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 2008-12-16

00:01:27lucentkugel: ooooh I want to try :)
00:02:36lucentkugel: I'll send you the Clip2GB after a delay, I need paycheck money and pay my monthly rent first before other things ;)
00:03:06kugelno need to rush, we have time :)
00:03:23lucentthe other night I tried to drive back to my apartment, and my car stopped and slid backwards down the hill in the snow
00:04:00lucentit was exciting, I guess. I never had used tire chains before then. So I think I will walk to the post office to send the Clip ;)
00:05:10lucentanytime I think that there is something I can complain about, I just think about my car being helpless in the snow... and I don't think I want to complain so much
00:05:55lucentkugel: I did leave some feedback on your flyspray task submission, all is working on my 8GB fuze
00:06:11lucentkugel: one question for you though, can we add buttons support to the bootloader?
00:06:53kugellucent: there's button support for the bootloader
00:07:03kugelit prints text if you press any button
00:07:14lucentkugel: oh okay, yes I think that is working
00:07:33lucentkugel: on the Clip if I boot and hold "home" button or another, then it goes to the OF
00:07:43lucentkugel: is that part implemented?
00:08:24lucentI could not make the 8gb Fuze boot to the OF using the rb bootloader
00:08:28kugelpress |<< just after turning it on (you need to be fast, since dual boot is before the bootloader)
00:08:51lucentkugel: got it, just tried and it is working as you say. Thanks
00:11:33 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
00:14:08kugelgevaerts, lucent:
00:19:33 Join Strife89 [0] (n=michael@
00:21:26gevaertskugel: why the #ifdef SANSA_FUZE? Is the code likely to be different for *200?
00:21:56kugelno idea, it's safer this way
00:22:03gevaertsMaybe, yes
00:22:16kugelI don't want to know what happens if you mess with the interrupts based on wrong assumptions
00:22:30gevaertsIndeed. Especially for bootloaders
00:22:49*gevaerts will test this code on his clip before committing
00:22:51kugelbootloader shouldn't get hit, as it doesn't define mutlivolum or hotswap
00:23:06gevaertsNo, but still
00:23:21kugelyes, you're right :)
00:24:14kugellol, I never thought I'd get it working (i.e. mess with sd driver and get useful results), especially not within a few hours
00:24:19kugelbut I'd great help
00:25:46 Quit faemir (Remote closed the connection)
00:26:44Zagordoes anyone object to me committing a buildable database generator during feature freeze? it doesn't touch rockbox code other than adding/moving a few #ifdef __PCTOOL__
00:28:13*Llorean has no objection.
00:28:14lucentkugel: I get a panic "card not initialised" when booting my 8GB Fuze
00:28:28*gevaerts doesn't object either
00:28:31 Quit Strife89 (Read error: 104 (Connection reset by peer))
00:28:37 Join Strife89 [0] (n=michael@
00:28:41kugellucent: with or without card inserted?
00:28:54lucentwith card inserted, Fuze panics on boot
00:29:00lucentwithout card inserted, Fuze boots okay
00:29:05lucentinsert card, Fuze panics
00:30:08lucent8GB Fuze, 8GB microSDHC card
00:30:14kugelit works for me
00:30:17lucentI don't have another capacity of card to try
00:30:46gevaertsBoth normal builds and bootloaders built with this patch don't brick a clip
00:31:14kugellucent: can you please install a recent svn bootloader, just to be sure?
00:31:24lucentkugel: sure!
00:31:36lucentkugel: just vanilla SVN, should I apply any patches?
00:31:59kugeltry without patches first
00:32:34*gevaerts thinks he'll commit regardless of lucent's issue. The code looks clean (or at least the changes don't reduce cleanlyness), it works better for some people, and it doesn't work worse for others
00:32:39kugeland if that doesn't work, apply my button and lcd patches and recompile bootloader and main build
00:33:41kugellucent: If it doesn't help I can only imagine that your card is super slow or something (even though I already have a budget one)
00:33:44 Quit aneqrs (Read error: 104 (Connection reset by peer))
00:34:13lucentkugel: class 6 microSDHC hm, that's what it is
00:34:38kugelis that better than class 2?
00:34:39lucent"Transcend microSDHC Class 6 8GB"
00:34:58lucentclass 1 is slow, class 2 is faster, class 6 is fastest available to date
00:35:00*kugel assumes it's class 2, since he sees a tiny 2 within a ring on the card
00:35:10kugellucent: oh, maybe it's too fast then :P
00:35:31kugelI assume it works in the of?
00:35:40lucentI can say it does work with OF, yes.
00:36:13lucentI'm grabbing a vanilla SVN repo and will try what you suggest
00:37:19 Join fdinel [0] (
00:37:48 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
00:38:36 Quit tvelocity (Remote closed the connection)
00:38:51***Saving seen data "./dancer.seen"
00:39:54gevaertskugel: have you seen what the code would do after the goto that comes after the panic?
00:40:21 Quit ender` (Read error: 104 (Connection reset by peer))
00:40:23kugelanother panic
00:40:47gevaertsThis code wants to make sure :)
00:41:10kugeland, after that panic, reset initialization of the card
00:41:23kugelgevaerts: apparently ;)
00:41:45 Quit Chesteta ()
00:42:26 Join ender` [0] (
00:46:46 Quit ron_o (Connection timed out)
00:47:51kugelgevaerts: bad news, I've forgotten a tiny thing
00:48:10Unhelpfulwait, panic(); goto label; panic();?
00:48:35kugelUnhelpful: and code after the last panic
00:49:30kugelUnhelpful: this way, he wants to make sure, that goto can't be considered harmful ,9
00:50:35kugeleh, a yellow
00:50:43gevaertsI think that all those panics probably need to go once things start to stabilise a bit. The code calling this does check for errors
00:51:19kugelgevaerts: I think once the code is stable, those panics won't happen anyway, and thus they can stay imho
00:51:52kugelgevaerts: you need to make another post-commit, I'm so sorry
00:51:52gevaertskugel: they can happen due to e.g. a bad card. You don't want to panic for that if you can avoid it
00:52:24kugela bad card? /me pokes at lucent
00:52:44*lucent peers into crystal bal
00:53:06lucentI don't know boss, I think the card is okay, or maybe OF is just really lucky
00:53:19kugelhave you tried svn bootloader now?
00:53:32lucentabout to compile, sorry for the wait
00:53:52kugelSo, I have made sure, that both bootloader and build *only* patched with my patch work for me
00:54:10 Quit fdinel (Read error: 104 (Connection reset by peer))
00:54:35 Join Strife1989 [0] (n=michael@
00:54:45 Quit ender` (Read error: 113 (No route to host))
00:54:53 Quit Strife89 (Read error: 110 (Connection timed out))
00:54:55kugelseeing this post-commits, adding the sd_enabled was a nightmare ;;
00:55:18 Nick Strife1989 is now known as Strife89 (n=michael@
00:55:49lucentkugel: I will be installing bootloader from plain SVN r19446, is that correct?
00:56:42kugellucent: you might take the latest, bootloaders *shouldn't* be touched by this patch
00:57:35 Join ender` [0] (
00:57:37kugelgevaerts: you wait for the build table? ;)
00:58:04 Quit Strife89 (Client Quit)
00:58:26kugelsorry about the yellow btw: I still get this massive warnings due to ubuntu 8.10, so it happens that I overlook warnings
00:59:26gevaertsNo problem
00:59:51*gevaerts wants to know why r19450 doesn't start building
01:02:45 Quit ender` (" I like work: it fascinates me. I can sit and look at it for hours. -- Jerome K. Jerome")
01:03:58kugelgevaerts: the weird thing. playback is still unreliable for ogg/mp3, as I still get panic for FIFO FULL on the internal storage, even though I'm playing from microsd
01:04:49gevaertskugel: probably related to the playlist control file then
01:05:55kugeldeactiving crossfeed seems to help
01:06:09kugelso, it might have something to do with slowness of decoding
01:06:14lucentkugel: okay, I have bootloader from plain SVN r19450 on my Fuze. Which firmware should I try, just plain svn r19450?
01:06:24kugellucent: yeps
01:06:37 Quit lasser (Read error: 60 (Operation timed out))
01:07:13 Quit Thundercloud (Remote closed the connection)
01:08:37 Quit herrwaldo ("Konversation terminated!")
01:08:57 Join CaptainKewl [0] (
01:09:03kugelhm, I'm kinda wish funman was here to see recent progress (especially on buttons)
01:10:31lucentkugel: okay, well the vanilla SVN has no working buttons ;)
01:10:34kugelmp3 is unplayable
01:10:45lucentmy backlight turns off very quickly
01:10:45kugellucent: sure, but does it panic?
01:10:57kugelinserting usb turns it on
01:11:05kugelif not, expect it to be panic'd
01:11:23lucentkugel: booting with 8GB microSD card = panic
01:11:52lucentbooting without microSD card = OK
01:12:00lucentinserting microSD card = panic
01:12:27kugelhm weird.
01:12:35kugel"uninitialized card" yes?
01:13:18kugellucent: I'm sorry, I can't really help. It seems card related. It works just fine here
01:13:30kugeland I can only test with cards I own
01:14:47gevaertslucent: can you apply and try again? That should give a bit more information
01:15:43kugeloh yes, the return value tells much
01:16:17kugelgevaerts: I think it's funny how rockbox reboots when I play mp3
01:16:47gevaertskugel: try turning the volume down. If it helps on m200, it may help for fuze :)
01:17:21 Nick fxb is now known as fxb__ (
01:19:00kugel"MCI_DATA_TIMER(drive) = 0x1000000; /* FIXME: arbitrary */" :/
01:21:20lucentgevaerts: will try, one minute while I make the update
01:23:06 Quit soap ("Changing server")
01:25:29 Quit Zagor ("Clint excited")
01:25:32 Quit ZincAlloy ("CGI:IRC (EOF)")
01:26:31kugelgevaerts: given that flac goes w/o problems (but still needs decoding, unlike wav), I think the panics are at least related to decoding speed
01:26:54kugelmaybe the clocking is not sync'd very well or something
01:30:13gevaertsPossibly. That's an area that I'm not familiar with at all
01:31:34lucentgevaerts: "card not initialised -5"
01:32:20kugel-5, weird
01:34:21kugelgevaerts: another note: "while(!(card_info[drive].ocr & (1<<31)) && max_tries−−);" isn't it wrong in this way? I mean shouldnt't max_tries come first?
01:34:56 Quit PaulJam (".")
01:35:22kugel(in init_sd_card())
01:35:50gevaertskugel: I don't think so. It's late though, so I could easily be wrong
01:36:45kugellucent: can you try to comment the line "return -5" out? it seems isn't cid isn't needed
01:37:08kugelthat's line 259 in ata_sd_as3525.c
01:37:22lucentkugel: looking at it, let me try... one minute please
01:37:47gevaertskugel: careful. If you comment that out but not the if, the next if will become conditional
01:37:59kugelah yea
01:38:14lucentI know enough C, but good thing to remind me
01:38:25gevaertsSo add an empty ; in place of the return
01:38:28kugelgevaerts: I know he's good in C :)
01:38:56gevaertsAnyway, I'm off to sleep now. Goodnight
01:39:05kugelbut those errors can get very painful, indeed
01:39:11kugelgevaerts: night, thanks again for the help
01:39:11lucentthanks for the help and advice tonight gevaerts, best dreams for you!
01:42:52lucentkugel: uh... I need to check again if I wrote my code correctly, because now it returns -6
01:44:17 Join Thundercloud [0] (
01:44:50kugeldoing ";" instead of "return -5;" should do ti
01:45:00lucent ; /* return -5; */ /* 8GB microSD on Fuze doesn't need CID? */
01:45:16lucentyeah that is what i have, now it returns -6
01:45:53kugeluhm, well
01:46:26lucentshould I not send CID? It's still evaluated even if we don't return from it
01:46:51lucentif(expression) does evaluate expression, in this case it's '!send_cmd(...)'
01:46:52kugelyou might try
01:47:27kugelit's sending a command, and getting a response. That shouldn't matter unless we actually do something witht the response
01:49:26 Part toffe82
01:49:54kugellucent: you can also look at send_cmd, there's 2 ways of returning false
01:50:17lucentkugel: ha, well if I remove that part out, the Fuze doesn't get very far ;P
01:50:20kugelone of which is that the card timed out
01:50:49kugelok, then it might be needed ;) I haven't looked to far, just within the file
01:50:58lucent"ATA error: -6" "Press ON to debug"
01:51:41 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
01:53:50lucentlooking at send_cmd now
01:54:41kugellucent: you can also add mci-delay() here and there, could very well be that we're too quick for the card
01:55:45kugelwell, 50% success rate with my patch ain't good. Gotta find some more testers
01:56:01 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
01:56:21saratogapost counts on the forums seem to be a bit off
01:56:40saratogait seems that when posts are deleted sometimes the post count is decremented and sometimes it is not
01:56:58kugellucent: Ok, I wish you good luck! I find the driver pretty straight forward, so if you're confused you might just take a look at a SD manual/doc
01:57:25rashersaratoga: Probably if the post is still in the "Removed post" "forum"?
01:57:32kugelsaratoga: do you have a working fuze, and a microsd?
01:57:59JdGordonsaratoga: also new ports doesnt count posts
01:58:23kugelJdGordon: I thought posts in unsupported builds don't count
01:58:41kugelhow did I else get 50?
01:59:19 Quit MethoS- (Remote closed the connection)
01:59:28JdGordonor that one
02:00:12 Join hd [0] (
02:01:44rasherSeems more likely
02:01:50 Quit HellDragon (Read error: 54 (Connection reset by peer))
02:02:05saratogaJdgordon: new ports does count, and removed posts isn't supposed to
02:02:18saratogathough I'm looking at someone with 7 posts, all in removed posts, and a post count of 4
02:02:47saratogakugel: yeah let me try it out
02:04:15 Nick hd is now known as HellDragon (n=jd@Wikipedia/HellDragon)
02:04:32 Join mc2739 [0] (
02:04:47kugelsaratoga: my patch works fine on my fuze+microsd, but apparently not on lucent's combo
02:04:59saratogai've got a 2GB and a 1GB here
02:05:05lucentthey're different combos though
02:05:15*lucent hopes saratoga's combo works
02:05:22kugellucent: Hm, you haven't touched the partitions of your fuze do you?
02:05:34lucentkugel: nothing weird like that, it's all stock configuration and filesystems
02:05:37saratogano i haven't
02:05:51saratogaoh sorry misread that last question
02:05:56kugelI've touched them (my fuze is actually only 500MB as of now), to workaround the problem with reading past 1GB
02:06:31kugelit could be, that rockbox has a problem with the layout of your fuze's internal memory and messes the SD up
02:06:33mc2739kugel: my 4gb card works on e200v2 on first insert - remove and reinsert and I get DATA CRC FAIL
02:06:57kugelmc2739: was it the same gpio pin?
02:07:28 Quit BHSPitLappy (Read error: 60 (Operation timed out))
02:07:46mc2739I used svn - did not change gpio
02:08:07kugelbut the #ifdef SANSA_FUZE I suppose
02:08:13mc2739just added E200V2 to ifdefs
02:08:25 Quit Thundercloud (Remote closed the connection)
02:09:07kugelmc2739: ehh, I get those too now. It was supposedly fixed before. I haven't had this when it went into svn
02:09:52 Quit mofux (Remote closed the connection)
02:10:06mc2739wait, I am not vanilla svn, let me try again
02:11:18kugelmc2739: I got those too now. Only god knows why
02:14:23lucentkugel: I think I should try to cause a panic without any remmovable SD interaction
02:16:11*kugel doesn't understand this
02:16:19kugelit worked the whole time, now not anymore
02:16:20saratogaoddly my fuze doesn't boot, just gives me a panic, SD: Data timeout
02:16:35kugelah, more news of not working
02:16:42 Join Aurix_Lexico1 [0] (
02:16:42saratogathats with no SD card
02:16:45kugelsaratoga: w/o microsd?
02:16:58kugelthat's weird
02:17:09lucentkugel: your SD patch might be good but with some unrelated problem that always exists
02:17:26lucentlike how you said your Fuze is modified to be 500mb
02:17:40lucentmine is not modified and I see the errors
02:17:56lucentsaratoga: is your Fuze in stock configuration, any changes to the filesystems?
02:19:26kugellucent: what's currently driving me crazy is that I get those CRC failures, which I should've fixed before it was committed
02:19:53saratogai have not made any changes to the fuze, i just updated the bootloader and .rockbox to current svn
02:21:10lucentkugel: my Fuze device's internal memory is using 506M only with files
02:21:20lucentso maybe I am avoiding the bug because of that
02:22:55 Quit Aurix_Lexico (Read error: 110 (Connection timed out))
02:24:20kugelsaratoga: and what happens if you boot with microsd?
02:25:18saratogakugel: the same
02:25:55 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
02:26:15kugelsaratoga: that's really weird though. can you try r19446, to see if it's even related to my patch?
02:29:16saratogakugel: sure
02:31:49 Quit XavierGr ()
02:33:55 Quit avacore (Read error: 110 (Connection timed out))
02:34:12saratogakugel: it also happens with r19446
02:34:38kugelsaratoga: uhm, seems to be related to your fuze then
02:35:38kugelmc2739: I think I can get it work again
02:36:40mc2739kugel: what needs to be changed?
02:37:02kugelmc2739: try this
02:38:54***Saving seen data "./dancer.seen"
02:39:32 Join avacore [0] (
02:43:31kugelmc2739: does it work?
02:46:36mc2739kugel: yes, that fixed it
02:48:31kugelI always assumed that sd_enable will get us into trouble before the sd driver is fully working
02:49:07kugelmc2739: so, you microsd is working? you don't have lucent issues?
02:50:04kugelmc2739: he basically gets a panic whenever his card is inserted, so
02:50:11lucent(clarifying: My experience is a panic when booting with or inserting microSD card)
02:50:59mc2739kugel: everything looks good - copied files to and from microsd
02:51:27mc2739remove and insert multiple times - no problems
02:53:31kugelmc2739: wonderful. Let me upload a diff with the (a slightly different, but the correct one imho) and the proper #ifdefs
02:54:24mc2739wait, one problem now - playing mp3 causes reboot - cannot find .rockbox directory - microsd is now the only drive
02:56:20kugelmc2739: I have that too, but I couldn't ever get mp3 to work, also other formats don't cause that, so I don't think it's related
02:57:08Aurix_Lexico1is there any interest in a turret defense game?
02:57:55kugelsaratoga: would be nice if you could commit
02:58:15kugelor, mc2739 poke someone else :)
02:58:40mc2739kugel: its not just the reboot, but when I gor to rockbox info, the microsd shows up as int and msd not present
02:58:55Aurix_Lexico1because I have 1100 lines written already
02:59:13kugelmc2739: those are only mp3 problems correct? other formats don't do that
02:59:50mc2739yes, the mp3 causes the reboot
03:00:08rasherAurix_Lexico1: We rarely reject working plugins, but keep an eye on keeping it screen-independent, so it works on different players
03:00:24kugelmc2739: do you think it's related?
03:00:27mc2739but the microsd should not become the internal drive on the reboot
03:00:43saratogakugel: ok going in
03:01:02Aurix_Lexico1yeah, it should be fairly easy to change it to different screen sizes
03:01:15kugelmc2739: I've tried some mp3's now
03:01:37kugelmc2739: some just freeze, some reboot with perfect internal memory, some without memory at all. it's just weird
03:02:04Aurix_Lexico1only thing is I don't have a rockbox'able player to test it :/
03:02:34mc2739ok, maybe we should wait to see if fixing mp3 playback corrects this issue
03:02:44kugelmc2739: yep, that's what I think too
03:03:09kugelsaratoga: thanks a lot
03:04:42mc2739kugel: have you tried to get the radio working yet?
03:05:01kugelI'm still under the impression that calling init_pl180_controller() is kinda wrong, but I didn't get it working without
03:05:12kugelmc2739: no, I was busy making microsd today :)
03:05:20kugelplanned for tomorrow though
03:06:02 Quit HellDragon (Client Quit)
03:06:06 Join HellDragon [0] (
03:06:19kugelmc2739: should be easy if you got it working
03:06:33*kugel goes to bed
03:06:41kugellucent: good luck again!
03:07:06rasherAurix_Lexico1: heh, well that's not absolutely necessary of course, because of the sims. details the currently working Rockbox targets. Unfortunately, none are in production. has Sansa E200s that are guaranteed to work with Rockbox, but only ship to the US. Otherwise, used is your best bet.
03:07:54 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111319]")
03:09:20rasherAurix_Lexico1: Strike that, froobi appears to be sold out on v1 Sansa E200s.
03:09:56 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:10:09 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
03:10:26 Join aarcane [0] (
03:15:12 Join ajonat [0] (n=ajonat@
03:20:58 Join sbhsu [0] (
03:21:06 Join slact [0] (
03:26:49 Join soap__ [0] (
03:27:16 Join fdinel [0] (
03:29:31lucentsaratoga: thanks for committing, sorry I replied to that bug in duplicate of what you had to say
03:30:40slactGreetings. I have sucessfully registered on the rockbox TWiki (thanks for fixing the validation error, whoever did that) and would like to request write permissions to add a theme. My wiki name is LeoPalmer
03:31:38saratogaslact: added you
03:31:54lucentah, same here, TWiki username EricShattow
03:33:11soap__slact, lucent just make sure you leave contact information along with the theme, thanks.
03:33:40lucentsaratoga: are FS #9623 and FS #9645 ready for committing?
03:33:52slact−− on the wiki page, or in the theme zip?
03:34:12saratogalucent: ask kugel
03:34:22saratogai spent this evening trying to figure out why my fuze won't boot
03:34:23 Join Chronon [0] (
03:34:25 Nick soap__ is now known as soap (
03:35:02lucentkugel was looking for a sponsor to commit those patches
03:35:17soapslact, preferably inside your theme (where you put the license info) as well as on your wiki userpage.
03:35:57saratogalucent: do you remember when I can check the logs
03:36:18Unhelpfullucent: then it sounds like he considers them ready...
03:37:04lucentsaratoga: FS #9623 says it right inline "Someone volunteers to commit?"
03:37:37saratogalucent: yeah then he updated it
03:38:07saratogaas far as I'm concerned fuze stuff can go in whenever kugel says its ready since hes really the only one seriously hacking it right now
03:38:12Unhelpfulmaybe commit, but leave open for updates? it's not like it's one a supported target.
03:38:26saratogafair enough
03:39:54 Quit aarcane (Client Quit)
03:40:20mc2739how about committing an e200v2 patch? FS #9569
03:41:52lucenthm, does FS #9569 patch improve things for your player, mc2739
03:42:40mc2739yes, it fixes the lcd_update_rect function which is buggy in svn
03:44:44saratogamc2739: ok that looks safe enough
03:45:57saratogaSVN is absolutely crawling tonight
03:46:50 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
03:49:04ChrononThere's a strange bug that I didn't notice before in the HTML manuals:
03:50:09Unhelpfulsaratoga: it's git-svn users!
03:50:20ChrononIt seems to be bad handling of ligatures by the Latex to HTML converter. But I hadn't noticed this behavior until today.
03:51:08saratogai'm leaving all these AMS sansa FS entries open since it seems like people keep finding new bugs
03:51:20lucentagreed to that
03:51:55lucentthere's some doubt that remains about how the SD driver is working
03:52:57lucentI mean that maybe SD handling bugs are causing problems for other areas
03:55:03 Join Hillshum_ [0] (
03:55:53mc2739saratoga: thanks
03:59:50 Quit saratoga ("CGI:IRC (EOF)")
04:07:18 Join blkhawk- [0] (
04:12:31 Quit mc2739 ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
04:13:33soapslact, beautiful, a model of what a theme upload should contain.
04:13:38soapOne thing, a tiny nitpick.
04:13:54soapdid you look at the wikipage after you edited it?
04:14:18slactdid i miss something?
04:14:20soapSee how EscapeBoD is highlighted?
04:14:38slactah, i see.
04:14:43soapThat is because TWiki ASSUMES all CamelCase words are WikiWords and attempts to make a link to them.
04:14:53slactwill link it proper.
04:14:58soapYou have two options - one is
04:15:07slacti'm already there.
04:15:19soapyes - to make it a proper link - that is going above and beyond the call of duty and proves what a swell guy you are.
04:16:05soapThe other is to make it !EscapeBoD - the ! is invisible in this context and says "Hey TWiki! The following word isn't a WikiWord, Ok? No linking!"
04:22:34slactok, that should do it. thanks for the tip
04:24:47 Quit blkhawk (Read error: 110 (Connection timed out))
04:25:16 Nick blkhawk- is now known as blkhawk (
04:25:19 Quit perrikwp (" ajax IRC Client")
04:27:42soapthanks for the WPS
04:27:49 Quit Hillshum_ ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
04:28:10 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
04:32:58Unhelpfulmy CPP-fu is apparently insufficient, how do i make sure i'm including the correct bmp.h, now that plugin.h includes the "core" bmp.h? :/
04:33:35lucentwow, not fun
04:33:58hobbsUnhelpful: #include "filename" searches relative to the source file before it searches the include path.
04:33:59lucent#include "./relative/path/to/file.h"
04:34:01Unhelpfulthat's creating... a *tiny* bit of fallout in my postfreeze branch.
04:34:27Unhelpfullucent: i think that pattern might break the deps generation in build system?
04:36:06Unhelpfulhrm... it would seem, since plugin.h gets the "right" one when sliding_puzzle uses it, apps/recorder is *in* the include path?
04:38:11rasherHrm, maybe the display invert option should have an option to not invert albumart?
04:38:41rasherWhich would infact invert albumart when displaying, since the display invert is a hardware feature, I believe?
04:38:57***Saving seen data "./dancer.seen"
04:43:02Unhelpfulprobably easy enough to invert-while-blitting, i think all our display formats can be inverted with a ^= ~0
04:44:12lucentrasher: I'm failing to think of a useful case for that, could you explain more?
04:45:11 Quit JdGordon (Read error: 60 (Operation timed out))
04:45:48rasherlucent: The display invert feature is nice if you have a WPS that is dark and you want light (or the opposite). For example, I think the default theme looks nice inverted, but that makes albumart look silly
04:46:00 Quit jhulst (Remote closed the connection)
04:46:07rasherSo the display invert option could be "Yes / No / All but albumart"
04:46:32lucentoh you are referring to colors not rotation
04:46:39rasherAh indeed
04:46:57lucentI'm wondering mistakenly why anyone would not rotate the album art and have a rotated display
04:47:31Unhelpfulmerely making the one in plugin.h <bmp.h> is no good... does the pluginlib dir get added to the inc path when building plugins that use it?
04:47:48lucentin that event I can't think of a useful case for album art colors to be inverted
04:47:56 Quit ajonat (Read error: 110 (Connection timed out))
04:47:59lucentwell nothing sober anyways
04:48:27Unhelpfulthe relative include works, trying build after clean to see if it breaks things :/
04:48:48lucentinverting the general color output should not touch album art IMO
04:49:10rasherlucent: I'm inclined to agree
04:49:15lucentif it does, restoring the album art to expected colors would be great
04:50:01lucentrasher: a notable exception now that I think about it some more... monocrhome displays
04:50:14Unhelpfulhow do we draw the album art? i think either an invert flag there, or walk the buffer and invert album arts when the setting is toggled, is the way
04:56:30 Join kaineo [0] (n=456d9d67@gateway/web/cgi-irc/
04:56:37kaineohey guys
04:58:05lucentrasher: side note, FS #9647 enhance script, it is ready for commit... objections about make flags raised on the task are unrelated to the patch because make obeys MAKEFLAGS
04:58:53lucenta script like this only need be concerned if we are doing some active denial of certain make flags, to check MAKEFLAGS for a flag and then erase it
04:59:04lucenthi kaineo
05:00:38kaineois there an estimate on when rockbox would work with sansa e280v2?
05:00:50lucentkaineo: it sort of works right now
05:01:04kaineosort of?
05:01:16lucentkaineo: very crashy, not all features, mp3 playback causes reboot
05:01:34lucentrockbox bootloader and firmware build and run on it though
05:01:40kaineoCurrent Status: Rockbox does NOT run on the v2 players.
05:01:56lucentyeah I wouldn't dare update that page ;P
05:02:04ameyerdepends on your definition of "run"
05:02:10lucentit's not going to be me that updates it, that's for damn sure
05:02:14kaineoso its very buggy then eh?
05:02:19 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
05:02:34lucentit's in-progress
05:02:40kaineoman, that sux and i just bought it for 50 too :( 50% off
05:02:46soapit isn't so much buggy as not everything needed to be worked out about the hardware is worked out.
05:03:12kaineoso if i install it now, would i be able to undo it?
05:03:28 Quit miepchen^schla (Read error: 110 (Connection timed out))
05:03:31lucentkaineo: yeah, actually it might help us to have another tester for some specific features
05:03:56lucentkaineo: are you comfortable with C code as in commenting out certain parts and modifying debug printfs ?
05:04:12kaineoalrite, so how would i install rockbox on this player? i updated firmware to enable MSC now
05:04:36kaineono im not, but i am a major in computer science so i might be able to figure something out if i look at it
05:04:49lucentkaineo: it's not a straightforward install, you'll need to get comfortable with building rockbox from SVN trunk, can you do that?
05:05:05lucentthere's some TWiki docs about it
05:05:13kaineothen i need to research from the beginning, let me read up on that
05:05:13 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
05:05:15lucentwhich OS are you using?
05:06:05lucentoh, that's tricky, I don't know anything about vista
05:06:51lucentI could help better if you said "Ubuntu" ;)
05:09:00kaineoi dont like ubuntu
05:09:04kaineodoesnt work with games :(
05:09:15lucentkaineo: the e2x0v2, Clip, and Fuze are all called Sansa AMS because they use similar IC AMS AS3525, if you look in rockbox SVN trunk there's some activity to get these devices working
05:10:44lucentkaineo: it would be very nice to get more testing with external microSDHC access and these devices, notably
05:10:59kaineoquestion: what is SVN trunk?
05:11:16lucentoh, are you familiar with the "SVN" software tool?
05:11:28 Quit Aurix_Lexico1 (Read error: 110 (Connection timed out))
05:11:32kaineoimma have to mess around w/ it
05:12:16lucentit is a source code management system, many programmers check in and check out bits of code over the network using "Subversion" (svn for short)
05:12:54lucentthis is much easier to explain if you're using Ubuntu or a similar Linux Distro, so I think I'll just leave you to figure it out from there ;)
05:14:24 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
05:14:25kaineoi think it'll be next week that i can take a look at it
05:14:28kaineofinals this week
05:14:50lucentgood luck on that, and nice find - 8GB player for $50
05:16:45 Quit Chronon ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
05:26:21kaineoquestion: what happen if i install rockbox on and it doesnt work out well, how would i remove it?
05:28:29lucentkaineo: the answer depends on which hardware you're installing to, for as3525 Sansa targets, the current method is to modify the stock firmware
05:28:57lucentso you would just jump into MSC mode, pop in an unmodified firmware upgrade file, and then upgrade back to OF (original firmware)
05:29:09kaineofor the e280 sansa
05:29:35lucentinstallation is similar, rockbox bootloader piggybacks onto a modified firmware upgrade
05:29:49kaineoso u gotta find out where the installations are going?
05:30:14lucentI don't know, it makes more sense if I could explain it better
05:30:52kaineoso to sum it up, would it be possible to uninstall if i install it now using the rockbox utility?
05:31:21lucentwith the rockbox bootloader and as3525 Sansa targets, the default is to look for rockbox firmware and load it unless a button press is active. If a button press is active, skip and load the original firmware
05:31:22kaineoselecting device J (sansa e280)
05:31:54lucentyeah it's not supported though
05:31:59lucentI've heard that it could work
05:32:11lucentit works as suggested for my Clip and Fuze
05:32:46kaineoand so, if it doesnt work - would it brick the e280?
05:33:02lucentI don't know about that, it's possible
05:34:05lucentthere's a real danger to bricking your player when modifying the OF upgrade file to insert the rockbox bootloader, and using that as a firmware upgrade
05:34:30kaineoah i see, dang i cant afford to test it then
05:34:33kaineoits my only mp3 now
05:35:30saratogatheres not much point in testing a port that we already know doesn't work
05:37:03lucentkaineo: if you know a friend who is more familiar with Linux and software development, and they have an as5323 target with a microSD card, don't be shy to mention that testing is needed for this stuff
05:37:23lucentI can't really suggest that you try it if you're not comfortable with the risks involved ;)
05:37:44lucentthere is a #rockbox-community channel for discussions not directly related to development, too
05:38:27kaineok ill keep that in mind
05:38:31kaineothanx for being so nice lol
05:39:25 Quit kaineo ("CGI:IRC")
05:40:26lucentI'm all confused tonight, I just said as5325... whoops. I meant the arm/as3525/sansa-* targets
05:41:22 Join pixelma_ [0] (n=pixelma@rockbox/staff/pixelma)
05:44:17 Quit Horscht ("We don't make mistakes, we just have happy little accidents")
05:47:51lucentrasher: I think FS #9601 is obsoleted by other bugs, FYI. It's a minor observation.
05:52:42 Quit saratoga ("CGI:IRC (EOF)")
05:56:46 Quit pixelma (Read error: 110 (Connection timed out))
05:58:11 Quit amiconn (Read error: 110 (Connection timed out))
05:59:13 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
06:19:05 Quit DataGhost (Nick collision from services.)
06:19:13 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
06:30:35 Quit Seed ("cu, Andre")
06:35:00 Quit esthar (Excess Flood)
06:35:20 Join esthar [0] (
06:39:02***Saving seen data "./dancer.seen"
06:59:36Unhelpfulbig batch of non-critical scaler updates @ FS #9458, and time for bed
07:05:32*Bagder detects that the last series of 86 builds took less than 5 minutes to build
07:06:53JdGordon1Mbit upload is more than enough yeah?
07:07:03JdGordonto do more than just sim builds i mean
07:08:21Bagderoh sure
07:09:02JdGordonthen things will get a bit faster hopefully in a week :)
07:25:39 Join einhirn [0] (
07:34:51 Quit GodEater_ (Remote closed the connection)
07:36:26 Quit DataGhost (Nick collision from services.)
07:36:34 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
07:36:59 Quit Nibbl (Read error: 110 (Connection timed out))
07:51:21 Join kugel [0] (n=chatzill@unaffiliated/kugel)
07:53:33kugellucent: hi. did you have luck with your card?
07:56:15lucentkugel: I have been distracted, did not try much yet
07:56:22lucentkugel: I am now hacking on buttons for the Fuze
07:56:37lucentkugel: there is yet the "Home" and "Hold" buttons to figure out
07:56:58lucentkugel: do we have a guess which GPIO these are on?
07:57:09lucenthow does someone find that out anyways, look at the circuit?
07:57:24kugellucent: not at all. I've been watching all pins (that's how I found the microsd one)
07:59:17kugelthere's 2 possibilities which come to my mind: a) home, hold and scroll wheel are read with another algorithm/protocol, and/or b) we need to set some other gpio pins before reading them (like the keymatix scan thing on m200v4/clip)
08:00:32kugelbut in fact, "direct gpio connection" isn't given
08:00:48J-23IMO we need to modify keymap to use only up, down, right, left and select and analyse pin states with I/O monitor
08:01:00kugelJ-23: I did that
08:01:24lucentJ-23: FS #9645 has a modified keymap
08:03:03J-23first I need to check why newest SVN revision (even with Domonoky's patches) shows grey screen at boot
08:03:54lucentkugel: I wonder where these buttons physically connect to though
08:04:11lucentmaybe it can help narrow to what we should look at
08:04:21kugellucent: I think saratoga did some analysis of the physical connection
08:04:24lucent(wow, bad English on my part)
08:05:15kugelwhat I know is, that the button on the wheel, and the wheel itself (and the button light), are connection with some kind of ZIF with a cable to the mainboard
08:05:15lucentkugel: does he share the findings on a forum I can look at?
08:07:19kugelnot that I know of
08:07:25kugelyou might find in in the logs
08:07:53J-23kugel: should your patch work if I just replace apps/keymaps/keymap-e200.c with keymap-fuze-reduced.c?
08:08:55J-23s/your patch/it/
08:13:32 Quit kachna (Read error: 60 (Operation timed out))
08:14:20lucentkugel: I like the work that you've done already on the buttons for the Fuze, it should be committed to trunk but I think there's some clean-up to do
08:15:22 Join Rob2222 [0] (
08:15:34lucentkugel: otherwise, I was able to make rockbox panic with "SD timeout" when installing it to higher areas of the internal flash (above 1GB I am guessing)
08:17:25 Quit daurn ("Cyas")
08:17:31lucentinteresting news for the Fuze - a new OF was released the other day
08:21:34 Join Nibbler [0] (
08:22:53kugellucent: really?
08:23:05kugellucent: that's the panic saratoga had yesterday
08:23:23kugelI already installed the new of :)
08:23:36kugellucent: which cleanup?
08:24:10kugelJ-23: just throw it in the folder (no need to delete old keymaps). Edit sources to include the reduced one
08:24:18kugelapps/SOURCES that is
08:24:50lucentkugel: minor things, some /* thing */ missing from the end of an #endif, a missing ! in a comment which makes it unclear, overly-verbose { } if structure
08:26:06lucentand it mixes two different features, one being the working button code, the other being a keymap which fits the only buttons we have working right now
08:26:35kugellucent: you mean the ifs that do { } for one line?
08:26:51lucentkugel: what's the preferred style for rockbox code?
08:27:16kugelthere's no guideline on {} in one-line ifs afaik
08:27:46kugellucent: the keymap isn't supposed to be committed ;) It's for "testing purpose"
08:27:58scorchelucent: you will likely want to read
08:28:24kugellucent: but you're right about the comments, should be fixed
08:31:20 Join ameyer_ [0] (
08:31:25kugellucent: also, before committing any of the button stuff, I think domonoky should have a opinion on that
08:31:39kugelhe basically made the code, and so he should decide about it
08:33:07 Quit ameyer (Nick collision from services.)
08:33:11 Nick ameyer_ is now known as ameyer (
08:33:42 Quit Rob2223 (Read error: 110 (Connection timed out))
08:33:59kugelJ-23: are you running clean vanilla svn?
08:34:31J-23no, but I have same problem when I try without Domonoky's patches
08:35:31kugelJ-23: which patches? didn't you have a fuze?
08:35:31 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
08:35:39J-23no, e200v2.
08:35:53kugeloh, my mind is lying to me!
08:36:19kugelJ-23: you probably have to update the bootloader
08:36:26J-23I tried with changing bootloader.
08:36:30kugelclocks changed recently
08:36:52 Quit linuxstb (Read error: 60 (Operation timed out))
08:37:08kugelwell, clean svn build and clean svn bootloader should work
08:37:09J-23and patches are #9569 and Domonoky's buttons
08:37:35kugelJ-23: the former one is committed
08:38:28 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
08:39:03***Saving seen data "./dancer.seen"
08:39:20lucentkugel: I don't understand, why did you change button_read_device(void) ?
08:39:52lucentjust moving #ifdef's around and such
08:40:22kugellucent: ??
08:40:47lucentkugel: at the very bottom of firmware/target/arm/as3525/sansa-fuze/button-fuze.c
08:41:12 Join daurnimas [0] (
08:41:23kugelbutton_int was only for the bootloader
08:41:35kugelbut it's now needed in the main too
08:41:47J-23hm, clean SVN works :)
08:42:09lucentoh okay, why keep the #ifndef BOOTLOADER though?
08:44:24lucentmaybe I am asking silly questions but it is part of the code that doesn't make sense to me, why CPP around the light handling code
08:44:29J-23kugel: should your Fuze reduced keymap work on e200 too?
08:45:18kugelJ-23: yes
08:45:59J-23I tried to just replace keymap-e200.c with keymap-fuze-reduced and now only button that works is USB ;)
08:46:13kugellucent: because that's what covered by it (when it's implemented) is probably not for the bootloader. That was before me
08:46:24kugelJ-23: have you changed apps/SOURCES?
08:46:57kugelJ-23: also, "replaced" how? I told you to not delete anything, only place the file in the keymap folder and change apps/SOURCES accordingly
08:47:06J-23replaced file
08:47:50lucentmakes sense now
08:47:53kugelJ-23: you also need to put #if 0 around the various SCROLLWHEEL defines in firmware/export/config-e200v2.h
08:49:30J-23wheel_accel* and have_scrollwheel?
08:50:00 Join ender` [0] (
08:52:09lucentkugel: sorry, it makes me crazy to look at #ifndef FOO .... #endif /* FOO */
08:52:41lucentbecause looking at #endif /* FOO */ is the code preprocessed with or without FOO?
08:53:04lucentit's nothing to do with you directly
08:53:19J-23kugel: I did everything as you said, but it still doesn't work
08:53:50lucentJ-23: without any keymap modification, left and right buttons should be working
08:54:08J-23no buttons work, even left/right
08:54:24lucentsomething else is the problem then
08:54:32J-23button driver?
08:55:04kugellucent: change it, and update the patch :)
08:55:15lucentkugel: okay, will do
08:55:18kugelwas an accident if I did it that way
08:56:40 Quit BHSPitLappy (Remote closed the connection)
08:57:05 Quit daurnimator (No route to host)
08:57:09 Nick daurnimas is now known as daurnimator (
08:58:11 Quit daurnimator (Client Quit)
08:59:04J-23I put all scrollwheel code in button-e200v2.c into #if 0, still doesn't work
09:03:34kugellinuxstb: saratoga intended to keep those tasks open
09:05:41linuxstbkugel: I know, but IMO we don't want completed tasks left open. They can easily be re-opened, or new patches should simply be posted as new tasks.
09:06:16kugelJ-23: read what I write
09:06:34kugelJ-23: I wrote in firmware/export/config-e200.h
09:06:45kugelconfig-e200v2.h even
09:06:50 Join petur [50] (n=petur@rockbox/developer/petur)
09:06:50 Quit gevaerts (Nick collision from services.)
09:07:02 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
09:07:48 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
09:07:54J-23I modified button-e200v2 *after* trying with modified config-e200.h
09:07:58J-23hm, e200?
09:08:03J-23not -e200v2?
09:09:08kugeldo make clean after changing it
09:10:49 Quit GodEater (" ajax IRC Client")
09:10:55 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
09:12:19J-23I reverted my changes in button-e200v2.c and modified config-e200{v2,}.h, still doesn't work
09:13:58kugelJ-23: show me your svn diff please
09:15:11 Quit jhulst (Read error: 113 (No route to host))
09:17:31 Join n1s [0] (n=nils@rockbox/developer/n1s)
09:18:33 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:22:25kugellucent: eh
09:22:39lucentkugel: not so good? :/
09:22:51kugellucent: You mean those "keymap code"?
09:23:14kugellucent: I think I explained why I put the #if 0 around the scrollwheel defines
09:23:49lucentkugel: I know, but that is to get reduced keymap working, isn't it?
09:24:20lucentI'm running my fuze with the patch I posted and plain SVN only, it works fine and the buttons are good
09:24:22kugelnot only
09:24:46kugelthere's no need to define scrollwheel until it's working
09:25:01kugelchances are that it'll just bug us
09:25:09kugelwithout giving a gain
09:25:29lucentah okay
09:26:18lucentI went for what I thought could be comitted without affecting too many things
09:26:38kugelwe affect what we need to affect in such an early stage
09:26:57kugelit's not like we can remove the #if 0 once we get the scrollwheel working
09:27:33lucentI agree
09:28:10 Join lasser [0] (
09:29:24 Join kachna [0] (
09:31:26lucentI want to also see a simplified keymap put into effect that would #if\ 0 out the scrollwheel, at least until some progress is made to fix the SD timeout problems
09:31:42lucentI mean, as a different patch
09:31:44kugelJ-23: don't edit firmware/export/config-e200.h
09:32:04kugelJ-23: now you need domonky's patch, and it should be working
09:32:34kugellucent: why? the normal fuze keymap is untouched
09:33:12kugelmaybe I'm not sure what you mean
09:33:21kugelanyway, I got to go, see you later
09:33:33 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
09:34:39J-23works! :)
09:37:18lucentJ-23: it may be helpful to comment on relevent tasks in flyspray, mention your hardware configuration and what svn rev + patches were applied
09:37:47J-23in comment to some task?
09:38:31lucentlike reply to a flyspray bug and say "yes it worked" or "no it did not work" and mention which player hardware you're testing with, if in normal or bootloader build
09:38:38lucentjust a suggestion anyways
09:38:46LloreanYou should post if it doesn't work.
09:38:58lucentit's kind of tough to track which bugs are outdated, or which ones could be applied
09:39:03LloreanIf it works exactly as it does for the patch author, just saying "yes it works" doesn't really accomplish anything.
09:42:30JdGordonspeaking of whcih...
09:42:35n1sif you have a different target it might though
09:42:39JdGordonpixelma: did you get aaround to testing the presets patch?
09:43:16Lloreann1s: If you have a different target, it's not "exactly" as it does for the patch author, though.
09:43:41 Join Thundercloud [0] (
09:55:11 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
09:55:59 Join ameyer_ [0] (
09:56:28 Quit ameyer (Nick collision from services.)
09:56:33 Nick ameyer_ is now known as ameyer (
10:05:28 Join daurnimator [0] (
10:07:15 Nick daurnimator is now known as zz_daurnimator (
10:07:46 Join kugel [0] (n=chatzill@unaffiliated/kugel)
10:07:54J-23kugel: everything works :)
10:08:19pixelmaJdGordon: seems better on my c200 but I just realised that I should try on my OndioFM too because the difference there is that you can't reach the presets list directly from the radio screen and have to go through the context menu (in case there are differences). Currently compiling and will try as soon as that's ready (and put on the Ondio)
10:08:41 Nick zz_daurnimator is now known as daurnimator (
10:12:57 Quit daurnimator (Remote closed the connection)
10:19:22 Quit Jabone_ (Remote closed the connection)
10:21:17J-23why GPIO monitor is available only on PP-based players?
10:21:40J-23argh, nvm
10:25:53 Quit kugel (Remote closed the connection)
10:29:13pixelmaJdGordon: looks like editing presets works better with it and no breakage even on the OndioFM
10:30:53 Quit Thundercloud (Read error: 104 (Connection reset by peer))
10:31:53pixelmaediting/deleting is where I saw problems before - the numbers of presets shown in the radio screen were still the old ones when returning unless you exited and restarted radio (IIRC, it was late). That looks correctly now
10:33:13 Join wpyh [0] (n=william@
10:35:11 Join kugel [0] (n=chatzill@unaffiliated/kugel)
10:39:04***Saving seen data "./dancer.seen"
10:41:04J-23hm, my e200v2 freezes at MP3 playback
10:41:53LloreanJ-23: Playback is still in early stages on the AMS sansas.
10:42:18J-23thanks for that information :)
10:42:34B4gderJ-23: time to debug then!
10:43:04J-23time to relax with game plugins and Wave playback
10:43:17J-23and radio scanning ;p
10:43:41B4gderoh someone jumped on the pctool at once
10:43:53*B4gder refers to dev list
10:43:57LloreanJust saw it
10:44:02LloreanLooks like he'd already been working on it, and we missed it.
10:44:16B4gderah indeed
10:44:17LloreanB4gder: I think we should release an 3.1 binary of the pc database tool when we do the 3.1 release.
10:44:33B4gderthat'd be cool, if the tool is fine enough for it
10:45:03LloreanWell, yes, if it's 'working'
10:45:09LloreanIt could be a 'beta' for 3.1 :)
10:47:38ZagorI looked at his patch, but it changes _way_ too much to be acceptable
10:48:07Zagorbut I will encourage him further
10:50:02J-23hm, I tried to play video
10:50:17kugel\o/ first success on c200v2
10:50:33J-23"*PANIC* \n SD: Internal storage: RX OVERRUN, RXR FIFO FULL," ;)
10:51:04LloreanJ-23: That SD issue is also expected.
10:51:10LloreanYou can get it during playback too
10:51:23B4gderanother opportunity to do some debugging!
10:52:18 Join ameyer_ [0] (n=ameyer17@
10:56:03linuxstbB4gder: What do you think about adding the database tool (and maybe other things - e.g. checkwps) to the automatic builds in some way?
10:56:24B4gderthat's a good idea
10:56:38B4gderwe need to just consider how it is done the best way
10:56:55 Join mofux [0] (
10:56:58kugelmaybe only built daily, but automatic build would be nice to keep them wokring
10:57:01LloreanIf that's the case, I would think we might want rbutil builds too.
10:57:10B4gderyes, we do
10:57:44linuxstbLlorean: Maybe, but changes to core Rockbox can't break rbutil. checkwps and the database tool look quite delicate.
10:57:50Lloreanlinuxstb: What's checkwps? I seem to have missed that, I thought it was a command-line switch for the sim.
10:58:21linuxstbLlorean: It's a standalone utility to validate a wps.
10:58:43linuxstbI wrote it mainly for future use in our mythical themes site.
10:58:48LloreanAaah, right, that one.
11:00:53*linuxstb wonders if the database should skip unplayable files - e.g. high compression level ape - or if that will just cause more problems than it solves...
11:01:29LloreanProbably just confuse things
11:01:49LloreanIf someone reports their database not showing all files, it's already confusing enough without another reason why it could be skipping them.
11:01:49 Join PaulJam [0] (
11:01:58linuxstbYes, I think so too. Plus it would require target-specific database building.
11:02:27 Quit SUSaiyan (Read error: 110 (Connection timed out))
11:02:45n1shmm, maybe it should have a hwcodec switch that only includes files the archoses can handle though
11:05:40pixelmaI'm not even sure what the database would do on target currently with e.g. an ogg on my OndioFM...
11:07:52linuxstbIt should just ignore it - no code relating to the swcodec-only formats is compiled.
11:08:18linuxstbSo for example, .ogg isn't in the list of audio-file extensions.
11:11:07 Quit ameyer (Read error: 110 (Connection timed out))
11:12:09Zagorhum, does anyone know the difference between utils/songdb and tools/database?
11:13:02B4gderit's probably the same thing from different times
11:13:16LloreanIs one of them the old tool from pre-tagcache?
11:13:48Zagortools/database uses tagcache
11:14:07Zagorand songdb too...
11:14:22Zagorit rather looks like songdb is a little more evolved
11:14:53Zagorso I should probably have put my effort there instead. uhh...
11:15:04 Quit linuxstb (Read error: 104 (Connection reset by peer))
11:15:16kugellinuxstb: can you quickly look if my asm code is correct for setting the button light? I assume it's on D7 on c200v2 too
11:15:37kugelgone while typing
11:18:26n1sZagor: isn't songdb the one from the patch, it doesn't seem to be in svn unless i'm blind
11:19:09Zagorah, you're right! It's in my dir because I have been looking at his patch before, and svn revert didn't remove it.
11:20:02 Quit kugel (Remote closed the connection)
11:20:13 Join SUSaiyan [0] (
11:23:09Zagorwhy do we have both utils and tools? is there a difference intended?
11:23:23B4gderyes, tools is for building Rockbox
11:23:30B4gderutils are just "related" stuff
11:23:38Zagorright, so database and checkwps should be moved out of tools
11:24:00B4gderI would say so, yes
11:24:38 Join robin0800 [0] (
11:34:00 Join kugel [0] (n=chatzill@unaffiliated/kugel)
11:36:11 Quit Nibbler (Read error: 110 (Connection timed out))
11:36:30 Join DerDome [0] (
11:58:13 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
12:03:17SlasheriZagor: great that you fixed the tagcache tool! i think that is the right tool to develop, because it uses tagcache core functions to build the db
12:03:29Slasheriso the format should always be correct
12:10:58 Join RockRabbit [0] (n=3aac9a01@gateway/web/cgi-irc/
12:13:29 Quit robin0800 (Remote closed the connection)
12:13:34RockRabbitBagder, would like to get a copy of the AMS5325 datasheet if possible. I am working on the Sansa c250 - discovered the first key today (now on the hardware mapping page). I have bought an extra c250 to do the work (I now have three - two V2's and a V1). I understand that this would be for my use only and I will give it to no one else under any circumstances.
12:14:07kugelRockRabbit: hey. have you tried the buttonlight code?
12:14:46RockRabbitNot yet - its late here (Australia) and I will be hitting the sack soon. Will probably try it tomorrow.
12:15:18kugelRockRabbit: you might re-read my post, I've posted shit and editted several times (should be correct information now)
12:16:07 Quit DerDome ("Leaving.")
12:17:33RockRabbitYes I read it only a minute before posting here. Looks pretty straightforward. I assume the button light is valuable for tracing code execution?
12:18:05kugelRockRabbit: yep. You can let it blink and stuff. Much better than delay (not as good as lcd of course)
12:18:53RockRabbitIs there any significant danger of that code bricking the player?
12:19:34kugelI don't think so
12:20:16kugelRockRabbit: you might want to wait for someone to review it again if you feel unsure, but I quite doubt it can brick your dap.
12:20:58RockRabbitThought so. Good. I would like to see the datasheet if I can before doing too much more. I dont like working in the dark. Been boning up on ARM assembler but I dont think I really need too much skill in that area - its the hardware that I need to know about in detail.
12:21:33kugelthe datasheet doesn't tell you how sandisk used pins, but it helps to understand how gpio's working and stuff
12:23:21kugelRockRabbit: if you want to start using C, you should probably start integrating c200v2 code into the target tree, of course with disabling almost every feature
12:24:02RockRabbitYes thats what I want to know about. I did send an email to Sandisk requesting that they send hardware information to the Rockbox project. I fancy a laugh every now and again. You never know - sometimes miracles can happen.
12:24:41kugelworth a try, but don't hold your breath for an answer
12:24:53RockRabbitThink i'll concentrate on getting the buttons mapped and the backlight first. I know zilch about the ROckbox C code, so thats gonna be a lot of work and take a while.
12:26:18 Join robin0800 [0] (
12:27:18kugelRockRabbit: it would definitely usefull if you edit the rockbox source so that you can at least build a bootloader. The bootloader can contain pure C
12:27:41kugeland the bootloader will be executed by the dual boot (which you have found a button for)
12:28:40RockRabbitI plan to do that, but that will take some time.
12:29:10RockRabbitAnyway, have to go now. Will post further results on the forum. Bye.
12:29:20kugelRockRabbit: it doesn't take too long actually
12:31:30 Quit RockRabbit ("CGI:IRC (EOF)")
12:33:17 Quit GodEater (" ajax IRC Client")
12:33:23 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
12:39:07***Saving seen data "./dancer.seen"
12:39:18 Join debber [0] (n=983e6da3@gateway/web/cgi-irc/
12:40:36debberI am trying out the ./rockboxui and I have problems with the database. When I try to access it it always says 'Building database... 351 found (PREV to return)
12:47:04debberIs anyone alive?
12:47:29Zagoryes, but apparently noone has a quick answer to your problem
12:49:11debberI followed the guide to run the rockboxui (configured to ipod video) and I cannot access the database nor play an mp3 (through) the 'Files' menu
12:49:27Zagorwhat happens when you play the mp3?
12:49:44debberI am running Linux-ubuntu
12:50:09debberIt stops immediately. Goes from 'pause' to 'stop'
12:50:53debberok, the was a faulty mp3
12:52:34debberI get it. If you have a faulty mp3 in your simdisk folder, the database never actually initializes. It keeps on saying 'Building...' but never finishes.
12:53:16n1sdebber: that sounds like a bug, could you please file a bug report and attach the file?
12:54:08debberyup, I will investigate and report it.
12:54:43 Join moos [0] (i=moos@rockbox/staff/moos)
12:56:24linuxstbdebber: Also, you need to restart Rockbox to fully initialise the database. However, I'm not sure if there's a way to cleanly shutdown the sim...
12:59:22 Quit goffa_ (Read error: 60 (Operation timed out))
12:59:30JdGordonpixelma: ok great... ill commit it then
13:00:50debberif you have an empty file as mp3 in you directory (and nothing else) the database fails to get initialized. A rather uncommon scenario, but nevertheless wrong behaviour.
13:01:00 Part linuxstb ("Leaving")
13:01:27*debber is out for a meeting
13:01:35 Join Aurix_Lexico [0] (
13:03:17 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:04:46pixelmathere is already a bug report for the database failing to initialise without any music files at all. I'd guess the latter is related
13:05:18 Quit kharo (Read error: 110 (Connection timed out))
13:05:27 Join kharo [0] (
13:06:11pixelmafailing to initialise might be the wrong term... the building process hangs
13:06:42 Join daurnimator [0] (n=fake@unaffiliated/daurnimator)
13:06:48n1syes, i think that only invalid music files gets the same result
13:07:03 Quit kachna ("Konversation terminated!")
13:09:31 Quit einhirn (Read error: 104 (Connection reset by peer))
13:09:58n1smaybe it's time to poke Slasheri about that, FS #9093 :)
13:17:45 Quit kugel (Read error: 104 (Connection reset by peer))
13:18:31JdGordonback on the topic of checkwps and the database tool... wasnt there an idea/plan to make a proper 'lib' out of the internals so changes to the core are less likely to break things?
13:18:36bxcracerwhat is the status of clip audio playback ?
13:19:16Lloreanbxcracer: If you're interested in working on it, status updates and instructions for getting started are in the wiki and the forum thread.
13:19:24LloreanOtherwise, we'll let you know when it's ready for users.
13:19:35bxcracerok, thanks.
13:19:44wpyhkugel: is the DATA TIMEOUT error from the bootloader or the firmware?
13:22:35 Join kachna [0] (
13:25:03 Join AndyI [0] (i=AndyI@
13:25:06 Quit AndyIL (Read error: 145 (Connection timed out))
13:29:38 Join kugel [0] (n=chatzill@unaffiliated/kugel)
13:41:07 Quit moos (Read error: 104 (Connection reset by peer))
13:46:35 Nick fxb__ is now known as fxb (
13:48:46 Quit Aurix_Lexico ("Leaving.")
13:55:24 Join fredddy [0] (
14:01:03 Quit robin0800 (Remote closed the connection)
14:04:12 Quit jhMikeS (Nick collision from services.)
14:04:18 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:04:48 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
14:06:09 Quit nplus (Read error: 110 (Connection timed out))
14:09:53 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
14:19:11 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:22:16 Quit linuxstb ("Leaving")
14:23:20 Join kachna|lappy [0] (
14:25:15 Join nplus [0] (n=nplus@141.25.Globcom.Net)
14:26:30 Quit kachna (Read error: 110 (Connection timed out))
14:30:38 Join robin0800 [0] (
14:37:54 Join LambdaCalculus37 [0] (i=44a04303@rockbox/staff/LambdaCalculus37)
14:39:08***Saving seen data "./dancer.seen"
14:42:29 Quit kugel (Read error: 113 (No route to host))
14:49:22 Quit nplus (Read error: 110 (Connection timed out))
15:06:13 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:08:49debberCan u use rockbox to set a rating on a song?
15:08:55 Quit CaptainKewl (Remote closed the connection)
15:13:09PaulJamyes, the rockbox database has a rating feature.
15:14:25debberI am trying to find the code that can set a rating on a song (to add it to the context menu), but so far I have not yet found a pointer to do it. (first time in contact with the code).
15:14:44 Join Esdian [0] (i=user@
15:27:02 Quit bmbl (Read error: 110 (Connection timed out))
15:38:07 Join goffa [0] (n=goffa@
15:40:04n1shmm, does anyone use the "playlist catalog" ?
15:41:59Unhelpful"the what"?
15:41:59*LambdaCalculus37 does
15:42:19LambdaCalculus37Unhelpful: It's an option in the Playlists context menu.
15:43:10Unhelpfuli vaguely recall having seen it. i'd probably have dug through the manual for what it is if it sounded like something i felt was "missing" from my Rockbox Experience (TM)
15:43:52n1sLambdaCalculus37: if i use "Add to New Playlist" to create a new playlist, start playing that playlist and then add more tracks through the catalog the current playlist isn't updated, but if i stop and resume it is, and resumed in the wrong place, do you think this is intended?
15:45:44n1sUnhelpful: i think the main point is being able to modify non-current playlists
15:46:16Unhelpfulohhhh. yeah, i pretty much don't use playlists at all, aside from the current one.
15:46:28LambdaCalculus37n1s: I'm not entirely sure. Let me try that out.
15:46:45*wpyh doesn't use playlists, but thinks n1s may have found a bug
15:47:45LambdaCalculus37Argh... fine time for my iPod to report a flat battery. :/
15:49:17debberHow can u set a song rating using rockbox?
15:50:30n1sdebber: maybe by checking the manual?
15:53:06ZagorI think we ought to discuss default settings for the release. I want backlight timeout raised to 30, and the just mentioned "Warn When Erasing Dynamic Playlist" might be something new users would find helpful too.
15:54:47n1sZagor: sounds sane, i though the warning thingy was on by default though
15:55:45n1sturns out i was wrong :)
15:55:52Zagoryeah, I just checked too :)
15:56:24n1sI'd say mail on the dev list would be good for such discussion though, if you want to change others than those two
15:56:26PaulJamn1s: when you experienced the resuming at the wrong position, did you have shuffle enabled?
15:56:48n1sPaulJam: don't know, can check again
15:57:08Zagorn1s: good point. I don't know any others off the top of my head, but I bet there are some more.
15:58:53n1sPaulJam: seems to resume in the right place with shuffle off so yeah, that's probably it
16:00:51PaulJamn1s: i think this is expected then. afaik rockbox stores the random seed along with the playlist name and playlist position, so when the playlist length changes you get a different shuffle with the same random seed.
16:02:26n1sPaulJam: yes, however i would expect the current playlist to get updated when you add stuff to it through the catalog but maybe that is impossible...
16:04:01n1salso i find the catalog to be too limited to be truly useful, it can only add to the end or create new lists...
16:05:05n1sIMHO the current playlist handling and the catalog should be merged so you can do what you can with the current playlist with any playlist in the catalog
16:07:02n1sIt seems like a bit of a bolt on and not as integrated as it could be
16:10:03 Quit kachna|lappy (Read error: 110 (Connection timed out))
16:17:28wpyhfunman, kugel: my Sansa Fuze can't access files >1GB (known problem). looking at MTABLE.SYS, it references files on "mmc:0:\MUSIC\...", so I think it may be treating the internal storage as MMC rather than SD. To access >2GB MMC cards, we need to use the new standards (MMC v4.2 or later). I'm looking at it but I'm not experienced in this area, so you might want to take a look instead.
16:20:15 Quit perrikwp (" ajax IRC Client")
16:20:28 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
16:21:24 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
16:21:38*pondlife hears talk of playlists
16:23:17rasherBagder: the buildsystem doesn't mind unavailable clients, right?
16:23:21 Join Jaykay [0] (
16:25:37 Part wpyh
16:27:10Jaykayi have a question regarding the "anti-skip buffer" setting.... isnt this only possible with targets with a hard drive?
16:35:46ZagorJaykay: that settings simply says how much earlier the buffering should start than normal
16:36:03Zagorsince flash targets buffer too, the setting is valid. but rather useless...
16:37:13Jaykayzagor: i think the setting is not present on my sansa e200.... bit still present in the manual, so it should be removed.
16:37:15Zagorthere is a number of aspects in the code that shows traces of rockbox' harddrive origins
16:37:17Jaykay(from the manual)
16:37:24 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
16:37:31 Quit debber ("CGI:IRC")
16:37:58ZagorJaykay: ah, ok. right then it's simply a manual bug
16:38:18Jaykayshould i post it somewhere?
16:38:34Zagorfile a bug report
16:39:09***Saving seen data "./dancer.seen"
16:40:44Unhelpfulohhhh. yeah, i pretty much don't use playlists at all, aside from the current one.
16:40:47 Quit evilnick (" ajax IRC Client")
16:46:24 Join hillshum_ [0] (
16:46:24 Quit hillshum (Read error: 104 (Connection reset by peer))
17:00:27 Quit Zagor ("Client exiting")
17:00:59Jaykayanother question: i should do some benchs with aac files. because i dont use aac, i downloaded the rfee audio converter, converted an album into aac
17:01:19Jaykaycopied it to the player and tried to play it..... rockbox did nothing.
17:01:38Jaykayi thought ok might be a problem with .aac
17:02:12rasherJaykay: I don't believe rockbox plays raw aac - it needs to be in an m4a container
17:02:29Jaykaysame procedure with m4a..... now the player says root and below this root.... but doesnt play anything
17:03:05Jaykayah above the first root theres a rectangle (?)
17:03:55rasherI think this is the point where you file a bug report and attach or link the file
17:04:30 Join kachna|lappy [0] (
17:04:48Jaykaywhere should i upload the file? its bigger than 2mb....
17:06:17 Join toffe82 [0] (
17:17:19rasherJaykay: I've sent you a pm
17:17:58Jaykayi see, first im trying some other converters
17:18:21Jaykaybut..... actually rockbox should play it when itunes can play it
17:20:31rasherAt the very least, it shouldn't hang, so however you look at it, it's a bug
17:33:29 Join {phoenix} [0] (
17:45:18Jaykayrasher: i think rockbox' behaviour is as expected, when a music file cant be read, jump to the next readable or quit
17:45:25Jaykaythats what it does
17:45:41Jaykayit seems do be a problem of the decoder, decoding with itunes worked
17:45:50 Quit mofux (Remote closed the connection)
17:50:00 Join kugel [0] (n=chatzill@unaffiliated/kugel)
17:52:12rasherJaykay: Okay, so the file is rejected properly - but it shouldn't be. Still better than if it hung
18:00:31 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
18:06:19 Quit petur ("work->sports")
18:09:44 Quit kharo (Read error: 110 (Connection timed out))
18:10:30 Join kharo [0] (
18:17:16 Join Schmogel [0] (
18:18:38 Quit {phoenix} (Remote closed the connection)
18:19:11LloreanJaykay: Rockbox won't play many AAC files. Usually problems with the metadata layout will prevent them being played.
18:20:03 Join gregzx [0] (
18:20:04Jaykayi was just trying the third converter, didnt work.....
18:20:11 Join MethoS- [0] (
18:20:57LloreanJaykay: Have you tried using iTunes to convert?
18:21:11Jaykaythis works
18:21:31Jaykayif you can tell me how to set the bitrate of the outcomes, ill use itunes
18:21:39Jaykayi was told i should use high bitrates
18:21:47Jaykaythe highest i got was 128-.-
18:21:49EsdianJust making sure, Standerd Playlist made with Winamp work with rockbox'ed players?
18:22:26scorcheEsdian: if it is an m3u, then sure
18:22:42Jaykayand i was trying .ogg files from one of the converters.... didnt work too
18:22:58 Join karashata [0] (n=karashat@
18:23:09 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:24:43Esdianthanks, I've been using m3u cause that was the default on my winamp. ^.^
18:30:04Jaykaynow i have an converter whose oggs are working.....
18:31:28n1soggenc works well for me...
18:31:36 Join faemir [0] (
18:33:06 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:33:38Jaykayi wont use any oggs after those benchs :) its just for some guys who meant those benchs will help and they will work with aac or ogg files...
18:33:53 Join bertrik [0] (
18:34:11 Part Esdian
18:34:26n1sJaykay: what are you benchmarking?
18:35:20Jaykaybattery runtime with these oggs with cabbiev2, then the same with a empty wps and maybe with the mnu
18:35:40Jaykayi think you should know of this
18:35:52Jaykayit was a diskussion with kugel
18:37:08n1sah, i don't read all of the logs so i probably missed it then :9
18:37:31Jaykayi thought you discussed with us
18:37:42Jaykaybut i dont really know :)
18:38:14bertrikn1s, can you try the very latest si4700.c from the sansa clip radio patch and see if it still works on your gigabeast? It looks like I fixed the fm chip initialisation problems I initially found on the clip.
18:38:33n1sbertrik: sure
18:38:44bertrikI'd like to commit it if it still works on your gigabeast, after the freeze ends
18:39:14***Saving seen data "./dancer.seen"
18:39:34n1swe should branch tomorrow, no?
18:40:13 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
18:40:26kugelBagder: can you rename the SansaV2 page to SansaAMS or AMSSansa or the like?
18:41:34bertrikn1s, I have no idea to be honest. To make radio fully work on the clip, there's more stuff to do, like properly enabling the LINE2 input, so there's no real hurry.
18:42:52 Quit B4gder (Read error: 60 (Operation timed out))
18:43:59domonokyyes, we need to include LINE-IN2 support into the as3514.c driver, to get radio on all ams-sansas :-)
18:44:36*bertrik considers adding a TODO list to the SansaV2 wiki
18:47:38bertrikto explain in a little more detail what is needed for features that are marked "partial" in the port status for example, and specifically what is needed for full radio support for example
18:48:08kugeldomonoky: hey, do you have a microsd? ;)
18:48:36domonokyyes, a todo list would be nice. @kugel, no i dont have a micro-sd card.
18:49:23kugeldomonoky: oh, but I guess you noticed anyway that I got microsd and hotswap working?
18:49:44domonokyyes, i read it. congratulations :-)
18:50:25Jaykayis it possible to software-detect wether phones are plugged in or not?
18:50:38Jaykayon some ipods it works with the OF
18:50:40LloreanJaykay: Depends entirely on the hardware.
18:50:48kugeldomonoky: actually, my intermezzo (which failed) with GPIO isrs for the button was actually useful. I could just copy and paste those for the hotswap isr ;9
18:51:20JaykayLlorean: is something like stopping playback when plugging out the phones implemented?
18:51:27Jaykay(in rockbox)
18:51:39LloreanJaykay: Pausing, yes.
18:51:46kugeldomonoky: another question: how do you feel about committing the button drivers? They're incomplete, but yet enable us to actually use most aspects of rockbox
18:52:00LloreanJaykay: A quick glance at the manual for the iPods would've told you this.
18:52:30kugelbut I'm not sure myself so..
18:52:55JaykayLlorean: ok, sorry. the main question was wether this depends on the hardware or not..... and the answer is not in the manual :)
18:52:59 Quit robin0800 (Read error: 110 (Connection timed out))
18:52:59domonokykugel: i think the fix for correctly reading the buttons, should go in. everything elses should wait/stay out.
18:53:04bertrikI was a bit worried about getting only silence (not even noise) on the clip radio around 96 MHz, it turns out that the OF has the same behaviour...
18:53:29LloreanJaykay: Well, I answered that question without referring you to the manual...
18:53:36kugelI think I can agree with that :)
18:54:29JaykayLlorean: right, i think we're finding a problem that does not exist so thanks for the first answer and sorry for the second question ;)
18:54:29bertrikMy guess is that a harmonic of one of the clocks from the clip itself is received on the radio (96 MHz is four time the clock crystal frequency of 24 MHz) and that this signal is not FM modulated, so you hear silence instead of noise
18:55:56 Join miepchen^schlaf [0] (
18:59:41 Join tyfoo [0] (
19:00:06n1sbertrik: it works on the beast, but sometimes when switching between presets the tuner doesn't switch , so the radio screen shows the new station but the old one is still playing, never had this before but not entirely sure if it's caused by your patch
19:02:28bertrikn1s, ok thanks for trying
19:02:47n1shave you ever had similar behavior?
19:03:00bertrikI wonder if the failure to tune may be because of simply waiting 60 ms instead of polling for the tune result
19:03:15LambdaCalculus37wpyh: (for the logs) On the Clip, MTABLE.SYS appears to be a database of sorts.
19:03:57n1sbertrik: ah, that may be it, i'll try a longer delay
19:03:59bertrikn1s, no I've never seen that. With an earlier patch I did have bad tuning in general and sometimes the first preset would not tune, but I think I can blame that on the incorrect initialisation sequence
19:05:07bertrikthe delay is because the si4702/3 datasheet warns against i2c bus traffic during tuning when the internal oscillator is used
19:09:16 Quit tyfoo ("Carpe diem")
19:09:43 Join BigBambi [0] (
19:09:55bertrikI'll have another look at how the clip OF deals with tuning the si4702
19:12:03n1sstill not reliable with a 200ms delay but it works better
19:15:01 Join tyfoo [0] (
19:15:09 Join {phoenix} [0] (
19:17:43 Quit linuxstb ("Leaving")
19:18:21 Join mofux [0] (
19:21:18bertrikI checked both the si4700 and the si4702 and both claim 60 ms max for tuning, but only when the channel spacing is 100 kHz or 200 kHz
19:22:32 Join tvelocity [0] (
19:26:39n1syeah, i see the same, and i use 50kHz channel spacing of course... also that the time is per channel I don't know what that means when you select a freq and try to tune?
19:27:34 Join Thundercloud [0] (
19:35:07 Join Seed [0] (
19:35:31 Quit MethoS- (Remote closed the connection)
19:37:29n1shmm, i wonder if writing directly into READCHAN would work, not that it seems possible to write beyond register 6 on the beast...
19:38:29 Part pondlife
19:40:25 Join aneqrs [0] (
19:41:58n1sbertrik, how does it behave with the loop in the tune function for you? sometimes the delay is pretty noticable on the beast using that so maybe the tune actually takes quite a long time?
19:42:59n1sor we use an gpio interrupt for this...
19:44:07lucentthe AMS AS3525 datasheet is available from AMS pretty simply, it's just a form to fill out on their website ;)
19:46:03 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
19:51:14kugelbertrik: I'm listening to radio here
19:51:36 Join Chesteta [0] (
19:53:02 Join bxcracer_ [0] (n=bxcracer@
19:53:39 Quit bxcracer (Read error: 110 (Connection timed out))
19:55:17bertrikn1s, the algorithm I see in the clip OF is basically the following: check the band setting, do something with the number 50, set the channel + tune bit, write to i2c, wait 70 ms (this is a guess), read 4 bytes from i2c, clear tune bit, check if CHAN is the desired value and exit if it so or else retry all of this up to 25 times
19:55:26 Join GodEater_ [0] (n=ge@rockbox/staff/GodEater)
19:56:25bertrikn1s, I haven't experimented with the loop much
19:56:42n1sbertrik: does it do anything with the bytes read from i2c?
19:58:12n1sand does it really check CHAN and not READCHAN, afict CHAN shouldn't change from the value you write there
19:58:18bertrikn1s, I think it compares the CHAN bits read back from the CHANNEL register with the desired value
19:58:49bertrikn1s, ah you're probably right
19:59:45bertrikit combines 2 bytes at offsets 2 and 3 from some byte buffer and compare it so some value, that must be the _read_ buffer :)
19:59:47kugelbertrik: what's the issues on radio? seems to work nicely so far on my fuze
20:00:10bertrikwe're trying to make sure it works correctly both on the beast and on the ams sansas
20:00:18 Quit Chesteta (Read error: 60 (Operation timed out))
20:00:47n1sbertrik: are all i2c reads bad during a tune operation?
20:01:09bertrika difference between those is that it seems that the ams sansa use the internal oscillator while the beast uses an external clock signal. Also the beast has a si4700, while the ams sansas have a si4702
20:01:48 Part LinusN
20:02:40bertrikn1s, I guess so
20:02:59n1si read back the chip id a while ago and it seems to actually be a 4701 so it should even have RDS :)
20:03:30n1sbertrik: then i can see why they are doing that because the STC bit is cleared when setting TUNE low :/
20:03:47kugelbertrik: what about HPCM? I think it can be enabled for all ams sansas now
20:04:03kugelthe fuze needs it, and would be very surprised if the c200v2 wouldnt
20:06:15*kugel focus to fix the next issue: buttonlight
20:06:48lucentside question, I'm going to request a datasheet for the as3525... Which is it, the AS3525A or AS3525B for Clip / Fuze ?
20:07:56n1sbertrik: i can try implementing something similar to their implementation
20:07:56kugellucent: just poke Bagder, he might give it to you
20:08:15kugellucent: this way you also get surely the one we're using
20:09:22bertrikn1s, actually the code does seem to check STC in addition to comparing CHAN with READCHAN. The registers with the TUNE bit low are not written back to the chip until after those comparisons have been made
20:10:09bertrik(the STC check was disguised in a LSL#25, BMI construction)
20:10:40n1sso that is basically what the loop i wrote does but with a longer delay and a check for the channel (and a timeout)
20:14:07 Quit {phoenix} (Remote closed the connection)
20:15:19bertrikehm, yes :) except the OF code checks the READCHAN and keeps toggling the TUNE bit inside the loop
20:15:56bertriknot sure anymore if the OF tunes if in the most efficient way...
20:24:11n1sthis is only slightly more reliable...
20:24:35*n1s tries the beast OF to se if it can tune nicely
20:26:19n1sthey do tune in a new preset _fast_
20:26:55n1sthey also use 100kHz spacing
20:27:18kugelbertrik: I'm gonna upload the diff to make your radio patch work with fuze and e200v2. Can you include those if you update the patch the next time? (sorry for wrong channel in the first place)
20:27:29bertrikkugel, ok
20:28:04bertrikn1s, that pastebin looks pretty optimal to me
20:28:55n1sbertrik: maybe it just doesn't work well with 50kHz channel spacing?
20:29:43 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
20:29:47bertrikn1s, maybe... I wish they just specified something like that in the datasheet
20:30:54amiconnWhich country uses 50kHz channel spacing for terrestrial broadcast, btw?
20:31:13*amiconn believes there might be one, but doesn't remember
20:31:22 Join __lifeless [0] (n=lifeless@
20:31:33bertrikAFAICT, the "european" setting uses 50 kHz spacing, but all stations in the netherlands are on 100 kHz spacing
20:33:20*kugel doesn't know any xx.x5 station
20:33:44amiconnIn .de, 50kHz spacing is only used for cable
20:34:25amiconnHence 100kHz spacing is sufficient in rockbox. There is no target with an antenna input
20:34:59n1sthere are no such stations here either but rockbox tries to tune such frequencies for the europe setting
20:35:11rasheramiconn: Italy uses 50kHz steps
20:35:34bertrikrasher, reading ? ;)
20:35:45rasherWouldn't dream of it.
20:39:18***Saving seen data "./dancer.seen"
20:41:28 Quit mud-rb (Remote closed the connection)
20:41:32amiconnI wonder whether we should have 2 settings for Europe then
20:41:48 Join mud-rb [0] (
20:42:55amiconnUnrelated to any tuning problems, scanning in 100kHz steps will only take half the time, i.e. it is more convenient if you know there won't be any xx.x5 stations
20:42:57bertrikI wonder if using 50 kHz spacing instead of 100 kHz can have a negative effect
20:44:14kugelrasher: theoretically, or are there really xx.x5 stations?
20:44:14Jaykayi noticed there is a alarm-function on the ipod but not on some other players.... why?
20:44:58gevaertsJaykay: for the same reason that some players have recording settings and others don't
20:45:01rasherkugel: I'm quite sure there are .x5 stations. Wikipedia says so, and I seem to recall noticing it when visiting
20:45:21kugelJaykay: the answer is always the same: a) no-one didn't care enough to implement or b) the hardware can do it
20:46:06kugelcan't* rather
20:46:15 Quit _lifeless (Read error: 110 (Connection timed out))
20:46:17Jaykaykugel: i know, the question was what answer was right, so gevaerts: thanks^^
20:46:47lucentnice, I'm finally registered on the TWiki as EricShattow
20:47:05lucentmay I have edit access to ?
20:47:39kugelsure, wait a second
20:47:46pixelmathere are some targets which could have alarm (hardware wise) but don't have it yet because the implementation isn't done completely yet
20:48:09kugellucent: promise to not spam!
20:48:29lucentkugel: right!
20:48:40kugelJaykay: are you johannes linke?
20:48:53Jaykaywell spotted ;)
20:49:02kugelthere seems to be an issue with your twiki account
20:49:51Jaykaykugel: i didnt notice, what is the problem?
20:50:01kugelyou have no page. everyone has a page
20:50:23pixelmaI know of the X5, M5 and H300. A patch for adding alarm wake-up is in the tracker but it still needs work on the M5/X5 bootloader
20:50:26kugeli.e. doesn't exist
20:50:29n1sbertrik: with that tune loop i posted, using 100kHz spacing works much better than 50kHz spacing but still not 100% so maybe we need to do some similar fiddling with READCHAN as the clip OF...
20:50:40lucentdo we know anyone with a "v2 Fuze" ? i.e. OF version 02.01.22 or 02.01.15a
20:52:02Jaykaykugel: i can live without it, as long as theres something not working either..... never mind
20:52:13 Join {phoenix} [0] (
20:52:33bertrikn1s, the clip OF way just looks rather complicated to me and the datasheet suggests that STC should be enough .... argh
20:53:26n1sapparently STC gets set even if no new channel was tuned in sometimes...
20:54:58bertrikthe datasheet says that there is an initial tuning during power-up, maybe you see the STC from that tuning ?
20:55:36 Join Chesteta [0] (
20:56:01kugelI wouldn't give too much on the sansa of tuning
20:56:07kugelradio is quite bad in the of imho
20:56:31n1sbertrik: STC gets set low by setting TUNE low so that would only affect the first tune
20:58:07n1skugel, maybe so but that is only an opportunity for rockbox to be better!
20:58:18kugelsure :)
20:58:36n1salso the tuning and autoscanning works perfectly fine in the beast OF
21:01:09 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
21:01:49 Join Zagor [242] (
21:02:41 Join massiveH [0] (
21:08:44 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
21:09:02 Join ajonat [0] (n=ajonat@
21:10:40n1sbertrik: works pretty well
21:10:56 Part pondlife
21:11:38n1sor rather it never fails to tune a preset
21:12:02bertrikso the check on the channel is really needed?
21:12:20n1syes, it seems so
21:12:28 Join massiveH_ [0] (
21:12:39n1swhich seems weird because that either means that STC is set too early or that it's not set low at all :/
21:13:06 Join arohtar [0] (
21:14:17 Join vertic39 [0] (
21:14:38 Quit massiveH (Nick collision from services.)
21:14:40 Nick massiveH_ is now known as massiveH (
21:14:59bertrikn1s, ok I'll try it on the clip too. Basically it means that STC can not really be trusted :/
21:16:30 Quit faemir (Read error: 54 (Connection reset by peer))
21:16:41 Quit arohtar (Client Quit)
21:17:16 Join faemir [0] (
21:17:51n1syeah, or the update of READCHAN is delayed and is canceled if TUNE is set low before it's done?
21:18:17*lucent grins
21:18:47lucentkugel: I bet that's why listening to David Bowie's "Ziggy Stardust" sounded like karaoke
21:19:26lucentit was pretty funny, and yeah changing the L-R balance in settings would alternate vocal parts
21:19:27 Quit Chesteta ()
21:19:32kugelthe earlier you reported that, the earlier I'd told you the fix ;)
21:19:50lucentah - old school recordings. The time before engineers compressed the beans out of everything
21:24:00lucentkugel: I mentioned it and Llorean went off on me some diatribe about how it's in progress and I shouldn't complain
21:24:16bertrikkugel, I think it should be #if (CONFIG_CPU != AS3525) rather than just #if !defined(AS3525)
21:24:16 Join vitja [0] (n=vitja@
21:24:31lucentbertrik: does != work in CPP ?
21:24:32bertrikAS3525 is #defined as 3525 :P
21:24:35Lloreanlucent: I said one sentence on the fact that it's in progress and not expected to work properly.
21:25:02bertriklucent, I assume so
21:25:05kugelbertrik: maybe, I haven't thought much about it, but rather just did what the comment above said
21:25:22n1syes != works in the preprocessor
21:26:57 Join FlynDice [0] (
21:27:20kugelbertrik: and the comment exactly said "defined(AS3525)" (yes it was lacking the !
21:27:42bertrikthe comment was wrong and I removed it
21:27:50kugelok :)
21:28:31bertrikI'll commit your HPCM fix tonight, ok?
21:28:33kugelthen I apparently didn't update the one tree where I did that quite some time
21:29:37Unhelpfulamiconn: the first 90% of hq greyscale is mostly done. then comes the debugging, ie the other 90%. output_row is going to take a void* for the data to output, so that the signature doesn't have to change, but it will point to a row of uint32_t on greyscale targets, and struct uint32_rgb on color ones
21:30:18bertrikactually, I wonder why this HPCM thing is enabled first then disabled later
21:31:44lucentguessing, would it have anything to do with mono-channel playback?
21:32:32lucentbertrik: also the Fuze at least used to support audio through the dock connector in older OF
21:32:32 Quit vertic23 (Read error: 110 (Connection timed out))
21:32:50lucentthrowing ideas up, here. I don't really know
21:32:58 Join Aurix_Lexico [0] (
21:33:42kugelbertrik: the other place is at closing the driver? That shouldn't really matter I suppose?
21:34:20bertrikdisabling the HPCM saves a little power (1.5 mA or so IIRC)
21:35:07Jaykaythere are two patches and which do very similar things, but the codes are totally different.... and the second one is much bigger than the first....
21:35:14Jaykaynow the noob-question: why?
21:35:36Jaykayoh i forgot: i suggest closing of one of these two patches.....
21:35:39 Join mib_u322u6yu [0] (i=d8ef4249@gateway/web/ajax/
21:35:46 Nick mib_u322u6yu is now known as MarcGuay (i=d8ef4249@gateway/web/ajax/
21:36:49 Quit fredddy (Remote closed the connection)
21:36:55MarcGuayJaykay: Hi there. I've noticed that you're trying to help clean up the tracker - good idea. I'm wondering though if it's really valuable to close patches where although the code is currently out-of-date, it could still prove useful in the future.
21:37:14lucentJaykay: August of 2006? that's pretty old
21:37:23 Nick ameyer_ is now known as ameyer (n=ameyer17@
21:37:25MarcGuayAnd if not the code, perhaps the implementation method.
21:37:51MarcGuayIt might be more helpful to look over the bugs and see if they're still happening.
21:37:56 Join akur [0] (
21:38:37bertrikn1s, your code works fine here (but so did a single 60 ms delay)
21:38:56MarcGuayBut that might just be my opinion. Patch tasks have been closed before because they're not going to be implemented in the way the patch currently exists...
21:39:08Jaykaymarcguay: why should i not suggest closing of outdated patches: i didnt really understand
21:39:22lucentMarcGuay: what do you think of closing "wontfix" and adding a note to the wiki with a link to the bug in FS and a quick explanation? people could re-open if they are having trouble
21:39:32LloreanMarcGuay: Patches aren't around to be "helpful." They're seen as either working toward, or being a candidate for inclusion. If they're neither, they should be closed.
21:39:43LloreanIn the case of both of these patches, it's kinda a case of "we still haven't decided what to do with short-Record"
21:40:25Jaykayat least one of them can be closed (imo)
21:40:38Lloreanlucent: When patches are closed, they can have a closing note. There's not much reason to put a note in the wiki as well, is there?
21:40:43n1sbertrik: the problem i think with this is that the datasheet doesn't actually say _what_ tune does, it seems to do something and then just copy CHANNEL into READCHAN...
21:40:49 Quit tvelocity (Read error: 60 (Operation timed out))
21:40:57MarcGuayI didn't even look at the tasks, honestly, it just seemed that maybe the philosophy was flawed.
21:41:14LloreanMarcGuay: Which philosophy?
21:41:29MarcGuay"This patch is old so it should be closed"
21:41:30 Join tvelocity [0] (
21:41:40 Join LeAstrale [0] (i=5737c599@gateway/web/ajax/
21:41:49LloreanThey're rarely closed simply for age.
21:41:57Jaykaymarcguay: i never said this at the only reason
21:42:13MarcGuayPardon my myopia.
21:42:14LloreanUsually they're closed if it would be more work to resync than reimplement, or because they're old *and* going in the wrong direction (ie: the author shows no interest in correcting problems preventing it from inclusion)
21:43:05Jaykaythere was one author who didnt know what his patch does... he copied it from another developer/volunteer/whatever
21:43:17Jaykaythose patches are useless
21:43:33bertrikn1s, AN232 says about STC "Seek Tune Complete (STC) illuminates when the RSSI threshold is met in tuning activi-
21:43:33bertrikties (Tune Up, Tune Down, Seek or Scan). STC bit 0Ah;14 is set high to indicate STC
21:43:33bertrikoperation is complete and successful.
21:43:33DBUGEnqueued KICK bertrik
21:44:06LeAstraleQuestion: Have any of you tried installing Rockbox on one of the new 4th generation iPod nanos ? I have been unable to find any info newer than 2nd gen
21:44:10 Join maffe [0] (
21:44:17BigBambiLeAstrale: Won't work
21:44:29n1saha, so tune _does_ take the threshold into account.
21:44:30lucentLeAstrale: guessing that Apple locked down their hardware not to accept 3rd party firmware
21:44:33BigBambiLeAstraleAnd it is stated as not working on
21:44:36LeAstraleBigBambi: *sighs*
21:44:37 Quit maffe (Client Quit)
21:44:50lucentn1s: "should" at least until tested ;P
21:44:52BigBambiLeAstrale: New undocumented hardware, encrypted firmware
21:45:00Lloreanlucent: Pretty much all hardware we run on arguably has some restrictions in place preventing third-party software.
21:45:21LeAstraleBigBambi: I was unable to find it on the site. I will just have to wait then. Thank you all for the quick response
21:45:21lucentLlorean: why are you telling me this?
21:45:25 Join elinenbe [0] (n=elinenbe@
21:45:33BigBambiLeAstrale: Incidently, tells you this on the front page
21:45:42BigBambii.e. that it doesn't work
21:45:47JaykayLlorean: is one of the patches going to be closed?
21:45:51Lloreanlucent: So you don't need to be guessing?
21:45:58BigBambiLeAstrale: And I wouldn't get your hopes up waiting
21:46:00n1sLlorean: more like pretty much all hardware is buggy or doesn't do what the documentation says :/
21:46:04LeAstraleBigBambi: I'm sry about that, my apologies
21:46:15bertrikn1s, yeah maybe this can be interpreted that as to say that the STC bit is only one of several conditions of a successful tune
21:46:16BigBambiLeAstrale: It isn't a problem :)
21:46:25LeAstraleSeems i'm gonna have to use .ALAC then :/
21:46:33LloreanJaykay: Can't really close one until a decision is made about whether or how the feature will be implemented.
21:46:53 Part LeAstrale
21:46:59Jaykaybut they're duplicates.... some kind of
21:47:08BigBambinot exactly
21:47:20n1sbertrik: the an230 says that it "indicates" it same as a gpio interrupt (which they wait for in each of their examples before checking any bits)
21:47:32LloreanJaykay: You yourself noted they're not identical. So to close one, we have to decide that one's not the code we want to use. Until we know what we DO want to use, that decision's pretty hard to make.
21:48:27MarcGuayJaykay: Like I said, I think a better use of your time would be to root through the open bugs.
21:48:30CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:48:30*lucent throws a firecracker aat Llorean
21:48:59LloreanJaykay: The idea is to keep around tasks that may be of use to the developers. Tasks shouldn't be closed unless they're deemed useless or more effort to use than re-create in one form or another
21:49:13Lloreanlucent: A reminder, this is an on-topic channel.
21:49:28BigBambiAlthough, isn't one short press to view playlist and the other short press to view playlist and long to record. Given we already use long press to go to record, then they are essentially the same? I've only looked at the summary, not the patch btw
21:49:36Jaykaymarcguay: ill do that.... if you think it helps better
21:49:37lucentJaykay: I'm up for helping with tracker clean-up, it's easier for me though if I can test patches locally i.e. with hardware I already own or pctools
21:49:48MarcGuayLlorean: Doesn't that contradict slightly with your previous statement that "Patches aren't around to be "helpful.""?
21:49:53Jaykayllorean: ok....
21:50:18LloreanMarcGuay: They aren't around to be helpful in a general sense. As in "we don't leave them open on the off chance they'll be helpful to users, once we know we won't use them"
21:50:22LloreanSorry if I was a bit unclear.
21:50:38bertrikn1s, ok I don't know anymore what to think of this STC bit
21:50:48MarcGuayGot it. I heard it as a response to another angle.
21:50:50n1sme neither
21:51:08LloreanMarcGuay: Sorry. I should try to be a bit clearer.
21:51:25MarcGuayLlorean: Without tone it's difficult.
21:52:01elinenbeHow's rockbox coming along on the iphone?
21:52:05Jaykaylucent: i also can test the patches...... but not now as im doing battery benchmarks :)
21:52:13*lucent :)
21:52:22rasherelinenbe: It's not really, afaik.
21:52:45pixelmabtw. I read the mail and I thought about the "use white as transparent mask on greyscale targets" one because I know amiconn had some ideas about it but more in the lines of what LinusN's comment said back then - using a different colour for this. Don't know details but wanted to point him to it
21:53:17Jaykaylucent: i also have some patches wich could be committed imo, but ill wait for 3.1
21:54:07LloreanJaykay: Are they bug fixes or features?
21:55:13elinenbeHow's rockbox coming along on the ipod nano v2 and ipod nano v3 and ipod nano v4?
21:55:21JaykayLlorean: features, except 2. i thought they were fixes but gevaerts said i should wait
21:55:35krazykitelinenbe, also not coming along at all. please stop asking for status reports and check the relevant wiki pages and forum threads
21:56:10 Quit massiveH ("Leaving")
21:56:17Jaykayboth are working fine
21:56:48elinenbekrazykit: ah!
21:57:06*lucent groans a bit
21:57:15lucentelinenbe: I'll explain in privmsg :)
21:57:23BigBambiI'm sure he knows
21:59:41 Quit LambdaCalculus37 (" ajax IRC Client")
22:08:01 Join moos [0] (i=moos@rockbox/staff/moos)
22:08:02 Quit MarcGuay (" ajax IRC Client")
22:08:24 Join mib_sbh9gpmd [0] (i=d8ef4249@gateway/web/ajax/
22:08:30 Nick mib_sbh9gpmd is now known as MarcGuay (i=d8ef4249@gateway/web/ajax/
22:09:02 Quit slact (Remote closed the connection)
22:09:40bertrikn1s, this is what the clip OF does as far as I understand, except that the OF uses a for-loop instead of do-while:
22:10:14 Join stsquad [0] (
22:11:11stsquadIs is possible to create messages that take a parameter in the lang files? i.e. Save file as "%s" type thing?
22:12:33Zagorstsquad: yes
22:13:29Zagorthere are already 28 such strings used
22:13:58stsquadZagor: example?
22:14:19Zagor"Inserted %d tracks (%s)"
22:14:31Zagorgrep the lang files for %
22:14:46stsquadZagor: I see now, thanks!
22:16:11Unhelpfulhow does that work when languages require different orderings? or does it, um, not work very well, then?
22:16:56ZagorUnhelpful: not well. the parameter order must be maintained.
22:17:11rasherI'm not aware of any strings with multiple replacements where order might be different across languages
22:17:19 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
22:18:41domonokywhat does voice do for this strings ?
22:18:58stsquadWhat's the best example of implementing a multi-option menu selection in the code? I'm looking at replacing the erase dynamic playlist warning as discussed on the ml
22:19:19rasherdomonoky: Use static strings
22:23:27jhMikeSgevaerts: On Windows at least, there's a problem with the USB charging only config in that if the device is not recognized, it never gets configured and so never gets 500mA.
22:23:39 Quit faemir (Read error: 54 (Connection reset by peer))
22:23:53 Quit Zagor ("Clint excited")
22:24:29 Quit elinenbe (" HydraIRC -> <- Try something fresh")
22:25:58 Quit ajonat ()
22:27:11gevaertsjhMikeS: interesting. So we basically will need a real nop-driver (like HID or audio). I'm not sure if it's urgent though, I don't think Windows will really hard-limit the port to 100mA
22:27:50 Join MethoS- [0] (
22:28:07 Part akur
22:28:20jhMikeSPerhaps not, but I want the charging to observe things. An inf file wouldn't be enough?
22:29:07gevaertsI don't think so. Not sure though
22:30:41gevaertsFS #8747 is a possible "real" solution to this, although HID would probably be better
22:34:06gevaertsjhMikeS: by the way, did you see r19418?
22:34:30jhMikeSNot specifically. I'll check it out.
22:35:53jhMikeSah, yes. I did find a way to not require audiobuffer reinit that really worked better overall. hehe.
22:36:56gevaertsI could revert it, but there were no binsize differences for that one
22:37:07bertrikjhMikeS, are you familiar with the HPCM setting in the as3514 codec? we first turn it on in audiohw_preinit but it is later disabled in audiohw_postinit
22:37:54bertrikWouldn't it be better to just turn it on or off in audiohw_preinit and not modify it later?
22:37:56jhMikeSSomeone else added the turnoff. I would think you could just set it appropriately in audiohw_preinit.
22:38:28 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111319]")
22:38:47 Join petur [50] (n=petur@rockbox/developer/petur)
22:39:05bertrikok, I'll code that and try it on my e200 (which allows it be off) and my clip (which needs it on)
22:39:22***Saving seen data "./dancer.seen"
22:39:27 Quit {phoenix} (Remote closed the connection)
22:40:12jhMikeSThe only time HPCM helped on e200v1 was during shutdown as part of preventing pops.
22:43:08 Quit XavierGr (Nick collision from services.)
22:43:19 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:44:52 Quit ameyer (Remote closed the connection)
22:49:50 Quit FlynDice ("Konversation terminated!")
22:50:00stsquadIs it possible to have menu text with a dynamic parameter (as opposed to just messages with a dynamic paramter)?
22:51:05 Quit XavierGr (Nick collision from services.)
22:51:15 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:52:15 Quit domonoky (Read error: 104 (Connection reset by peer))
22:53:20bertrikjhMikeS, I changed it to use the proper setting in audiohw_preinit (off for e200v1) and I'm not hearing any pops on shutdown or stop/start playback
22:53:30 Quit krazykit (Read error: 104 (Connection reset by peer))
22:53:53bertrikI hear a rather silent pop on powerup, but that's before even the sandisk logo shows, so it's nothing we can do about
22:54:30 Join krazykit [0] (
22:54:51bertrikon the clip I do hear a pop when starting/stopping playback, but I never paid any attention to pops on the clip before now
22:55:10jhMikeSPreviously, if audiohw_close didn't close it properly, it would be a huge bang on repowering. Apparently it needed to discharge some caps.
22:57:34 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
22:57:47 Quit linuxstb (Nick collision from services.)
22:57:50 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
23:02:54 Quit linuxstb ("Leaving")
23:04:25 Quit jhMikeS (Nick collision from services.)
23:04:28 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
23:06:43 Quit petur ("Zzzz")
23:13:59 Join lazka [0] (
23:14:05 Part lazka ("cya")
23:18:04 Quit karashata ("G'bye everyone!")
23:23:05 Join baio [0] (
23:24:45 Quit BigBambi (Remote closed the connection)
23:28:31 Join BigBambi [0] (
23:35:28 Join MiniGod [0] (
23:35:48MiniGodis there an emulator for iPod?
23:36:06 Nick MiniGod is now known as _MiniGod (
23:37:43 Quit Thundercloud (Remote closed the connection)
23:37:58 Join herrwaldo [0] (
23:38:28bertrikkugel, I can commit the fmradio-i2c sources, but I'd like to hold off committing the other radio stuff until after the freeze because it may also impact the gigabeast
23:38:50rasherThe beast isn't a realease target either
23:39:14_MiniGodis there an emulator for iPod on windows?
23:39:28Bagder_MiniGod: we don't have emulators for anything, we have simulators
23:39:42_MiniGodsorry, emulator then.
23:39:48_MiniGodsimulators lol
23:39:55rasherBagder: We have FS #6524!
23:41:07Llorean_MiniGod: There's a pretty major difference.
23:41:26_MiniGodi dont know the difference
23:42:01_MiniGodah. now i found the simulators
23:42:13LloreanEmulators fake the hardware. Our simulators just look like the player, but underneath there's no fake hardware, they just use the PC like normal software does.
23:50:33 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
23:54:02 Quit bluebrother ("leaving")
23:57:12 Quit ender` (" A computer program will always do what you tell it to, and seldom what you want it to.")
23:57:33 Quit BigBambi (Read error: 113 (No route to host))

Previous day | Next day