipherI'm just used to being able to find things easier for windows
07:44:49Calcipherfelt like more of a hassel when I was using mandrake
07:45:07emeralddespecially after they butchered it into mandriva
07:45:11CalcipherI was told by many people already, that Mandrake sucks hard
07:45:22Calcipherso that might be a bad intro to linux
07:45:37emeralddI used it for a while, and did pretty well with it
07:45:40emeralddurpmi is nice
07:45:46emeralddyum is decent on fedora
07:45:58emeralddand this is getting well off topic ;)
07:46:25Calcipherhehe very true
07:46:43 Join webguest99 [0] (i=80b1e623@gateway/web/cgi-irc/
07:46:51emeralddTake a look at a recent fedora, and with that, and a clock reading 1:46 am, I'm out of here ;)
07:47:01webguest99can someone help me with rockbox
07:47:04 Quit emeraldd ("using sirc version 2.211+KSIRC/1.3.12")
07:47:26webguest99is there anyway i can just start up my normal ipod without rockbox starting or do i have to uninstall it completly
07:47:54aminHow do I play videos on rockbox ? Ive transferred few videos thru itunes .. but it doesnt display those videos as it has displayed songs transferred thru itunes
07:49:35 Quit jhulst (Read error: 113 (No route to host))
07:50:09 Quit eigma (Read error: 113 (No route to host))
07:50:53 Quit webguest57 ("CGI:IRC")
07:51:11 Quit RoC_MasterMind (Connection timed out)
07:51:58 Join RoC_MasterMind [0] (
07:54:34 Quit webguest99 ("CGI:IRC (EOF)")
07:57:14Lloreanamin: In all seriousness, you need to read the manual and look things up in it before asking in here. Over the last few days you've asked many questions that are answered in the manual.
08:00:51aminLlorean I checked it but there it's not mentioned how to bring untagged tracks at the bottom when I go in database > tracks
08:02:04LloreanYour last question was about video...
08:02:18 Join J-23 [0] (
08:02:39LloreanAs for your iTunes music, iTunes doesn't always tag songs, it just puts the info in *its* database. Use a better tagging program
08:03:15aminI used godfather
08:03:25aminand still I cant see tagged music in rockbox
08:05:01 Join ptw419 [0] (n=ptw419@
08:05:04LloreanYou said Rockbox is showing that the title tag is empty
08:05:09LloreanThis means something is still wrong with it.
08:05:44aminis there a possibility that I can select all the untagged files and Just delete them
08:06:14LloreanYou can just create a playlist of all the untagged files, then use that to look them up.
08:06:54 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:06:55 Quit webguest51 ("CGI:IRC (EOF)")
08:07:23aminIve also checking abt the artwork of the album .. like the when I play a certain album track it shows the image of the artist but in rockbox .wps it does not
08:08:27 Quit RoC_MasterMind ("Leaving")
08:08:42 Join JdGordon [0] (
08:14:32 Join MrRichard [0] (i=Tentacle@
08:14:39MrRichardHello irc.
08:14:53aminnow this is strange .. I merely have 2-3 untagged tracks if i go thru Database > Artist .. but theres a huge list of untrack songs if I go thru Database > Tracks .. is this is a bug in .rockbox ?
08:15:51MrRichardI think I've had similar problems, amin
08:15:57MrRichardWhat did you use to tag your songs?
08:16:01 Join ptw [0] (
08:16:18MrRichardOh okay. I haven't a clue then.
08:16:28aminwhat did u use ?
08:17:49Lloreanamin: "Untagged" under "Artist" would mean they don't have the artist tag, "Untagged" under track would mean they don't have the title tag, I think.
08:18:33LloreanAnd please respect our channel guidelines.
08:19:07aminI havent dis respect any, intentionally
08:19:46 Join Rob2222 [0] (
08:19:50Llorean"Use clear, grammatical, correctly-spelled English. "Words" such as "thx", "u", and "kewl" are not acceptable."
08:20:05MrRichardI've been trying to sort my database issue out for a while now, and still to no avail.
08:21:14MrRichardIf my database will not initialise [ie gets stuck on 'found... 0'], and we'll assume it's because of some crazy tag; if I remove the file with the malevolent tag, will my database work, or will I have to reformat entirely?
08:21:17 Quit ptw ()
08:21:36LloreanMrRichard: If you can find the problematic file, then you should just need to "initialize now" again and wait.
08:22:23MrRichardllorean: excellent, thank you.
08:22:44MrRichardJust to make it easier to find - what sort of things am I supposed to be looking for that may be cuasing this problem?
08:22:58LloreanI really don't know.
08:23:03MrRichardOh okay.
08:23:12LloreanIn all honesty, nothing should cause that sort of behaviour. So once you identify what *is* causing it, you should file a bug report about it.
08:23:29MrRichardWill do. thanks. Nobody seems to be able to help :P
08:24:10LloreanUnfortunately, nobody really can help with problems that they haven't experienced and solved before
08:24:23 Quit ptw419 (Read error: 110 (Connection timed out))
08:24:55MrRichardThat's what worries me. Something must be heinously wrong if it is some sort of exotic error
08:25:20 Join SirFunk [0] (
08:26:32LloreanMore likely it's probably some very obvious and silly mistake.
08:26:39LloreanLike using > instead of >= somewhere
08:26:47MrRichardLet's hope.
08:26:53LloreanBut until there's a clear sign where to look it'd be nearly impossible to spot
08:27:06MrRichardI'll test now, by deleting some files and seeing what changes
08:28:10MrRichardOut of curiosity, would strange characters, such as japanese characters possibly cause any issues?
08:28:31LloreanThey *shouldn't*
08:28:36Lloreanbut that's a should not, not will not. :)
08:29:01MrRichardYeah. I've just heard of certain players not being overly keen on foreign characters
08:29:14LloreanRockbox works the same on all players, though.
08:31:01aminLlorean ive updated all the tags now, with track names. will they be automatically synced if I connect my ipod to itunes
08:31:20LloreanI don't use iTunes, so I have no idea.
08:31:33MrRichardamin: try to get into the habit of drag+dropping your music.
08:31:49MrRichardas opposed to using iTunes
08:33:52aminMrRichard drag and drop from the itunes or just to the f:\ipod
08:37:34 Quit Rob222241 (Read error: 110 (Connection timed out))
08:39:36hcsjust wondering, it'd be silly to record in stereo on a sansa, as it only has one microphone? (other than FM, of course)
08:41:02J-23Can I use multiboot - original firmware and Rockbox?
08:41:07J-23(Sansa c240)
08:41:56Lloreanhcs: It'd be silly to record on the sansa, since it's at half the usual sample rate iirc. :-P but yeah, stereo from the internal mic is probably not useful
08:42:31hcsLlorean: well, considering I don't have a microphone for the iPod it's my only choice at the moment
08:43:03LloreanBut yeah, it should just be a mono input
08:45:51J-23or restore original firmware without booting Rocbox?
08:47:23LloreanJ-23: The c200 should dual boot just fine...
08:49:05J-23I need MSC mode?
08:54:04J-23How I can configure dualboot?
08:55:03 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
08:55:51J-23Bootloader install returns "No sansa found"
08:56:11LloreanAre you using sansapatcher or rbutil?
08:56:31J-23rbutilqt on Windows
08:56:35LloreanTry sansapatcher
09:00:10J-23it works! :)
09:01:28J-23But can I restore original firmware without starting RB? (i.e. when Rockbox crash)
09:13:55 Join webguest03 [0] (i=54d15b74@gateway/web/cgi-irc/
09:14:37***Saving seen data "./dancer.seen"
09:15:30J-23How I can play music from old firmware?
09:17:46IsolinearIs there a way to see a list like but that goes even farther back in time?
09:19:20J-23I have uploaded music to my Sansa with MTP, how I can copy it to Rockbox?
09:22:42LloreanIsolinear: Look at the links at the bottom of the list...
09:23:15LloreanJ-23: The music is still there, exactly where you put it. As long as it's not copy-protected, and is in a format Rockbox can play, just use database, or browse to it.
09:23:38IsolinearHaha, I hadn't quite scrolled far enough... Thanks.. :)
09:26:03JdGordonLlorean: changed your mind with the keymap?
09:26:40LloreanJdGordon: Been a hell of a day.
09:27:02LloreanThanks for reminding me.
09:35:04LloreanJdGordon: I'll commit it tomorrow. The patch is ready, but I really ought to fix the manual before I do so.
09:36:36LloreanI also want to come up with a commit message that will present minimal confusion, and I'm not sure I can manage that tonight. :)
09:37:37scorche"Things changed."
09:37:57Llorean"Things changed, you'll figure it out" ;)
09:39:17MrRichardstill not having any luck with these database issues...
09:39:37MrRichardIve been deleting suspicious files/folders one by one
09:39:45MrRichardbut so far nothing's worked.
09:39:53scorcheMrRichard: do a binary search
09:40:01MrRichardHow would I go about doing that?
09:40:39scorchedelete half of your files and see if it is apparent...if not, check the other half....repeat until you ahve teh file(s)
09:41:08MrRichardThat's similar to what I'm doing. But I'll try that now.
09:41:42MrRichardWhat I'm doing is deleting files, deleting database.tcd, then reinitialising the database to see if it works. Is this what I should be doing>
09:42:33aminLlorean if I want to play videos on rockbox, do I need to copy it in .rockbox/videos or it can play those videos that are present in applefirmware
09:43:09MrRichardamin: Someone correct me if I'm wrong, but I'm pretty sure you can put the video anywhere and then just open it respectively.
09:43:16Lloreanamin: Remember what I said about reading the manual?
09:43:36 Join ddalton [0] (
09:43:48aminthanks MrRichard
09:43:52ddaltonAnyone got the docs for the e200 rtc?
09:44:06ddaltonI am thinking of buying one and want to get the wakeup alarm working...
09:44:22MrRicharde200? Is that the sansa?
09:44:43aminyes Llorean I do remember and after reading it, I still could not figure out how to play videos on rockbox because I've transferred the videos in applefirmware but rockbox cannot locate those videos through Files
09:45:04Lloreanamin: And the manual says you need to convert videos specifically for rockbox...
09:45:10ddalton250, 260, 270 and 280
09:45:20ddalton2, 4, 6 and 8 gb
09:45:38aminLlorean thanks, I will carefully read it again
09:45:39MrRichardamin: llorean is correct, this is specified in the manual.
09:46:29ddaltonwho did the port?
09:46:34ddalton(sansa e200)
09:47:57*ddalton Svn is annoying ddalton
09:49:02ddaltonI know I should know the answer to this. But how do I delete a dir in linux?
09:50:18Jddalton: rm -rf
09:50:32ddaltonoh ok thanks.
09:51:24ddaltonCalcipher: around?
09:52:41Calcipher•ddalton• yup
09:52:55MrRichardThis is getting ridiculous, I've deleted nearly my entire library and still my database will not work
09:53:00Calcipherits almost 4am here, but not so bright me is still awake
09:53:32 Quit webguest03 ("CGI:IRC")
09:53:46ddaltonCalcipher: ok well. Im just going to read your comment on that fs report. I would think a list for the wps info would be best. What do you think?
09:54:17ddaltonlots of info coming at you isn't good trust me. I use screenreaders and its not a nice thing...
09:55:31ddaltonand configurable buttons won't be accepted...
09:55:34MrRichardDatabase info states "DB Ready: No" "Progress: -1% (0 entries)" and "RAM Cache: No" Anyone have /any/ idea what may eb causing this?
09:55:42Calcipher•ddalton• I understand, thats why I'm suggesting having the items that are included on the "speak wps" function be pre selected from a list of information you want it to read each time
09:56:07 Part J-23
09:56:30Calcipherso if you want it to read name, elapsed/duration, artist, album, next
09:56:52ddaltonCalcipher: Hang on. How many settings do you want!
09:56:57Calcipheror any combination of possible items that you can set in settings
09:58:13Calcipher•ddalton• what do you think about that idea? that way if you want less information you can specify what fields it looks to read to you on wps
09:58:20MrRichardI only have a quarter of my library left on the device now ...
09:58:36MrRichardShould I have Load to RAM selected?
09:59:15ddaltonCalcipher: I would say. Just speak the elapsed time and length. You can go to track info for more information...
09:59:16Calcipher•ddalton• its almost like a configurable alternate voice driven wps if you think about how different wps will display different information
09:59:36ddaltonhave you seen sdoyon's patch (in 6323?)
09:59:40ddaltonid3 info voicing.
10:00:08Calcipherno I haven'te
10:00:19ddaltoncheck it out.
10:00:35MrRichardif I delete the .playlist_control file in .rockbox, will it delete any playlists I have made?
10:00:40CalcipherI have never tried any patches, don't even know how to install them yet
10:01:14ddaltonyou need to setup your build environment. worth doing.
10:01:20ddaltonbut just read the description.
10:01:23MrRichardI struggled with patching
10:01:26MrRichardToo many errors...
10:01:49ddaltonnot if they have been resynced
10:02:01ddaltonI guess it helps me that I understand how to resync them...
10:03:34ddaltonso you can get more info that way. But for quick info maybe this one...
10:03:42aminany one know some other site for themes except
10:03:47ddaltonlet me write one up for the wps and you can tell me what you think.
10:03:57ddaltonI just need to figure out how to find the elapsed time...
10:08:15MrRichardOh man. I've basically emptied my library and still database won't work...
10:08:53CalcipherI personally only see that time function useful if your listening to a long file like an audiobook or cast
10:09:40ddaltonwell. 1. To configurable. 2. Lots info at once isn't good
10:09:53 Quit Bagder ("*plopp*")
10:11:04Calciphertoo configurable?
10:11:19ddaltonyes. we don't like lots of settings in rockbox. :-)
10:11:37ddaltonask JdGordon why
10:12:11Calcipherwell its not like its up to me anyway
10:12:49CalcipherI've never used any of those patches, I noticed there are a few interesting ones added by stephan doyon
10:13:09ddaltonok going for dinner. ill see if I can write up a patch. But very busy tonight.
10:13:12MrRichardI give up
10:13:16MrRichardIm reformatting
10:13:37ddaltonget your dev environment working first because you will need it to test this out if I write it
10:15:23MrRichardWhile I'm reformatting, can anyone recommend a good .flac ripper/converter
10:15:24 Join Thundercloud [0] (
10:17:18 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
10:17:29aminwhich is the folder that .rockbox copy from backdrop or backdrops
10:18:03JdGordonwhat sort of configurabilty do we want to allow for the statusbar?
10:18:32JdGordonthe default icons there are way too small on the mr500 so i figured i may as well fix it for all targets instead of just that one
10:20:27 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
10:22:26 Join Siku [0] (
10:22:48Calcipher•JdGordon• what status bar changes are you planning?
10:23:28JdGordonim not really sure how far to take it.. custom icons and colours i guess
10:23:36linuxstbJdGordon: IMO we want it to be as configurable as possible. But at the very least, I think we need custom height, colours (fg/bg), and ability to display info as either text or icons.
10:23:41Calcipher•JdGordon• also, have you looked at any of the earlier comments on the recent voice feature request I made?
10:24:20Calciphernice, I would love a more visible status bar
10:24:41linuxstbAre there any E200 users running the Rockbox bootloader instead of the Sansa bootloader? (sansapatcher -bl option) If so, is it working well?
10:24:42Calcipherso allowing configurability will make that possible
10:25:01JdGordonlinuxstb: i am... almost no speed difference, but it does work fine
10:25:13linuxstbJdGordon: Have you lost recovery mode?
10:25:20JdGordoni dont tihnk so
10:25:27*JdGordon will check
10:25:57linuxstbI'm thinking it may be nice to make that the default install method.
10:25:57CalcipherI don't even remember what bootloader I'm using.
10:26:29Calcipher•linuxstb• what method is that?
10:26:36 Join davina [0] (
10:26:38JdGordonlinuxstb: yeah, looks like I have lost recovery...
10:26:45linuxstbIt would also make it easy to extend e200rpatcher to install the bootloader - it will simply need to embed a bootloader image in the uploaded application, and write it to the start of the firmware partition.
10:27:06linuxstbJdGordon: OK, that makes sense - it seems recovery mode is a feature of the sandisk bootloader.
10:27:18Bagderit is, yes
10:27:22JdGordonwe dont really need rec mode though do we?
10:27:43linuxstb_we_ don't, but users seem to find a way to need it...
10:28:30linuxstbAlthough most problems seem to be caused by people doing wrong things in recovery mode, so disabling it might actually help ;)
10:29:30JdGordonhaah yeah
10:29:53Bagderhehee indeed
10:29:55linuxstbamiconn: Have you started runtime tests with HD power-off? I'm about to start a reference test on my Color (from before your commit).
10:30:37 Quit amin ()
10:30:53ddaltonCalcipher: got homework to do but will see if I can write you up one sometime this week.
10:31:05linuxstbWhat about the standard firmware upgrade procedure - copying a firmware file to the FAT32 partition? Is that a feature of the main firmware?
10:31:19 Join pepie34 [0] (
10:31:21ddaltonJdGordon: BTW do you have the sansa e200 rtc docs?
10:31:39Bagderlinuxstb: I would assume it is at least
10:32:14CalcipherI'm heading to bed now, but I would like to hear what you all think of some suggestion I made, and would like to hear feedback on how or if its possible
10:32:29Bagderbtw, the view does indeed look a lot like the ipod video:
10:32:44Calcipherthe item is FS #8083, and a wonderful misspelled title
10:33:13ddaltonlinuxstb: or do you have them?
10:34:10linuxstbddalton: No
10:35:15 Join zicho [0] (
10:35:37linuxstbBagder: So it seems the View has TV-Out, but sandisk aren't selling a cable for it?
10:35:46Bagdervery strange
10:36:10JdGordonnot selling *YET*
10:36:49Calcipherhe he
10:37:23JdGordonBagder: I tihnk we are about the same point with the mr500 that the e200 was when we started offering svn builds... doom works!
10:37:50*linuxstb begs that we write documentation before announcing it's supported and offering downloads
10:37:50Bagderso you want mr500 builds added to the build and daily tables?
10:38:11Bagderyeah, but we can add it to the build table anyway, just not providing any zips
10:38:14JdGordonwould be nice.. but actually i tihnk linuxstb has a good point
10:38:35linuxstbIt still had a few warnings last time I built it (yesterday)
10:38:45*JdGordon looking to hopefully get help writing the touchpad code for all the screens :p
10:38:50JdGordonyeah, they can be cleaned up
10:39:02JdGordonBagder: no need for a bootloader build if you do add them
10:39:09Bagderhaving it in the build table usually helps cleaning up the warnings ;-)
10:39:09linuxstbWhat does installation involve for the mr500?
10:39:20Bagderso a sim and a normal build then
10:39:34JdGordonits not a simple install proccess yet :(
10:39:37linuxstbJdGordon: Why don't you need a bootloader build?
10:39:50JdGordonthe OF loads the main rockbox binary
10:39:58linuxstbThat's friendly of it...
10:40:05JdGordonyou needs to install a hacked OF first
10:40:19JdGordonthen create a remix thing which then gets replaced with a hacked .svg
10:40:33Calciphernight folks
10:41:00Calcipherthat sounds like the process for the linux install on the mrobe
10:41:04ddaltonJdGordon: if you get them can you send them to me?
10:41:31linuxstbHow is the OF hacked? Is there a tool available to patch it, or do you just download it from somewhere?
10:41:36JdGordononly certain peo[ple got them
10:41:51JdGordonIm not sure
10:42:01JdGordoneigima and kkurbjun will know about that more than me
10:42:01ddaltonJdGordon: who has them? how can I get them?
10:42:10*JdGordon not sure who has them
10:42:18linuxstbJdGordon: So presumably you were given a copy of the hacked firmware to install?
10:42:23JdGordonwhy do you want them? you really think you'll be able to do anything usefull with them?>
10:42:32*ddalton Does his homework
10:42:55ddaltonok better go now.
10:42:58 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
10:44:18Bagderok, two mr500 builds added to the build table
10:45:17JdGordonyay :)
10:45:29*JdGordon gets ready to fix yellow
10:46:01JdGordonBagder: oh, is there a reason the 2 tables are in a different order?
10:46:20JdGordonoh.. ok :p
10:46:35 Join bertrik [0] (
10:46:43Bagderbut the reason is so lame I don't dare to tell ;-)
10:47:06JdGordonyou dont want to fix it? its annoying :p
10:47:28Bagderreally, the column order in the delta table is meant to be controlled by a sort order table, but it bugs
10:47:44Bagderthe order of the columns in the main build table is simply alphabetical
10:47:49Bagderof the full build name
10:47:55linuxstbJdGordon: Are you sure the m:robe OF is hacked? I've just been reading about the Linux installation, and my first impression is that it's a specific version of the OF, which has a known exploit.
10:48:18linuxstb(I could be wrong...)
10:48:26JdGordonI'm pretty sure its hacked.... but yeah, i cant really remember
10:49:11linuxstbIf it's hacked, then I can't see why we can't just modify it to insert our bootloader code into the image.
10:50:04JdGordonwe havnt got up to that yet.. but that is the plan
10:50:22JdGordononly eigima has jtag connected to his, and he may have damaged it today which isnt very good :p
10:50:53 Quit pepie34 ("Ex-Chat")
10:53:23 Quit zicho (Remote closed the connection)
10:53:33linuxstbWho needs jtag anyway... ;)
10:53:49Bagdergee, the problem with the non-working column sort order was that I modified the wrong script... :-P
10:56:47amiconnlostlogic: Battery runtime should now be better than before, but I never said it would be better than the apple OF on 4th gen and up
10:57:01amiconnThat requires quite some more work
10:57:39amiconnOn PP5002 (1st...3rd gen ) we are on par or even a bit better than the OF
10:58:44linuxstbWhat sort of runtime do you get on your PP5002 targets?
10:59:43 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
11:03:12*linuxstb wants an email "diff" program to read Rocker's emails...
11:08:47 Join ender` [0] (
11:12:08 Join ilgufo [0] (
11:12:42JdGordonthe mrobe wasnt put in a nice spot :p it was good having all the hwcodecs together
11:13:10linuxstbI was just about to ask the same - was that intentional? (Bagder?)
11:13:28amiconnlinuxstb: I my test in August, I got 16 hours 3 minutes out of my 2nd gen
11:13:49 Join Bagder_ [0] (
11:13:54amiconnIt has a 1600mAh battery instead of the stock 1200mAh, so "normalized" it would mean 12 hours. Apple promises 10 hours
11:14:19amiconnAnd there's another (slight) improvement that went in after this runtime test
11:14:27linuxstbamiconn: That's very respectable... I assume you haven't jumped through the hoops required to test the OF runtime?
11:14:40***Saving seen data "./dancer.seen"
11:14:56amiconnNo, as that would require installing iTunes, which refuse to do
11:15:01linuxstbBagder_: Is the position of the m:robe in the build table intentional?
11:15:03amiconn(insert "i")
11:15:10linuxstbamiconn: There are also lots of alternatives...
11:15:17JdGordonBagder_: oh.. re your camera question... dont bother [aying a premium for sdhc... you'll run out of battery before your 3gb chip is filled anyway
11:16:46*linuxstb hopes MoB is stable enough to do a runtime test
11:17:41*amiconn wonders whether he should perform a runtime test on the G5.5 and the mini G2
11:18:41linuxstbHopefully someone will...
11:18:52 Join desowin [0] (
11:20:29amiconnlinuxstb: On 2nd gen, disk poweroff and the PP internal device switch-off each provided about half of the runtime improvement
11:24:42 Join GnosticE [0] (
11:25:27 Part Llorean
11:30:17 Quit Bagder (Read error: 110 (Connection timed out))
11:33:35Soapthat much? Wow. Just because of the Nano's poor runtime I guessed the HDD was less of a factor.
11:34:14SoapMy Video's battery is growing old, but I should be able to do some runtime tests this week.
11:35:55desowinamiconn: half of whole runtime? so if it was 8 hours it gets to 12?
11:36:12 Quit MrRichard ()
11:37:28 Quit ilgufo ("So Long, and Thanks For All the Fish -")
11:38:09 Join merbanan [0] (n=banan@
11:38:09 Quit amiconn (" HydraIRC -> <- Organize your IRC")
11:40:42 Join mf0102 [0] (n=michi@
11:43:34 Join spiorf [0] (
11:51:38JdGordonBagder_: no mrobe in the deltas table..
11:55:20JdGordonBagder_: and when you get a chance.. can you move SimMouseInterface to TouchpadInterface please?
11:57:24JdGordonpixelma: you round?
11:58:04pixelmayeah, but lunch break shortly, bbl
11:58:14JdGordonok dw
11:58:42JdGordoni was going to ask about the manual and where to add the touchpad instructions in... but it can wait
12:01:06pixelmaok, did you already have a look at the platform files? But as I said, I'm afk for a while now
12:02:30 Part pixelma
12:03:22 Join Buschel [0] (
12:04:03Buschelwow, just saw that switching off HDD was submitted for PP502x! will build now and do a battery bench :o)
12:04:13Buschelamiconn: thanks!
12:05:57bertrikBuschel: also a lot of work was done yesterday/last night on playback and buffering by lostlogic
12:08:23Buscheldoes anybody know how much current we save approximately?
12:11:47 Join Orc [0] (n=Orc@
12:12:05Orcsorry, Ineed help(
12:14:16OrcI update original firmware, after use Ipodpatcher and copy .rockbox on player (iPod nano).
12:14:41Soapsounds like you are on the right track so far.
12:14:55OrcRockBox is running, plugins and conf running
12:15:15Orcbut anu track dont playing
12:15:45 Join ackbahr [0] (
12:16:11Soapwhat happens when you play a track?
12:16:53Orcany button dont work
12:18:05OrcI understand because my Rockbox was stabilyty before backup
12:18:08 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
12:18:32Bagder_JdGordon: delta should appear now, wiki page renamed
12:18:33SoapUnfortunately for you a percentage of the Nanos out there seem to have problems with some recent changes in Rockbox. This is a (more or less) known issue and is being (more or less) looked into. I think there is a very good chance you have one of the effected Nanos.
12:18:35 Nick Bagder_ is now known as Bagder (
12:18:41*Bagder vanishes again
12:21:07Orcsorry, I understand this... This is a (more or less) known issue and is being (more or less) looked into
12:21:58OrcI can change original firmware on 1.1 version?
12:22:29SoapI honestly don't know what you can or can not do with the original firmware.
12:24:36 Nick parafin|away is now known as parafin (
12:24:45 Quit ackbahr (Remote closed the connection)
12:25:44 Join Domonoky [0] (
12:28:34 Join lee-qid [0] (
12:29:10 Join k-man [0] (n=SK@unaffiliated/k-man)
12:29:18k-mani have rockbox on my ipod
12:30:01k-mani loaded all the songs using the ipod OS and I'm just using the rockbox db to access the music, however each song appears 4 times in rockbox
12:30:15k-manlike the db adds the song each time it updates even if its already in the db
12:30:50SoapIs every song showing up exactly four times?
12:31:46OrcI repair, but dont know how
12:32:13OrcSoap: Its mystics
12:32:53k-manSoap, any ideas?
12:33:15SoapFour copies stumps me.
12:34:05Domonokyk-man: you could try deletin the *.tcd file in the rockbox dir, and rebuild the db.. and see if it helps :-)
12:34:20SoapHey. Soap being stumped isn't the end of the world, there are 146 other people here, and most all of them know more than Soap!
12:34:21k-manDomonoky, ah, ok, i'll try that
12:35:01 Join lazka [0] (
12:35:30pondlifeHmm, I just got my WPS displaying -32 of 375 (where -32 is the playlist position)
12:35:37pondlifeCan't repro though
12:36:06Domonokyk-man: also good a good test is, making a playlist of duplicate songs, and check the paths, if the duplicates in the db, are really the same file, or if your files are more then one time on the harddrive
12:37:26 Join sasukee_ [0] (n=SK@unaffiliated/sasukee)
12:38:47 Nick sasukee_ is now known as k-man__ (n=SK@unaffiliated/sasukee)
12:40:00k-man__do i just drop the rockbox files over the old ones to upgrade?
12:40:11 Quit Orc ("Leaving")
12:41:31 Join moos [0] (
12:42:34ompaulwhat do I do with a scrobbler.log to cause it to be uploaded?
12:44:15Domonokyompaul: you have to use a tool to submitt it (for example QtScrobbler)
12:44:40ompaulDomonoky, thanks
12:47:02BigBambik-man: yes
12:47:31 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
12:51:49JdGordonBagder: thanks
12:54:20JdGordonpondlife: how badly do we want 8084 fixed?
12:54:21 Join ameeen [0] (
12:54:34JdGordonit cant be done without a semi-nasty hack
12:54:35 Quit k-man (Read error: 110 (Connection timed out))
12:54:49pondlifeNot badly, but I'd like Set Song Rating to work reliably
12:55:09ameeeni saved a .mpg file in the folder: f:\ .. I can access the folder thru Fiels menu of rockbox but I cant see that video ?
12:55:44JdGordonill do the hack fix and you can decide if its worth it...
12:55:51pondlifeAlso, maybe Bagder could deny that task closure request...
12:56:00Domonokyameeen: change the view mode of rockbox.. see manual.. :-)
12:58:22JdGordonwouldn anyone object to having a proper screen for the song rating? that inline menu is farking horrible!
13:00:26JdGordonpondlife: should fix it.. up to you to commit...
13:00:29ameeenDomonoky I checked the manual, there's nothing mentioned abt from where I can get the mpeg player plugin :|
13:00:55pondlifeJdGordon: haha, no I'll let you take the full blame this time :)
13:01:09JdGordoncompiling to test on my sansa
13:01:11pondlifeA non-hacky fix would be better
13:01:17JdGordonnot possible
13:01:23pondlifeSurely it's possible
13:01:27JdGordonits this or nothing
13:01:39Domonokyameeen: change the view mode of the filebrowser in rockbox, then you can see the file, and play it... (how to chnage the view mode is in the manual)
13:01:52JdGordonwell... its this.. or a much larger hack
13:02:28JdGordonit does work though
13:04:05ameeenDomonoky yes I found it and chnged the mode to: view all files.. now I can see the video file and since Ive selected it to play, it gave an error: missing packet start code ... prefix: 00 at 1B28
13:07:12 Join Frazz [0] (
13:08:06pondlifeJdGordon: You've borked the voicing again ...
13:08:25pondlifeThat list commit, to redraw on no action
13:08:36JdGordonoh bloody hell :'(
13:08:41pondlifeI'll fix it
13:08:47pondlifeI have an idea
13:08:55davinawill rockbox work on the new ipod nano 4gb ?
13:08:56JdGordonits gotta be done properly...
13:09:08JdGordononce and for all...
13:09:10davinacool ty
13:09:19pondlifeYou want to see the time update? Was that the reason?
13:09:42 Join Sedgewick [0] (
13:09:43JdGordonyeah.. we need to get it to redraw often.. but not stop voicing
13:11:41linuxstbameeen: How did you create the video?
13:11:58ameeenlinuxstb I converted it into mpg video
13:12:08k-man__how is support for album art going?
13:14:44***Saving seen data "./dancer.seen"
13:14:50ameeenjpegs are working just fine, but taking too long to load .. but no complains.. vieos are not working at all
13:14:56 Quit Frazz ("Leaving")
13:17:27 Quit hcs ("Leaving.")
13:19:28 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
13:19:29JdGordonpondlife: that hack added bassically 0 delta.. so im happy with it :p
13:20:21amiconnbass, bass .... ;)
13:21:50JdGordontreble, treble... ?
13:30:21JdGordonpondlife: ok so i was wrong :p there is a better way to do it :p
13:30:40pondlifeWith a green delta?
13:31:36JdGordonprobably i higher red...
13:31:54JdGordonnope.. not working
13:32:44pondlifeHmm, the list isn't calling the get_talk routine when I move the selection...
13:32:55pondlifeSo my fix won't work
13:33:00pondlifeIsn't it meant to?
13:33:58JdGordoni tihnk list_do_Action is supposed to call that automatically
13:34:21pondlifeNot working here - in the info screen (on simulator, admittedly).
13:34:34pondlifeinfo_speak_item() is only called once, at initial enty,.
13:35:24pondlifeMaybe that's sim thing, I'll put it on H340.
13:37:47JdGordonrevert my commit... its not worth the trouble..
13:38:07JdGordona real proper fix needs to be found
13:38:12 Join lazka_ [0] (
13:38:25 Join ApotheoZ [0] (
13:41:06JdGordonpondlife: is the list supposed to speak every redraw?
13:41:49 Join jack [0] (i=52266cfe@gateway/web/cgi-irc/
13:42:29pondlifeEvery time the selection moves
13:43:11JdGordonso why does redrawng kill talking?
13:43:18jackhow did u install on 30gig ipod
13:43:34pondlifeI don't know now
13:44:03JdGordonit must be calling talk_shutup() too early
13:44:17pondlifeIt doesn't call that, does it
13:44:38 Quit jack (Client Quit)
13:45:14JdGordonit does.. as it should... just its too often
13:45:17pondlifePreviously it was calling info_speak_item() repeatedly
13:45:36JdGordon... or that could be the problem :p
13:45:44pondlifeStarting a new bit of speech before the previous got spoken.
13:46:01pondlifeBut now it's not calling it much at all :)
13:46:13pondlifeI've got to go, leave it with you
13:46:29JdGordonI found the problem :)
13:46:37JdGordonit is calling shutup too often
13:47:52JdGordonim not sure why though
13:48:28 Part pondlife ("Gone")
13:48:49 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
13:48:55 Join miepchen^schlaf [0] (
13:49:08 Quit jhMikeS (Nick collision from services.)
13:49:14 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
13:50:35 Quit ameeen ()
13:51:21 Quit lazka (Read error: 110 (Connection timed out))
13:53:17moosJdGordon: still red deltas ;P
13:53:35JdGordonyeah, i was about to say thats why hacks are better than good code :p
13:54:08mooshehe :)
14:00:46amiconnHmm, any e200 owner with a microsd card around?
14:01:30*JdGordon is
14:01:51amiconnI'd like to know whether the e200 finds the microsd card if it's plugged before booting
14:01:59amiconn(i.e. not hot-plug detection)
14:02:25JdGordonyeah, works fine
14:02:56JdGordonthere is a bit which is set if the card is in which is checked on boot
14:03:20amiconnI'm trying to make microsd work on c200
14:04:26amiconnSomehow even cold plug doesn't work although there is code for it...
14:04:48JdGordondisk_mount_all() eventually checks if the card is in
14:04:56 Quit k-man__ (Read error: 110 (Connection timed out))
14:04:57JdGordonotherwise its done though hotplug
14:07:57Buschelamiconn: do you think my patch (8075) will be submitted? or are you too much concerned about the stack size?
14:08:36amiconnjhMikeS: The PP5022 (and I'm quite sure the PP5020 too) also has the GPIO interrupts split like what you defined for PP5024 - and there's GPIO2_INT in addition to GPIO0_INT and GPIO1_INT
14:08:56amiconn0 is for GPIO A..D, 1 for E..H and 2 for I..L
14:09:20amiconnI'll move that to pp5020.h - but then pp5024.h becomes useless. Should I remove it?
14:18:05 Quit ApotheoZ ()
14:19:29 Quit billytwowilly (Read error: 110 (Connection timed out))
14:19:30 Quit Sedgewick (Read error: 110 (Connection timed out))
14:21:10 Quit davina (Remote closed the connection)
14:23:11 Join davina [0] (
14:23:20 Quit bertrik ("bye")
14:32:03 Quit mf0102 ("Verlassend")
14:33:41 Quit lee-qid ("aufwiederbyebientotsayonara")
14:33:49 Join lee-qid [0] (
14:35:24 Quit davina (Remote closed the connection)
14:36:29 Join karashata [0] (
14:37:15 Join davina [0] (
14:41:08 Quit GnosticE ()
14:43:15 Join bluey [0] (
14:51:32 Quit BigBambi (Read error: 113 (No route to host))
14:52:45 Quit lazka_ (Read error: 110 (Connection timed out))
14:55:27 Quit bluey ("This computer has gone to sleep")
15:00:02 Join Frazz [0] (
15:00:05 Join bluey [0] (
15:01:46 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
15:07:00pixelmaBigBambi: you might want to read the description of the patch you pointed to in the forums :)
15:07:30BigBambipixelma: OK, why?
15:07:37BigBambiDoes it do embedded?
15:07:57pixelmare. getting the lyrics from a tag
15:08:14BigBambiI will go search for it
15:09:01BigBambipixelma: Are you rtalking about 2999?
15:09:18pixelmanope, fs #7432
15:09:20BigBambiI didn't know that existed, I was talking about 7432
15:09:36BigBambipixelma: Ah, got it now
15:10:17BigBambiI never actually read the patch before, and when it was mentioned around the forums it was as a snc/lrc viewer
15:10:23BigBambiI will edit, cheers for the pointer
15:12:00 Join MajorC [0] (
15:12:33 Join darkapostrophe [0] (
15:14:45***Saving seen data "./dancer.seen"
15:15:01*amiconn summons low_light
15:15:10 Part bluey ("Leaving")
15:16:43 Quit JdGordon (Remote closed the connection)
15:17:58*amiconn wonders what the sansa OF uses rijndael for
15:26:38lostlogicamiconn: sorry for misquoting. I thought that battery spin down was the major obstacle between us and OF-esque runtimes.
15:26:43lostlogicer drive spin
15:26:54amiconnOne of 2
15:29:27 Quit Dark_Apostrophe (Read error: 110 (Connection timed out))
15:29:27 Nick darkapostrophe is now known as Dark_Apostrophe (
15:31:36 Quit Frazz (Read error: 104 (Connection reset by peer))
15:42:52 Quit karashata ("Leaving.")
15:43:08amiconnlostlogic: If disk powerdown were the only issue, iPod Nano battery runtime would be fine in rockbox. But it isn't.
15:43:18 Join karashata [0] (
15:43:40 Join bertrik [0] (
15:43:41lostlogicamiconn: *nod*
15:48:35amiconnHaha: the sansa OF contains debug strings for their database...
15:49:03lostlogicbertrik: is 8077 dead once and for all?
15:49:14lostlogicfrickin' bug.
15:49:18amiconnNo, things like this: Database::createRecord(long*, Database::Record*)
15:49:33lostlogicoohhh like full debug symbols, lol
15:49:33amiconnLooks like at least their database is written in C++
15:49:56bertriklostlogic: yes, looks like it, I tried but cannot reproduce it anymore
15:50:38bertriklostlogic: what was the root cause of the bug?
15:51:00lostlogicthere wer at least two root causes either one of which would have kept it from working
15:51:33lostlogicthe first one that got fixed was a livelock between codec and buffering caused by a copied and not updated yield_codec implementation from pre-MoB world
15:52:07lostlogicthe second was a flash specific problem where the lack of disk spin-up time was assumed to allow a _zero_ buffer watermark as a legal value, which is clearly not reasonabl.
15:52:52lostlogicalong the way a good half dozen or more other minor race conditions that were recoverable on other targets were resolved... each time with me thinking it would kill that bug *sigh*
15:53:11bertrikah, I noticed that usefl didn't drain all the way to 0 indeed
15:53:31bertrikwell, nice to have them fixed too
15:53:50lostlogicneither buffer should ever reach zero, if they do they aren't really doing their buffering job
16:08:59 Quit qwm (Remote closed the connection)
16:09:13 Join qwm [0] (
16:14:00 Join Lear [0] (i=chatzill@rockbox/developer/lear)
16:15:14 Join billytwowilly [0] (
16:15:46 Join ilgufo [0] (
16:16:03 Join emeraldd [0] (
16:16:03 Quit spaceinvader ("Server upgrades: hopefully back in an hour or so")
16:20:25 Join Daniel_S [0] (i=57b0f772@gateway/web/cgi-irc/
16:21:06 Quit spiorf (Read error: 110 (Connection timed out))
16:23:04 Quit qweru ("moo")
16:23:40 Quit Daniel_S (Client Quit)
16:23:46 Join Daniel_S [0] (i=57b0f943@gateway/web/cgi-irc/
16:24:52 Join lazka_ [0] (
16:25:15 Quit pixelma (Read error: 110 (Connection timed out))
16:26:14 Join spiorf [0] (
16:30:55 Join WalterEgo [0] (
16:31:59 Join bluey [0] (
16:34:25WalterEgoCouldn't see it mentionned in svn comments nor in the irc logs... Does the recent commits regarding disc spinofff (or poweroff?) on the PP502x affects the iriver H10 too? (the description says iPods and even though one h10 file gets changed, for all I know it might just be an if/else statement to skip over the added parts..(?))
16:43:31desowinI don't think so (I've looked at this diff), but I'm not authority
16:45:46WalterEgoThanks.. So I'll suppose it does not. Maybe later, or maybe it just isn't needed..
16:47:14 Quit Daniel_S ("CGI:IRC (Ping timeout)")
16:48:53 Quit moos ("Glory to Rockbox")
16:49:57 Join dandin1 [0] (
16:56:11 Quit animeloe (Read error: 104 (Connection reset by peer))
16:57:24 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
16:58:03 Join cooz [0] (
16:59:10 Join Daniel_S [0] (i=57b0d20b@gateway/web/cgi-irc/
16:59:43amiconnWalterEgo: No, H10 already had disk poweroff. The change in the H10 file is due to making something more general
17:02:20Soapamiconn: If you don't mind spending the time answering - why did the H10 already have disk poweroff, when the iPods didn't? Was it a previously REd find, or does the H10 not use the same PP ATA as the iPods?
17:03:32amiconnDisk poweroff is not handled by the ata controller, but via gpio ports
17:03:43amiconnAnd the ipods use different ports than the H10
17:04:00amiconnThe G5 even uses a port that wasn't know at all before last week
17:05:49Buschelbtw, the current runtime measurements show savings of ~8mA through the disk poweroff
17:06:08Buschelexpected more from the HDD's datasheet
17:06:19Soapwhat was the power consumption before disk poweroff?
17:06:36Buschel~50mA on ipod video 5.5G
17:07:12amiconnYou measured it?
17:07:53Buschelyou can calculate it via runtime and battery capacity
17:08:30 Join pixelma [0] (
17:08:47BuschelHDD's datasheet claims to use 70mW at 3.3V in sleep mode -> gives ~21mA
17:08:57 Join zlp [0] (i=zlp@
17:08:58Soapstill seems signigicant.
17:09:28Buschelyep, the current reduction is still great and will gain in longer runtime :o)
17:09:43 Join Heinz [0] (
17:09:49merbananBuschel: does rockbox last longer then the Apple firmware ?
17:09:56WalterEgoThanks. Man I wish I could understand those registers and assembly stuff and find out about how to turn off features 'cause, really, two-three hours battery life and no (useable) EQ makes for a rather poor H10 port.. YET, I can't live without crossfeed. That feature alone is worth using rockbox for (like, thanks whomever got it in!!).
17:10:20Buschelmerberan: no, not for 5.5G
17:10:26karashataWalterEgo: you only get 2 to 3 hours on the H10?
17:10:48WalterEgo5gb... The 20gb, I hear, has much better life out of its battery.
17:10:57Buschel5.5G 30GB's runtime should now go up to ~9.5h (~8h before)
17:11:23karashataahh, yeah, I'm getting on average around 12 hours or so with my 20GB H10
17:11:25amiconnWalterEgo: Small H10?
17:11:28merbananBuschel: do you know the time for the Apple firmware ?
17:11:36amiconnAh, yes
17:11:43Buschelapple sais 14h
17:11:49karashatathough, I did replace the battery with a 1700mAh one, since the stock battery died
17:11:49*amiconn also gets only ~3hours out of his 6GB
17:11:50WalterEgoamiconn - exactly.. I can live with it, though. :)
17:11:54zlphello people, Im having troubles trying to play mp3 on rockbox (iPod Nano), I can even play videos already with the audio track in mp3, but when I try to play a .mp3 it says "No codec for..."
17:13:50Buschelamiconn: is the ata controller also switched off when disk is powered off? or is it running "idle"?
17:13:59amiconnAtm it's not
17:14:14WalterEgozlp - Have you ever played mp3 correctly with rockbox and something just went borked or is it a new, first time install?
17:14:28amiconnWe need to learn how to init it properly. And I have the strong suspicion that this would help us solving the Nano problems
17:14:28Buschelhmm, do we know if switching it off may also save some more power?
17:14:47***Saving seen data "./dancer.seen"
17:14:48amiconnYes, but then we need to init it when powering it up again
17:15:13zlpWalterEgo - no, it is first time install, made with the official installer
17:15:16Buschelis this difficult? as far as i remember this was done for sansa?
17:15:19 Quit Daniel_S ("CGI:IRC")
17:15:24WalterEgozlp - I'm no expert.. I just remember having that sort of problem once when I updated the .rockbox files after a while... I had to change the bootloader together with it (there had been big changes)
17:15:53WalterEgozlp - Ohh... Then sorry, I don't know how I could help.. You tried different mp3 files, I guess?
17:15:55amiconnBuschel: Sansa is *not* ata
17:16:24 Quit ilgufo (Read error: 110 (Connection timed out))
17:16:39Buschelamiconn: maybe it wasn't the sansa... i'm not sure :/
17:16:41zlpWalterEgo - I have just downloaded the current build for nano
17:17:25emeralddAny news on the zogar's progress with usb?
17:17:51 Join eigma [0] (n=cat@
17:17:52zlpWalterEgo - yes, I tried... but thanks!
17:18:32WalterEgozlp - Wish I could help.. Tried rebooting the device and trying again?
17:19:55zlpWalterEgo - ajam, and I tried building the database and playing the mp3 from there, and also tried to play mp3 from the files browser
17:20:33 Quit bluey (Connection timed out)
17:20:35 Quit davina (Remote closed the connection)
17:20:38 Join Shagnar [0] (
17:21:01Buschelamiconn: found it. i talked about fs #7036 (was for e200)
17:21:20Shagnarhello guys. i want to start a new wiki page for the port of the iriver s10... (see ) - can anyone give me further assistance?
17:22:25 Join davina [0] (
17:45:00 Join Maha` [0] (i=tabac@gateway/gpg-tor/key-0xB9002659)
17:47:44 Join Caleb [0] (n=ichigo@unaffiliated/caleb/x-000007)
17:47:53Calebaww sweet
17:48:02 Join Hilikus [0] (
17:48:20Hilikuswhen will cover art be integrated with the main build?
17:48:33 Join mf0102 [0] (n=michi@
17:48:56krazykitHilikus, when it gets coded.
17:49:13BigBambiHilikus: When someone writes it in the correct way
17:49:16Hilikusit is, im using it already
17:49:24BigBambiHilikus: *correct way*
17:49:25Hilikusi see BigBambi
17:49:53Shagnarsecond try... : i want to start a new wiki page for the port of the iriver s10... (see ) - can anyone give me further assistance?
17:49:58krazykitHilikus, but the main thing holding it back has been implented now, so there's good news on there
17:50:01 Join illissius- [0] (
17:50:15BigBambiShagnar: Have you registered for the wiki?
17:50:28Hilikusyes, thats why i was asking, i heard the buffer thing
17:51:14BigBambiHilikus: That will make it possible to do in the correct way, now someone just has to do it
17:51:14krazykitHilikus, the primary concern right now, though, seems to be doing bugfixing for the Metadata on Buffer commit
17:52:21 Join ptw419 [0] (
17:53:34ShagnarBigBambi no, not yet
17:53:45BigBambiShagnar: Well that would seem a goos first step
17:53:46 Join toffe82 [0] (
17:54:29*Caleb needs to get his ipod from his gf
17:56:08 Part zlp
17:56:57ShagnarBigBambi you're right
17:57:02Shagnari just did so... but:
17:57:02ShagnarTWiki detected an internal error - please check your TWiki logs and webserver logs for more information.
17:57:03Shagnarillegal characters in file name component 'TillEichinger.2265176128 of filename path'TillEichinger.2265176128
17:57:20 Quit eigma (Read error: 113 (No route to host))
17:57:24Shagnarthe link in the email just worke
17:57:41BigBambiWhat is your wiki name, I need to give you write access
17:58:00Shagnarthanks for helping :)
17:58:00BigBambiOne mo
17:58:05bertrikamiconn, in powermgmt-e200.c there's a multiplication by BATTERY_SCALE_FACTOR and then a shift, can't we make this simpler by simply multiplying by 5?
17:58:45BigBambiShagnar: OK, you should have write access now
17:59:24Shagnarokay, thanks. where can i start now? didn't see a link or something...
18:00:09BigBambiIf you type the name of the new page into the go box, it won't find it and give you options to create it
18:00:15 Quit jumpatrain (Read error: 104 (Connection reset by peer))
18:00:18BigBambiHave a look at for naming conventions
18:00:33Shagnarvery good, thank you once again!
18:00:35BigBambiAnd then I'd probably link it to that page as well
18:00:38BigBambino problem
18:00:42 Join eigma [0] (n=cat@
18:01:19 Part Hilikus
18:06:44 Quit illissius` (Read error: 110 (Connection timed out))
18:08:36 Join bluey [0] (
18:09:04 Quit Thundercloud (Remote closed the connection)
18:09:15 Join zicho [0] (
18:17:20 Join RoC_MasterMind [0] (
18:17:41amiconnThe sansa OF uses libtomcrypt
18:18:02 Join Frazz [0] (
18:19:53 Join roolku [0] (
18:21:19 Quit spiorf (Remote closed the connection)
18:21:38 Join donutman25 [0] (n=chatzill@
18:22:21 Join saratoga [0] (i=41becb3b@gateway/web/cgi-irc/
18:23:15 Join midgey [0] (
18:24:11saratogaamiconn: is there any documentation showing all the known GPIO functions and their bits?
18:24:42amiconnIn fact we should document everything we found
18:25:06 Quit Shagnar (Read error: 110 (Connection timed out))
18:25:13amiconnsaratoga: Hmm, did you fix & update your G5 treble/bass patch
18:26:24saratogaamiconn: wrong guy?
18:26:51saratogai only have a sansa + 3G Ipod
18:27:03amiconnHmm, then that was safetydan
18:27:10amiconnBut he doesn't have a G5 either
18:27:25saratogaso theres no good resource for pin assignments on the PP players?
18:28:45amiconnFor one PP itself is secret, and then the port pin assignments are target dependent
18:29:12amiconnThose we know are sort of documented - in the rockbox code
18:29:43 Join webguest00 [0] (i=4a469ecd@gateway/web/cgi-irc/
18:30:06saratogathats what i figured
18:30:36webguest00i'm using an e200 with rockbox and when i play an mpeg video, the sound moves very quickly and the video runs nomal. How do i fix this?
18:31:59saratogawhat sample rate is the audio in your video?
18:34:56 Quit bluey (Read error: 110 (Connection timed out))
18:35:19 Part Heinz
18:35:25webguest00how do i figure that out?
18:35:55 Quit RoC_MasterMind ("Leaving")
18:36:14BigBambiat webguest00
18:36:54webguest00on my desktop, the video plays normal but after i transfer it to the player, then it start to speed up
18:37:06BigBambiwebguest00: Are you in windows?
18:37:16saratogaprobably because the samplerate isn't 44100hz
18:38:00webguest00bigbambi: yes
18:38:26BigBambiwebguest00: 1) You should have specified the audio sample rate when you converted the video, 2) to find out what it is try virtualdub (ask google how to get it and how to use, it is off topic here)
18:39:56webguest00okay, thank you
18:42:13 Quit EnterUserName (Read error: 104 (Connection reset by peer))
18:42:42saratogai think you can just click on file properties in windows to see sample rate
18:43:02 Join maxkelley_ [0] (
18:43:06BigBambiOh, OK, that'd be easier
18:43:12 Join EnterUserName [0] (
18:44:30 Quit maxkelley (Connection reset by peer)
18:44:57webguest00i right-clicked the video file and hit properties but it only showed type of file, location, and size.
18:49:16WalterEgowebguest - When you right-click video and hit properties, there are two tabs to the dialog box. General (the one you saw) and Summary. Hit summary and there's your audio samplerate.
18:50:11webguest00there is only one tab for me and it is general
18:51:00 Quit saratoga ("CGI:IRC")
18:51:06WalterEgo... oh.
18:55:55 Join petur [0] (
18:56:36 Join Llorean [0] (n=llorean@
18:57:12 Quit idnar (Nick collision from services.)
18:57:15 Join idnar_ [0] (n=mithrand@unaffiliated/idnar)
18:57:30 Quit courtc (Read error: 104 (Connection reset by peer))
18:57:32 Join MethoS- [0] (
18:57:33 Join Thundercloud [0] (
18:57:34 Join courtc [0] (
19:00:11 Quit Weiss (Read error: 104 (Connection reset by peer))
19:01:53 Quit webguest00 ("CGI:IRC")
19:02:16 Join ilgufo [0] (
19:03:28 Quit WalterEgo ()
19:07:12 Quit Thundercloud (Remote closed the connection)
19:08:02 Join kugel [0] (i=kugel@unaffiliated/kugel)
19:08:30emeralddhere's a fun one, how do you get the battery cover off on a c250?
19:10:12 Join Thundercloud [0] (
19:13:06 Quit Thundercloud (Read error: 104 (Connection reset by peer))
19:13:34 Join Thundercloud [0] (
19:14:49***Saving seen data "./dancer.seen"
19:17:46*linuxstb passes emeraldd a hammer
19:18:16emeralddI want to be able to put it back on again ;)
19:18:48*linuxstb passes emeraldd some duct tape
19:19:02 Join Shagnar [0] (
19:19:18lostlogiclinuxstb: I'm trying once more to make buffer fds 'normal'
19:20:00emeralddwhen is anything normal?
19:20:26Shagnarhey BigBambi, ive just added the information I got
19:20:33Shagnar see
19:20:37Shagnarcould you add that to the "future ports" page at ?
19:20:40BigBambiShagnar: OK, cool
19:21:30linuxstblostlogic: It wasn't as straightforward as it sounded?
19:22:11*linuxstb finishes a pre-HDD poweroff commit battery benchmark on his ipod Color - 6h 09m...
19:22:17BigBambiShagnar: done
19:22:21 Quit MethoS- (Read error: 110 (Connection timed out))
19:23:35Shagnarthanks, BigBambi
19:23:42BigBambiShagnar: Now you just need to find some interested parties to do the rest of
19:23:45BigBambiNo problem
19:25:44 Quit illissius- (Read error: 110 (Connection timed out))
19:25:59Shagnaryes, i think this could be a problem since i don't know any people for this task
19:26:26BigBambiBest idea is to try and drum up support on a fan site such as misticriver (in the case of iriver)
19:31:31Shagnarok i'll try that
19:33:30lostlogiclinuxstb: no −− due to the way tracks were cleared −− hids were _zeroed_ to mark them cleared, often by a memset.
19:33:32 Join bluey [0] (
19:33:43 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
19:34:02linuxstblostlogic: Hmm, so do you still think it's a good idea to make them "normal" ?
19:34:21emeralddI'm thinking that the adc-target.h values for the c200 are wrong,
19:34:48emeralddnot sure what they should be but the values don't seem to line up with definitions
19:34:59lostlogiclinuxstb: yes, I do.
19:35:21lostlogicI think I have it worked out properly now −− the memset only made sense because the track struct used to contain an mp3_info
19:36:00linuxstbOK, so you're removing the memset?
19:39:01pixelmaemeraldd: and try a few times. It's very hard on mine too, but eventually it'll work ;)
19:39:36 Quit GodEater (Read error: 104 (Connection reset by peer))
19:39:46 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
19:40:50lostlogicNico_P: how do you feel about converting handle_ids to have 0 as a valid handle?
19:40:56lostlogic(I have the patch done and running on my ipod)
19:41:11Nico_Plostlogic: wasn't that in one of your commits?
19:41:16Nico_Pwhich you then reverted
19:41:23lostlogicNico_P: had to be reverted because I fscked it up that time
19:41:39lostlogicforgot about the memsets used to clear track_info structs
19:41:47 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
19:41:48Nico_PI have nothing against it, but is it useful?
19:41:48lostlogicbut those memsets were put in because track_info contained mp3_info
19:42:09Nico_Pit's nice to be able to memset the track info structs to clear them
19:42:42lostlogicit's more consistent with fds and lets checks for errors on an hid be combined with is valid checks
19:42:48lostlogicwhich is why it's the standard for fds
19:42:54lostlogic(at least that's why afaik
19:43:28*lostlogic posts patch now that it works, changes codecs and rebuffers
19:45:01Nico_PI still need to catch up with all your recent commits :p
19:45:24*Nico_P 's vacation is over, so more time to rockbox
19:46:34lostlogicNico_P: hehe :) yeah... I've abused the hell out of it
19:46:59 Join mg [0] (i=d8ef51d5@gateway/web/cgi-irc/
19:47:03lostlogicbut it's much tighter now I think... 8077 forced me to beat like 10 gremlins to death.
19:47:16 Join linuxstb_ [0] (n=chatzill@rockbox/developer/linuxstb)
19:48:15lostlogicNico_P: the big remaining bugs are gapless (how do we strip tags correctly in the new world?? in playback?) and runtime data
19:48:23lostlogicneither of which matter to me so I'm not as inclined to fight with them
19:48:53Nico_PI'll have a close look to the changes and then try to squash those
19:49:34 Join maxkelley [0] (
19:52:25 Quit mg ("CGI:IRC (Ping timeout)")
19:53:11Nico_Plostlogic: anyway, thanks a lot for doing all this
19:53:20 Join mg [0] (i=d8ef51d5@gateway/web/cgi-irc/
19:53:41 Quit bluey (Read error: 104 (Connection reset by peer))
19:53:43 Quit Thundercloud (Read error: 104 (Connection reset by peer))
19:54:03lostlogicNico_P: it's my joy, I was just thrilled you committed MoB to give me something to chew on :)
19:54:03 Quit mg (Client Quit)
19:54:23 Join mg [0] (i=d8ef51d5@gateway/web/cgi-irc/
19:56:07 Part pixelma
19:58:11 Quit maxkelley (Connection reset by peer)
19:58:17 Join maxkelle1_ [0] (
19:58:21 Nick maxkelle1_ is now known as maxkelley (
19:59:10 Quit ilgufo (Remote closed the connection)
19:59:28 Quit maxkelley_ (Connection reset by peer)
19:59:49 Quit mg ("CGI:IRC (Ping timeout)")
20:01:51 Nick idnar_ is now known as idnar (n=mithrand@unaffiliated/idnar)
20:02:42 Nick fxb__ is now known as fxb (
20:02:46 Join ilgufo [0] (
20:04:44eigmaanyone know how to get gdb to fill an area of memory?
20:12:45lostlogicInteresting −− my changes apparently make ARM happy and coldfire unhappy
20:13:09 Quit roolku ()
20:13:27 Join bobbo [0] (
20:13:33linuxstb_lostlogic: Are there any "char"s involved?
20:14:28bobbois anyone here good with the Sansa e200 Series?
20:14:29 Quit J3TC- (".•«UPP»•.")
20:14:33lostlogica couple of bools, which I imagine have the same implications
20:15:08linuxstb_I mention chars, because they're unsigned on ARM, and signed on Coldfire (or vice-versa, I forget...)
20:15:24 Quit dandin1 ()
20:15:36lostlogicoh, I had guessed it had to do with the 1byte size
20:17:10 Join bluey [0] (
20:17:41 Quit Siku (Read error: 145 (Connection timed out))
20:18:55 Quit bobbo ("Lost terminal")
20:19:50 Join bobbo [0] (
20:20:56bobboI have a Rockboxed Sansa e280, which i managed to cause FS damage to. It still boots but i lost all the music, but Nautilus says the space is still being taken up, Does anyone know how to fix this?
20:21:36bobbowill give those a go
20:21:57Avehoo careful
20:22:07Avemkdosfs will wipe everything on the block device
20:22:13Avelike ANY mkfs
20:22:43Averead manpage and lets be careful out there
20:22:55bobboso if i make a copy of the rockbox files and the original firmware, mkdosfs it, then put the files back on, it should be back to normal?
20:22:58scorchein other words, fsck first ;)
20:23:15 Part bobbo (""IRSSI Screwed, will be back"")
20:23:27 Join bobbo [0] (
20:23:36scorchewell, if you lost all the music, why bother copying files?
20:23:39 Part bluey ("Leaving")
20:25:32bobboscorche: to keep my rockbox install/themes/settings etc.
20:25:56bobbofsck has come up with errors, which action should i take? (1,2 or 3?)
20:31:17 Join stripwax [0] (
20:31:20BigBambibobbo; perhaps pastebin the message?
20:31:42bobbowill do
20:34:16BigBambibobbo: I don't see anything
20:34:36bobbodamn, im in a tty, will switch to gui, be back in a sec
20:34:42 Part bobbo
20:35:32 Join bobbo [0] (
20:35:46bobbothat should be it
20:36:17BigBambibobbo: OK, I see it now but I don't know the answer
20:36:19BigBambiSorry :)
20:36:25bobbono problem
20:36:31 Join n1s [0] (
20:37:26eigmait's just the disk label that's different
20:37:28eigmatry #3
20:38:52bobboim guesing ill just have to go the way of backing up and repartitioning, time for brickage :D
20:38:57 Join spiorf [0] (
20:40:47bobbomy MP3 player is now full of fsck*.recs :S
20:41:40SlasheriNico_P: hmm, the database statistics still seems to be broken
20:42:02lostlogicSlasheri: it's on flyspray, and is Nico's next 'victim'
20:42:09Slasheriah, good :)
20:42:12lostlogic(that and gapless issues)
20:43:01lostlogicthe problem with saving the stats has to do with figuring out when to trigger a copy-back into the buffer handle so that the unbuffer callback can write them (afaik)
20:43:07lostlogicI have no idea about the problem with display
20:44:43bobbothanks everyone thats help, god knows whats happened but its fixed now :D
20:44:45 Quit courtc (Read error: 104 (Connection reset by peer))
20:44:54 Join courtc [0] (
20:44:54 Part Llorean
20:44:57BigBambibobbo: Goodey
20:45:10Slasheriit's enough that the track info is only valid while calling the database unbuffer event
20:45:19bobbocya guys
20:45:21 Part bobbo ("Leaving")
20:45:42Slasheriatm it seems that track data from db isn't loaded either (and then db engine refuses to process unbuffer callbacks also)
20:46:09lostlogicah, I disable database on my builds so I have no idea about the loading
20:46:38lostlogicbut I know that the unbuffer callback breaking has to do with the elapsed time not being copied back to the mp3_info for that track which is on buffer
20:47:17Slasheriah, then it can't work
20:47:24 Join roolku [0] (
20:47:35roolkuhi Slasheri :)
20:47:45Slasheriroolku \o
20:48:19roolkuSlasheri: what do you think about the idea to fill in the runtime info from get_metadata (FS #8082)
20:48:54Slasherihmm, i will check that
20:51:53Slasheriroolku: that looks like a good idea, i don't think it will cause troubles
20:52:35roolkuit's a shame it causes a special case for HWCODEC though
20:52:40Slasherias long as tagcache generator doesn't lock up :)
20:52:53Slasherihmm, it does?
20:53:15roolkuas far as I know HWCODEC doesn't use get_metadata
20:53:37roolkuso I left the callback function for HWCODEC
20:53:58Slasherihmm, really? if it still uses direct mp3 metadata callbacks, i think it should be converted to use get_metadata
20:54:10Slasheribecause that is compiled for hwcodec too, iirc
20:54:27roolkuhm, maybe I am mistaken
20:54:45Slasheriwell, i don't know about that :)
20:54:55LearI think buffering in mpeg.c doesn't use get_metadata, but the tagcache would still use it.
20:55:02roolkuI thought it only used the function in id3.c
20:57:06Slasheriyep, it has a few mp3info callbacks
20:57:17Slasheribut it seems quite straightforward to convert those to use get_metadata
20:57:27Slasheriall the structs are the same
20:59:12roolkubut that would mean that firmware code calls application code, wouldn't it? I thought this was a no-no?
21:00:30Slasheriah, hmm. that is true. But it will be calling application code anyway, through those callbacks
21:01:13LearYes, but callbacks are another thing. If they are registered by app code, that's ok.
21:01:16roolkuyes, I never quite got this policy - that it was okay to use callbacks
21:02:08Slasherione solution would be to create a get_metadata callback for the hwcodec
21:02:23LearSometimes firmware code has to call app code, no way around it. Callbacks avoid direct dependencies.
21:02:33roolkuit seems a bit longwinded to introduce callbacks for get_metadata, but yes, it would be a solution
21:02:34Slasherithat is probably because firmware part should be independent from app code while compiling
21:02:47 Join Thundercloud [0] (
21:04:24Calcipherah yes, slept most of the day away, good morning afternoon, evening, night
21:04:27 Quit stripwax (Read error: 110 (Connection timed out))
21:05:30 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
21:05:38 Quit advcomp2019 (Nick collision from services.)
21:05:42 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
21:06:50roolkuconsidering that the on_buffer call backs for HWCODEC are uneffected by the recent MOB changes and work (?) wouldn't be the solution from FS #8082 be the least intrusive?
21:07:09*roolku is a little hesitant to make HWCODEC changes as he can't test them
21:07:25 Quit darkless ("Leaving")
21:12:35 Quit Lear ("ChatZilla [Firefox]")
21:13:23 Quit TMM ("Ex-Chat")
21:13:31 Join nanok [0] (n=nanok@
21:14:50***Saving seen data "./dancer.seen"
21:14:55emeralddany pointers on disassembling firmware in mi4 format?
21:18:05linuxstb_Decrypt it with mi4code, then remove the first 512 bytes (e.g. dd if=decrypted.mi4 of=firmware.bin skip=1), then use "arm-elf-objdump -m arm -b binary -D firmware.bin"
21:18:33 Quit n1s ()
21:20:17 Quit MajorC ()
21:20:46emeralddthe objdump gives me a seg fault :(
21:23:07linuxstb_Where did you get it from?
21:23:52emeralddfrom gentoo using the crossdev script
21:24:12emeralddbinutils 2.18
21:26:25linuxstb_You could also try the arm disassembler in Rockbox SVN - utils/disassembler/arm/
21:27:35 Quit bertrik ("bye")
21:27:41 Join lespea [0] (
21:28:32emeralddspotted that, are there any issues with alignment and non-executable data?
21:30:14linuxstb_What kind of alignment issues would there be?
21:30:54emeraldddoes it properly handle string constants and the like?
21:31:22linuxstb_No, I don't think it deals with strings at all.
21:31:30linuxstb_But feel free to improve it ;)
21:32:22emeralddjust so long as constants don't upset the disassembly of instructions
21:34:02 Join bertrik [0] (
21:34:53 Quit emeraldd (Remote closed the connection)
21:36:29linuxstb_All ARM instructions are 32-bit, so it's not an issue. Both the Rockbox dissassembler and objdump just process 32 bits at a time, starting at the beginning of the file.
21:37:22 Quit lespea (Read error: 104 (Connection reset by peer))
21:44:58 Quit amiconn (" Try HydraIRC -> <-")
21:48:32 Quit midgey ()
21:48:53 Join midgey [0] (
21:48:57 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:49:10 Nick ilgufo is now known as gufaway (
21:52:52 Join lespea [0] (
21:54:03 Quit linuxstb_ ("ChatZilla [Firefox]")
21:56:53 Join Guile`` [0] (n=Guile@
21:58:48Calcipherwhat do you guys think of having an option under voice section that allows using a quick press of the rec button to read the current focused item to the user, when in menus, file browsing, wps and where ever voice reading functionality can be applied?
21:58:56CalcipherI was looking at the manual and noticed the mahor function of the rec button is a long press to get to record mode
21:59:16Calcipheras for the sansa e200 series players for example
22:00:53BigBambiCalcipher: What would you do for players without that button? Why not just have it read the currently selected item?
22:01:06 Quit Shagnar ()
22:05:10Calcipherwell I don't know other targets since I only own a sansa e280r, but being able to be prompted where you are on the player is a huge improvement, because using the interface without only voice output, this would add a more intuitive way of knowing whats going on
22:05:43Calcipherwithout having to change your focus or leave a menu or screen like the wps screen
22:06:23BigBambiBut voice tells you where you are now (or so I thought)
22:06:36CalcipherI have a friend who is a blind rb user and he thought it would be a very helpful option as wel
22:06:54BigBambiSo how would that work in WPS? There sin't a focussed item
22:07:29Calcipheryeah I know, so in wps the voice reading ability would need to be added
22:07:48BigBambiBut how would you know what to voice?
22:07:51CalcipherI guess thats the most important section
22:08:01 Join Domonoky_ [0] (
22:08:33 Join Shagnar [0] (
22:08:51Shagnardoes anybody know what the "Sunny 546" oscillator chip is?
22:09:01Calcipherwell ddalton suggested a list view that you could scroll through and have different fields read from each
22:09:17BigBambiHow would that work?
22:09:25CalcipherI guess with a tag like "By : artist"
22:09:56BigBambiSo add a complete duplicate set of tag to the WPS parsing as essentially voice tags?
22:10:01Calcipherand next line you can scroll to "from : album:
22:10:36Calcipherwell are wps read at all?
22:10:57BigBambiNot that I know of, but I don't use voice
22:11:06CalcipherI just mean if its going to read an item, it should indicate what field its reading from
22:11:11 Quit gufaway ("So Long, and Thanks For All the Fish -")
22:11:38BigBambiAgain, I come back to how
22:11:53CalcipherGuessing the most intuitive first available item to be read would be simply the current song title
22:12:12BigBambiThis is almost getting back to the reading out of order argument
22:12:31BigBambiVoice should always voice in the same order as on screen, otherwise it is confusing
22:12:42Calcipherwell as a menu that can be accessed from the wps?
22:12:48BigBambithis leads to a problem with WPS, as graphical info can't easily be voiced
22:13:25BigBambiThere is already the tag info screen
22:13:27CalcipherI see so just have it read whatever order the wps displays its info
22:13:34BigBambiFrom the context menu
22:13:51BigBambiWhat about conditionals in that case?
22:14:00BigBambiIs the tag screen voiced?
22:14:17Calcipherlet me check
22:16:21BigBambiSo the solution to me seems to be voice that screen - much easier than than trying to do the WPS
22:16:36BigBambiThen all that is left is track position etc
22:17:17BigBambiPerhaps that screen could be expanded (or another devised) that had playlist position, track position etc
22:17:42BigBambiAs I would imagine that things like conditionals would make the WPS extremely hard to voice
22:18:49Calcipherit just seemed to me having a preset items to include for reading, regardless of wps, if thats easier
22:19:10BigBambiI would not want to voice things independant of what is on screen
22:19:25BigBambiThere are plenty of sighted people who use voice and that would be plain confusing
22:19:26Calcipherso if you selected that the voice wps function would only include title artist album next
22:19:41Calcipherthats all it would ever read on the quick key press if enabled
22:20:21CalcipherI see
22:20:42CalcipherI was just looking for a one button press way to hear whats on the wps
22:21:00 Join Zagor_ [0] (
22:21:10BigBambiWell feel free to sketch out a plan. You can find out what keys are used and not on the other players via the manual. I would prefer a screen accessible via the context menu in the WPS, but I'm not in charge :)
22:21:24BigBambiSpeaking of people in charge, welcome Zagor_
22:22:12*Calcipher hails lord Zagor_ , master of universal serial bus
22:22:12 Nick Zagor_ is now known as Zagor (
22:22:40Zagorheh, if only...
22:23:28 Join barrywardell [0] (
22:24:04 Quit Frazz ("Leaving")
22:24:08Calcipher•BigBambi• do you see what I'm trying to achieve atleast? simply a quick button press read of whats playing
22:24:49Calcipherddalton says that too much info being read is no good, and I agree, but wouldn't the amount of info depend on the wps your using?
22:24:59BigBambiCalcipher: Sure, I see what you are getting out. I'm just playing devils advocate. Plenty of players don't have the free button to press, and I don't like voicing things not on screen
22:25:33BigBambiCalcipher: Yeah, but how would you deal with a conditional that displayed next track then next artist for 2 seconds each?
22:25:52 Quit Domonoky (Read error: 110 (Connection timed out))
22:26:15BigBambiHence I would prefer something like a screen accesible from WPS quick screen that just listed e.g. track, artist, position, whatever that could be read like a normal screen
22:27:22CalcipherI guess treat the conditionals as if they were on an ordered list? I'm trying here, I'm not familiar with how these functions work in the first place, thats why I welcome constructive feed back
22:27:52BigBambiCalcipher: It is all possible I guess, but difficult
22:28:09Calcipherthat sounds good too
22:28:44BigBambiI can see the advantage of voicing what is on the WPS, but jut wonder if the aim could be achieved more easily on a slightly different tack
22:29:09Calcipherso if it was a different screen that was accessible, I guess it can be possible to have the items to be included or excluded on this screen be configurable, right?
22:29:18BigBambiAnything is possible :)
22:29:37BigBambiYou would just have to persuade the powers that be that it is a good idea :)
22:29:43Calcipherso it basically be a list of fields, that you set in settings somewhere
22:30:21BigBambiThat would be my idea on the spot with no thinking about it. A blind WPS as it were
22:31:10BigBambiYou could write a BWPS that on each line just had e.g. track, artist, position, .... that then displays and is read like a list
22:31:11preglowNico_P: why isn'
22:31:15Calcipherwell I have been discussing different ways of achieving some kind accessible wps like interface since I made that feature request, hopefully any heads interested will see some of the good ideas mentioned on the logs
22:31:26preglowNico_P: why isn't metadata stripped from files in the buffer anymore?
22:31:34Calcipherand they're more qualified on deciding what works best
22:31:39preglowNico_P: that will indeed break gapless for mp3
22:32:27Nico_Ppreglow: because the function that did it wasn't trivial to adapt to the buffering code and a FIXME in it said it should be moved to metadata.c and made more generic
22:32:54Nico_PI didn't see any huge breakage from not porting it and then kinda forgot about the issue
22:33:21 Join webguest15 [0] (i=d559880c@gateway/web/cgi-irc/
22:33:41 Join cynic [0] (
22:33:42CalcipherWell its out of my hands, I'm going to the gym and enjoy my rb'ed Sansa, thanks all of you
22:34:49webguest15Hello Zagor! Keep up the good work! Could you update with a fresh patch?
22:35:06Zagoryeah, I was just about to do that
22:35:07cynicis there any news on rochbox on the ipod classic?
22:35:12webguest15zagor: nice
22:35:45scorchecynic: no
22:36:54 Quit cynic (Client Quit)
22:37:17 Join cynic [0] (
22:38:21preglowZagor: so, how's progress?
22:38:42 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
22:39:09Zagornothing new since friday, been busy over the weekend
22:42:04Nico_Ppreglow: any ideas on how to get rid of the tags?
22:42:31preglowNico_P: not really, no, don't even know how the old code worked
22:42:41preglowNico_P: but i don't see how it can be a difficult problem
22:43:10Nico_Pit used to seek around in the file and wind the buffer backwards... it's not very difficult and could quite easily be added in buffering.c
22:43:24Nico_Pbut IMHO it's not the buffering code's work
22:43:28preglowit's not
22:43:33preglowit surely is metadata.c work, yes
22:46:37 Part Domonoky_
22:48:18 Quit merbanan (Remote closed the connection)
22:48:55 Quit petur ("*plop*")
22:52:09lostlogicNico_P: I was thinking that when playback starts a new track (ie when the codec requests the first packet from that track) playback could fiddle with the buffer
22:53:44 Join safetydan [0] (n=safetyda@rockbox/developer/safetydan)
22:54:59Nico_Plostlogic: how? that doesn't sound safe to me
22:55:10*amiconn wonders whether the ipod hdd power control is worth a major changes entry
22:56:36 Join Jon-Kha_ [0] (
22:56:40 Quit Jon-Kha_ (Remote closed the connection)
22:56:40 Quit Jon-Kha (Read error: 131 (Connection reset by peer))
22:56:54lostlogicNico_P: audio_update_trackinfo seems like the place to do it to me. The audio thread could request a pointer to however much data it needs to determine teh tag size, then bufadvance the correct final amount
22:57:30*Buschel still runs his battery bench, his ipod video did never run this long :o)
22:57:40lostlogicI forget, did it need to fiddle on both the front and the tail or just the front or just the tail?
22:57:50midgeyamiconn: i'd say yes
22:57:58Nico_Plostlogic: I think both but I'm not sure
22:59:05amiconnBuschel: How long so far? :)
22:59:29Buschelamiconn: hits 10h in 5mins
22:59:54amiconnIs this a single or dual platter model?
22:59:56 Quit mf0102 ("Verlassend")
23:00:16 Quit tedrock (Client Quit)
23:00:20Buschelamiconn: and still at 14%... it's a 5.5G 30GB model with a standard 400mAh battery
23:00:23barrywardellamiconn: i noticed in the logs you said the H10 has two bits in GPO32 set. My H10 doesn't, it only has 1 bit
23:00:45amiconnbarrywardell: Yours is the big one?
23:00:58lostlogicNico_P: ugh
23:01:09 Join SoapSud [0] (
23:01:22amiconnbarrywardell: My 6GB has bit 24 and bit 29 set
23:01:43barrywardellyeah, my 20GB just has bit 29.
23:01:46 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
23:02:17amiconnMaybe tuner power is on one of the GPO32 pins?
23:02:20barrywardellI haven't heard of any other differences in the GPIO between H10 versions though
23:02:39*amiconn doesn't remember what the problem with the tuner was
23:02:56lostlogicyeah, I was just lookin' at it −− that's trimming off of the end only
23:02:57 Quit Thundercloud (Read error: 104 (Connection reset by peer))
23:03:06 Join tedrock [0] (
23:03:12lostlogicso i was backwards in my rememberance
23:03:16barrywardellamiconn: I'll try bit 24.
23:03:54amiconnBuschel: Whoa, singnificantly more than what was measured on 30GB Videos so far :)
23:04:18amiconnBuschel: What format(s) and bitrate(s) are you using?
23:05:21preglowmusepack :D
23:05:39Buschelamiconn: i guess it will shut down at ~10:40h, so 40mins to go. former runtime was 8:40h maximum −− using mpc @179kbps average
23:05:57Buschelpreglow: exactly :P
23:06:03amiconnVery nice extra 2 hours :)
23:06:18*amiconn should bench the 80GB at some point
23:06:52amiconnThe drawback is that I often have ideas which need testing on units which are currently running a benchmark ;\
23:07:05Buschelamiconn: battery is of 580mAh on 80G, correct? so, it should have ~3h extra time
23:07:08 Quit SoapSud ("ChatZilla [Firefox]")
23:07:41amiconn11 hours seem to have been common on 60/80GB so far, and ~7.5 hours on 30GB
23:07:54amiconn(using less exotic formats than musepack) ;)
23:08:30lostlogichmm, now I'm confused... how can the decoder correctly ignore the id3v2 tag stored at the beginning of the file but not the id3v1 or ape tags at the end?
23:09:23Buschelamiconn: there you can see what's happening if you're using the wrong format ;)
23:09:55amiconnlostlogic: The metadata reader doesn't know about an id3v1 tag at the end *if* id3v2 has priority and exists
23:10:10amiconnThen it doesn't check for id3v1, saving a (costly) seek
23:11:31lostlogicamiconn: but do the codecs even use the metadata reader? I'm trying to figure out how when the codec fires up it knows to skip past the id3v2 data at the beginning (up to 256mb of it according tot he spec)
23:12:15amiconnAfaik the track loader doesn't load the id3v2 data, as the start of the actual audio data in the file is already known
23:13:11Buschelhmm, i must lower my estimated end of runtime... it just loaded via HDD and drained to battery :/ -> maximum 20mins to go
23:13:39lostlogicamiconn: "track loader" ?
23:14:52***Saving seen data "./dancer.seen"
23:15:15Shagnaranybody here who has some information about samsung ICs?
23:16:35Zagorwebguest15: new patch uploaded to tracker
23:17:08BagderShagnar: samsung docs is mostly not public
23:17:26Shagnaroh noes :-x that means.. no chance of getting any information?
23:17:47BagderShagnar: welcome to the world of Rockbox ;-)
23:17:52Shagnarwhat would also mean no chance to port RBX on a samsung based player?
23:18:02Bagderwe have ports to numerous players without docs
23:18:13Shagnarvery good
23:18:13Bagderit just means it is harder
23:18:21Bagderand more work
23:18:37Shagnarokay... yes that's the point. but where to start then?
23:18:39amiconnYeah, RE is time consuming
23:18:48*amiconn currently wades through the c200 OF
23:19:01BagderShagnar: reverse engineering, disassembly of the OF etc
23:19:16 Join tihoc4n [0] (
23:19:39Bagderbut really, if you need to ask you're probably not the man for it
23:19:46amiconnI noticed that it's really helpful that we have so many different PP502x targets now. It helps to get a better understanding of the chip
23:19:47Shagnarwhat does OF mean? original firmware?
23:19:57 Join mguay [0] (i=d8ef51d5@gateway/web/cgi-irc/
23:20:05BagderShagnar: yes
23:20:25Shagnarhow do you mean that.. if i need to ask? i don't know how to learn it else.. can you give me a hint?
23:20:39barrywardellamiconn: no luck with bit 24 and the h10 radio :(
23:20:41BagderShagnar: a hint for what?
23:21:00Shagnarstarting with RE, ...
23:21:07amiconnbarrywardell: Looks like a H10 OF disassembling session is necessary...
23:21:29BagderShagnar: you need to figure out how the hardware works. Without docs, the only way to do that is to investigate the hw and to disassemble code
23:21:49barrywardellamiconn: yeah, I suppose looking for gpiob bit 2 is a good place to start
23:22:17amiconnIirc the H10 OF contains a bunch of debug strings
23:22:30Shagnarokay. what means it's really to much... but who could make that?
23:22:36amiconnThat should help tracing back where the corresponding code is
23:22:47BagderShagnar: a skilled dedicated person
23:22:52Bagderor several
23:22:54bertrikdon't you worry about legal issues around picking apart the original firmware?
23:23:01Bagderbertrik: no
23:23:07Bagderthere are none in most countries
23:23:22Bagderfor this purpose anyway
23:23:36mguayNico_P: Still having that problem with audio stopping on the Sansa if you need someone to test. It's really odd how it choose certain files to lock-up on.
23:23:51lostlogicmguay: no you aren't, I fixed it, you're a liar.
23:24:00bertrikmguay: aaaaaargh, thought that was fixed
23:24:05Calebheh seems most my songs have crappy tags
23:24:06lostlogicmguay: ... ... anything particular about the files that are still breaking?
23:24:09Shagnaryes okay... so that's where my work ends. I can't go any further, all i got is written down on
23:25:03*amiconn learned most of what he knows about RE during his rockbox work
23:25:30barrywardellamiconn: there are some 'Tuner' strings in there
23:26:15*Caleb wonders how he can get his tags to show right
23:26:21*lostlogic pokes mguay <−− I didn't mean it, you're not a liar...
23:26:33lostlogicCaleb: use foobar2000 or easytag or some other such program to make them consistent
23:26:59Calebwouldnt i have to manually edit them with easytag?
23:27:25 Nick parafin is now known as parafin|away (
23:27:44lostlogicCaleb: often not −− if they have 12314134123 sets of tags and filenames you can usually train it to coallesce all of the junk info into a consistent 1 set of tags + filenames + directories withou ttoo much hassle
23:28:26 Quit Shagnar ()
23:28:39Calebahh ok
23:28:49Calebman i love pacman
23:28:50lostlogicmguay: really really, I'm curious what problem you're still having... I really really don't think you're a liar!!
23:29:00 Quit mguay ("CGI:IRC (Ping timeout)")
23:29:40lostlogicwell that explains it
23:30:39 Join random_desu_is_s [0] (
23:33:14SoapBuschel: don't fret those last 15 drops of battery running away. I really do think the 90-10% discharge curve is much more telling than the 100-0.
23:34:26 Join mguay [0] (i=d8ef51d5@gateway/web/cgi-irc/
23:34:26 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:34:45lostlogicmguay: what's special abou tthe files that still freeze... and I assume you are running absolute latest build?
23:34:58Buschelsoap: but i want to have the _real_ runtime in total. for estimation of current consumption the first2 or 3 hours are enough
23:35:18Buschelbtw, ipod did shut down now -> 10:28h runtime
23:35:46amiconnSoap: I don't think a 90-10 curve tells more. It's rather imprecise compared with the full 100-0
23:36:05 Join random_desu_is_s [0] (
23:36:47Soapamiconn: the 90-10 discharge time removes much of the variability caused by battery aging (which makes the peak voltage drop faster) and the +-5% dips and spikes caused by a late HDD spin-up.
23:37:22amiconnImho the opposite is true. How do you tell when you're at 90%, or 10%?
23:38:03amiconnIf you use the discharge curve, you put much more uncertainty on this than if you just let it run until it shuts down
23:38:04 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:38:13SoapI'm not sure I am following the question - but I think you are saying that the 90% and 10% voltage numbers are arbitrary and thus my reasoning is flawed?
23:38:47amiconnThe voltages aren't arbitrary, but they just represent the *average* battery at an *average* temperature
23:39:08SoapI follow so far.
23:39:12amiconnAnd the discharge curve is much more steep at theirs ends than somewhere inbetween
23:39:43SoapIt is that steep part of the curve I am trying to trim out. Isn't that non-linear part more noise than signal?
23:39:56 Quit mguay ("CGI:IRC (Ping timeout)")
23:40:07Buschelamiconn: so, at a result your HDD poweroff saves 8mA (for my usecase)
23:40:18amiconnStill not bad...
23:40:38Buschelno, it's really good progress!
23:40:39SoapIsn't it that part of the discharge curve which varies more by age, (last) charging conditions, and spikes in power consumption?
23:40:47amiconnThe typical saving from hdd poweroff obviously depends on the generation as well as on the size
23:40:59 Join random_desu_is_s [0] (
23:41:03amiconnOn my 2nd gen, the hdd draws ~15mA when in standby
23:41:15amiconnBut that's an MK2003GAH
23:41:42amiconn(and rockbox has hdd poweroff on 1st..3rd gen for some months now)
23:42:05*amiconn should really do a battery bench on mini 2nd gen
23:42:31amiconnI'm curious how much standby current this microdrive needed...
23:42:53Buschelamiconn: the MK3008GAL should draw 21mA in sleep and 36mA in standby by specification. it seems to be much better than that...
23:43:36amiconnWell, these are the worst-case values according to the datasheet I presume?
23:43:49amiconnI'd expect typical values to be better
23:45:13preglowwe removed the disk poweroff setting, yes?
23:45:14 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:45:25*lostlogic wishes mguay would 'stick' to the channel
23:45:46pixelmaroolku: in your last commit you mentioned something about the gather runtime feature on hwcodec. Did you actually take a look into the code or is that what you assumed? I'm wondering because it doesn't gather anything on my Ondio...
23:45:52linuxstblostlogic: (if you're still wondering) The codecs skip id3v2 tags at the start based on the id3->first_frame_offset value set by get_metadata()
23:45:55 Join random_desu_is_s [0] (
23:46:02amiconnBuschel: Specs for the MK2003GAH state 60mA in standby and 25mA in sleep mode
23:46:11lostlogiclinuxstb: ah, thanks much!
23:47:22 Quit random_desu_is_s (Connection reset by peer)
23:47:35Buschelamiconn: which drive is this?
23:48:01 Join random_desu_is_s [0] (
23:48:05preglowamiconn: any reason why buschel's two other lcd speedup patches aren't commited?
23:48:07Buschelamiconn: for 2nd gen?
23:48:09*amiconn just mentioned it 7 minutes ago
23:48:14 Quit BigBambi (Remote closed the connection)
23:48:57amiconnpreglow: The ine that asm'izes lcd_update completely is hiding too much imo. The other one considerably increases stack usage, and I'm undecided about it
23:49:08amiconnNeeds checking the video thread stack in mpegplayer
23:49:10lostlogiclinuxstb: that also tells me why some particular mp3s were pausing for people
23:49:11 Quit random_desu_is_s (Read error: 104 (Connection reset by peer))
23:49:25linuxstblostlogic: So what's changed wrt that?
23:49:26lostlogiclinuxstb: that request could easily be > guardbuf
23:49:36linuxstbAh, the seek to skip it?
23:49:44lostlogiclinuxstb: right
23:50:00pixelmaroolku: I understand that the commit is not related to it but I thought (hoped) that you could have seen something weird or so
23:50:16lostlogiclinuxstb: I'm definitely starting to feel like we should continue the move toward posix-esqua apis... wherein the read-esque functions would be more free to return not-the-full request
23:50:18 Quit jhMikeS (Nick collision from services.)
23:50:23preglowamiconn: by how much does it increase stack use?
23:50:24 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:50:33amiconn4 times
23:50:33linuxstblostlogic: Why does it matter if it's > guardbuf ?
23:50:38linuxstblostlogic: (if it's just a seek...)
23:51:02amiconnIn fact just 2 times against svn, because the chroma buffer calculation in svn is wrong
23:51:02lostlogiclinuxstb: well at least the first codec I looked at first does a read of tha tsize
23:51:06amiconn(but to the safe side)
23:51:12lostlogiclinuxstb: not sure why it does that jus tto throw it away but it does
23:51:18linuxstblostlogic: Ouch, that's a bug then IMO.
23:51:19 Quit eigma (Read error: 113 (No route to host))
23:51:32lostlogiclinuxstb: my inclination as well
23:51:35linuxstbAh, mpa.c ?
23:51:46lostlogicwhich explains why mp3 was freezing on people and not other codecs
23:51:58lostlogic(of course now t he buffering code is much more resiliant to obnoxious clients ;))
23:53:18 Quit webguest15 ("CGI:IRC")
23:53:39Buschelamiconn: we could just #ifdef both the byte-stack vs. the word-stack as well as the str/ldr vs. stm/ldm. if there any problems we can easily switch back...
23:53:56amiconnWe can always switch back thanks to svn
23:54:16 Quit Nico_P (Remote closed the connection)
23:54:26amiconnBut G5 lcd is not the only thing I'm busy with, so I can't check everything again and again :\
23:54:46Buscheli can guess that
23:55:01 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
23:55:48roolkupixelma: does it not gather the data or does it not display the results?
23:56:02 Join _Caleb_ [0] (
23:56:10pixelmahow could I check?
23:56:50 Quit Caleb (Nick collision from services.)
23:57:22roolkupixelma: you could export your database and have a look at the text file
23:57:28 Nick _Caleb_ is now known as Caleb (
23:58:08pixelmawill try
23:58:08roolkupixelma: or use one of the tagnavi menus that make use of runtime data
23:58:50pixelmathat's what I did last time, IIRC.

Previous day | Next day