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 2009-01-04

00:00:26gevaertsZagor: It's clean and unclean in different ways I think. I like this ability to look at SOURCES to see what gets built, and in that sense the new doom solition isn't very clean. It's all a set of tradeoffs of course
00:01:21Zagorgood point
00:03:20UnhelpfulZagor: the pluginlib bmp loader needs to call the scaler, which doesn't even exist in core on mono.
00:03:52ZagorUnhelpful: yes but why doesn't it exist? isn't that by your choice?
00:04:27ZagorI mean why do we compile the same code in different ways on mono only?
00:06:00UnhelpfulZagor: the scaler is only used in core by albumart, which is essentially not worth doing on mono without greylib - i tried, dithering it to mono is just not nice at the resolutions our mono targets have.
00:06:48amiconnAlso mono often means lowmem - and that's not only the archoses
00:07:18Zagorwhich plugins use it on mono?
00:08:46Unhelpfulnone at all, since it's not available at all on mono right now - which is why i'm trying to make it available in pluginlib
00:09:49Unhelpfuli think that using the core object "directly" should be fine for resize.o, and having a pluginlib bmp.o that depends on the core bmp.o will work as long as their deps are the same
00:10:50Zagordon't add .o -> .o deps. if you make a #include "file.c" solution, add the container.c to OTHER_SRC so it gets its' own deps.
00:11:41saratogait definately looks like at least some of the nk.bin file is still on the gigabeatS firmware partition after installing rockbox
00:12:19saratogamy partition dump has >100KB segments of the nk.bin still on the firmware partition
00:13:14Unhelpfuland i suppose .c -> .o deps are bad, as well... i don't really like including a C file, though. :/
00:14:01Unhelpfulalthough, the wrapper file would also solve the issue of the pluginlib having another bmp.c already
00:14:37saratogais there some good tool for comparing binary files?
00:15:25Zagorsaratoga: does the partition use a file system or is it dd dump-and-run?
00:15:37Unhelpfulit's TFAT
00:16:03UnhelpfulZagor: couldn't the #include solution just put the wrapper in SOURCES, and call it done?
00:16:42ZagorUnhelpful: yes it would. for your case, I'd say that is the best solution. everything else gets messy.
00:17:19Unhelpfuli think i have it pretty much solved... resize.o and albumart.o will just be built directly from the core files, as you did with doom. bmp.c gets a wrapper. seems least messy overall.
00:17:44Zagorsaratoga: in that case, isn't natural that pieces of old files remain on the partition?
00:19:08saratogaZagor: I think most of the file is there
00:19:11saratogain order
00:19:28saratogaat least the header is located in the dump file_length bytes away from the footer
00:19:38saratogathough i haven't yet figured out a good way to test if its all there
00:20:10Zagorthat too doesn't feel strange to me. first write new file, then flag old as deleted. especially for a transaction-safe system, that ought to be the norm.
00:20:31saratogayes this is exactly what I was hoping !
00:20:34n1sZagor: that's what we're hoping for
00:20:46Zagoraha, haha. I misunderstood.
00:20:50 Join MethoS-- [0] (
00:20:56saratogaif so we can resolve the dual booting issue on the gigabeat
00:21:12n1sand the uninstallation
00:21:39n1sprovided the user doesn't lose his backup :)
00:21:40saratogais there an easier way to dump the middle of a very large binary file other then writing c code to do it?
00:22:10Bagderor even hexdump
00:23:22amiconnn1s: That's something we need to document. If the user then doesn't keep it, it's his own fault. Same situation as with the original flash dump when flashing an archos
00:23:35saratogais hexdump less confusing then DD?
00:24:10Unhelpfulsaratoga: man split
00:24:11saratogaless confusing enough to figure out how to get it in cygwin?
00:24:36BagderI have no idea ;-) I figure it would be available there
00:25:17Unhelpfulor split −−bytes=<blocksize> -d <input>
00:26:39 Join Beta2K [0] (
00:27:52 Quit agaffney (Read error: 104 (Connection reset by peer))
00:28:07 Join agaffney [0] (n=agaffney@gentoo/developer/agaffney)
00:29:33 Join unstable [0] (i=unstable@tor/regular/sid)
00:29:40 Quit {phoenix} (Remote closed the connection)
00:29:53unstablebleh, this RefreshDatabase delay on e280 v2 is terrible, I've been waiting 5 minutes so far.
00:30:12unstableIs there a long delay like this on the e280 v1, with or without rockbox?
00:31:42ameyerthere is a longish delay. I'd say more like 30s with a filled 2gb card
00:31:46 Quit t0mas ("bedtime!")
00:32:01unstableIt's an 8gb card, with only 200 megs on there.
00:33:31ameyeralso, I *think* the of refreshes the internal memory too, but the rockbox bootloader does something to stop that
00:33:39ameyeron the v1 I mean
00:34:25ameyer5 minutes seems a bit long
00:34:56unstableI have nothing on the internal memory, i just have music on the external. This refreshdatabase sits there for several minutes, I haven't been patient enough to get past 6 minutes so far. heh
00:35:32 Quit MethoS- (Read error: 110 (Connection timed out))
00:37:09bertrikunstable, reading the sd card on my nintendo ds got a lot faster after reformatting with a larger sectors/cluster amount, maybe this helps on the sansa OF too
00:37:39unstablebertrik: How big? you have a mkfs.vfat command you used?
00:38:20bertrikno, I formatted it on windows
00:39:11 Join krazykit [0] (
00:40:27unstableThis chipset on the e280 v2, is the same as the fuze?
00:40:42Bagderunstable: yes
00:41:00unstableSo if rockbox dev team gets fuze to work with rockbox, this e280 v2 will work as well?
00:41:10 Quit Beta2K_ (Read error: 110 (Connection timed out))
00:41:11Bagderprobably, yes
00:41:17Bagderthey're not identical
00:41:45 Join domonoky1 [0] (
00:42:13unstableIs there a page somewhere saying what is done, and what needs to be done? Does rockbox run on it and do basics like play music?
00:42:39Bagderthere's a wiki page and a forum thread for that
00:42:54*Bagder runs off to bed
00:48:49***Saving seen data "./dancer.seen"
00:49:23saratogai just recovered an nk.bin from a gigabeatS without taking it apart
00:50:02saratogaended up just using fopen to do it :)
00:50:48saratogavery nice of toshiba to defracgment their fat firmware image before flashing it
00:51:44saratogait was as simple as searching for the start of the file, then copying file_length bytes to a new file
00:51:57 Quit flydutch ("/* empty */")
00:55:14LloreanSo we can probably have our own tool create a dual booting GigabeatS without needing a separate nk.bin file around?
00:55:25saratogaLlorean: correct
00:55:51saratogasendfirm could probably just dump the nk.bin during the install easily enough
00:56:44saratogasilly question: can I just fopen the sdb1 directly without needing to dd it?
00:58:57 Quit domonoky (Read error: 110 (Connection timed out))
01:00:13 Quit n1s ()
01:03:52Zagorsaratoga: at least open(), but fopen ought to work too.
01:04:08Zagorin linux, that is. cygwin might be different.
01:04:46saratogai'm switching over to linux now
01:06:54 Quit ender` (" The problem with political jokes is they get elected. -- Henry Cate, VII")
01:13:58 Join Setox[tm] [0] (
01:14:29Setox[tm]hey anyone awake and know about the iriver h320?
01:15:26Setox[tm]i got a small issue maybe someone will know what to do or know i just trash it finally
01:15:45Setox[tm]ok so i wanted to put new mp3s in my iriver h320 when i decided to format it since i was going to load it with all new songs i formated it though the computer other than doing it though the iriver it self.. i come back and i see it is stuck at the main screen saying VER 1.29 K
01:17:30gevaertsHave you tried resetting it?
01:18:26Setox[tm]yeah and still same thing
01:18:58Setox[tm]when it first happened i said ok reset but its like it just wont boot over to the os at all
01:19:13gevaertsHow did you format it?
01:19:42Setox[tm]i had it hooked up though windows and i went to the drive and told windows to format it
01:20:15Setox[tm]i been looking around that maybe there is something i can press while i turn it on to force into a mode or something
01:20:20 Join Strife89 [0] (n=michael@
01:20:22gevaertsOK. My guess is that you formatted it as NTFS instead of FAT32.
01:20:26Setox[tm]but i have been unable to find anything
01:21:03Setox[tm]i might have u know i did not even pay much attention
01:21:23Setox[tm]that makes sense on why it is stuck at loading...
01:22:17 Quit Zagor ("Leaving")
01:22:22Setox[tm]i guess my only options is get a external case to format it to fat32 or another hd
01:22:36*gevaerts doesn't know
01:23:52gevaertsThose options would certainly work, but maybe there is an easier way
01:24:30Setox[tm]yeah thats what i been trying to figure out
01:24:47Setox[tm]if they only built a usb safe mode or something into this
01:24:55gevaertsThe rockbox bootloader has a USB mode...
01:25:04saratogacould someone suggest why this tiny bit of simple c code doesn't work for me:
01:25:14saratogait thinks every block is a match to the header
01:25:23saratogamaybe I'm just not doing memcmp right?
01:25:23 Quit tessarakt (Read error: 113 (No route to host))
01:25:49Setox[tm]u know i was thinking about that but it would need to connect to it though usb and cause of the way i formated it won't even show up as a drive
01:27:04gevaertsWhy wouldn't it?
01:27:29gevaertssaratoga: memcmp returns 0 on match
01:27:59Setox[tm]dunno when i put the usb in it just goes to the loading... and wont show up as an external hd
01:28:27*gevaerts decides to leave this to the h300 experts
01:29:02Setox[tm]haha like who?
01:30:40Setox[tm]so now all day i been looking for new mp3 player thats cheap and i found the V-Touch vl-875 but i can't find out whats the limit that i can put as for size of the expandable memory
01:31:39LloreanSetox[tm]: Things like that don't really belong here. This channel is for questions relating to Rockbox and using Rockbox.
01:32:40Setox[tm]yeah i know it does not and all the places that have things about the h320 are dead though the rockbox project is still alive and there is a rockbox for the 320 and maybe someone has come accross this
01:32:44Setox[tm]that is why i am here
01:33:00LloreanSetox[tm]: This isn't a place for "I didn't know where to ask" questions.
01:33:38Setox[tm]ok how about i had the rockbox in there
01:33:41LloreanIf you're having problems with the original firmware installed, you should google for communities around that player, or contact the makers of the player.
01:33:42Setox[tm]does that help ya?
01:34:01gevaertsSetox[tm]: if you had rockbox installed, you wouldn't have this problem...
01:34:09Setox[tm]i did look around and did not find much help at all and this is a discontinued item
01:34:43 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
01:34:56LloreanSetox[tm]: Just because it's discontinued doesn't mean you should come here and ask about it. If you're going to bother someone who doesn't support it, at least bother iRiver because they actually *made* the software it's running and know how it works.
01:35:43 Join ChadFM [0] (
01:35:58 Quit Llorean (Read error: 104 (Connection reset by peer))
01:36:20 Join Llorean [0] (
01:39:50saratogathis is driving me nuts
01:40:08saratogaI read a char * array from the disk
01:40:28saratogahow should I compare that to the hex string I've taken from the top of the nk.bin file?
01:41:21 Join fenugrec [0] (
01:42:29saratogamaybe i'm thinking about this wrong
01:43:01gevaertssaratoga: why don't you initialise the header from hex values instead of as a string?
01:44:28 Quit Setox[tm] ()
01:46:49gevaertssaratoga: something like
01:47:59 Quit MethoS-- (Read error: 104 (Connection reset by peer))
01:48:41 Join MethoS- [0] (
01:49:02 Quit domonoky1 (Read error: 104 (Connection reset by peer))
01:50:34 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
01:52:15saratogagevaerts: that did it, thanks
01:52:33saratogai was just too lazy to convert all those symbols
01:52:55saratogathough i guess it would have been a pretty simple regex
01:53:10*gevaerts pretends to have converted them by hand
01:57:37 Quit tvelocity (Remote closed the connection)
01:58:44*Strife89 wonders how the v2 ports are coming. :)
01:59:31_Auron_slowly but surely from what I can tell
02:01:28 Quit tyfoo ("Carpe diem")
02:01:52saratogaanyone with a gigabeat S around?
02:02:03Strife89_Auron_: Good to hear. :)
02:04:35 Join Thundercloud [0] (
02:05:39 Quit gregorovius ()
02:06:03 Join Blackfrosch [0] (
02:07:53 Quit Blackfrosch ("Leaving.")
02:10:19gevaertssaratoga: I have one, but I'm going to sleep now. I can test things tomorrow though
02:12:48 Quit Strife89 ("Shower time. :D")
02:22:14 Quit bertrik ("Leaving")
02:22:21saratogagevaerts: ok i'm going to commit what i have for the night
02:22:38 Join lee321987 [0] (
02:23:05lee321987Say I put a 1000 song playlist on my DAP in a format that RB decodes at 250% realtime. I fully charge my player, start the playlist, turn off all lights/displays, set the DAP down and let it play till it's drained. Then I do the same thing only with a codec that RB decodes at 500% realtime. How much difference might there be in play time?
02:23:52JdGordonif its that high neither should do any boosting
02:24:18LloreanJdGordon: Depends on the player. I'd imagine 250% realtime still means some boosting on coldfire.
02:25:22JdGordonah yeah...
02:26:06lee321987How much difference could a codec that needs boosting vs. one that doesn't make?
02:26:32JdGordonif its always boosting vs never.. then could be almost half the runtime...
02:26:38Lloreanlee321987: In a practical sense it's impossible to predict, since % realtime doesn't map directly to % time spent boosted, and VBR can change this (since "never boosting" at a CBR doesn't mean a ABR of that rate means it'll never boost) and your WPS etc can also affect things.
02:26:51 Join kiddecks [0] (
02:28:15lee321987System > Debug > CPU Frequency −−- is where I find out if I'm boosting right?
02:28:25 Join OverThere [0] (n=d0f80ecf@gateway/web/cgi-irc/
02:28:34 Join HellDragon [0] (
02:28:46JdGordonyeah, but the info in the audio debug screen is more useful if you really care that much
02:29:07JdGordonin that screen it will esitmate how much your boosting while decoding
02:29:31OverThereThe WPS for the dark geek nano theme won't load. Rockbox will, however, load any WPS besides that one. Does anyone know what the problem might be?
02:29:42lee321987"View Buffering Thread" right?
02:30:01JdGordonOverThere: most likely its using old tags, so it needs to be updated
02:30:03saratogawhy does SVN want to say a text file I created is executable
02:30:04JdGordonlee321987: yes
02:30:07LloreanOverThere: It's probably outdated and needs to have an old tag fixed. The CustomWPS wiki screen contains current syntax.
02:30:08saratogaand how do i reverse this
02:31:16OverThereJdGordon: okay, thanks
02:34:39 Nick fxb is now known as fxb__ (
02:35:31 Join MysticKitsune1 [0] (
02:36:29lee321987my screen cuts off part of the "boost ratio" −− it reads: "0.0% (3" −− what might the rest of that line read? And if it stayed at 100% that would cut my run time in about half right?
02:37:06saratogadepends on the target
02:37:24MysticKitsune1this is an intriguing bug: if your playing any file from the external card on an e250, then stop partway through, shut the dap off, then turn it on again and go to resume playback, or hit play, the file wont resume, but if its in internal memory, it plays
02:37:48saratogayou'll get about 65% of the runtime for mp3 in that case
02:38:10JdGordonMysticKitsune1: silly question... but is the card still in the dap?
02:38:52MysticKitsune1BTW its Gartral, just on an Alt
02:38:54kadobansaratoga: svn propdel svn:executable <filename> should turn off executable i think
02:39:42JdGordonworks fine here...
02:40:16lee321987thanks for info guys. I wish you all a good year.
02:40:16saratogaalright thanks
02:40:22 Quit lee321987 ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
02:40:23 Quit saratoga ("CGI:IRC (EOF)")
02:41:12MysticKitsune1hmm, ohh, let me update, cause it doesnt work as of 19663
02:41:27MysticKitsune1yepp, works great now
02:41:50 Quit moos ("Rockbox rules the DAP world")
02:43:36MysticKitsune1i wonder what caused it in the first place, cause it dosnt look like you changed any playback or resume functionallity
02:44:51JdGordonglitch in the matrix or planets out of alignement :p
02:45:30MysticKitsune1JdGordon: your always complaining about ME being OT, now its my turn!
02:45:56LloreanHe was answering your question. Basically, what caused it was likely some random fluke problem.
02:46:43MysticKitsune1not likly, cause i managed to reproduce it 4 times before i updated
02:47:15LloreanYou reproduced it 4 times on the same flaky build, though.
02:47:54MysticKitsune1nope, accross 19661-19663
02:48:14MysticKitsune1but 19664 works, so the points moot now
02:48:27LloreanWell, if you think it wasn't random flakyness you should track down when it started.
02:48:46LloreanIf it's a real bug that's just covered up at the moment, it should be documented by someone who can reproduce it.
02:48:50***Saving seen data "./dancer.seen"
02:49:14MysticKitsune1too do that ill need to dig through the archives
02:49:29MysticKitsune1if you even maintain a complete archive
02:49:42LloreanSVN is a complete archive.
02:50:18MysticKitsune1cant compile
02:51:03LloreanIf you can install software on the computer you're using, you can.
02:51:32MysticKitsune1if you would be so kind as to compile... say the last 10 builds, ill try it, or if you can provide me a compiling enviroment that works on a PSP...
02:51:58MysticKitsune1or morespecifically, the PSPs hardware
02:54:35MysticKitsune1and my main system isnt thermally stable enough to mingwm with gcc, i tryed, and i cant compile a third of anything without it shutting down from overheating, or freezing up due too the same reason
02:55:46*scorche|sh hands MysticKitsune1 a fan
02:56:21MysticKitsune1i have an internal firon coolant pump and bridges, 2 system fans, and a cooling plate, and it STILL overheats
02:56:29LloreanMysticKitsune1: Might I recommend you actually clock it as a stable rate then?
02:56:43MysticKitsune1its not the proc, its the MOBO
02:57:00 Join Willwolfe [0] (
02:57:04kadobanMysticKitsune1: sounds like either your heat sink isn't seated corrected/wrong amount of thermal paste/or you're overclocking too hard
02:57:24MysticKitsune1no overclocking, in fact, i had too underclock it
02:59:12LloreanStill, why'd you try mingw anyway? That's not one of the recommended methods in the first place.
02:59:27MysticKitsune1and i know what the problem is, its too many high powered componts crammed too close together, i was worried when i built it, and it has serious stability issues at "native" clocks, and the proc has a fan sink, with nearly a whole tube of thermal paste, way more than it needed for the size
03:00:01MysticKitsune1cause cigwin bluescreens it
03:00:11 Join itcheg [0] (i=62db4767@gateway/web/ajax/
03:00:21MysticKitsune1and VMware runs ungodly slow
03:00:47MysticKitsune1it took me 4 days to compile about 12% under vmware
03:01:23Unhelpfulsounds like you made some bad choices... and vmware shouldn't be that slow even before x86 supported VM properly... but maybe you should try colinux?
03:01:30scorche|sh...something isnt right there...cygwin should be loads slower than vmware
03:01:51scorche|shassuming you have any modern sense hardware
03:02:43 Quit OverThere ("CGI:IRC (EOF)")
03:02:44MysticKitsune1it was built to game on, and it works, mostly, for that, i have no idea why its so slow with any hybrid systems, but all i know is, it wont compile anything
03:02:48kadobanMysticKitsune1: "way more thermal paste than is needed" is NOT a good thing
03:02:55MysticKitsune1may i post the specs?
03:03:00kadobanbelow or above the recommended amount is bad
03:03:17scorche|shwe should probably move this to #rockbox-community at this point..
03:03:24LloreanThis is wandering far from topic, since these are really problems above and beyond the usual compiling issues
03:06:20 Quit Thundercloud (Remote closed the connection)
03:08:41 Join OverThere [0] (n=d0f80ecf@gateway/web/cgi-irc/
03:09:33OverThereHi again. Can the %bc tag for WPS screens be used as a conditional like %?bc<true|false>
03:20:04Unhelpfulit looks more and more like the wrapper C files are the only solution that allows conditional compilation and preserves the dependency chain
03:25:20 Quit OverThere ("CGI:IRC (EOF)")
03:30:09 Join OverThere [0] (n=mobile@
03:31:36OverTherethe %bt tag is showing 0h0min on a full charge. i'm using a nano. pleas help?
03:40:44 Quit MysticKitsune1 ("Leaving.")
03:41:46 Quit ChadFM (" HydraIRC -> <- *I* use it, so it must be good!")
03:43:29 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:52:03 Quit DerDome (Nick collision from services.)
03:52:06 Join DerDome1 [0] (
03:52:18 Nick DerDome1 is now known as DerDome (
04:06:18 Join Darksair [0] (n=user@
04:08:10 Join blkhawk- [0] (
04:08:59 Quit kiddecks ("kiddecks has no reason")
04:17:09 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
04:20:46 Quit fdinel (Read error: 104 (Connection reset by peer))
04:24:36 Quit blkhawk (Read error: 110 (Connection timed out))
04:25:07 Nick blkhawk- is now known as blkhawk (
04:26:17 Quit splinter (Read error: 110 (Connection timed out))
04:32:25 Quit OverThere (Read error: 110 (Connection timed out))
04:32:56 Quit fenugrec ("Leaving")
04:40:14 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:43:28 Join nevr [0] (
04:48:51***Saving seen data "./dancer.seen"
04:51:03 Join Barahir_ [0] (
05:02:32 Quit jhulst (Read error: 113 (No route to host))
05:02:33 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
05:08:22 Quit Barahir (Read error: 113 (No route to host))
05:20:21nevrWhat is the first partition for on a 1st gen iPod nano?
05:28:38LloreanThe firmware, if it's windows formatted.
05:31:38nevrIf that was trashed, would the device be recoverable?
05:31:54 Quit Nimdae (Read error: 104 (Connection reset by peer))
05:34:51nevrExcellent, thanks for the info
05:34:54LloreanDisk Mode is stored in a ROM so you still have access through that.
05:38:07nevrCould that be done away with, made in to one partition, and have rockbox installed in that? (I'm just trying to get a feel of how the iPod hardware is set up)
05:41:52 Quit nplus (Remote closed the connection)
05:47:13 Quit jhulst_ (Remote closed the connection)
05:51:15nevrI guess I could try it...nevermind the question...
05:52:17 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
05:57:47 Quit timc`` ("Leaving")
06:00:49 Quit jhulst (Remote closed the connection)
06:01:43JdGordonanyone want to talk about possible theming options?
06:02:22JdGordonHow does a WPS tag to say "put the menu in this viewport" and the rest of the screen will keep getting updated as the WPS regardless of which screen you are on?
06:02:58JdGordonthe other option is "this vieport is the statusbar" tag and that is the only viewport updated automatically
06:07:06 Join schNibbler [0] (
06:11:38 Join pixelma_ [0] (n=pixelma@rockbox/staff/pixelma)
06:11:57 Quit amiconn (Nick collision from services.)
06:12:00 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
06:12:59 Quit pixelma (Read error: 110 (Connection timed out))
06:20:07LloreanJdGordon: I like a "draw the menu in this viewport" tag.
06:20:14Lloreani've been in favor of it since viewports were first discussed.
06:21:32 Quit itcheg (" ajax IRC Client")
06:21:49 Join itcheg [0] (i=62db4767@gateway/web/ajax/
06:23:27 Quit Nibbl (Read error: 110 (Connection timed out))
06:25:08Unhelpfulgiving up on this for now... scaling for mono targets is working in sim, but fails to build for target because somehow my objects have refs to memset and memcpy... i don't even use the latter, and checking the preprocessed source shows that i'm only accessing the former through the plugin API
06:25:20Unhelpfulany ideas? maybe gcc is doing something clever?
06:27:14 Quit itcheg (" ajax IRC Client")
06:32:26*JdGordon has no idea how to do this :p
06:32:32JdGordonespecially with remote lcds...
06:37:53 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
06:38:24saratogaUnhelpful: doesn't gcc need memcpy for all files?
06:46:21 Part aurix_lexico
06:48:56***Saving seen data "./dancer.seen"
06:49:36 Join fenugrec [0] (
06:49:38 Join Aurix_Lexico [0] (
07:06:06 Quit Willwolfe ("ChatZilla 0.9.84 [Firefox 3.0.1/2008070208]")
07:07:43 Join itcheg [0] (i=62db4767@gateway/web/ajax/
07:09:31 Quit saratoga ("CGI:IRC (Ping timeout)")
07:12:34 Quit fenugrec ("Leaving")
07:15:18 Quit itcheg (" ajax IRC Client")
07:27:32 Quit MethoS- (Remote closed the connection)
07:41:35JdGordonhmm... now I just need someone who can be bthered fiddling with the .wps :p
07:45:58 Quit Aurix_Lexico ("Leaving.")
07:52:58 Quit Llorean (Read error: 54 (Connection reset by peer))
08:00:04 Quit BHSPitLappy (Remote closed the connection)
08:04:17 Join JWest [0] (n=457dac63@gateway/web/cgi-irc/
08:04:32JWestHello All..
08:04:46JWestNewbie for Rockbox!
08:04:56lucentwhat do you want
08:05:07JWestJust D/L and installed latest/stable on my Cowon iAudio 5XL.
08:05:27lucenthm, I'm not familiar with that target. is it working?
08:05:29JWestI have done the download, but dont have the charger... so I can't complete the last step.
08:05:50JWestInstall went fine. Did all the parts... the bootloader, fonts, etc...
08:06:08JWestbut the last step is to restart on the charger.. which I dont have with me.
08:06:23JWestSo.. it boots up under the Cowon sourced firmware instead.
08:06:48JWestThe last step is to shutdown, and restart on the charger... i guess thats a trigger somehow?
08:07:12lucentyou could try rebooting while it's plugged into a computer that is loaded only in the BIOS setup screen? maybe that acts like a charger
08:07:14JWestAnyhoo.. how can I restart w/o a charger? It seems happy to charge off USB... FWIW...
08:08:45ameyersome computers provide USB power while turned off.
08:12:18 Quit JWest ("CGI:IRC (Ping timeout)")
08:13:30 Join JWest [0] (n=457dac63@gateway/web/cgi-irc/
08:13:52JWestSorry there.. my web client locked up..
08:15:52JWestIf I Shutdown...any restart will take it into Cowon FW.
08:17:41JWestAny hoo... Has anyone a clue how to invoke the Boot loader when the 5XL is seen under /media/IAUDIO?
08:17:57JWestFedora 10, Kernel 2.6.27.something.
08:18:51JWestAnyone on line tonight?
08:21:14kadobanwhat do you mean by "invoke the bootloader" exactly?
08:22:12JWestWell... I have loaded the software on the device, I see the .firmware in /media/IAUDIO... and hte other installables went fine (as far as I know!).
08:22:41JWestJust now sure how to invoke the Rockbox bootloader ...
08:23:22JWestWhen its connected to USB.. it (Cowon FW) says (Connected)... and it I turn it off, it says (CHarging).
08:23:59JWestThe instructions say to use the Power Supply and restart, but I don't have that with me now.
08:24:03 Quit SUSaiyan (Read error: 54 (Connection reset by peer))
08:24:36lucentJWest: when you get the power supply, could you please test and then let us know if there is a need to use the power supply?
08:24:44JWestSo when I turn it on and off, it doesn't boot from Rockbox, it boots from Cowon.
08:25:17JWestOK.. was hoping to find a way to boot to Rockbox :^)
08:25:33JWestI guess Power Supply is the trick here?
08:26:12lucentI own a Sansa Fuze, I don't know about your target
08:26:36JWestAhh... hoping to get on Rockbox. Looks neat.
08:26:49JWestConsensus is to follow directions... (DUH :^)
08:27:03 Join bluefoxx [0] (
08:29:08JWestThanks. Good night.
08:30:09 Quit JWest ("CGI:IRC (EOF)")
08:30:20 Join SUSaiyan [0] (
08:30:25bluefoxxis there a tool out there for converting a wps file to the new standard?
08:40:58 Join Llorean [0] (
08:43:03 Part toffe82
08:47:27 Join trippedoutspider [0] (n=4a2cbc8a@gateway/web/cgi-irc/
08:47:36trippedoutspiderhi all
08:48:58***Saving seen data "./dancer.seen"
08:50:31trippedoutspiderany1 here able to help me install rockbox on a sansa c250?
08:53:30kadobanthe installation instructions are on the website. you can always ask questions if you don't understand something or run into trouble
08:53:41trippedoutspiderok well i ran into trouble lol
08:54:03kadobanhehe, what happened? i don't have that player, but i'm sure someone will be able to help
08:54:35trippedoutspideri had to switch from auto detect to msc so it would give me a letter drive to enter into the utility
08:54:58trippedoutspiderbut when i go to install i says disc access denied
08:55:24trippedoutspiderhowever there is no security tab under properties for (L:)
08:57:35ameyerI think you may need admin/root/(the equivalent for your OS of choice) rights, depending on your setup.
08:57:37trippedoutspiderso i tried the manual install, it only had one directory on it "tmp" should i extract into the directory?
08:58:16kadobantrippedoutspider: you generally extract into the root of the drive (so into L:\ in your case)
08:58:48kadobanso not in tmp, but you extract to the directory above that
08:59:05ameyerhrm, any chance your c250 is a v2?
08:59:27trippedoutspiderno it 100004a or
09:00:21*ameyer wonders what sort of error message you'd get if you tried to use rbutil to install on a v2 sansa
09:00:49*JdGordon thinks the wps is going to need some work to get wps-"statusbar" working nicely
09:01:24kadobantrippedoutspider: good sign
09:02:22trippedoutspidermy user acount is admin, only account on pc, but im running vista
09:02:54Lloreanameyer: I think we detect v2 Sansas and report to the user that they're not compatible.
09:03:26kadobantrippedoutspider: does it work when you do the manual install? (in other words, does it allow you to copy over the files?)
09:03:44Lloreankadoban: A manual install isn't simply copying over files.
09:04:05kadobanLlorean: i know, i meant the first step that he was trying
09:04:16Lloreantrippedoutspider: When you run RBUtil does it pop up the little UAC request for admin rights (or have you run it by right clicking and choosing to run as administrator)?
09:05:13 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:05:23trippedoutspideryes it lets me extract to drive L;
09:06:20trippedoutspiderat this point my sansa "drive L:" has two folders .rockbox and tmp
09:07:40 Join timc [0] (n=aoeu@
09:11:01trippedoutspideri think i got it workin now
09:51:34 Join flydutch [0] (
09:54:48 Quit XavierGr ()
09:55:28 Join Agneees [0] (
09:58:49 Quit trippedoutspider ("CGI:IRC (EOF)")
10:00:39 Join daurnimator [0] (n=quae@unaffiliated/daurnimator)
10:02:09 Quit bluefoxx (Read error: 104 (Connection reset by peer))
10:12:58 Join bertrik [0] (
10:16:34 Quit ameyer (Read error: 60 (Operation timed out))
10:24:24 Part Agneees
10:44:30 Join tyfoo [0] (
10:46:59 Join Blackfrosch [0] (
10:49:00***Saving seen data "./dancer.seen"
10:52:16 Join obo [0] (
10:57:49 Join fredddy [0] (
11:04:47 Join ajb` [0] (
11:07:13 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
11:09:38 Join miepchen^schlaf [0] (
11:12:28 Join tyfoo2 [0] (
11:15:11 Quit tyfoo (Read error: 60 (Operation timed out))
11:17:14 Join ibseo [0] (
11:22:01 Quit ajb (Read error: 110 (Connection timed out))
11:24:52 Join {phoenix} [0] (
11:52:25 Join nplus [0] (n=nplus@141.25.Globcom.Net)
12:00:45 Nick Barahir_ is now known as Barahir (
12:06:43 Quit nplus (Remote closed the connection)
12:26:26 Join esthar [0] (
12:26:31 Join n1s [0] (n=nils@rockbox/developer/n1s)
12:27:23 Join culture [0] (
12:28:39 Join gartral [0] (
12:29:00gartral im not sure if this is known, but the e200's rtc is updated by the OFW when its connected via MSC
12:29:56bertrikI'll try it out, but I don't know of any mechanism that could be responsible for synchronising the RTC
12:30:21bertrikI can imagine that MTP offers some kind of mechanism for that, but not MSC
12:30:26Lloreanbertrik: iPods are updated from iTunes through a custom SCSI inquire, if I recall. Maybe if the sandisk software is running?
12:31:03gartralright, thats what i mean
12:31:06 Quit gevaerts (Nick collision from services.)
12:31:17 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:32:08Lloreangartral: You mean MTP, or you mean "when the sandisk software is running"?
12:32:37gartralwell, its done it in MSC so far, havent tryed MTP yet
12:32:38bertrikhmm interesting, but I won't believe it until I see it :)
12:36:29gartralok, does NOT do this in MTP mode
12:37:00gartralat least, mine does
12:38:56bertrikgartral, MSC didn't update my RTC, but I don't have any sandisk software installed. What sandisk software do you have installed?
12:45:21gartralnone, just connect via USB >.>
12:45:35 Nick tyfoo2 is now known as tyfoo ( would that work? it must have been a coincidence
12:46:27gartralOFW is 01,02.18A
12:46:35 Join Riku [0] (
12:49:03***Saving seen data "./dancer.seen"
12:49:54JdGordonZagor: if you come online tonight... is it still broken? can you give me more info? which wps your using etc...
12:49:57 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
12:54:01bertrikgartral, if it's being done with a SCSI command, it should be easy to capture it with a USB sniffer
12:58:53 Join moos [0] (i=Mustapha@rockbox/staff/moos)
13:03:52 Join Thundercloud [0] (
13:04:10oborasher: do you have an example track that triggers FS #9747?
13:05:30 Quit crwl ("update")
13:05:49 Quit DerDome ("Leaving.")
13:09:28 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:15:27pixelmaobo: does the scrobbler log get its data from the database? Because I think that the database has some weird tag guessing...
13:17:00obopixelma: it gets it directly from the ID3 struct... which I'd guess is where the database also gets its info from
13:23:26gartralwould it be at all possible to kill the scroll wheels backlight on the e200s at rolo?
13:24:33gartralits just... uncomfortably bright for me
13:24:51_Auron_very blue
13:25:16gartralvery very bright, iv compared it too my moms, mines alot brighter
13:26:29Lloreangartral: What do you mean by "at rolo"?
13:26:38gartralbut hers has "plasticy" scroll wheel, mines slightly newer, and has the smooth spinning wheel
13:26:44gartralat bootloader
13:26:45LloreanRoLo is the part of Rockbox that executes firmware images from within Rockbox.
13:26:46 Join AndyIL [0] (i=AndyI@
13:26:51LloreanThe bootloader is _not_ RoLo
13:27:04gartraloh, i was undr the impression it was bootloader
13:28:00LloreanPeople often get this impression. Did it come from some specific document?
13:28:46gartralnah, me masr the connection between LiLo=LInuxLoader RoLo=ROckLoader
13:30:05LloreanTry to stick to the terms we use in the manual for things, it'll make you much easier to understand.
13:30:14gartralbut anyway, is there a way to accomplish the effect im looking for, should i make a formal feature request?
13:30:31LloreanThere are no longer "feature requests" anyway.
13:30:48gartralthe option still exists in FS
13:31:34LloreanYou sure you can actually create one?
13:32:00gartralohh, nvm, wow, shows how much attention i pay when i made bug requests, lol
13:32:20*gartral thwacks himself with his DAP
13:32:23 Part Blackfrosch
13:35:32gartralanyway, is it possible to disable scroll wheel backlight in the bootloader?
13:36:15 Join TMM [0] (
13:38:39 Quit AndyI (Read error: 110 (Connection timed out))
13:39:58Unhelpfulsaratoga: "need" in what way? i wasn't getting unresolved symbols at link time for memcpy and memset before moving these files into pluginlib... i still can't figure out why the objects expect to find either :/
13:47:23 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
13:48:53kadobanUnhelpful: i'm not sure i understand your problem (and probably wouldn't even if you explained it), but it couldn't be fixed by the weird MEM_FUNCTION_WRAPPERS in apps/plugin.h, right?
13:50:08Unhelpfulthat it might... hrm.
13:51:32gartraldoes anyone know if the bootloader can even interact with the scrollwheels backlight?
13:51:52Unhelpfulthe comment seems to suggest that i have the problem those are trying to solve
13:52:17 Join GodEater [0] (n=ge@rockbox/staff/GodEater)
13:52:51 Quit ibseo (No route to host)
13:53:27Lloreangartral: The bootloader is just a stripped down version of Rockbox, in a way. It can be made to do anything we know how to make Rockbox do, if it becomes necessary to do so. We usually try to have it do a minimal amount
13:54:48 Quit DataGhost (Nick collision from services.)
13:54:56 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
13:57:19gartrali know, but wouldnt it be fairly simple to make it not power the scroll-wheel backlight?
13:57:43LloreanIf you knew all that, why'd you ask if it was even possible?
13:57:51gartralim willing to do it, if i could ask somone to build it
13:57:53LloreanPlease, ask the question you mean to ask. I can't guess what you're really asking.
13:58:28gartralok, point, how would i do that?
13:59:48LloreanFind the code that does what you want and figure out where you want to put it? I'm not sure what you're asking.
14:00:15gartrali would assume the code is in button-e200.c, assume me a coding noob...
14:00:30gartralthis is my first procect in real C
14:02:46 Quit jhMikeS (Nick collision from services.)
14:02:52 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:03:03LloreanI'm not really gonna walk you through this. Investigate and spend some time reading through the code and such.
14:03:54gartralwell i am, and it didnt expect you to walk me through, just point me in the direction
14:05:32gartraland thank you for being so patient with me, i relise im tough too deal with...
14:06:50kadobangrep -R BUTTON | grep LIGHT comes up with some interesting things. i'd start there if you don't know what to do
14:07:00kadobangrep -r sorry
14:07:10gartralno grep, winblows
14:07:23gartralmanual looking >.>
14:09:23t0maswho's in charge of the default settings at the moment btw? :-)
14:09:28 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
14:10:00t0masI found one that imho needs changing..
14:10:01Lloreant0mas: Nobody specific. We just discuss them (best on the list where it can be discussed non-realtime)
14:10:27*t0mas unsubscribed from the list a year ago.. so I'll hop on again
14:10:43LloreanWhich setting?
14:10:59t0mason ipods the "load database in memory" setting should be on by default
14:11:23LloreanThis can use a considerable amount of RAM
14:11:39t0mastrue, but that is something an experienced user can decide for themselves..
14:11:45t0masnormal users are probably not going to change it (it sounds intimidating)
14:12:11kadobangartral: windows should be able to search in files, at least it could in XP last time i used it. but at the risk of hand holding a bit too much, i'd look in export/config-e200.h and see if it's as easy as not defining something (no idea if it'll work, haven't tried)
14:12:14Lloreant0mas: Maybe the description in the manual should be improved then?
14:12:18t0masand when you're playing a song, press select to go to the database.. and press back to get to another folder (common use path? selecting a new song?) you have to wait quite long when it is disabled
14:12:29gartralit also dosent work too well on the e200s, all i found it did was slowed things down
14:12:35t0masor... change the default to on on small players like the ipod nano?
14:13:15t0masbecause waiting so long for the next screen to appear makes normal users press the button twice.. which skips back to far after an even longer delay?
14:13:30t0mas(at least that's what my girlfriend did)
14:13:52LloreanMaybe there should be a splash to let them know it's doing something, then.
14:14:05t0masthat would be oke too
14:14:24t0maslike.. the splash it shows when you select a song..
14:14:31t0mas"searching..." or something like it?
14:14:32gartralkadoban: if i make the appropriate changes, would you be so kind as to build it?
14:15:31 Join tvelocity [0] (
14:15:35gartralBGrep said that file doesnt exist
14:15:36kadobangartral: sorry, i can't really. i don't have a clean copy of the svn at all, and i have few ways to share the results, and i don't have my e200 with me
14:16:08gartralkadoban: email? lol
14:17:38kadobanmeh. you really just need to find a way to build :) you end up doing it a million times to figure out anything non-trivial
14:18:54Unhelpfulseriously, it's already very easy to spend more time building that coding *without* having to send it off to somebody else to do
14:18:59gartralwell, vmware worked... paaaaiiiinfully slowly, it wouldve taked 8 days to compile the whole rockbox firmware for JUST the e200s
14:20:14kadobani have no idea how it can even possibly take that long. i have an old 200 mhz pentium 2 or something laptop that can do builds, and it doesn't take anywhere near that...
14:20:21 Quit fredddy (Remote closed the connection)
14:20:44gartralkadoban: could it be qemu?
14:21:12Lloreangartral: If you're using qemu, you're not using vmware...
14:21:22gartralvmware image...
14:21:55LloreanYes, but it's emulated differently. It's a different situation.
14:22:05LloreanTelling us "vmware is slow" when you're not even actually running the application is misleading.
14:22:31LloreanHave you even actually tried vmware?
14:22:54t0masgartral: what kind of setup of qemu do you have?
14:23:11t0masit can run in actual x86 emulation mode (indeed painfully slow.. and not intended for what you are trying to do)
14:23:24 Join gregzx [0] (
14:23:32t0masor in some accelerated mode iirc.. which is intended for just running another OS
14:23:43gartralyes, vmware itself ran into the thremal stability problem i was talking about last night, causing my system too overheat
14:23:51 Join ibseo [0] (
14:24:18gartralaccelerated mode, but it was still SLOW
14:24:22kadobani imagine that anything that actually runs well enough to task the processor is going to run into that...i guess you need to fix that first
14:24:51Lloreangartral: Well, I'd recommend not getting into trying to code anything until after you've got compiling working locally
14:25:10gartralonly real fix is a new proc, and i have 0 money this monthtoo spare, for anything
14:25:14LloreanYou're probably going to need a lot of compile/test/improve cycles.
14:25:58t0masgartral: but you are running Windows right?
14:26:08gartralyes, at core
14:26:10n1sgartral: sounds more like you need to fix the cooling but this is off topic here
14:26:14 Quit schNibbler (Remote closed the connection)
14:26:16t0masisn't the cigwin toolset still availible?
14:26:36n1syes it is
14:26:44 Join itcheg [0] (i=62db4767@gateway/web/ajax/
14:26:52gartralcigwwin seems too have too little ram with just one gig, and adding more ram would only further heat troble
14:27:13t0maswhat? 1gb not enough?
14:27:19t0mashave you disabled the windows swap file?
14:27:40t0mas(disclaimer: I'm not a windows user.. last time I booted windows was months ago..)
14:28:06t0masbut it has some option to enable the "pagefile" as it's called in windows.. set it to automatic?
14:28:22t0masand make sure you have some diskspace left.. then it should run..
14:28:24LloreanI've successfully used cygwin with 1gb of RAM plenty of times.
14:29:03gartral300 gigs externally, 60 internally, 2 for page file on each
14:29:23t0masno way you can't run cygwin then..
14:30:08Lloreant0mas: His system is massively unstable and he's apparently got a nonstandard Windows install. Basically, it's best just to leave it at "he can't compile and can't fix it."
14:31:15gartralyea, thats good
14:31:24t0masgartral: or ?
14:31:34gartralas i said last nigh, "it works for what it was built for"
14:32:00t0mastake a week to learn to use it.. hang around in their respective IRC channels.. and you'll have a stable and working system :-)
14:32:37t0masbut indeed don't try to do any development while unable to actually build.. nobody is a good enough programmer to write code without testing
14:33:38gartralright, but im shur the issue is its an OLD "newark" 64 bit proc
14:33:57t0masmaybe Linux can handle that better? :-)
14:34:50gartralit wont get past lilo, or grub...
14:35:19t0masyou did install for the right architecture?
14:35:23gartraland noone i talked too said they could do anything
14:35:35gartralyes, 64 bit win
14:35:56t0masok.. over to #rockbox-community?
14:36:13gartral64 bit ubuntu, slack, debian, solaris, and zenwalk is ambigus
14:41:08 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:43:18 Quit kadoban (Remote closed the connection)
14:45:29Unhelpfulhow much memory does our vmware image expect?
14:49:05***Saving seen data "./dancer.seen"
14:49:49 Join kadoban [0] (
14:51:20 Quit gregzx (Read error: 104 (Connection reset by peer))
14:54:43 Quit ibseo ("quit")
14:57:15*Unhelpful is not really liking this MEM_FUNCTION_WRAPPERS business...
14:57:29Unhelpfuli just don't see how any plugin can know whether or not it will need them. :/
14:59:16 Join crwl [0] (
15:02:49 Nick fxb__ is now known as fxb (
15:06:00*Unhelpful notices that, at least in sim, calling rb->button_get(true) no longer seems to work...
15:12:33 Join MethoS- [0] (
15:24:44 Part gartral
15:35:19ajb`Who looks at build fixes? I've just uploaded FS #9758 to fix a problem with non git-svn repos.
15:35:59*Unhelpful is sure he already has a patche for that
15:37:53 Quit Acksaw (Read error: 104 (Connection reset by peer))
15:38:08 Join Acksaw [0] (
15:42:00 Join MethoS-- [0] (
15:44:49Unhelpfulooh, but this is nicer/more complete
15:51:27 Join TheSphinX^ [0] (
15:51:55 Quit MethoS-- (Remote closed the connection)
15:52:08 Join stripwax [0] (
15:52:46 Quit TheSphinX^ (Remote closed the connection)
15:53:02 Join TheSphinX^ [0] (
15:55:09 Quit MethoS- (Read error: 110 (Connection timed out))
15:58:16Unhelpfulajb: it looks good, but i don't have time to do a proper read now
16:02:20 Join Schmogel [0] (
16:06:15 Join ender` [0] (
16:06:43ajb`Unhelpful: no rush - at least my build works now
16:07:10 Nick esthar is now known as lymeca (
16:07:13ajb`Are many core devs using git now?
16:09:10Unhelpfuli use git via git-svn, i know that at least GodEater uses "pure" git, and i've seen more than a few patches that clearly came from git
16:09:23 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
16:11:52ajb`Any more comments on FS #9677 (Change erase dynamic playlist warning)?
16:12:30Unhelpfuli haven't seen it. i'll have to look later.
16:13:30Unhelpfulif anybody would care to comment on FS #9759... there are parts that are ugly hacks a bit, but the other equivalent hacks were uglier.
16:18:04ajb`The conditional stuff seems to grow doesn't it
16:18:43ajb`I guess a lot of this is boilerplat for becomming a Plugin?
16:19:48Unhelpfulin many ways. plugins don't have direct access to core functions
16:20:52Unhelpfulso a source file that depends on any functions at all from core is pretty much going to need a solution like that API macro to build in plugins or core
16:21:25 Quit Thundercloud (Remote closed the connection)
16:23:27Unhelpfulgevaerts: maybe this belongs here, and not the other channel. could the rule used for pluginlib sources be rewritten so as to apply to any .o in plugins/lib that builds from any .c file? and would that build command then apply if a rule just added a dep for a new .o file on a .c file somewhere?
16:23:37ajb`I would posit the BMP_LOAD macro is extraneous
16:24:16Unhelpfulit isn't. the core bmp loader is different from the pluginlib one.
16:24:38Unhelpfulwould an ifdef around the actual call be better?
16:25:32Unhelpfuli'll have to check back later, but i'll scroll back for suggestions
16:25:42ajb`Unhelpful: possibly, as you already have an #if LCD_DEPTH == 1 there
16:26:25ajb`it's a styalistic thing at most. Which is about all I can comment on at the moment. Otherwise the changes look OK to my unexperienced eye
16:27:07gevaertsUnhelpful: I don't know. I'd have to study the rule in detail, but I'm not really makefile guru
16:27:51 Quit itcheg (" ajax IRC Client")
16:44:32 Quit stripwax (Read error: 104 (Connection reset by peer))
16:47:21 Join stripwax [0] (
16:48:00 Join toffe82 [0] (
16:49:06***Saving seen data "./dancer.seen"
16:55:15toffe82anybody with a gigabeat S ? I would like to know what version of libmtp is working fine.
16:55:32Unhelpfulgevaerts: apologies, i thought you were one of the people who "got" make... i often don't.
16:55:46toffe82I have it working with the last version 0.3.0 but it give some error message
16:57:45 Join LqR [0] (
16:58:47LqRI'm researching possible new DAPs (or preferrably old, used devices), which ones work best with rockbox these days?
17:00:34LqRI've heard the iPod video port would be working nicely, but since my catastrophic experiences with iPod Photo a couple of years ago I've developed a bit of an iPod phobia.
17:00:59LqRAh, I knew there was a guide somewhere. Thanks! :)
17:01:52toffe82LqR: you can have a look there also :
17:02:25toffe82there are more player but not supported for the moment , like the gigabeat S
17:08:49 Join itcheg [0] (i=62db4767@gateway/web/ajax/
17:16:27 Join Lynx_ [0] (
17:20:52t0mashelp! ;-)
17:21:18t0masI am trying to get the ipod to work on an alarm-clock (made for ipod) when using Rockbox
17:22:05t0masthe problem is that the device pauses the player when it goes to "alarm clock" mode.. and then enables it at the wake-up time
17:22:21t0masthe apple firmware gets this in some way, pauses.. and starts playing when the alarmclock asks it to
17:22:34 Join Nibbl [0] (
17:22:36t0masbut rockbox returns to the menu.. and does nothing when the alarmclock tries to start playing
17:23:23t0masany suggestions as to how to debug this? It seems to trigger a "headphones_inserted" when the alarm sounds.. and a headphones disconnected when it shuts off to wait for the correct time
17:26:57 Quit bertrik (Remote closed the connection)
17:29:34stripwaxt0mas - are you using the 'pause on headphone unplug' / 'resume' settings in rockbox?
17:29:54stripwaxactually - how are you sure it's triggering a 'headphones_inserted'?
17:30:53t0masI am using that indeed
17:31:01t0masand that's why I know it's triggering them
17:31:35t0masif I switch to a song manually.. and then wait for the alarm time to roll around (before the player shuts down for inactivity) it start playing
17:31:53t0masthat might be because it "presses the play button"
17:32:27t0masbut I'm not sure how it does it with the apple firmware.. might be using other external commands (I guess it can send a lot of commands?)
17:32:30stripwaxt0mas - it sounds like it might be. but ipod accessories usually use the accessory protocol to pause/unpause
17:32:39 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
17:32:57t0masok, how does Rockbox react to that?
17:33:00stripwaxseems odd that it would use the headphone unplug as a way to pause (unless that's a side effect of something else).
17:33:01t0maswhen it's in the menu?
17:33:15t0maswell.. unplug is a pause trick in apple firmware
17:33:23stripwaxt0mas - if rockbox is already playing a track when you insert it into the dock, does it work?
17:33:35stripwaxthe apple firmware already has a 'pause' command in the accessory protocol
17:33:37t0maswait.. I'll get it an try some of those things
17:33:53stripwaxi don't remember offhand what the status of the apple accessory protocol in rockbox is
17:33:54t0masdoes the play command from an accessory work as a "resume" in rockbox?
17:34:13t0masthere is quite a list of working accessories.. which is why I wanted to check if this one worked :-)
17:34:15stripwaxshould work as play or unpause
17:34:18 Join gregzx [0] (
17:35:09 Quit toffe82 (Remote closed the connection)
17:35:36t0masok, but not resume?
17:36:06stripwaxMaybe I'm not sure what you mean by resume
17:36:16stripwaxif it's stopped, play plays. if it's paused, play should unpause
17:36:17t0masstripwax: do you know where it's handled? I'm looking for it in the code.. but didn't find any protocol handling yet :)
17:36:40stripwaxt0mas - here's the thread with the serial protocol patch discussion. the patch seems to be closed, but I don't know what the current status is
17:36:54stripwaxwow, that's some crazy link
17:37:07t0masI mean: if it is in the main menu, does a play/unpause command from the accessory-port start some music?
17:37:10stripwaxah. patch was committed in r19585!
17:37:32stripwaxt0mas - I don't know, but I would expect it to
17:37:34 Join ZincAlloy [0] (n=d9eeec31@gateway/web/cgi-irc/
17:39:05t0masI'll get the radio/alarm thingy and try
17:40:25 Quit gregzx (Read error: 104 (Connection reset by peer))
17:40:51 Join gregzx [0] (
17:41:27 Join toffe82 [0] (n=chatzill@
17:42:39toffe82this is the error I have using sendfirm with the last version of libmtp :
17:42:55t0masstripwax: pause/play and skip work as intended when the ipod is on the WPS
17:43:14stripwaxt0mas - hm, from the patch thread link I posted - there's this comment - "There is limited support for any IAP commands outside of the while-playing-screen. It sounds like sometimes the iPod is receiving the keydown command but not the keyup command. Not sure why that is happening, the best way to diagnose is to insert a hardware serial logger in between the devices; that's how I do it."
17:43:14t0masbut.. when it is in some kind of menu.. the pause/play button seems to work as a "select"
17:43:52toffe82to make it work I just delete the last parameter of the Send_File_From_File as it is not use anymore
17:44:07stripwaxmaybe the rockbox support is only fully enabled while on the wps screen. that's a guess, and sorry I don't have better info
17:44:46t0masstripwax: when I press pause/play in the main menu it definitely works as "select"
17:45:07t0masbut skip-back/previous doesn't work to exit a menu
17:45:15t0maswhile the button on the ipod does
17:45:31saratogatoffe82: for what its worth, I used the precompiled sendfirm on windows last night without issue
17:46:01toffe82saratoga: it is on linux with ubuntu 8.10
17:46:13stripwaxt0mas - I'm reading the thread just as quickly as you are, to see if there are any clues.
17:46:16t0masstripwax: I guess I'll have to change some menu-code to get it to do the effect of selecting "resume playing" or "now playing"
17:47:24stripwaxLooks like Ryan Sawhill has exactly the same problem on 29 October 2008
17:48:19stripwaxAnd Ryan Press' response immediately after it, saying there is limited support for accessory protocol commands outside of WPS.
17:48:53 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
17:49:36 Join gregzx [0] (
17:50:08stripwaxbut quite why play gets interpret by rockbox as Select when in a menu, I don't know. Actually, probably what's happening, is that the protocol sends a different code for Play than pressing the real Play button. in the WPS, both are interpreted as Play. in the menu, the accessory Play button is actually mapped, in the rockbox code, to Select
17:50:22stripwaxi'm guessing. I should shut up and look at the code
17:50:41saratogalooks like the Zune 2009 bug happens on the Gigabeat S too because of shared code with MS
17:50:43 Quit ender` (" Light a fire for a man, and he's warm for a day. Light a man on fire, and he's warm for the rest of his life.")
17:50:58saratogamaybe this will mean cheap Gigabeasts on ebay
17:51:09toffe82how do you see it ?
17:51:18t0massaratoga: lol.. that was the bug of the week man.. why in hell would they have put that while loop around the leap-year calculation? expecting years of 800 days? :P
17:51:37t0masstripwax: so it seems.. I'm looking into what the patch changed.. seems to do some button mapping
17:52:04stripwaxwell, that patch is now in svn :)
17:52:23t0masuhu.. but it's easier to search in the patch
17:52:41stripwax:) true. probably firmware/export/iap.h and apps/keymaps/keymap-ipod.c are the main ones here
17:53:11stripwaxok, i'm going to stop saying stuff until I've looked at a piece of code this year :)
17:53:41stripwaxsaratoga - what's the bug
17:53:42t0masI'm reading iap.c
17:54:02Lloreansaratoga: Well, the bug cures itself a day later anyway, apparently.
17:54:22saratogastripwax: the clock driver freezes the player
17:55:08saratogatheres a bunch of people posting yesterday and today about it on ABI, so maybe the Toshiba one doesn't fix itself
17:55:35t0masstripwax: ok.. I've got something in keymap-ipod.x
17:55:53toffe82ther is no date or calendar on the S so how can it be affected ?
17:55:54t0masin remote_button_context_standard
17:56:01t0masit maps BUTTON_RC_PLAY to ACTION_STD_OK
17:56:11t0maswhich is not really "ok" for an ipod I guess
17:58:06t0masstripwax: the normal play button does map to "start playing and show the WPS" from the menu.. so if the mapping of the BUTTON_RC_PLAY would be the same as the normal play button it should work
17:58:21t0masnow the big question is.. why is it mapped to SELECT / OK instead of PLAY?
17:58:27*stripwax shrugs
17:58:38t0masas the button is labeled play
17:58:46Lloreant0mas: In what screen?
17:59:06t0masmain menu
17:59:13LloreanDoes the remote have an okay button?
17:59:21Unhelpfulajb`: i see what you mean now. i would still prefer the macro, since it doesn't duplicate all of the function arguments. that seems like it will be better if the loader arguments change, down the road.
17:59:40LloreanOkay is more essential than Play in the menu, because you can choose "Resume Playback" with a 'select' but you can't do anything BUT resume playback with a Play
17:59:57t0masLlorean: mine doesn't
18:00:08t0masbut it doesn't work in a menu anyway
18:00:17t0masbecause next/previous don't work for selecting menu items
18:00:18Lloreant0mas: You can't "select" menu items?
18:00:29LloreanSo it seems that's the real problem.
18:00:39t0masbecause the buttons aren't supposed to do that..
18:00:45t0masat least.. not in the apple firmware
18:00:53LloreanWe aren't trying to duplicate the apple firmware...
18:01:05t0masI know
18:01:07LloreanThe goal should be to expose as much of Rockbox's functionality still.
18:01:21LloreanThis means they should allow navigation of the menu if there's enough buttons.
18:01:26t0masthis is not a "remote" in the sense that you put it inbetween the player and the earbuds..
18:01:48LloreanIf you just want to control playback, don't leave the WPS, though...
18:01:49t0masbut accessories aren't actual remotes with buttons in that sense of the word
18:01:52t0masthey send commands..
18:02:02t0masand if an accessory sends the "play" command.. the ipod should play music
18:02:19t0masthat's what it does with the OF.. and that's what people creating those accessories intended right?
18:02:36stripwaxLlorean - the remote unpauses the ipod. on original firmware, that works, because play does the unpause. on rockbox, when you stop you leave the wps
18:02:47t0masbecause accessories like alarm-clocks send a play command after booting the player..
18:03:11Lloreant0mas: Rockbox has its own wake-up alarm
18:03:18LloreanSo that's not a very good argument at all.
18:03:28stripwaxyeah, the real problem is that the command is sent automatically by the remote, not when you press a button.
18:03:42LloreanIt's send automatically if you use an unnecessary feature from the sound of it.
18:03:53LloreanCan't the iPod itself handle waking up with our firmware?
18:03:59 Join MethoS- [0] (
18:04:10 Quit MethoS- (Remote closed the connection)
18:04:25t0masLlorean: that wouldn't work.. because the device it's in is an alarm clock
18:04:30stripwaxt0mas - if you set the wakeup alarm in rockbox itself, does it still work if docked and the alarm goes off?
18:04:38t0maswhich powers up the ipod.. enables it's speakers.. and then sends a "play" command
18:04:42Lloreant0mas: You mean to say you can't just play music in the device? It ONLY works as an alarm clock?
18:05:05t0masif the device is set to ON.. and not to alarm-clock.. I can get rockbox to play music
18:05:06LloreanIs there some reason the alarm clock's speakers need to be powered down beforehand?
18:05:13t0masand then play/pause and next/previous work
18:05:16LloreanThen set the device to ON, then set Rockbox's alarm...
18:05:18t0masthere is..
18:05:23t0masit emits a lot of light when it is on
18:05:27t0maswhich is irritating for an alarm-lcok
18:05:46LloreanI don't see why all other remotes should be crippled just because you don't want to put a little electrical tape over some lights. :-P
18:06:13t0masbut.. the remotes don't work for using a menu now anyway..
18:06:19saratogapresumably some sort of workaround here is possible
18:06:29saratogacouldn't you just remap the command if you're not in the WPS?
18:06:37t0masindeed.. sync clocks in rockbox and the alarm.. set rockbox to alarm 1 minute after the alarmclock
18:06:38Lloreansaratoga: That's the problem
18:06:56Lloreansaratoga: If you're not in the WPS, "Play" on the 'remote' currently acts as 'select' acts on the physical unit, it sounds like.
18:07:01t0masso the clock powers it up.. asks it to play (and fails because Rockbox doesn't respond to play commands) and then rockboxs decides to play because of it's own clock
18:07:06t0masbut that seems rather user-unfriendly?
18:07:15 Join MethoS- [0] (
18:07:18t0masLlorean: indeed
18:07:33t0masand it should not respond to "play" commands by selecting a menu option..
18:07:39t0masactually.. the button is not mapped in the code at all
18:07:48t0masbut in some way it acts as select
18:07:50Lloreant0mas: I think most iPod remotes would allow navigating the menus
18:07:58 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
18:07:59LloreanIt's non-remote docks that don't
18:08:09LloreanSince they rarely adjust the iPod's volume
18:08:10t0masok.. I'm going to make it worse..
18:08:14t0masthere's a remote with this thing too :P
18:08:21t0maswhich actually has way more buttons
18:08:39LloreanThe iPod remote has +, -, >>, <<, and "Play"
18:08:40t0masbut with that too.. there is no way to navigate the menu's at the moment
18:08:48t0masthis device too
18:08:52Llorean+ and - should let you navigate the menus
18:09:01LloreanIf they control the iPod volume, and not the dock volume
18:09:02t0masvol+, vol-, skip-next, skip-back, play/pause
18:09:11t0masthey control the dock volume I guess
18:09:12LloreanThe problem is, yours probably control your speaker volume
18:09:36saratogado ipod remotes use the iap?
18:09:42Lloreansaratoga: They use the dock port
18:09:45LloreanI can only assume they do.
18:09:54t0masipod volume has no effect on the dock.. in the apple firmware you can't change the volume when it's in this dock iirc.. in rockbox I can change the volume.. but it has no effect
18:10:21Lloreant0mas: But the apple remote can change the headphone volume.
18:10:28saratogaeven if they use the iap they almost certainly use a different iap mode number, so perhaps the button could be remapped for whatever mode clocks and stereos use
18:10:28LloreanRemember, we're not JUST talking about docks here.
18:10:39Lloreansaratoga: They probably use simple remote.
18:10:53Lloreansaratoga: there's not necessarily a reason for clocks to use advanced remote.
18:10:58saratogaLlorean: IIRC even in simple theres a lot of different modes
18:11:18LloreanI just don't think it's necessarily good to assume they'll use a different mode.
18:11:32LloreanInstead we should do something like renaming "car adapter" mode to "wake-up" mode
18:11:46LloreanAnd having it ignore "Play" commands issued within X seconds of bootup (user configurable?)
18:12:04LloreanSince it'll resume playback itself anyway in that mode, IIRC.
18:12:14t0maswait.. I can try that
18:12:20t0mashow do I get it to car adapter mode?
18:12:41LloreanThere's an option somewhere, should be in the manual. I never use it, and it may not be finished for iPods yet.
18:12:53LloreanIt's an old feature that hasn't necessarily been tweaked for newer hardware much.
18:13:49t0masok.. maybe a new option?
18:14:05t0masbecause when I set the "startup screen" to "resume playing" it works fine as an alarm clock
18:14:12t0masbecause the clock starts the ipod
18:14:34t0masso a mode of "shutdown and resume playing only the next time we boot" but then named better might be a solution :)
18:14:56LloreanWhy not just fix/improve car adapter mode?
18:15:07t0masI'm looking for it :)
18:15:17t0masI guess it's supposed to do something like that
18:15:24LloreanExcept it's not one-time only
18:21:50 Quit saratoga ("CGI:IRC (EOF)")
18:22:20 Quit TheSphinX^ (Read error: 60 (Operation timed out))
18:25:36 Join Aurix_Lexico [0] (
18:26:25stripwaxwhat's SYS_FOURHERTZ ?
18:26:46*stripwax apologises for being away from rockbox for so long.. happy new year!
18:30:38 Join TheSphinX^ [0] (
18:31:57 Quit TheSphinX^ (Remote closed the connection)
18:34:24 Join fdinel [0] (
18:38:12 Join ender` [0] (
18:42:20 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:49:08***Saving seen data "./dancer.seen"
18:59:19 Join _fdinel [0] (
19:02:33 Quit Darksair ("People who are zhuangbility want to show their niubility but only reflect their shability.")
19:09:46 Quit fdinel (Read error: 110 (Connection timed out))
19:10:28 Nick _fdinel is now known as fdinel (
19:10:45t0masstripwax: I've been away longer I guess.. about since ehm.. 2006 or something
19:12:19 Quit t0mas ("running off.. time to get a beer with the girlfriend :-)")
19:12:51 Join tapiocapudding [0] (n=chatzill@
19:13:16 Join MOHDMACH [0] (
19:16:08 Join einhirn [0] (
19:16:50 Quit Lear (Read error: 110 (Connection timed out))
19:21:07 Quit itcheg (" ajax IRC Client")
19:21:23 Join Willwolfe [0] (
19:21:51 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
19:24:19 Join itcheg [0] (i=62db4767@gateway/web/ajax/
19:27:02 Quit ZincAlloy ("CGI:IRC (Ping timeout)")
19:28:01 Quit Nibbl (Remote closed the connection)
19:30:09 Quit itcheg (" ajax IRC Client")
19:30:31 Quit MOHDMACH ()
19:31:57 Quit LqR ("ERC Version 5.3 (IRC client for Emacs)")
19:35:00 Quit tapiocapudding ("ChatZilla 0.9.83 [Firefox 3.0.5/2008120122]")
19:36:43 Join itcheg [0] (i=62db4767@gateway/web/ajax/
19:45:52 Quit MethoS- (Remote closed the connection)
19:48:43 Join MOHDMACH [0] (
19:50:26 Quit Willwolfe ("ChatZilla 0.9.84 [Firefox 3.0.1/2008070208]")
19:50:49 Join m0f0x [0] (n=m0f0x@
19:56:12 Quit itcheg (" ajax IRC Client")
19:56:23 Join MethoS- [0] (
19:56:42 Join bertrik [0] (
20:01:43 Join kugel [0] (n=chatzill@unaffiliated/kugel)
20:08:28 Join likemindead [0] (
20:09:15*likemindead is away: I'm busy so STFU.
20:11:09*Bagder makes a note
20:14:55 Join Seed [0] (
20:15:12 Quit bertrik (Remote closed the connection)
20:15:52 Join merbanan [0] (n=banan@
20:19:12*likemindead is back (gone 00:09:58)
20:19:26 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
20:21:27 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
20:23:22 Quit mcuelenaere (Read error: 131 (Connection reset by peer))
20:23:24 Join mcuelenaere_ [0] (n=mcuelena@rockbox/developer/mcuelenaere)
20:24:42 Quit {phoenix} (Remote closed the connection)
20:26:28 Quit jhulst (Read error: 60 (Operation timed out))
20:28:51 Nick mcuelenaere_ is now known as mcuelenaere (n=mcuelena@rockbox/developer/mcuelenaere)
20:29:44 Join {phoenix} [0] (
20:30:27 Join itcheg [0] (i=4ae91e22@gateway/web/ajax/
20:33:39 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:36:41*amiconn starts to think that SYS_FOURHERTZ is somewhat nasty, and there should be another solution
20:38:48*likemindead is installing 3.1 & is xexcitedx
20:41:24 Join massiveH [0] (
20:42:42 Join Keripo [0] (
20:44:02KeripoI happened to stumble across engadget's article on the Cowon S9 and I was wondering about people's opinion on it as an MP3 and video player
20:45:00KeripoBeen looking for a good FLAC player for a while since my iPod's dying - opinions on the Cowom S9?
20:45:40 Part massiveH ("Leaving")
20:45:52krazykitKeripo, it doesn't run rockbox.
20:46:23krazykitand, as noted in the IrcGuidelines, non-rockbox discussion is offtopic.
20:47:09likemindeadI'm using the latest Rockbox Utility in Linux and I'm getting "Permission for disc access denied!"
20:47:10Keripoah, I was kinda unsure whether to ask here or in #rockbox-community
20:48:16krazykitlikemindead, are you running with administrator/root permissions?
20:49:09***Saving seen data "./dancer.seen"
20:49:54 Part Keripo
20:50:03krazykitthen do so
20:50:28domonokylikemindead: rbutil needs raw-disc access to install the bootloader on some players, so you should run it as root. (or add the correct rights to your user)
20:50:52likemindeadWill do.
20:53:51 Quit itcheg (
20:53:51 Quit tvelocity (
20:53:51 Quit AndyIL (
20:53:51 Quit rasher (
20:53:51 Quit Seed (
20:53:51 Quit tyfoo (
20:53:51 Quit SUSaiyan (
20:53:51 Quit ch4os (
20:53:51 Quit crashd (
20:53:51 Quit JdGordon|zzz (
20:53:51 Quit goffa (
20:53:51 Quit courtc (
20:53:51 Quit larstobi (
20:53:51 Quit Acksaw (
20:53:51 Quit Riku (
20:53:51 Quit solexx_ (
20:53:51 Quit J-23 (
20:53:51 Quit Slasheri (
20:53:51 Quit jon-kha (
20:53:51 Quit Krisinu (
20:53:51 Quit desowin_ (
20:53:51 Quit blahrus (
20:53:51 Quit parafin (
20:53:51 Quit freqmod_qu (
20:53:51 Quit MethoS- (
20:53:51 Quit Aurix_Lexico (
20:53:51 Quit domonoky (
20:53:51 Quit crwl (
20:53:51 Quit kadoban (
20:53:51 Quit timc (
20:53:51 Quit pixelma (
20:53:51 Quit _Auron_ (
20:53:51 Quit write__erase (
20:53:51 Quit unstable (
20:53:51 Quit amigan_ (
20:53:51 Quit lastebil (
20:53:51 Quit scorche (
20:53:51 Quit Hadaka (
20:53:51 Quit BlakeJohnson86 (
20:53:51 Quit XavierGr (
20:53:51 Quit mcuelenaere (
20:53:51 Quit MOHDMACH (
20:53:51 Quit TMM (
20:53:51 Quit moos (
20:53:51 Quit daurnimator (
20:53:51 Quit thegeek (
20:53:51 Quit soap (
20:53:51 Quit jfc (
20:53:51 Quit reacocard (
20:53:51 Quit trisiak (
20:53:51 Quit ment (
20:53:51 Quit pabs (
20:53:51 Quit flux (
20:53:51 Quit fred_2 (
20:53:51 Quit markun (
20:53:51 Quit loswillios (
20:53:51 Quit Neovanglist (
20:54:20NJoinXavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:54:20NJoinitcheg [0] (i=4ae91e22@gateway/web/ajax/
20:54:20NJoinmcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
20:54:20NJoinSeed [0] (
20:54:20NJoinMethoS- [0] (
20:54:20NJoinMOHDMACH [0] (
20:54:20NJoinAurix_Lexico [0] (
20:54:20NJoindomonoky [0] (n=Domonoky@rockbox/developer/domonoky)
20:54:20NJoinAcksaw [0] (
20:54:20NJoincrwl [0] (
20:54:20NJoinkadoban [0] (
20:54:20NJointvelocity [0] (
20:54:20NJoinTMM [0] (
20:54:20NJoinAndyIL [0] (i=AndyI@
20:54:20NJoinmoos [0] (i=Mustapha@rockbox/staff/moos)
20:54:20NJoinRiku [0] (
20:54:20NJointyfoo [0] (
20:54:20NJoindaurnimator [0] (n=quae@unaffiliated/daurnimator)
20:54:20NJointimc [0] (n=aoeu@
20:54:20NJoinSUSaiyan [0] (
20:54:20NJoinpixelma [0] (n=pixelma@rockbox/staff/pixelma)
20:54:20NJoinunstable [0] (i=unstable@tor/regular/sid)
20:54:20NJoinsolexx_ [0] (
20:54:20NJoin_Auron_ [0] (
20:54:20NJointhegeek [0] (
20:54:20NJoinJ-23 [0] (
20:54:20NJoinparafin [0] (
20:54:20NJoinsoap [50] (n=soap@rockbox/staff/soap)
20:54:20NJoinwrite__erase [0] (
20:54:20NJoinch4os [0] (n=ch4os@gentoo/user/ch4os)
20:54:20NJoinjfc [0] (
20:54:20NJoinreacocard [0] (i=reacocar@
20:54:20NJoinKrisinu [0] (n=Krisinu@
20:54:20NJoinBlakeJohnson86 [0] (
20:54:20NJoingoffa [0] (n=goffa@
20:54:20NJoinpabs [0] (
20:54:20NJoinJdGordon|zzz [0] (n=jonno@rockbox/developer/JdGordon)
20:54:20NJoinment [0] (
20:54:20NJoinscorche [0] (i=Blah@rockbox/administrator/scorche)
20:54:20NJoinblahrus [0] (n=blahrus@
20:54:20NJoinHadaka [0] (
20:54:20NJoinloswillios [0] (n=jan@unaffiliated/loswillios)
20:54:20NJointrisiak [0] (
20:54:20NJoinmarkun [50] (n=markun@rockbox/developer/markun)
20:54:20NJoincourtc [0] (n=court@unaffiliated/courtc)
20:54:20NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
20:54:20NJoinfreqmod_qu [0] (i=quassel@2001:700:300:1430:213:d3ff:fee9:5ed0)
20:54:20NJoinamigan_ [0] (
20:54:20NJoincrashd [0] (
20:54:20NJoinjon-kha [0] (
20:54:20NJoinflux [0] (
20:54:20NJoinlastebil [0] (
20:54:20NJoinfred_2 [0] (
20:54:20NJoindesowin_ [0] (n=desowin@
20:54:20NJoinlarstobi [0] (n=larstobi@
20:54:20NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
20:54:20NJoinNeovanglist [0] (i=Neovangl@
20:54:38 Join bertrik [0] (
20:55:38likemindeadOkay, progress is at 100%, but the icon below still says "Abort." Do I close the window or wait a bit longer?
20:59:31likemindeadAh.... nevermind. It finally finished after the themes not being available error. Thanks!
21:07:12 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:08:41kugeljhMikeS: was there a decision on the backlight now?
21:09:08 Join gregzx [0] (
21:10:03 Part likemindead ("Making like a tree...")
21:12:01 Quit jhulst (Read error: 113 (No route to host))
21:13:42 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
21:13:48 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:14:36saratogadoes anyone remember where the MrH document about PP's memory and cache performance was put?
21:16:07linuxstbsaratoga: Maybe one of these?
21:16:13mcuelenaeresaratoga: ?
21:16:32 Join fredddy [0] (
21:19:43saratogamcuelenaere: thanks
21:21:20 Quit itcheg (" ajax IRC Client")
21:21:41 Join notlistening [0] (
21:21:50 Join itcheg [0] (i=4ae91e22@gateway/web/ajax/
21:22:09 Quit timc (Read error: 60 (Operation timed out))
21:22:30saratogashame he doesn't compute the MB/s though
21:22:50notlisteningUSB or parallel port ARM JTAG cables, any huge differences between them ?
21:23:32saratogaUSB ones are really expensive, otherwise i don't think so
21:23:36Unhelpfulaside from the obvious ones? and the usb one *might* do some of the work for you?
21:23:57saratogadoes the Ipod Video have 128k of IRAM?
21:24:30notlisteningah well the obvious ones? I have never used JTAG before :P
21:24:34Unhelpfulsaratoga: the parallel one will have you twiddling bits... then again, it's hard to say if the USB one will have any kind of intelligent controller in it, or if it will have you twiddling bits over USB?
21:25:00Unhelpfulnotlistening: the obvious difference would be that one requires a parallel port, the other a usb port. obviously. ;)
21:25:10saratogaah it defines PP5022
21:25:45notlisteningoh haha thought there might be something more technical than that ;)
21:25:48saratogaeither way your software will handle the actual USB or parallel interface
21:25:56saratogaso you shouldn't have to care about the differences
21:25:58 Join cmdbbq [0] (
21:26:44Unhelpfuli've only used it for a bricked wrt router before... actual debugging and such is another matter
21:26:50notlisteningWould someone mind looking at a website and suggesting if the parallel ones is suitable, it has been suggested that speed might be an issue?
21:27:36Unhelpfulonly for large volumes of data. you do *not* want to have to flash a device over it.
21:28:04notlisteningnah not doing that just debugging
21:29:02notlisteningI am just being tight about the cost ;)
21:29:52Unhelpfuli'd imagine parallel will likely suffice, if you have an old beater of a PC around with the port
21:30:11Unhelpfulit will likely require the "newest" bidirectional PP standards ;)
21:30:25 Quit cmdbbq (Remote closed the connection)
21:32:21 Quit itcheg (" ajax IRC Client")
21:32:59 Join cmdbbq [0] (
21:33:43notlisteningabout a 2000 / 2001 machine i have dedicated to development, umm is there a way to tell?
21:35:20kugelsaratoga: I'd have a plugins patch, but it's quite inconsistent
21:35:36kugelfor the fuze
21:36:19gevaertsUnhelpful: I do know make a bit, and I can understand it with some effort, but I've never really looked at the rockbox build system
21:36:33 Join Blackfrosch [0] (
21:36:39Unhelpfuli'd imagine your software might say what you need, probably EPP or ECP... i'd be shocked to find a system with a parallel port that doesn't support them
21:36:59 Join timc [0] (n=aoeu@
21:37:31notlisteningthanks Unhelpful
21:37:44Unhelpfulgevaerts: the only thing i can think of is to stuff the pluginlib object commands into a variable, and use that for the extra pluginlib objects from core. i don't think there's any other way for rules with different target/dep patterns to have the same build command?
21:38:27gevaertsI can't think of a way
21:38:30*kugel tries pf on his fuze
21:38:33Unhelpfulthat aside, there's also the issue of not being able to cleanly support building those objects conditionally in the makefile - SOURCES is preprocessed, so any of the config defines are available there
21:38:51Unhelpfulkugel: it has what, 2MB RAM? good luck. ;D
21:39:03kugelfull plugin buffer
21:39:35Unhelpfulohhhh. yeah, should work fine, then. why did i think it was some ultra-lowmem thing?
21:39:52kugelit's slow
21:40:01kugelit boosts, doesn't it?
21:40:37 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
21:40:52Unhelpfulkugel: it does. the dynamic reflections add 3 multiplies per reflection pixel. i might be able to take it down to two with enough precision to still work.
21:41:12kugelit's slower than the e200
21:41:24kugelwho has 1/4 MHz
21:41:38kugelsame architecture
21:41:50Unhelpfulnot transposing also adds a multiply per output pixel. i know a cleaner way to transpose the images, with a scaler output plugin... i just need to write one.
21:43:43 Quit Blackfrosch (Read error: 104 (Connection reset by peer))
21:44:36kugelUnhelpful: another note: none of the covers is obviously broken
21:45:14Unhelpfulgevaerts: i'll go ahead and commit the c-file-include version, then. if somebody comes up with a nicer way, later, at least the part for making it compilable in the right place is already "done"
21:45:19Unhelpfulkugel: should they be?
21:45:43kugelDid you forget the bug?
21:46:16UnhelpfulOH. did you back out the chunk-size change?
21:46:38UnhelpfulBM_MAX_WIDTH applies to *all* bitmap loads
21:46:57kugelwhat I have now is svn+plugins
21:47:22Unhelpfulhrm. are the background/wps bmps still bad?
21:48:34Unhelpfuli'd be surprised if they "just got better" in svn, unless somebody found whatever the issue was, and fixed it...
21:50:21unstableNot using git or mercurial yet?
21:51:34Unhelpfulthe "central" repo is still subversion. at least some of the devs use git on their end.
21:51:48gevaertssaratoga: I've just run beastdump on my beast and I get (not unexpectedly) a different md5sum. This is of course a beast with previously installed rockbox and a slightly more complex history than the average one
21:53:42 Quit fredddy (Remote closed the connection)
21:54:54kugelUnhelpful: my unscaled 150x150 covers are fine too
21:55:26Unhelpfulkugel: they will be, it's just that the scroll animation will cause some z-axis issues
21:55:33kugeleven though I can't view all for sime reason, pf stops substituting the empty slide with the covers at the end
21:56:17Unhelpfulbasically, the side images are rendered from the outermost in, and then the center image last. if they're small enough that the slides never intersect, fine.
21:56:33notlisteninganother quicky is a 128x128 pixel LCD with 4096 colours okay for rockbox same as nokia 6610?
21:56:53Unhelpfulif not, there's a point in the animation where the previously-center cover suddenly "pops" behind the new one, because it just got demoted from center to stack cover
21:56:54kugelUnhelpful: I don't understand
21:57:53 Join FOAD_ [0] (
22:00:32 Quit blithe ("Rebooting.")
22:01:08 Join robin0800 [0] (
22:02:00kugelUnhelpful: I was saying that all pixels of my 150x150 covers seem to be fine
22:02:01 Join blithe [0] (
22:02:24*Unhelpful has misunderstood, then :)
22:06:20kugelUnhelpful: doesn't that make you curious?
22:06:47kugelfrom what I've seen, every yellow is yellow, ever black and grey is black and grey
22:07:13Unhelpfulit does. what happens to the broken bg bitmap if you load another theme, then load cabbie again?
22:09:17kugelUnhelpful: hm, I have my doubts about a storage bug
22:09:28Unhelpfulbut, *where*, then?
22:09:43kugelUnhelpful: I did a small experiment
22:10:14kugelI rotated the cabbiev2 bg from my e200 (same resolution, but portrait).
22:10:25kugelI don't get the weird pixels
22:10:36kugelbut the same behavior at the rockbox logo which I get in the wps
22:10:46 Part cmdbbq ("later")
22:11:05kugelsee here:
22:11:24kugelthe ROCK in rockbox is weird
22:11:41kugelthat's exactly what I get with the rotated e200 cabbie bg
22:11:45Unhelpfulyes, it looks to me like some of the lines have been offset a pixel or two?
22:11:55 Join mcuelenaere_ [0] (n=mcuelena@rockbox/developer/mcuelenaere)
22:12:21Unhelpfulthe curious thing about that is that the entire line is read in at one go...
22:12:45 Quit mcuelenaere (Read error: 113 (No route to host))
22:13:22Unhelpfulit's almost like a couple of bytes have been dropped from the line on read
22:13:37domonokynotlistening: of course its ok, rockbox also runs on players where only a 2 line character display is available..
22:13:55 Quit FOAD (Read error: 110 (Connection timed out))
22:13:56 Nick FOAD_ is now known as FOAD (
22:14:09kugelUnhelpful: so, that shows that this behavior is exactly reproduce-able on 2 different files
22:14:30jhMikeSkugel: I thought something had been agreed upon with #defines already.
22:15:00kugeljhMikeS: yea, but I'm still curious about the bool setting thing
22:15:31kugelI could integrate that in my code police patch if wanted
22:16:36kugelUnhelpful: and if the same behavior on 2 different files can be noticed, it's unlikely a storage issue, is it?
22:16:50 Quit mcuelenaere_ (Read error: 113 (No route to host))
22:16:56jhMikeSI'm not sure we need a fading define at all for a custom algorithm but just a setting type define. In common cases, the setting type define can be implied by it.
22:17:37Unhelpful... i don't see why that makes it not storage. if large reads are losing/corrupting data, why would it only affect one file?
22:17:38kugelexactly, your define is basically only needed for the setting
22:17:57kugelUnhelpful: because the curroption is exactly the same?
22:18:17 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
22:18:21kugelif the corruption was different, ok, but this is entirely the same
22:18:31Unhelpfullooks like lines shifted horizontally? all that takes is losing two bytes.
22:18:42 Join Zagor [242] (n=bjst@rockbox/developer/Zagor)
22:19:00kugelplus, all my album art is read correctly
22:19:31kugelhundreds of bitmaps, 2 with the same corruption, every other without
22:19:39kugelthat doesn't seem like a storage problem to me
22:19:39 Join petur [50] (n=petur@rockbox/developer/petur)
22:20:16kugelmeh, but it stills seems to be limited to the fuze, that's so weird
22:20:25Unhelpfuli have a really hard time believing it's a loader problem, though, when simply changing the chunk size "fixes" it
22:22:10kugelUnhelpful: want a screendump from the rotated e200 bg? The weird pixels in the Rockbox lettering are at the same position
22:22:54Unhelpfulsure. i'll take a look, but i'm still pretty sure this is not my bug.
22:25:21 Quit kugel (Remote closed the connection)
22:25:41 Join kugel [0] (n=chatzill@unaffiliated/kugel)
22:26:15 Quit blithe ("Lost terminal")
22:26:25 Join blithe [0] (
22:33:04kugelUnhelpful: hm, your last commit looks weird to me
22:33:07 Join meyerman [0] (
22:33:26kugelI don't think #ifdef PLUGIN and the stuff within should be in a core file
22:33:45Unhelpfulkugel: no way not to, other than having two files.
22:34:58kugellooks nasty :/ but if you say so it's probably right
22:35:34Unhelpfulit does look nasty... but i rather thought that forking all of bmp.c and resize.c was nastier.
22:36:39kugelhm, couldn't you include a api.h which only has the #define API which is different in a plugin?
22:37:14kugelbut I see there's more dependent on plugin
22:37:26kugelanyway, have you compared the images?
22:38:37 Join ameyer [0] (n=ameyer17@
22:38:49Unhelpfulyes, a few things are different about the bmp loader, also the API define has to be set when the bmp.c/resize.c are compiled - the pluginlib versions of them can't go directly to core for their files.
22:38:59n1snotlistening: we already support a player with 128x128pixels lcd the H10 5/6GB
22:39:12Unhelpfulerm, s/files/functions
22:39:43n1sLlorean: I think the only thin "Car adapter mode" does is turn the player off if the charger is removed so it powers down when you shut down the car
22:40:15kugelUnhelpful: hum? if you define API in the plugin lib file it's available at compile time?
22:41:23Unhelpfulbmp.c and resize.c need to be compiled with the proper version of API defined, and they're compiled into pluginlib, not as part of a plugin
22:41:59 Part blippe
22:42:36UnhelpfulAPI *also* needs to be defined when they're built in core, though
22:43:31moosUnhelpful: just in case you miss them, yellow builds
22:43:34Unhelpfulbut first, i need to fix some yellows :/
22:43:42mooshehe :)
22:44:07 Join cemysce [0] (
22:44:32kugelI thought the pluginlib is compiled into any plugin it uses it
22:45:10Unhelpfulkugel: it's compiled into a library, and linked into any plugin that uses it.
22:45:21kugelyea, I meant that
22:45:47Unhelpfuli think i yellowed every *non*-mono target?
22:45:54Unhelpfulor many of them, anyway
22:48:03 Join MethoS_ [0] (
22:48:10*Unhelpful sees the problem :)
22:49:10***Saving seen data "./dancer.seen"
22:51:52Unhelpfulkugel: moving the API define into somewhere like plugins/lib/core.h or similar would be a nice cleanup, to avoid defining it twice... but there's really no way to get around using the API macro in the core files, aside from maybe an include that would wrap the functions in API calls under their original names... the plugin would then need to make sure that it includes that :/
22:53:17 Quit bertrik ("Leaving")
22:53:26 Join casainho [0] (n=chatzill@
22:53:47 Quit Zagor ("Leaving")
22:53:59kugelUnhelpful: yea, I see the api need
22:54:09casainhohello :-)
22:54:13 Quit MethoS- (Read error: 60 (Operation timed out))
22:54:33kugelbut apps/plugins/lib/core_bmp.c could like like
22:55:11casainhoI am evaluating an chip stereo audio dac and I would like to know if an IC with SPI bus protocol at 10MHz would be sufficient?
22:55:30Unhelpfulkugel: i like that... but we'd have to then put the API define somewhere in core, right?
22:56:00 Quit MethoS_ (Read error: 131 (Connection reset by peer))
22:56:16kugelthe api could be in bmp.h/c
22:56:28kugelthat's what's the #undef is fore
22:57:44 Quit XavierGr (Nick collision from services.)
22:57:55 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:58:11Unhelpfulit appears in resize.c as well... that's where i was going with the plugins/lib/core.h idea, that an #undef/#define like that could happen there, and that the API define for core could be somewhere "else"
22:59:18Unhelpfulalbumart.c will need to get the same treatment to support pictureflow-on-mono, which is the request that this is trying to fill, anyway... so the question then is, where does the macro belong?
22:59:35kugelUnhelpful: why is much code #ifndef PLUGIN? only because what you did is only for mono?
22:59:58kugelbmp.h? they all include it afaik
23:00:03 Quit {phoenix} (Remote closed the connection)
23:00:18kugelor some "api.h"
23:00:24 Quit m0f0x (Read error: 110 (Connection timed out))
23:00:42Unhelpfulyes, that's where i was going, there should be a "proper" place for such a def if we want to make more core code shared with plugins.
23:01:30scorche|shdomonoky: ping
23:01:31Unhelpfulthe #ifndef plugin is mostly stripping out the native-output code. scaled native output is worthless on mono, and unscaled native output is available from the core bmp loader
23:01:48Unhelpfulthe pluginlib one can be smaller if it only supports scaled loads
23:01:53kugelso only to make plugins a little smaller?
23:02:12kugelgood enough I suppose
23:02:29*Unhelpful sees that he got dinged by some spurious yellows again
23:02:40kugeljhMikeS: so there was no definite decision about dropping the setting, right=
23:03:37kugelrather, simplifying to yes/no
23:05:22 Join Rob2222 [0] (
23:10:24 Quit petur ("Zzzzzz")
23:10:58*pixelma wonders about Febs' mail in the dev list
23:11:21Bagderit puzzles me too
23:11:33BagderI figure he meant to actually add contents too...
23:12:26 Join tyfoo2 [0] (
23:13:23pixelmaor he accidentally pressed "reply" or some such
23:15:28casainhoBagder: can you please help me on audio codec IC?
23:15:57Bagdercasainho: I don't think I'm the man to ask for that, sorry.
23:16:01casainhoBagder: I would like to know if this module would be ok to play audio:
23:16:11 Quit einhirn (Read error: 104 (Connection reset by peer))
23:17:01 Join countrymonkey [0] (n=4b05639a@gateway/web/cgi-irc/
23:17:05casainhoBagder: don't you know about the audio "engine" on Rockbox? - I just would like to know if audio goes to audio drivers in PCM or something like that
23:17:11Bagderhardware decoder sounds wrong though
23:17:33Bagderyes, the audio chip basically gets pcm
23:18:12casainhoBagder: so, that module may work, as it can work as a slave and get PCM from a master :-)
23:18:43casainhoBagder: what about LCD? Do you think this would work?
23:18:52Bagder"Not recommended for new projects"
23:19:35Bagder12bit color is unusual, but sure
23:19:54 Quit bmbl ("Woah!")
23:20:51casainhoBagder: well, it's just a prototype, I just need to convince Olimex or other big company to workout a hardware player after seeing our prototype, I don't mind if I use an "old" stereo IC codec
23:21:23casainhoBagder: I would like to know if other targets uses SPI or I2C for LCD..?
23:21:28saratogagevaerts: that makes sense, I'd only expect this to work if you haven't written anything else to the firmware partition
23:21:33Bagderwell, making a new design with a chip that already is discontinued seems like a risk to me
23:21:54gevaertssaratoga: it does indeed. It was worth a try though
23:21:57toffe82I have a strange bbehavior on 2 gigabeat S, the battery is almost full, I switch off the player and try to switch it on and I have the icon of the OF on black background showing empty battery and the screen flashing between white and black like a full discharge battery
23:22:01 Quit countrymonkey ("CGI:IRC (Ping timeout)")
23:22:23Bagdercasainho: btw, the irivier h1x0 series have the S1D15E06 LCD which might be similar (controller-wise)
23:22:29casainhoBagder: but is just temporary, we will never make the final prototype −− the company that will build it must do it!! and is more a question of making new drivers
23:22:44 Join Davide-NYC [0] (
23:23:18casainhoBagder: hmmm, okok. I started today reading the drivers of LCD on RB source... :-)
23:23:18saratogai still need someone with a fresh gigabeat to try it and confirm that my results are typical
23:23:38casainhoBagder: okok - thanks, it's all for now :-)
23:23:51saratogaanyway assuming this all works, what is needed to get the gigabeat S supported
23:24:07Unhelpfulsaratoga: i could restore mine with the "stock" firmware, then install the bootloader... but i guess that's not fresh enough? :/
23:24:30 Quit Nico_P (Remote closed the connection)
23:25:03toffe82saratoga: would it be a problem if the nk.bin installed on the player is already the rockbox one, the md5 will not be the same ?
23:25:03saratogaUnhelpful: i think we'd really need to restore it with an image dumped from a fresh gigabeat firmware partition
23:25:41saratogatoffe82: this only works if you've installed the rockbox bootloader a single time and then not touched the firmware partition at all
23:26:10saratogaif you just installed and then didn't do anything else to the firmware partition, it should work
23:26:21saratogaif you've reinstalled the bootloader it probably will not
23:26:22Unhelpfulsaratoga: ... actually, it can recover from a wiped partition table... is "after it formats itself" good enough?
23:26:51toffe82saratoga: what do I have to do to try ?
23:26:51 Quit tyfoo (Connection timed out)
23:26:51saratogaUnhelpful: it can do that after you've erased the original nk.bin?
23:26:51 Nick tyfoo2 is now known as tyfoo (
23:27:17saratogatoffe82: compile and run the beastdump.c program in utils/MTP
23:27:22Unhelpfuli would assume it did that without accessing the files, since it detected the table being corrupted :/
23:27:46 Join saratoga2 [0] (n=9803c6dd@gateway/web/cgi-irc/
23:28:02saratoga2where does it get the new nk.bin from then?
23:28:02 Quit saratoga ("CGI:IRC (EOF)")
23:28:35Unhelpfulit starts recovery mode, and asks you to give it one... so it would still be the one that was dumped, which i have a copy of from before we "removed" it from wiki
23:28:59saratoga2Unhelpful: ah ok, you can try that but i'm not sure if its the same as a fresh image
23:29:12saratoga2this program assumes that the nk.bin file is not fragmented
23:29:16gevaertsIt would be very useful to know...
23:29:34saratoga2which is may be if you readd the nk.bin rather then directly flash the entire partition
23:29:54kugelsaratoga2: is one allowed to wrote to the OF partition on the beast?
23:30:11saratoga2kugel: yes its just a normal tfat partitoin and it mounts in windows by default
23:30:18saratoga2at least when in bootloader usb mode
23:30:29kugelwithout rockbox too?
23:30:31Unhelpfulare we sure that they *do* image the whole drive?... then again, the recovery process is not convenient to use for initial install, i suppose
23:30:41saratoga2without rockbox you can get no UMS mode
23:30:58toffe82saratoga2: you have to put the player in recovery ?
23:30:59kugelif the nk.bin dump is safer if one isn't allowed to write OF partition, why allow it?
23:31:00Unhelpfulkugel: without rockbox, you can send an appropriate file via MTP, which it will treat as the new firmware
23:31:15kugelon the v1 sansas you cannot write to it either
23:31:36gevaertskugel: you can...
23:31:43saratoga2Unhelpful: I expect that they image all drives the same out of the factory, but thats only an assumption
23:31:52Unhelpfulkugel: if you try to mess with the on-disk partition table too much, the OF will format it for you.
23:32:11Unhelpfulsaratoga2: yes, thinking about it, the idea of using the device format/restore feature in the factory is absurdly silly
23:32:19kugelgevaerts: you can? you mean it's mounted in windows/whatever if you use usb with rockox?
23:32:42saratoga2kugel: its not partitioned so you can't mount it in windows
23:32:42gevaertskugel: it's not FAT, so you don't get a filesystem, but you can access it
23:32:46saratoga2but you can still write to it
23:32:46*Unhelpful vaguely remembers talk of filtering the MBR portion of the drive in rockbox usb?
23:33:01*gevaerts vaguely remembers that as well :)
23:33:10Unhelpfulback in the very early days of beast usb support
23:33:14kugelwell, anyway, I was talking about to hide it for the normal user, so the dump will be fine for the time being
23:33:15saratoga2anyway, what needs to happen to get the beast supported?
23:33:20kugelisn't that possible?
23:33:24gevaertsI still don't really like the idea of not showing what's really on disk
23:33:38saratoga2kugel: i think the best bet is to just dump the nk.bin during the install so they don't get a chance to screw it up
23:33:47Unhelpfulgevaerts: nor do i, and there's no way at all to do the "right" thing if they try to write to it.
23:34:17Unhelpfulsaratoga2: it would have to be on first boot, i'd think - the "new" firmware doesn't have disk access until it's installed and booted
23:34:37kugelsaratoga2: what's better? hiding 150MB of storage, or have users not being able to uninstall rockbox because they lost the dump for some reason?
23:34:40saratoga2Unhelpful: the install instructions say it reboots after sending the new firmware, at which point it comes up in UMS
23:34:40gevaertskugel: it is possible, by presenting a fake MBR. It won't help though, as once rockbox is installed that partition isn't pristine anymore anyway
23:34:48kugel(150MB is what I remember)
23:34:54saratoga2this didn't happen to me, I had to manually force a reboot, but in theory it could all be done at once
23:35:05toffe82saratoga2: how do you compile it ? :)
23:35:16gevaertskugel: if you upgrade the rockbox bootloader just once, you've lost recoverability
23:35:17saratoga2gcc beastdump.c
23:35:49kugelgevaerts: why?
23:36:04Unhelpfulsaratoga2: the util could also send the rockbox.tar before sending the firmware
23:36:06saratoga2then after you run it type "md5sum nk.bin" to make sure your nk file is ok
23:36:10kugelis the bootloader overwriting parts of the OF?
23:36:13 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121622]")
23:36:41saratoga2kugel: this is absically an undelete utility
23:36:46 Join itcheg [0] (i=62db4767@gateway/web/ajax/
23:36:54saratoga2you can only undelete a file if you haven't touched the file system since deleting it
23:37:04saratoga2otherwise it gets overwritten by subsequent files
23:37:11kugelyes, I know
23:37:29*gevaerts thinks
23:37:34Unhelpfulkugel: if you write to the partition later with an OS filesystem driver, the remnants of the old nk.bin may be overwritten
23:37:35kugeland if you don't overwrite it ever, it will always be recoverable
23:37:52gevaertsHow destructive is the firmware patching? Can't the rockbox bootloader be patched out cleanly again?
23:38:03kugelUnhelpful: and that's why I propose hiding that partition
23:38:18Unhelpfulgevaerts: we can't *make* the patched rockbox bootloader without an original nk.bin
23:38:21kugelat least in a way that it doesn't get automounted for the normal user
23:38:44gevaertsUnhelpful: sure, but right now we're talking about trying to make sure the user doesn't lose the file afterwards
23:38:46Unhelpfulkugel: and then your bootloader is install-once? i know they don't move as fast as the real thing, but bootloaders *are* still a moving target.
23:39:24gevaertskugel: just writing the patched bootloader during normal install will probably already make recovery impossible
23:39:28Unhelpfulgevaerts: the point is, you have to install a firmware the first time that contains no part of the original nk.bin
23:39:52Unhelpfulonce that boots, you can try to recover nk.bin and create a dual-boot bootloader...
23:39:54*gevaerts is slowly getting his head around it all
23:39:56kugelUnhelpful: well, that's why I was asking if it's needed that the bootloader overwrites parts of the of (Sorry, I'm not into patching the beast so much)
23:40:13gevaertskugel: yes it is
23:40:31Unhelpfulgevaerts: your question is, then, "could we extract nk.bin from the dual-boot bootloader"?
23:40:32gevaerts"overwriting parts of the of" means writing to a FAT partition in this case
23:41:00saratoga2when the gigabeast formats itself is the data partition lost or only the firmware partition?
23:41:04gevaertsUnhelpful: yes. That way keeping the file safe for a longer time is less of a problem.
23:41:05kugelso it would be at least nice to have a safe nk.bin until the bootloader is updated
23:41:24gevaertssaratoga2: don't count on keeping the data partition once the OF starts having fun
23:41:37amiconnWhy not undelete the original nk.bin on the fw partition, but naming it differently?
23:41:47gevaertskugel: yes, but that's part of the installation process
23:42:14Unhelpfulamiconn: if the device goes into recovery mode, it's probably gone, then :/
23:42:36amiconnYeah, hence I think that just dumping it onto the pc is better
23:42:52amiconnBut the fw partition should be hidden regardless, if only to reduce confusion
23:43:16*kugel agrees
23:43:19amiconnThis can be done by just modifying the partition type on-the-fly (to "hidden fat32")
23:43:33Unhelpfuli don't know how flexible recovery mode is. i know it can format the data partition without needing a fresh nk.bin upload. i know it can recover from the partition table being corrupted (thanks gevaerts ;), presumably without accessing data on the partitions
23:43:38amiconnThis way someone who knows what he's doing can still mount it
23:43:39kugelUnhelpful: recovery mode is something from the of I suppose?
23:43:57kugelin that case, you can just undelete/dump it again after recovery mode, can't you?
23:44:02 Quit stripwax (Read error: 104 (Connection reset by peer))
23:44:18 Quit merbanan (Remote closed the connection)
23:44:42Unhelpfulkugel: recovery mode is clearly started from something earlier than the OF nk.bin. i would think flash.
23:45:00saratoga2does it save the nk.bin when it formats?
23:45:03kugelah ok
23:45:28Unhelpfulsaratoga2: if there is one already? i would rather doubt it does, and i'm not sure how to check
23:45:40toffe82saratoga2: in recovery, it just erase the nk.bin
23:46:22Unhelpfuli *suspect* that it formats the fw partition when it decides you need to reinstall the fw
23:47:06saratoga2then it asks you for a new nk.bin?
23:48:24Unhelpfuland right now, the only way to have a beast nk.bin is to take the drive out, or get one from somebody who has :/
23:49:11saratoga2how odd that they would do that but not give a link to replacement nk.bin files
23:49:42gevaertsI think we should look at actual "normal" FAT undelete. If that works it's likely to be more reliable than current beastdump
23:49:47amiconnNot the only oddity in the beast's OF...
23:50:02saratoga2gevaerts: will they work on TFAT?
23:50:15Unhelpfuli agree, there. the process is clearly designed to be relatively user-friendly, it has nice on-screen prompts that detail what it's doing, and what it needs you to do
23:50:16gevaertssaratoga2: "if that works" :)
23:50:19toffe82there is a MTP function which seems to be supported by the beast, it is copy, I tried to use it to copy the nk.bin to the other partition but couldn't do it or perhaps didn't understand really how it works
23:50:50saratoga2beastdump works if the file is intact and continous, I'm curious when a regular undelete utility would do any better for TFAT?
23:51:30Unhelpfultoffe82: no good, really, since MTP doesn't expose the "true" filesystem, but rather a collection of file objects
23:51:50gevaertssaratoga2: I don't know. Another way would be storing md5sums of all disk blocks (clusters?) of nk.bin, and scanning the entire firmware partition
23:51:51Unhelpfulbeast doesn't even store them in the folders that it tells you you're creating or copying to
23:52:08Unhelpfulgevaerts: that seems safest.
23:52:22gevaertsUnhelpful: the problem is that we're talking about 6000 or so blocks
23:52:29toffe82Unhelpful: but if you know the id of the files it could work ?
23:52:37saratoga2gevaerts: that could be done
23:52:48saratoga2but i figured it wasn't worthwhile
23:52:55Unhelpfultoffe82: not all disk files are MTP files.
23:53:26saratoga2i think that would only help if you installed rockbox, uninstalled using your saved file, then lost it and finally decided to reinstall
23:53:29Unhelpfulthe beast in OF exposes a music and a pictures folder, and an albums folder, and not a whole lot else.
23:53:48 Join new [0] (n=58f417e5@gateway/web/cgi-irc/
23:53:49toffe82yes, but the firmware is transfered by mtp and have an ID
23:53:50 Join solexx [0] (
23:53:52gevaertssaratoga2: it all depends on how many people reinstalled their OF from whatever source and then lost their nk.bin file
23:53:53saratoga2i'm content for now to simply tell people to not lose the nk.bin if they want to reinstall
23:53:56newhi there
23:54:07newi need some help could you help me?
23:54:24Unhelpfultoffe82: when you send it a new firmware file, it is a new mtp file, yes, but the installed firmware is not exposed as an mtp file
23:54:29saratoga2and anyway, if someone lost the nk.bin they could always download a new one using google
23:54:45saratoga2what matters is that we have a legit way to do it so that we don't tell people to distribute nk.bin files
23:54:59saratoga2if other people want to do that then its there business as far as I'm concerned
23:55:09gevaertsnew: it would help if we knew what sort of help :)
23:55:59newi installed rockbox, it works but when i join games or applications like metronome i can't exit
23:56:15newi'm reinstalling right now
23:56:31Aurix_Lexicowhat player do you have?
23:56:40Unhelpfulthere aren't actual "music" or "pictures" folders on disk at all, either, those are just names for collection objects in the mtp object hierarchy. the beast version stores all MTP objects under /contents ore such, in folders with hex names that i would imagine are partial file hashes or similar
23:56:52newipod nano 2 GB
23:56:58*amiconn points new to the fine rockbox manual
23:58:59newi tried again

Previous day | Next day