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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2007-01-04

00:00:03linuxstbThere are two independent caches, one per core
00:00:15linuxstb(8KB code and data)
00:00:37Shawn_Kcache is used for variables?
00:01:04linuxstbcache is used to cache memory accesses. So if the COP writes to RAM, that will be cached and the main CPU won't see the change...
00:01:30*preglow hugs bus snooping
00:01:42*linuxstb looks around, but doesn't see any
00:02:27 Quit webguest93 ("CGI:IRC")
00:02:37Shawn_Kso how do you flush the cache?
00:03:04Shawn_Kor rather, update
00:03:23preglowit's not a viable way to do it
00:03:28preglowshared data should be in iram
00:03:35preglowand there should be as little of it as possible
00:03:52dan_a(IRAM is uncached)
00:04:04preglowwould be no point in caching it anyway
00:04:12Shawn_Kok, so how would I go about storing data in iram?
00:04:44linuxstbThere is only 48KB of it available to core Rockbox, so every byte is precious...
00:04:51dan_aDeclare the variable with IDATA_ATTR or IBSS_ATTR (if the data should be in BSS)
00:04:51 Join lavid [0] (
00:05:35linuxstbIn fact, is the core able to use the extra 32KB on the PP5021?
00:06:07dan_alinuxstb: treats all PP502Xs the same
00:06:09 Join gotthardt [0] (i=86868802@gateway/web/cgi-irc/
00:06:27linuxstbdan_a: I just noticed that..
00:06:27dan_aWe ought to change that at some point
00:06:53Bagderhey gotthardt, welcome aboard
00:06:55 Quit ender` (" Be wary of strong drink. It can make you shoot at tax collectors and miss. -- Robert A. Heinlein")
00:08:04linuxstbdan_a: BTW, I've got the suspicion that the failure of mpegplayer to reliably start the video thread is just the video thread freezing due to my sloppy code. I'll let you know.
00:08:53 Join webguest17 [0] (i=91740560@gateway/web/cgi-irc/
00:12:28 Quit x1jmp (Remote closed the connection)
00:15:44 Quit bluebrother ("nite.")
00:18:19 Join jlcox [0] (i=dad6250f@gateway/web/cgi-irc/
00:18:22 Quit WilliamC ("ChatZilla 0.9.77-2006121112 [Firefox]")
00:18:23 Quit nls_web ("CGI:IRC (EOF)")
00:18:56 Quit jlcox (Client Quit)
00:19:03 Quit lee-qid ("aufwiederbyebientotsayonara")
00:21:43Shawn_Kwow, after beginning to understand CO-P, it makes alot of sense
00:21:54Shawn_Know just to test these changes I made
00:22:06Shawn_Kif my iPod doesn't freeze of crash....sucess
00:22:08 Join hcs [0] (n=hcs@rockbox/contributor/hcs)
00:23:01Shawn_KI like saying CO-P
00:23:05Shawn_Kinstead of COP
00:23:22*dan_a gets triggered by COP, but not CO-P
00:23:29dewdudei put some 96/24 flac on my ipod last night...skipped like mad
00:24:00preglowdewdude: not exactly a surprise
00:24:02Shawn_Kit works
00:24:11Shawn_KI got the scroll thread put on COP
00:24:38dewdudepreglow, yeah. i was surprised it didn't crash. however...after increasing anti-skip buffer and cranking the disktimeout up...they started playing properly.
00:24:51Shawn_Khmm, increasing the tick was didn't help anything, I actually think it made it slower >.<
00:24:53preglowdewdude: really? that does however surprise me
00:25:43dewdudeyeah, it surprised me to. but i noticed with 5 sec anti-skip and 5-sec'd skip...every 5 seconds. increasoing the anti-buffer alone did nothing, i think the trick was the disk timeout..i'm wondering if the files are just..physically too large to fit into memory
00:25:49Shawn_Knope, not CPU boost was at 0 and it seems to lag unless it's boosted to 1
00:25:50dewdudei mean, a couple are just under 3000kbps
00:26:10 Join safetydan [0] (i=cbca159f@gateway/web/cgi-irc/
00:26:26dan_aShawn_K: Does scrolling work when it's on the COP? Can you make playback skip by scrolling when the scrolling thread is on the COP? When it is on the CPU?
00:27:08Shawn_Kdan_a: I think so, but I made a bug in another optimization...and it just kicked in
00:27:15Shawn_Kso gimme a sec to reboot and check
00:27:49preglowdewdude: chunks should fit in memory, but it does sounds a bit weird that they play seamlessly at all
00:28:02Shawn_Kdan_a: yep, it works
00:28:35Shawn_Kdan_a: if you want a copy of the build, I can send it to ya...
00:29:07LloreanShawn_K: From a development standpoint, "the build" doesn't really give him anything
00:29:19LloreanHe's not really interested in just playing with it.
00:29:24dan_aShawn_K: I only have a 3G iPod - music skips on that without any help from me :D
00:29:36dewdudepreglow: doesn't it? from everything i read skipping is caused by cpu speed...but what got me is usually when i'd have that happen, like with an mp3 file using software EQ..everything would lag...volume changes, directory scrolling...when playing the flacs, even with thier skipping, the UI didn't lag up at bit. I don't know what any of it means..or if any of it would be of any use to anyone, but i thought i'd share my findings.
00:30:04*safetydan kicks sourceforge
00:30:07safetydanglad we moved off that
00:30:13preglowdewdude: it's just that last time i checked, even 44.1/24 was a bit slow, so i'm just surprised that files that are twice as cpu intensive work without skipping
00:30:31dan_aShawn_K: But I'm planning on committing the kernel_on_cop patch to CVS soon, and knowing it can do useful things makes me smile
00:30:52Shawn_Kouch, bad idea to sick codecs and scroll on same CPU....I'll have to find better threads to choose
00:30:58dewdudetwice as cpu intensive as in how? the FLAC decoding for the 96/24 or the 44/16 resampling rocobox does in addition to that
00:31:06Lloreanpreglow: The problem maybe has to do with the file being a higher data rate than the PCM buffer, rather than anything else?
00:31:47Shawn_Khow can i set a minimal CPU boost of 1 for the iPod?
00:31:57Shawn_Kin code
00:32:16dewdudewouldn't that kind of information be in the wiki?
00:32:28Shawn_KI'm afraid to mess w/ the boost in code w/o having 2nd opinion/help
00:32:34Shawn_Kok, sure...I'll look then
00:32:56dewdudeit's as i always say..RTFW
00:34:09Shawn_Kjust as I thought in first place...only descriptors of actual info on modding it
00:34:46Shawn_Kalso, I saw how the boost for the 1st and 2nd core are linked....can those be separated?
00:34:48safetydanShawn_K, boost is either on or off. The count is just the number of times boost has been called
00:34:51dewdudewell, at least you looked
00:35:02Shawn_Kok, thanks a perfect fix for scroll would be to enable the boost when the scroll is touched....
00:36:19dan_aShawn_K: We don't know how to use the clock skipping function on the PortalPlayer chips. As I understand it, this can run the COP at the same speed as the CPU, half speed, quarter speed, etc.
00:36:30LloreanThat wouldn't solve the UI sluggishness during playback with high CPU use, nor would it be good for battery life.
00:36:36LloreanWhich is why alternate solutions are being looked for.
00:36:40preglowdewdude: both combined
00:36:58Llorean"Boosting on scroll" has been suggested nearly a year ago. :-P
00:37:00lostlogicany new word on committing some version of kernel on cop?
00:37:07preglowdewdude: 96 is more than twice 44.1, and the resampling adds something too
00:37:10dewdudeheh, i spoke a tad too that i've turned EQ on my 2989kbps flac is skipping
00:37:21lostlogic(I always run it now and after traveling last week, it clearly has better battery life)
00:37:25dewdudepreglow: i'm an audio engineer, i'm aware of the bitrates of 96 vs 44
00:37:25preglowLlorean: i don't see why that in itself should have anything to do with it
00:37:26Lloreanlostlogic: I hear it still makes 4Gs freeze.
00:37:37preglowdewdude: sure, just pointing it out since you asked
00:37:39lostlogicah, crappy.
00:38:01Lloreanpreglow: I was just curious if the buffering code could handle reading more from the compressed buffer than it's putting into the PCM, not just less.
00:38:17dan_alostlogic: It seems a little unstable, but I'm back at home now, and hopefully will get the problems sorted soon
00:38:20LloreanWell, buffer/codec/playback
00:38:33dewdudeahh, i was just looking for clarification to what you said..becuase as far as actual decoding of the 96/ least on my PC, it doesn't actually take any more cpu to decode that back to a 96/24 wav. but maybe that's a bad example.
00:38:35Shawn_Kanyways, I found an improvement on the line drawing function
00:38:56lostlogicdan_a: cool
00:39:17Shawn_Kvery little, but it will always help when drawing vertical or horizontal lines
00:39:28Shawn_Kno drawback for other coordinates
00:39:36LloreanOn the scale between iPods and PCs, a 1% difference in CPU speed on a PC could overwhelm an iPod. :)
00:40:16Shawn_Ksad that PCs software is as slow as it is
00:40:32preglowdewdude: oh? weird, it should take roughly twice as much cpu
00:40:43preglowdewdude: if only because you've got twice as many samples to process
00:40:56preglowthings are more complex on a pc cpu, of course
00:41:18dewdudeX86 or even PPC vs. ARM7 isn't a fair match
00:41:47dewdudethen again, i know little about either of the three..i leave that to the people that can actually code.
00:41:47safetydanShawn_K, have you got a patch posted on the tracker for that change?
00:42:09Shawn_Ksafetydan: I will have it up within a few days
00:42:11 Quit Febs (Read error: 110 (Connection timed out))
00:42:38Shawn_Ksafetydan: it's such little improvement, no rush only helps when not drawing diagonal lines
00:42:48linuxstbShawn_K: Is is this patch?
00:42:48dan_aShawn_K: What are you doing? Sending the patch through the post?!
00:43:07 Join barrywardell [0] (n=barrywar@
00:43:58Lloreanlinuxstb: No, those are the changes his build includes on the unsupported builds forum
00:43:59Shawn_K....eeks who posted that old crap of mine from yesterday :P
00:44:24Shawn_Knope, that's nothing, that was me stripping out unneeded code
00:44:30Shawn_KI actually made REAL optimizations
00:44:46linuxstbClipping is unneeded?
00:45:16preglowdewdude: arm7 doesn't really cease to surprise me just yet, though. i had to code assembler optimised routines to make our coldfire based players decode even 44.1/24 at 120mhz...
00:45:19Shawn_KI feel it should be done by the function calling it....clipping should be separate from drawing
00:45:46linuxstbYou should talk to amiconn - he wrote most/all of the graphics code.
00:45:55Shawn_Kok, will do
00:46:02Shawn_Kis he afk?
00:46:06dewdudewell, i believe the DS is based on ARM7 and ARM9 cores..if that could be any indication of how good that little CPU is
00:46:17preglowdewdude: it is
00:46:23linuxstbShawn_K: We'll find out now that I've mentioned his name...
00:46:41preglowin some ways, it's really quite ordinary, but then it just starts performing in some areas, for some reason...
00:46:55preglowit's probably the cache doing it, a cache always complicates stuff
00:47:31 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
00:47:36Shawn_Klinuxstb: I'm only removing things that (i believe) don't relatively have much use in the current 1-bit drawing system
00:47:53 Quit BHSPitMonkey (Connection timed out)
00:47:56amiconnShawn_K: lcd_drawline() _is_ for diagonal lines. Using it for vertical or horizontal lines is just silly
00:47:58Shawn_Klinuxstb: I hope to create some of my own API to account for what I remove...
00:48:09dewdudeyeah. like i said..i know NOTHING about the actual workings of the CPU's..i can't code to save my life...cept for BASIC
00:48:17Shawn_Kamiconn: yes, I know, but some people are dumb enough to do it >.<
00:48:19amiconnlcd_vline() and lcd_hline() are dedicated for these tasks and are far more optimised
00:48:57Shawn_Kamiconn: only reason I did the optimization was because I found it and no drawback on drawing diagonal lines
00:49:12LloreanShawn_K: Does it increase code size?
00:49:24Shawn_KLlroean: yes
00:49:27 Quit Juice^ (Read error: 60 (Operation timed out))
00:49:28LloreanThen it has a drawback.
00:49:49LloreanIt would be better to make sure the function isn't abused.
00:49:55Shawn_KLlorean: when I said no draw back, I meant as far as operation
00:50:12safetydanAlso, is the change actually measured to be faster?
00:50:14LloreanCode size is a drawback in regards to operation, as it decreases memory available for other things, including audio buffer.
00:50:38LloreanIf you add code to do something that's already done well elsewhere, at the cost of even a few bytes, it's somewhat silly.
00:50:44amiconnRemoving clipping is certainly not what we want
00:51:06Shawn_Kwhy can't there be a separate clipping function?
00:51:30Shawn_Kit would make sense not to even attempt to clip if the user knows it doesnt need to be
00:52:08Shawn_K*programer is better woird that user
00:52:21amiconnThe clipping simplifies lots of other things
00:52:42Shawn_Kmaybe I should just PM you for this ^^
00:52:58amiconnSome code already relies on clipping, and viewports will require it even more
00:53:18 Quit mirak (Read error: 54 (Connection reset by peer))
00:53:58linuxstbShawn_K: I think this is of general interest...
00:54:05 Join Angry [0] (
00:54:29Shawn_Kas i said to him...
00:54:30Shawn_Kso, what if I added code to draw w/o all the clipping failsafes
00:54:36Shawn_KI just feel it would make sense when drawing certain parts of the UI, etc
00:54:45amiconnAnd then I can't see how removing the fill optimisations will make things faster...
00:55:21Shawn_Kwell, I am new to this...I know i did make some mistakes
00:55:23dan_aAre any Flyspray admins about? Can I be given developer access?
00:56:59Shawn_Keither way, I may be doing some things wrong, but I am really trying to help w/ good intentions, if you seem mistakes I am making, let me know
00:58:26 Quit lavid ("Leaving.")
00:59:18linuxstbIf your interest is in writing a graphical UI, why not write the UI first, then optimise it once you see what the bottlenecks are? i.e. don't optimise too early...
01:00:12Shawn_Ki may, idk atm
01:00:31linuxstbIt's just that if you change the graphics API, you'll need to implement it for all LCD types...
01:00:58 Join debauched_sloth [0] (
01:01:01 Quit strabes (Remote closed the connection)
01:01:26Shawn_Kok, then I wont change, I'll see if I can add on for optimizations only generated when compiling for devices that support it
01:01:40Shawn_Kor changes will have same effect
01:03:15 Quit dpassen1 ()
01:04:44Shawn_Kyeah, well...maybe I should just fork off since I have such different views than you guys...I don't want to create any firction
01:05:39***Saving seen data "./dancer.seen"
01:06:33 Join dpassen1 [0] (
01:06:57 Join norbusan [0] (
01:07:01 Quit hcs ()
01:07:22Shawn_Kno comments on yes or no?
01:07:27dan_aShawn_K: I don't think forking would be a sensible thing to do - I'm sure there are many optimisations which can be done, but on the other hand a lot of things are the way they are for sensible reasons
01:07:47 Join webguest26 [0] (i=47cd8985@gateway/web/cgi-irc/
01:08:02Shawn_Kok, I just see alot of friction between ideas...that's only reason
01:08:03 Quit webguest26 (Client Quit)
01:08:23Shawn_KI like to challenge things and I understand not everyone likes that
01:08:30 Join webguest37 [0] (i=47cd8985@gateway/web/cgi-irc/
01:08:44webguest37is the iaudio u2 supported?
01:09:05dewdudewebguest37, did you read the website?
01:09:20webguest37yes i thought it was the last time i checked
01:09:29webguest37cant see it now
01:09:56LloreanIt never was.
01:10:04dewdudeyeah. just the X5.
01:10:56webguest37oh guess i aws reading this and thought it was for some reason
01:11:00LloreanShawn_K: You can challenge things all you want, but trying to do something a new way won't work unless you resolve all the conflicts relating to why it was done the old way. Faster graphics at the expense of breaking several other things doesn't make sense unless you fix the other things too.
01:11:08 Join hcs [0] (n=hcs@rockbox/contributor/hcs)
01:11:10Shawn_Kok, gotcha
01:11:16linuxstbShawn_K: I don't see any friction. You're challenging things, we're responding.... If you write patches which prove themselves better than the current way of doing things, then I don't see a problem with them being accepted.
01:11:26Shawn_Kjust as long as you guys are willing to lead in right direction
01:11:35 Join Mouser_X3 [0] (
01:11:41 Nick Mouser_X3 is now known as Mouser_X (
01:11:56LloreanShawn_K: Notice that they explained why they did things the way they were before, rather than saying "Changing things is stupid, don't do it."
01:12:17preglowooh, more usb love
01:12:55barrywardella small bit
01:12:57Shawn_Kok, anyways...amconn made me realize that I did remove one optimization of his :P...i fixed that
01:13:05dewdudeheh, i've started booting into ipod firmware to transfer stuff because the emergency disk mode is SLOW
01:13:08dan_aShawn_K: Challenging things is excellent, but not when done for its own sake. I also don't see any friction: you've had an idea for an optimisation, the guy who wrote that code says that what you are doing wouldn't be an optimisation. Show things that say there's an improvement in features, performance and/or battery life and some core developers will probably be all for your changes. Some will be against them too
01:14:11Shawn_Kare videos decoded on the "codecs" thread?
01:14:27Shawn_Ki havent tried anything with MPEG4 yet
01:14:44webguest37is there a patch to play xvid on ipod 5 gen ?
01:15:02linuxstb(to webguest37)
01:15:03dewdudewebguest37: read the website.
01:15:14webguest37no pdf viewer either hmmm
01:15:20dewdudeif you don't see it there, then check the wiki, if it's not there, search the forums, if it's not there, then ask
01:15:31Shawn_K>.> pdf is fairly complex...
01:15:37linuxstbShawn_K: No, videos are currently outside the playback/codec code, in a standalone plugin.
01:15:45Shawn_Kthat will probably be a while for that one
01:15:53Shawn_Klinuxstb: ok, thanks
01:16:10Shawn_Klinuxstb: I want to make sure all my COP changes are compliant
01:16:30Shawn_Kwell, at least not restruicting on video
01:17:08Shawn_Kalso, maybe I can get that volume error I found fixed ^^
01:19:13 Join CloFan [0] (i=46b2848a@gateway/web/cgi-irc/
01:19:54amiconnreboot, brb
01:19:54 Quit amiconn (" HydraIRC -> <- Leading Edge IRC")
01:20:35CloFanI have a question... when I connect my ipod 4g non-color to my computer, it boots into apple's firmware instead of staying in Rockbox's. I see the USB cord on the rockbox screen, then the thing reboots and goes into the apple software. How can I fix this, or is it normal?
01:21:40dan_aCloFan: That is normal - Rockbox does not have it's own USB support for iPods yet, so goes into Apple's Disk Mode
01:22:02CloFanAlrighty, thanks Dan! Just wanted to make sure it wasn't doing somethign wrong :)
01:24:54 Join amiconn [0] (
01:24:55*preglow watches his ipod refuse to boot rockbox all of a sudden
01:27:43pregloweh, what the hell
01:27:49preglowit just took a minute for the bootloader to work
01:27:53 Quit norbusan (Remote closed the connection)
01:28:26dan_abarrywardell: I'm just having a look at your USB plugin... I don't think that the config option should be USBOTG_MX31 - As far as I can tell, the USB is a design by Innovative Semiconductors, used by both PortalPlayer and Freescale in their chips
01:29:00barrywardelldan_a: ah, ok. what do you think it should be called?
01:29:55dewdudeCloFan, you're actually better off booting completely into the apple firmware than just emergency disk mode if you're transferring a lot of data. i found EDM to be painfully slow.
01:30:22barrywardelldan_a: USBOTG_ISP1504?
01:30:40preglowlinuxstb: is there something funky with the bootloader you bundled with ipodpatcher when i tested it a while ago?
01:30:54dan_aThe Linux driver refers to it as ARC, so either USBOTG_ARC, USBOTG_IS or USBOTG_ISI would seem the best fit to me. Why ISP1504?
01:32:02Lloreandewdude: The slow EDM is mostly on Nanos, with a few exceptions that seem to be on individual iPods rather than target categories.
01:32:23barrywardelldan_a: never mind. ISP1504 is the transceiver driver in the freescale code
01:32:27dewdudereally? mine must be an exception then.
01:32:44LloreanVery possibly
01:32:47LloreanNotice I said "mostly"
01:32:54hcsbarrywardell: would it be helpful if I got a dump from your plugin on my ipod photo?
01:33:48barrywardellhcs: it would be interesting. maybe we could see if the ipod is running the usb and wasting battery
01:33:57dan_abarrywardell: Of course, it is completely up to you what you call it :D
01:34:02dewdudewell, i know when i was converting some FLAC's to MPC last night directly on the ipod...EDM yields me about 4x realtime where as booting into the ipod firm yields about 11x...doing the same process on the local drives yields about 12x. I also noticed when copying a few hundreed megs of data that EDM took about 2 minutes per 100 megs
01:34:07dewdudeerr....1 minute
01:34:09barrywardelldan_a: i think USBOTG_ARC is a good choice
01:34:36preglowbarrywardell: doesn't look like debug info changes much when doing setup and run here
01:34:38Lloreandewdude: On Nanos it can take 7 minutes to copy an already extracted Rockbox full build onto it, for me.
01:35:01LloreanIn Apple_OS the same operation takes less than half a minute
01:35:04dewdudewow. mine's not that slow. but i do notice a HUGE difference between EDM and ipod on myunit
01:35:10barrywardellpreglow: no? then it's possible that the ipod is already running the usb chip
01:35:42preglowbarrywardell: should the ipod reboot when inserting the usb cable while running the plugin?
01:36:25amiconnbarrywardell: That would not be really surprising, given the fact that both the H300 and the X5 usbotg chips are active by default, and need to be put to sleep explicitly
01:36:33Soapdewdude: can you ABX 96/24 to 44.1/16 or is it just convience?
01:36:33barrywardellpreglow: i didn't take account of that (the h10 doesn't reboot by default), so yes. but maybe i should disable it
01:37:10preglowbarrywardell: wait, it does change
01:37:16preglowbarrywardell: but it reboots, so i can't check dmesg
01:37:54preglowamiconn: i really wonder if the usb part in the pp chips needs to be put to sleep to
01:37:56dewdudeSoap: since my 96/24 is vinyl content, 44.1/16 is actually more more resolution than needed..i just have a theory about preserving as much of the original analog waveform as possible.
01:38:07barrywardellpreglow: try commenting out the code in usb_enable() in firmware/target/arm/usb-pp.c
01:38:34barrywardellor wait a minute and I'll have a new patch ready
01:39:08dewdudeit's also a matter of convenience...often times i'm loading stuff up as i'm about to walk out the door..and doing a downsample to 44.1/16 takes a little bit of time.
01:39:58SoapMM or MC cartridge?
01:40:26dewdudemc's are too expensive, slightly overrated, and i don't have a preamp that'll take one.
01:40:59linuxstbpreglow: I can't remember giving you anything especially funky...
01:41:00preglowi get a ton of dmesg output
01:41:19preglowlinuxstb: the bootloader uses about fifteen seconds to load a rockbox.ipod now
01:41:27preglowlinuxstb: think i'll just reinstall it
01:41:59preglowah, no
01:42:00preglow[17700.943838] usb 5-4.4: new high speed USB device using ehci_hcd and address 16
01:42:00 Quit webguest37 ("CGI:IRC (EOF)")
01:42:03preglow[17701.339134] usb 5-4.4: device descriptor read/64, error -71
01:42:05preglowthat seems to be it
01:42:46preglowand then more of the same
01:42:58barrywardellpreglow: it's possible that connecting with the usb plugin messes up the usb on your computer (it did on my mac)
01:43:02barrywardella restart fixed things
01:43:27dan_abarrywardell: I'm sure you know this already, but your plugin seems to work on the Sansa
01:43:28barrywardelli experienced very slow usb after using the plugin until I restarted
01:43:35preglowbarrywardell: yeah, got that a ton of times, don't think it happened now, though
01:44:00barrywardelldan_a: great. i hadn't got around to testing
01:44:57preglowbarrywardell: nah, usb still works
01:45:45Arathisdoes anybody work on rockboy for the H10?
01:45:55LloreanArathis: No.
01:46:25LloreanYou're welcome to though.
01:47:06ArathisWould do it, but I have no idea of C :/
01:48:17LloreanThe problem is, the people who are interested in it, like you, either can't do it, or are unwilling to spend the time on it.
01:50:06 Join BHSPitMonkey [0] (
01:53:38debauched_slothI'm working on the Gigabeat port and am looking at doing my first commits and have some questions about how to organize some of the code
01:53:50debauched_slothlinuxstb has been kindly helping out and suggested that I ask here
01:54:17debauched_sloththe full set of patches is here:
01:54:29 Quit webguest17 ("CGI:IRC")
01:54:46debauched_sloththere are probably three commits in there, one having to do with mmu, one with ATA DMA, one with the LCD driver
01:55:01debauched_slothcomments welcome
01:55:15debauched_slothwhat has me stumped is what to do with the code in lcd-16bit.c
01:55:25debauched_slothas you can see, I've got some GB specific stuff in there
01:55:57debauched_slothI can fork it off, clearly, but I will end up with about as much duplicated code, and about as many #if defined GIGABEAT's
01:56:11debauched_slothI would take functions like lcd_clear_display and fork them entirely
01:56:47ArathisLlorean: that's very often the problem of opensource anyway
01:56:54Shawn_Kwhat command operates the "audio" thread in rockbox?
01:57:43Shawn_Kbetter said, where can I find the code for the audio thread in the source?
01:58:06LloreanArathis: This is true. But also remember, the primary focus of Rockbox is not playing games, but audio features, so it's rare that a core programmer's going to work on Rockboy, and even rarer that one will work on it for a single target (though there's likely little H10-specific stuff to do for it, and optimizations would fall in the PortalPlayer family rather)
01:59:05ArathisThat's why I'm sad about it, but wouldn't blmae anybody for it or somethin
02:02:33amiconndebauched_sloth: I'm not sure what to do in this special case yet, so just some questions and general hints
02:03:04amiconn(1) Avoid to check for models (e.g. #ifdef TOSHIBA_GIGABEAT_F), check for features instead
02:03:37 Quit CloFan ("CGI:IRC (EOF)")
02:03:41amiconn(2) Is there a reason not to use dma blit?
02:03:50amiconnIf not, the check could go away. kiss.
02:04:21linuxstbShawn_K: It's all in apps/playback.c
02:05:30amiconnIt looks like the only function using dma blit atm is lcd_clear_display()
02:05:38debauched_slothamiconn, I left that option in there until I can get some more widespread testing
02:05:53debauched_slothno, there is another one, lcd_update - same option
02:06:25amiconnMaybe those functions should be implemented in the target part of the lcd driver, and #ifdefed out in the general lcd driver if a target does special optimisations
02:06:49amiconnDon't forget the simulators though, they'll probably need the standard function
02:07:43debauched_slothamiconn: yes, I can do that (and missed the sim, thanks) but it will involve duping some code
02:07:59debauched_slothI would need to pull init, set_fg, etc
02:08:16debauched_slothbasicalyl everywhere I have an ifdef, that function gets pulled to the driver in the target tree
02:08:24debauched_slothI am fine with this, but it's some duplication
02:09:03amiconnThe init part specific to the gigabeat could go into lcd_init_device(), which is target specific anyway
02:09:25debauched_slothgood point
02:09:38debauched_sloththat just leaves the set_*, which are minimal
02:09:53debauched_sloththen I can reuse virtually all of that driver
02:10:40 Part Angry
02:10:47amiconnThe if(!wordcount) return; can be removed from copy_read_sectors()
02:11:17amiconnThe ata driver ensures wordcount to be a non-zero multiple of 256; other target's implementations also rely on that
02:11:46debauched_slothfigured an easy catch if there way a bug, could generate a multimegabyte overwrite
02:11:54debauched_slothwas a bug
02:12:25amiconnATA writes are limited to 256 sectors anyway (128KB)
02:12:40amiconn...unless lba48 is supported & enabled
02:12:41debauched_slothno, the DMA may go a little nuts
02:12:55amiconnAh, you mean when zero is passed...
02:13:02debauched_slothyes, undefined
02:13:24debauched_slothprobably hangs that function. probably should, actually
02:13:55amiconnIs there a reason not to use dma in the bootloader?
02:14:14debauched_slothat this point, other than caution, not really
02:14:25amiconnOk, so this could go away later
02:14:30debauched_slothlike the bootloader taking the simplest path possible
02:14:41amiconnImho it's better to use the same code in bootloader and core
02:14:45debauched_slothyes, absolutely
02:15:25amiconnI thought there might be a special reason, like the of not starting when the bootloader changes certain hardware settings
02:16:32debauched_slothit used to need a correctly setup irq handler, which happened a little later in bootloader
02:16:34debauched_slothnot any more
02:16:52amiconnIs the explicit refresh in __backlight_on() really required?
02:17:16debauched_slothwhen I turn off the lcd, I stop blitting to the LCD framebuffer
02:17:26debauched_slothso when its coming back on, I need to update the screen to the last state
02:17:32amiconnI would have guessed it is, but I think __backlight_on() isn't the right place
02:17:46 Quit muesli__ ("ich will Kühe!!!")
02:17:58debauched_slothwhere would you put it/
02:18:18amiconnOn iriver H300 and iaudio X5, we also power off the lcd when backlight is switched off, and have to refresh its internal buffer when powering on again
02:18:29*amiconn checks where and how this is handled
02:19:28 Join strabes [0] (
02:19:41amiconnAh, yes. Check firmware/target/coldfire/iriver/h300/lcd-h300.c
02:20:18amiconnThere is a function lcd_enable(), which calls lcd_update() when enabling the lcd
02:21:12amiconnJust checking from where this is called..
02:21:15 Join Febs [0] (
02:22:12debauched_slothyep, same idea
02:23:39amiconnThe idea is that switching the backlight might be made independent from the backlight (not implemented on H300, but on X5), for 2 reasons:
02:24:35amiconn(1) Switching the backlight back on might take a while. Even if it's just a fraction of a second, it might annoy users. So the X5 allows to set a delay which starts after switching off the backlight before the lcd is also switched off
02:25:02amiconnThis allows for fast reactivation of the display during this delay time
02:25:32Shawn_Kanyone know what function starts the "audio" thread?
02:25:36Shawn_KI cannot seem to find it
02:25:37amiconn(2) The display might be readable without backlight in very bright environments. If that's enough for a user, he might desire to keep the display active
02:26:01debauched_slothI can go that route, the thing is that I don't think that switching it off is really something you can tell apart from turning off the backlight and the display is not readable, at all (that I can see anyway) in any light
02:26:18amiconndebauched_sloth: lcd_enable() is called from the platform specific backlight_*.c files
02:26:19debauched_slothone problem I have had debugging this code is telling if people actually have it
02:26:25Lloreandebauched_sloth: Try it with a pure white background, and black text in a large font.
02:26:27debauched_slothI'll take a look
02:27:04debauched_slothI'll try it, but I'm at large font with gray background and in a bright light, nothing
02:27:14debauched_slotheasy to test
02:27:43goffayeah.. its pretty bad w/ no backlight
02:27:54amiconnAhaha, on h300 (without the delay) __backlight_on() and __backlight_off() just call lcd_enable() with the appropriate parameter
02:27:56Shawn_Knm...I found it in playback.c
02:28:20amiconnSo nothing wrong with your way, just that using an explicit function would make things a bit clearer
02:28:42 Quit Ajaxinc (Read error: 110 (Connection timed out))
02:28:59debauched_slothcan't see a think
02:29:05debauched_slothI really can't tell the difference
02:29:17debauched_slothbut you are right, cleaner, I will move it into lcd_enable
02:30:31amiconnAn mmu in an embedded system seems to be a strange beast...
02:30:57debauched_slothhow so?
02:31:02 Quit akaias (Read error: 110 (Connection timed out))
02:31:08debauched_slothall this manual cache coherency stuff?
02:31:40amiconnsystem_reboot() should go into system-meg-fx.c
02:31:58debauched_slothsystem_reboot needs a great big comment saying it doesn't work but should
02:32:04debauched_slothI'll yank it to there
02:33:33amiconnand invalidate_icache() should go into the appropriate system-target.h
02:33:41 Join akaias [0] (
02:34:00debauched_sloththose last two, I was copying what was in place
02:34:15debauched_slothI gather the general direction is to split to target when possible, so I'll do so
02:34:18amiconnThe whole idea behind target tree is to reduce or even remove target specific code from the generic firmware/ files
02:34:33debauched_sloththe move is in progress?
02:34:39amiconnYes, still
02:34:40 Quit dpro (Read error: 110 (Connection timed out))
02:34:50amiconnEspecially the old targets aren't moved completely
02:34:50debauched_slothwas not sure and did not want to rock the boat
02:35:01amiconn(i.e. archos and some iriver h100/h300 code)
02:35:16 Join rkostynu [0] (
02:36:34linuxstbThe X5 port is probably the best example of target-tree usage - it was never not target-tree.
02:37:03amiconnata-meg-fx.c is Copyright (C) 2006 by Linus Nielsen Feltzing ???
02:37:18*amiconn somehow doubts that - seems to be a c&p mistake
02:38:08Shawn_Kwhat uses more cpu, the codecs or the audio?
02:38:16Shawn_KI'm guessing codecs
02:39:00debauched_slothnot sure who created that file
02:42:42debauched_slothah, here is an interesting problem, maybe a bug
02:42:59debauched_slothlcd_init_device calls lcd_clear_display *before* calling lcd_init_device
02:43:14debauched_sloththis seems reversed - is reversed for my puposes
02:44:10amiconnHmm. It's not reversed for a 'traditional' lcd interface
02:44:17amiconnlcd_clear_display clears the framebuffer
02:44:40amiconnlcd_init_device() performs the first transfer of the framebuffer to the lcd.
02:44:43debauched_slothIf we want to fill with anything other than zero for init, won't work
02:44:57debauched_slothfor us, anyway, my stuff aside, we need to fiddle the framebuffer format
02:45:01amiconnSo if it's not cleared beforehand, garbage would be transferred
02:45:24debauched_slothhere the framebuffer is in main memory
02:45:46amiconnWhat I'm talking about is the rockbox framebuffer
02:45:54amiconnThe lcd has its own internal buffer
02:46:23debauched_slothyou are right
02:46:36debauched_sloththis works for me, too, since I can do that initial fill w/o DMA
02:46:39gotthardtFor the GB the LCD shares main memory
02:46:52linuxstbamiconn: I didn't think any of our drivers did an lcd_update() as part of the lcd_device_init() though.
02:47:39debauched_slothyes, my issue is that I do the lcd_clear_display using DMA and need some setup for that
02:47:49debauched_slothnot a problem if I do initial fill using non-DMA
02:48:15amiconnlinuxstb: H300 does, H1x0 does...
02:48:36linuxstbAh, the _display_on() call?
02:48:39debauched_slothI have to setup those fill patterns
02:48:46debauched_slothI can init them statically, I guess
02:48:50debauched_sloththe 16-byte fills?
02:49:15 Quit gotthardt ("CGI:IRC")
02:49:34amiconnlinuxstb: Yes on H300. On H1x0, lcd_init_device() calls lcd_update() directly
02:50:13linuxstbIs there a reason why it's only in those two drivers?
02:50:34 Quit jhMikeS (Nick collision from services.)
02:50:40 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
02:51:48amiconnArchos does too, but X5 doesn't seem to do it. Strange...
02:52:03 Quit rkostynu ("Chatzilla 0.9.77 [Firefox]")
02:52:53 Join fasmaie [0] (
02:53:29Shawn_KI just switched a few more threads over to COP to allow the "audio" thread more CPU
02:53:39amiconnIpod doesn't either, but then we don't shut down the lcd on ipods (yet)
02:53:42debauched_slothpulling these functions into the LCD target driver is going to involve non-static'ing some stuff
02:53:44Shawn_Kanyways, I'll have patches for my stuff by this weekend hopefully
02:53:51debauched_slothfor this target anyway
02:54:02Shawn_K <−− get it here
02:54:06Shawn_Ksee ya all, I g2g
02:54:10 Quit Shawn_K ()
02:54:55*Llorean sighs.
02:55:01 Join shnee [0] (
02:55:20amiconndebauched_sloth: That's sometimes necessary, and I don't think it's a problem
02:55:30debauched_slothI could add hook functions in there as well
02:55:39debauched_slothlet me hook at the end or beginning, like that
02:55:41debauched_slothugly, too
02:55:53amiconnA related note is that it's usually better to declare functions in .h files rather than in individual source files
02:56:26amiconn...and the worst thing is to declare functions within other functions...
02:56:34*amiconn is referring to extern void lcd_clear_display_dma(void);
02:56:43 Join Mouser_X3 [0] (
02:56:49debauched_sloththat just got pulled to where it does not need a prototype
02:56:54 Quit Mouser_X (Nick collision from services.)
02:56:56 Nick Mouser_X3 is now known as Mouser_X (
02:57:49debauched_slothall this other stuff I will move into...hmm
02:57:51 Quit fasmaie ("Apples have meant trouble since eden.")
02:58:05amiconnAnother hint, if dma blit is used in core but not in the bootloader, you could just check for !defined(BOOTLOADER) instead of doing a runtime check
02:58:32debauched_sloththat is there because this port is early and I want more testing (with ability to isolate issue) before making it a default
02:59:02debauched_slothpretty much all the feature on/off flags are there for that reason
03:00:00*amiconn prefers writing and using test plugins for performing heavy testing, and then remove alternatives before commit
03:00:20debauched_sloth**sloth wants full coverage testing of a video driver change before making it default
03:00:34debauched_slothwe'd end up in the same place, I reckon
03:00:57 Quit dan_a (Read error: 110 (Connection timed out))
03:01:39debauched_slothwhere do you suggest I place the extern defs for the stuff I am pulling?
03:01:43amiconnI often use alternatives (using #if 0 / #if 1) during such tests, but those usually never see the light of cvs
03:02:14debauched_slothI want this to maybe see the light of users, most especially the DMA blit stuff
03:02:31debauched_slothif someone sees display artifacts, I want them to be able to disable function and isolate to the blit
03:02:37 Quit phrozen77 ("Some people say that I must be a horrible person, but that's not true. I have the heart of a young boy... in a jar... on my d)
03:03:31amiconnThe problem is that users usually don't test things unless you either force them to (by making it default), or explicitly request test (e.g. via the mailing list)
03:03:46debauched_slothgood point
03:04:14debauched_slothbut if I break one of their games....? bet I hear about that right quick :)
03:04:38amiconnIf there's an older daily build without the new feature, and the feature causes problems in the new build, users will (1) go back and (2) complain
03:05:35 Part mat
03:05:46debauched_slothI note that the lcd_fastpixelfuncs are not static...are they used elsewhere?
03:05:48***Saving seen data "./dancer.seen"
03:05:51amiconnOf course it's still useful to be able to produce a build without that feature for testing purposes, and sometimes keeping old code enclosed in #if 0 can be helpful for that
03:05:53debauched_slothnever mind, I can answer that
03:07:23 Join fasmaie [0] (
03:07:43fasmaielinuxstb: did you get the information about the 4G ipod
03:07:48 Quit funky ("leaving")
03:08:19linuxstbfasmaie: Yes, thanks. But I haven't had chance to investigate yet.
03:09:13fasmaielinuxstb: That's fine, I just wanted to be sure you noticed the posting
03:09:23fasmaielinuxstb: Hope it helps
03:09:52amiconndebauched_sloth: Hmm, you're right, and that also goes for lcd_pixelfuncs and lcd_blockfuncs
03:10:12barrywardellhcs, preglow: did you get a dump of the usb debug info?
03:10:52 Quit jhMikeS (Nick collision from services.)
03:10:56 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
03:11:30hcsno, I'll check it out with the new patch now
03:11:40amiconnThe idea was that these functions might be needed for some complementing drawing functions which shouldn't go in the core, but the only such drawing function is xlcd_filltriangle() - and that one does just use lcd_vline() or lcd_hline() depending on target
03:11:59barrywardellcool, thanks
03:12:34debauched_slothI'm going to define a new function, lcd_prepare_backdrop()
03:12:43barrywardellhcs: logs before and after doing the setup and run would be interesting.
03:12:49debauched_slothfor everywhere except GB, that will do nothing
03:12:55debauched_slothfor GB, it can clean the cache
03:13:08debauched_sloththis will let me pull a lot less symbols over, and it is less dup'ed code
03:13:08 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
03:13:13debauched_slothany issues with that?
03:13:22 Quit amiconn (Nick collision from services.)
03:13:22 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
03:14:28amiconndebauched_sloth: Then just #define lcd_prepare_backdrop() for everything except gigabeat in the appropriate .h file
03:14:41debauched_slothok, will do
03:15:02 Part pixelma
03:16:39 Quit fasmaie ("Hey! Where'd my controlling terminal go?")
03:17:22linuxstbWould lcd_init_backdrop() bea better name?
03:17:30amiconnlcd_update_rect() does only take y and height into account on gigabeat?
03:17:31linuxstb^be a
03:18:11debauched_slothit's not really an init
03:18:19debauched_slothyes, right now, only why and height
03:18:33linuxstbWhat will it do?
03:18:37debauched_slothlcd_prepare_backdrop is more like getting it ready to display, which is literally what this is doing
03:18:41 Quit Rob2222 ()
03:18:58 Join Rob2222 [0] (
03:19:07linuxstbSo it will be called from lcd_set_backdrop() ?
03:19:38 Join gotthardt [0] (
03:20:02linuxstblcd_device_set_backdrop() ?
03:20:13debauched_slotheven better
03:20:23debauched_sloththat's not what it does
03:20:35debauched_slothnot setting the device backdrop, getting the backdropped prepped for the device
03:20:49debauched_slothmaybe lcd_device_prepare_backdrop?
03:22:18linuxstbI guess that works.
03:23:02debauched_slothnow lcd_clear_display
03:23:21debauched_slothit sets scrolling_lines to zero
03:23:30debauched_slothfair enough, if I duplicate that code, I can do that as well
03:23:40debauched_slothbut I sort of hate that that logic is in two places
03:24:26debauched_slothok, that code has been pulled
03:24:33 Join Alexinc [0] (n=Ajaxinc@
03:24:51debauched_sloththe stuff that I made non-static, where would you suggest that I put it? lcd.h seems wrong. New file...lcd-16bit.h?
03:25:48debauched_slothextern int drawmode;
03:25:48debauched_slothextern fb_data* lcd_backdrop;
03:25:48debauched_slothextern unsigned fg_pattern;
03:25:48DBUGEnqueued KICK debauched_sloth
03:25:48debauched_slothextern unsigned bg_pattern;
03:25:49debauched_slothextern long lcd_backdrop_offset;
03:25:49***Alert Mode level 1
03:25:49debauched_slothextern volatile int scrolling_lines;
03:25:52debauched_sloththat's the list
03:32:06hcsbarrywardell: what order should I perform these steps in?
03:32:22hcsi.e. when to plug in
03:32:52 Part Llorean
03:34:23barrywardell1) dump log 2)plug in 3) setup 4)run 5)dump
03:35:06hcshold down menu while plugging in?
03:35:50***Alert Mode OFF
03:36:09barrywardellshouldn't be necessary
03:36:11 Quit strabes (Read error: 110 (Connection timed out))
03:36:17barrywardellthe patch disables the system reboot
03:36:27hcsit brings up the usb plug logo
03:36:32hcsuntil I unplug
03:36:39 Join Thundercloud [0] (
03:37:22 Join jdong [0] (n=jdong@ubuntu/member/jdong)
03:37:30barrywardellah, ok. in that case yes, hold menu
03:37:34jdongis the coprocessor patch safe to test out on my 5.5G ipod?
03:37:49jdongand does it require anything new, like a new bootloader?
03:38:34hcswant the end of dmesg, too?
03:39:30barrywardellyeah, sure
03:41:17 Quit jhMikeS (Nick collision from services.)
03:41:22 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
03:43:13debauched_slothhave that down to two new imports, fg_pattern and bg_pattern
03:46:46 Quit Alexinc ()
03:49:09hcsbarrywardell: uploaded my logs
03:49:36barrywardellgreat, thanks
03:50:34hcsthe usbdebugbefore/setup/run series is without connecting it to anything
03:52:36barrywardelllooks the same as i have for my h10
03:53:21barrywardellit looks like it's running in low power mode
03:53:47barrywardelland stopped
03:56:39barrywardellhcs: thanks for running the tests. they confirm that the usb controller is in the same state on the ipod and h10
03:56:52 Quit Thundercloud (Remote closed the connection)
03:57:05barrywardellbed time...
03:57:13 Quit barrywardell ()
04:09:11 Part debauched_sloth
04:12:57 Quit Arathis (Read error: 110 (Connection timed out))
04:13:50 Join Nico_P [0] (
04:15:42 Join heapcat [0] (i=4840e8db@gateway/web/cgi-irc/
04:17:59heapcatI am looking at doing my master's thesis using rockbox. In particular something realated to the DataBase system. Is there any other documentation besides what is on the wiki?
04:18:37hcsnot that I know of, I learned what little I know studying the source
04:21:09heapcatDo you know of any status on projects involving syncing the database externally. Like with a sync program? Or do you just have to let the system re-initialize?
04:21:25hcsI have not heard of such projects.
04:22:24safetydanheapcat, there's been talk of someone writing an external program to generate the database off the device, but no one's actually done it. I think most people don't think it'd be that useful
04:22:50heapcatAt the simple end of my thesis I would like to allow the users to easily rate their tracks/files they listened to ...
04:22:59hcsthere is such an external program, it is in cvs
04:23:25heapcaton the more complex end I would like to allow their ratings to be combined to generate recommendations
04:24:01safetydanhcs, ah, I must have missed when that got updated for the new format
04:24:21heapcathcs: what is the name of the program/plugin?
04:24:29hcsuh, how new is the new format (I may not have heard about it yet...)?
04:24:47safetydanhcs, about a year
04:25:05hcsheapcat: tools/
04:26:05safetydanheapcat, I believe you can already rate songs within Rockbox and generate playlists with those ratings
04:26:12hcsthe rating system in there so far is based on playtime
04:26:22hcsthere is no explicit rating in the db
04:26:54safetydanhrm... maybe I should actually use the system before talking about it :)
04:27:23heapcatWhat I am looking at doing is collecting statistics about how users listen to podcasts and then collect that information to make recommendations
04:28:09safetydanHow would you tell a podcast from any other audio file?
04:28:43heapcatI figured at the the syncing of the files there might be some ID3 type tag.
04:28:44jdongisnt the genre Podcast?
04:29:05heapcatFor example the title could be Podcast:Rockbox is greate
04:29:10heapcaterr, great
04:29:13safetydanjdong, no idea, never listened to a podcast
04:30:59heapcatLet's make it more simple. Say you are listening to any audio file that you would like to recommend to others. I may rate it a 5 out of 5, but not want to e-mail to my friends.
04:32:47dewdudepodcasts are just basically RSS delivered media files...most commonly mp3 or aac.
04:35:12heapcatHave you ever used Pandora ( you have three choices for music they recommend? You like it/hate it or fast forward the song.
04:36:10 Quit merbanan (Read error: 110 (Connection timed out))
04:37:51heapcatHow would I collect/generate this information in Rockbox? Would it be a plug-in that records the name/id of the song and what I did (like/hate/fast forward).
04:40:28Soapthere is a hack out there that allows you to record your Pandora listening habits on Last.FM.
04:41:13SoapReading about that might give you insight into how Pandora listening habits are monitored and give you some ideas on how else you could use said information
04:41:56SoapI'm not sure, though, what you would do with said statistics in rockbox.
04:42:44 Join Criamos [0] (
04:42:55dewdudei don't think he wants to use the info in rockbox, he wants to generate with rockbox
04:43:05safetydanheapcat, if you've looked at those are already supported in Rockbox
04:43:06dewdudeyou'd have to code a plug-in of some sort.
04:43:25safetydanI imagine supporting Pandora type ratings could be done in a similar manner
04:43:30Soapahh, well there is already the Last.FM log in rockbox, you could build upon that.
04:43:43Soapoh, safetydan beat me to the punch.
04:44:50heapcatThe idea then would be to have a plug-in write to a log (Last.FM or other) that could track user responses,etc right?
04:44:54 Join Alexinc [0] (n=Ajaxinc@
04:50:01dewdudewell, if you had looked..and as it's been said...rockbox can generate a log.
04:52:06 Join gop [0] (
04:52:14gopmy ipod keeps rebooting
04:52:23gopsaying loading orginal firmware
04:52:26gopipodmini 2g
04:52:37gopbut it won't allow me to try boot in rockbox
04:52:51gopany idea
04:55:18 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
04:55:51 Quit Mouser_X (Read error: 110 (Connection timed out))
04:56:17 Quit gop ("Leaving")
04:56:20Soaphold switch on?
04:56:46Soapwait, nevermind, you say it is caught in a reboot loop.
04:57:01Soapdoes it ever successfully load the apple firmware?
04:57:48 Join BobJonkman [0] (
04:58:17Alexinche left
05:04:35OverandHas there been any work on an NSF player?
05:05:03hcsI've got unofficial builds if you'd like to try it out
05:05:54***Saving seen data "./dancer.seen"
05:07:15OverandAcutally, I would- I'm running on an iPod G5, though I'm using a nightly, haven't compiled anything myself.
05:07:22OverandIs it a plugin or a codec?
05:07:37OverandI'm moving the HVSC over right now (;
05:08:38LloreanOverand: Beware the HVSC collection. Do not combine with Database. :-P
05:08:50OverandLlorean: not surprising that it could cause issues.
05:09:00jdongnsf = nes sound format...
05:09:05LloreanIt's a few more songs than Database likes to handle.
05:09:06jdongwow you learn something new every day I guess :)
05:09:06Overandjdong: yeah
05:09:19LloreanNSF worked well when I tried it in the past.
05:09:24Overandso something cute I've noticed-
05:09:32LloreanI'm too lazy to roll my own builds most of the time, so I've been hoping it'll get committed.
05:09:52Overandcopying the HSVC over to my ipod works great when it's booted as an iPod, and drags ass when it's in emergency disk mode- looks like emergency disk mode doesn't impliment caching or something
05:09:55CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
05:09:55*jdong is still overly fascinated with coprocessor builds of rockbox
05:10:00hcshvsc is awesome!
05:10:06jdongamazingly it actually does feel faster already :D
05:10:19Overandliek, 20 minutes to copy vs. 120- i think it's a matter of how tiny and many the files are
05:10:23Overandjdong: awesome
05:10:38*Overand wishes he could contribute to rockbox beyond ideas for skins/backgrounds and occasional testing
05:10:46OverandI'm just not really a coder, unfortunately.
05:11:31scorchethen learn
05:12:04Overandscorche: It's been on the back burner for years- one of the issues is that I still think in an awful combination of turbo pascal, VB6,, and PHP 4
05:12:08Overandwhich is just about as bad as it gets
05:12:16 Join Daishi [0] (
05:12:19Overandreally, my main gig is as an audio engineer though
05:12:27Overandand i have way, way too many hobbies
05:12:30*hcs learned dBase IV scripting first
05:13:03hcslet this be a lesson to keep your children away from manuals
05:16:23 Join gopp [0] (
05:16:30goppone question
05:16:34goppmy ipod just keep rebooting
05:16:37goppI can't stop it
05:16:41gopploading orginal firmware
05:18:15 Join Mouser_X3 [0] (
05:18:20 Nick Mouser_X3 is now known as Mouser_X (
05:18:41 Quit Nico_P (Remote closed the connection)
05:20:39goppmy ipod
05:20:39goppjust keeps
05:20:48goppwon't even allow me to enter in diskm ode
05:21:32goppor is it becuase of low battery
05:22:00hcsthe original firmware takes a fair amount to load
05:22:35hcstry plugging it in and as it is coming on holding down the key combination for disk mode
05:22:40tehsmohcs: so I think the database has changed since the most recent itdb2tc was released
05:22:54tehsmounfortunately I don't know enough about how it's stored to look at what's changed
05:23:13hcstehsmo: oh, ok, does still work? we could look at that
05:23:34dewdudegopp: when your battery gets low enough, you'll get the apple logo, then it'll blip out. my ipod has gotten to that point like, twice.
05:23:56hcsthe offline tagcache generator
05:24:21tehsmoI see, it's in tools. I'll check
05:25:13goppdewdude so I need it to get charged
05:25:23goppall it say is loading orginal firmware
05:25:36hcsis the hold switch on?
05:25:48dewdudewell, check your hold switch. but, plug it up
05:26:08dewdudeif it loads original firmware, fine...let it.
05:26:47Overandalright man
05:26:58Overandhcs - i'm rocking out to the solstice nsf
05:27:07Overandyeah man
05:27:56goppit says loading orginal firmware and it also say partion 1 : 0x0b 3869
05:28:07goppvesion 200 something something
05:28:09dewdudethat's normal.
05:28:14goppand it it bleeps apple logo
05:28:34dewdudeyou mean it rebooted again?
05:28:41goppit keeps rebooting
05:28:48goppnone stop
05:28:51dewdudeeven plugged up to USB or firewire or whatever?
05:29:29dewdudeemergency disk mode key-combo work?
05:29:31 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
05:29:46goppselect and play / pause
05:29:48*dewdude shrugs
05:30:21 Quit Daishi (Remote closed the connection)
05:30:21goppI must be stupid
05:31:43goppif I unplug
05:31:48goppthe hardrive
05:31:57goppor is thier some reset button in the logic board
05:32:29goppmy ipod just keep rebooting
05:32:53LloreanDisk mode will work if you get the battery to have enough charge.
05:32:56LloreanWhat type of iPod is it?
05:33:12goppipod mini gen 2
05:34:43LloreanPlug in the iPod. Hold Menu+Select until the screen blanks, don't depend on the reset it's doing itself. IMMEDIATELY hold Play+Select. If you don't see it *try* to go to diskmode before restarting (if you still see the normal things it shows right now) then you did it wrong.
05:34:47Overandhm, SID playback sounded a little off.
05:35:07hcsNSF may, too
05:35:22Overandthe few songs i played sounded reasonable
05:35:26OverandI didn't give it a hardcore listen.
05:35:34OverandI'm wowing my co-workers at the studio, heh.
05:36:03Overandi mean, HVSC - "Agent X II - The Mad Prof's Back" actually did play wrong, I've heard it on a real 64 and in various other players, but SID playback is a tough job
05:36:24hcsin the source there is a warning that the sound quality is shitty
05:36:37hcsmost of my work on the nsf player was hacking out accuracy...
05:37:55goppLlorean nope
05:38:15Lloreangopp: Nope to what?
05:38:31goppoh shit
05:38:35goppgot in disk mode
05:38:40*Llorean said you would.
05:38:43gopphow do I mount it
05:38:49gopplinux cli
05:38:49LloreanLet it charge a bit.
05:39:08LloreanIf it was a low battery, you should make sure that's fixed before you go spinning the disk all around
05:39:25goppyea battery bit low
05:39:29gopphow long do I leav eit
05:39:59goppI have to do it all from cli
05:40:09Overandhcs- ultimately it's mostly square, tri, noise, etc
05:40:15OverandI mean, yeah, PCM channel and all of its fun
05:40:29gopphmm I was never around for c64
05:40:37OverandI still remember when there weren't any NES emulators that did the PCM channel
05:40:38hcsexpansion chips, FDS, also
05:40:43goppthe most advance system I used them was my nintendo
05:40:45Overandwhen NESticle started supporting it, it was a BIG DEAL.
05:41:09goppLlorean your a god
05:41:21OverandWhat about his "a god?"
05:41:34goppwhat command mount -t /dev/what do I use in linux
05:41:38LloreanI've seen a couple hundred people claim disk mode doesn't work, and every single time it's worked in the end.
05:41:50goppthe problem with this ipod
05:41:59goppis the hold switch and top bazzel is broken
05:42:01goppand missing
05:42:08goppso I made a litte object to press hold
05:42:13goppmy friend gave me it
05:42:23goppand it also low on battery
05:42:25OverandI *do* wish emer. disk mode wasn't 'required' just due to the transfer speed difference.
05:42:45Overandbut one can always reboot in iPod mode if need be
05:42:55Overand(As I did to copy the HSVC)
05:43:08goppdon't not dissconnect ipod
05:48:55tehsmohcs: yeah, is currently broken as well
05:49:12hcstehsmo: ah, through the diffs of the database we must go, then
05:49:12tehsmothe ipod says "database not ready" when I copy the created db files over
05:49:45hcsmy quest for tomorrow, perhaps
05:50:03tehsmohave you used vimdiff before?
05:50:27goppthis wierd arstric girl
05:50:31goppalways went crazy in high school
05:50:34goppis on myspace
05:50:37tehsmoit might help, it's a really nice way to visually see the differences betwene two files
05:50:37goppwow she in college
05:50:42goppShe was arstic but smart
05:50:44goppsavant like
05:51:00goppshe would have wierd mood swings and look like she had problems, but she got tops grades
05:51:19goppshe would go crazy in class, act out
05:51:30goppI just saw her on myspace
05:51:46LloreanDoes this have *anything* to do with Rockbox?
05:51:47goppalso came from a upper class family
05:51:50goppopps sorry
05:51:54goppMy rant
05:53:26goppLlorean how long do I need it so it charged
05:55:21LloreanUsually it's a good idea to give it quite a while, make sure it's well charged. You can probably use it earlier.
05:56:37goppwhat command do I use to mount -t /dev/
05:56:39goppin linux
05:56:50goppmount -t ntfs /dev/
05:57:19 Join AMDBCG [0] (
05:59:54 Join fejfighter [0] (
06:00:14gopphey one last question I have a ipod cig light charger
06:00:23goppwill that charge the ipod while I drive
06:00:29goppin rockbox
06:00:52AMDBCGquestion: how do you boot into the regular ipod firmware?
06:01:08AMDBCG(forgot the shortcut)
06:01:23scorcheAMDBCG: see the FAQ
06:01:55AMDBCGok (it WAS my only question though...)
06:02:11bagawkgopp: when I first saw that I thought you said you had a cigarette lighter for the ipod :P
06:03:06hcsthere's one built in, that's where all the power is going
06:03:32Mouser_XAh! That explains it all!
06:03:54goppa ac charger for cig ligther
06:03:56Mouser_XStupid undocumented features...
06:08:14*AMDBCG found the ipod faq
06:10:03*AMDBCG wonders how to get a good sound from the iPod mic :
06:20:01 Quit AMDBCG (Read error: 104 (Connection reset by peer))
06:24:30 Part safetydan
06:28:25 Join Mouser_X3 [0] (
06:28:34 Quit Mouser_X (Nick collision from services.)
06:28:39 Nick Mouser_X3 is now known as Mouser_X (
06:29:33 Join BHSPitMonkey [0] (
06:29:53 Quit excitatory (Remote closed the connection)
06:32:39 Join Rob222241 [0] (
06:35:01 Join piroko [0] (
06:36:52 Part Llorean
06:40:44 Quit heapcat ("CGI:IRC")
06:40:50 Quit piroko ("leaving")
06:42:35 Join CriamosAndy [0] (
06:50:59 Quit Rob2222 (Read error: 110 (Connection timed out))
06:58:50 Quit Criamos (Read error: 110 (Connection timed out))
06:59:07fejfighterjust a question about the irivers
07:00:01fejfighterwhat has changed in cvs to change the status of the eeprom driver here ( to 100%,
07:00:03fejfighterin particular for the H300
07:04:46fejfighteror realistically, what has cased the need for the change in the wiki
07:05:24 Join strabes [0] (
07:05:57***Saving seen data "./dancer.seen"
07:09:01 Join FOAD [0] (
07:14:53 Quit EspeonEefi ("Leaving")
07:17:32 Quit FOAD_ (Read error: 60 (Operation timed out))
07:24:24 Join dewdude_ [0] (
07:26:10 Quit mathgl ("Quitte")
07:32:12 Quit dewdude (Read error: 145 (Connection timed out))
07:36:01 Join Mouser_X3 [0] (
07:36:05 Quit Mouser_X (Nick collision from services.)
07:36:07 Nick Mouser_X3 is now known as Mouser_X (
07:40:13 Join decayedcell [0] (i=3ba749de@gateway/web/cgi-irc/
07:40:39 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
07:44:10 Quit jdong ("Ex-Chat")
07:45:33 Quit gopp ("gop: god giveths and democarts taketh away!")
07:46:28 Quit strabes ()
07:47:09 Quit fejfighter ()
07:50:51 Join Julian [0] (i=182cd144@gateway/web/cgi-irc/
07:51:10Julianwow alot of people
07:52:35Juliandoes anyone know if it possible to play video on a ipod nano with rockbox?
07:53:38scorcheit is
07:53:54 Join merbanan [0] (
07:54:46 Quit Febs (Read error: 110 (Connection timed out))
07:56:37 Join cheriff [0] (
07:57:30Juliando u know how?
07:58:00cheriffhello everyone, just dropped in to day that I installed the nightly build onto my iHP-120 and its going great! congratulations on a job being well done
07:58:45cheriffI havent used rb for several months - from when you couldnt simultaneously run a plugin + codec. Has come a long way
08:01:20Julianawesome thanks
08:01:33Julianrockbox is pretty cool!
08:06:10 Join ender` [0] (n=ender@
08:19:49 Join juxtap [0] (
08:43:45 Quit Julian ("CGI:IRC (EOF)")
08:44:20 Join dan_a [0] (n=dan-mirc@
08:46:36 Join safetydan [0] (
08:52:07 Quit Jsunu ("Chatzilla 0.9.77 [Firefox]")
09:00:12 Join kaaloo [0] (
09:00:38 Nick kclaf is now known as kclaf2K7 (
09:00:45 Join Mouser_X3 [0] (
09:03:14 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:06:06***Saving seen data "./dancer.seen"
09:17:55 Quit Mouser_X (Read error: 110 (Connection timed out))
09:18:13 Quit gtkspert (Read error: 110 (Connection timed out))
09:24:38 Quit CriamosAndy ("( :: NoNameScript 4.03 :: )")
09:25:06 Quit dan_a ()
09:29:43 Quit kaaloo ("Leaving.")
09:30:04 Join bluebrother [0] (i=QKSImARc@rockbox/staff/bluebrother)
09:30:21 Quit BigBambi ("Leaving")
09:42:09 Join GodEaterWeb [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:45:49 Join kaaloo [0] (
09:50:24 Quit Mouser_X3 (Read error: 110 (Connection timed out))
09:54:54 Quit Nimdae (Read error: 104 (Connection reset by peer))
09:54:56 Join Nimdae_ [0] (
09:56:33 Join lavid [0] (
09:57:40peturbah... stkov dircache :(
10:10:11 Join spiorf [0] (
10:12:55peturos stacks at a limit: dircache 95% and codec 97% (playing mp3 while checking)
10:13:51 Quit juxtap (Read error: 104 (Connection reset by peer))
10:20:55 Join BobJonkma1 [0] (
10:23:06 Join Ribs [0] (n=ribs@
10:26:23 Quit softi_42 (Read error: 60 (Operation timed out))
10:28:14 Join triplah_ [0] (
10:28:42triplah_are gen2 ipos the new smooth ones?
10:29:05GodEaterWebthey have the anodized aluminum cases if that's what you mean by "smooth"
10:29:17GodEaterWebI've not noticed that the 1st gen ones were particularly rough....
10:31:06triplah_hmm ok
10:31:15triplah_by smooth i meant rounded edges
10:31:25triplah_just want to make sure i can put rockbox on my nano
10:31:52GodEaterWebwell if buy one of the "smooth" ones - you won't be able to
10:32:11 Join softi_42 [0] (
10:32:36 Part BobJonkma1
10:36:48triplah_nah i've got got one of the old ones
10:37:42triplah_gtg :)
10:37:45 Quit triplah_ ("Leaving")
10:38:49 Join debauched_sloth [0] (
10:40:14 Quit BobJonkman (Read error: 110 (Connection timed out))
10:45:49 Join jba [0] (
10:49:01 Quit gotthardt ("Chatzilla 0.9.77 [Firefox]")
10:51:53amiconndebauched_sloth: Are you Greg?
10:52:01jhMikeShmmm...don't know why code exists in dircache.c to align audiobuf since it's always kept aligned after buffer_init and buffer_alloc calls. :\ I'm seeing if I can spot excessive stack usage too.
10:52:12debauched_slothyes, amiconn, that is I
10:53:23amiconnJust a remark regarding your 03:31 commit: It's better to name the target .h files just *-target.h instead of *-<target_name>.h
10:53:47amiconnOnly a minor thing atm, since the gigabeat is the only target featuring an mmu atm
10:54:06amiconnCan be renamed after the svn move, next week or so
10:54:21debauched_sloththank you, renaming in CVS is always a joy :)
10:55:25debauched_slothI'll rename the new one I just created
10:55:42debauched_slothI'll have a new patch with the changes we discussed last night in a few minutes here
10:56:30debauched_slothif you could take a look, I'd appreciate it
10:57:06 Part cheriff
10:57:17debauched_slothok, so the file should be just mmu-target.h?
10:57:20 Quit Nimdae_ (Remote closed the connection)
10:57:22debauched_slothfor example
10:57:34 Join Nimdae [0] (
10:58:24jhMikeSamiconn: Thought about that re: the hardcore splitting of playlist/audio api. I'm probably going off the deep end with that thought. :)
11:01:19amiconndebauched_sloth: yes
11:02:00debauched_slothgood, I'll regenerate that patch in a minute here
11:02:08amiconnThe .c files can be named per target, no problem with that. But naming the .h files per target would create unnecessary #include ifdefing
11:02:22debauched_slothgot it
11:02:42debauched_slothwhere are these target files included, so I caninsert mmu in there?
11:03:58debauched_slotham I correct in assuming there should be a system-target.h somewhere?
11:04:19amiconnThe include paths are also hierarchical, a .h file in firmware/target/<arch>/<manuf>/<model> overrides one in firmware/target/<arch>/<manuf> with the same name, and one in firmware/target/<arch>/<manuf> in turn ovverrides one in firmware/target/<arch>
11:05:15debauched_slothok, so I find system-target.h in coldfire, but none in ARM
11:05:34amiconnYes, because there is no general system-target.h for arm
11:05:36debauched_slothmy changes are ARM specific, but dependant on an ARM9 CPU
11:05:47jhMikeSnot sure what the objection is to have a single xxx-target.h with the #ifdef-ing in there and other source can just include that
11:06:06debauched_slothand reboot() is pretty much gigabeat specifc, since it depends on the S3xxx chip
11:06:10***Saving seen data "./dancer.seen"
11:06:28debauched_slothI'll create a system-target.h in the final directory? in meg-fx?
11:06:40amiconnYes I think that's the best location atm
11:06:49debauched_slothand can I include mmu.h in there?
11:07:07debauched_sloththat one should see very limited play outside of the target, I should think
11:07:10amiconnIn case other targets use the same cpu, it can be moved / adapted later
11:07:26debauched_slothwould prefer to keep new stuff in our branch of the target tree where possible
11:08:18amiconnThe system stuff isn't yet moved to target tree for the arm targets it seems...
11:08:21*jhMikeS doesn't wanna hear nothin' about some target with a MMU...just bad timing ;)
11:08:38debauched_slothamiconn: any formatting requirements for copyright?
11:08:43debauched_sloth(C) 2006-2007
11:08:47debauched_sloth(2) 2006,2007
11:08:53debauched_slothI've seen both used
11:09:09amiconnNo special requirements I know of
11:09:36debauched_slothjhMikeS: you know the week that MMU has given me? ;)
11:09:47debauched_slothI don't want to hear of it either, not any more
11:10:36debauched_slothnot seeing any includes of system-target.h, where do I turn that on
11:10:41debauched_slothwill grep
11:11:57amiconnsystem.h includes it
11:12:08debauched_slothyes, inserting there now
11:12:18debauched_slothneed to move the asm out of there anyway
11:13:40jhMikeSdabauched_slot: no, but I'm thinking it could maybe make some playback stuff easier to accomplish
11:14:33jhMikeSoy my typing, sorry debauched_sloth :)
11:14:54debauched_slothjhMikeS: how? at a conceptual level, yes, but if you split our buffers up, you could really make DMA a joy to behold, since it is at the phys, not virt level
11:15:10debauched_slothright now DMA is a very simple thing to do, and that is very nice
11:16:56debauched_slothok, commited new system-target.h
11:16:57jhMikeSwas thinking it would be nice to mirror the audio buffer next to itself to handle wrapped requests more easily. but that's dreamland I guess.
11:16:58 Join My_Sic [0] (
11:17:26debauched_slothat the higher level, yes, you could do that, but at some point someone has to reassemble them
11:17:26 Quit My_Sic (Read error: 104 (Connection reset by peer))
11:17:30 Join My_Sic [0] (
11:17:40debauched_slothmaybe not so much of a complication, we have to kick off a series of DMAs today
11:19:42debauched_slothyesterday's battery bench ran for 15 hours
11:19:45 Join Wiwie [0] (
11:19:47jhMikeSI'm thinking that codecs with special needs will have to be wrap aware and do the assembly themselves
11:19:54debauched_sloththree days running: 11, 13, 15
11:20:09 Join MarcoPolo [0] (
11:20:19debauched_slothI'm thinking that reassembling packetized data is probably not so bad, if each packet is of a reasonable size
11:20:37debauched_slothI mean, if you are thinking 32 bytes, I have issues - 4K, better
11:21:11amiconnI think it's simply out of question for some non-streaming formats (i.e. mods)
11:21:20jhMikeSWell I was thinking about segmenting the buffer but don't want to waste a lot of memory
11:21:43debauched_slothnon-streaming because you are splitting across packets?
11:21:47amiconnJust make the buffering wrap earlier if the format requires it
11:21:57amiconnNo memory wasted under normal conditions
11:22:56linuxstbjhMikeS: What's the problem with the current method of buffering?
11:23:01 Quit jba (Read error: 110 (Connection timed out))
11:23:14jhMikeSThat's the rub: when and how much since anything can be there. I can't see a reason a looping file can't be larger than memory.
11:24:08 Join jba [0] (
11:24:10amiconnjhMikeS: A simple looping file format can, but a mod file cannot.
11:24:20debauched_slothok, I've updated my patch,
11:24:25amiconnThe replayer needs _random_ access to the whole file, without wrap
11:24:35debauched_slothstill not done cleaning up, but the general drift of the changes we discussed should be in there
11:24:37jhMikeSlinuxstb: If a codec needs contiguous memory larger than the guard size, it can't get it for one
11:24:48jhMikeSunless the request doesn't wrap
11:25:18debauched_slothsince we are discussing playback, who should I ask about playback in general?
11:25:39debauched_slothI am interested in making a couple of mods to make this the best audiobook player out there
11:26:02amiconnjhMikeS: Hence my idea to make the guard buffer size dynamic, and let the track loader decide which size is needed.
11:26:31debauched_slothheh, you could use guard pages, let the track loader assume it has infinity
11:26:40debauched_slothsorry...I'm MMU happy this week :)
11:26:59debauched_slothit won't last, I assure you
11:28:02linuxstbjhMikeS: Daling with non-streaming codecs is not a good reason for changing the current method for streaming codecs IMO.
11:28:05amiconn0 byte for simple pcm formats, some small size (e.g. frame size) for compressed streaming formats, and the whole file size for non-streaming formats (which need it)
11:28:07 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
11:28:44amiconnHandling midi present an even bigger problem though....
11:29:33 Join trypt0 [0] (
11:30:34jhMikeSlinuxstb: I'm thinking of how to make them both happy at the same time. The mod codec may have to deal with things in a non-flat manner AND I don't want to change things for the streaming codecs.
11:30:39 Join Mouser_X3 [0] (
11:31:17amiconnVariable guard buffer size doesn't really change things fopr streaming codecs...
11:31:31debauched_slothhow big are the midi files we are talking about, anyway? I have no experience with super large midi files
11:31:43amiconnHandling mod non-flat is out of question imo
11:31:56linuxstbI'm not sure we even need to go as far as a variable guard size - just flag certain codec types as needing a contiguous buffer, and load them in that way.
11:32:18amiconndebauched_sloth: The problem isn not that the files are large, in fact they are small
11:32:54debauched_slothI'm not seeing the problem, then - the output buffers are (conceptually) linear
11:32:55jhMikeSWould like to see a mod file decoder and know exactly why non-flat is out of the question.
11:33:15amiconnBut playing midi files requires a synthesizer, and for half-decent sound it needs a set of instrument samples (the patchset)
11:33:55amiconnThis patchset can be quite large, and it needs to be loaded somewhere
11:34:10debauched_slothbut, again, the output buffers (and I apoligize if I am jumping in late here) are still conceptually linear
11:34:13 Join trip_ [0] (
11:34:18debauched_slothexcuse typos, early here
11:34:29trip_anyone know how to get banshee to treat rockbox as a normal DAP?
11:34:34jhMikeSyes, that's one point I had in mind but we're not doing DMA transfers directly out of the filebuf
11:35:10linuxstbdebauched_sloth: We're talking about the input buffer - i.e. the large (e.g. 28MB) buffer containing the compressed audio data.
11:35:16debauched_slothah, got it
11:35:49debauched_sloththe one that freaked me out when I DMA'ed the ATA reads since it drained down alarmingly
11:38:15 Join webguest08 [0] (i=c25a1226@gateway/web/cgi-irc/
11:38:22 Quit hcs (Nick collision from services.)
11:39:48 Quit webguest08 (Client Quit)
11:41:19 Quit Wiwie (Read error: 110 (Connection timed out))
11:41:34jhMikeSIf the patch set is larger than memory you're hosed though, that's true. That bit must be flat of course or at least wrapping should occur on a sample boundary only. Point taken.
11:46:23*safetydan listens to high pitched daemon screams in doom
11:46:36safetydanI don't think my pcm_set_frequency stuff is working in the sim
11:46:47safetydanThough the screams are rather amusing
11:46:48jhMikeSMixing up mp3s/wavs etc. with that stuff is a bit of a problem. In that case the buffer should have another mode of operation where the codec gets to control it as if it stole it.
11:47:32jhMikeSsafetydan: samplerate running a bit high? :)
11:47:33trip_so no one knows about using banshee with rockbox?
11:47:53 Nick trip_ is now known as triplah_ (
11:48:01safetydanjhMikeS, just a tad
11:49:02amiconnjhMikeS: Exactly, and the same applies to module formats; they can be compared to a midi file with the patchset included
11:55:19debauched_slothamiconn, linuxstb: any comments on that patch set?
11:55:21jhMikeSThen playback doesn't buffer the file, the codec does. Playback just starts the codec and gives all the resources to it. It can decide how it wants to go about things.
11:56:40jhMikeSBTW: How much guard buffer is needed for a streaming format like mpg video?
11:58:00amiconnGiving the whole buffer to the codec would be bad. This way it would not be (easily) possible to buffer multiple non-streaming tracks, or a mix of streaming and non-streaming
12:00:02 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
12:00:10 Join barrywardell [0] (n=barrywar@
12:01:38 Quit Mouser_X3 (Read error: 110 (Connection timed out))
12:02:28jhMikeSamiconn: Do we need gapless playback between mp3 and say MIDI? If I let a MIDI codec have the whole buffer it can load it's patches once and then it can lookup and load all the files in a row marked for that codec. It seems quite handleable in that way when I think about it.
12:03:37jhMikeSFor others with patches included, it can just load next file into the buffer flat or fail if it doesn't fit.
12:03:55*linuxstb adds a FPS table to the PluginMpegplayer wiki page and invites people to post test results with the Elephants Dream sample videos
12:04:36LloreanHow big is the midi patchset currently?
12:04:46amiconnjhMikeS: We certainly don't need gapless, but we still want to minimise disk spinups
12:05:04LloreanIsn't the full set larger than 16mb?
12:05:08amiconnit is
12:05:19amiconnWe need a smaller set for the h100 and x5
12:05:31LloreanCould we compress it?
12:06:04linuxstbI thought the midi codec was already struggling with speed?
12:06:31amiconnThe on-disk size is irrelevant, and in memory it needs to be uncompressed
12:07:30jhMikeSamiconn: sure, me too each large file as big as the buffer will require spinups anyway and no part of the current one is discardable
12:08:22amiconnnon-streaming != big
12:08:42amiconnAnd mixing streaming and non-streaming should be possible
12:09:19amiconnImho it should be possible to play e.g. .mod - .mp3 - .mod without intermediate spinups
12:11:03 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
12:11:12 Join miepchen^schlaf [0] (
12:14:44 Quit barrywardell ()
12:16:33 Join barrywardell [0] (n=barrywar@
12:19:06debauched_slothcvs is complaining about conflicts existing, but I am just not seeing any
12:19:11jhMikeSBut I'm thinking of MIDI as well where the (likely somewhat large) patchset should only be loaded once for a train of midis. Not sure I want to take a significant hit in the core for that. I suppose some attributes must be defined for what to do when the file is now current.
12:19:14debauched_slothanyone have any experience with that?
12:20:03jhMikeSI've gotten plenty of conflicts. I just search for the <<<<<<<. Doing another cvs up should show a "C" next to the file.
12:20:26debauched_slothit shows as modified not conflicting
12:20:40debauched_slothand I see no <<<<or >>>>
12:22:26jhMikeSdoing what exactly? I usually cvs up right before committing things to make sure I'm current.
12:22:33 Join dpro [0] (
12:24:48 Join Angry [0] (
12:25:44 Part linuxboy
12:26:56debauched_slothah, now I see it
12:27:03debauched_slothSOMETHING in my commit set is modified
12:27:06debauched_slothwill go looking
12:27:19debauched_slother. conflicting, of course
12:36:40 Quit GodEaterWeb ("CGI:IRC")
12:36:45 Join GodEaterWeb [0] (i=c2cbc95c@gateway/web/cgi-irc/
12:36:49 Quit Alexinc ()
12:43:47 Join Nico_P [0] (
12:46:08 Join Arathis [0] (
12:46:18debauched_slothwhen does the automated build process get kicked off?
12:46:25debauched_slothis there some flood control?
12:47:55GodEaterWebon the channel ?
12:48:07debauched_slothon CVS
12:48:14GodEaterWeboh - no clue
12:48:23GodEaterWebI wouldn't have thought so though
12:48:27GodEaterWebgiven the limited commit access to it
12:51:02 Quit safetydan ("Ex-Chat")
12:51:39 Quit decayedcell ("CGI:IRC (EOF)")
12:53:45jhMikeSOk, my take is to keep the logic in the core simple and small as possible, not to be hyper paranoid about disk spinups but I can see from google-ing that a .mod tends to be small and will not often need flattening. Those MIDIs are another story though and will likely not get mixed with other stuff. I guess the only thing left is to figure out how to make playback do _everything_ anyone can think up to throw at it. :)
12:55:21debauched_sloththanks all for the help with the patch - especially amiconn and linuxstb. I'll not be such a pest moving forward, just wanted to make sure I was being a good citizen
12:57:06 Quit MarcoPolo (Remote closed the connection)
12:58:51debauched_slothwhoops, broke some stuff, fixing
13:02:44 Quit Nico_P (Remote closed the connection)
13:03:45 Join tj2master [0] (i=a@
13:06:14***Saving seen data "./dancer.seen"
13:06:57 Quit GodEaterWeb ("CGI:IRC (EOF)")
13:07:16jhMikeSdebauched_sloth: heh...happens all the time...sims hate me the most
13:07:41debauched_slothpoor gotthardt spent a few hours fixing it last night, too
13:08:44 Join hannesd [0] (
13:09:22jhMikeSno worry, when I updated recording my score was almost 10k :P
13:09:36debauched_slothheh, I have 100 servers at work
13:09:44debauched_slothmight help the build speed a touch?
13:09:58debauched_slothmight upset everyone if Iinstalled tools on them, too...alas
13:12:57jhMikeSwell, one build per server would be nice. won't need more servers than builds.
13:14:21 Quit Nimdae (Read error: 60 (Operation timed out))
13:14:36debauched_slothnext try for the build
13:24:02 Join My_Sic [0] (
13:26:11 Join funky [0] (n=repulse@unaffiliated/funky)
13:26:11 Quit My_Sic (Read error: 104 (Connection reset by peer))
13:28:10 Quit Rob222241 ()
13:31:12debauched_slothoko, better
13:37:01 Join muesli__ [0] (n=muesli_t@
13:40:07 Quit jba (Read error: 110 (Connection timed out))
13:40:47 Part tj2master
13:42:02 Quit zylche (Read error: 104 (Connection reset by peer))
13:42:16 Join zylche [0] (
13:43:19 Join Mouser_X3 [0] (
13:44:51preglowno one can accuse the gigebeat port of stalling
13:44:54bluebrotherwow, Greg only on the recent cvs table :)
13:46:08linuxstbAll in the space of 63 minutes...
13:46:55debauched_slothsorry guys :) those have been queing up for a while
13:47:21bluebrotherlooks like there has been some hidden code somewhere ;-)
13:47:52debauched_slothmarkun was doing the commits, but he's off to Paris and is busy so we've been sort of stuck
13:48:13bluebrothernice to see the progress
13:48:32 Join Nico_P [0] (
14:00:21peturdebauched_sloth: did Bagder also give you FlySpray access? I think should get closed, no?
14:02:58debauched_slothdon't seem to have it, tried to login, no dice
14:03:07debauched_slothyes, that should be closed
14:04:23 Quit Ribs ("eh eh ehhhh!")
14:05:28 Nick Mouser_X3 is now known as Mouser_X (
14:14:53 Quit barrywardell ()
14:16:05 Part Angry
14:16:30 Join adam1302 [0] (i=cbab4195@gateway/web/cgi-irc/
14:16:58 Part adam1302
14:20:08 Join MarcoPolo [0] (
14:22:37 Quit zylche ("-")
14:27:17 Join norbusan [0] (
14:30:46 Quit akaias (Remote closed the connection)
14:32:56 Join Rob2222 [0] (
14:34:07 Quit petur ("stkov")
14:38:16 Part debauched_sloth
14:43:09 Part ScoTTie
14:46:27 Join Everybody| [0] (
14:47:05 Join BigBambi [0] (
14:48:47 Quit Mouser_X (Read error: 110 (Connection timed out))
14:57:26 Quit Rob2222 ()
14:58:03 Join anewuser [0] (i=anewuser@unaffiliated/anewuser)
15:00:05 Join Rob2222 [0] (
15:01:04 Quit Everybody|away (Read error: 110 (Connection timed out))
15:03:13 Quit kaaloo ("Leaving.")
15:06:15***Saving seen data "./dancer.seen"
15:06:24 Join My_Sic [0] (
15:07:33 Quit Everybody| (Read error: 110 (Connection timed out))
15:11:28 Join akaias [0] (
15:12:40 Part norbusan
15:19:45 Join phrozen77 [0] (
15:32:40 Join Febs [0] (n=chatzill@rockbox/administrator/Febs)
15:34:03 Join debauched_sloth [0] (n=gwhite@
15:36:58Febslinuxstb: you around?
15:37:45 Join Wiwie [0] (
15:44:21 Quit Wiwie ("Miranda IM! Smaller, Faster, Easier.")
15:53:43 Join habana [0] (i=58a10615@gateway/web/cgi-irc/
15:56:02 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
15:57:47 Join norbusan [0] (
15:59:17 Quit spiorf (Remote closed the connection)
16:01:05 Join gamkiller [0] (
16:01:25 Quit Rob2222 ()
16:02:32 Join BigBambi_ [0] (
16:02:32 Quit BigBambi (Read error: 104 (Connection reset by peer))
16:02:52gamkillerHello lads. Lately, my iRiver iHP120 has been freezing when I select a song to play. I have to wait for the battery to go flat before I can "try again". I was wondering if this might be a known problem in later versions of rockbox, or if it might be a hardware fault?
16:05:25bluebrotherthere is a reset button at the bottom of the player
16:05:35bluebrotherit's a small hole, use a paperclip to press it.
16:05:50bluebrotheralso, try updating to the latest version and reset your settings
16:06:58gamkillerbluebrother: Yes, um haha.
16:07:01gamkillerI tried the reset button
16:07:08gamkillerthe pin I used slid right in to the hole.
16:07:19gamkillerI cannot seem to locate the actual button in there.
16:07:31bluebrotherit's a bit tricky to use.
16:07:31gamkillerMaybe they forgot it in mine ^_^
16:07:50bluebrotherthen the only way to find out would be opening it up and look
16:07:56gamkillerbluebrother: But yeah, I'll try the latest version when I get home from work.
16:08:00bluebrotheranyway, my h120 works like a charm.
16:08:17gamkillerYeah I've been using rockbox for as long as I can remember on it
16:08:36gamkillerIt makes me all and fuzzy inside :).
16:09:53 Join zylche [0] (
16:11:23 Join Criamos [0] (
16:12:58 Join perl|work [0] (
16:21:36 Join BigBambi [0] (
16:22:53 Quit BigBambi_ (Read error: 104 (Connection reset by peer))
16:23:22 Join tj2master [0] (i=a@
16:29:37 Join Mouser_X3 [0] (
16:34:53 Join juxtap [0] (
16:34:55 Part norbusan
16:36:18 Join Arathis_ [0] (
16:38:02habanayour preffered tool to clean a sansa partition (linux/windows) ? i dont manage to repair file errors today and numerous files are broken:(
16:39:00 Quit tvelocity ("Ī‘Ļ€ĪæĻ‡ĻŽĻĪ·ĻƒĪµ")
16:40:16 Quit Mouser_X3 (Read error: 145 (Connection timed out))
16:50:11 Quit Arathis (Read error: 110 (Connection timed out))
16:50:30 Nick Arathis_ is now known as Arathis (
16:52:24 Join GodEaterWeb [0] (i=c2cbc95c@gateway/web/cgi-irc/
16:53:26 Join x1jmp [0] (
16:53:58GodEaterWebhabana: I'd say my favourite linux tool for cleaning any partition with linux is the "rm" command
16:54:52 Join Ghoulunatic [0] (
16:55:00 Join _Veseliq_ [0] (n=veseliq@
16:58:15 Quit MarcoPolo (Remote closed the connection)
16:58:24 Join x1jmp_ [0] (
17:01:50 Join norbusan [0] (
17:03:43 Quit trypt0 ("Leaving")
17:05:15 Join PaulPosition [0] (
17:06:18***Saving seen data "./dancer.seen"
17:06:49 Join tvelocity [0] (
17:10:10 Join My_Sic [0] (
17:11:31 Quit norbusan ("Leaving.")
17:11:53 Join norbusan [0] (
17:12:07 Quit x1jmp (Read error: 110 (Connection timed out))
17:14:19 Quit norbusan (Client Quit)
17:14:38 Join norbusan [0] (
17:14:55 Join spiorf [0] (
17:17:24 Nick x1jmp_ is now known as x1jmp (
17:17:38 Part norbusan
17:18:09 Join norbusan [0] (
17:18:17 Part tj2master
17:18:54Arathisis there any possibility to make a static link on the newest file of a sourceforge project? so you don't need to edit it every time there is a new version
17:19:49habanaGodEaterWeb: lol, why not rm -rf ? way more clean. But "rm" alter files not partition, i presume
17:22:41habanafinally, deleted/recreated the partition.
17:23:13habanaCould you point me a good source for flash spec ?
17:25:59 Quit BigBambi (Read error: 104 (Connection reset by peer))
17:38:06 Part norbusan
17:39:37 Quit habana ("CGI:IRC (EOF)")
17:44:04 Join strabes [0] (
18:04:25GodEaterWebamiconn : Can you check on the value of word 64 again - I'm not sure on the ATA-7 spec's meaning when it says "If the serial interface is implemented, bits (1:0) shall be set to one." Does this mean that usually devices which support both PIO modes 3 AND 4 only set bit 1, and leave bit 0 as zero ?
18:04:58 Join trypt0 [0] (
18:05:24 Join votetrev [0] (
18:05:27votetrevhey guys!
18:05:51 Join lowlight [0] (i=c730180b@gateway/web/cgi-irc/
18:07:46GodEaterWebalso - I can't see where it says a value of 00 is reserved in bits 7:0 in word 47. It just says "if the serial interface is implemented, this field shall be set to 16 or less."
18:09:12 Quit ender` (" A computer program will always do what you tell it to, and seldom what you want it to.")
18:10:05GodEaterWeboops. just realised I was reading the PACKET device identiy device info - ignore me
18:10:11votetrevanyone in here that could help me with a question?
18:10:18 Join lee-qid [0] (
18:11:03 Quit lee-qid (Client Quit)
18:11:13GodEaterWebor possibly just not the right bit of the table. Sigh. I may need glasses.
18:11:20 Join lee-qid [0] (
18:11:28GodEaterWebvoterev trick to say until we know the question
18:11:44 Quit Arathis (Read error: 54 (Connection reset by peer))
18:11:56votetrevif i want to use my regular apple os AND the rockbox OS with my same music....what is the best possible way to do this?
18:12:39GodEaterWebuse Rockbox's database feature
18:12:57bluebrotheruse itunes to satisfy the apple os (or any other apple db tool) and Database for Rockbox
18:13:18votetrevany links on how to use rockbox's database feature?
18:13:41votetrevyeh i downloaded that...
18:13:42*GodEaterWeb didn't think the manual was so hard to find
18:14:03votetrevbut when i did what it told to on there.....i still cant quite figure out where to even go to play music in rockbox i guess
18:14:07 Join ender` [0] (n=ender@
18:14:40GodEaterWebif you can't do that after reading the manual, then perhaps rockbox is not for you
18:15:07votetrevlol :P i know im starting to feel that way
18:15:19votetrevbut i am by no means a beginner
18:15:19 Join charles [0] (i=5743add3@gateway/web/cgi-irc/
18:15:26votetrevim wondering if something didnt go right with my flash
18:15:29bluebrotherthen the manual should have all information needed for you
18:15:31votetrevwell my flash went right....
18:15:41votetrevbut i think maybe when i tried to add a theme it messed something up and i am unaware of it
18:15:59charlesIs there anyone that can help me with linked lists....
18:18:07GodEaterWebvotetrev: I find it hard to believe that adding a theme would somehow stop you playing music
18:18:11 Join gotthardt [0] (i=86868804@gateway/web/cgi-irc/
18:18:31x1jmpI made a modification to rockbox's dict plugin, it's very useful to read wikipedia on rockbox devices
18:18:37 Join nuge [0] (n=ideas@unaffiliated/nuge)
18:18:50x1jmpI'd like to get some feedback, a description is on misticriver at
18:19:27votetrevgodeaterweb.....before the theme though i got right into the music and it was playing showing the music information/ when i try to bring it up after loading a new theme no matter what i cant get any music going...?
18:19:38votetrevi dont know ill keep messin around on it and let u guys know what i figure out :P
18:19:43GodEaterWebx1jmp: you'd be better off putting your post on rockbox's forums
18:20:15GodEaterWebvotetrev: are you using a standard rockbox build, or one of the evil album art ones ?
18:20:31x1jmpis it more frequently visited than MR?
18:21:07charlesIs there anyone with coding knowledge around here with a few spare minutes to help me out?
18:21:22GodEaterWebcharles: what are you working on ?
18:21:31votetrevcharles what language?
18:21:54charlesIm trying to make a file browser closer to the windows explorer
18:21:56votetrevi know a LITTLE on C using a bleeding edge rockbox build, the most recent one
18:22:15bluebrothercloser to windows explorer? Why that?
18:22:23charlesWell I have a good knowledge of some other languages , just took a crashcourse on C on the net...
18:22:40charlesWell I figured it would be nice to be able to have icons and such
18:22:42GodEaterWebwhy does rockbox need another file browser ?
18:22:48GodEaterWebit already has icons
18:23:05 Join Arathis [0] (
18:23:06charlesYeah, well, lets just say I don't like it so much ;)
18:23:32bluebrotherI guess most people (and especially devs) won't like a icon-browser
18:23:51charlesI suppose so
18:23:56charlesThats why I'm trying to make it as a plugin
18:24:18*GodEaterWeb hmms loudly - the ATA-7 spec appears to be confused about bit 13 in word 106
18:24:22 Join Wiwie [0] (
18:24:32charlesBut anyway, thats not really important, the important thing is that apparently you can't dynamically allocate memory within rockbox
18:24:38bluebrotherI think it would be better to make it possible using bigger icons in the browser as it is now −− when using large fonts the icons are tiny
18:24:44charlesSo I'm wondering how to make linked lists work
18:24:52charlesYes exactly
18:24:57bluebrotheryou need to allocate the memory static.
18:25:10GodEaterWebin other words - there IS no way to do link lists in rockbox
18:25:17charlesAlso I think on iPods and other big screen devices it would be a great benefit to be able to show files one next to another
18:25:23bluebrotherbut you can steal the plugin buffer
18:25:31bluebrotherwhich of course won't work with a plugin :)
18:25:37 Quit lowlight ("CGI:IRC (EOF)")
18:25:38charlesBluebrother, how would something like that work?
18:25:47charlesI'm still very new to C as I said =)
18:25:54GodEaterWebAmiconn: read page 49 in the ATA-7 spec
18:26:03nugewill nanos (2nd gen) be supported in the (near or distant) future? if so, is there somewhere i can track the progress of the port?
18:26:09bluebrotherthere are functions to do so. But I'm not familiar with it.
18:26:25bluebrothernuge: no, and most likely no.
18:26:26GodEaterWebnuge: very unlikely - no-one is working on it
18:26:30charlesWell thanks anyway...Is there anyone who is familiar with them?
18:26:38nugeok thanks, too bad :/
18:26:38bluebrotherat least, don't hold your breath −− the 2g are completely different
18:26:50 Join lowlight [0] (i=c730190b@gateway/web/cgi-irc/
18:26:53GodEaterWebcharles: hang around here a while longer for some of the more experienced devs to show up
18:26:58charlesnuge, I would try if I were you
18:27:03charlesThey support 2g
18:27:07GodEaterWebno they dont'
18:27:13votetrevthere is no way to run ipl on 5.5g yet is there?
18:27:13bluebrothernano 2g? I don't think so
18:27:14nuge1st gen only i think
18:27:15GodEaterWebthey're stuck at the same point we are
18:27:20charlesOh well - I thought so
18:27:24GodEaterWebvotetrev: 30GB works fine
18:27:25charlesI wasn't sure - sry about that
18:27:30votetrevoh yeh?
18:27:31GodEaterWeb80GB is what I'm working on :)
18:27:35votetrevshit i could not for the life of me get it going
18:27:38bluebrothernano 2g is completely different in terms of hardware. Which is the problem
18:27:45votetrevi wanted to triple boot with ipl, rockbox, and apple os
18:27:46GodEaterWeband it's firmware is encrypted
18:27:51votetrevbut couldnt get it going right
18:27:57bluebrotherat least one of the major problems. Another is the fact that the firmware is completely encrypted
18:28:07GodEaterWebthere's definitely an echo in here
18:28:24bluebrothermaybe I'm a bit slow atm.
18:28:30GodEaterWebs'okay :)
18:28:45GodEaterWebright - I think I'd best be off home
18:28:49votetrevgodeaterweb how tough is it to get all 3 os's running on a 30gb 5.5g?
18:28:50maquisany ideas for why they encrypted the 80GB but not the 30GB? (not that i'm complaining...)
18:28:55bluebrotherI guess I'll leave now to get some food and return later. Let's see if the echo is still there then :)
18:29:00GodEaterWebvotetrev: no clue I've never tried
18:29:06votetrevdont care for ipl?
18:29:09*GodEaterWeb vanishes in a puff of http
18:29:12 Quit GodEaterWeb ("CGI:IRC")
18:29:18bluebrothermaquis: they encrypted the nano2g, not the 80gb 5.5g
18:29:21votetrevi dont really know why i would use it either...just wanted to try
18:29:31bluebrotherthe 80gb 5.5g has issues with ATA.
18:29:36maquisbluebrother: ah...
18:29:56maquisbluebrother: hopefully that doesn't mean that they're planning on encrypting future releases...
18:29:59lowlightcharles: check the plugin api for plugin_get_buffer & plugin_get_audio_buffer
18:30:00maquisit's not very nice
18:30:04 Join Mouser_X3 [0] (
18:30:13bluebrotherand to be precise, the ipod firmware has been encrypted since the start. The only difference is that previously the bootloader wasn't
18:30:35bluebrotheranyway, back later.
18:30:38charlesOk thx lowlight
18:31:08maquishonestly, from my pov, the worst feature of the ipod original firmware is the fact that you have to use itunes / gtkpod to add music rather than just using cp
18:33:10 Join Rob2222 [0] (
18:33:27votetrevgodeaterweb...can you have rockbox create a database for you with a plugin or something similar or does it have to be done on your computers end of it?
18:34:09charleslowlight: You don't happen to know a bit of code where this is used , so that I could figure out how to use it myself?
18:35:05Nico_Pguys... anyone want to give me impressions on ?
18:36:01charleslowlight: Nevermind. Found it.
18:37:01 Quit _Margot_ (Read error: 104 (Connection reset by peer))
18:37:02 Quit blue_lizard (Read error: 60 (Operation timed out))
18:37:29 Join blue_lizard [0] (
18:37:32lowlightcharles: jpeg plugin uses both
18:38:28charlesYeah thanks, I just found that out too
18:38:45charlesWell I gotta go study again...freaking exams ;-) See ya all folks and thanks.
18:41:59 Quit charles ("CGI:IRC (EOF)")
18:42:29 Join bluey- [0] (
18:43:00 Quit votetrev ()
18:46:17 Join RedBreva [0] (
18:46:35 Quit Nico_P (Remote closed the connection)
18:47:28 Quit RedBreva (Client Quit)
18:47:37 Quit trypt0 (Read error: 110 (Connection timed out))
18:49:51 Join MarcoPolo [0] (
18:50:25 Join Dr4g0n [0] (
18:50:53 Join RedBreva [0] (
18:50:55 Join Angry [0] (
18:53:25AngryIs there any possibility to trace values over the lcd display??? the logf only makes sense with a remote
18:54:47AngryI suppose no answer means stupid question ;-)
18:56:42 Quit Arathis ("[rl_bot quit]")
18:58:21 Join hcs [0] (n=hcs@rockbox/contributor/hcs)
18:59:27 Join Thundercloud [0] (
18:59:58 Join Arathis [0] (
19:00:30 Join mathgl [0] (
19:00:51 Join GodEater [0] (
19:04:59 Quit hcs ()
19:06:19***Saving seen data "./dancer.seen"
19:06:55 Join DreamThief|off [0] (
19:07:05 Nick DreamThief|off is now known as DreamThief (
19:07:30lowlightAngry: it's been suggested in the past to us a subsection of the main lcd screen for logf output...but so far no one is up to the task
19:08:13Angryah ok.. so i'll continue try and errror^^
19:09:36lavidIs anyone here who is working on the Sansa port? I'd like to help with the DAC. Would looking at the way the DAC is used in the OF be a good place to start?
19:13:50lowlightlavid: the dac seems to have been id...
19:23:43 Quit bluebrother ("leaving")
19:23:57 Join bluebrother [0] (i=khjBglfv@rockbox/staff/bluebrother)
19:28:35 Quit bluey- ("Leaving")
19:31:51lavidlowlight: okay, so where do I/we go from there? I have a good deal of assembler programming experience, some C, but no embedded C. I'd really like to help, as I mentioned
19:32:18 Join Araknis [0] (i=48aeac7b@gateway/web/cgi-irc/
19:33:01 Quit strabes (Remote closed the connection)
19:33:49 Quit Mouser_X3 (Read error: 110 (Connection timed out))
19:33:57lowlightsee if you can get a datasheet and figure out how to use it
19:35:07lavidlowlight: I found a product brief by googling the chip, but I assume you're referring to a full blown data sheet, right?
19:35:58*lavid grumbles
19:36:40lowlightthere's a request form on the website...I thought someone got one, but others were denied
19:37:23lavidyeah, I see the reference to someone getting one. Do we know who that was?
19:39:07 Join Rondom [0] (
19:39:21lowlightdon't know, maybe MrH
19:40:45lexi'm maybe on high
19:40:57lexi accidentally bought one month subscribtion to wow
19:41:53GodEaterah, the even more addicitve alternative to evercrack
19:42:06GodEaterkiss your social life goodbye
19:42:10GodEaterand your suntan
19:44:21 Quit Wiwie ("Miranda IM! Smaller, Faster, Easier.")
19:48:09lavidlowlight: okay, I submitted a request for the datasheet. Let's see how quickly I get denied. In the interim, would looking at the OF's usage of the DAC be helpful?
19:50:51GodEatercan't hurt
19:52:17lavidadditionally, looking at the diagram for the as3515, it seems like it has the battery charger as well as the backlight
19:53:24lavidbut I thought that was all on the PP chip
19:54:06 Join trypt0 [0] (
19:56:14linuxstbFebs: ?
19:58:08FebsI wanted to check to see if you had seen forum thread about people having difficulty booting to the Apple firmware with the new bootloader.
19:58:30FebsI have the same problem.
19:58:35linuxstbOn what ipod?
19:58:39Febs5g 60GB
19:58:56FebsRecall, however, that I installed the new bootloader directly from the old, not from a clean firmware.
19:59:17FebsI haven't had a chance to restore the Apple firmware to see if that makes a difference.
19:59:28linuxstbAh, I thought the problems were only affecting the older, greyscale ipods.
20:00:19FebsIt took me a while to notice the issue. Apparently I don't use the Apple firmware much! ;)
20:01:46linuxstbHave you seen other reports of 5g problems?
20:02:21FebsNot that I can recall.
20:06:21lowlightlavid: I think the assertion is that the dac parts are similar between the 2 chips...and of course you could try to look at the disassembly of the OF to figure out how to use the dac
20:08:03 Join BobJonkman [0] (
20:08:11 Part BobJonkman
20:09:28 Join Everybody|away [0] (
20:09:37lavidlowlight: okay, thanks. I'm going to start poking around
20:10:02 Join trypt0_ [0] (
20:12:34lowlightgood luck
20:12:42lavidha, thanks
20:13:40lavidin terms of seeing what others have found on this front... where should I look? I have daniel's page, the forums and this chatroom... anything else?
20:15:06 Quit phrozen77 ("Some people say that I must be a horrible person, but that's not true. I have the heart of a young boy... in a jar... on my d)
20:15:36lowlighttalk to dan_a and barrywardell, they are doing most of the sansa developing
20:15:44 Join Vyrus [0] (
20:15:45 Quit Vyrus001 (Read error: 54 (Connection reset by peer))
20:15:56 Join cynicalliberal [0] (
20:16:17lavidneither of which are on, I see.
20:16:23lavidokay, thanks again
20:18:25 Quit Everybody|away ("( :: NoNameScript 4.02 :: )")
20:19:08 Join Everybody [0] (
20:25:31 Join an-user [0] (i=52432d71@gateway/web/cgi-irc/
20:25:35an-userhi !
20:25:57 Quit lee-qid ("aufwiederbyebientotsayonara")
20:26:30an-usersorry to bother you, but I've got a problem after a failed flash of rockbox on a H10 (like many other, I guess)
20:27:09an-userthe file rockbox.h10 seems to be corrupted and I cannot replace it with another, because pressing the O key no longer let me inter in the UMS mode
20:27:39 Quit trypt0 (Read error: 110 (Connection timed out))
20:28:12an-userso my question is : is there something that could allow me to enter UMS mode from the rockbox bootload with a corrupted rockbox.h10
20:29:03 Nick Everybody is now known as Everybody|ingame (
20:29:09goffaon boot is there an option to press a key to enter recovery mode?
20:29:19 Join votetrev [0] (
20:29:23goffalike on the gigabeat i think you can hold menu an-user
20:29:32goffabut not sure on h10
20:29:36votetrevhas ANYONE in here succesfully gotten a rockbox theme working on the 5.5g ipod video
20:29:49votetrevi can get them working but then they wont show any of the song information....its really bugging me
20:30:24 Quit lowlight ("CGI:IRC (EOF)")
20:30:34an-usergoffa: before this failure, the "O" key let me enter in recovery mode
20:30:56goffaah...try that on boot.. if not.. i dunno
20:31:06goffayou'll probably have to wait for a h10 dev
20:31:24an-user(more precisely : removing battery, plug USB, turn the device on with O key pressed)
20:31:42 Quit GodEater (" HydraIRC -> <-")
20:31:47 Join lowlight [0] (i=c730190b@gateway/web/cgi-irc/
20:32:30an-userso I'll wait ^.^ do you have an idea of when they will be there?
20:32:44an-user(btw, thanks for your help)
20:33:36votetrevanyone in here using a 5.5g ipod?????
20:33:53goffaunfortunately no
20:34:00goffai don't even know who was working on that
20:34:31goffai have an h140, x5l, and a gigabeat
20:34:36 Join GodEater [0] (
20:36:20votetrevlucky bass :)
20:36:32PaulPositionan-user: I may not be ble to help much but I'll do what I can...
20:37:02PaulPositionan-user: First things first, though : There is NO flashing of rockbox on H10 devices...
20:37:30lowlightan-user: have you tried holding "Next" when powering on until the USB screen appears (from
20:37:55PaulPositionlowlight: Next / -o- refer to the same buttons. So he did try.
20:38:43lowlightah, ok
20:38:45an-userPaulPosition: so the flash would still work if I could make it ignore the HD ?
20:38:56PaulPositionan-user: A few things.. You didn't forget to reset the player before trying the ums trick, did you? (ie, take out battery (5 or 6gb) or insert clip into 'reset' hole (20gb))
20:39:09 Quit votetrev ()
20:39:26PaulPositionan-user: The flash and the UMS mode has nothing to do with what's on the disc, and rockbox only installs to disc on H10.
20:39:28an-userthe soft in flash, sorry (talking in english is not easy for me)
20:39:43an-userno, I didn't forgot
20:40:14PaulPositionan-user: What happens exactly when you try to connect through the UMS trick?
20:41:26an-userinstead of giving hand to the emergency mode, the player try to load rockbox and fail to do that
20:43:33PaulPositionErr... Fails with an error message? Or if not, what's the last line it shows?
20:44:00an-userRockbox error: -5
20:44:24an-userthe last line is "Rockbox error: -5", the first being "Rockbox boot loader"
20:44:34PaulPositionan-user: And excuse me if it sounds patronizing, but you're *really* sure you reset the player correctly and that it wasn't connected through usb/powercord while you did that reset?
20:45:07PaulPositionan-user: wait a second, checking out things.
20:45:57an-userPaulPosition: I understand well that you ask me for that, I know that often the problem is the user -_^, but I'm sure
20:46:43an-userI'm going to redo step by step the trick 8t time, and ask my brother to do so, so I'll be even more sure
20:47:31PaulPositionan-user: Because, really, that's the only thing I've ever seen stop ums trick from working right. If you press right (next, select, -o-, watever), hold it, powerup still holding it, it shouldn't even notice that rockbox is installed.. It's pure firmware in flash.. :o
20:49:47an-usershame on me
20:50:03an-usermy brother succeeded ...
20:50:16PaulPositionNah, we prefer you go with the 'Eureka'. It works? Fine!
20:50:35 Join jdong [0] (n=jdong@ubuntu/member/jdong)
20:51:11jdongregarding USB ports only letting 100mA power draw with no device registering, is that a USB standard or a software setting?
20:51:22jdongi.e. could I override that in the OS, or is it hardcoded to the USB controller
20:51:42 Quit Thundercloud (Read error: 60 (Operation timed out))
20:52:18an-userI don't like being a lame... again : thanks for your help, you saved a brand new device
20:52:35PaulPositionyour welcome
20:54:00 Quit an-user ("you're really nice, thanks !")
20:54:18 Join yossa [0] (
20:55:27 Quit lowlight ("CGI:IRC (EOF)")
20:56:06jdongis there any way for Rockbox to report charging current on the iPod
20:56:22jdongor otherwise gauge if the battery is charging at a appropriate rate?
20:56:53 Join kaaloo [0] (
20:57:09PaulPositionjdong - On wps or you mean for debugging purposes?
20:57:20jdongPaulPosition: for debugging purposes
20:58:15PaulPositionjdong - info->debug->battery ...
20:58:42 Nick Everybody|ingame is now known as Everybody (
20:58:58jdongPaulPosition: I see detailed voltage info
20:59:03jdongbut nothing about charging current
20:59:27PaulPositionUp and down (and maybe left right) on that page pages through lots of stuff
20:59:49 Quit GodEater ("User pushed the X - because it's Xtra, baby")
20:59:52 Join petur [0] (n=petur@rockbox/developer/petur)
20:59:58PaulPositionMaybe there's what you look for, if not, sorry maybe someone has better idea.
21:00:15jdongno, unfortunately it doesn't deliver the info I'm looking for
21:00:18jdongit comes a bit close
21:00:33jdongbut I'm looking for current (mA) info
21:00:48jdongand I don't have an ammeter to hook up around the USB cable either :D
21:00:55jdongthough that would probably be the best method, lol
21:02:02PaulPositionjdong: heh. Can't this be deduced (induced?) from using battery bench and looking at its log? (I know nothing of electricity.. :p )
21:02:37jdongPaulPosition: well, charging current doesn't directly equal change in voltage in any predictable fashion
21:02:54jdongand also charging percentage is very very unreliable/unsmooth
21:03:20jdongthe closest I could get is a full discharge, then recharge in stock firmware and then recharge in rockbox
21:03:24jdongbut that sounds like 2 days of work :D
21:04:07jdongof course if lsusb can tell me how much current is drawn in realtime...
21:04:07jdong(but it doesn't)
21:04:56 Join Thundercloud [0] (
21:05:28 Nick Everybody is now known as Everybody|BRB (
21:06:21***Saving seen data "./dancer.seen"
21:06:24 Join krz [0] (
21:07:29 Join norbusan [0] (
21:08:19 Part norbusan
21:08:21 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
21:08:52 Join Zero-Coo|_ [0] (
21:09:19Zero-Coo|_can anyone help me ;/ ?
21:10:06 Join skibone [0] (i=904a6c4c@gateway/web/cgi-irc/
21:10:43LloreanZero-Coo|_: We won't really know that until you ask your question and find out if someone here knows the answer.
21:11:00Zero-Coo|_i bought the new nano ipod
21:11:16jdongnot supported
21:11:16Zero-Coo|_the rockbox support the new ?
21:11:29jdongI have 4 of them
21:11:30Zero-Coo|_hmm what can i do to put videos on it?!?!
21:11:42jdongtrade 2 for a 30GB video
21:12:00 Quit Criamos ("( :: NoNameScript 4.03 :: )")
21:12:01jdongsorry, there really isn't much support for anything other than stock firmware on the 2G nano
21:12:02Zero-Coo|_on future
21:12:15jdongmaybe in the future... but it appears like the firmware is encrypted
21:12:24jdongand there doesn't seem to be much progress yet in decrypting it
21:12:26Zero-Coo|_oo duhhhh ;/
21:12:30 Part tvelocity ("Ī‘Ļ€ĪæĻ‡ĻŽĻĪ·ĻƒĪµ")
21:12:37jdongthere's a lengthy forum topic on this
21:12:43Zero-Coo|_i listen something about ipodlinux ?
21:12:47Zero-Coo|_what is this?
21:12:53jdongipodlinux = running linux on the iPod
21:13:00jdongit also doesn't support the 2G nano :(
21:13:02maquisZero-Coo|_: ipodlinux won't have anymore luck on the new nano, i'm betting
21:13:02Zero-Coo|_yeah but the new ones?
21:13:11jdongno supporto :(
21:13:19Zero-Coo|_i spend 110 pounds
21:13:23jdongipodlinux doesn't support FAT32 5.5G ipods either
21:13:24Zero-Coo|_buying that ipod
21:13:31jdongZero-Coo|_: I'm sorry to hear that :(
21:13:40Zero-Coo|_thnx anyway for the help:P
21:13:43jdongI too am anxiously awaiting progress on the 2G nano
21:13:43Zero-Coo|_bye bye
21:13:45 Join tvelocity [0] (
21:13:47jdongas I said, I got 4
21:13:52jdongand they collect dust right now
21:14:08jdonggonna probably give them away as gifts
21:14:19Zero-Coo|_c ya in the "future" :Pp
21:14:25jdongk :)
21:14:28 Quit Zero-Coo|_ ()
21:14:51skiboneHi gang - one day newbie here running rb on a Gigabeat. Is it possible to delete the currently playing file?
21:15:11jdongLlorean: do you know if it's possible for some USB controllers to allow 500mA current draw for rockbox?
21:15:23jdongLlorean: I swear on my Centrino Duo laptop Rockbox and stock firmware charge at the same rate!
21:15:25 Join Mouser_X3 [0] (
21:15:39 Quit krz (Read error: 131 (Connection reset by peer))
21:16:04jdongskibone: I don't think any flavor of Rockbox allows that at the moment... it definitely didn't work on my iPod
21:16:21*maquis hopes for jdong's sake that he's right, because a centrino duo laptop would be a sad thing to lose
21:17:02jdongmaquis: let me add a "that" after the 2nd word of that sentence so I don't bet anything :D
21:17:31maquisjdong: that does make a small difference
21:17:38jdongyes it does :D
21:17:51skiboneWell, why the heck not?!? ;-) Drat. I may have to stick with the nasty Gigabeat firmware. I love to load up my player with all kinds of music and then delete the crappy music along the way.
21:18:25PaulPositionskibone: gimme a second, I *may* have an idea..
21:18:41Lloreanjdong: Some people have reported Rockbox charging at much faster speeds (either from the wall, or from computers) and I think there may be USB controllers that provide the full 500, independent of what the device requests, but I'm just guessing.
21:18:54jdongmmm, ok
21:19:08jdongso it's a hit-or-miss anomaly then
21:19:16Lloreanskibone: Have you *tried* deleting the currently playing file?
21:19:25Lloreanjdong: I believe you can delete the currently playing file, actually.
21:19:41*jdong goes and tries it
21:19:47jdongI swear it didn't work the last time I did it!
21:20:26LloreanIf you set Rockbox to follow the playlist, then just open the context menu and choose "delete", if the file is entirely in buffer it should delete fine. If part of the file is off-buffer it may not work wholly right.
21:20:28PaulPositionskibone: just use the general settings->file view->follow playlist option so when you back into browser it highlights the currently playing song. Stop playback, then hold select on the song and choose delete.
21:20:43 Join norbusan [0] (
21:20:52 Join krz [0] (
21:20:55 Part norbusan
21:21:05jdongwow yes it does work
21:21:23jdongand interestingly it continues to play
21:21:43*jdong fast forwards half an hour to see if it's the buffer
21:22:09 Part Angry
21:22:35jdongyes it's definitely the buffer!
21:22:41skiboneHmmm - well, on the Gigabeat, playback stops when you go back to the file browser. It's not supposed to do that, right? I think I read in the gigabeat logs that they are working on that.
21:25:08 Quit crwl (Remote closed the connection)
21:25:41 Join crwl [0] (
21:26:06 Join norbusan [0] (
21:27:53AraknisCan anyone help me with how to Disable CPU frequency scaling for the kernel_on_cop_6.diff for the iPod Photo like linuxstb (Dave Chapman) did with his iPod Photo
21:28:10jdongLlorean: it's pretty glitchy though, like I tried to stop (which I have auto-bookmarking) playback and rockbox hard-locked
21:28:31jdongat this point I don't think it's something we should tell people to do :D
21:28:39Lloreanjdong: That's mostly because there's an open bug of "If the last song is deleted while playback is stopped, and then resume attempts, a hard freeze will happen"
21:28:48jdongah :)
21:29:03AraknisI tried removing #define HAVE_ADJUSTIBLE_CPU_FREQ from config-ipodcolor.h and it let me continue into Rockbox however the Mpegplayer would just have sound and saying buffering (I never seen a video clip just sound)
21:32:51 Nick DreamThief is now known as DreamThief|off (
21:32:56LloreanAraknis: I have to ask, are you the person who asked the same thing on the Patch Tracker on the Coprocessor task? It's not really a discussion forum, and should be kept more on the topic of discussing the actual task, not asking how to get videos working.
21:34:00PaulPositionAraknis - Are you trying to make a useable build out of this or do you just want to preview what mpegplayer will look like when the patch is working?
21:34:37 Quit skibone ("CGI:IRC (EOF)")
21:35:06 Quit trypt0_ ("Leaving")
21:38:26AraknisYes, I am the person who asked the same thing on the Patch Tracker
21:38:47AraknisI had no idea it was not a discussion forum, others looked like there were
21:39:24 Quit Rob2222 ()
21:39:32AraknisI was wanting a useable build, to see what the fps would look like
21:39:54 Quit MarcoPolo (Remote closed the connection)
21:40:32PaulPositionAraknis: If all you want is to try and see, there is an 'old' version of patch, that syncs with daily cvs from 29 dec. (or maybe 28th)..
21:40:57LloreanAraknis: The patch tracker is for discussion of the specific patches. Your question didn't really relate to the kernel_on_cop patch's state, so much as mpegplayer's state and the CPU scaling problem on the 4G
21:41:09PaulPositionAraknis: But the resulting build isn't too stable, and the mpegplayer can't be quit except by resetting player.
21:41:15 Nick Everybody|BRB is now known as Everybody (
21:41:23linuxstbAraknis: Yes, starting mpegplayer is unreliable for me, but normally works more than 50% of the time. If it doesn't work for you, you'll just need to wait until mpegplayer is more stable.
21:41:50*Llorean wishes everyone would just have some patience regarding both Mpegplayer and Gigabeat/Sansa installs.
21:42:21AraknisWithout a Patch it seems stable - does not freeze but I get a lot slower framerate than yours linuxstb
21:42:37*ender` yawns
21:42:45*bluebrother doesn't understand why one could mix up the tracker with a discussion board
21:42:58bluebrothermaybe someone should add this to the tracker frontpage ...
21:43:39AraknisSorry, Llorean I am also a noob and guess I was excited to try it out after reading about it on the wiki
21:44:14Lloreanbluebrother: It is *technically* a discussion board, just a very, very focused one ideally
21:44:27bluebrotheryeah, it's a tracker :)
21:44:43 Quit thegeek (Read error: 113 (No route to host))
21:44:45Lloreanbluebrother: Can I ask you to do a quick manual change?
21:45:32LloreanAt this page, right below the anchor is a "Backup Original firmware Bootloader" and "Upload the Rockbox Bootloader" section:
21:45:38LloreanThey should go in the Gigabeat installation section for the bootloader.
21:46:22 Join thegeek [0] (
21:46:23LloreanSomewhere before the "Upload" bit should also be a Download the Rockbox bootloader step, that should say we don't *yet* have one posted.
21:47:09AraknisI am sorry for messing up your tracker, I guess I should have posted in the Forum.
21:47:15LloreanRight now a lot of people think that because the Bootloader Installation section of the manual is empty, all a Rockbox install is is a copy of the .zip and an extraction.
21:51:16 Quit dpro (Read error: 110 (Connection timed out))
21:52:50 Join BigBambi [0] (
21:53:31 Join Osse [0] (
21:57:58Ossecan anyone help me please ? Imk trying to install rockbox but have fell short embarrasingly quickly :S , i cant run the iPodpatcher.exe =/
21:58:55bluebrothercan the gigabeat dual-boot?
21:59:02 Quit gotthardt ("CGI:IRC (EOF)")
21:59:20jdongOsse: a more detailed explanation of what goes wrong trying to run ipodpatcher could be helpful
21:59:24goffabluebrother: yes
22:01:20bluebrotherwhat happens if someone puts in a bootloader with a wrong name? Will this break the gigabeat or only trigger some recovery mode?
22:01:36goffahavent attempted that
22:01:42goffahaven't heard reports
22:02:03goffaer wait
22:02:07goffathat's fine
22:02:14goffafirmware haven't attempted
22:02:30goffayou have to hold menu if you put a wrong name bootloader on
22:03:12 Quit jdong (Remote closed the connection)
22:07:41 Join Rob2222 [0] (
22:11:58 Join christie [0] (n=hypermoz@
22:12:13 Join decayedcell [0] (i=3ba7bf28@gateway/web/cgi-irc/
22:14:40OsseAFT VERSION
22:14:41OsseAFT VERSION
22:14:42DBUGEnqueued KICK Osse
22:14:42OsseThe following instructions refer to the “installation folder.” For Windows users, the
22:14:44OsseThe following instructions refer to the “installation folder.” For Windows users, the
22:14:44Osse”installation folder” is assumed to be a folder in the root (top- level) of the C: drive
22:14:47Osse”installation folder” is assumed to be a folder in the root (top- level) of the C: drive
22:14:49Ossecalled rockbox (you will obviously need to create this folder yourself). For Mac OS
22:14:51Ossewhoops >.<
22:14:53Ossecalled rockbox (you will obviously need to create this folder yourself). For Mac OS
22:14:55OsseX and Linux users, the “installation folder” is assumed to be the Desktop folder. But
22:14:57OsseX and Linux users, the “installation folder” is assumed to be the Desktop folder. But
22:14:59Osseyou can use any folder you wish. Note that these files should be saved onto your
22:15:01Osseyou can use any folder you wish. Note that these files should be saved onto your
22:15:05Ossecomputer’s hard disk, not on your Ipod.
22:15:05Ossecomputer’s hard disk, not on your Ipod.
22:15:07Osse1. First, download the ipodpatcher tool to your installation folder. You can download
22:15:10Osse1. First, download the ipodpatcher tool to your installation folder. You can download
22:15:11Ossethe ipodpatcher tool for your operating system at http://download.rockbox.
22:15:13Ossethe ipodpatcher tool for your operating system at http://download.rockbox.
22:15:16 Quit Osse (Excess Flood)
22:17:01christiehi everyone:-) are there any gigabeat f users?
22:17:13liniquick question - i did a sim build for ipod video
22:17:28liniwhen I start the sim it says - no .rockbox directory
22:17:46Lloreanlini: Did you do a "make install" after "make"?
22:17:52bluebrotherlini, make install
22:17:54liniah thanks
22:18:04liniforgot about that
22:19:39 Join Alonea [0] (
22:19:42 Join Lars_G [0] (n=lars@unaffiliated/lars-g/x-000001)
22:19:44Lars_GHi all.
22:19:52Lars_Gquestion, who here are rockbox regular devs?
22:20:59 Nick krz is now known as sD (
22:21:17peturdefine regular. Produce code or be around or ...
22:21:17 Nick Everybody is now known as Everybody|away (
22:21:20BHSPitMonkeyLars_G, why don't you check that list on the site.
22:21:27 Nick sD is now known as kraizee (
22:21:32Lars_Gpetur: Produce code
22:21:41Lars_GBHSPitMonkey: It'd be very slow to cross compare :)
22:21:48BHSPitMonkeywhy is "regular" a requirement?
22:21:56Lars_Gi just want to ask to the regular devs here, what do you code in? (as in editor or ide)
22:22:06BHSPitMonkeyemacs? :P
22:22:10bluebrothergvim :)
22:22:18BHSPitMonkeyMS Word
22:22:27Lars_GBHSPitMonkey: masochist
22:22:44BHSPitMonkeyusing a blinking cursive font
22:22:56LloreanI use ConText.
22:22:59peturcopy con myfile.c
22:23:12CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
22:23:12*Lars_G googles context
22:23:22bluebrotherchanging bit using permanent magnets :)
22:23:34LloreanIt's a windows text editor that isn't so bad, really.
22:23:49Lars_GOk highlighting editor for windows.
22:23:56BHSPitMonkeyI've heard of this "Programmer's Notepad" for windows
22:23:56bluebrothergvim :)
22:24:04debauched_slothI use SlickEdit
22:24:10linuxstbLlorean: "isn't so bad, really" - a glowing recommendation....
22:24:15Lars_GMe too, I should've said I'm on linux but oh well
22:24:28Lloreanlinuxstb: Well, when you say "windows text editor" there's so much negative weight just in the term...
22:24:30Lars_Gbluebrother: you can only vote once :)
22:24:30bluebrotherProgrammers Notepad is nice. I used it at uni where I hadn't a good editor. It can do code folding.
22:24:47bluebrotherLars_G, I voted for both operating systems :)
22:24:59linii use SciTE
22:25:03Lars_GAh well.
22:25:15bluebrotherOn Windows I used to use CrimsonEditor
22:25:17LloreanI've actually started looking at this DevCPP thing for windows, but haven't really dove into it yet.
22:25:47LloreanI've been wondering if I could wrangle it into compiling Rockbox, but right now for it to cross compile you have to set it up to dedicatedly cross compile.
22:25:49Lars_GSorry to bring the "poll" here, it's just I'm on the look for a C/C++ ide, and rockbox it's a great source of info, since you use a non-completely-common build system, you mix C and other kinds of code, and have a pretty complex structure for a small project
22:25:54bluebrotherI used DevC++ 1 1/2 years ago. It's kinda nice but I didn't liked it too much
22:26:05 Nick christie is now known as coloboque (n=hypermoz@
22:26:14 Join Soader03 [0] (
22:26:58Lars_Gok thanks for the feedback
22:27:25*pixelma seconds Llorean - surely amiconn would do so too :)
22:28:24bagawkLars_G: the tools rockbox uses to build are very common...
22:29:20Lars_Gmake might be, but your configure didn't look like autotools to me
22:29:42Lars_GSlickEdit looks good, but if it's over $10 I can't pay it right now
22:29:51peturpixelma: on which one?
22:29:59 Quit Araknis ("CGI:IRC (EOF)")
22:30:00bagawkWhy not just use vim?
22:30:23 Quit decayedcell ("CGI:IRC (Ping timeout)")
22:30:46Lars_Gvim is ok but when a project goes beyound 6 files, I get frustrated handling them all on vim, even with the project plugin
22:31:27 Quit PaulPosition (Read error: 110 (Connection timed out))
22:31:32 Join Phul [0] (
22:31:41 Part Soader03
22:31:42PhulIAudio G3/U2 has rockbox?
22:32:15Phulwhy not
22:32:19 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:32:22bagawkLars_G: I can see that at times, but between mutiple terminals and the use of :e#, things work well for me
22:32:31LloreanPhul: Because nobody with the hardware has written it. Are you going to?
22:32:40Phulwhy not
22:33:01LloreanBecause they keep sitting around saying "why not" instead of doing anything.
22:33:04*Lars_G gently pushes Phul to dislodge the needle
22:33:32PhulI would have been trying to help
22:33:39Phulbut pretty busy this days.
22:33:50Lars_Grockbox's kern's monolitic right?
22:34:09LloreanPhul: It doesn't need help. It needs someone to *do* it.
22:34:22PhulWhat qualifications needed for writing those stuff?
22:34:39LloreanRight now no work is going on, at all, in that area, because there are no developers for that hardware.
22:34:44LloreanAt least, none who speak here.
22:35:10bluebrotherLlorean, bootloader instructions added. I hope it's understandable ;-)
22:35:53kaalooHi all, just wanted to let you know I made a small donation. Keep up the good work !
22:36:01Phulwell i'm up to the task
22:36:05Lloreanbluebrother: Hey, if it's not understandable, that's fine too, as long as they're aware that the bootloader install process isn't just empty space. :)
22:36:06Phulaccording the first paragraph
22:36:12 Join webguest34 [0] (i=53e2d3c0@gateway/web/cgi-irc/
22:36:37bluebrotherI tried my best to make it confusing *g*
22:38:00 Quit kaaloo ("Leaving.")
22:38:14Lloreanbluebrother: Looks good enough to me.
22:39:46Aloneaok, I was curious and wanted to look at the code, but I haven't done it in ages, so, bear with me here. I just wanna see if I will be able to help some on the project at some point in time. So, I guess first I would have to ask is what file do I download first? The Source Archive or justthe CVS for my player?
22:39:51Phulhmm looks intersting, I might consider it in future.
22:40:14Lloreanbluebrother: Now, we just need a prebuilt gigabeat bootloader officially available and I'll open up a Toshiba installation forum.
22:40:27*Lars_G laughs at the guide
22:40:33Lars_Gtrace the connections with a multimeter.
22:40:35bluebrotherI just read your argumentation for this
22:40:38*Lars_G cries with laughter.
22:40:57bluebrotherLars_G, what guide?
22:41:03 Quit cynicalliberal ("Chatzilla 0.9.77 [Firefox]")
22:41:03LloreanAlonea: The CVS for your player is a precompiled build. You can either follow one of the install guides, and check out our sources from CVS, or if you just want to skim for now the source archive should be adequate.
22:41:05Lars_Gbluebrother: New port one.
22:41:19bluebrotherah. That's a nice one. I like it :)
22:41:20PhulWhat about U3
22:41:21 Join Rob222241 [0] (
22:41:23Phuldoes it have rockbox?
22:41:34Lloreanbluebrother: I feel offering installation support as a forum should wait until there's a clearly defined install process that doesn't include compiling. :)
22:41:51LloreanPhul: ALL officially supported targets are on the front page of the site
22:41:52Lars_Greversing the schematic of a DAP (or anything with several high pin count ICs) with a DMM is not something the weak can do without going insane
22:41:52bluebrotheryeah, I agree with that
22:42:17LloreanLars_G: Fortunately nobody's accused our core devs of being fully sane in years.
22:42:26bluebrotherhmm, I don't believe Rockbox porting is for the weak
22:42:47FebsHehe. This may be one of the all time classic user mailing list quotes: "Perhaps you can tell me more about your unit privately. "
22:42:49Aloneaalright. really I just wanted to see if I could make sense of it enough to fix a couple things. Or at least make a learning experience of it.
22:42:51Lars_Git's not bluebrother but that step is specially not for the stable, weak, or someone with a spouse
22:43:54bluebrotherI'm still wondering if I happen to get insane one day ...
22:44:34LloreanAlonea: The archive should work for that, but using CVS to check out the sources will allow you to keep your local copy up to date more easily.
22:44:40 Join Osse [0] (
22:44:52Ossesorry about that earlier =P
22:45:20Ossei cant open the bootloader file with ipodpatcher , can anyone help ? =(
22:45:26 Quit Phul ()
22:45:36LloreanOsse: What error message do you get?
22:45:46Aloneaokies, thanks Llorean. Any suggestions on what program would be best to read the stuff in?
22:45:52Osseit just says ...
22:46:00LloreanAlonea: It's mostly C code, so any good C editor.
22:46:04 Quit Rob2222 (Read error: 110 (Connection timed out))
22:46:12 Join votetrev [0] (
22:46:18 Join Rob2222 [0] (
22:46:24Osse[ERR] Couldnt open input file bootloader
22:46:33Osseusing win32 cmd
22:46:36 Quit lost|X40 (Read error: 104 (Connection reset by peer))
22:46:51linuxstbThe instructions will tell you type something like "bootloader-ipodvideo.ipod"
22:47:05Ossewell yeah thats what i did
22:47:14linuxstbNo you didn't....
22:47:18linuxstbDid you add a space?
22:47:19LloreanNot if that was your exact error message.
22:47:47Aloneaok. I guess thats where I was confused. Used to more java than C, but I will try my best.
22:47:52 Quit amiconn (Read error: 110 (Connection timed out))
22:47:52 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:48:10votetrevanyone in here that could help with a simple question about a theme for the 5.5g ipod video
22:49:22webguest34Hi guys! I'm curious about how to make recursive functions not to try to open too many subfolders. Would something like "if opendirs<MAX_OPEN_DIRS..." be possible?
22:50:07webguest34Background info:
22:51:14Ossei cant get whats wrong =/
22:51:15LloreanOsse: Was what you gave use the *exact* error message then, or did you shorten it?
22:51:53LloreanBecause if you typed "bootloader-ipodvideo.ipod" it should say "[ERR] Couldnt open input file bootloader-ipodvideo.ipod"
22:51:55Osse[ERR] Couldnt open input file bootloader-ipodvideo.ipod
22:51:59Osseis the exact one
22:52:02LloreanWhy did you tell us something else, then?
22:52:23LloreanIt's very hard to help you if you give us incorrect information, because that information can then tell us you did the wrong thing wrong.
22:52:46LloreanWhen you type 'dir' without the 's, and hit enter, is bootloader-ipodvideo.ipod in that list?
22:52:48Osseright , fair enough , well what i just typed is the exact error message
22:53:34linuxstbYou haven't saved the bootloader-ipodvideo.ipod file into the same folder as ipodpatcher.exe
22:54:20Osseno it isnt there , but i was running the program / commands in the same command line in the cmd window , wont that work ?
22:54:30Ossethe file is in the same directory as the exe
22:54:58LloreanOsse: It matters where you run the .exe from, the file has to be where you run it, not where the .exe is
22:55:35Osseahh , thanks =)
22:55:44Osseso i need to run it from the directory its in then
22:56:11LloreanI believe the instructions tell you to do so.
22:56:30LloreanYes, they do.
22:59:33Osseok , thanks for the help =)
23:00:43 Quit Rob222241 (Read error: 110 (Connection timed out))
23:01:23webguest34Anyone? ...I'll rephrase:
23:01:29webguest34In "dir.c" i find declaraion of "MAX_OPEN_DIRS 8" and "static DIR opendirs[MAX_OPEN_DIRS]". Can these be used in the code of a plugin?
23:02:45 Part coloboque
23:03:56 Quit BigBambi ("Leaving")
23:04:05Bagderdebauched_sloth: if you create a flyspray user I'll add you to the dev group which give you the "proper" rights
23:04:08Osseok ... so installed it , and now i get the loadng screen but i get error "-1" as it tries to load the rockbox firmware , and just loads the ipod firmware instead
23:04:23Ossefirst time ive done this , my apologies for being a noob
23:05:28LloreanOsse: You aren't following the installation instructions...
23:05:31linuxstbwebguest34: Why not just check the return value of opendir() ?
23:05:39Ossesorry , what did i miss :S
23:05:41LloreanOsse: It tells you to then extract a onto the device.
23:06:12Osseoh god i missed a page :o
23:06:14Ossesorry !
23:06:14LloreanThis is the second question you've asked that's resulted from you not carefully following the instructions. Take your time and read them carefully.
23:06:23***Saving seen data "./dancer.seen"
23:06:33Osseyeah im just dying to get this on , ill read it properly
23:06:38Ossethanks again <3
23:07:19 Quit Mouser_X3 (Read error: 60 (Operation timed out))
23:07:31 Join Rob222241 [0] (
23:07:54webguest34linuxstb: exactly my thought. But I still haven't fixed me an environment to try it myself yet...
23:08:24webguest34So it seems quite easily fixed then?
23:09:00webguest34At least to avoid a plug crashing because of it not "beeing careful enough"
23:09:37 Join lostnihilist [0] (
23:10:01linuxstbIt looks like disktidy already does that check - that's the error people are referring to in that forum thread.
23:10:48webguest34Well.. I'm "mlind"
23:11:02webguest34And Disktidy does crash
23:11:15linuxstbHow do you mean "crash"?
23:12:06webguest34I have more than 8 subdirectories, and when disktidy gets too deep it just stops and says it encountered an error.
23:12:18linuxstbThat's not a crash...
23:12:19LloreanThat's not a crash then, that's a "graceful exit upon error"
23:12:39webguest34Ok ok
23:12:46LloreanI mean, what were you planning to make it do, if it couldn't open any more directories because of the limit?
23:13:05linuxstbI agree a more specific error message could be useful though.
23:13:16webguest34It could just skip those deeper subdirectories
23:13:48webguest34...instead of not cleaning the rest of my disk
23:14:18LloreanYou're going to have to run disktidy again either way.
23:14:20webguest34...that doesn't have as deep structure as a few portable apps that I keep in there
23:15:22webguest34I seldom dig into those folders with RockBox, but it annoys me to see the stuff in the root folder
23:15:55 Quit kraizee (Read error: 131 (Connection reset by peer))
23:16:00 Join sD [0] (
23:16:21LloreanIt just seems strange to me that you want a plugin to perform the process in an incomplete manner as the default behaviour.
23:16:26Osseok sorry but im stuck again , i dont know if ive missed something but i apologise if i have
23:16:44LloreanBut it should be changeable to do that fairly easily, I'd imagine
23:16:54Ossei cant see where it says to extract a to the ipod , when was i supposed to do this ?
23:17:23LloreanOsse: Depends, what instructions are you reading?
23:17:25linuxstbAre you following the manual or the IpodInstallationBeta wiki page?
23:17:26webguest34Thanks. I'll see if I get the time to set up everything to try it...
23:17:45Osseits a pdf file , for the ipod video
23:17:54Ossethe instructions are specific to it at least
23:18:11LloreanOsse: 2.2.2
23:18:16LloreanOsse: The step BEFORE installing the bootloader.
23:19:02Ossethanks =D
23:20:44 Quit hachi ("And I begin to wonder... The dreams I can't remember when I wake in the morning - where in the world did they go?")
23:20:59 Part perl|work
23:22:08 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
23:22:10 Quit yossa ("Chatzilla 0.9.77 [Firefox]")
23:22:25 Join miepchen^schlaf [0] (
23:23:49 Quit Rob2222 (Read error: 110 (Connection timed out))
23:24:51 Part webguest34
23:28:06 Join hachi [0] (
23:28:13 Part Llorean
23:33:57 Quit petur ("sssssssssss---------PLOP!")
23:35:51 Part norbusan
23:38:07 Part debauched_sloth
23:40:18 Nick DreamThief|off is now known as DreamThief (
23:41:57 Quit votetrev (Read error: 60 (Operation timed out))
23:45:32 Join BobJonkman [0] (
23:47:11 Quit x1jmp (Remote closed the connection)
23:49:06 Quit triplah_ ("Leaving")
23:52:33 Join BobJonkma1 [0] (

Previous day | Next day