Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2008-12-21

00:00:12amiconn...actually call grey_ub_gray_bitmap_part() internally, with the buffer as the source
00:00:55amiconngrey_show() is effictively no-op if the state doesn't change
00:01:48Unhelpfuli think nearly everything but the menu is going to need to be greylib... unless we want to change the look of the tracklist, and get rid of the gradient effect, etc, anyway.
00:01:57 Quit mcuelenaere (Read error: 110 (Connection timed out))
00:02:39*amiconn never saw pictureflow in action yet
00:03:07 Part kremlin
00:03:10amiconnAll I know is that it throws a misleading error message if there are no cover bitmaps at all
00:03:18amiconn(like on my targets)
00:03:21Unhelpfuli looked at how it works. the issue is that i need to be able to mix direct writes to a buffer with lcd_* draws to it, and do all of that painting *before* it goes to screen. buffered mode with direct buffer access pretty much covers that need
00:03:47Unhelpfuldid you try the test greylib bitmap viewer?
00:03:48kugelamiconn: I believe that was fixed recently
00:04:18amiconnNot yet - didn't have time for that today
00:04:39*amiconn hopes to be able to try it in the next few days
00:05:32 Quit bmbl ("Woah!")
00:06:26Unhelpfulno hurry, i'm reasonably confident it will work on hardware pretty much the same as in sim. nothing i've written is very close to the hardware, the only target-only problem i can recall seeing was that alignment issue in the scaler
00:07:59amiconnDoes it use the audio buffer, or only the plugin buffer?
00:08:29saratogadidn't the database used to require a reboot to enable?
00:08:29 Quit QUICKSTART (Read error: 104 (Connection reset by peer))
00:08:51Unhelpfulthe test loader, or pictureflow?
00:08:57 Quit bertrik (Remote closed the connection)
00:09:26Unhelpfulthe test loader uses a static buffer, large enough to load an LCD-size image + the scaler overhead
00:09:41 Quit {phoenix} (Remote closed the connection)
00:10:11Unhelpfulpictureflow as i'm working on it uses static buffers to load/scale images, but still loads the actual display bitmaps into the audio buffer.
00:12:28amiconnHmm, there's an interesting question: 2bpp greyscale targets have the scaler in the core, and the greylib output function plugs into it. But what about the mono targets with greylib?
00:13:07*amiconn suggested the pluggability, but now tends to think that it would be better to compile the scaler twice, once for the core, and once for the greylib
00:13:38 Quit BigBambi (Read error: 113 (No route to host))
00:14:30amiconnThe question is where to put code that is shared between core and plugins so that it is obvious
00:14:36Unhelpfuli actually don't think we should do that, since calling a function per output row actually saves code on color targets, by sharing the buffer->lcd color conversion between the two scalers
00:15:06 Join BigBambi [0] (
00:15:21amiconnYeah, but mono targets don't have a scaler in the core, yet we might want to use it in plugins
00:15:33bluebrotherdomonoky: tagged. Building w32 and linux binaries right now
00:16:15Unhelpfulyes, that's where things get tricky. one thought i had on that was that we pass a scaler function to the loader. we can just use NULL pretty much everywhere, and on grey/color targets it can use the core scaler if passed NULL
00:16:26amiconnThere is in fact already a plugin that shares core code (which incidentally isn't used in the core), but that one's nasty
00:16:31Unhelpfulon mono targets, the scaler can be in pluginlib, and we can pass it to the loader
00:18:07Unhelpfulor, we could build the loader+scaler in pluginlib, on mono targets. i have no idea how best to handle that.
00:18:52amiconnWhy 2 scalers, btw?
00:20:06Unhelpfulfor up vs down. they're different algorithms, each only useful going one direction
00:20:42amiconnHmm, same probably applies to the greyscale scalers too, then?
00:20:43 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
00:21:20Unhelpfulthe greyscale scalers are exactly the same scalers, except that they operate on uint32_t values instead of uint32_rgb
00:22:03Unhelpfulit's even cleaner in the vertical scalers, which were already just summing or multiplying whole rows, and doing it by working with uint32_t and tripling the width
00:22:33Unhelpfulthe vertical scalers just have the for() ifdef'd for the width
00:23:00 Part robin0800 ("Konversation terminated!")
00:23:07Unhelpfuldinner time here. i'll be back later, but probably not for at least a couple of hours.
00:28:22*hillshum just noticed /tools/genlang has no license info
00:28:41hillshumonly (c)Daniel Stenburg
00:29:48 Join einhirn [0] (
00:35:02bluebrotherBagder: the rbutil binaries are at
00:35:49bluebrotherplease upload to the download server (maybe not before Rockbox 3.1 is available?)
00:36:49bluebrotherunfortunately the w32 binary got somewhat bigger ... seems switching to Qt 4.4.3 (instead of 4.4.0 which is still used for the linux binary) caused that :/
00:40:58*domonoky starts the macos rbutil build.. :-)
00:41:10kugeldomonoky: reducing the lcd clock removed the sporadic int_btn == 0
00:41:39***Saving seen data "./dancer.seen"
00:41:59*bluebrother curses this stupid webspace not allowing to use +Indexes
00:42:54domonokykugel: the datasheet says something about not using the dbop read command, while the dbop fifo isnt empty (from a lcd dbop write) maybe check fifo status before trying buttons might help ?
00:43:18kugeldomonoky: where? I didn't read that.
00:43:41kugeldomonoky: although I read that "read data valid" needs to be polled before, and I do that
00:44:12kugeldomonoky: but, the nop loop should be enough to get the fifo empty, but I'll look into it
00:48:09domonokykugel: in the middle of page 90
00:51:30fdinelkugel & domonoky: so how is it going with the buttons? :P
00:51:56fdinelhow come?
00:52:20kugelfdinel: I can only get 3 buttons out of it, power, hold and down
00:52:28fdinelfrom DBOP?
00:53:44kugelfdinel: this is the c code I made of it
00:54:00 Quit lasser (Read error: 60 (Operation timed out))
00:54:22 Quit Horscht ("")
00:54:34domonokyfdinel: i would be much more interessted in details for the wheel... ( there has to be some enable pin or alike).
00:54:35kugeldomonoky: no change
00:54:58kugelyea, I suspect we're lacking a proper init func
00:55:12 Quit MethoS- (Remote closed the connection)
00:55:19kugelas we can't even read the buttons over dbop we get fine with direct gpio
00:56:49kugeldomonoky: and you surely don't get the hold button read? that's surprising me
00:57:08domonokykugel: yes, no sign of the hold button.
00:58:03kugelfdinel: maybe we set up the lcd wrongly. Especially our dbop_init does less than the OF's version
00:59:03fdineljust to be sure, are you both on the fuze?
00:59:09kugelfdinel: which version are you disassembling?
00:59:32kugelI'm 1.01.15
00:59:40fdinel1.1.11 I think
01:00:01fdinelwill you guys be around in about an hour?
01:00:10kugelhm, so offset
01:00:35kugel's are unlikely the same
01:00:35kugelfdinel: yes, if you tell me it's worth it? ;)
01:00:35domonokyfdinel: no, i am on e200v2. But it should be the same.
01:00:35fdinelcause I gotta go for about an hour
01:00:53fdinelwell it depends on what you think is worth it :P
01:01:02fdinelI'll be back later then
01:01:04kugelfdinel: heh, did you read the latest forum post?
01:02:04fdinelwell I missed the last 6-7 posts, I'll read when back, see ya later :)
01:02:34 Join wa4chq [0] (
01:02:56*kugel is curious if this guy is really willing to donate a fuze
01:03:57 Quit ZincAlloy ("CGI:IRC (EOF)")
01:06:10wa4chqhey all....yesterday I discovered rockbox....was given a sansa e260r and couldn't get much to work....rockbox solved that....great piece of work
01:09:09advcomp2019wa4chq, yea rockbox is great
01:10:09*hillshum wishes he had a working target
01:10:47 Join Hillshum_ [0] (
01:12:46wa4chqout of the box, I could not get stuff to pictures, no vids and barely tunes....rockbox is my christmas present come early!
01:16:07Llorean_Auron_: Do you mean "Do you really have an e260r" or "What is an e260r" or what? It helps if you include other words...
01:16:34 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
01:16:42domonokybluebrother: little problem with your changes to utils.cpp, you used statfs for mac, and didnt include a header for it. and on the mac i am now, there is only statvfs (like we use on linux).
01:19:14 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
01:22:27_Auron_what is that model
01:23:44Llorean_Auron_: It's a Sansa e200 series, but they come preloaded with a different firmware that makes installation a bit more complicate.
01:23:46advcomp2019_Auron_, an e200R series is the e200 series with rhapsody firmware
01:25:37Hillshum_seems to me the R has a bit more hardware too
01:25:38saratogakugel: I read it as "I have a fuze, let me know when I can run rockbox"
01:25:44saratogawas just about to delete it actually
01:27:08 Quit _Auron_ (Read error: 131 (Connection reset by peer))
01:28:00 Join _Auron_ [0] (
01:28:17wa4chqits the e200 series like advcomp2019 said...on the back it say e260 with a big R..
01:29:17wa4chqfront of unit has e200 written on it
01:30:00mooswa4chq: this big "R" might be for Refurbished...
01:30:08wa4chqi take that back...the front does not have the model number...just the back...
01:30:25amiconnlinuxstb: ping
01:31:34wa4chqmaybe, but the R is not screen printed like the e260 is...
01:32:00 Quit freddy__ (Remote closed the connection)
01:32:09LloreanHillshum_: What are you basing "it has more hardware" on?
01:32:41Hillshum_what i seem to recall reading some months ago
01:33:05LloreanYou said "it seems to me" so I thought perhaps you'd actually looked.
01:33:29linuxstbamiconn: Hi
01:33:37kugelsaratoga: I read "disposal"
01:33:46kugeland that's more than 1 player
01:34:04kugelso I read it as: I'm willing to test, and give 1 or 2 as a donation
01:34:15saratoga"at my disposal" is a saying for "available for my use"
01:34:15amiconnlinuxstb: While fixing 8 bit support in libdemac, I found something fishy in decoder.c
01:34:25wa4chqhave a merry christmas all...later
01:34:31 Quit wa4chq ("leaving")
01:34:38LloreanHillshum_: Where did you read that the hardware is different?
01:35:02saratogaHIllshum: its not different, which is why you can convert back and forth between them
01:35:08amiconnFrames flagged APE_FRAMECODE_STEREO_SILENCE are handled different for mono than they are handled for stereo. To me this looks like one of them must be wrong
01:35:23kugelsaratoga: Hm ok, I'm no native english, but I knew it as something with "giving away" (and kinda agreed to that)
01:35:37Hillshum_Not sure. May have been somewhere on, maybe abi, maybe sansa forums\
01:35:42saratogakugel: yeah its odd phrasing
01:35:53amiconnFor the mono case, entropy_decode() is called before the 'return 0', for stereo it's not called
01:36:03LloreanHillshum_: It'd be handy if you don't tell people random things you've read somewhere else, especially if you can't really cite them. It leads to misinformation that leads to trouble down the line.
01:36:04kugelok :) But hey, if we ask him nicely :)
01:36:26kugeli bet a fuze or e200 would be useful in the hands of fdinel or another dev
01:36:30saratogaoh i just deleted his post, want it back?
01:36:57*Hillshum_ was partially making sure his info was correct
01:37:14kugelsince I think as long as I'm alone (you always seem to busy with other stuff) on that we don't get much further, at least not in a appealing time frame
01:37:31 Quit aneqrs ()
01:37:34LloreanHillshum_: If you're asking, end the sentence with a question mark. That's how you let other people know you're not trying to tell instead. :)
01:37:36kugelsaratoga: no it's ok, you probably understood that better
01:38:49LloreanAre we really expecting the 2mb of RAM on the Clip/AMS c200 to be a significant problem at this point?
01:38:54 Quit Galois (Remote closed the connection)
01:39:00*amiconn has 8 bit working, but not commited yet
01:39:15LloreanI thought Zagor's test showed we could play with _really_ small buffers now, which basically means the RAM is most likely to only get in the way for Voice, I think?
01:39:19kugelLlorean: I don't
01:39:31kugelhence my answer in the forum
01:39:32Lloreankugel: You make it sound in your post like you think the c200's color screen will be a hurdle
01:40:13linuxstbamiconn: Yes, that does look wrong... Have you tested a silent file?
01:40:21kugela hurdle to be mastered
01:40:34linuxstbamiconn: I'll try it now.
01:40:37amiconnThat's a good idea I think. Will try...
01:41:00Lloreankugel: I don't think it's a problem at all. It's less than 30k for it, right?
01:41:03amiconnThe else {} part can be resolved, that's not a problem
01:41:17kugelI meant that a color screen is more memory demanding than a greyscale/mono one, but not that it'll rockbox unable to run just as fine
01:41:31kugeli have no numbers
01:41:36Hillshum_thats what i read it to mean
01:42:02 Join domonoky1 [0] (
01:42:04Lloreankugel: Well, you've described it as a "problem" though
01:43:04saratogai don't see much sense speculating now
01:43:07linuxstbamiconn: Hmm, looks like I already had some silent files as part of my demac test suite - and they decode OK.
01:43:14LloreanI'd almost say "The c200 has a color screen that will require a little more room, but there's no reason this should be a problem."
01:43:24*Hillshum_ agrees
01:43:42Lloreansaratoga: I can free enough RAM on my clip for the c200's screen buffer if it's 100k or less, so I don't really see it as speculating. :)
01:44:05linuxstbamiconn: (just using the standalone demac - with its crc checking of the decoded data)
01:44:14kugelI think in context with what I said above it's clear what I mean
01:44:44Lloreankugel: I think from the fact that it wasn't clear to me, that it's obviously not entirely clear.
01:45:05linuxstbLlorean: It's not just the screen buffer. It will be other things like wps buffer, backdrop buffers (there are two), plus embedded bitmaps (rockbox logo, usb screen logo)
01:45:59Lloreanlinuxstb: Even given all of those, we can free nearly 400k of compressed audio buffer and still have playback, so I wouldn't call it a "problem"
01:46:03kugelLlorean: Ok, I rephrased it a bit, should be better now
01:46:27amiconnlinuxstb: Those code paths are simply never taken...
01:46:31 Join Galois [0] (
01:46:54bluebrotherdomonoky1: interesting. Unfortunately I can't test on mac at all −− I used a BSD VM to try building and it worked fine
01:47:02kugelLlorean: but I really meant it to not be a major problem, seems I phrased it unclear
01:47:13bluebrotherbut if OS X supports statvfs I don't see a reason why not to use it
01:47:24Lloreankugel: What feature restrictions do you see as "likely" then?
01:48:26kugelsome plugins maybe, nothing in the really
01:48:26 Join ghindo [0] (
01:48:31kugelin the core
01:48:50linuxstbamiconn: Which code path(s)? The stereo version of APE_FRAMECODE_STEREO_SILENCE ?
01:48:58kugelI haven't thought too much, might be that there's no restriction in features at all
01:49:13Lloreankugel: Then maybe you shouldn't call them "likely" if you don't even know what might be impacted...
01:49:36kugelLlorean: and again, it's speculation
01:49:57kugelnothing I wrote is set in stone
01:49:58Llorean"Likely" is not a very speculative word.
01:49:59amiconnlinuxstb: Both. I added a printf("silence\n"); to them (quick hack, only works for standalone demac). It's never printed for both stereo and mono silent files
01:50:14kugelLlorean: it's one of the most speculative words I know
01:50:25Lloreankugel: If you say something is "likely" in English it means you think the odds are in favour of it happening
01:50:32LloreanIt doesn't mean the same as "are possible" or "may happen"
01:50:40*Hillshum_ agrees
01:50:41LloreanIt means "I think it will happen, but am leaving room in case it doesn't."
01:51:32kugeland there's always speculation in it
01:51:34 Quit frinkazoid (Remote closed the connection)
01:51:47BigBambilikely means it'll happen bar some odd events
01:51:47domonoky1Bagder: rbutil mac binarys at:
01:52:06Lloreankugel: Yes, but "likely" indicates you have some information affecting the odds.
01:52:06BigBambiNot much speculation involved in my opinion, just that you are not absolutley sure
01:52:11 Quit jhulst (Remote closed the connection)
01:52:53kugelLlorean: I expect at least jpeg viewer and pictureflow to not run as nicely as on a highmem target
01:53:35Lloreankugel: They don't run as nicely on a 16mb target as 64mb target, in that sense. At least, jpegviewer doesn't.
01:54:08kugeland they'll run even worse on a 2MB target
01:54:27kugelnot "worse" as in "bad", just as in "not as fine"
01:54:39linuxstbamiconn: It's printed for me for stereo silence - both APE_FRAMECODE_PSEUDO_STEREO and APE_FRAMECODE_STEREO_SILENCE are set, so the first (correct) implementation is used - i.e. the one that calls entropy_decode()
01:55:05Lloreankugel: Would you call Rockbox on the iPod 5G 30gb "Feature restricted" compared to on the 60gb then?
01:55:13amiconnEh? How did you create the silence files?
01:55:44linuxstbI can't remember - they've been sitting on my hard disk since I developed demac.
01:55:47kugelLlorean: if he can open bigger jpegs and/or zoom deeper, then yes. Not that it matters
01:55:47Lloreankugel: If the plugins actually do what they're supposed to do (view as large an image as the player has RAM for, for jpeg, and whatever pictureflow does for it) Rockbox isn't "Feature restricted"
01:55:51amiconnI think the second implementation will be triggered if one of the channels is silent - and then things go very wrong...
01:55:59 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
01:56:46*amiconn curses Audacity
01:57:03*Llorean thinks calling Rockbox "feature restricted" based on plugins is silly anyway.
01:57:19LloreanPlugins are add-ons, not really part of the "featurset" to me, at least.
01:57:38 Quit domonoky (Read error: 113 (No route to host))
01:57:40kugelAnd to me they're part of the featureset
01:57:58LloreanPlugins are separate applications, with their own features, though.
01:57:58bluebrotherplugins don't extend Rockbox itself ...
01:58:01saratoganeeds less rhetoricla arguments
01:58:05 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
01:58:06LloreanIt's like saying "Doom" is part of the "feature set" of Windows XP
01:58:16bluebrotheras in they are run from the "OS Rockbox".
01:58:24bluebrotherbut they don't change the OS
01:58:43amiconnThe 'silence' isn't actually silent...
01:58:47_Auron_Llorean: in the case of XP, don't you mean solitaire? :P
01:58:48LloreanAs long as all the core features of Rockbox work, I wouldn't say it's "feature restricted", but some of its plugins may be feature restricted, sure.
02:00:44 Join toffe82 [0] (
02:02:33linuxstbamiconn: I normally use "dd if=/dev/zero ..." and then sox to convert that to wav.
02:06:21 Quit bluebrother ("leaving")
02:06:36Hillshum_kugel: "restrictions (no that matter though" that no should be none
02:08:20amiconnlinuxstb: It also decodes ok if I remove the entropy_decode() call. COuld you verify?
02:10:57 Quit jhulst (Remote closed the connection)
02:11:16 Quit Hillshum_ ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
02:11:25linuxstbamiconn: Yes, I'll test now. How big is your file?
02:11:52amiconn428 bytes (30 seconds of 16 bit stereo silence)
02:12:10amiconnThe mono version has exactly the same compressed size, btw
02:12:46linuxstbIt gives me a segmentation fault - a 17MB wav file
02:13:18 Join EspeonEefi [0] (
02:14:06 Join JdGordon_ [0] (n=jonno@rockbox/developer/JdGordon)
02:14:08linuxstbamiconn: This is my test file -">
02:15:05amiconnI made a test file with silence in just one channel, and tried it with demac.exe. That does indeed trigger the second "silence" condition, and yields crc errors
02:17:28linuxstbamiconn: That's strange - looking at entropy_decode(), it memsets both channels to zero if the APE_FRAMECODE_STEREO_SILENCE flag is set - so if that flag is set when only one channel is silent, the bug is bigger...
02:17:33amiconnlinuxstb: Hmm, it does indeed produce a segfault, just that the windows cmdline doesn't show it
02:17:53 Quit planetbeing (Read error: 104 (Connection reset by peer))
02:17:59amiconnAPE_FRAMECODE_STEREO_SILENCE is actually two flags combined
02:18:31amiconnThe original implementation has separate left silence and right silence values
02:19:02kugelfdinel: re?
02:19:33amiconnlinuxstb: With the call to entropy_decode() it doesn't segfault
02:20:17amiconnThose silence handling bugs seem to be non trivial - I should probably commit the 8 bit handling fixes first
02:20:36 Quit JdGordon (Read error: 110 (Connection timed out))
02:20:51linuxstbamiconn: Looking at the original MAC source, the problem is that (as you say) the STEREO_SILENCE flag is actually LEFT_SILENCE (1) and RIGHT_SILENCE (2) combined.
02:20:51 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
02:21:21 Quit hillshum ("Leaving")
02:23:44 Join planetbeing [0] (
02:23:48linuxstbamiconn: Does your file just have one of those two silence bits set?
02:24:01amiconnIt should
02:24:24amiconnIt's basically your standard 16 bit test file, with the right channel silenced (using Audacity)
02:25:25linuxstbI ask because I'm not sure if the original MAC decoder deals with that case.
02:26:22amiconnIt definitely triggers the APE_FRAMECODE_STEREO_SILENCE check in the stereo case in demac
02:28:05 Quit einhirn (Read error: 54 (Connection reset by peer))
02:32:48 Quit saratoga ("CGI:IRC (EOF)")
02:32:49 Quit domonoky1 (Read error: 104 (Connection reset by peer))
02:33:07 Join JdGordon_ [0] (
02:33:58 Join itcheg [0] (i=62db4767@gateway/web/ajax/
02:35:23amiconnlinuxstb: I suspect another bug - pseudo-stereo (non silent) looks like it might not work as it should
02:36:03amiconnI hope to find some time to look into this tomorrow
02:36:49 Quit Schmogel (Read error: 104 (Connection reset by peer))
02:37:47linuxstbamiconn: Can you upload your current test track (one silent channel) somwhere?
02:41:40***Saving seen data "./dancer.seen"
02:45:29 Quit JdGordon (Read error: 110 (Connection timed out))
02:48:00kugellinuxstb: can you please have a look at (the mkamsboot patch)?
02:50:34 Quit ghindo (Client Quit)
02:54:23 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
02:56:48 Quit jhMikeS (Read error: 110 (Connection timed out))
02:58:22 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
02:58:50 Join agaffney_ [0] (n=agaffney@gentoo/developer/agaffney)
02:58:59 Quit agaffney (Connection reset by peer)
02:59:15 Nick agaffney_ is now known as agaffney (n=agaffney@gentoo/developer/agaffney)
03:02:29 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
03:12:17 Quit moos (Read error: 110 (Connection timed out))
03:22:09 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111319]")
03:23:01 Quit JdGordon (Read error: 54 (Connection reset by peer))
03:25:25 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
03:31:42 Join JdGordon_ [0] (
03:34:38 Quit JdGordon (Read error: 60 (Operation timed out))
03:40:26 Join kugel [0] (n=chatzill@unaffiliated/kugel)
03:52:53 Join QUICKSTART [0] (
03:56:15 Quit QUICKSTART (Remote closed the connection)
03:57:42 Join DogBoy [0] (n=john@unaffiliated/dogboy)
03:58:28daurnhi all, hows the new sansa work going?
04:03:23 Quit toffe82 (Read error: 110 (Connection timed out))
04:25:40 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
04:26:20Unhelpfulrasher: it uses greylib. it's semiworking! :D
04:26:40Unhelpfulit compiled, and runs, but the covers are all screwed up
04:26:43 Join blkhawk- [0] (
04:29:48Unhelpfuli also ended up defining a few macros that may be of use for plugins that want to be color-or-greylib, ie, that calculate either a brightness or a packed pixel value from an RGB triple
04:30:22Unhelpfulwith G_ and N_ for greylib or native colors, that are the same, of course, on color
04:34:21ameyerhrm, is sound failing to work on the clip for anyone else?
04:35:08*ameyer is using r19475, for what it's worth
04:35:28ameyerand this is on two separate 2GB clips
04:35:41 Join midgey [0] (n=tjross@
04:36:28ameyeryeah, I know I'm about 40 revisions behind
04:36:37*ameyer makes a new build to fix that
04:38:24 Quit JdGordon (Read error: 131 (Connection reset by peer))
04:41:27 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
04:41:41***Saving seen data "./dancer.seen"
04:42:20Lloreanameyer: Sound won't work on the stock SVN build for clip because the buffers need rearranged still.
04:42:46 Quit blkhawk (Read error: 110 (Connection timed out))
04:43:05ameyerI got the impression it used to work on stock svn
04:43:12ameyerwell, sort of
04:43:41 Nick blkhawk- is now known as blkhawk (
04:43:45ameyerwith the flash buffering patch, it seems to work as long as you're using vorbis
04:43:57LloreanIt worked before one of the larger features was compiled in.
04:44:15LloreanThe flash buffering patch is one way of rearranging the buffers (or rather, removing the need for them)
04:44:40ameyeralthough FLAC seems to fail big time with the flash buffering patch
04:44:51*ameyer suspects the sd driver is a bit too slow
04:45:07kugelfdinel: Hm, waiting for you was NOT worth it ;) Hopefully you have some more time tomorrow
04:47:06ameyerI also get the impression the flash buffering patch doesn't apply particularly cleanly
04:51:45 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
04:57:15 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:00:03midgeydo we have any native speakers of spanish or hebrew in the room?
05:07:43ameyerok, as far as I can tell, after you apply the flash buffering patch on the clip, vorbis and mp3 play, flac fails spectacularly, and the AAC codec fails to load.
05:08:04ameyersort of.
05:08:23Lloreanameyer: Comments about the patch should go in the task on the tracker.
05:08:55*ameyer checks to see if the AAC codec's even getting built
05:09:40ameyerthat's... strange.
05:11:00*ameyer assumes the aac codec "should" get built for the clip
05:12:58*ameyer notices it "doesn't" get built
05:14:32Lloreanameyer: Is that unpatched?
05:14:55ameyerpatched. I'm building an unpatched build to confirm
05:15:11LloreanCould you please make comments about the patched version on the task, as I asked already?
05:15:52ameyerI will.
05:16:23ameyeralthough I'm not sure any of that is related to the task, other than FLAC.
05:17:53LloreanWell, part of it is also "please don't fill the logs up with rambling about things that should just be a comment on the task." If you're using a patch, and you experience new strange behaviour, comment there so it gets to the authors.
05:18:12LloreanAnd, since the Clip is in-dev, the best place to talk about strange behaviour unpatched for it is in the forum thread.
05:18:55ameyeryou know what, screw this
05:18:57 Quit ameyer ("leaving")
05:30:39midgeybah, i made a mistake bringing the serbian lang up to date a few days ago, the strings with the new sources weren't considered
05:35:48 Join midgey_ [0] (n=tjross@
05:37:14 Join GodEater [0] (i=765a2d7f@gateway/web/ajax/
05:45:21 Quit amiconn (Read error: 60 (Operation timed out))
05:46:26 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
05:54:41 Quit midgey (Read error: 110 (Connection timed out))
06:01:13Unhelpfulanybody with a greyscale target and album art available to test #9689?
06:01:40 Quit pixelma (Read error: 110 (Connection timed out))
06:01:56 Join pixelma [0] (n=pixelma@rockbox/staff/pixelma)
06:03:30 Join Minthe [0] (
06:06:00 Quit itcheg (" ajax IRC Client")
06:08:48 Join itcheg [0] (i=62db4767@gateway/web/ajax/
06:14:04JdGordonmidgey_: hey, whats the story with plugin localisation stuff?
06:14:14*midgey_ hides
06:14:32midgey_i took on too large of a workload this semester
06:14:43midgey_i have to update the patch to the new build system
06:15:34midgey_i'll likely have some down time over the next few days so hopefully i can sync/update it by the end of december
06:15:43midgey_that's the plan at least
06:15:48JdGordonsounds good :)
06:16:01 Quit itcheg (" ajax IRC Client")
06:16:13JdGordoneven if you cant, upload what you have so someone can look at it if they have time
06:16:39midgey_i'll be without internet so i might be yelling at make while i attempt to sync it
06:17:04midgey_the latest patch on the tracker is mostly up to date, i think i made a few changes after the deadline
06:31:36 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
06:36:55 Join toffe82 [0] (
06:37:10 Join massiveH [0] (
06:37:25 Quit massiveH (Read error: 104 (Connection reset by peer))
06:41:42***Saving seen data "./dancer.seen"
06:43:18 Quit mc2739 ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
06:46:44 Quit Aurix_Lexico (Remote closed the connection)
06:47:13 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
06:47:34 Quit jhulst (Read error: 104 (Connection reset by peer))
06:50:56 Part toffe82
06:52:33 Quit XavierGr ()
06:53:46midgey_well hopefully my fiasco with srpski.lang is finally over
06:54:06 Quit GodEater (" ajax IRC Client")
06:54:46midgey_i'm heading off for a week. if i broke something, someone please fix it and slap me around when i get back
06:57:02 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
07:00:07 Quit midgey_ ()
07:00:09 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
07:06:25 Nick Minthe is now known as xsloader (
07:07:15 Join ameyer [0] (
07:22:39lucentameyer: hi, I had owned a 2gb Clip
07:22:48lucentameyer: what format files did you attempt to play on it?
07:27:25ameyerWith the flash_buffering patch, I've tried vorbis, mp3, flac, and AAC. As I understand it (thanks Llorean), the AAC codec doesn't get compiled for a target unless the memory size is >2MB, so AAC failed. FLAC died with some sort of SD RX buffer underrun. MP3 and Vorbis sort of worked. Without the flash_buffering patch, sound doesn't actually work. If my slow SD driver and/or buffering issue theory is correct, shn, wav, and w
07:27:58*ameyer is probably missing something and should take that to the patch tracker anyway
07:32:22 Quit EspeonEefi (
07:32:22 Quit FOAD (
07:32:22 Quit ender` (
07:32:22 Quit _lifeless (
07:32:22 Quit J-23 (
07:32:22 Quit goffa (
07:32:22 Quit DogBoy (
07:32:22 Quit shodanX (
07:32:22 Quit preglow (
07:32:22 Quit DaCapn (
07:32:22 Quit Zambezi (
07:32:22 Quit tim__b (
07:32:46NJoinDogBoy [0] (n=john@unaffiliated/dogboy)
07:32:46NJoinEspeonEefi [0] (
07:32:46NJoinFOAD [0] (
07:32:46NJoinender` [0] (
07:32:46NJoin_lifeless [0] (n=lifeless@
07:32:46NJoinJ-23 [0] (
07:32:46NJoingoffa [0] (n=goffa@
07:32:46NJoinpreglow [0] (
07:32:46NJoinDaCapn [0] (
07:32:46NJoinshodanX [0] (
07:32:46NJoinZambezi [0] (
07:32:46NJointim__b [0] (
07:35:03*ameyer should reload rockbox on one of his clips and try the flash_buffering patch with low-bitrate FLACs.
07:35:42Lssok i think i screwed my ipod by loading the wrong theme for it now i can see any words....
07:36:28Lsshumm nevermind its working now weird
07:36:37 Quit ender` (" I went to the cinema, and the prices were: Adults $5.00, children $2.50. So I said, "Give me two boys and a girl."")
07:39:05lucentameyer: it should be noted that presently, the Sansa AMS targets puke if you feed them data > 0.97gbyte
07:39:48lucentameyer: so be sure and like empty out the clip, do an OF format, then load rockbox, then load music and keep it < 0.5gb ish for testing I recommend
07:41:44ameyerI thought it only puked with >.97GByte for targets with > 2 GByte storage
07:42:05lucentjust my experiences
07:42:25lucentI forgot whom, but there was a report of no error with a 1GB Clip
07:42:41lucentsame test on my 2GB clip resulted in error
07:43:33ameyerI'm done with testing rockbox on my clip, honestly.
07:43:57lucentokay, sorry to bring that up then
07:44:10lucentI had 2GB Clip, and have an 8GB Fuze
07:44:29lucentshipped the Clip out to kugel maybe he can have some fun making it work :)
07:45:54ameyeractually, let me try that
07:47:19lucentameyer: also although it's a bit off-topic, unless you already know about it, there -is- a way to read the MTP files from OF on the fat partition, in case you need to back it up just copy the whole block device
07:48:09lucentyour songs saved from MTP mode (like the pre-loaded music) are in a funky directory that has "Directory" and "Volume Label" attributes set
07:48:21lucentso it gets ignored by most fat filesystem drivers
07:48:26lucentbut that's easy to work around
07:48:37lucentjust saying in case you didn't already know or needed to know
07:48:41ameyerare those files in folders that start with ##?
07:52:17ameyerprobably should throw a musepack file and a shorten file onto the clip while I'm testing your theory
07:52:44ameyeralthough codec type doesn't exactly fit into either of our theories
07:54:22lucentmp3 is known to cause a crash or reboot on Sansa AMS targets
07:57:30ameyerI don't think that's all Sansa AMS targets
07:57:43ameyeror at least the wiki makes it sound like a e200v2 thing
08:00:53 Quit jhulst_ (Remote closed the connection)
08:02:19ameyercrap, I deleted the bootloader file
08:05:48 Join MethoS- [0] (
08:08:45ameyerwell, there goes my theory, sort of.
08:10:12ameyerwav seems to "play", so it's obviously not just bitrate
08:10:16 Join Rob2222 [0] (
08:28:08 Quit Rob2223 (Read error: 110 (Connection timed out))
08:31:14xsloaderExcuse me, but is there a WPS tag to show mono or stereo?
08:41:46***Saving seen data "./dancer.seen"
08:44:02ameyerxsloader: if there is one, it's not documented.
08:45:00JdGordonxsloader: you can use the setting dislpay tag for that
08:45:03JdGordon%St iirc
08:45:31ameyerI think he means whether the file he's playing is stereo or mono
08:45:38xsloaderI checked the website but I could not find
08:45:46xsloaderstereo or mono.
08:45:55JdGordonumm... dunno
08:46:26ameyeralthough I shouldn't have assumed that xsloader is a he or assumed that xsloader didn't mean the value of the "channel configuration" setting
08:47:28xsloaderI mean the music files.
08:48:50ameyeras far as I can tell, it can't be done.
08:51:08xsloaderSo I cannot know whether I'm listening with mono channel though channel conf value is stereo.
08:52:45xsloader#If the music file is mono, channel conf value is ignored (treated as mono), right?
08:53:38 Quit slact ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
08:57:14ameyernot exactly. mono files would be played as "dual mono"
08:57:19 Join blithe_ [0] (
08:57:52ameyeralthough, for what it's worth, if you set the channel configuration value to mono, you get dual mono
08:58:10 Quit blithe_ (Client Quit)
09:23:37 Join bertrik [0] (
09:26:04 Quit BHSPitLappy (Remote closed the connection)
09:26:58 Quit xsloader ("Leaving...")
09:30:44 Quit ameyer ("leaving")
09:34:31 Join MethoS-- [0] (
09:50:24 Quit MethoS- (Read error: 113 (No route to host))
09:53:19 Quit MethoS-- (Connection timed out)
10:21:01 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
10:25:30 Join n1s [0] (n=nils@rockbox/developer/n1s)
10:32:16 Join Schmogel [0] (
10:33:38 Quit Schmogel (Client Quit)
10:35:28 Join Darksair [0] (n=user@
10:35:44 Join wthww [0] (
10:36:19wthwwI might have missed it on the site or wiki, but is there any sort of visualizer for rockbox?
10:37:48bertrikwthww, not that I know of
10:38:09bertrikyou could consider the oscilloscope plugin as a visualiser
10:39:47wthwwah, yeah.
10:40:13wthwwI just picked up two used ipod videos, and really wanted to play flac on them, so, yay RockBox.
10:41:04bertrikI think the emphasis of rockbox is mostly on audio playback. Fancy animated graphics drain the battery, especially if the lcd backlight needs to be kept enabled.
10:41:36wthwwYeah, I was just wondering :)
10:41:50***Saving seen data "./dancer.seen"
10:41:51wthwwI know flac playback drains it too, but i think its worth it
10:43:34Lssi think its rockbox that drains it
10:43:57Lssisnt flac playback the most optimised lossless format for rockbox?
10:49:30 Quit BigBambi (Read error: 104 (Connection reset by peer))
10:52:43 Quit bertrik (Remote closed the connection)
10:56:11 Join bertrik [0] (
11:01:05 Quit bertrik (Client Quit)
11:03:40 Join __lifeless [0] (n=lifeless@
11:06:03JdGordonwhat do people think about saving the filename of the current track in the playlist control file, so if when you resume a dirplay (and hopefully, eventually DB) playlist, if files are deleted it can try and find the correct file instead of starting the wrong track?
11:07:23Lssim not getting it if its deleted whats there to try finding?
11:07:48 Join {phoenix} [0] (
11:09:28JdGordonwell there is 2 problems, first if you delete a track that was before the current track it will resume the track after current.. and the 2nd problem is obviously not resuming the corect track if the current one was deleted (and not being able to know that that happened)
11:09:32*JdGordon back in 10
11:12:51 Quit Horscht ("Snak 5.3.3 Unregistered copy. Evaluation period is over. Program will now quit. Thanks for using Snak.")
11:13:03 Join bertrik [0] (
11:18:51 Quit _lifeless (Read error: 110 (Connection timed out))
11:19:04 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
11:23:25 Quit Horscht (Client Quit)
11:28:05 Quit __lifeless (Remote closed the connection)
11:28:24JdGordonalso, anyone wanna comment on fs#9680? "loading a empty .cfg takes a while"... the fix is simple and tiny, but how often would somone load a empty .cfg?
11:33:17 Join sbooz [0] (
11:33:30sboozwow, seriousness going on in #underthebridge
11:33:38sboozspecial guest star... Sky__
11:33:41sboozact now
11:33:48 Part sbooz
11:35:37 Join MethoS- [0] (
11:37:09n1sJdGordon: i think fixing the resume after delete thing would be nice, since i've heard that behaviour reported several times before, fs#9680 seems like a very unusual situation but i don't feel strongly about it :)
11:39:00JdGordonI'd like to move playlist resume completly into the .playlist_control file so the filename can be stored easily.. but after a while the playlist would get heaps of these save lines...
11:40:22JdGordon... which isnt really a problem because only the last one would be used.. but still not very nice
11:41:37JdGordonespecially if bookmarks ever get merged with the playlist control like I'd like, then there will be "bookmarks" which were not actually created by the user
11:42:07 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
11:42:45 Join _lifeless [0] (n=lifeless@
11:44:57 Join moos [0] (i=Mustapha@rockbox/staff/moos)
11:48:51 Quit _lifeless (Remote closed the connection)
11:53:31 Quit JdGordon (Remote closed the connection)
11:53:53bertrikn1s, we discussed some si470x methods some time ago. What method gave you the best tuning performance on the gigabeast?
11:54:53 Quit Darksair (Read error: 104 (Connection reset by peer))
11:55:21 Join JdGordon [0] (
11:56:58 Join _lifeless [0] (n=lifeless@
11:57:52 Join cagnulein [0] (
11:58:02 Quit cagnulein (Client Quit)
11:58:32 Join stripwax [0] (
11:59:43 Join Jaykay [0] (
12:01:28Jaykayi get an error when i try to attach a file in the sansaruntime page in the wiki
12:01:36 Join einhirn [0] (
12:01:52Jaykayor better, i dont get an error, the address of the empty page is;def=topic_access;param1=change;param2=access%20not%20allowed%20on%20web
12:02:12JdGordondid you log in?
12:02:40Jaykayi tried to, i get it after this login-message where i should put my account name and passwort
12:03:57JdGordonwhat is your wiki name?
12:04:19JdGordonhas someone fixed you up so you can edit?
12:04:24Jaykayi got it one time already after my registration, but i dont know how i solved it
12:04:36Jaykayfixed you up?
12:04:52Jaykaybut yes i can edit
12:05:05JdGordonok, then i dunno :p
12:09:21JaykayJdGordon: can i attach files to emails to the dev-mailinglist?
12:09:43JdGordonyeah, you should be able to
12:10:02Jaykayok ill try
12:15:09 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
12:23:05n1sbertrik: the loop i pasted worked well, didn't try your variant that set the TUNE bit in the loop, can do if you want
12:24:27bertrikn1s, yes please try that. It's how the clip OF does it and seems to work fine in rockbox too.
12:25:11bertrikthis is how I currently implemented it:
12:26:47 Part wthww ("Leaving")
12:37:11 Quit JdGordon (Read error: 131 (Connection reset by peer))
12:41:52***Saving seen data "./dancer.seen"
12:47:29 Quit HBK (Read error: 110 (Connection timed out))
12:48:24 Join HBK [0] (
12:50:00n1sbertrik: both variants work for me but your version is much slower
12:50:29n1salso i now managed to get hangs with both versions so a kind of max iteration check is probably necessary
12:50:45n1ssince using that got rid of the hangs
12:52:15 Quit J-23 (Remote closed the connection)
12:52:35 Join J-23 [0] (
12:59:38 Quit stripwax (Read error: 54 (Connection reset by peer))
13:00:36Jaykayis it right that its not possible with rockbox to create files?
13:00:59scorchewhat kind of files?
13:01:17Jaykaysimply a file in the file browser
13:01:33Jaykaynot a specific one with the text editor or recorder
13:01:45Jaykaymaybe abc.xy
13:01:48scorchethen no...i dont see why one would want to do that...
13:02:03Jaykayi want to create a empty .wps file
13:02:10 Join stripwax [0] (
13:02:14 Quit n1s (Remote closed the connection)
13:02:18Jaykayand have to connect the player to the computer
13:02:35Jaykayits not much work but with rockbox it would be easier :)
13:02:56scorcheand why dont you use the text editor for this?
13:03:22Jaykayhm.. i could use it
13:03:31Jaykaybut it could be done easier
13:06:25pixelmait was noticed recently that a completely empty .wps will result in the WPS parser rejecting it and you
13:06:38pixelmayou'll get the built-in WPS
13:07:30 Join ilbozo [0] (n=5cec2cc4@gateway/web/cgi-irc/
13:07:38Jaykayi know
13:07:45Jaykayi was the one who noticed it ;)
13:08:04Jaykaybut i was wrong, i wanted to create a empty .cfg file
13:08:07Jaykaysorry :)
13:08:30 Join stoffel_ [0] (
13:08:52scorchewhy would you want to do that?
13:09:34Jaykaybecause of a "bug" i reported
13:09:40Jaykayit was not really a bug
13:09:59Jaykaybut a littleannoyance
13:11:49 Quit FOAD ("I'll be back")
13:12:08 Join FOAD [0] (
13:12:24 Quit stoffel_ (Client Quit)
13:13:28 Quit ilbozo ("CGI:IRC (EOF)")
13:18:48 Quit moos ("Rockbox rules the DAP world")
13:23:10 Join AndyI [0] (i=AndyI@
13:26:56 Join n1s [0] (n=nils@rockbox/developer/n1s)
13:33:13 Join lasser [0] (
13:34:25 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
13:35:21 Quit AndyIL (Read error: 110 (Connection timed out))
13:41:19amiconnlinuxstb: around?
13:46:43 Join gregzx [0] (
13:46:48 Join miepchen^schlaf [0] (
13:57:05 Quit lasser ("ChatZilla 0.9.84 [Iceweasel 3.0.4/2008112309]")
14:00:57bertrikn1s, sorry for the delay (was afk). On the clip, tuning does not seem that critical so let's use what works well on the gigabeat. The only thing I think is probably necessary on the clip is to leave in the 60 ms delay during tuning, to satisfy the erratum for tuning with use of the internal oscillator.
14:02:26n1sbertrik: ok, as i said i managed to get it to hang with both versions before i added a max iteration check so i think that's a good idea
14:04:33 Quit nplus (Remote closed the connection)
14:05:09*JdGordon wonders why gevaerts didnt rename ata_idle_notify.h when he did ata->storage rename?
14:05:21Jaykayis there any way to download the whole trunk-directory with svn from an older revision?
14:05:27Jaykaynamely r19396
14:05:40JdGordonadd a -r<revision> to the svn command
14:05:48JdGordonshould work for both svn up and svn co
14:06:49JaykayJdGordon: thanks
14:07:21Jaykaywith svn co all files are overwritten?
14:07:55n1sif you already have a checkout, "up" is better
14:12:03 Quit n1s ()
14:13:42 Join stoffel_ [0] (
14:22:27gevaertsJdGordon: good point...
14:23:14JdGordonrealised trying to resync an old patch :)
14:23:52*gevaerts finds an excuse
14:24:12gevaertsThat was only the first part of the storage rework. There's more to come :)
14:29:06 Join nplus [0] (n=nplus@141.25.Globcom.Net)
14:29:28 Quit stoffel_ ("leaving")
14:31:22 Quit domonoky (Read error: 104 (Connection reset by peer))
14:36:07 Quit {phoenix} (Remote closed the connection)
14:41:55***Saving seen data "./dancer.seen"
14:46:13 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
14:51:32 Join Horschti [0] (
14:51:46 Quit Horscht (Nick collision from services.)
15:00:04 Quit Zom (Read error: 110 (Connection timed out))
15:00:40 Join culture [0] (
15:03:30 Quit MethoS- (Remote closed the connection)
15:04:41 Quit Dieterbe ("leaving")
15:07:11 Join Dieterbe [0] (
15:08:13 Quit Dieterbe (Client Quit)
15:09:16 Join MethoS- [0] (
15:09:28 Join Dieterbe [0] (
15:09:33JdGordonanyone have any ideas how to remove a command from the playlist control file in the code?
15:10:05JdGordonthe best I can come up with is finding the command to remove and change the first letter to a comment...
15:30:46 Join ender` [0] (
15:30:59 Join Darksair [0] (n=user@
15:34:54 Join fredddy [0] (
15:36:19 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
15:40:10 Join MethoS-- [0] (
15:47:23 Quit MethoS- (Read error: 145 (Connection timed out))
15:49:14 Join kugel [0] (n=chatzill@unaffiliated/kugel)
15:50:31 Quit wpyh (Remote closed the connection)
16:01:11*gevaerts decides to go through the commits that were done on the 3.0 branch to see if there's anything that needs to be redone for 3.1
16:08:02 Join Nibbler [0] (
16:15:52 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
16:16:36 Quit fredddy (Read error: 110 (Connection timed out))
16:17:42 Join fredddy [0] (
16:30:27 Join Aurix_Lexico [0] (
16:33:08rasherJdGordon|zzz: Why didn't you commit r19523
16:33:19rasherJdGordon|zzz: Eh, why didn't you commit r19523 to the release branch?
16:33:39 Nick fxb__ is now known as fxb (
16:39:58 Join freddy__ [0] (
16:41:21 Join itcheg [0] (i=62db4767@gateway/web/ajax/
16:41:59***Saving seen data "./dancer.seen"
16:42:22 Quit fredddy (Read error: 60 (Operation timed out))
16:44:58 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
16:48:47 Join tyfoo [0] (
16:49:59 Quit tyfoo (Client Quit)
16:54:42 Join stoffel_ [0] (
16:55:38 Quit Darksair ("Everything that has a beginning has an end.")
17:01:32 Join ZincAlloy [0] (n=d9eedb12@gateway/web/cgi-irc/
17:03:06 Quit Richlv (Read error: 104 (Connection reset by peer))
17:06:33gevaertsWill the list of release-targets be the same for 3.1 as for 3.0?
17:08:01LloreanI believe so.
17:09:05 Quit EspeonEefi ("さよなら")
17:09:53 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
17:13:00 Join EspeonEefi [0] (i=eefi@STRATTON-TWO-EIGHTY.MIT.EDU)
17:14:04gevaertsWill there be new flash images for archos?
17:14:37LloreanThere ought to be, I think
17:15:31gevaertsOK. I'll update archos-flashing.tex accordingly then
17:23:17 Join fdinel [0] (
17:37:05 Join Zom [0] (
17:38:40 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
17:40:08 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
17:44:17 Join {phoenix} [0] (
17:44:21 Join __lifeless [0] (n=lifeless@
17:46:10 Quit _lifeless (Read error: 60 (Operation timed out))
17:47:04 Quit stRR (Read error: 104 (Connection reset by peer))
17:59:07 Quit itcheg (" ajax IRC Client")
18:01:03 Quit jhulst (Remote closed the connection)
18:05:20gevaertsrasher: yellow...
18:06:32kugelfdinel: hey
18:13:32 Quit ZincAlloy ("CGI:IRC (Ping timeout)")
18:14:15 Quit bertrik (Remote closed the connection)
18:16:10amiconnrasher: Ugh, loadsa yellow...
18:18:52gevaertsThe 3.1 manual should be in good shape now, i.e. it has all branch-specific changes that the 3.0 manual has, and fixes in 3.0 that somehow got overlooked in trunk are now merged back.
18:19:27gevaertsThe only differences between the trunk manual and 3.1 are (a) metronome for c200,player and ondio, and (b) version numbers
18:20:52 Part Llorean
18:24:36 Join itcheg [0] (i=62db4767@gateway/web/ajax/
18:25:33 Join Jaykay [0] (
18:26:00 Join massiveH [0] (
18:26:18 Quit EspeonEefi ("さよなら")
18:27:18 Quit massiveH (Client Quit)
18:28:09 Quit itcheg (Client Quit)
18:31:00 Join bertrik [0] (
18:40:01 Join bertrik_ [0] (
18:41:04*gevaerts forgot some differences
18:41:52gevaerts(c) removed fixmes, and (d) a different changelog (trunk has everything, 3.1 has only things relevant to released targets)
18:42:02***Saving seen data "./dancer.seen"
18:43:57 Quit stoffel_ ("leaving")
18:51:19 Quit bertrik (Read error: 113 (No route to host))
18:52:54 Quit einhirn (Read error: 54 (Connection reset by peer))
18:53:31 Join EspeonEefi [0] (i=eefi@STRATTON-TWO-EIGHTY.MIT.EDU)
19:00:03 Join aurix_lexic1 [0] (
19:08:45 Join tyfoo [0] (
19:13:23 Join Agneees [0] (
19:13:54*jhMikeS will wait for the "vented with flame" FS tasks now :P
19:18:55 Quit bertrik_ (Read error: 104 (Connection reset by peer))
19:20:44 Join bertrik [0] (
19:21:31kugeljhMikeS: so, what's missing for the beast to be supported?
19:21:54 Quit {phoenix} (Read error: 60 (Operation timed out))
19:22:03jhMikeSkugel: I don't know. Nothing major I can think of.
19:22:14Agneeeshi everyone
19:22:16jhMikeSInstallation is still though?
19:22:26kugelisn't the battery runtime rather low, is it?
19:23:15 Join {phoenix} [0] (
19:23:25jhMikeSIt is but PP targets were "supported" when abyssmal. No target that runs <20hrs can be "supported" now? :)
19:23:41AgneeesI'm just curious (didn't found this in FAQ): how long does it take when you start working on a new player? is it something like 6 months? 6 years? something in between?
19:23:56 Quit freddy__ (Read error: 60 (Operation timed out))
19:24:52kugeljhMikeS: yea, I don't think that's a showstopper, I was just curious
19:25:11kugelwell, then I think it's the installation, the dual boot in particular
19:25:19kugel"[To be completed - there is currently no source for the required original firmware =nk.bin=] "
19:25:31 Join freddy__ [0] (
19:25:45jhMikeSkugel: There isn't really any need for it to keep it running now.
19:26:26kugelwhat about uninstallation? If that's easy I think dual boot isn't critical
19:26:40lucentAgneees: not to be rude, but how long would it take you to add support? :)
19:26:59Agneeeswell, I'm trying to find what I could de
19:27:09lucentAgneees: which hardware?
19:27:47kugelAgneees: you can't estimate that, it depends on the hardware and how accessible it is (including hurdles due to the OF)
19:27:53Agneeessansa fuze
19:28:33lucentAgneees: we need maybe some donations to cover the cost of getting a fuze to interested developers, and it's not a promise
19:28:58Agneeesapparently, according to the only thing I could help with would be donate money!!
19:29:20jhMikeSkugel: It's a breeze if OF isn't needed. Just get a bootloader and use sendfirm. I see a simple windows version now too.
19:29:49scorchelucent: and if money is donated, nothing is promised that that money might go to such a purpose..
19:30:21lucentAgneees: yeah... that's the official line. If/When the target is considered a candidate for a supported target, a manual will need to be written
19:30:27kugeljhMikeS: I mean how's reverting from rockbox to the OF? I see that "TODO: How can a user uninstall Rockbox? We need a source for the original nk.bin..." which makes me think there's no way to
19:30:31Lsshumm i have some unicode tags and filenames that dont show up
19:30:32lucentso skills with LaTeX are a plus
19:30:36Lssis that normal?
19:30:50scorcheLss: are you using a unicode font?
19:31:03Lssi selected unicode
19:31:10Agneeeslucent: so, in the meantime, I can learn LaTeX!
19:31:23scorcheAgneees: it isnt tough =)
19:31:37jhMikeSkugel: If you have either the original nk.bin or can use the updater it's easy too.
19:31:52Agneeesscorche: at least, I already know lilypond
19:32:04kugelso uninstalling ain't a problem? If so, the comment is missleading
19:32:28AgneeesI'd help with the French translation, but apparently, it already taken care of
19:32:32jhMikeSNo, you can uninstall rockbox from rockbox if you want.
19:33:18kugelthat's cool
19:34:12jhMikeSit will still be on the HD though. but really it's just as easy as dumping nk.bin onto the small partition
19:34:23 Join lasser [0] (
19:35:31 Quit Xerion (" ")
19:42:58 Quit Nibbler (No route to host)
19:44:56 Join Xerion [0] (
19:46:49 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
19:49:12fdinelkugel: hey :)
19:50:04gevaertsI think there are two important issues to be solved with the beast. One is uninstallation, the other is that there is no manual.
19:50:38gevaertsThe manual issue is (according to precedent) not enough to call it unsupported, but it is enough not to release for it
19:51:02gevaertsI have no idea about the uninstallation issue
19:52:50lucenthi fdinel
19:53:46fdinellucent: hi!
19:56:30Lssi find the ui on the ipod too small how do i make the font size and icons larger?
19:56:56gevaertsHave you read the manual?
19:58:32Lsssort of i couldnt find anything on increasing font size though
19:59:36 Quit Aurix_Lexico (Read error: 110 (Connection timed out))
19:59:41gevaertsYou need to choose a different font. The font names include size information
20:00:31Lssbut there is only 1 unicode?
20:02:22 Quit culture (Read error: 60 (Operation timed out))
20:04:06Lssi see so it cant actually display kanji?
20:04:59gevaertsLss: for the complete story see">
20:05:59 Join mc2739 [0] (
20:11:00 Quit write__erase ("Leaving")
20:16:33 Quit mc2739 ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
20:24:22 Join karashata [0] (n=karashat@
20:27:09 Join mc2739 [0] (
20:28:48 Join Schmogel [0] (
20:39:33 Join massiveH [0] (
20:39:56 Quit massiveH (Client Quit)
20:42:05***Saving seen data "./dancer.seen"
20:48:31 Join stripwax [0] (
21:01:36 Join MarcelLotze [0] (n=548565a0@gateway/web/cgi-irc/
21:03:16MarcelLotzeGood evening, I wanted to ask if somebody can give me (Username MarcelLotze) write permission for the wiki, because I wanted to add some FM presets, as I mentioned long ago here:
21:07:43 Quit MarcelLotze ("CGI:IRC")
21:08:53jhMikeSbloody bools
21:13:16bertrikin fmradio_i2c.h ? ;)
21:21:48 Quit TMM (Remote closed the connection)
21:22:06 Join TMM [0] (
21:22:38kugelfdinel is a god!
21:23:03*scorche coughs
21:23:13kugelI got the wheel working!
21:23:18kugelon my fuze
21:29:58Unhelpfulgevaerts: unless there's a way to execute code within the firmware update, *before* the bootloader is written over the existing nk.bin, i don't see how we can make uninstall work
21:30:19Unhelpfulor unless there's a "real" gigabeat s recovery tool from which nk.bin can be extracted
21:31:53gevaertsUnhelpful: I'm not sure if we need to support uninstallability at all, but precedent says that there are 28 targets where we can do it and none where we can't, so full support without it will be a hard decision I think
21:32:51mc2739FS #9641 can be closed - code is now included in FS #9611
21:34:19kugelmc2739: please try and report back
21:35:05gevaertsmc2739: done
21:35:59mc2739kugel: trying now
21:36:17mc2739gevaerts: thanks
21:36:20kugelmc2739: copy button-fuze.c into button-e200.c after applying
21:36:24Unhelpfuludivdi3 and udivsi3 are... libgcc 64-bit and 32-bit unsigned multiply, respectively?
21:36:33mc2739kugel: ok
21:37:38kugelmc2739: i hope you saved the patch, I removed it
21:39:22mc2739kugel: yes, I have it saved.
21:39:59kugelso, baiscally only home button is missing
21:44:11 Join Juice^ [0] (
21:45:45 Join robin0800 [0] (
21:51:13 Quit karashata ("G'bye everyone!")
21:52:05 Join einhirn [0] (
21:52:11 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
22:02:30 Join ZincAlloy [0] (n=d9eedb12@gateway/web/cgi-irc/
22:03:19kugelmc2739: results?
22:08:24 Join Thundercloud [0] (
22:15:48fdinelkugel: thanks for the cheers :P
22:15:54fdinelgod, no less :D
22:16:21fdinelso, what did I miss? which buttons/wheel are you able to read now?
22:16:32kugeleven though, if I kept disassembling for the next few days, I'd probably gotten to that too
22:16:39kugelthe wheel
22:17:02kugelI can read all buttons combined with the direct gpio read, except for the home button
22:17:59Unhelpfulgevaerts: maybe somebody who knows "how" we fake the firmware install could say if a firmware update file could be used to dump the existing firmware...
22:18:09fdinelkugel: what does the wheel reading looks like? and you're telling that you still can,t read all the buttons using DBOP?
22:18:21kugelyes, I do
22:18:48kugelBits 9 and 10 of DBOP_DIN change as follows: Clockwise rotation 00 -> 01 -> 11 -> 10 -> 00
22:18:49gevaertsUnhelpful: we upload the firmware file over mtp, and it overwrites the original
22:19:03kugelother way around for counter-clockwise
22:19:09gevaerts(that or remove the disk and access it directly)
22:19:43fdinelthat's great :) and for the buttons, what is the mapping? same as we saw using GPIO?
22:19:50Unhelpfulso the overwrite happens before any chance to execute code? that's not helpful to us at all
22:19:56fdinel(C0 = bit0 and so on)
22:20:06kugelfdinel: what's great?
22:20:28kugelI can now read the wheel, hold and power (down not anymore for some reason) with dbop
22:21:22kugelfdinel: do you spot another function with just sets a gpio pin? maybe that enables the power on gpioc
22:22:27fdinelI didn't see it up to now, I'll have a look later... But I gotta go again :)
22:22:34fdinelgreat news for the wheel though
22:22:40kugeloh, you're really one of the busy ones
22:22:56fdinelfamily matters, I've got two young kids ;)
22:22:59fdinelsee ya later
22:26:35 Join opteek [0] (
22:26:45opteekHi all
22:27:24opteekI have a Sansa e260 that I had rockbox installed and working perfectly and then it randomly bricked on me
22:27:33opteekand for some reason it looks like I can
22:27:38opteekcan't get into manufacture mode
22:27:38opteekat all
22:29:08 Quit feisar (Remote closed the connection)
22:29:51 Join massiveH [0] (
22:30:11 Quit massiveH (Client Quit)
22:30:57gevaertsopteek: I've never heard of an e200 not being able to get to manufacturer mode. Are you sure it isn't working?
22:30:58 Join Acksaw [0] (
22:31:20opteekgevaerts, im in ubuntu and fail to get anything from lsusb about the sansa
22:31:37gevaertsopteek: are you running lsusb as root or with sudo?
22:31:56gevaertsThat would explain not seeing it
22:32:09opteekbut i can see other usb devices
22:32:18 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
22:32:22saratogakugel: congrats on the wheel
22:32:30saratogai tried those pins a couple weeks ago but got nothing
22:32:32saratogai must have screwed up
22:32:49saratogayeah the DBOP in pins
22:33:00saratogathe 4 that aren't shared with GPIO
22:33:02gevaertsopteek: I know, but you won't see the sansa in manufacturer mode with normal user rights
22:33:11saratogaany idea how to get an interrupt out of them?
22:33:41kugelsaratoga: not at all. you can't generate an interrupt from DBOP_DIN
22:34:37kugelI'm right now building a very basic button driver with wheel and hold button support, let's see how this turns out
22:34:42saratogathats what i was afraid of
22:34:49saratogai wonder how well polling works
22:34:54opteekgevaerts, ok i su'ed and tried again but still having no luck unfortunately
22:35:26gevaertsopteek: does dmesg say anything useful?
22:35:41rashergevaerts: don't you need to "initialise" with e200tool?
22:36:18gevaertsrasher: you do, before you can upload the bootloader, but lsusb should show it before that
22:36:26rashergevaerts: I'm not sure of that
22:36:44gevaertsHow else would e200tool find the sansa?
22:37:03rasherI don't know, but I'm pretty sure I remember not seeing it
22:37:09gevaertsas root?
22:37:17kugelsaratoga: probably not too wheel
22:37:24rashergevaerts: I'll have to pass on that
22:37:26kugelwell even
22:37:57opteekgevaerts, would i have to boot linux with the sansa connected via usb?
22:38:05gevaertsno, why?
22:38:05opteekto get anything useful from dmesg
22:38:06saratogakugel: got to run, but good work again!
22:38:12 Quit saratoga ("CGI:IRC (EOF)")
22:38:42gevaertsJust connect it following the SansaE200Unbrick and see if lsusb or dmesg see it
22:38:45rasherI'm pretty sure "e200tool init" did it for me
22:39:52gevaertsIt's always worth trying of course, but that sounds like black magic to me
22:41:55 Join petur [50] (n=petur@rockbox/developer/petur)
22:42:07***Saving seen data "./dancer.seen"
22:42:12mc2739kugel: still here?
22:43:13mc2739had to manually update button-fuze.c
22:44:30 Quit XavierGr (Nick collision from services.)
22:44:34mc2739I think I got everything patched properly, but on testing, counter continuously increments
22:44:41 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:44:49 Quit Acky (Connection timed out)
22:45:22kugelmc2739: there should be some number right to the counter
22:46:01kugel0x00000843 or so
22:46:07mc2739btn = 0x00000843 - when down is pressed btn = 0x00000803 - when power is pressed btn = 0x00000943
22:46:22kugeland when you turn the wheel?
22:46:24mc2739no other buttons cause a change in btn
22:46:34mc2739no response from wheel
22:47:48rashergevaerts: init shouldn't do any harm, should it?
22:47:59kugelmc2739: why did you have to apply button-fuze.c manually? it was a patch against svn
22:48:04gevaertsNot as far as I know
22:48:20*gevaerts admits to never have used e200tool with his e200 yet
22:49:03kugelmc2739: can you pastebin your button-e200v2.c? just to be sure
22:49:48mc2739kugel: it wouldn't apply, I'm at svn r19509
22:50:27kugelmy patch was against r19510
22:51:14kugelmc2739: well, your button-fuze.c apparently was edited (maybe only slightly)
22:52:10mc2739kugel: I did svn revert -R before applying
22:52:31kugeland I did svn diff :S
22:53:04kugelmc2739: hm, I pasted the wrong patch anyway
22:53:25kugelor no, it was the correct
22:53:59kugelmc2739: paste your button-e200v2.c please
22:57:42kugelmc2739: looks correct to me, seems our ways split in terms of hardwaremapping :(
22:58:20mc2739kugel: that's what I was afraid of
22:58:54mc2739I'll look for this code in the e200 firmware to identify differences
22:58:57kugelyou might try another pin instead of "GPIOA_PIN(1) = (1<<1);"
22:59:04kugelthat's a good idea
22:59:23Unhelpfulhrm, anybody happen to know what "accurate" reflection would be for a polished black surface? it seems to me that the linear fade used in pictureflow is probably not "right" ;)
22:59:40 Quit opteek (Read error: 110 (Connection timed out))
22:59:49kugelmc2739: we found in the fuze of a function which only sets/clears that pin, in the 0x4xxx address room, maybe you find something similar
23:00:52kugelmc2739: I may forward the disassembly of button_int (that's what my code is based off) I got from fdinel, you can search for that too
23:01:20mc2739is that what was in you diff?
23:02:07mc273949d4: through 4a29:
23:04:32kugelseems I already forwarded it :)
23:04:38 Quit Acksaw (Success)
23:04:58 Quit JdGordon|zzz (Read error: 104 (Connection reset by peer))
23:07:14 Join JdGordon|zzz [0] (n=jonno@rockbox/developer/JdGordon)
23:10:31 Join moos [0] (i=Mustapha@rockbox/staff/moos)
23:10:35 Quit petur ("Zzzz")
23:23:45 Quit planetbeing (Remote closed the connection)
23:24:28 Join culture [0] (
23:25:10 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
23:25:59Unhelpfulamiconn: several problems. the renderer renders columns, not rows, so the angle can change per-pixel, and two, the reflections are pre-rendered, by inverting and fading the bottom of the cover image
23:26:44 Join totofdu33 [0] (n=56c970d7@gateway/web/cgi-irc/
23:27:01totofdu33hi everybody
23:27:24totofdu33I've a problem with my archos jukebox recorder .....
23:27:37totofdu33is there anybody ?
23:27:56 Quit Agneees (Read error: 113 (No route to host))
23:29:08gevaertstotofdu33: what is the problem?
23:29:08 Quit totofdu33 (Client Quit)
23:29:09Unhelpfulsomebody may know about your problem. i do not. you also haven't said what your problem is yet, which makes it quite hard to help you.
23:32:04 Join totofdu33 [0] (n=56c970d7@gateway/web/cgi-irc/
23:32:20t0mastotofdu33: "somebody may know about your problem. i do not. you also haven't said what your problem is yet, which makes it quite hard to help you."
23:32:22totofdu33anybody here ?
23:32:29t0maswas say right after you quit :)
23:32:47totofdu33ok, sorry
23:33:54t0mas(I don't have an archos either.. but out of 135 people around at least somebody is likely to be able to help)
23:33:54totofdu33so, I've just install the Rockbox 3.0 but my archos recorder says: "No rockbox directory" and "installation incomplete"
23:34:09t0masand you do have a rockbox directory on the drive?
23:34:26t0masthen I think we have found the problem?
23:34:36t0maswhat installation instructions did you follow?
23:34:49totofdu33I've just done 2 things
23:35:11Unhelpfulfirst of all, the directory will be name ".rockbox", unless we do things *very* differently on archos
23:35:29t0masUnhelpful: it is named .rockbox in archos
23:36:22totofdu33install the new rockbox and also install the last daily build
23:36:41t0masbut how exactly did you install it?
23:36:52totofdu33drag and drop ....
23:37:03t0masok, but there is no .rockbox folder on the device?
23:37:47gevaertstotofdu33: try using the GUI installer from
23:37:52t0masand you did a manual install
23:37:54Unhelpfuljust unpack the archive on your device, or use the utility (preferred)
23:39:08kugelHmm...keymap issue again ;;
23:39:11totofdu33maybe because it's a 30Go whereas the original was only 20Go)
23:39:51totofdu33the utility didn't recognize my archos
23:40:20gevaertsWhat OS are you using?
23:40:41totofdu33on archos or on my computer ?
23:40:52totofdu33mac osX
23:41:38*gevaerts looks around for Mac users
23:41:55gevaertsAre there unzip tools for Mac that behave decently?
23:42:18pixelmathen you'd probably wouldn't see the .rockbox directory even if it's there
23:42:18t0masI have no idea.. but does anybody know if a Mac hides directories starting with a dot?
23:42:20t0maslike Linux?
23:42:48gevaertsFrom what I understand the default Mac unzip doesn't handle the dot correctly
23:42:49pixelmait does per default, as far as I know
23:43:06totofdu33With rockbox 2.5, I could see it
23:43:55pixelmathat too, but I thought that's one thing Rockbox Utility would handle (the unzipping)
23:44:31gevaertsCan't you point Rockbox Utility to your Archos manually?
23:45:50linuxstbgevaerts: The OS X "unzip" utility is just the standard one. I've no idea what the GUI apps do though.
23:46:10gevaertslinuxstb: I meant the GUI things. Command line would be the next step
23:46:45kugelahhh, I hate this keympas
23:47:05totofdu33Shall I try to install rockbox utility on my father's computer wich is a PC and then try to install Rockbox 3.0 on my archos ?
23:47:10kugelthe scrollwheel clearly works (volume changes in the wps), but it doesn't want to scroll in menus/filetree
23:49:01gevaertstotofdu33: first try to configure it manually on your mac
23:49:34kugelbertrik: ping
23:50:18gevaertstotofdu33: what exactly did you try the first time? i.e. in what way didn't Rockbox Utility fine the archos?
23:50:59totofdu33I'm just going to do it again and I tell you
23:54:11 Quit stripwax (Read error: 104 (Connection reset by peer))

Previous day | Next day