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 2009-02-22

00:01:16 Quit {phoenix} (Remote closed the connection)
00:02:32gevaertsUnhelpful: try "stty -icrnl < /dev/ttyUSB0"
00:07:18 Quit ender` (" The most important lesson we can learn from history is the inability of humans to learn lessons from history. -- Aldous Hu")
00:12:25 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
00:12:25 Quit pixelma (Nick collision from services.)
00:12:30 Quit amiconn (Nick collision from services.)
00:12:31 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
00:12:39 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
00:12:49 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
00:14:26 Join CaptainKewl [0] (i=jds@207-237-172-77.c3-0.nyr-ubr4.nyr.ny.cable.rcn.com)
00:15:09 Quit faemir ("Lost terminal")
00:15:55rasherWhatever happened to translations/voice of plugins?
00:17:44Unhelpfuladd an igncr to that recipe, and you get unix-ish text :)
00:20:25n1srasher: dunno, an unfinished patch was posted to the tracker a while ago and iir midgey(?) said he would pick it up again at some point, then silence
00:21:57 Quit petur (Remote closed the connection)
00:23:05n1sFS #9067 is the patch, last post on aug 13
00:25:02n1stoo bad it wasn't finished as it looked like it had come a long way and was pretty close to finished
00:25:43rasherI wouldn't say all hope is lost. I'd expect that area to have relatively few changes since then
00:26:37rasherAnd midgey did seem interested in continuing the work, last I heard
00:27:55n1sno, all hope is not lost, it just seems to have stalled :)
00:28:49n1si'd expect it to need a bit of adjusting for the overhauled buildsystem
00:29:47rasherTrue that, but the code changes are probably more or less still valid
00:30:09n1swow. We are getting close to 400 open patches...
00:32:09***Saving seen data "./dancer.seen"
00:33:44 Join faemir [0] (n=daniel@88-106-169-118.dynamic.dsl.as9105.com)
00:33:52 Quit n1s ("Lmnar")
00:35:13soappeople are really wanting a "rockbox" handout. if Mr. Somewhere wants to make one I'll Kinko's a few score coppies.
00:36:04 Quit FOAD (Remote closed the connection)
00:36:39Unhelpfulpf yields at the end of render_slide, and after lcd_update. render_slide calls surface, which looks for a cached slide in a linked list, and posts an event to load the slide if it's not in cache, returning the empty slide in that case. i'm guessing that queue_post, and queue_wait* yield?
00:36:55rashersoap: a single page of short and sweet information?
00:37:19amiconnUnhelpful: queue_post doesn't, queue_wait does of course
00:37:22rasherTop paragraph of wikipedia with a few more screenshots and a huge logo?
00:37:42Unhelpfulhonestly, i'm not even sure the locks around the "to load" queue are needed, as i understand things here...
00:38:01rasherOr am I misunderstanding the meaning of "handout"
00:38:23Unhelpfulhow does sim handle threads... not bare platform threads, i would think?
00:38:35 Join FOAD [0] (n=dok@dinah.blub.net)
00:39:31amiconnThe sim simulates rockbox threading (i.e. cooperative threading) using sdl threads which are interlocked
00:39:38Lloreansoap: Maybe just the beginning of the WhyRockbox page? (Including the table)?
00:39:48amiconnYou should probably ask jhMikeS for details
00:40:08 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
00:40:25soaprasher - yes
00:40:35rashersoap: whch size?
00:41:00soapwith "Rockbox.org" in bold
00:41:19gevaertsrasher: probably one of those weird US sizes
00:41:41rashergevaerts: Indeed
00:41:49rasherLetter?
00:41:55Unhelpfuli'm mostly trying to figure out what the difference could be. there must be a reason it works on sim, and if it can't be threads, or alignment, since buffer_alloc.c aligns allocations at 32 bits... i still lean toward buf_* writing outside their buffers, but various tests to catch that on sim did nothing, and there's no sign of covers ever being "wrong" in sim, either.
00:42:07soapwith "Rockbox.org" in bold
00:42:23soap8.5x 11
00:43:03 Join FlynDice [0] (n=jack@12-46-55-89.seatac.seattwa.wayport.net)
00:43:31rasherThat appears to be "Letter"
00:44:41 Quit lymeca (Read error: 110 (Connection timed out))
00:45:57amiconnUnhelpful: If it is alignment you should get a data abort on arm (unless you're only accessing it with ldm/stm, in which case your data will silently get rotated iirc)
00:46:31Unhelpfulyes, there's that too, the fact that it simply locks silently :/
00:46:47amiconnThe write outside would be easier to test on a target that offers some sort of memory protection
00:47:11rashersoap: having a go..
00:48:54Unhelpfulamiconn: heh. i tried that strategy, replacing the use of plugin_buffer for buf_* with a static page-aligned array on sim, and using mprotect to lock a page at each end against any access. no crash. :/
00:49:24amiconnEither an arm with MMU (I didn't deal with mmu programming so far though), or coldfire / sh1 (modifying the memory guard thing - I did something like this once)
00:51:15 Quit faemir ("Changing server")
00:51:17rashersoap: would double-sided work?
00:51:39rasherI'm having a hard time fitting much on one page
00:51:54 Join faemir [0] (n=daniel@88-106-169-118.dynamic.dsl.as9105.com)
00:51:58soapmore expensive to print Â:(
00:53:17gevaertsWill this be colour or B&W?
00:55:39soapcheap! :(
00:56:33rashersoap: I expect the rockbox fund might be willing to reimurse you
00:58:05soapcolor 2 sided !
00:58:20LloreanHahaha
00:59:15soapOR 2 fliers
00:59:53LloreanOne about Rockbox, and one concentrating solely on "Doom on your iPod", how to play, etc.
01:00
01:01:43soaplol
01:02:00soapLinux Gazette wants do to an article
01:02:05LloreanAwesome
01:04:10rasherrasher.dk/rockbox/handout.pdf">http://rasher.dk/rockbox/handout.pdf
01:04:55rasherThat obviously wants to be colour..
01:06:23linuxstbrasher: "We believe that you should never need to go through a series of menus for an action you perform frequently." We don't?
01:07:03LloreanThat's what our WhyRockbox page says, it seems.
01:07:13linuxstbAnd "We also believe that you should be able to configure almost anything about Rockbox you could want" ?
01:07:43linuxstbLlorean: Hmm, who wrote that?
01:08:02rasherMidkay, it seems
01:08:25 Quit evilnick ("http://www.mibbit.com ajax IRC Client")
01:08:46rasherOkay, maybe that isn't really the text we want..
01:08:52LloreanPossibly not.
01:09:24linuxstbrasher: Are you sure? It looks like MarcGuay to me - http://www.rockbox.org/twiki/bin/rdiff/Main/WhyRockbox?rev1=38;rev2=37
01:09:24rasherThe first bit is fine.. Everything after "We believe" goes
01:09:45linuxstbThe text in that diff sounds much better to me than the replacement...
01:09:50rasherlinuxstb: http://www.rockbox.org/twiki/bin/view/Main/WhyRockbox?rev=1
01:10:25rasherI think MarcGuay just moved some stuff about
01:10:29rasherCheck the diff further down
01:10:40linuxstbAh yes...
01:11:01rasherOkay, so, taking suggestions on a few sentences to put instead of those?
01:11:11*Llorean edits that page slightly.
01:11:26Lloreanrasher: I've just replaced those sentences on that page, I *think* it's more suitable
01:11:49rasherLooks good
01:12:10*rasher gets rid of all that bold text
01:12:56soapthanks!
01:13:22rashersoap: don't run off to kinkos just yet!
01:13:51rasherNew version: rasher.dk/rockbox/handout2.pdf">http://rasher.dk/rockbox/handout2.pdf
01:14:07Unhelpfulwell, i've added a ton of empty DEBUG_OUT("");, and of course, it's not crashing. :)
01:14:27 Quit intrados (Read error: 54 (Connection reset by peer))
01:14:35LloreanDoes anyone know if TWiki offers group restrictions for editing certain pages? Like... maybe for ReleaseTodo we could disable editing by non-devs since we want to avoid it becoming a feature request thing?
01:14:41 Quit SirFunk__ (Success)
01:14:45linuxstbrasher: You're missing whitespace in "customizable.It"
01:15:11 Join SirFunk__ [0] (n=Sir@208-15-25-145.netsync.net)
01:15:41 Join intrados [0] (n=intrados@cpe-71-67-133-85.woh.res.rr.com)
01:15:57linuxstbrasher: And "OGG" is a container format, not a codec... (and written Ogg...)
01:17:33rasherBut we want Vorbis here, right?
01:17:42rashersoap: are 1cm margins enough?
01:18:20 Quit FlynDice (Read error: 113 (No route to host))
01:18:29linuxstbrasher: Maybe say something like "MP3, WMA, AAC, Ogg Vorbis and FLAC"
01:19:30linuxstbAlthough it may be worth including the entire list - audio playback is Rockbox's most important feature.
01:20:24saratogai wouldn't put the game formats
01:21:01saratogais there a disk spin up when you change audio codecs during playback?
01:21:02linuxstbWhy not?
01:21:12linuxstbNo, the codec binary is buffered.
01:21:51saratogain the audio buffer?
01:22:02linuxstbYes, and then copied to the codec buffer.
01:22:05saratogaah ok
01:22:23saratogai just wouldn't bother listing the emulated formats individually since theres a lot of them and not many people know what they are
01:22:45rasherSupport for MP3, WMA, AAC, Ogg Vorbis, FLAC, Musepack, A/52 (AC3), ADX, Speex, AIFF, ALAC, Wavpack, Shorten, Monkey's Audio, SID, MOD, NSF, NSFE, SPC, MIDI, SAP
01:22:53rasherThey'll spot the ones they care about though
01:23:43*rasher is unsure whether to add "MPEG Audio" separately
01:23:58LloreanMP1/2/3 maybe?
01:24:08LloreanOr MP3/2/1 since people will notice MP3
01:24:41rasherGah, that pushed it into being 4 lines..
01:24:52 Join lightbulbjim [0] (n=jim@203.171.93.108.static.rev.aanet.com.au)
01:25:17scorche|shso does anyone want to have a guess at user numbers? (for press)
01:25:39rasherscorche|sh: I think Bagder blogged about that at one point
01:25:52 Quit tmzt (Connection reset by peer)
01:26:05LloreanWe have at least 10,000 people who've posted in the forums and not had their posts removed as spam.
01:26:44scorche|shabout download numbers from one server, but as a guess for total users?
01:27:25rasher10k forum posters. I'd say that *at most* 1 in 5 users posts on the forum
01:27:59rasherEven if some of those posters aren't actually users (new ports etc)
01:29:01rasherNew version: rasher.dk/rockbox/handout3.pdf">http://rasher.dk/rockbox/handout3.pdf
01:29:40saratogasome idiot on wikipedia deleted the rockbox screen shot, anyone with an account feel like fixing it?
01:31:14rasherApparently you need to be "autoconfirmed"
01:32:51saratogai guess they didn't like the album art
01:33:38Unhelpfulwas it "The Virgin Killer"? ;P
01:34:01LloreanThey have album art and book covers all over wikipedia, the usual policy is "it has to be small and low resolution"
01:34:52linuxstbLlorean: In what context? Discussing that album/book, or elsewhere (like the Rockbox image) ?
01:35:56LloreanUsually the page for the album or book, yes.
01:36:37*linuxstb sees there's still a Sansa screenshot with album-art on the Rockbox wikipedia page
01:37:28gevaertsShouldn't the handout list supported players?
01:37:41rasherAnd this: http://en.wikipedia.org/wiki/File:IPod_Classic_6th_Generation.jpg ?
01:37:59rashergevaerts: I was thinking about that
01:38:02saratogaeven better, wiki has an actual copy of the album art: http://en.wikipedia.org/wiki/Wednesday_Morning,_3_A.M.
01:38:17saratogaso they object to our screen shot of a scaled down version, but the high res is ok
01:38:37rashersaratoga: But that's with a fair use rational that it's discussing the work in question
01:38:43rasherI see where they're coming from.
01:38:51LloreanI've seen album covers in other articles though
01:38:54rasherSurely there are albums with free album art
01:38:58rasherTo avoid this completely
01:39:01LloreanFor example, there's a couple album covers in the Album_Covers article
01:39:04 Quit JdGordon (Read error: 110 (Connection timed out))
01:39:14LloreanBut yeah, maybe we should find some free cover art
01:39:17LloreanOr even just use a placeholder image
01:40:54soapinfoweek interview tomorrow
01:42:54saratogawe could use defiance ohio, since they Creative Commons their stuff
01:45:34 Join JdGordon [0] (n=jonno@123-243-140-31.static.tpgi.com.au)
01:49:33 Join lymeca [0] (n=lymeca@157-157-71-184.dsl.dynamic.simnet.is)
01:50:35soapa
01:51:31 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
01:51:52 Join FlynDice [0] (n=jack@c-24-19-225-90.hsd1.wa.comcast.net)
01:52:29 Quit bmbl (Client Quit)
01:53:11rasherNew version on its way with supported players listed
01:53:34rasherrasher.dk/rockbox/handout4.pdf">http://rasher.dk/rockbox/handout4.pdf
01:54:20gevaertsrasher: maybe s/not /work is going on for the / ?
01:54:49gevaertsor just "not yet"
01:55:07rasherYeah
01:57:13 Quit kachna (Read error: 113 (No route to host))
01:57:22rasherv5 uploaded
01:57:48Unhelpfulshould editing the config-<target>.h suffice to change the plugin buffer?
01:59:00gevaertsrasher: looks good to me (except that I think that http URLs should start with http:// :)
01:59:33rasherI considered that.. but then it looks all techy and gross
01:59:55gevaertsThe capital R looks a bit weird to me though
02:00
02:00:23*gevaerts doesn't see domain names as english sentences, so he doesn't think english grammar applies to them
02:01:03LloreanUnhelpful: It has for me in the past, at least.
02:01:27rashergevaerts: better now? (v6)
02:02:10rashersoap: go go gadget Kinko's!
02:02:23gevaertsrasher: if I'm allowed to pick one last nit, maybe the logo is a bit close to the text...
02:02:25soapthanks again
02:02:44rashergevaerts: the whole thing is a bit cramped.. I'd need to make the logo smaller?
02:02:44gevaertsThat's really minor though
02:02:56rasherNot sure that'd look terribly nice
02:03:03rasherIt's already a bit weird not being full-width
02:03:20gevaertsrasher: maybe just moving the text down a tiny bit to move it closer to the table?
02:03:29rasherI'll see
02:04:48rasherYeah, that works..
02:05:32*rasher is doing this really horribly in openoffice.org
02:05:41rasherAs a curiousity, the .odt is nearly exactly the same size as the pdf
02:05:57rasherrasher.dk/rockbox/handout7.pdf">http://rasher.dk/rockbox/handout7.pdf
02:06:18 Join Lss [0] (n=Lss@cm246.delta140.maxonline.com.sg)
02:07:00 Quit `Assassin (Read error: 104 (Connection reset by peer))
02:07:01 Join z35_ [0] (n=z35@h69.149.18.98.dynamic.ip.windstream.net)
02:07:12*gevaerts has more suggestions that take lots of space, so he'll shut up now :)
02:07:46 Join flyinghippo [0] (n=flyinghi@pool-173-72-208-154.clppva.east.verizon.net)
02:08:07 Nick z35_ is now known as help (n=z35@h69.149.18.98.dynamic.ip.windstream.net)
02:08:18 Nick help is now known as z35_ (n=z35@h69.149.18.98.dynamic.ip.windstream.net)
02:13:48 Nick z35 is now known as Guest49058 (n=z35@h78.29.89.75.dynamic.ip.windstream.net)
02:14:28 Quit bertrik ("Leaving")
02:15:53 Nick z35_ is now known as z35 (n=z35@h69.149.18.98.dynamic.ip.windstream.net)
02:21:11 Join blkhawk- [0] (n=blkhawk@f051004010.adsl.alicedsl.de)
02:23:01 Quit mcuelenaere ()
02:25:31Unhelpfulhrm, after i get resizing of allocated buffers done, does this sound like an interesting feature, as well? i think it might be useful to have a buf_out, or whatever might be an appropriate name, that moves the data for a handle before all other allocated buffers, moves the the start of the allocation pool to after that data, and return a pointer to the start of the data.
02:26:07Unhelpfulbasically, convert a previously dynamic allocation to a static one. could be useful for some things i've seen in PF, already.
02:29:36 Quit Guest49058 (Read error: 110 (Connection timed out))
02:29:52Unhelpfullike generating the cache of album titles, which is basically a big collection of null-separated strings, and an array of structures that point into it. in svn, these are both static allocations. in my local branch, i'm taking the strings off of the plugin buffer, but the index is still a static allocation. with a buf_resize and buf_out, both could be allocated as much space as they need, and then the pushed out of the buffer, with thei
02:29:52Unhelpfulr allocation overhead returned to the buffer pool, and their addresses fixed.
02:32:12***Saving seen data "./dancer.seen"
02:34:52 Quit blkhawk (Read error: 110 (Connection timed out))
02:35:09 Nick blkhawk- is now known as blkhawk (n=blkhawk@f051004010.adsl.alicedsl.de)
02:36:38 Quit Rob2222 ()
02:41:24 Quit domonoky (Read error: 104 (Connection reset by peer))
02:50:58*linuxstb beastpatches his beast
03:00
03:02:34 Join Darksair [0] (n=user@221.221.151.10)
03:02:56 Quit jhMikeS (Nick collision from services.)
03:03:02 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
03:06:42linuxstbjhMikeS: Do you think the untar stuff can be removed from the Beast bootloader now?
03:08:15linuxstbjhMikeS: Also, are you happy for a beast bootloader to be released now? Or do you have any imminent changes planned?
03:08:23 Quit tyfoo (Read error: 104 (Connection reset by peer))
03:13:13Unhelpfullinuxstb: the untar stuff is (one way) around the "file not found" error
03:22:38 Quit flyinghippo ("Nettalk6 - www.ntalk.de")
03:25:40 Join `Assassin [0] (n=blank@71-8-56-51.dhcp.leds.al.charter.com)
03:33:00linuxstbUnhelpful: Hmm, is that worth keeping it for? What are the other ways? I had that error once, but doing a recovery fixed it.
03:36:31Unhelpfulrecovery can fix it, but that's proven quite unreliable for mine. formatting it can fix it, but if the parameters are not right it will just force a recovery, and even if they are, that may be a cause of the random recovery problem some have
03:38:25linuxstbSo you think it's worth keeping?
03:46:15soapOntario Linux Fest (october) really really wants Rockbox to be there and mingle with their embedded group
03:48:07Unhelpfullinuxstb: hard to say. i've also had the tar "trick" fail to resolve the issue. there must me some reason that files written by our bootloader can then not be read by our bootloader, but i have no idea what the cause might be :/
03:49:03linuxstbsoap: How much ($$?) do they want us? ;)
03:52:43soapThey want to hook Rockbox up with horny^W willing embedded devs.
03:55:45jhMikeSlinuxstb: I have no objection to releasing it but the untar stuff has proven to be a nice backup to getting a build in place and/or running. For instance, fixing the 'File Not Found' problem unless that has been solved in a more geniuine manner?
03:56:38linuxstbjhMikeS: OK, I'll leave it then. If you haven't fixed that problem, I don't think anyone has...
03:57:07jhMikeSI think some modification to the FAT driver to better handle this TFAT stuff will be needed for that.
03:57:19linuxstbI've noticed a "to-do" to implement charging in bootloader USB mode - is that still a to-do?
03:58:31jhMikeSyes, it's till a TODO. I'm wondering if it's even desirable.
03:58:45linuxstbWhat's the downside?
03:59:00Unhelpfulbeast charging is a bit complex, isn't it?
03:59:06jhMikeSnone but there's no monitoring for it so you can't tell what it's up to
03:59:26jhMikeSUnhelpful: Tell freescale to fix the mc13783 and make it easier :)
03:59:43linuxstbYou mean there's no monitoring in general?
03:59:43Unhelpfulbeast has quite a bit of memory, doesn't it? is there a reason it has the 512KB plugin buffer instead of 1MB?
04:00
04:00:01jhMikeSlinuxstb: the power thread would but there's no indicator on the screen about it
04:00:14jhMikeSUnhelpful: It has 64MB
04:00:15linuxstbBut that's fixable though?
04:01:01Unhelpfulmaybe a "mini-statusbar" could draw a battery or plug icon, without adding too much code or data?
04:01:12jhMikeSlinuxstb: Yeah. Throw some little GUI widget together. The rest is simply starting power management and turning if off properly. I wouldn't bother starting that unless a USB session will be started.
04:01:19linuxstbIs code/data an issue though?
04:01:55jhMikeSno issues other than letting the user know what its doing
04:02:35jhMikeSbinsize isn't an issue at all for the bootloader. it would take quite a bit more size to notice any load time change.
04:03:05linuxstbAlthough if rockbox is present on the device, is there a reason to go into bootloader usb mode (by default) ?
04:03:32linuxstbI guess we would want the option (on a button press), but by default couldn't we just start Rockbox?
04:03:44Unhelpfullinuxstb: filesystem repair, perhaps?
04:03:51 Join Barahir_ [0] (n=jonathan@Zc78b.z.pppool.de)
04:03:53jhMikeSYeah, if I'm developing and mess up rockbox too much :)
04:04:18linuxstbThat's why I said there would still be usb bootloader mode, but only on a button press.
04:04:48linuxstbMainly because Rockbox USB mode has a shiny logo...
04:05:36jhMikeSI don't know, other targets just do USB mode if the cable is inserted.
04:06:28jhMikeSFrankly I'd find it annoying to have to hold a button. What makes the beast special or require any different treatment in that regard?
04:06:38linuxstbYes, it would be inconsistent...
04:07:13linuxstbIt just seems nicer from a "normal user" point of view for Rockbox to still load when usb is attached.
04:08:01linuxstbDoes the USB detection distinguish between a computer and a USB charger?
04:09:49 Join blkhawk- [0] (n=blkhawk@g228071036.adsl.alicedsl.de)
04:12:48jhMikeSIt won't at this point, it'll just sit and wait for the bus reset.
04:13:43jhMikeSI've never checked a USB charger on it actually (though I have one for a cell phone).
04:16:14jhMikeSI did check it by insulating the data lines though so I guess that's the same?
04:16:32 Quit AndyI (Read error: 60 (Operation timed out))
04:17:37jhMikeSIf so, there's the matter of it not drawing full current anyway unless the host says it's ok.
04:17:44 Join AndyI [0] (i=AndyI@212.14.205.32)
04:18:12 Quit perrikwp ("http://www.mibbit.com ajax IRC Client")
04:19:49 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/mibbit.com/x-edb41a76baf6b7ce)
04:20:03soapfair to say the d2 "holdup" is primarely one of internal flash?
04:20:33linuxstbjhMikeS: I've just tested it - in Rockbox itself it works fine (the charger icon appears), but in the bootloader it says "Waiting for USB".
04:20:40 Quit Barahir (Read error: 110 (Connection timed out))
04:20:59linuxstbsoap: Yes, I think so.
04:21:39saratogafrom what I understand, the flash issue seems to be making some progress
04:24:09jhMikeSlinuxstb: yeah, you'll get around 90mA max (highest charger setting <= 100mA) unless we should just drop this "compliance" stuff and not care like most devices (which would fix driver popups too).
04:24:43linuxstbjhMikeS: But what's different between the bootloader and Rockbox?
04:25:54soapare those issues something that someone with experience WRT raw flash access could help?
04:26:08 Quit blkhawk (Read error: 113 (No route to host))
04:26:23jhMikeSThe core waits for the BUS reset and doesn't inform threads about USB unless it gets that. The bootloader waits in a loop for the USB driver to broadcast SYS_USB_CONNTECTED.
04:26:30linuxstbsoap: I don't think so - the issue is reverse-engineering the telechips flash translation layer.
04:26:47 Nick blkhawk- is now known as blkhawk (n=blkhawk@g228071036.adsl.alicedsl.de)
04:27:16linuxstbsoap: But I think shotofadds would welcome any help...
04:27:27jhMikeSThe bootloader only initializes these extra things if it finds the port is powered (USB4V4S)
04:28:40soapohh - we don't have to totally replicate? We can use?
04:29:51linuxstbsoap: ?
04:30:17soappoorly worded question? :)
04:31:05linuxstbYes ;)
04:31:15soapWill the telechips provide the flas access if only we use it right, or are you saying we need to learn from how it does it?
04:32:16***Saving seen data "./dancer.seen"
04:33:29linuxstbWe can access the raw data on the flash (that was implemented months ago).
04:33:47soapbut write?
04:34:23linuxstbThe issue is that if we want to maintain compatibility with the OF, we need to use the same wear-levelling algorithm.
04:35:17linuxstbshotofadds has said in the past that he didn't expect to understand it well enough to implement writing. But I think he's getting more confident. And yes, I'm sure he knows how to physically write to the flash.
04:36:32soapCurious as I've gotten the question thrice today "what specalist does Rockbox need hooked up with to finish the d2"
04:38:25linuxstbMaybe you should post that question to the forum thread - I'm not sure how to answer that....
04:39:20linuxstbBut even after the flash is working, there are other major things to work on - e.g. DAB radio support. That's likely to require a lot of OF disassembly.
04:47:06UnhelpfulFS #9919 is updated with my lastest work on playback-friendly pictureflow. i'd appreciate if somebody could give it a test, it works well on my e200 and beast, but needs plugin buffer size changed to 1MB on targets that use a smaller one.
04:49:58 Quit tvelocity (Remote closed the connection)
04:50:58linuxstbUnhelpful: Do any targets use a 1MB plugin buffer?
04:52:58 Quit jhMikeS (Nick collision from services.)
04:53:04 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
04:55:37Unhelpfullinuxstb: not very many. i've talked to Nico_P about a plugin_get_audio_buffer-like method that would allow the caller to claim only part of the buffer, and then rebuffer and resume playback. another alternative would be to allow rb->bufalloc to force the tail of the buffered data out to satisfy allocations, without forcing a complete rebuffer.
04:57:11Unhelpfulgrep says 1MB plugin buffers are the norm on creative and onda targets, as well as m:robe 500
04:58:48Unhelpfulbeing able to allocate from the audio buffer would probably be preferable to increasing the "normal" size of plugin buffer. the code and static allocations take up most of the 512KB buffer on ARM, so there's fairly little left for the cache
04:59:32 Join Barahir [0] (n=jonathan@Xd269.x.pppool.de)
05:00
05:00:41Unhelpfuleven if "no-fail bufalloc" ends up being the way to go on swcodec targets, the FS #9916 stuff may still be useful on hwcodec, where bufalloc doesn't exist
05:01:35soapthanks for your time, linuxstb
05:04:30 Quit gromit`` (Read error: 110 (Connection timed out))
05:06:28 Quit Aurix_Lexico (Remote closed the connection)
05:12:24 Quit miepchen^schlaf (Read error: 101 (Network is unreachable))
05:16:13 Quit Barahir_ (Read error: 110 (Connection timed out))
05:40:46 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
05:43:46lostlogicanyone built UISIM on mac with macports SDL (instead of Fink)?
05:52:56 Quit Zoxc ()
05:57:34 Join webguest67 [0] (n=62c6b8b8@gateway/web/cgi-irc/labb.contactor.se/x-36ef191a8829a183)
05:58:08webguest67I have a quick question, How do i reset my Sansa e 200 to the way it was?
05:59:36webguest67?
05:59:45saratogadepends what you did to it
06:00
06:00:02 Quit mirak (Remote closed the connection)
06:00:04webguest67Only installed Rockbox, and thats about it.
06:01:00saratogaif you're asking about uninstalling, the uninstall section in the manual is a good place to begin
06:01:23webguest67Alright, nevermind then. Ill go check that out and come back if i have a problem.
06:02:02webguest67I uninstall rockbox, the boot loader and the rockbox itself right?
06:03:46webguest67?
06:04:24 Quit webguest67 ("CGI:IRC (EOF)")
06:12:02 Join tmzt [0] (n=tmzt@adsl-99-164-41-91.dsl.akrnoh.sbcglobal.net)
06:21:19 Quit lightbulbjim ("leaving")
06:22:17 Quit Lss (Read error: 110 (Connection timed out))
06:29:53 Quit CaptainKewl (Read error: 110 (Connection timed out))
06:32:20***Saving seen data "./dancer.seen"
06:32:37 Join midgey [0] (n=tjross@c-71-205-31-207.hsd1.mi.comcast.net)
06:33:41midgeyrasher, n1s, et al: I'm still alive, I just haven't had much time for Rockbox work. I'll be giving plugin translations some much needed love over the course of next week.
06:44:23 Join jack_ [0] (n=jack@c-24-19-225-90.hsd1.wa.comcast.net)
06:46:18 Quit fdinel ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
06:49:10 Join lightbulbjim [0] (n=jim@203.171.93.108.static.rev.aanet.com.au)
06:59:05 Quit FlynDice (Read error: 110 (Connection timed out))
07:00
07:05:21 Quit midgey ()
07:08:25 Quit awake_ (Read error: 104 (Connection reset by peer))
07:16:12 Quit intrados (Remote closed the connection)
07:26:19 Join awake_ [0] (n=Zzz@c122-106-37-8.rivrw1.nsw.optusnet.com.au)
07:48:42 Quit saratoga ("CGI:IRC (EOF)")
08:00
08:13:51amiconnUnhelpful: Imo the plugin buffer shouldn't be increased further. Plugins are not core functionality.
08:14:57amiconnThat said, PF could use an intelligent method using its own allocator. When music playing, use only the pkugin buffer in order to not interrupt playback, but when music is stopped to begin with, use the audio buffer for better caching
08:15:39amiconnSelecting some music to play from within PF can be fixed independently of all this
08:16:37 Join kachna [0] (n=kachna@r4ax178.net.upc.cz)
08:16:49amiconnOn lowmem targets (hwcodec and swcodec) PF will have to use the audio buffer, but still can be made to start the selected music
08:19:22 Join n1s [0] (n=n1s@rockbox/developer/n1s)
08:26:17 Join J-23 [0] (n=zelazko@unix.net.pl)
08:32:24***Saving seen data "./dancer.seen"
08:51:41 Quit BHSPitLappy (Remote closed the connection)
08:58:01 Join alexbobp [0] (n=banmenot@66.112.249.147)
08:58:20 Join LinkinX64 [0] (n=linkinx6@c-76-101-167-191.hsd1.fl.comcast.net)
08:58:23 Part LinkinX64 ("Leaving")
09:00
09:29:05lostlogicfinally got the UISim mostly building on os x, but still can't get the codecs to link, ugh
09:38:16 Quit z35 ("Leaving")
09:54:18*JdGordon doesnt remmber having problems with th sim and osx... that was pre the make rework though
09:58:37 Quit lymeca (Read error: 110 (Connection timed out))
10:00
10:03:57lostlogicweird, lots of probs here −− first codec_crt0.c would crash the compiler (weird assembler error), so I switched manually froj /usr/bin/gcc to /usr/bin/gcc-4.2 in the makefile and that got me most of the way there
10:04:16lostlogicnow it can't link codecs, a symbol in codec_crt0.c isn't found "_ci"
10:08:32lostlogicJdGordon: shouldn't you like already be on a plane or something? :-P
10:10:32 Quit `Assassin (Read error: 110 (Connection timed out))
10:16:26 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
10:19:50LloreanCan we do something to get rid of that prompt for drivers when we plug in a Rockbox SWUSB target that's just wanting to charge and not mount?
10:21:17 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
10:21:47amiconnIiuc this would require to implement a dummy HID driver in rockbox (or another class that windows knows how to handle generically)
10:22:35amiconnMaybe it could even be UMS, but without exposing the drive. gevaerts?
10:23:30JdGordonlostlogic: 18 hours to go or something...
10:32:26***Saving seen data "./dancer.seen"
10:32:37 Join spitfire_ [0] (n=quassel@abjf68.neoplus.adsl.tpnet.pl)
10:35:18 Part spitfire_
10:35:44 Join spitfire_ [0] (n=quassel@abjf68.neoplus.adsl.tpnet.pl)
10:41:15 Join merbanan [0] (n=banan@c-83-233-243-9.cust.bredband2.com)
10:43:55 Join flydutch [0] (n=flydutch@host25-43-dynamic.5-87-r.retail.telecomitalia.it)
10:50:02 Join bertrik [0] (n=Bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
11:00
11:03:21 Join maddler [0] (n=maddler@static-217-133-171-24.clienti.tiscali.it)
11:04:21 Join {phoenix} [0] (n=dirk@p54B464D0.dip.t-dialin.net)
11:06:09 Join ender` [0] (i=krneki@foo.eternallybored.org)
11:15:09 Join Toni1 [0] (n=91fec189@gateway/web/cgi-irc/labb.contactor.se/x-33378c996cbf879c)
11:15:59Toni1I just commited a fix for the data aborts in FS #9827.
11:16:04 Quit Darksair ("Emacs = ESC-Meta-Alt-Ctrl-Shift")
11:17:33Toni1Misalignment and wrong placing of a memory handle should be avoided by this commit.
11:24:29n1sgreat :)
11:26:14Toni1It was an old bug, which only occured in rare cases, but got obvious with bigger album art memory handles.
11:28:56 Quit Toni1 ("CGI:IRC (EOF)")
11:29:05n1sToni1: shouldn't the bug report be closed now then?
11:30:57 Join Cylvre [0] (n=47572ce2@gateway/web/cgi-irc/labb.contactor.se/x-f25e825e26a19f20)
11:32:51 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
11:32:51 Quit Cylvre (Client Quit)
11:35:19 Join Cylvre [0] (n=47572ce2@gateway/web/cgi-irc/labb.contactor.se/x-3d1f66f46502e086)
11:37:37CylvreI'm having a problem with my Sansa e260: The database information in Rockbax is showing the data from when I first installed Rockbox after I load completely different data on the player and refresh the database in the original firmware. Database is correct in original firmware and the new files play properly.
11:37:42 Quit stripwax (Read error: 54 (Connection reset by peer))
11:38:17CylvreAny suggestions?
11:39:14LloreanRockbox uses its own database
11:39:32LloreanRefreshing it in the original firmware won't affect Rockbox either way. You need to have the Rockbox database update, as described in the manual.
11:41:10 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
11:42:53CylvreAh, was confused there. I thought with the need to transfer while in the original firmware it utilized that database as well. Thanks Llorean.
11:43:49LloreanIs there something in our documentation that gave this impression, that we can correct?
11:46:38CylvreNo, I had confused myself into thinking it when I was trying to explain the process of uploading music to the player to my technologically illiterate mother. All of her questions got me turned around and I see now she and I both ended up a bit confused.
11:51:28 Quit Cylvre ("CGI:IRC (EOF)")
11:57:14 Quit thegeek (Read error: 113 (No route to host))
11:58:02 Join thegeek [0] (n=nnscript@s243b.studby.ntnu.no)
12:00
12:03:13 Join Darksair [0] (n=user@221.221.151.10)
12:12:44 Quit gevaerts (Nick collision from services.)
12:12:49 Join gevaerts_ [0] (n=fg@195-144-092-166.dyn.adsl.xs4all.be)
12:16:45 Join Rob2222 [0] (n=Miranda@p4FDCD606.dip.t-dialin.net)
12:22:48 Quit merbanan (Remote closed the connection)
12:32:29***Saving seen data "./dancer.seen"
12:33:46 Join fml [0] (n=4fd3e5ba@gateway/web/cgi-irc/labb.contactor.se/x-558ecdb7acb22f4c)
12:34:50fmlHello. Is there a make target that would build all the fonts?
12:36:05fmlpixelma: there are some fon patches in FS #9931 for the fonts introduced by in in r13325
12:37:07 Quit kachna (Remote closed the connection)
12:38:06 Join MethoS- [0] (n=lem@host-091-097-240-187.ewe-ip-backbone.de)
12:42:52 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
12:51:05 Join gregzx [0] (n=chatzill@cmn9.neoplus.adsl.tpnet.pl)
12:51:31 Quit {phoenix} (Remote closed the connection)
12:53:25JdGordonbluebrother: (for the logs) jdgordon.info links wont work for the rbutil osx builds soon... change it to gordonmob.com (the rest being the same)...
12:53:41 Join The-Compiler [0] (n=compiler@unaffiliated/the-compiler)
12:54:02The-CompilerUh, hi there, I wasn't here for some months to years now :O
12:59:55 Join Buschel [0] (n=AndreeBu@p54A3D269.dip.t-dialin.net)
13:00
13:06:18 Join Thundercloud [0] (n=thunderc@84-51-130-71.judith186.adsl.metronet.co.uk)
13:08:36bertrikThe-Compiler, I hope you're running a current version of rockbox then, quite a few things have improved in the past year
13:08:36 Join at0m|werk [0] (i=a548c80b@gateway/web/ajax/mibbit.com/x-13b2f3d81aa8f292)
13:10:54The-Compilerbertrik: I'm not running rockbox at all since albumarts came up, because then my iPod broke (meh, I'll never get an apple player again) and I couldn't really afford another player (I'm student) as I already had my Nintendo DS running homebrew as my MP3 player
13:11:27at0m|werkhi, do i need to put .rockbox folder on the first partition? IAW, can the first partition be ext3 and the next vfat?
13:11:59The-Compilerat0m|werk: I assume you're talking about an iPod?
13:11:59linuxstbat0m|werk: What device are you running Rockbox on?
13:12:12at0m|werkh340 w 60gb disk
13:12:33linuxstbRockbox will mount the first FAT32 partition it finds, and only that partition.
13:12:53at0m|werkoki so if the first one isnt vfat, that'd be ok
13:12:59linuxstbBut the original firmware may be different...
13:13:48at0m|werkthanks. but who would want to go back to orig when there's rockbox :)
13:14:21The-Compilerhow true :D
13:17:04 Quit spitfire_ (Read error: 54 (Connection reset by peer))
13:22:30*Buschel checked FS #9863 while going through open patches/bugfixes and verified it
13:22:50Buschelanything speaking against submitting it?
13:23:59 Join fml [0] (n=4fd3e5ba@gateway/web/cgi-irc/labb.contactor.se/x-2816017389f2cdfa)
13:24:14fmlJdGordon: you here?
13:24:23JdGordonhey
13:25:13fmlJdGordon: there's still a bug with the status bar in the (at least) clock plugin. Latest build (r20081) on sansa e200.
13:25:35JdGordonwhats the problem?
13:26:21fmlMy setting is "status bar=ON". When I call the menu from the clock plugin, the status bar isn't drawn. I.e. I see the remnant of the clock screen where the SB should be. This happens not verey time though.
13:26:30fml*every
13:26:42*kadoban points to FS #9922
13:27:48fmlkadoban: ah, I searched for "status bar" but oversaw that report!
13:27:56JdGordonyeah, ive seen the fix for that.. havent done any rb work for a while though.. maybe later in the week unless someone wants to commit that before me
13:28:06JdGordonim not sure its the best fix.. but it works
13:28:21JdGordonthe general problem is there is too much screen clearing
13:28:54kadobanoh, does the statusbar get drawn and then cleared off or something?
13:29:38JdGordonyes
13:29:54JdGordonand because it only redraws if it needs to it sometimes doesnt happen
13:30:06JdGordonon rtc targets the worst is 1min between redraws
13:30:14JdGordonon non-rtc it may apear to never redraw
13:30:15fmlBTW: we have the wonderful clock plugin. Does it have the countdown functionality?
13:30:27*JdGordon doesnt know
13:30:33kadobanyeah, that makes sense. if i get some time, i'll see if i can find the better fix for 9922
13:31:45JdGordonadding the forced redraw when the menus enter i fine... but not great
13:32:37JdGordonalthough... there should be blank where the bar should b... not remnants of the old screen
13:33:07kadobanyeah, that's what i was just wondering about
13:34:54 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
13:44:55 Quit J-23 ("ZNC - http://znc.sourceforge.net")
13:47:59 Join lasser [0] (n=chatzill@W8028.w.pppool.de)
13:49:46 Quit perrikwp ("http://www.mibbit.com ajax IRC Client")
13:52:14*linuxstb implements on-the-fly MBR fixing for the beast, but isn't sure if it's clean enough to commit - linuxstb.cream.org/rockbox/validate_mbr.diff.txt">http://linuxstb.cream.org/rockbox/validate_mbr.diff.txt
13:52:49*linuxstb pings amiconn and gevaerts_ for comments...
14:00
14:01:22 Quit Horschti ("Verlassend")
14:07:24 Quit JdGordon (Read error: 110 (Connection timed out))
14:08:29amiconnlinuxstb: The bootflag fixing can be done simpler: *p &= 0x80;
14:09:01linuxstbDoesn't that make it always active though? I've no idea what that flag actually means in practice...
14:09:37amiconnNo, it just masks the MSB
14:10:03*linuxstb drinks some more coffee and tries to wake up
14:10:17linuxstbYes, that would be better.
14:13:04linuxstbBigBambi: Can I just give you a binary of beastpatcher - to test it works on your system? I can give you the source as well if you want it.
14:13:15BigBambilinuxstb: yep, no problem
14:13:25BigBambiI can do linux or windows as you wish
14:14:46BigBambihmmm, one thought though - I suspect I need to be in the OF to do this?
14:14:59BigBambiIn order to have MTP ness
14:15:08linuxstbYes ;)
14:15:25 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
14:15:30BigBambiI'll have to stick a dual-boot bootloader on first then
14:15:34BigBambigive me a minute to make one :)
14:15:55linuxstbOr just copy the OF back
14:16:11linuxstbRunning beastpatcher will restore the RB bootloader.
14:16:14BigBambiThat won't wipe the drive?
14:17:23linuxstbI can see that it should be any different to putting a dual-boot bootloader there.
14:17:32BigBambiOK then :)
14:17:39linuxstbBut don't blame me...
14:17:52BigBambiI have it all backed up of course, it is just a 120 GB drive takes a while to fill :)
14:17:55BigBambilinuxstb: I won't
14:18:01BigBambiand then we will know!
14:18:07linuxstblinuxstb.cream.org/rockbox/beastpatcher-test1.tgz">http://linuxstb.cream.org/rockbox/beastpatcher-test1.tgz - the Linux binary
14:18:18linuxstb(I haven't done the Windows part yet)
14:18:33linuxstbYou just run it without any parameters.
14:19:07linuxstbThe bootloader does the on-the-fly MBR fixing, but I assume you've already fixed that.
14:20:27BigBambiI have
14:20:37BigBambiBut I can unfix it if you like
14:21:24linuxstbIt's OK, that's not important.
14:21:32BigBambiOK
14:22:19linuxstbI'm curious if you get the same libusb errors that I get. It works, but both libusb and libmtp give errors - I'm guessing due to the beast rebooting after the firmware is uploaded, but before beastpatcher cleanly exits.
14:22:48linuxstbCurrent sendfirm gives the same problems, but I can't remember them with sendfirm in the past (maybe it's libmtp 3.0)
14:24:16 Join JdGordon [0] (n=jonno@123-243-140-31.static.tpgi.com.au)
14:24:17 Quit JdGordon (Read error: 104 (Connection reset by peer))
14:24:40BigBambiOK, well I have the OF back :(
14:26:05BigBambihmmm, the OF won't connect to my laptop
14:26:27linuxstbamiconn: Any other comments on that patch? I don't like the fact that it hacks target-specific code into the usb storage driver, but it's relatively unobtrusive....
14:26:54BigBambiah, different cable fixed it
14:27:39BigBambilinuxstb: Worked perfectly
14:27:46linuxstbAny error messages?
14:27:54BigBambilinuxstb: yep
14:27:59BigBambione mo
14:29:12BigBambilinuxstb: http://pastebin.com/m5bb936c6
14:29:18 Join J-23 [0] (n=zelazko@unix.net.pl)
14:29:45 Join MethoS-- [0] (n=lem@host-091-097-245-241.ewe-ip-backbone.de)
14:29:47linuxstbYes, that's exactly what I get.
14:30:00BigBambilook like the standard it has already shut off
14:30:19fmlIn the clock plugin, binary mode, I think that the bitmaps are used the wrong way. I'd rather use a black one for "1" and an "empty" one for "0". Any opinions?
14:32:32***Saving seen data "./dancer.seen"
14:33:10linuxstbBigBambi: OK, thanks.
14:33:34 Quit MethoS-- (Read error: 104 (Connection reset by peer))
14:34:37BigBambilinuxstb: no problem
14:35:23linuxstbYou may want to change the install instructions now - I think you can remove the partition table section (I'll commit my patch after others have commented).
14:35:36BigBambilinuxstb: OK, will do
14:35:40 Quit Buschel ()
14:36:20linuxstbIt will be similar to the ipodpatcher/sansapatcher instructions - although maybe distributing the Linux binary in a .tgz is better than a binary the user needs to chmod (Windows will just be a .exe, OS X a .dmg)
14:36:32BigBambiRighto
14:36:58BigBambiI don't think a .tgz vs having to chmod makes much difference
14:37:15BigBambiWe shall either have to give instructions on how to extract or how to chmod
14:37:25BigBambialthough maybe extract is more gui
14:37:51Unhelpfulamiconn: a major problem is that PF takes up most of the .5MB buffer with its code and static allocations. this has not improved *much* by using smaller types for data sizes, and using a dynamic allocation for the album text cache helps, but will probably actually hurt if you have enough albums with long enough names
14:38:17BigBambilinuxstb: I note beastpatcher didn't need root - will it need admin on Windows?
14:38:39 Quit Acksaw ("Erection reset by queer")
14:39:04 Join Toni1 [0] (n=91febfe1@gateway/web/cgi-irc/labb.contactor.se/x-6bf0a3bd532975b2)
14:39:16Toni1Two comments on AA:
14:39:18Unhelpfulthere is room to cache about 6-7 covers in the plugin buffer on sansa, and it is a good deal worse on targets with larger screens.
14:39:50 Join jw5801 [0] (n=jw@CPE-124-191-3-57.vic.bigpond.net.au)
14:40:01linuxstbBigBambi: I've no idea, but I would guess not.
14:40:03Toni11)Do we really need to buffer the original album's bitmap track-times?
14:40:09BigBambiOK, cool
14:40:32 Join {phoenix} [0] (n=dirk@p54B464D0.dip.t-dialin.net)
14:40:47Toni12) Is it correct beahviour, that in case of a temporary buffer full state, the AA is silently ignored for that track?
14:41:20amiconnUnhelpful: Why does PF need to buffer album names?
14:41:38amiconnI would think it could read them from the database query result as needed
14:42:02Unhelpfulit only queries the database at init, perhaps the original author thought the DB too slow?
14:42:17linuxstbCouldn't the DB be on disk?
14:43:16amiconnThe database result is written to the dirbrowser buffer afaik. It only needs to hit the disk if there are more entries than this buffer can take (chunked browsing)
14:43:59BigBambilinuxstb: Will the error messages be surpressed, or shall I put a note in saying not to worry about them?
14:44:02 Join bertrik_ [0] (n=bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
14:44:14 Quit MethoS- (Read error: 113 (No route to host))
14:44:35amiconnIt would mean that browsing the individual tracks for an album would hit the disk (new db query), but it would potentially save a lot of ram iiuc
14:45:16amiconnNote that I didn't look up what the PF code actually does
14:47:30linuxstbBigBambi: I don't think I can suppress them, so yes probably best to mention them. I need to call the function that displays those error messages in order for the beast to accept the file.
14:47:42BigBambiOK, no problem
14:47:42 Quit Toni1 ("CGI:IRC (EOF)")
14:48:11Unhelpfulerm, sorry, i've overstated things, i believe it only queries DB for *album titles* at init. tracks are queried when an album is selected, and there is a bit of a delay.
14:51:21Unhelpfuli think it might be helpful, perhaps, to have a custom linker script that puts init-time-only code and data above the rest, with a pointer marking the division.
14:51:23fmlHow should I provide a patch for a bitmap?
14:52:00linuxstbYou can't - you just need to attach the new file(s)
14:52:56fmllinuxstb: ok, will do
14:54:59*linuxstb commits the first version of beastpatcher
14:59:13Unhelpfulhrm, that's only about 15KB, not even a whole cover. almost all of the space used by init-time-only belongs to the splash image
15:00
15:04:39Unhelpfuli'm not actually suggesting we increase plugin buffer to 1MB, only that that's needed to effectively test it, right now. going forward, it either needs to be able to force rebuffer and steal part of audio buffer, or have something like core bufalloc that can clear some buffered data to make sure it succeeds
15:07:00 Quit jw5801 (Read error: 110 (Connection timed out))
15:07:19amiconnThe splash image could be loaded from disk. At the time the plugin starts, the disk is spinning anyway
15:08:35amiconnLoad it, display it, then discard it. It will stay on the lcd until cleared
15:09:01Unhelpfulthe rest of the savings is pretty trivial
15:09:36amiconnProbably only pays of on colour targets - greyscale and mono native bitmaps are pretty small
15:09:40 Quit fml ("CGI:IRC (EOF)")
15:09:40amiconn*off
15:09:50 Join Aurix_Lexico [0] (n=comrade@c-68-56-205-239.hsd1.fl.comcast.net)
15:11:11 Join balug_ [0] (n=dvg@HSI-KBW-078-042-132-156.hsi3.kabel-badenwuerttemberg.de)
15:15:04mcuelenaeregevaerts_: I'm having 'attempt to access beyond end of device' dmesg errors with USB mass storage; could this because sd_get_info() doesn't give the right info?
15:15:06 Join flyinghippo [0] (n=flyinghi@pool-173-72-208-154.clppva.east.verizon.net)
15:27:17 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
15:31:55 Join CaptainKewl [0] (i=jds@207-237-172-77.c3-0.nyr-ubr4.nyr.ny.cable.rcn.com)
15:32:41 Join mcuelenaere_ [0] (n=mcuelena@d54C4BB07.access.telenet.be)
15:32:41 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
15:36:15 Join midijunkie [0] (n=Miranda@pD95442C4.dip0.t-ipconnect.de)
15:41:32 Quit Thundercloud (Read error: 104 (Connection reset by peer))
15:43:36 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
15:47:58 Nick mcuelenaere_ is now known as mcuelenaere (n=mcuelena@rockbox/developer/mcuelenaere)
15:48:35 Nick jack_ is now known as FlynDice (n=jack@c-24-19-225-90.hsd1.wa.comcast.net)
15:49:34 Join robin0800 [0] (n=robin080@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
15:51:01 Join bne [0] (n=ben@78.151.60.220)
15:52:47 Quit bne (Client Quit)
15:53:42 Quit bertrik_ (Remote closed the connection)
15:54:04 Join miepchen^schlaf [0] (n=miepel@p579EC58B.dip.t-dialin.net)
15:59:33 Quit bertrik ("Leaving")
16:00
16:02:39 Nick gevaerts_ is now known as gevaerts (n=fg@195-144-092-166.dyn.adsl.xs4all.be)
16:03:17gevaertsmcuelenaere: either sd_get_info gives a too low value or your filesystem lies (== is corrupt)
16:03:30mcuelenaereprobably the second
16:03:42mcuelenaerealthough RB can mount of it..
16:05:22gevaertslinuxstb: I'm not sure about the "validate_mbr" name, it seems to just mean "confirm validity" to me, not "fix it".
16:06:39gevaertsOn the cleanness issue, a generic approach would be a lot more complex, as it also has to handle changes in start sector and size changes, so I agree that as long as there's only this one, an ad-hoc fix is best
16:07:30gevaertsfor the record: I still don't really like the idea of changing data on the fly, but I seem to be alone there, so go ahead and commit this :)
16:21:03 Join Zoxc [0] (i=Zoxc@ti0128a340-dhcp0438.bb.online.no)
16:24:20n1slinuxstb: do you think it would be worth it to make beastpatcher handle future bootloader updates (i.e copy the file over msc to the firmware partition) ?
16:24:22 Quit balug_ ("Ex-Chat")
16:26:07n1samiconn: i see you tried the tool in UDMAonUSB, do you think the same thing could be done on the device (by rockbox)?
16:29:30n1slinuxstb: also why not for example make the beast bootloader writhe the fixed mbr to disk and be done with it?
16:29:41n1s*write even
16:29:53*gevaerts agrees with n1s :)
16:30:53gevaertsPeople seem to have this idea that it's OK to patch firmware and flash boot roms , but DON'T TOUCH THE MBR! THAT'S DANGEROUS!
16:32:35***Saving seen data "./dancer.seen"
16:34:45 Quit Seedy ("cu, Andre")
16:34:55Nico_Pis it possible to also hide the firmaware partition somehow? having two disks appear when you plug the beast in is annoying
16:35:57BigBambiYou lot are just trying to make me rewrite the install instructions again :)
16:36:05rashern1s: did you notice midgey stopped by?
16:36:16gevaertsAh yes, that one. That was IIRC actually the reason people proposed on-the-fly patching
16:36:46gevaertsIIRC just marking the partition as hidden makes the OF (or the rom?) angry
16:36:53n1srasher: i did not, todays logs?
16:37:03rashern1s: yup, around 6:00
16:37:26gevaertsBigBambi: we have to have some purpose in life :)
16:37:43 Join Seed [0] (n=ben@bzq-84-108-232-45.cablep.bezeqint.net)
16:37:52Nico_Pgevaerts: I guess we could write the "active" flag fix and patch the "hidden" flag on the fly
16:38:31n1srasher: that sounds like good news! :)
16:38:36rasherIndeed
16:38:42gevaertsNico_P: yes. The hidden flag is a bit more tricky though, as you do want to access that partition for updating bootloaders, so there needs to be a way around this
16:39:43n1sbutton press in the bootloader maybe and always hidden in rockbox?
16:40:08gevaertsthat works
16:41:20 Join bs66_1 [0] (n=sysuser@95.209.20.5.bredband.tre.se)
16:42:32 Quit midijunkie ("?(???~~)?")
16:44:54 Join Chesteta [0] (i=Chesteta@dyn244.wireless-104.ndsu.NoDak.edu)
16:48:21 Join stoffel [0] (n=sfr@p57B4C888.dip.t-dialin.net)
16:49:23 Quit bs66_ (Read error: 60 (Operation timed out))
16:49:43 Join ibseco [0] (n=ibseco@BAH7b4c.bah.pppool.de)
16:49:57 Quit ibseco (Read error: 104 (Connection reset by peer))
16:58:27 Join SirFunk_ [0] (n=Sir@208-15-25-145.netsync.net)
16:59:16 Join eightfold [0] (n=qwerty@c213-89-119-90.bredband.comhem.se)
16:59:24 Quit eightfold (Remote closed the connection)
16:59:49 Quit SirFunk__ (Read error: 60 (Operation timed out))
17:00
17:01:50 Join jaykay [0] (n=chatzill@p579E7481.dip.t-dialin.net)
17:03:08 Join rocko [0] (n=rocko@c-67-167-117-152.hsd1.il.comcast.net)
17:17:37 Join itcheg [0] (i=62db4767@gateway/web/ajax/mibbit.com/x-59272c37d00d2a78)
17:20:28*mcuelenaere got response from http://www.dealextreme.com/details.dx/sku.12753
17:20:30mcuelenaerewoops
17:20:38 Quit stoffel ("leaving")
17:20:44mcuelenaeres/http://www.dealextreme.com/details.dx/sku.12753/Carlos E. Vidales/
17:20:49 Join stoffel [0] (n=sfr@p57B4C888.dip.t-dialin.net)
17:21:29BigBambimcuelenaere: He responded via FM?
17:21:49mcuelenaerenah, just a bad copy-paste :)
17:23:07CtcpPing from gevaerts!n=fg@195-144-092-166.dyn.adsl.xs4all.be
17:24:04CtcpPing from gevaerts!n=fg@195-144-092-166.dyn.adsl.xs4all.be
17:25:21 Quit gevaerts ("Reconnecting")
17:25:31 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
17:25:32aviswould it be difficult to compile the usb test build for the e270 along with album art support ? the one that i found, for usb test build, doesn't seem to have support for album art. either that, or i dont know how to turn it on. :)
17:25:57BigBambialbum art support is there
17:26:01kadobanavis: album art is supported in the default build for a long time now
17:26:04BigBambiand has been for a long long time
17:26:15BigBambiHave you set up the covers properly?
17:26:24BigBambisee www.rockbox.org/wiki/AlbumArt
17:26:33avisthanks. i'll look at that
17:27:10Unhelpfulamiconn: on e200, it's still not enough for even one whole cover - cover cache items are about 24KB for e200, they run about 50KB on beast.
17:27:50Unhelpfulprobably even worse on landscape targets :/
17:27:50rasherBigBambi: or the manual!
17:28:58BigBambiyes!
17:29:04BigBambiI forgot it was now in there :)
17:29:48 Quit at0m|werk ("mibbit.com: reboot: runaway helpsvc.exe")
17:33:25 Quit miepchen^schlaf ()
17:37:51 Join mcuelenaere_ [0] (n=mcuelena@rockbox/developer/mcuelenaere)
17:37:59 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
17:42:59 Nick mcuelenaere_ is now known as mcuelenaere (n=mcuelena@rockbox/developer/mcuelenaere)
17:43:03 Quit Darksair ("Use the Force, Luke!")
17:43:26 Quit lasser ("ChatZilla 0.9.84 [Iceweasel 3.0.6/2009020409]")
17:50:15 Join balug [0] (n=dvg@HSI-KBW-078-042-132-156.hsi3.kabel-badenwuerttemberg.de)
17:53:13 Quit flyinghippo ("Nettalk6 - www.ntalk.de")
18:00
18:04:35 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:07:09 Join sarixe` [0] (n=sarixe@ool-43540968.dyn.optonline.net)
18:08:12 Join bombardier [0] (n=mcizmek@93-139-83-210.adsl.net.t-com.hr)
18:08:20bombardierhello
18:12:58 Join tyfoo [0] (n=tyfoo@77-20-31-238-dynip.superkabel.de)
18:16:56 Join dfkt_dt [0] (i=dfkt@unaffiliated/dfkt)
18:18:56bombardierCan somebody confirm me the status of Fuze port? I would like to contribute to development, but I don't want to brick my player on my first attempt. By following the forum disscussion, I got the feeling that the development has passed the point where some work can be done without the fear of bricking the unit. Is this correct?
18:20:23 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/mibbit.com/x-3e8592ae9ac733c7)
18:20:24 Quit perrikwp (Remote closed the connection)
18:20:45 Join bertrik [0] (n=bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
18:22:05 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/mibbit.com/x-c4e290fe6e4af83c)
18:22:49linuxstbbombardier: Nothing is guaranteed, but the dangerous code (which is the dualboot part of mkamsboot, not Rockbox itself or even the Rockbox bootloader) hasn't been changed for months, so it should be safe.
18:24:04gevaertslinuxstb: did you see the discussion earlier about beast USB and the MBR?
18:24:11linuxstbYes, I'm just about to respond...
18:24:33 Quit sarixe (Read error: 110 (Connection timed out))
18:24:44linuxstbI meant "validate" as in "make valid", but I see your point - any suggestions?
18:24:56gevaerts"fix"?
18:25:04gevaertsIt's a very minor thing though
18:25:16linuxstbI know, I'm working up to the bigger things ;)
18:25:53*gevaerts gets nervous :)
18:26:14 Join Gnutoo [0] (n=gnutoo@host76-59-dynamic.24-79-r.retail.telecomitalia.it)
18:27:32linuxstbRegarding writing the fixed mbr back - that would be slightly more complicated. i.e. we would need to test if it was invalid, and then only write it back in that case. I'm not really sure if there's any difference "morally" between secretly fixing the disk by writing to the MBR, or secretly fixing the disk on the fly...
18:28:33linuxstbn1s: Regarding beastpatcher, I think we should just keep it communicating with the beast via MTP. rbutil could potentially handle bootloader updates, once the first bootloader install it done.
18:30:01Gnutoohello, there is a creative zen 4GB with points at the supermarket...but if I understood well: 1)the is no port in progress for the 4GB model 2)the others models only have the bootloader working...so Here's my question...is the port active?...and how long should I wait for having something that works more or less?
18:30:32gevaertsI think that if we do more on the fly fixing (e.g. marking the firmware partition as hidden), the active flag can easily be done the same way, If it's just the active flag I'd really prefer to just fix it on disk
18:30:36BigBambiThere are no timelines
18:31:06BigBambiGnutoo: It is impossible to say. Even if people were working on it constantly they might hit an unforseen block.
18:31:09 Join toffe82 [0] (n=chatzill@expo-06.socallinuxexpo.org)
18:31:10Gnutooyes I know that there are no timelines...but I've no idea of how long a port usually takes
18:31:11gevaertsGnutoo: if you really want a port to work, you shouldn't wait for it :)
18:31:27BigBambiGnutoo: There is no usual time
18:31:31Gnutoook
18:31:36BigBambiGnutoo: Every single one differs
18:31:50Gnutoook thanks a lot
18:31:51BigBambiI wouldn't buy it with the intention of a port
18:31:58gevaertsAlso, current work is on the ZVM. I have no idea if that's even similar to the other zens
18:32:08krazykitunless you're planning on working on the port yourself, that is
18:32:36***Saving seen data "./dancer.seen"
18:33:02Gnutooand I bet I can't help because rockbox is complicated....(I only know ANSI C for now...and I will learn GNU/Linux kernel userland interfaces soon)
18:33:18BigBambiRockbox is NOT linux
18:33:23Gnutoos/GNU\/Linux/minix
18:33:29Gnutooyes I know...that's the problem
18:33:29scorche|shor that
18:33:40gevaertsrockbox is nearly 100% pure ANSI C
18:33:42rasherRockbox is pretty much jus tANSI C
18:33:45Gnutoodoes it use cooprating threading?
18:33:46 Quit dfkt (Read error: 110 (Connection timed out))
18:33:48Gnutoook
18:33:55BigBambiIt is written in C (mostly) and from scratch
18:33:56Gnutoos/cooperating/cooperative
18:33:58Gnutoook
18:34:30scorche|shhttp://www.rockbox.org/twiki/bin/view/Main/RockboxArchitecture
18:35:12gevaertsThreading is cooperative, yes
18:36:52linuxstbgevaerts: I guess I'm in favour of the on-the-fly solution because its the simpliest. I don't have any strong views though. I believe amiconn is in favour of on-the-fly?
18:38:45gevaertslinuxstb: I think so, yes
18:44:15 Join miepchen^schlaf [0] (n=miepel@p579EC58B.dip.t-dialin.net)
18:44:22 Part linuxstb ("Leaving")
18:45:11mcuelenaereGnutoo: currently there's no support at all for the Creative ZEN
18:45:28mcuelenaeresome people have looked into it, but no code
18:48:31Gnutoomcuelenaere, ok thanks a lot I won't buy it
18:48:39 Join WillDS [0] (n=nds@d199-126-180-153.abhsia.telus.net)
18:49:30mcuelenaereGnutoo: I don't want to discourage you, you could also look into porting yourself ;)
18:49:39Gnutoook
18:49:54*mcuelenaere tried porting Rockbox to the ZVM
18:50:37 Quit Seed (grisham.freenode.net irc.freenode.net)
18:50:37NSplitgrisham.freenode.net irc.freenode.net
18:50:37 Quit Zoxc (grisham.freenode.net irc.freenode.net)
18:50:37 Quit FOAD (grisham.freenode.net irc.freenode.net)
18:50:37 Quit maraz_ (grisham.freenode.net irc.freenode.net)
18:50:37 Quit Tuplanolla (grisham.freenode.net irc.freenode.net)
18:51:54WillDSI'm working on a Nintendo DS port. It isn't going so good.
18:52:02NHealgrisham.freenode.net irc.freenode.net
18:52:02NJoinSeed [0] (n=ben@bzq-84-108-232-45.cablep.bezeqint.net)
18:52:02NJoinZoxc [0] (i=Zoxc@ti0128a340-dhcp0438.bb.online.no)
18:52:02NJoinFOAD [0] (n=dok@dinah.blub.net)
18:52:02NJoinmaraz_ [0] (n=maraz@xob.kapsi.fi)
18:52:02NJoinTuplanolla [0] (n=jani@unaffiliated/tuplanolla)
18:52:48 Quit CaptainKewl ("( www.nnscript.de :: NoNameScript 4.02 :: www.XLhost.de )")
18:52:58stripwaxWillDS - are you trying to port the entire rockbox operating system to the Nintendo DS, or just the rockbox application?
18:54:14WillDSJust the app. It needs a good media player. It will run on a microSD card.
18:56:08 Quit mcuelenaere ()
18:56:24stripwaxDoes the NDS devkit have a port of SDL already?
18:57:05stripwaxOthers - what is the status of RockboxAsAnApplication?
18:57:35WillDSI am not sure about SDL. Is it required?
18:58:18kugelhttp://www.dcemu.co.uk/vbulletin/showthread.php?t=13306
18:58:44 Join fdinel [0] (n=Miranda@modemcable204.232-203-24.mc.videotron.ca)
18:59:24WillDSI can easilly get apps running on DS, my IRC app is on my DS.
19:00
19:00:05 Quit _lifeless (Read error: 104 (Connection reset by peer))
19:00:19BigBambiWillDS: The Rockbox simulator uses SDL
19:00:21krazykitWillDS, well, the simulator uses sdl, so having sdl would make as-an-app a whole lot easier
19:00:36 Join _lifeless [0] (n=lifeless@90.151.39.37)
19:01:12WillDSOK, I'll Check. Thanks.
19:01:22 Quit WillDS ("ClIRC - IRC client for Nintendo DS")
19:02:42 Join jaykay_ [0] (n=chatzill@p579E7481.dip.t-dialin.net)
19:05:28 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
19:08:33 Part linuxstb ("Leaving")
19:08:34toffe82BigBambi: I try to run the beastpatcher and the S is stuck on the screen firmware update restoration required ???
19:08:45 Quit Gnutoo (Read error: 104 (Connection reset by peer))
19:09:07 Join Gnutoo [0] (n=gnutoo@host76-59-dynamic.24-79-r.retail.telecomitalia.it)
19:10:34BigBambitoffe82: Odd - I restored the OF (to the original version, not the downloadable version), connected in MTP mode, ran beastpatcher, beast turned off, turned it on and it booted Rockbox with the new bootloader perfectly
19:12:28amiconnn1s: No it can't. Bridge configuration can only be done via USB
19:13:29toffe82I try this
19:14:23 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
19:14:25amiconngevaerts: Imo the active flag should *not* be fixed on disk, as it seems to trigger the reformat of the data partition if the beast decides to go into recovery mode
19:15:01amiconnI didn't fix my active flags, and had no reformat happen when I (once) encountered recovery mode
19:15:18gevaertsamiconn: hm, that's a good reason...
19:15:21 Join gartral [0] (n=Gartral@adsl-75-33-69-103.dsl.bcvloh.sbcglobal.net)
19:16:05 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
19:16:23amiconnIt's not for sure that it triggers this behaviour, but at least it seems to be likely
19:17:08 Quit jaykay (Read error: 110 (Connection timed out))
19:18:12 Join Horschti [0] (n=Horscht@xbmc/user/horscht)
19:18:31gevaertsAs long as we're not sure, I guess playing safe is the best thing to do
19:19:29 Join mirak [0] (n=mirak@81-66-70-98.rev.numericable.fr)
19:20:33*linuxstb goes to commit before anyone changes their mind
19:20:53toffe82BigBambi: I think I have a problem because I do it in a vmware machine
19:20:55bombardierAbout fuze build... Do I have this right? I use mkamsboot with original firmware and rockbox bootloader and then copy that file and .rockbox directory from 'normal' build to root of the device.
19:21:27BigBambitoffe82: wouldn't surprise me - USB is odd in those
19:22:21kugelbombardier: yes
19:22:30toffe82I think it is always connected
19:22:32 Join MethoS- [0] (n=lem@host-091-096-214-082.ewe-ip-backbone.de)
19:23:08bombardierkugel: should I format the device before?
19:26:43kugelbombardier: probably. if rockbox doesn't boot (stuck at "load firmware"), the .rockbox folder is behind the first 1 GB, which we cannot read yet
19:28:08bombardierkugel: and if it gets stuck? does usb connection still work or is device bricked?
19:30:51linuxstbbombardier: You can always start the original firmware, and hence get a usb connection.
19:31:59Nico_Plinuxstb: what do plan on doing for the hidden flag on the firmware partition?
19:33:04bombardierlinuxstb: thanks. I figured that was the case but just wanted to double check. :)
19:33:08linuxstbNico_P: You mean to mark the firmware partition as hidden?
19:33:24Nico_Pyes
19:33:38linuxstbNothing at all...
19:34:11Nico_Pcouldn't we patch it on the fly as well?
19:34:31Nico_Phaving both partitions show up when plugging the beast in is annoying most of the time
19:35:09linuxstbIt doesn't annoy me, but then I disable automounting...
19:35:58 Quit dfkt_dt (Read error: 110 (Connection timed out))
19:35:59 Quit kugel (Nick collision from services.)
19:36:04 Join kugel [0] (n=kugel@rockbox/developer/kugel)
19:36:14 Join dfkt_dt [0] (i=dfkt@unaffiliated/dfkt)
19:36:20kugelbombardier: then you simply boot the OF
19:36:33 Quit Horscht (Read error: 110 (Connection timed out))
19:36:34linuxstbNico_P: I don't mind if someone else wants to implement that, but it's not my "must be done before release" list.
19:36:45bombardierbootloader file is bootloader-fuze.sansa i presume. not the bootloader.bin
19:37:32Nico_Plinuxstb: I was suggesting it because AFAIK it's a minor change to what you just committed, and I think it would be appreciated by most
19:38:50linuxstbNico_P: I'm assuming we don't _always_ want that behaviour though. Making it conditional isn't a minor change - unless it was simply always done in Rockbox, and never in the bootloader.
19:39:12Nico_Pyes, that's true
19:40:58toffe82same problem running direct from linux
19:41:25toffe82it shut down back on and ask for a firmware restore
19:41:51 Quit balug ("Ex-Chat")
19:45:17toffe82any idea for the beast, I try with v1.1 and and v1.2 and get the same error
19:45:27toffe82the transfer looks to be ok
19:46:02 Quit miepchen^schlaf ()
19:46:04bombardierkugel: is the bootloader file called bootloader-fuze.sansa or bootloader.bin? I have both of them in build directory
19:46:40 Join miepchen^schlaf [0] (n=miepel@p579EC58B.dip.t-dialin.net)
19:47:13 Join z35 [0] (n=z35@h69.149.18.98.dynamic.ip.windstream.net)
19:47:43 Join Horscht [0] (n=Horscht@p4FD4EF7C.dip.t-dialin.net)
19:47:52linuxstbbombardier: You need to use the first one. mkamsboot should reject the wrong one though.
19:48:07kugelbombardier: the first
19:48:12bombardierok
19:48:37bombardierand what if i want to revert back to of? how do i accomplish that?
19:51:50kugeljust drop a unpatched OF on the drive
19:52:34bertrikthere's a magic key to hold to boot to the OF (probably the '|<<' key)
19:52:58bombardierkugel: that still works? I thought that when i replace bootloader there is no way back
19:53:20kugelyou don't replace the bootloader
19:53:31kugelwe just inject our bootloader into the OF
19:54:04 Quit Chesteta (Read error: 110 (Connection timed out))
19:55:15 Quit dfkt (Read error: 110 (Connection timed out))
19:55:20 Join Lear [0] (i=chatzill@rockbox/developer/lear)
19:56:32 Quit MethoS- (Read error: 110 (Connection timed out))
19:58:39 Quit toffe82 ("ChatZilla 0.9.84 [Firefox 3.0.6/2009011913]")
19:59:38bombardier'read failed (image)'
19:59:54bombardierany ideas?
20:00
20:01:14linuxstbwhat says that?
20:01:46bombardierrockbox boot
20:02:22 Quit Horschti (Read error: 110 (Connection timed out))
20:02:34bombardierit says Loading rockbox.sansa and in next line it reports this error
20:03:33linuxstbDid you unzip rockbox.zip onto your Fuze?
20:03:44bombardieri formated the device and copied firmware and .rockbox directory
20:04:03bombardierit has to go in root directory?
20:04:06 Quit Seed ("cu, Andre")
20:05:25bombardieri formated it using system menu on device. is that ok?
20:06:16linuxstbThat all sounds OK to me. Maybe kugel has some ideas.
20:06:56 Quit dfkt_dt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
20:07:10kugelbombardier: should be. I'd just try again, remember that we cannot read behind 1GB
20:08:21 Nick Beta2K___ is now known as Beta2K (i=1000@d36-124-26.home1.cgocable.net)
20:08:27 Join Seed [0] (n=ben@bzq-84-108-232-45.cablep.bezeqint.net)
20:09:18bombardierit works now
20:09:37bombardiermaybe i forgot to unmount it so it didn't write the cach
20:10:15 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/labb.contactor.se/x-ce9e602eff5ab748)
20:10:53saratogai saw someone ask about a DS port
20:10:59saratogaI looked into that
20:11:11saratogathe linux port already has all the drivers written under the GPL
20:11:27saratogabut doing rockbox as an app really makes no sense
20:11:52saratogasince we have the driver source code theres no reason to call SDL, just call the drivers directly
20:12:05linuxstbIsn't it the wrong GPL though?
20:12:24saratogain linux?
20:12:50 Join CaptainKewl [0] (i=jds@207-237-172-77.c3-0.nyr-ubr4.nyr.ny.cable.rcn.com)
20:13:22gevaertsMaybe v2 only, but we can't use them straight away anyway
20:13:30kadobanisn't there already GPL v2 only code in rockbox? (i assume that's what the drivers are if they're in linux)
20:14:09gevaertsthere shouldn't be
20:14:28saratogathe driver files I looked at say V2 or later
20:14:43saratogaso it looks like the dslinux people like us
20:15:16saratogaugh heres a V2 only one
20:15:24saratogaso some of them could be used at least
20:19:55BigBambiIs CREDITS-MANUAL in use?
20:20:10BigBambiIt seems slightly odd to have different credits to me
20:20:38saratogagevaerts: some of it might be usable directly, like the SD and CF driver command functions
20:21:43gevaertshm, true, but then those are trivial enough to reimplement without thinking
20:25:36 Join bombardi1r [0] (n=mcizmek@93-139-111-85.adsl.net.t-com.hr)
20:26:33*BigBambi adds himself anyway
20:32:37***Saving seen data "./dancer.seen"
20:33:43 Part gartral
20:35:27kadobanWas FS #9067 (plugin localization) abandoned for a good reason, or did it just stall because the developer got distracted? I'm wondering if it's worth trying to finish up, or if there were fundamental problems with it.
20:37:18BigBambikadoban: stall
20:37:31gevaertsSee last night
20:37:33gevaertss log
20:37:36BigBambikadoban: And midgey was in here earlier saying he was hoping to continue in the coming week
20:37:51BigBambikadoban: I'm sure he'd appreciate any help
20:38:11kadobanoh, hmm. i guess i'll take a look then
20:38:31kadobani could probably at least get it working with the new build system...
20:39:24 Quit bombardier (Read error: 110 (Connection timed out))
20:42:35 Quit J-23 (Read error: 110 (Connection timed out))
20:46:31 Quit bombardi1r ("leaving")
20:55:32 Quit saratoga ("CGI:IRC (EOF)")
21:00
21:01:45 Quit flydutch ("/* empty */")
21:04:21 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
21:04:55 Nick jaykay_ is now known as jaykay (n=chatzill@p579E7481.dip.t-dialin.net)
21:05:16 Quit robin0800 (Read error: 110 (Connection timed out))
21:11:03 Quit domonoky (Read error: 104 (Connection reset by peer))
21:13:50 Join MethoS- [0] (n=lem@host-091-097-241-086.ewe-ip-backbone.de)
21:14:52 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
21:15:34 Quit kugel (Nick collision from services.)
21:15:38 Join kugel [0] (n=kugel@rockbox/developer/kugel)
21:20:33 Join meyerman [0] (n=chatzill@76-191-132-123.dsl.dynamic.sonic.net)
21:21:58 Join MethoS-- [0] (n=lem@host-091-096-212-131.ewe-ip-backbone.de)
21:22:54 Join gartral [0] (n=Gartral@adsl-75-33-69-103.dsl.bcvloh.sbcglobal.net)
21:23:35 Quit Nico_P (Remote closed the connection)
21:24:48 Join Horschti [0] (n=Horscht@xbmc/user/horscht)
21:25:44Unhelpfulamiconn: i'm starting to think that LRU might be the wrong caching scheme entirely. perhaps instead of the draw thread queue a list of slides for the load thread to load, it ought only to say which slide it's on, and let the load thread try to cache as many slides as possible on either side...
21:26:39Unhelpfuland the cache would only flush covers to make room if they were closer to the center slide than an image in the cache
21:28:37 Join bombardier [0] (n=mcizmek@93-139-111-85.adsl.net.t-com.hr)
21:29:06Unhelpfula linked list could still be used, i supposed, with the highest priority item being "center" instead of head or tail, and the head or the tail being freed when loading new slides
21:30:34 Join kachna [0] (n=kachna@r4ax178.net.upc.cz)
21:31:45gartralwell were on AA, how does rockbox handle not-quite-squared art?
21:32:56Unhelpfulgartral: the same as other art, it scales to the largest possible size where source aspect ratio fits within the constraints of the display size. :P
21:33:22 Quit MethoS- (Read error: 113 (No route to host))
21:33:53gartralbut that would produce a distorted looking image, wouldnt it?
21:34:03 Quit mirak ("Ex-Chat")
21:34:11rashergartral: no, it preserves aspect ratio
21:34:27gartralohh, right, nvm, i can't read today
21:38:16 Join tessarakt [0] (n=jens@e180072242.adsl.alicedsl.de)
21:38:59Unhelpfulis there a better data structure for handling this "cache a centered view of part of a list" thing? it might not even need linked list at all, i'm thinking, perhaps just an array of cache slots, and the cache is indexed by album number % cache size?
21:42:46 Quit Horscht (Read error: 110 (Connection timed out))
21:43:45 Quit MethoS-- (Read error: 113 (No route to host))
21:45:34kugelUnhelpful: linked list, with a special pointer to the center slide, I'd say
21:46:16 Join JdGordon [0] (n=jonno@123-243-140-31.static.tpgi.com.au)
21:52:16 Quit bombardier ("leaving")
21:54:15n1slinuxstb, Nico_P: I suggested earlier that e could always hide the fw partition in rockbox usb and have a button press unhiding it in the bootloader usb, any thoughts on that?
22:00
22:04:56 Join flyinghippo [0] (n=flyinghi@pool-173-72-208-154.clppva.east.verizon.net)
22:07:14 Join fml [0] (n=4fd3e5ba@gateway/web/cgi-irc/labb.contactor.se/x-9dc4bba271bc1b03)
22:08:12 Join robin0800 [0] (n=quassel@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
22:09:43 Join robin0800_ [0] (n=quassel@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
22:09:56fmlHello. There were recently changes to the wiki page SansaE200Unbrick. The words "you might have to rename" are used. I think the instructions should be more direct, ie. "rename X to Y if ..." Because an unsure and scared user is made even more unsure by "might". Should I or should I not? I have not had such problems hence wouldn't like to correct the page myself.
22:09:59 Quit robin0800 (Read error: 104 (Connection reset by peer))
22:10:04linuxstbn1s: I was also thinking that the bootloader shouldn't even enter USB mode by default - only if Rockbox couldn't be loaded. So if you turned on a beast with USB attached, it would start Rockbox normally, unless it couldn't, or a button was pressed. This could mean that bootloader USB mode always exposed the firmware partition.
22:10:28gevaertslinuxstb: sounds like a good idea
22:10:44linuxstbgevaerts: jhMikeS didn't like it though...
22:10:59linuxstbI guess he frequently breaks Rockbox, so likes bootloader usb mode...
22:11:19linuxstbHe also pointed out that this isn't how other Rockbox targets with bootloader usb behave.
22:11:49gevaertslinuxstb: he also doesn't want charging in the bootloader. Other targets do that!
22:12:32linuxstbfml: That page has always been vague - IMO it should be explicit as possible.
22:13:19gevaertsI think bootloader behaviour is always going to be varied anyway
22:13:57fmllinuxstb: exactly my point. This page describes things to do in a critical situation and should be as unambiguous as possible.
22:14:21n1slinuxstb: either way sounds fine to me but i think that the "standard" usb mode should hide the fw partiton both for avoiding annoyance and slightly less risk of messing up the firmware
22:14:53linuxstbn1s: I'm just trying to keep things simple...
22:15:10linuxstbi.e. Bootloader USB = always show, Rockbox USB = always hide.
22:15:23linuxstbSo we just do it with #ifs
22:16:11n1syes, that is simple but I, for example, use bootloader usb quite frequently so always showing the fw partition in the bootloader would defeat the point if users do that too
22:16:29linuxstbWhy do you use bootloader usb?
22:16:58linuxstbI mean is it simply because you have usb plugged when turning on? (that's the behaviour I would also like to change)
22:17:11BigBambiI do quite often just because not infrequently I plug in then turn on
22:17:18n1sgrab the turned off player connect usb, drop some tracks on, disconnect and go
22:17:44linuxstbSo if the bootloader loaded Rockbox in that situtation (and hence Rockbox USB mode), it would be fine?
22:18:02n1syeah, i think so
22:18:09BigBambifor me, sure
22:22:55 Quit ChanServ (shutting down)
22:23:46amiconngartral: I hope you appreciate r20080 ;)
22:24:05 Join robin0800 [0] (n=quassel@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
22:24:47Unhelpfulkugel: i'll need to access the ends as quickly as the center, though. i think an array, with the index wrapping at either end, might be simpler... linked list definitely makes it easier to expire slides if the index has moved by a large amaount, though. :/
22:25:21 Quit robin0800_ (Read error: 104 (Connection reset by peer))
22:25:22amiconnlinuxstb: I also prefer bootloader usb mode, just for the fact that it's much fatser to enter usb this way from poweroff status
22:26:11amiconn(even if the dircache bug with usb-on-startup will hopefully fixed in main rockbox)
22:26:12BigBambiamiconn: Surely negligable compared to file transfer times
22:28:18 Quit JdGordon (Read error: 60 (Operation timed out))
22:29:06linuxstbWe also have the problem that the bootloader doesn't either charge or correctly deals with a usb charger (it freezes saying "waiting for usb")
22:29:13linuxstbs/deals/deal/
22:29:34kugelUnhelpful: why is fast access to the ends necessary? these aren't visible anyway
22:29:59 Join ChanServ [0] (ChanServ@services.)
22:29:59Mode"#rockbox +o ChanServ " by irc.freenode.net
22:29:59DBUGsent MODE #rockbox -o ChanServ
22:29:59***Server message 485: 'logbot ChanServ #rockbox :User is immune from kick/deop'
22:30:00gevaertsThe waiting for usb thing should be fixed I think. Either implement charging in the bootloader or timeout and go on
22:30:50 Quit fml ("CGI:IRC (EOF)")
22:31:05Unhelpfulkugel: fast access to the ends is necessary because they are the first ones to *free* when space is needed.
22:32:20linuxstbgevaerts: I'm just keen on releasing a bootloader asap (to try and get the Beast into 3.2). The behaviour I suggested seems easiest to implement, but I'm happy for someone else to do more. Or we just release as-is.
22:32:27kugelUnhelpful: then have special pointers to both ends too?
22:32:28 Quit robin0800 (Read error: 104 (Connection reset by peer))
22:32:40***Saving seen data "./dancer.seen"
22:35:36Unhelpfulit's circular at present, so one end is the other... adding a pointer to the middle would probably be sufficient, the real question is what to do when the index has moved so far that all of the cached slides are invalid - how to detect that, and how to handle it :/
22:36:23gevaertslinuxstb: I actually like what you suggested regardless of ease of implementation...
22:38:37 Join gromit` [0] (n=gromit@ALagny-154-1-69-174.w86-203.abo.wanadoo.fr)
22:42:26BigBambilinuxstb: Without beqst qttqched it sqys:
22:42:28BigBambi[INFO] Found device "<unknown> - <unknown>"
22:42:49BigBambi[INFO] Device version: "<unknown>"
22:42:59linuxstbBigBambi: Ah yes, that's a bit that isn't implemented yet... (the scanning)
22:43:03BigBambiand then press i etc
22:43:24BigBambiso would you like me to try with it attached~?
22:43:28 Quit stripwax (Read error: 104 (Connection reset by peer))
22:43:32linuxstbBigBambi: Yes please.
22:43:38 Join Thundercloud [0] (n=thunderc@cpc3-hem18-0-0-cust53.lutn.cable.ntl.com)
22:43:44BigBambiOK
22:44:45BigBambiHmmm, windows have given me a new hardware prompt
22:44:49 Quit LambdaCalculus37 ("Fwump")
22:44:56BigBambiDO I need a new WMP for this?
22:45:07linuxstbI don't know. Did you ever try the Windows sendfirm?
22:45:16BigBambinope
22:45:38BigBambiI seem to remember needing WMP 10 or 11
22:45:46linuxstbYes, I think so too.
22:45:52BigBambiAnd I never use it, so I guess it is at 9 still
22:46:01BigBambiI'll try updating it
22:46:01linuxstbBut I guess a Windows Beast user will have that...
22:46:08BigBambiyes
22:46:21linuxstbUnless they bought it for Rockbox...
22:46:34 Quit gregzx (Read error: 110 (Connection timed out))
22:47:12BigBambiJesus, IE wants me to do something with search providers
22:47:34gevaertsSounds naughty!
22:47:46amiconnJust select "keep current provider" (or how that's called in the english version)
22:47:57 Quit fdinel (Read error: 104 (Connection reset by peer))
22:48:03linuxstbI think this is going to need some work by a Windows dev (i.e. someone with VC installed) - so it can be built as a single .exe, without that separate dll.
22:48:20BigBambiI need IE for Windows Update to update WMP that I never use, yuet I never use IE so it isn't set up either :)
22:48:54amiconnYou could download the standalone wmp installer
22:49:11BigBambiam doing now
22:49:24amiconn(but if you do this using firefox or similar, you will have to execute a wga checker program separately)
22:53:55BigBambihmmm, Windows still wants a driver for it :/
22:54:08 Quit kugel (Nick collision from services.)
22:54:13 Join kugel [0] (n=kugel@rockbox/developer/kugel)
22:55:57 Quit rocko ("Leaving")
22:56:21 Join Kevets [0] (n=Kevets@adsl-76-252-75-214.dsl.lgtpmi.sbcglobal.net)
22:56:50BigBambilinuxstb: Well, I now have WMP11 but it still wants a driver. I tried beastpatcher.exe anyway, and it claimed success but nothing changed
22:57:08linuxstbCan you try sendfirm.exe ? I think there's a binary in the wiki...
22:57:38BigBambilinuxstb: The good news is that the first time I tried to put the OF back in it went into the restore screen and I used beastpatcher from linux, it worked poerfectly and all my data was still there too
22:57:42BigBambilinuxstb: yep, will do
23:00
23:00:19BigBambilinuxstb: sendfirm also claimed success but nothing happened
23:00:52BigBambiMaybe I need to reboot for the WMP update to 'work'
23:00:59 Join Davide-NYC [0] (n=Davide-N@user-12ld9li.cable.mindspring.com)
23:01:01linuxstbBigBambi: Hmm... I guess we need to ask mcuelenaere
23:01:09BigBambiBut I'm encoding videos at the moment so can't
23:02:06 Join Iiphii [0] (n=ep@dsl-jklbrasgw1-fe03fb00-99.dhcp.inet.fi)
23:02:33*BigBambi heads back to trusty linux to rid his beast of the OF :)
23:03:02BigBambilinuxstb: I'm sure this is all because the S isn't appearing in Windows
23:03:07BigBambidue to the driver issue
23:03:22BigBambiIf still needed I'll try to sort that out tomorrow eveing, then try again
23:03:50linuxstbOK, thanks.
23:05:01 Part Iiphii ("Lähdössä")
23:08:09 Quit Gnutoo (Remote closed the connection)
23:12:52 Quit stoffel ("leaving")
23:19:48 Join Chesteta [0] (n=Chesteta@dyn53-149.res-hall.ndsu.NoDak.edu)
23:21:10Chestetahey I could use some help compiling (again); its probably a simple fix however I don't know why I keep having the problem I am having... is anyone around?
23:23:06kadobani've never seen this room empty...it's probably best to just ask your question :)
23:23:09 Join robin0800 [0] (n=quassel@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
23:24:14Chestetait seems that everytime I install cygwin and then set things up for rockbox (svn co svn://svn.rockbox.org/rockbox/trunk rockbox); then make a build directory (I am building for the e200v2) and then run make
23:24:23Chestetait runs the first time just fine and then i can run make zip
23:24:41Chestetahowever after runing an update (svn update); things never seem to compile properly agian
23:25:21kadobanwell that seems quite odd. does it give you any kind of reasonable error?
23:25:23ChestetaI have even tried making a new build directory (after the svn update) and configuring that to build for my target
23:25:32Chestetayes; one sec ill give you the error
23:25:45 Quit tyfoo (Read error: 104 (Connection reset by peer))
23:26:02ChestetaAR libalac.a
23:26:03Chestetamake: *** [/home/Matt/rockbox/build/apps/codecs/libalac.a] Error 1
23:26:26Chestetathat is what it gives me in my original build directory (trying to build the newer version)
23:26:46kadobanChesteta: if you look up a bit there's probably something more descriptive
23:26:49gevaertsChesteta: I'm pretty sure that that isn't the first error
23:28:38Chestetai cant go up further; there is too much history (it erased some of the begining) and the rest of the listing that I can see either says CC, LD, or AR
23:28:58Chestetafollowed by a filename ofcourse
23:29:31Chestetawhen I try to run make agian in that directory it does give me more information...
23:30:13gevaertsRun "make >out.txt 2>&1", and then put the contents of out.txt on a pastebin
23:30:22Chestetaok
23:30:35LearMaybe make that "make V=1 ..."?
23:31:32 Quit jaykay (Read error: 110 (Connection timed out))
23:32:32ChestetaI am running it now; its just wierd that it compiles properly right after reinstalling everything and then after an svn update it errors out, I have not applied any patches or anything else like that either :/
23:33:15Chestetahttp://pastebin.ca/1344414
23:34:09rasherSurely that's not all of it
23:34:11Chestetai have noticed that it always seems to be a codec that causes the error; in my build2 directory it was libmad that errored
23:34:31rasherDo this instead: make V=1 >out.txt 2>&1
23:34:32Chestetawhen i just ran what he posted thats all that was in the file...
23:34:37Chestetaok
23:34:42rasherokay, run make clean first
23:34:48rasherand paste *all* of out.txt
23:35:01Chestetais that what make v=1 will do?
23:35:16gevaertsV, not v
23:35:19rasherCapital V. Turns on verbose output
23:35:20kadobanthat makes it verbose
23:35:23Chestetaok
23:36:09Chestetaso i should make another build directory and give you the output or just use verbose mode?
23:37:19rasherIn your build directory, type "make clean; make V=1 >out.txt 2>&1"
23:37:29Chestetaok :)
23:38:35 Quit {phoenix} (Remote closed the connection)
23:41:55 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
23:46:28 Quit Lear ("ChatZilla 0.9.84 [Firefox 3.1b3pre/20090217034531]")
23:52:03Chestetaok; just warning you all its 398k
23:52:50Chestetahttp://pastebin.ca/1344431
23:53:17 Join intrados [0] (n=intrados@cpe-71-67-133-85.woh.res.rr.com)
23:55:56 Join miepchen^schlaf [0] (n=miepel@p579EC58B.dip.t-dialin.net)

Previous day | Next day