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

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

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

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

#rockbox log for 2007-07-25

00:00:10BrandoNTgameboy roms arent being recognized!
00:01:37PaulJamBrandoNT: please be more specific. where/how are they not being recognized
00:02:25DerPapsthe can'T find them in the filebrowser
00:02:29 Part jack|Argon ("Ex-Chat")
00:02:36DerPapstthe dir where he put them is empty
00:02:37 Part UNKNOWNFILE
00:02:46amiconnjhMikeS: Hmm. There is an appropriate note in fat.c line 604f
00:02:48 Join rotator [0] (n=e@rockbox/developer/rotator)
00:04:40 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
00:04:56 Join barrywardell [0] (
00:05:16BrandoNT *bowing to DerPapst*
00:05:59jhMikeSyes. I'm guessing the in_use flag is supposed to mark it?
00:07:52 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:08:05 Quit Bagder ("It is time to say moo")
00:08:20amiconnTwo moos in one day... ;)
00:08:30 Join thegeek_ [0] (
00:09:23 Quit spiorf (Remote closed the connection)
00:09:44 Quit atsea- (Read error: 104 (Connection reset by peer))
00:10:17DerPapstactionscript + php + mysql = endless fun...
00:10:22 Quit kubiix (Read error: 110 (Connection timed out))
00:10:42 Quit jgarvey ("Leaving")
00:10:59Kasperleit involves 3 of the most hated web technologies, how can that be fun
00:11:17*amiconn only hates one of them
00:11:26 Join sarixe [0] (
00:11:32*DerPapst bets that is actioinscript :P
00:11:42amiconnIt's mysql
00:11:59DerPapstwhat do you prefer instead of it?
00:12:44 Quit sarixe (Client Quit)
00:12:54amiconnI didn't have to work with any of these myself yet, but judging from our devs at work, I would prefer postgresql
00:13:43*DerPapst has never heard of it
00:13:48amiconnAt least it has a well defined single license
00:13:54 Join atsea- [0] (i=ariel@gateway/tor/x-e2056375aa5c1fe1)
00:14:34DerPapstwhat do you work?
00:14:41amiconnAnd performance with large databases/complex queries is better
00:15:00DerPapstso no codemonkey
00:15:25amiconnWell, it sometimes involves coding (not much though)
00:15:34*amiconn likes coding
00:15:38Kasperlei recently discovered i hate coding
00:15:46DerPapsti guessed that already ;)
00:15:50Kasperleit's hard, boring and tedious to do it correctly
00:16:27DerPapstamiconn: which company are you working for?
00:16:55 Part maffe
00:17:09JdGordonKasperle: depends what your coding
00:17:11 Join maffe [0] (
00:17:13JdGordonand the hard part is the fun
00:17:52*jhMikeS hates coding too but likes the rush when the code runs and so the latter motivates the former
00:17:55 Quit BrandoNT ()
00:18:18KasperleJdGordon: i wonder. right now i'm working on a bit that can fail at about every 5th line of code, for which error handling adds at least another 5 each time
00:18:29*amiconn likes when the system does what he wants and hence likes coding
00:18:36amiconnEven if it's sometimes tedious
00:19:59 Nick BHSPitMonkey_ is now known as BHSPitMonkey (
00:20:30*jhMikeS never knows how he's going to get anything to work...ever. ususally it all starts with just a blank stare and "I have no idea what I'm going to do".
00:20:45*amiconn wonders what's wrong with the iriver website :/
00:20:49*DerPapst 's brain hurts from all these php sql and flash stuff
00:21:00amiconnurgh, flash
00:21:24DerPapstactionscript −−> flas
00:21:33*jhMikeS does at least now the pizza just got here :)
00:21:53DerPapstbon apetit
00:22:12amiconnWell, flash itself isn't half bad, but the platform support sucks
00:22:28Kasperleflash is grossly overused on the www
00:22:43DerPapsthopefully adobe does a better job on this
00:22:48krazykitflash is an abortion
00:23:00krazykitflash 9 for linux is ludicrously unstable :-/
00:23:32bluebrothermaybe that's the reason for my frequent firefox crashes?
00:23:37DerPapsti bet the adobe guys haven't modified it much after they got it from the macromedia dudes
00:23:42amiconnfirefox crashes???
00:23:45Kasperlei can view flash 7 somewhat reliably through different emulation layers. and then these websites laugh into my face saying "what, you don't have flash 9? download it, FOR FREE, from .... YOU IDIOT"
00:24:33krazykiti think the FOSS flash things are only supporting flash 7 right now, working on 9. it'd be better for them to just open the damn thing and let someone write a proper plugin.
00:24:34amiconnI only had firefox freezes so far, all due to the adobe pdf plugin (but solvable by killing acrord32.exe)
00:24:54bluebrotherfirefox crashes once in a while here on linux. Usually when I have a bunch of tabs open
00:25:02bluebrother(say, around 30 or so)
00:25:09pixelmamy observation is that a flash site at least causes an (imo) unusal increase of ram use of firefox
00:25:11amiconnThat's on windows, with 20+ tabs opened all the time
00:25:18bluebrothernever had those issues on windows.
00:25:39krazykiti don't use firefox in windows
00:25:44bluebrotherunfortunately quite a bunch of sites require flash :/
00:25:46krazykitk-meleon is far faster
00:25:48Kasperlei once had a runaway flash viewer consume 1.2GB of RAM
00:25:50 Quit thegeek (Read error: 110 (Connection timed out))
00:25:55DerPapston my berold laptop some stipd flash ads create about 100% cpu load ^^
00:26:01Kasperlethough it might've been the emulation wrapper's fault ;)
00:26:07krazykitDerPapst, learn the magic of flashblock :D
00:26:10PaulJamfor me just opening firefox causes an unusual ram usage :P
00:26:40Kasperleyeah, firefox gets annoyingly slow after a while, when i look it's eaten 300-400 mb of ram
00:26:54bluebrothereclipse is much worse in regards of ram usage
00:27:27 Join saratoga [0] (i=980398fe@gateway/web/cgi-irc/
00:28:25amiconnJust consumes ~110MB of ram here (35 tabs open, and running for more than a week)
00:29:39 Join TheVrolok [0] (
00:31:16 Quit Rondom (Read error: 110 (Connection timed out))
00:32:42*DerPapst was working on a simple flash guestbook :)
00:34:09 Quit barrywardell ()
00:34:42DerPapsti'm doing a traineeship (<−−"Praktikum") with the main stress on flash and sql
00:35:08 Quit XavierGr (Nick collision from services.)
00:35:10DerPapsti know nothing about php and sql :D
00:35:11 Join XavierGr [0] (
00:35:36DerPapstthe "result" (100% stolen)
00:36:48 Quit JdGordon ("Konversation terminated!")
00:37:38 Part salty-horse ("Leaving")
00:37:52DerPapst^ should be fash 6 compartible
00:38:20 Join barrywardell [0] (
00:40:53pixelmathen no flash? :P
00:41:10bluebrotherwhy flash for a simple guestbook?
00:41:39*bluebrother wrote a simple guestbook in php some years ago
00:41:53DerPapstthis is just as an example on how to access a sql database with actionscript
00:43:24DerPapsthave to so something similar in my traineeship once it starts.
00:43:24 Quit Siltaar (Remote closed the connection)
00:43:50DerPapstand i don't want to go there knowing nothing ;)
00:44:21 Quit Xerion (Read error: 104 (Connection reset by peer))
00:46:44*bluebrother wonders if there is a html version of the gpl available for download or only embedded in the fsf pages
00:47:25 Quit Lear ("ChatZilla [Firefox]")
00:47:26*DerPapst bets google knows the answer
00:48:39 Join iamben [0] ( has a gpl.html which is in fact the text version surrounded by <pre> tags :(
00:50:29 Quit saratoga ("CGI:IRC (Ping timeout)")
00:52:59 Quit barrywardell ()
00:53:00 Join lymeca_ [0] (n=lymeca@unaffiliated/lymeca)
00:53:00Kasperlebluebrother: maybe you can render the LaTeX or docbook document to html?
00:54:52 Quit TheVrolok (Remote closed the connection)
00:55:14 Join jhulst [0] (
00:56:14*DerPapst spots 2 new enties in his guestbook :)
00:58:32*DerPapst builds something to eat
01:02:16 Quit obo ("bye")
01:04:56 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
01:06:04 Quit entheh ("^~")
01:08:34 Quit lymeca (Read error: 110 (Connection timed out))
01:09:12saratogaregarding the comments made earlier today about MP3 verses Vorbis performance, i've noticed that with WMA as well
01:09:18***Saving seen data "./dancer.seen"
01:09:38saratogaWMA seems largely insensitive to bitrate, while MP3 performance scales very closely with it
01:09:57saratogaanyone familar enough with MAD to suggest a reason?
01:12:00amiconnDoes it?
01:12:35amiconnjhMikeS: do you know whether PP has a store buffer?
01:13:26*amiconn had the impression that mp3 is rather insensitive to bitrate
01:13:32amiconnOn coldfire, that is
01:15:06 Quit Nico_P (Remote closed the connection)
01:16:38jhMikeSamiconn: store buffer for (not familiar with the term)
01:17:09amiconnFor decoupled writing of data (write pipeline)
01:17:51jhMikeSah...yes it indeed seems to if you're referring to the cache
01:18:25amiconnNo, independent of the cache
01:19:34jhMikeSwhatever it has, it definitely bursts data to ram once it's filled up (seen as long horizontal spans of the display being updated suddenly)
01:19:50DerPapstGood night everyone :)
01:20:00 Part DerPapst
01:20:52amiconnI wonder whether the write buffer (if it exists) is enabled for the register area(s), and is the cause for the strange problems I am observing
01:21:59 Quit PaulJam (Read error: 113 (No route to host))
01:22:19amiconnRead-modify-write sems to cure them, but the bits I am masking are always zero, so it doesn't make much sense, unless the read-modify-write forces the chip to flush the queue and hence synchronise the operation
01:24:51jhMikeShmmm...I do see that done in emu dumps
01:26:16jhMikeSare you talking about CPU/COP_INT_EN bits and such?
01:26:54 Join t94xr [0] (
01:29:07 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
01:29:08jhMikeSyou said register areas so I'm figuring that included
01:29:38t94xris there a site that that can help me load rockbox on a 1st gen nano?
01:30:30 Join kubiix [0] (
01:31:44 Quit dudleif_ (Read error: 110 (Connection timed out))
01:31:53amiconnMaybe included, but not the area I am looking at atm
01:32:38jhMikeSI'm wondering if it's for mutexing purposes actually
01:33:45saratogaon ARM at least, doubling the MP3 bitrate seems to about half the decode speed
01:34:35*amiconn wonders if we ever figure out how to do ide dma on PP502x
01:34:52krylonsaratoga: fixed my issue
01:34:53amiconnThe product brief states it supports UDMA4
01:35:12 Nick lymeca_ is now known as lymeca (n=lymeca@unaffiliated/lymeca)
01:35:43amiconnjhMikeS: I'm not talking about atomic r-m-w either
01:37:29 Quit jhulst (Read error: 110 (Connection timed out))
01:38:10pixelmat94xr: select your player and the desired manual (pdf or online)
01:39:01pixelmait has installation instructions too
01:42:32jhMikeSamiconn: I dunno why they'd cache those control regs. I just see a situation where reading marks a register busy for a certain number of clocks or until a write cycle happens.
01:43:00 Join Arathis [0] (
01:44:32amiconnjhMikeS: What I mean is that with a store buffer, write(reg1); write(reg2) might not be executed in order
01:44:56Arathisamiconn: got the same error again. just another address: Undefined instruction at 00058D40
01:45:22amiconnBut a sequence like write(reg1); read(reg2); write(reg2) is always executed in order because the second write has to wait for the result of the read
01:45:49amiconnArathis: Yes, it also crashed here...
01:46:02jhMikeShmmm...possible. will a nop in betteen have an effect?
01:46:18amiconnBut I have some further changes meanwhile
01:46:55amiconnA nop (or a streak of nops) might or might not have an effect, depending on how long that streak is and what the other core does meanwhile
01:47:37jhMikeSdoes any read at all make it ok and then just assigning it?
01:48:18amiconnI'll try that, but I will probably have to use asm volatile for the read, otherwise gcc will optimise it away the inl/outl macros use volatile? they probably should at that.
01:49:11amiconnThey do
01:49:54amiconnThe question is whether one could rely on a statement like inl(blah) not being optimised away
01:50:01*amiconn doesn't really trust gcc here
01:50:43jhMikeSIt seems to be ok with that one
01:54:28Arathisamiconn: the freeze was after about 1,5h here. good I started a bat-bench ;)
01:55:00amiconnI got >4 hours of stable runtime in a later test
01:55:09amiconnStopped that test for further experiments
01:56:00 Join DC1 [0] (
01:56:33Arathiswow. that's nearly half the battery-lifetime of the H10/20gb with RB :)
01:57:20amiconnI just started another test which will hopefully run all night
01:57:50lukasmi installed the arm-gcc tool chain on my debian etch system, added it to the $path. but make fails saying arm-elf-gcc : command not found.
01:57:53lukasmwhere the prob ?
01:57:55*jhMikeS prepares some more goats
01:57:56amiconnUSB connected for charging (I got a crash before with power connected, hence that shouldn't invalidate the test)
01:58:22 Quit t94xr (Excess Flood)
01:58:38amiconnjhMikeS: I just implemented the dummy read instead of and'ing/or'ing for real:
01:58:52 Join t94xr [0] (
01:58:57amiconninl(0x600060a0); outl(unknown2, 0x600060a0);
01:59:12amiconnThe inl isn't optimised away
02:03:24amiconnIf this theory is correct, it might explain a lot....
02:03:44krazykitlukasm, did you log out and log back in?
02:03:58jhMikeSweird that there'd a write-behind on that sort of thing
02:12:23lukasmkrazykit: yes i did. but i had to log-in 2times to get path var set...
02:12:26amiconnPerhaps it's supposed to be disabled for the register area, but isn't on H10?
02:12:50amiconnRemember that I didn't get a crash on ipod mini G1
02:14:36amiconnIt might also be the reason for the complex clock setup routine in the H10 OF
02:14:51 Join jhulst [0] (
02:15:26amiconnPerhaps they prefer the write buffer enabled in general for better performance, adding some workarounds where it would cause gotchas?
02:16:07*amiconn would _really_ like to have access to a complete PP datasheet :(
02:16:09jhMikeSwould such a thing really help very much?
02:16:39amiconnWell, are the i2s registers located in the register area?
02:16:43krazykitlukasm, does echo $PATH include the directory?
02:16:58jhMikeSif we had that we'd reveal all the secret stuff put in there for the recording industry protection
02:17:12lukasmkrazykit: now it does ; )
02:17:25amiconnjhMikeS: Somehow I doubt that. The datasheet, not the fdk
02:17:40krazykitlukasm, then it should work :P
02:17:53jhMikeSamiconn: afaik. i thought all that qualified as "register area"
02:18:26jhMikeSmaybe the fdk hides certain details for that purpose?
02:19:00amiconnI'm quite sure the drm is software only
02:19:35amiconnMaybe using a unique hw identifier, but that's nothing new nor necessary to keep secret
02:19:54jhMikeSI mean, AS3514 has such things to help with DRM but we don't spill that. I just observe that in searching for docs, absence seem to be more common in media applications.
02:20:20*amiconn wonders what a dac has to do with drm
02:20:43jhMikeSwhatever they decide it should :p
02:20:55amiconnWell, you can see the crappy recording capabilities as some kind of rec industry protection...
02:21:29jhMikeSthat too...and a shame for this player
02:30:49 Part pixelma
02:30:57 Quit DC1 ("If Obi-wan ain't home then I don't know what the fsck we're gonna do. I ain't got no other connections on Tattooine.")
02:32:13 Join Administrator_ [0] (
02:32:29 Nick Administrator_ is now known as Lizzard (
02:32:42 Quit Lizzard (Client Quit)
02:33:12saratogaamiconn: I believe the AS3514 has all sorts of random stuff stuck on it unrelated to the dac
02:33:30saratogaIIRC the clock and poweroff stuff are accessed via it
02:33:40saratogaat least I think I recall Toni saying something about that ages ago
02:34:51jhMikeSsaratoga: it's a mishmash chip. its init needs to be a different than svn currently does it since it's treated right now like a bunch of separate devices.
02:35:00 Join JdG|uni [0] (i=82c20d68@gateway/web/cgi-irc/
02:42:50 Quit t94xr (Read error: 104 (Connection reset by peer))
02:48:30 Quit JdG|uni ("CGI:IRC (EOF)")
02:50:21 Join FOAD_ [0] (
02:52:28krazykitsaratoga, heh, apparently the wma encoder will play the audio streams of wmv video
02:52:49krazykitsaratoga, er, decoder.
02:53:15 Quit lukasm (Read error: 110 (Connection timed out))
02:54:35saratogakrazykit: yes linuxstb made it that way
02:54:52saratogaapparently it was pretty trivial to do since all WMA files are inside ASF containers anyway
02:57:57 Join jhulst_ [0] (
02:58:36krazykiti see. nifty
03:04:49 Quit FOAD (Read error: 110 (Connection timed out))
03:04:49 Nick FOAD_ is now known as FOAD (
03:06:09 Quit jhulst (Read error: 110 (Connection timed out))
03:09:05 Join sarixe [0] (
03:09:23***Saving seen data "./dancer.seen"
03:21:32 Join BigMac [0] (n=mike@
03:38:34 Join ACSpike [0] (n=Aaron@
03:38:45 Join jhulst__ [0] (
03:45:24 Quit miepchen^schlaf ("Verlassend")
03:51:25 Quit jhulst_ (Read error: 110 (Connection timed out))
03:52:16 Join wrobbie [0] (
03:53:56 Quit Arathis ("Bye, bye")
03:57:19 Quit Thundercloud (Remote closed the connection)
04:14:39 Join jw123 [0] (i=4511161b@gateway/web/cgi-irc/
04:15:10jw123anyone here with twiki write access?
04:19:33ACSpikeare there any manufacturers that are supportive of rockbox or condone its use?
04:20:53TiMiD[FD]jw123: yes
04:24:11Battousaii think either sandisk or iriver donated a couple players
04:24:21Battousaino sheets though
04:25:03TiMiD[FD]iriver ... I don't think so
04:25:17Battousaiyeah i think it was probably sandisk
04:25:59Battousaihere we go,
04:26:15ACSpikeI got the crazy idea to buy an mp3 player today. I'm looking at a sandisk.
04:26:42Battousairockbox is usable on the e200 (not e200r)
04:27:19Battousaino usb support though, so you need to use the bootloader and original firmware to put stuff on it
04:28:01Battousaiwhich is trivial, plug it into a usb port on your computer while it's off and it'll boot into the original firmware on its own
04:28:14Battousaii don't know how well other players are supported ;)
04:28:43ACSpikeyou use an e200 series player?
04:29:15 Join billytwowilly [0] (
04:29:15 Quit borisyeltsin (Read error: 110 (Connection timed out))
04:31:54ACSpikedual boot sounds like a win win
04:32:25iambenBattousai: is it normal that i have to unplug and replug the usb cable after booting the default fw to connect to pc?
04:33:00Battousaii find that it is, when the database needs rebuilding in the original firmware
04:33:11Battousaiotherwise it works correctly for me
04:35:41 Quit jhulst__ (Remote closed the connection)
04:36:47iambenbut that will be every time for me, as the only reason i ever boot the default fw is to transfer files..
04:42:02saratogalinuxstb: I've mostly gotten LSP working
04:42:14saratogatheres some minor distortion, but I think its mostly in the noise coding stuff
04:42:20saratogatheres some still some overflows in that area
04:42:24saratogaprobably fix them tomorrow
04:46:17 Nick basscade1 is now known as sneakums (
04:49:46 Quit saratoga ("CGI:IRC")
05:01:05 Join kubiixaka [0] (
05:02:23 Quit sarixe ("Peace")
05:09:26***Saving seen data "./dancer.seen"
05:19:53 Quit kubiix (Read error: 110 (Connection timed out))
05:28:37 Part ACSpike
05:36:39jw123TiMiD[FD]: do you have the capability to remove users?
05:37:14 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
05:40:32jw123or does anyone here have the ability to remove twiki users?
05:40:49scorchewhat are you after?
05:43:35jw123removing me
05:43:50jw123since no one ever responds to emails
05:44:15jw123yeah, sorry, that's what I meant was the address
05:45:07scorchewhy do you need to be removed?
05:45:28jw123tired of my google results...resume shaping...
05:45:49jw123as I haven't really did anything for rockbox
05:45:50scorchei dont quite understand you
05:46:44jw123resumé I mean
05:47:17jw123all I've done was a basic wps
05:47:51jw123basically, my twiki page on rockbox comes up on a google search and I don't want it to
05:50:18jw123is it hard to remove a user from the twiki?
05:50:33scorcheonly 3 people are able to
05:51:47wishesmust be applying for a job at MS if you're scared of them finding your name on open source stuff :D
05:52:03jw123yeah especially if it wasn't related to coding...
05:52:13*wishes grins
05:52:50jw123no one else 'googles' theirself?
05:53:03wishesyeah i do, but its mostly stuff about me
05:53:06wishesmostly correct
05:53:06scorcheeverything on there is either rockbox stuff or not me
05:53:41wisheseverything about me is pretty much lug mailing archives :/
05:53:56TiMiD[FD]that's nice !
05:54:06jw123company I work for does with potential employees; not saying I wanna dissociate myself from rockbox, it's just what I have attached to me now is something I wish not to
05:54:06TiMiD[FD]has everyone her done it ?
05:54:09scorcheTiMiD[FD]: still catching up? =)
05:54:19TiMiD[FD]what ?
05:54:42TiMiD[FD]"catch up" ?
05:55:07wishesjw123, my boss hires employees based on irrelevant stuff
05:55:34*scorche thinks of a way to explain this without using other slang
05:55:47scorche"getting up to speed"?
05:55:59wishesits fun. the last employee got hired because when asked by "bad guy" 'What do you see yourself as when you look in the mirror ?" (they play good guy bad guy with 2 people in the interview .. my boss being the good guy)
05:56:01TiMiD[FD]no :(
05:56:06TiMiD[FD]I don't get it
05:56:08wishesand she answered "A princess from out of space"
05:56:19wishesshe got the job pretty much there :)
05:56:34scorchegoing back and finding the stuff you missed so that you eventually come back to the present in knowledge?
05:56:42wishesits weird having another female developer though, and shes pretty smart too :)
05:57:09TiMiD[FD]ah I get it half :)
05:57:25jw123lol, if I was a female I'd probably be insulted
05:57:43TiMiD[FD]with your explanation and my english jp dictionnary ...
05:58:08wishesjw123, whys that?
05:58:10TiMiD[FD]but I found this page only this mornig ...
05:58:25TiMiD[FD]already ordered the chips
05:58:44jw123wishes, you know how sensitive women can be...
05:59:08wishesi know guys that put me to shame (im female/developer/sysadmin)
05:59:40wishestheres one guy who just refuses to speak to me
05:59:40*scorche gets out his "offtopic" knife
06:00:05wisheshey we were still talking about development ... if i throw in the word "beer" does it become on-topic ?
06:00:18jw123he's probably scared of you. I love rockbox (to appease scorche)
06:00:48wishesme too :D
06:01:36wisheshow can i help out with the project? :)
06:01:58TiMiD[FD]I love rockbox too, it keeps me awake during working hours
06:02:02scorchehow would you like to/are able to?
06:02:56wishesive been dealing with linux soley for about 8-10 years, so i know it pretty well. I can test, i can probably sort out minor stuff, though i freely admit im not a major c/c++ code monkey
06:03:12wishesi have a 60gb ipod video
06:03:24TiMiD[FD]do you know latex ?
06:03:26jw123"code monkey", that's funny
06:03:59scorchewishes: there is this coming up:
06:05:15wishesoh that sounds like fun :)
06:05:52wisheswhat time zone? generic US ?
06:06:34scorcheit doesnt really matter =)
06:06:51wishescool. usually im a day ahead (NZ)
06:14:04jw123well if one of the "big 3" come...please delete the twiki account pointed to by thanks.
06:14:17 Quit jw123 ("boxing rocks")
06:17:35 Quit Nibbier (Read error: 110 (Connection timed out))
06:18:21 Join Nibbier [0] (
06:30:05 Quit crwl (Read error: 110 (Connection timed out))
06:33:01XavierGrTiMiD[FD]: I did the mod just 3 days ago
06:33:08XavierGrI have to say it is perfect
06:33:14TiMiD[FD]ah !
06:33:17TiMiD[FD]glad to hear it !
06:33:32TiMiD[FD]I'm a little afraid of frying my iriver
06:33:34XavierGrNow I am waking to the sounds of my H140 every morning (eh I mean afternoon :P)
06:33:41TiMiD[FD]last time I opened it it seemed so small ...
06:34:14XavierGrTiMiD[FD]: it is hard to fry it, believe me, at my first try I installed the chip upside up instead of upside down :\
06:34:16TiMiD[FD]I would still use my clock
06:34:36TiMiD[FD]well but the solderings on the main board ...
06:35:00XavierGryeah it is a little bit difficult but with practise it should'nt be a problem
06:35:17XavierGrtry to get a small soldering iron
06:35:21TiMiD[FD]oh I almost forgot I don't have an iron pane here
06:36:00XavierGrmy previous one made the mod quite difficult to apply, so I bought a smaller one
06:36:27TiMiD[FD]did you get the chips by samples ?
06:36:40XavierGrI just tried the USB charging mod too, but it didn't worked on my H115 at all
06:37:05XavierGrTiMiD[FD]: ah I have to thank bluebrother for that, he got 2 chips as samples and mailed the spare one to me.
06:37:08TiMiD[FD]I registered this morning on their website
06:37:17TiMiD[FD]ah nice :)
06:37:28TiMiD[FD]I used my company name, but put my personnal adress ...
06:38:01XavierGrhehe cool :P
06:38:45TiMiD[FD]I hope I'll receive them :)
06:40:50*amiconn just got up and found the H10 still playing
06:41:02amiconn>4h 30min w/o crash :)
06:42:02TiMiD[FD]you only sleep 4h30 a night ?
06:45:23TiMiD[FD]when I sleep less than 7hours I become associable
06:46:14TiMiD[FD]did you discovered something to stop the PP crashes ?
06:46:49XavierGrTiMiD[FD]: I get that when I sleep less than 12 hours hehe (kidding)
06:47:47amiconnYes, although I don't understand it completely
06:48:31TiMiD[FD]8-9 hours is a best, less than 6hours for a long period of time >> anaemia and depression
06:48:34amiconnWell, sleeping rather long one night doesn't make it a problem to sleep only 4 hours the other
06:49:26TiMiD[FD]not sleeping enought reduces your life expectancy
06:50:23XavierGrI have the weirdest sleeping habits, some days I will max out at 12 hours and some days no sleep at all like today
06:51:48TiMiD[FD]that happens ...
06:56:48 Quit rotator ()
07:09:29***Saving seen data "./dancer.seen"
07:17:13 Join kubiix [0] (
07:25:36 Join davina [0] (
07:43:52 Join spiorf [0] (
07:48:24 Quit Rob2222 (Read error: 104 (Connection reset by peer))
07:48:31 Join Rob2222 [0] (
07:48:48 Join rigel [0] (
07:49:23rigelso the last month or so of h10 20gb updates have just stopped working intermittently, where i have to press the reset button before my unit will turn on
07:51:02TiMiD[FD]try a fresh build from today, amiconn just said he could run his h10 without crash for 4h30
07:51:26TiMiD[FD]maybe hididn't commited this to svn
07:51:43amiconnNo I didn't
07:52:08amiconnI first want linuxstb to test something on his ipod color
07:55:02 Quit billytwowilly (Remote closed the connection)
07:55:18 Join billytwowilly [0] (
07:55:32TiMiD[FD]ah that's the kind of modifications that could break other targets ...
07:55:43amiconnThe H10 is now running for 6 hours w/o crash
07:55:55TiMiD[FD]that seems solved ?
07:56:07TiMiD[FD]is it your best runtime ?
07:56:15amiconnI don't think it will break the color, but I want to get a clue what this register really does
07:56:33TiMiD[FD]the color uses the same cpu ?
07:56:58amiconnAnd the register is handled the same on all ipods but the minis
07:57:10amiconnEven for the later (PP5021/PP5022) targets
07:57:43TiMiD[FD]so how would testing on an ipod help explaining what the register really does ?
07:57:51TiMiD[FD](I'm just curious)
07:58:20amiconnSo either it's some kind of enable bit(s) which are just handled the same in order to not require ifdefing (in the OF source/ PP fdk, whatever), or they have some real meaning (lcd transfer speed/clock source?)
07:58:48amiconnThe color is the only target where bits 30 and 31 of register 0x600060a0 are set to 1 after boot,
07:59:09amiconnand the ipl wiki says it has to do with lcd setup
08:00:13amiconnSo what I want to test on ipod color is: (1) does rockbox work when setting these 2 bits to 0 all the time? (2) *if* it works, does it affect lcd speed?
08:00:27rigelcan i install it?
08:00:30TiMiD[FD]ah I see
08:00:37rigelim wondering if its just my battery
08:00:49TiMiD[FD]I thought you were still working on the cpu frequency/voltage control registers
08:00:52rigelbut i cant run batt bench to shutdown
08:01:25amiconnTiMiD[FD]: Yes I do. These bits have to be set to 0 during clock switch or the ipod color will crash sooner or later
08:01:39TiMiD[FD]oh ok
08:01:44TiMiD[FD]that's weird then
08:02:12amiconnWell it's not weird when they are some kind of clock select for the lcd controller
08:02:46amiconnPLL output is unstable during reprogramming, so it shouldn't be used to clock anything in that phase
08:03:24TiMiD[FD]fortunately ... I don't put my hands at a level that low :)
08:03:47TiMiD[FD]isn't that frustrating ? I mean to stay for weeks on the same problem
08:04:07TiMiD[FD]if a program I make doesn't work I feel upset so ...
08:04:22amiconnNot if it works in the end
08:05:00amiconnWhen writing the MMC driver for the Ondios, it took me almost 4 weeks to get the SH1 SPI to do what I want
08:05:04TiMiD[FD]yes when it works you can feel relieved :)
08:05:19TiMiD[FD]SPI ?
08:05:25amiconnOf course not 4 weeks fulltime
08:05:49 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
08:06:00rigelwhered the pitch control go?
08:06:02amiconnMeh, SCI, Serial Communication Interface
08:07:25TiMiD[FD]the joys of reverse engineering :)
08:07:31amiconnSynchronous and asynchronous serial
08:07:44amiconnNo RE on that part, the SH1 datasheet is public
08:08:00amiconnBut even public datasheets are sometimes not very precise
08:08:51*amiconn now realises why he confuses SCI and SPI all the time
08:09:28amiconnThe MMC supports SPI mode, which is what is used in the Ondios as they are hooked up to the SH1 SCI
08:09:30TiMiD[FD]yeah same as with APIs I'm often obliged to plunge into the code of the function to understand what it does on non-trivial programs
08:09:37TiMiD[FD]writing documentation is an art
08:16:02 Join toffe82 [0] (n=chatzill@
08:16:10 Part toffe82
08:22:56 Join kaaloo [0] (
08:23:20 Part kaaloo
08:39:05 Join Siltaar [0] (
08:39:30 Quit idnar (Nick collision from services.)
08:39:33 Join idnar_ [0] (n=mithrand@unaffiliated/idnar)
08:39:43 Quit pilled (Read error: 110 (Connection timed out))
08:44:05 Join Rob222241 [0] (
08:51:57 Join pill [0] (
09:01:49 Join RaRe [0] (
09:02:34 Quit Rob2222 (Read error: 110 (Connection timed out))
09:09:32***Saving seen data "./dancer.seen"
09:16:02 Join JessicaIsConfuss [0] (i=4122e228@gateway/web/cgi-irc/
09:17:35JessicaIsConfussHello everyone.....I was wondering if anyone could help me out. I installed Rockbox onto my Ipod Mini and it worked fine! But the following day it reverted back to the old Ipod firmware but the rockbox files are still on the ipod harddrive? Help?
09:17:58JessicaIsConfussOh and the light doesn't come on now after it reverted
09:19:24 Quit RaRe` (Read error: 110 (Connection timed out))
09:20:53JessicaIsConfussAnyone Alive?
09:32:52 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
09:32:56markunJdGordon: ues
09:32:59markunyes :)
09:33:06markuneh, that was for JessicaIsConfuss
09:33:08JdGordonhey all
09:33:17markunmorning JdGordon
09:33:27*JdGordon cant handle this working thing :'( im buggered
09:34:35markunJessicaIsConfuss: how did you revert the installation?
09:35:26JessicaIsConfussoh hello
09:35:31JdGordonjhMikeS: crap, I wanted to tell you not to commit, ill have to play moer but unless something was fixed its actually worse than before
09:35:39*JdGordon hasnt put lastest svn on yet though
09:36:05jhMikeSJdGordon: other stuff was addressed majorly. what was the last one doing?
09:36:14markunwhat makes you say it's worse then?
09:36:18JdGordonshowing an empty file browser
09:36:22jhMikeSthat's fixed
09:36:27JdGordonok good
09:37:08jhMikeSI was getting it. It simply needs an init retry after an insert init failure.
09:37:08JdGordonthanks for fixing the rec thing
09:37:50jhMikeSI created the problem way back so it was my responsibility :)
09:37:53JessicaIsConfussI was wondering if anyone could help me out. I installed Rockbox onto my Ipod Mini and it worked fine! But the following day it reverted back to the old Ipod firmware but the rockbox files are still on the ipod harddrive? Help?
09:38:03JessicaIsConfussIve searched on the web and can't seem to find an answer
09:38:10JessicaIsConfussshould I just reinstall?
09:38:30scorcheJessicaIsConfuss: hold menu + select and tell me what happens
09:40:33JessicaIsConfussumm...nothing happens Scorche
09:40:36scorcheJessicaIsConfuss: is the hold switch on?
09:40:56scorchehold those 2 buttons for a bit till something happens
09:42:09JessicaIsConfussa song started playing
09:42:32scorcheyou are in the apple firmware at the moment, correct?
09:43:00scorcheare you holding the menu and select buttons at the same time?
09:44:20scorchewell, you are doing something incorrectly, as this should repeat: hold the menu and select (the middle button) buttons at the same time for around 10 seconds or until something happens
09:45:23JessicaIsConfussit restarted
09:45:23JessicaIsConfussthe apple firmware still came up
09:45:44scorchethen repeat the process of installing the bootloader as described in the manual
09:45:48scorchedo you use itunes?
09:46:15JessicaIsConfussi have been using Yamipod
09:48:26JessicaIsConfusswhat is it supposed to do when I hold menu and the select button?
09:48:32scorchereset the device
09:49:39 Join morrijr [0] (
09:49:58scorchewell, the ipod wont just "revert"....soemthing had to have happened
09:50:19 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
09:51:15JessicaIsConfussI don't understand why it will not restart I am holding the buttons down
09:51:44scorchedidnt you just say it did it?
09:51:56JessicaIsConfussit did it once but it wont let me do it again
09:52:06JessicaIsConfussits posessed
09:52:09scorcheit will let you provided you do it correctly...
09:55:57JessicaIsConfusswell whatever should I just reinstall rockbox?
09:56:11scorche[00:45:39] <scorche> then repeat the process of installing the bootloader as described in the manual
09:57:00JessicaIsConfussoh right I saw that earlier
09:57:41 Join pondlife [0] (
09:57:44linuxstbamiconn: I read the logs, but am not sure what you want me to do - IIUC, bits 30 and 31 of 0x60060a0 are set to 0 in current SVN (unknown2 & 0x3fffffff).
09:57:58*jhMikeS wonders if one of those goats was satanic and now a excorcist is needed :P
09:58:04amiconnThey are set to 0 _duringt clock changes_
09:58:21pondlifeGood morning linuxstb...
09:58:24amiconnI would like to know what happens when they are set to 0 all the time
09:58:25linuxstbamiconn: Ah yes, I didn't look down..
09:58:33linuxstbpondlife: Good morning.
09:59:14amiconnFirst, if the color still works at all. If it does, whether these bits affect lcd speed (test_fps)
09:59:18linuxstbamiconn: So remove the outl at the end of the function?
09:59:32JessicaIsConfussthere is something wrong with my select button where it doesn't want to stay held down no matter how hard I push it
09:59:36JessicaIsConfussim not restarted
09:59:38amiconnYes, that's probably the easiest method
09:59:48JessicaIsConfussretarded rather....well maybe
10:00:04linuxstbOK, I'll just do an LCD speed test first, then change and do another test.
10:00:49*amiconn 's H10 is now running for 8 hours without a crash, ~3h 30min of that on battery
10:01:05amiconnSo this will be the first (coarse) runtime test as well
10:01:17amiconnI expect something around 5 hours
10:02:20amiconnNot that good given that iriver states 12 hours (although measured with 128kbps mp3), and we're doing disk poweroff on the H10
10:02:49amiconnlinuxstb: It may be that the difference in LCD speed is only visible when boosted
10:05:49 Join miepchen^schlaf [0] (
10:06:12linuxstbHmm, test_fps isn't building for me..
10:06:25linuxstb(linker errors)
10:07:10JessicaIsConfussScorche reinstalling the loader fixed the problem
10:07:22JessicaIsConfussI have no idea what caused the problem in the first place
10:07:25JessicaIsConfussThanks for the help
10:07:40markunJessicaIsConfuss: yes, it is quite stange
10:08:23JessicaIsConfussmight have something to do with my Yamipod client
10:09:18amiconnlinuxstb: oh?
10:09:21*amiconn will check
10:10:25amiconnHmm, builds fine for H10
10:10:25 Quit JessicaIsConfuss ("CGI:IRC (EOF)")
10:11:23*amiconn tries building ipod color
10:11:25linuxstbIt's a weird error. I've done a make clean and reconfigured, but get this:
10:11:28linuxstb/usr/local/arm-elf/lib/gcc/arm-elf/4.0.3/../../../../arm-elf/bin/ld: warning: cannot find entry symbol _start; defaulting to 00008000
10:11:28linuxstb/tmp/ccEIzjdA.o:(.header+0x8): undefined reference to `plugin_start_addr'
10:11:28linuxstb/tmp/ccEIzjdA.o:(.header+0xc): undefined reference to `plugin_end_addr'
10:12:26amiconnThis is with unmodified svn?
10:13:00linuxstbYes, only file changed is apps/plugins/SOURCES
10:13:17*linuxstb slaps self
10:13:36linuxstbI had added "test_fps" instead of "test_fps.c" to SOURCES...
10:13:59JdGordonwell... the thought was there anyway :D
10:16:12 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:16:45 Join obo [0] (n=obo@rockbox/developer/obo)
10:16:56*amiconn wonders what 'make' tried to do with this
10:17:33*jhMikeS wonders if amiconn avoided saying "what make tried to make of this". :P
10:18:47 Join ganny [0] (i=9088689c@gateway/web/cgi-irc/
10:19:11 Quit atsea- (Read error: 104 (Connection reset by peer))
10:19:16 Join voltagex [0] (
10:19:31voltagexI still can't get RBUtil to compile on OS X
10:20:59 Join crwl [0] (
10:22:37linuxstbFor the logs, boosted (78MHz) is 50.5fps/197.0fps and unboosted (30MHz) is 19.5fps/75.5fps
10:22:38JdGordonany objections to changing chopper so the high score is displayed on the bottom of the lcd instead of the top line if the screen is narrow?
10:24:08jhMikeSJdGordon: none. how bout that ata change, eh? :) everything ok?
10:24:36JdGordonhavnt tested yet
10:24:46JdGordonsansa is too far away atm
10:24:55*JdGordon too tired to get out of chair :p
10:26:29jhMikeScrawl if you must :)
10:26:38pondlifeAnd you must!! :p
10:26:55JdGordonthat was too hard :'(
10:27:04JdGordoni managed to reach my bag without getting out of the chair
10:27:07*jhMikeS gets out the long whip
10:27:12*JdGordon now has pins+needles in his foot
10:28:11*pondlife thanks Bagder for the mug *and* T-shirt
10:28:12*jhMikeS thinks JdGordon should give up sewing
10:28:32JdGordonnot sewing... voodoo doll torture!
10:28:37linuxstbamiconn: After commenting out "outl(unknown2, 0x600060a0);" , FPS is 15.5/61.0 @ 78MHz, and 15.5/60.5 @ 30MHz
10:29:02amiconnWow... so these bits actually do what I though
10:29:17linuxstbSo that's good news?
10:29:21pondlifeJdGordon: I wondered why my arm was sore today. Now back to the buffering you go!!
10:29:35amiconnWith them set, the lcd controller (interface?) is clocked from the pll, and them set to 0, from the 24MHz xtal
10:29:37jhMikeSbut what do the bits _do_? set the lcd clock source?
10:30:02amiconnlinuxstb: I bet you also get 15.5fps with plain svn and clock set to 24MHz...
10:30:03jhMikeSguess that answers that
10:30:15amiconn19.5/15.5 ~ 30/24
10:31:03amiconnOk, so we should not drop these bits
10:31:09JdGordonjhMikeS: ok, installing new svn now
10:31:20amiconnlinuxstb: I wonder what happens when we set only one of the bits...
10:31:39amiconnCurrently we only know what 11b and 00b are doing.
10:32:06linuxstbOK, I'll give it a go. 10b coming up...
10:32:33JdGordonjhMikeS: much better :)
10:32:36jhMikeSIt doesn't just set one of the clock sources just like the other regs?
10:33:05jhMikeSJdGordon: awsome. no problem with abusing the insert?
10:33:20JdGordonnone yet
10:33:27amiconnI think it selects one of 4 clock sources, but which setting selects which source?
10:33:37*amiconn has a suspicion though
10:33:47jhMikeSI already figured that out and needs to refresh
10:34:20linuxstbamiconn: Am I right to just change the final outl to "outl(unknown2 & 0xbfffffff, 0x600060a0);" ? (my brain still isn't fully-functional after DevConPub2...)
10:34:38JdGordonjhMikeS: when the card is inserted but doesnt get shown, is that a card failure? or something just happens too quickly?
10:34:46amiconnThat should work
10:35:03amiconnYou can check the value of the register under 'View I/O Ports'
10:35:27*JdGordon doesnt like this comment... /* NOTE: This init isn't dual core safe */
10:35:34JdGordonwe are never going to go dual core are we? :'(
10:35:44amiconnI called it CLOCK_0xA0, not knowing its real purpose
10:36:07jhMikeSJdGordon: 1) it is a card failure 2) we sure as heck are going to do it. I'll do it sansa only if I have to.
10:36:20amiconnnooooooo ;)
10:36:28JdGordonthats what i like tot hear :)
10:36:36amiconnjhMikeS: Really, we should try to get it working on all PP targets
10:36:43pondlifenoooooo? Why not rrrrr?
10:36:45linuxstbamiconn: It froze at boot - I'm still staring at the Apple logo...
10:36:50amiconn(that includes PP5002)
10:36:59 Join Arathis [0] (
10:37:07jhMikeSwe'll that's what I want but I can't stand having a whole core going unused outside mpegplayer
10:37:14amiconnlinuxstb: Okay, so probably 10b selects an inactive clock source
10:37:35linuxstbTrying 0x7ff... now
10:37:39JdGordonwe dont have a filesize limit for the bootloaders do we?
10:37:58JdGordon... I mean, one we are in danger of passing
10:38:12jhMikeSif we has a datasheet the bits would probably be shown as "show apple logo"
10:38:20linuxstbJdGordon: Which bootloader? I expect there's a limit on iriver h1x0/3x0
10:38:31pondlifeJdGordon: File size, or memory size?
10:38:36JdGordonfile size
10:38:57JdGordonlinuxstb: hmm, yeah, I'm starting to remember something around 64k for our code?
10:39:01amiconnNot really unexpected... when I first tried setting the various core sources MrH figured out on my iPod, 3 of the 8 possible settings just hung
10:39:04JdGordonbut no problems on pp?
10:39:17linuxstbamiconn: Second test hung as well...
10:39:19amiconn(actually 4, but one was expected to hang without further measures)
10:39:36amiconnOkay, so at least I can document the register.
10:40:00amiconnAnd we should continue handling it, otherwise we'll use lcd performance which would be bad
10:40:01jhMikeSJdGordon: are you saying the SD card is not showing up alot or only during abuse?
10:40:45amiconnlinuxstb: Thanks again for testing
10:41:25JdGordonjhMikeS: both (?!), it seems if I put one in and its detected, then swap to my other one it will take a few tries for it to be shown
10:41:25linuxstbamiconn: You're welcome. Let me know if you need anything else - I'll be around all day, but am going out this evening.
10:41:51JdGordonbbs dinner
10:43:14TiMiD[FD]what's the difference between the bitmaps in bitmaps/native and bitmaps/mono ?
10:43:29TiMiD[FD]if I use a bitmap in native/ for the archos it works
10:43:42TiMiD[FD]and mono produces garbage
10:44:23*amiconn wonders whether 0x600060a0 will help speeding up lcd update on other pp targets
10:44:45amiconnThe minis don't handle the register in their clock change routine, so it probably won't help there
10:44:58amiconnBut the nano and video do
10:45:01jhMikeSe200 doesn't use it
10:45:33amiconne200 is probably rather special with its lcd dma
10:46:04jhMikeSnothing else can use lcd dma?
10:47:49jhMikeSthe use of flush_icache doesn't seem to be needed. invalidate_icache seems sufficient to keep the screen coherent.
10:48:58*jhMikeS tries the scaling again
10:49:13amiconnLooking at the lcd base addresses it might help on H10
10:50:13amiconn...on nano, and _perhaps_ on the greyscale ipods
10:50:39amiconnI'll test on H10 after finishing this (coarse) runtime test
10:50:41jhMikeSI hope this works and is fast since that should be the ticket for coherency between cores then...
10:51:24amiconnShouldn't take long since there are only 16% battery left now
10:52:24jhMikeSthe OF does an invalidate_icache when doing SD I/O so it seems promising but I have to wait though a full rebuild now :\
10:53:30*amiconn should try these bits on mini g2 meanwhile
10:54:19amiconnlinuxstb: I think we should also set them explicitly in our lcd init (after naming the register properly in pp5020.h)
10:55:01*amiconn would like to see a test on a nano as well
10:55:06linuxstbTiMiD[FD]: bitmaps/native are drawn with the normal lcd_bitmap function, and bitmaps/mono are drawn with the mono bitmap functions. So on Archos it shouldn't matter.
10:56:51*jhMikeS wonders how much trouble using DMA for audio would be...will attempt when the mood strikes
10:57:35TiMiD[FD]then I put everything under native
10:58:46Arathisamiconn: Are you already testing a new build? Or is the runtime test on another player?
10:59:35jhMikeSargh, still clicks
11:01:06pixelmaJdGordon: the highscore on the bottom is a bit irritating on archos - expecially because it's unnecessarily wide off of the screen's border
11:01:49pixelmaand I think I'd prefer it on the right side - but hard to tell
11:02:59pixelmaor maybe it was better when it disturbed in only one place so to speak (i.e. the top)
11:03:17amiconnArathis: A new build. It runs stable for >9 hours now, >4.5 hours of that on battery
11:03:45amiconnjhMikeS: Do you force a fifo refill before invalidating the cache?
11:04:14Arathiswow. great news
11:04:30jhMikeSI can do that but the solution is rather ugly. It only needs to have the thresholds changed and then reset.
11:05:08amiconnYes, but if it works...
11:06:09jhMikeSit does and needs no concern for pcm state. I should mention though that 30mHz is just unusable. 40 seemed better.
11:07:48 Quit ganny ("CGI:IRC (EOF)")
11:08:21 Join DerPapst [0] (
11:08:28DerPapstgood evening at all :)
11:09:22DerPapstamiconn: do you still have LinusN's iPod? If not i could probably do a test on my 5.5G iPod.
11:09:32amiconnYes I do
11:09:35***Saving seen data "./dancer.seen"
11:09:42amiconnI don't expect an effect on the video though
11:09:49DerPapstok then nvm ;)
11:13:21jhMikeSamiconn: I wonder if just writing some dummy data to a buffer might work...just enough to flush out the lines
11:14:04Arathisamiconn: If you have a build for me I could do a longtime test too. I'm away from the pc and player anyway the next eight hours
11:14:27amiconnIf you have a few mins I can make one
11:15:42pondlifeIs it well-known how much the iPod Video list scrolling (probably screen update) is improved when CPU is boosted? I saw it for the first time at DevConPub2 last night and it was very obvious.
11:17:44amiconnlcd update gets faster with cpu clock on all targets
11:17:56Arathisamiconn: thanks
11:18:52pondlifeamiconn: Yes, but this made the UI so much better, I really couldn't have imagined it. Just my first impression of a Rockboxed iPod....
11:20:10midkaypondlife: me! me! i definitely know the slow scrolling issue. i used to always go and boost the CPU when i turned my iPod on. (i later learned to just live with it..)
11:20:45amiconnScrolling is pretty okay imo since the multiple lcd update in menus got fixed
11:20:52JdGordonpixelma: oh ok, its just that with any font other than the default it was cut off on the sansa and very irritating
11:21:15pondlifepretty okay, but noticably slower than the OF still?
11:21:32*pondlife has little iPod experience.
11:21:34midkayamiconn: now the menus are better, yeah. (same speed as all other lists) but still overall the UI is rather slow, i think.
11:21:51pondlifeI was certainly happy to have an H300...
11:21:56midkaydefinitely noticeably slower than the OF. (except sometimes when playing music, the UI in the OF gets rather sluggish)
11:23:24amiconnpondlife: On H300 the difference is also quite noticeable
11:23:45amiconnThe only targets where the difference is practically unnoticeable are the non-color ones
11:23:58amiconnThe H10 just died from low bat. No crash
11:24:15amiconn4h 50min battery runtime
11:24:42pondlifeHmm, no crashing but reduced runtime?
11:24:51DerPapstnice :)
11:25:03pondlifeIs CPU scaling enabled?
11:25:07amiconnNot reduced... normal for a rockboxed PP5020 target
11:25:24DerPapstpondlife: yes
11:25:28pondlifeAh, ok. I thought a Rockboxed H10 was nearer 6 hours.
11:26:09amiconnDepends, maybe the 20GB versions run for 6 hours
11:26:25DerPapstthe test was all aboud clock changing :)
11:26:33amiconnWithout cpu scaling, it's very likely less
11:26:37ArathisH20 can run about 8 hours
11:27:18JdGordonwhat ever happened with the idea of putting all the pp targets under a single subfolder under arm/ in the TT?
11:27:23amiconnSetting the lcd clock bits has no effect on mini g2
11:27:45linuxstbThere's only one H10 test on the IriverRuntime page - a 6GB that ran for 4h 02m (128kbps VBR MP3)
11:28:07amiconnSo my test shows the improvement
11:28:23amiconn4h 50m (~200kbps VBR mp3)
11:28:36amiconn(I mean the improved runtime with clock scaling)
11:30:53linuxstbiriver appear to be claiming "up to 12 hours" for the 5GB and 6GB...
11:31:27amiconnwith 128kbps mp3
11:31:29ArathisI had ~13h with the OF on my 20GB if I remember correctly
11:31:42Arathiscould've been 15 though
11:33:04ArathisiRiver says 16h for the 20GB
11:34:52JdGordonpondlife: any reason you havnt commited 6266?
11:34:56Arathisamiconn: Is the build ready soon? Otherwise I'd have to go now
11:35:07amiconnYes, just uploaded
11:35:12amiconnSame URL as yesterday
11:35:25 Quit Bjoern-Erik ("leaving")
11:35:29JdGordonLlorean: you round?
11:37:22 Join Bjoern-Erik [0] (
11:37:43scorcheJdGordon, others who havent been able to attend the 3 meetings:
11:39:18JdGordonscorche: random number, between what?
11:39:49scorcheJdGordon: as in i assign a number to each wanting one, and see what numbers are generated and how they correlate..
11:39:58linuxstbamiconn: ?
11:40:22JdGordonumm... ok
11:40:23amiconnLCD update speedup on H10 by setting the lcd clock select bits to PLL: 10% at 30MHz, and 80% at 78MHz
11:40:35DerPapstwow indeed
11:40:40scorcheJdGordon: just say you want one and I will take care of it =)
11:40:57*amiconn wonders if Arathis will be around for long enough to test that on the 20GB
11:41:16amiconnMaybe it's better to do that tonight though, after testing clock change stabiliy...
11:41:28pondlifeJdGordon: Just because I was hoping the sim pause could be fixed first (FS #6267). That would make for a cleaner fix here.
11:41:31DerPapstthen start compiling fast like the wind :D
11:41:40JdGordonpondlife: ok
11:41:55pondlifeI'll resync, and might commit it anyway..
11:42:04amiconnNow if we also apply the special ipod color trickery involving the word count....
11:42:23Arathisamiconn: no. I'M off in about a minute ;)
11:42:38JdGordonpixelma: coo-eey..
11:43:13amiconnEven with the simple clock bit change, jewels suddenly becomes fun to play during music playback! :)
11:43:30JdGordonpixelma: how would you prefer chopper fixed?
11:43:51JdGordonjust put it back the way it was for wide screen targets?
11:44:40 Quit Arathis ("Bye, bye")
11:44:42JdGordon... or maybe just hide high score if the screen isnt wide enough?
11:46:26*amiconn wonders whether there's any other H10 20GB owner than barrywardell and Arathis
11:47:01pixelmahmm... maybe that - it could be displayed in the menu like in jewels (though that has a flaw on Archos too but that's a different story)
11:47:03*JdGordon sees if he can find one in another channell..
11:47:37JdGordon"maybe that" to which one?
11:48:04pixelmanot displaying the highscore in the game screen
11:48:13 Join Nick_Brackley [0] (i=7dff0332@gateway/web/cgi-irc/
11:48:20linuxstbTiMiD[FD]: You shouldn't add new functions to the middle of the plugin API struct, especially without bumping the plugin API version. They are normally added to the end of the struct and the PLUGIN_MIN_API_VERSION incremented.
11:48:31pixelmawasn't alienbiker99 an H10 owner too?
11:49:05 Quit pill (Nick collision from services.)
11:49:09pixelmaJdGordon: if that doesn't work, I think I prefer the "old way" on Archos at least
11:49:19 Join _pill [0] (
11:49:35JdGordonI dont think showing the high score in the menu is very good
11:49:50linuxstbTiMiD[FD]: Remove "_MIN" from my previous sentence...
11:50:42amiconnThat's a big no-no in fact
11:50:59TiMiD[FD]linuxstb: I know, but I forgot ...
11:52:04XavierGrmarkun: any idea on the USB charge on H100? :P
11:52:25TiMiD[FD]I made the commit before leaving work ...
11:52:56TiMiD[FD]anyway I'll probably recommit since a wrning popped out
11:53:32amiconnlinuxstb: 50fps->90fps fullscreen isn't bad, eh? :)
11:53:56TiMiD[FD]but for the plugins, always adding at the end of the structure makes the struct more and more messy
11:54:18amiconnYes, but it keeps backwards compatibility for a while
11:54:34JdGordonand it gts cleaned up when a commit is forced to break backwards compa
11:54:38TiMiD[FD]binary compatibility
11:55:06amiconnIf there is a necessary incompatible change, the min api version needs to be bumped, and whoever does that is required to sort the api
11:55:16TiMiD[FD]it gets cleaned when someone makes such a commit ... if the guy does it
11:55:20amiconnThis way it cannot become too messy
11:55:35TiMiD[FD]for example if I make changes to the screen api
11:55:40TiMiD[FD]which is included into the plugins
11:55:42amiconnThe comments in plugin.h clearly state what to do
11:55:49TiMiD[FD]it won't be binary compatible
11:56:12TiMiD[FD]even if I didn't changed anything into the plugin.h
11:58:15amiconnThen that rule applies too, as you changed the plugin api
11:58:32amiconn...even without changing a line in plugin.[ch]
11:59:39amiconnSo to fix that, either the new functions should be moved in another commit and the api version bumped, or both versions need to be bumped
11:59:52TiMiD[FD]I'll do it in my next commit
11:59:52amiconn(and remaining unsorted functions sorted)
12:00:18TiMiD[FD]but I wonder if it's still binary compatible after the compiler makes some memory alignments
12:00:32 Part maffe
12:00:45 Join maffe [0] (
12:01:43markunXavierGr: nope
12:04:38JdGordonpixelma: mind trying a patch before I commit?
12:04:57JdGordonive got it only displaying the highscore if the current score is < and if both fit on the screen
12:05:52amiconnHmm. It should be possible to achieve 118fps on H10 small and 94fps on H10 large
12:05:58linuxstbamiconn: I was wondering why mpegplayer seemed relatively slow on the H10 (compared to other PP targets) - e.g. the Nano can do 128x128 at 42fps, but the H10 could only manage 29fps. I guess you've found the reason...
12:06:10pixelmaI'll take a look - but you did shorten "highscore" and so on as before?
12:06:17amiconnlinuxstb: Hmm, so the nano has these bits set?
12:06:23amiconn(by the loader I mean)
12:06:34JdGordonpixelma: hmm, ill put that back in and then do a diff.. 2 min
12:07:03linuxstbamiconn: The Nano seems to perform almost the same as the Color in mpegplayer, so I would guess it does. I don't own a Nano though.
12:07:14amiconnHow many fps on nano?
12:07:24pixelmaJdGordon: is it again at the top?
12:07:40*amiconn would expect around 82
12:07:44amiconnI mean in test_fps
12:07:47JdGordonpixelma: yes
12:08:14linuxstbNo tests listed on LcdFrameRate...
12:08:30JdGordonpixelma: dcc?
12:08:38pixelmashould work
12:09:00amiconnAfaik Llorean does have a nano
12:09:06linuxstbYes, he does.
12:09:32scorchei do as well...just add test_fps to the sources or logf?
12:09:34*amiconn would like to know the CLOCK_0xA0 on nano
12:09:51linuxstbscorche: Yes, add it to sources, and then add your results to
12:10:14scorchereset settings as well, or shoudl that not matter?
12:10:39amiconnscorche: Just check 'View I/O ports' first, whether those bits are already set
12:10:53amiconnSettings don't matter for test_fps
12:11:30*amiconn guesses they're set though
12:12:18scorchei get C0000000 for CLOCK_0xA0
12:12:30amiconnAh, so they're set
12:12:46amiconnGood to know
12:12:56scorcheshall i still run test_fps?
12:13:37*scorche does it anyway for the wiki's sake
12:13:44linuxstbYes, it would be nice to have Nano measurements in the wiki.
12:13:44pixelmaJdGordon: the highscore wasn't shown in the game - though the splash said it was one in the end...
12:14:19JdGordonhighscore isnt shown if the current score is better
12:15:53pixelmaok - misunderstood but then I should see it if I start now and the score is less? If so it doesn't work, they are never shown
12:15:56TiMiD[FD]yaaa perfect a green line
12:16:02TiMiD[FD]now I can go to my date
12:16:09TiMiD[FD]15 minutes late >>>>
12:16:25DerPapstuh ohh :P
12:16:28pixelmaJdGordon: hmm... somehow it worked when I restarted a second time
12:16:47JdGordonyeah, it will show the hiscore untill the current score is better
12:17:00JdGordonsame here, but i think thats a bug with the sim saving?
12:17:02*JdGordon checks
12:17:59pixelmaright, seems to work now (and I tried in a sim) - testing on my Ondio but that'll take a bit longer...
12:19:27JdGordonalso, fixing the bug where resume game is selected when you get put back in the menu after dying
12:20:06linuxstbTiMiD[FD]: When adding to the end, you need to increment PLUGIN_API_VERSION, not PLUGIN_MIN_API_VERSION - as the comments describe. Your commit made the min API version greater than the current API version...
12:21:53 Join herbi_ [0] (
12:22:24amiconnlinuxstb: lcd_send_lo and lcd_send_hi are named wrong in lcd-color_nano.c
12:22:32herbi_hi guys, i installed rockbox on my ipod mini works fine just one quick q, if i have several directories full of mp3s in the root dir of the ipod, how can i tell rockbox to shuffle through the whole thing, atm i can only get it to shuffle in subdirs?
12:22:49amiconnIn fact lcd_send_lo means lcd_send_cmd and lcd_send_hi means lcd_send_data
12:23:10amiconnAnd the 2 different lcd types are actually 16bit and 8bit hookup
12:23:26scorcheherbi_: make a playlist of all the music and turn shuffle on
12:23:31JdGordonlinuxstb: TiMiD[FD]: if the screen access struct is modified then the PLUGIN_MIN_API_VERSION should probably be bumped anyway
12:23:45herbi_really? it cant handle just shuffling on its own?
12:23:49herbi_thats a bit retarded.
12:24:04pondlifeherbi_: Use the database?
12:24:14pondlifeSelect <All tracks>...
12:24:17linuxstbamiconn: Anything you can do to clarify that driver would be welcome...
12:24:22amiconnAnd I can probably reuse almost all code from there to improve the H10 lcd driver (8bit hookup)
12:24:46amiconnBad thing is that I can't test H10 large myself
12:24:59herbi_i cant believe it cant handle shuffling on its own
12:25:24herbi_tempt me to switch back to normal firmware
12:25:28pondlifeherbi_: It can, read the manual about database
12:25:35JdGordonherbi_: enable shuffle mode then do a recursive insert on the root of your disk and it will do what you want
12:25:59pondlifeYes, that's even simpler
12:26:21pixelmaJdGordon: seems to work correctly on my Ondio
12:26:56scorcheUGH...thanks to me updating with TiMiD[FD]'s commit, i get incompatible version...
12:27:02*scorche dives into plugin.h
12:27:29pondlifescorche: Looks like TiMiD[FD]'s gone... are you going to fix it up?
12:27:53scorchepondlife: i havent gotten commit access yet
12:28:03herbi_JdGordon, sorry i dont understand
12:28:11JdGordonread the manual then
12:28:12amiconnlinuxstb: How did you find out about the i/o count registers, btw?
12:28:14herbi_pondlife, so the database has to be rebuilt every time i change the tracks?
12:28:25amiconnOr is that also borrowed from ipl?
12:28:48linuxstbEverything is simply taken from ipl.
12:28:53herbi_well, rockbox is cool and shit, but it doesnt really do anything special that would convince meto take it
12:28:55scorchepondlife: i can give you a patch if you wish, but it is quite simple...
12:28:57 Part herbi_
12:29:28pixelmaa patient one...
12:29:35JdGordonhis loss....
12:29:40pondlifeWell, not mine
12:30:05*JdGordon puts out another call for anyone who knows how to pass data from the bootloader to rockbox?
12:30:14 Join antgel [0] (n=antony@pdpc/supporter/monthlybyte/antgel)
12:30:23JdGordonnot all targets have
12:30:25linuxstbJdGordon: There's no mechanism to do it afaik - so you need to invent one...
12:30:58JdGordoncan it be passed as an arg to main()? or does that involve yucky asm to fiddle with the stack?
12:31:20amiconnlinuxstb: Didn't you optimise the color lcd update?
12:31:28amiconnMaybe I don't rc...
12:31:30linuxstbNo, it was the video update I worked on.
12:31:44amiconnI mean the pixel count stuff
12:31:46 Quit thegeek_ (Read error: 110 (Connection timed out))
12:32:13linuxstbI may have tweaked it, but the code was definitely already there in a very similar form in IPL.
12:32:30*linuxstb goes to see what the original looked like
12:34:11linuxstbAll I changed was the 0x10000 - it was 64000 in the original code.
12:34:57linuxstbSo I wondered why it was such an odd value, and increased it slightly.
12:35:04antgelhi all, is there a page indicating sanity of builds? latest appears totally insane on my H300, but it's so long since i downloaded a bad build, i'm wondering if i might have done something wrong :)
12:35:18*amiconn should probably start a PP5020 wiki page
12:35:43linuxstbantgel: Insane in what way? It's known that all plugins are broken...
12:35:44amiconn...documenting all the stuff I figured out so far (includes some MrH knowledge)
12:35:58SlasheriJdGordon: IRAM on eeprom on H100
12:36:07Slasherieeprom is actually used with flashed rockbox
12:36:13antgellinuxstb: i can't listen to any music - it hangs. let me try with a fresh .rockbox directory
12:36:19DerPapstor extend iPLs wiki page :)
12:36:31linuxstbIs their server up at the moment? ;)
12:36:36JdGordonSlasheri: im hoping for a solution which would work on every target
12:36:46DerPapstno :'(
12:37:00DerPapstoh it is :D
12:37:01SlasheriJdGordon: well, iram sounds quite "global" solution for me :)
12:37:25linuxstbJdGordon: Just put a struct at a fixed location in RAM (SDRAM or IRAM, it doesn't matter), and read the values before that memory is overwritten by Rockbox.
12:37:51amiconnDerPapst: I won't touch the ipl wiki
12:38:08DerPapstwhy not? too messy?
12:38:17scorchepondlife: if you want to commit it...">
12:38:20JdGordonlinuxstb: in apps? how?
12:40:03DerPapsti hate it when windows refuses to start the task manager :-/
12:40:13linuxstbJdGordon: Just do a memcpy. Maybe the end of SDRAM would be a suitable place - IRAM will be overwritten by the startup code (copying icode functions there, clearing the ibss etc).
12:41:01JdGordonoh alright
12:41:30DerPapstquit "brb... reboot :-/"
12:41:32 Quit DerPapst ("So Long And Thanks For All The Fish!")
12:41:54amiconnSomewhat messy, yes, and the server is often down
12:42:11*amiconn still wasn't able to find the ipl svn
12:42:24amiconnI only know about the cvs, but that seems outdated
12:44:43pondlifescorche: Thanks, committed. Hope it was right!
12:44:55 Join tucoz [0] (n=tucoz@rockbox/staff/tucoz)
12:45:09JdGordonlinuxstb: how do I find out the end address of the sdram?
12:45:13scorcheit is only wrong if he broke the api, whhich it dodnt sound like it..
12:45:32tucozbluebrother, i saw that you were looking for a html gpl. In case you didn't find one already, check this page out:
12:45:48JdGordonlinuxstb: is that the DRAMSIZE define?
12:45:50amiconnlinuxstb, JdGordon: Why not just pass a parameter (in a register)?
12:46:06tucozbluebrother, you could simply copy the gpl that is shown on that page, using the html source.
12:46:15antgellinuxstb: this is too weird. some of my oggs play and some don't. they alll play with factory firmware. i want to try them with a rockbox sim as i'm not convinced my hardware is okay :|
12:46:19linuxstbtucoz: That's GPLv3 ;)
12:46:21scorchelinuxstb: what shall i put in "LCD Driver Revision" as all this is pre-svn?
12:46:29tucozlinuxstb, grrrrrr
12:46:39linuxstbscorche: I guess just put the SVN revision.
12:46:39JdGordonamiconn: I would if I knew how to
12:47:02tucozthis one then:
12:47:20linuxstbscorche: lcd-colornano.c says r12574 in the header, so you could quote that.
12:47:34scorchelinuxstb: so according to should i put 12574, ot the current revision that it was built under?
12:47:36linuxstbJdGordon: What information do you need to pass anyway?
12:47:39scorcheoh...alright =)
12:48:26scorchelinuxstb: the bootloader version afaik
12:49:27JdGordonyeah, some identifier from the bootloader, the whole version string would be nice... but not needed
12:49:34amiconnSo the H10 is in fact very similar to the ipod color (smaller display but with radio)
12:49:38pondlifelol - "Build should have been done 0m 45s ago"
12:50:03pixelmanothing new
12:50:40pondlifeI've not seen that one before..
12:50:47scorchelinuxstb: is there a special way of getting the info for the other MHz? or should i just change the boost counter?
12:51:04linuxstbYou need to manually change the boost counter.
12:51:53*amiconn thinks the plugin could set the clock itself
12:52:00amiconnAt least boosted/unboosted
12:52:10amiconnidle might be a problem
12:52:17markunantgel: are some of the files really old?
12:53:25linuxstbantgel: Also, do any of them have odd tags in them (e.g. id3v2 at the start) ?
12:53:50 Join DerPapst [0] (
12:54:14antgelmarkun: yeah, but the old ones play fine, it's my newly-ripped ones that don't work. also it appears that it could be something to do with non-ascii filenames or tags. the issue is weird, i haven't pinned it down yet. i'm not being very scientific
12:54:56*amiconn wonders why lcd_yuv_blit() for ipod color is not yet asm optimised
12:55:15scorchelinuxstb: and the transfer rate?
12:55:24antgellinuxstb: could well be - i think they have id3v2 tags as grip is not set to "Only tag files ending in '.mp3'"
12:55:33linuxstbAh, grip again...
12:55:37amiconnscorche: The transfer rate is a simple calculation
12:55:41linuxstbantgel: Yes, that will be the problem.
12:55:45DerPapstamiconn: iPLs svn is down. and systechs doesn't reply to any emails. wo don't have any access to the server.
12:56:15amiconnSo there's only the old cvs?
12:56:22linuxstbantgel: Do you have the "id3v2" utility installed?
12:56:47amiconnDoesn't seem to matter much as kernel development seems to have stopped long ago...
12:56:50antgellinuxstb: okay, at least you know the issue. :) i've been using grip since the year dot, and that setting has been like that for a long time, it all worked so i never changed it
12:57:06DerPapstone of iPLs core devs sets up a new server so we can put everything up there but i still don't have a idea how to get all databases off the server without any access :(
12:57:16linuxstbantgel: Something like "id3v2 -D *.ogg" should remove those id3v2 tags.
12:57:29antgellinuxstb: no, but i can grab it from the debian repos
12:57:31DerPapstamiconn: there is a little outdated svn availabe too
12:57:34DerPapstwait a sec..
12:58:07antgellinuxstb: ISTR that .ogg doesn't use id3, right? in which case the default grip setting seems a little insane
12:58:14linuxstbDerPapst: Has the IPL kernel moved to SVN?
12:58:28linuxstbantgel: Yes, it is insane...
12:58:39DerPapstlinuxstb: no
12:58:56DerPapstthe kernel is the only think that is still worked on in cvs
12:59:01amiconnAh, so no need for svn then
12:59:43*linuxstb does a "cvs update" in his IPL kernel tree for the first time in months and nothing changed...
12:59:47antgellinuxstb: having said that, some of my oggs work fine, and they were encoded with id3v2. is it pot luck?
12:59:58DerPapstif you are interested in lcd stuff hotdog might be interesting too. it accesses the lcd hardware directly for some targets
13:00:24DerPapstbut i think only the greyscale ones. not 100% sure though
13:00:32*tucoz reminds linuxstb of
13:01:14 Quit thomasg_ (Read error: 110 (Connection timed out))
13:01:37 Join thomasg_ [0] (
13:01:51linuxstbDerPapst: I seem to recall reading somewhere that free() doesn't do anything in IPL - do you know if that's right?
13:02:03*scorche should compare those dates with magazine articles for fun
13:02:13*amiconn doesn't even know what hotdog is
13:02:23linuxstbLCD drawing lib
13:02:30XavierGrok USB charging mod is behaving very weird
13:03:34XavierGrI managed to make it work, but only when powered off
13:04:16DerPapstlinuxstb: unfortunately that is correct :(
13:04:21XavierGrwhen powered on the green light comes up only when the USB jack is half-inserted
13:04:30linuxstbantgel: Are you sure that some Oggs with id3v2 tags are working in Rockbox? Reading the code, they shouldn't...
13:04:39tucozscorche, if you use trends for something like vista, you get a selection of news items aligned with the curve.
13:04:46linuxstbDerPapst: That seems a pretty major issue...
13:05:20antgellinuxstb: pretty much. i've just stripped the tags, so let me rsync everything and i'll report back
13:05:41DerPapstlinuxstb: it is indeed. one all memory is used the called app just crashed because it runs out of memory and then you have to reset the ipod.
13:06:05amiconnNothing new there... just the kernel code replicated
13:06:10*DerPapst joins a typo class
13:06:22pixelmaa typo class?
13:06:34linuxstbDerPapst: You don't need a class - you're doing very well with typos..
13:06:39*DerPapst gives up
13:06:39JdGordondoes (DRAMORIG + DRAMSIZE - sizeof(struct bootloaderversion)) look correct?
13:06:43JdGordon... for all targets
13:07:09linuxstbI would put it somewhere that didn't depend on sizeof(struct bootloaderversion)
13:07:29JdGordonwell, that ideally will never change
13:07:54linuxstbAnd give it a header with some magic (e.g. "BootInfo") and a struct version number and size.
13:07:54*amiconn thinks the ipl lcd code is a good example how the one-binary-runs-everywhere makes the code severely bloated and inefficient
13:07:57 Join amar [0] (
13:08:22amiconnInsert 'concept' where it fits
13:08:29scorchelinuxstb: added to the wiki page...quite fast
13:08:38linuxstbJdGordon: If you're going to add a transfer mechanism from bootloader to Rockbox, IMO it should be future-proof.
13:09:12JdGordonok, ill see if this works first and then bloat it out :)
13:09:31amarcould someone have a look at fs#7478
13:09:39***Saving seen data "./dancer.seen"
13:09:45 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
13:09:50linuxstbJdGordon: But feel free to tell me I'm over-complicating it if you think I am... ;)
13:09:56amarsimple bug, simple fix i think thanks
13:10:31JdGordonlinuxstb: we'll see... :) DRAMORIG and DRAMSIZE arnt defined for apps :(
13:10:34scorcheif it is simple, then why dont you make a patch?
13:10:40DerPapstonce the iPL kernel somewhat works on a new ipod develoment suddenly stops. and there are a lot of other issues with the politics about who is allowed to add code to the holy kernel
13:10:43pondlifeHe has
13:10:59*scorche got a php error and couldnt see it
13:11:15*scorche is still getting that error
13:11:19 Quit miepchen^schlaf ("Verlassend")
13:11:40amarscorche: I have
13:11:46scorcheamar: sorry for the canned response...i see that now =)
13:12:08*scorche places foot in mouth
13:12:25*DerPapst wants a picture
13:12:41*scorche is also sans-clothes, and so refuses
13:14:35pondlifeJdGordon: ?
13:15:01JdGordonDRAMORIG and DRAMSIZE arnt defined for apps :( and i dont really want to add more hacks than necassery
13:17:21JdGordonamiconn: how would I go about getting a value into a register and then out again?
13:17:32*JdGordon was hopiong to not have to touch asm for this
13:18:07amiconnWell, you need to
13:19:28JdGordon... how?
13:21:00amiconnlinuxstb: Seems that the lcd type in the color lcd driver is both an interface and a type indicator
13:21:20amiconntype 0 has different commands; would be helpful to find out the controller type
13:22:24linuxstbMy Color has a type 1 LCD. I remember when I first bought it, the LCD wasn't work, but someone in IPL figured it out. Before it was fixed, I seem to recall the LCD driver could write to a small rectangle in the LCD...
13:22:54amiconntype 1 is HD66789R - the same controller as in the H300
13:23:54amiconnThe lcd in the H10 is controlled like type 1, but is a HD66773R (like X5) in the fullsize, and a TL1771 in the small H10
13:24:15amiconnSo the commands are different, but the interface is the same
13:24:24 Quit iamben (Read error: 104 (Connection reset by peer))
13:24:59*amiconn is quite sure the clock speedup will work on the fullsize H10, but wants to see it tested before commit
13:25:11 Join iamben [0] (
13:26:03 Quit Siltaar ("Aurevoir -")
13:26:24 Join Entasis [0] (
13:28:16 Join Thundercloud [0] (
13:28:24amiconnlinuxstb: Do you know who has an ipod color with type 0 lcd?
13:29:22*amiconn thinks the lcd speedup will improve battery runtime as it significantly decreases boost ratio during playback
13:30:06*amiconn also thinks it shouldn't be hard to implement lcd sleep for ipod color (type 1 lcd only, unless someone finds out what controller type 0 is) and H10 small
13:30:16amiconn(H10 large already has that)
13:30:59JdGordon says r0 is the first function argument. does that mean I can change main(void) to main(void*) and so only have to worry about putting the write value there before its called?
13:31:26amiconnNo, since main isn't called directly by the bootloader
13:31:52amiconnYOu need to hack into crt0
13:32:15linuxstbamiconn: The only other person I know with a Color is (I think) hcs. I don't know which LCD he has.
13:32:47amiconnHmm. This should be displayed under 'View HW Info' imo
13:33:33*amiconn wonders whether he'll manage to rewrite 3 lcd drivers in parallel, with only one of 4 targets available locally for testing
13:34:12linuxstbamiconn: OK, I'll add it. I'm also cleaning up a very old hack of mine which passes the ipod_hw_rev (which is read from flash) from the bootloader to main Rockbox via a memory location in SDRAM. Now Rockbox remaps flash, it can just be read in Rockbox.
13:34:27amiconnnice :)
13:34:54*amiconn still thinks the ipod OF doesn't remap ram as a whole
13:35:06amiconnOnly a small part of ram seems to be mapped to 0x0
13:35:23linuxstbAny suggestions for where a #define for that memory location (IPOD_HW_REVISION) could live?
13:36:19linuxstbsystem-target.h ?
13:36:44amiconnFor the archoses we still have hwcompat.h; i wonder how this could be target treeified
13:37:12amiconnIn fact it's not different from the HW_MASK in archos, which is also read from rom
13:39:37amiconnDisplaying the hw revision should be #ifdef IPOD_ARCH
13:39:50*amiconn gets an odd hardware revision on H10
13:43:43 Join moos [0] (
13:45:51amiconnlinuxstb: As we know that type 1 is HD66789R, we could reprogram the directions in order to match type 0
13:46:01amiconn(no longer x and y juggling)
13:46:35amiconnHowever, I don't know what the OF will do when booted from a bootloader that does this
13:46:48*amiconn remembers the mirrored OF on ipod mini
13:47:10amiconnWe could undo this change when booting the OF
13:48:40 Quit tucoz (Read error: 104 (Connection reset by peer))
13:49:24linuxstbamiconn: Yes, I'll add some ipod #ifdefs in the debug screen as well.
13:49:52*amiconn is puzzled about line 224 in lcd-color_nano.c
13:50:01 Quit voltagex (Remote closed the connection)
13:50:30amiconnThe directions are already matched, just the nano is different from the color
13:51:18amiconnIt's still quite messy though
13:53:16linuxstbBTW, I remember davidc (I think) coming here to mention a bug in the IPL version of the type 0 lcd_update_rect. But he didn't give any details, and I wasn't around at the time to ask him for more info... So I don't know if our driver has that bug, but afaik no-one has reported lcd issues.
13:55:06antgellinuxstb: yep, all working great, and thanks for the unicode (hebrew names read LTR in iriver firmware)
13:55:55linuxstbamiconn: It was courtc - see around 17.55.26 in
13:56:17DerPapstno one has reported any issues on the iPL forums as well.
13:56:20linuxstbamiconn: But there was no subsequent commits from him fixing it in the kernel...
13:56:33DerPapstmaybe is is sored out already?
13:57:53linuxstbantgel: Glad it's fixed.
14:00:37DerPapstseems not :)
14:04:48 Join Siltaar [0] (i=c134d0e9@gateway/web/cgi-irc/
14:06:40 Quit Nick_Brackley ("CGI:IRC (EOF)")
14:11:03JdGordonnew plan.... would writing to some semi-random position in RAM work reasonably well? i.e about 0x100000 which is hopefull 1mb in?
14:12:34amiconnMost probably not
14:12:49amiconnCould be overwritten by nearly everything
14:13:06amiconnAlso keep in mind that you might need to pass on the bootloader version on rolo
14:13:09JdGordonit would be grabbed in app_main() so nothing else would have the chance
14:13:31*amiconn still isn't convinced that the bootloader version needs to be passed around at all
14:13:35JdGordonforgot about that one
14:13:45JdGordonits not essential, but it would be nice
14:13:53linuxstbIt _would_ have been nice...
14:14:16DerPapstespecially for support questions
14:14:30XavierGryay success, it seems that my hub usb port couldn't deliver enough amps to H115 for both HD and green light charging
14:15:37JdGordon... unless i target tree it and dont try to do it in a generic fasion?
14:21:53 Join PaulJam [0] (
14:22:54 Join lxx [0] (n=lexx@
14:23:02lxxhi all
14:23:52PaulJamJdGordon: i was just wondering, are you going to fix the behaviour when holding left in filetree/database or should i open a bugreport for that?
14:24:08lxxwhy if i change sennings in rockbox after i reboot ipod Apple OS starts not Rockbox what i need to do for Rockbox starts every time after rebooting ipod?
14:24:48JdGordonPaulJam: ok, i'll have a look now
14:25:02JdGordonI dont know if that feature will actually be able to stay though...
14:25:19linuxstblxx: The Apple firmware will start if you either have the hold switch on or the MENU key held as your ipod boots.
14:26:02linuxstblxx: Also, when you turn off your ipod in the Apple firmware, it doesn't actually turn off, it just sleeps. So turning it back on just wakes up the Apple firmware - the bootloader is never run.
14:26:07PaulJamI would miss it, but i guess i can live without it. it's propably just a matter of getting used to it.
14:26:26JdGordonno, ill see what i can do, it really is a nice feature to have
14:26:31JdGordonspecially for the blind
14:26:57lxxbut in Rockbox manual wrote i need turn on my ipod with select and menu button ?
14:28:35*JdGordon wonders how the heck that actually worked
14:31:51pondlifeIs a long left meant to return to the browser root? Or the main menu?
14:32:02JdGordonbrowser root
14:32:30lxxwhat instal to see this option ?
14:33:01bluebrotherlxx: the rockbox bootloader doesn't have a boot menu
14:33:08DerPapstthis is ipodlinux boorloader called Loader 2
14:33:10bluebrotherand IPL's Loader2 isn't supported by Rockbox
14:33:21bluebrotheryou can use that to load Rockbox, but it's not supported here.
14:33:25DerPapstbut you can use it to launch rockbox
14:33:52bluebrotherbut unless you want to use IPL there is no need to use Loader2
14:34:04bluebrotherin fact you can also boot IPL with Rockbox' bootloader ;-)
14:34:52JdGordonPaulJam: fixed, cleaning up then commit
14:35:09PaulJamcool, thank you :)
14:35:10*JdGordon cant belive that actually worked, this fix is how it should have been from the start
14:37:15JdGordonfix in svn
14:37:35lxx <bluebrother> and how i can boot IPL with rockbox bootloader?
14:38:25bluebrotherput the kernel to the FAT32 partition root with the filename "linux.bin"
14:39:11lxxand the how i can choose betwean OS`s?
14:39:16*DerPapst would never use the rb bootloader to boot ipl...
14:39:26bluebrotherthen hold Play while booting
14:40:08*bluebrother only used Rockbox bootloader for booting IPL
14:41:06lxxand what i see next the same menu like on picture or what?
14:41:16bluebrotheryou won't see any menu
14:41:32bluebrotheras I already said, the Rockbox bootloader doesn't have a boot menu
14:43:22DerPapstif you want a menu you have to use Loader 2
14:43:32JdGordonshould the default number of sides for dice be 6?
14:43:45*JdGordon thinks 20 is too high
14:44:52amari thought 6 was the number on most dice
14:45:11amar10 would also be a good choice
14:48:02linuxstbDerPapst: Do you know anything about the "sysinfo" struct used by loader2? IIUC, it's a struct stored at the end of IRAM and is used by Apple to pass info from the bootloader to the main firmware. It could well explain why we've never managed to get rolo to load the Apple firmware in Rockbox...
14:48:30linuxstbI've just dumped IRAM from the Rockbox bootloader, and it appears to be populated with that struct at the end.
14:49:40DerPapstno sorry. i don't know much about the internals of loader 2.
14:49:50JdGordonamar: I'm just adding config saving to dice, then ill commit your patch
14:51:14lxxRockbox working with kensington Digital FM Transmitter ?
14:51:15*amar wonders if 2 lines is worth calling a patch
14:51:40DerPapstthere have been smaller patches ;)
14:52:05DerPapstchanging only one symbol :P
14:52:43*Nico_P just spotted FS #7476
14:53:18Nico_PI like the idea
14:54:20JdGordonmore bloat :p
14:56:46Nico_PI don't think it would take up much code
14:57:47lxxRockbox and ipodlinux its only two alternative OSs for ipod or exist any other alternative OSs?
14:58:11amarout of curiosity if in future I find a bug then also write a fix should I regester a bug or a patch?
14:58:42JdGordonas a patch
14:59:18amarthanks - will remember that for future
15:00:56DerPapstlxx: there is another one but it isn't publich and i doubt it will be ever made public
15:02:46lxxwhat the name of that os? and what that os can ? what functions?
15:03:52DerPapstit's called ccOS and it is similar to the linux kernel. but i think the author just made it for the challenge.
15:05:21lxxOS suport Flac or APE ?
15:05:37XavierGrsome GPIOs are listed as unasigned on the wiki page for the H100, does that mean that someone could utilize them for signals writing the proper code?
15:07:40DerPapstOS? if you mean ccOS then no. Its just a kernel. iirc it doesn't even have any audio drivers nor button drivers. only an lcd driver for nano and photo/color and a not working one for 5G/5.5G
15:09:42***Saving seen data "./dancer.seen"
15:11:54lxxrockbox support uploading files and listen at the same time without usb unplug?
15:12:11 Join Febs [0] (n=chatzill@
15:13:11linuxstblxx: No - Rockbox isn't running when the ipod is connected to your computer via USB.
15:13:53lxxipodlinux can do that?
15:14:10linuxstblxx: No - ipodlinux isn't running when the ipod is connected to your computer via USB.
15:14:41JdGordonNico_P: ok, just reread that task, and yeah that will lead to some pretty crazy wps'
15:14:46markunlxx: but austriancoder is working on making it possible with rockbox
15:15:23linuxstbmarkun: But that's unlikely to allow simultaneous access to the disk by both the computer and Rockbox...
15:15:25Nico_PJdGordon: could be quite nice but I don't think the cost is too high
15:15:38JdGordonthat but shouldnt be there then
15:15:38Nico_PI'll try writing the patch
15:16:11lxxmarkun and when it can be possible?
15:16:21Nico_PJdGordon: sorry, what do you mean ?
15:16:32PaulJamlxx: even when rockbox has its own USB implementation i could imagine that it would be problematic to support that. because what would happen if for exanple rockbox and the computer are accessing/modifying the same file on the harddrive?
15:16:53JdGordonNico_P: you said "but i dont think the cost is too high", that but shouldnt be there...
15:17:16DerPapstlxx: infact i haven't seen any dap that supports that in its OF
15:17:45Nico_PJdGordon: ah yes I wasn't very clear... I meant that it's a good idea that probably won't be too costly
15:17:46PaulJamoh,i didn't read linuxstbs answer.
15:18:58 Join kaaloo [0] (
15:19:48 Join Jessica [0] (i=46ef580c@gateway/web/cgi-irc/
15:19:58markunDerPapst: but that alone wouldn't be a reason for rockbox not to support it of course
15:20:07JdGordonNico_P: its not just a matter of adding a new tag for it and basically copying the WPS_TOKEN_VOLUME case inget_token_value is it?
15:20:11 Part kaaloo
15:20:15markunlinuxstb: yes, could be problematic
15:20:15JessicaHi All. Just started playing with RB. You all did a nice job!
15:20:24markunthanks Jessica!
15:20:34Nico_PJdGordon: it's not much more complicated than that
15:20:39markunJessica: which player do you have?
15:20:47Nico_PJdGordon: although I'd say it's closer to the battery percentage tag
15:21:05Jessicaquick question. I've built a huge playlist. (1600+ songs) (markun: old archos recorder)
15:21:10Nico_PI'm thinking of adding a "progress percentage" tag that can be used in a conditional
15:21:32JessicaI get a bit o lag between songs. is that just cuz of the size of the playlist on this old player?
15:22:08DerPapstmarkun: it would be a neat feature indeed. but i already hear some devs complaining that it is either a PITA to implement or bloat ;)
15:23:20linuxstbNico_P: Have you looked at the patch that supports partial bitmaps in the WPS?
15:23:39Nico_Plinuxstb: no :( I like the idea but never got round to looking at the patch
15:24:02linuxstbDo I remember correctly that it doesn't solve the 52 images problem?
15:24:13linuxstbOr does it...
15:24:38Nico_Pdepends what you consider as a proper solution to it :)
15:24:51Jessicaactually, for a bit of clarity on my prev question, the lag happens when I skip to the next song, not simply when a song ends and goes to the next song.
15:24:53linuxstbSome of the SVN WPSs are now missing codec icons, and there aren't enough free image slots to add them.
15:25:16Nico_Pwell then it solves it
15:25:24Nico_Pbut for that issue I even prefer the other patch
15:25:35Nico_P(which I haven't looked at either)
15:25:41linuxstbWhat did that do?
15:26:18linuxstbJessica: Are you saying that when you manually skip to the next song, there's a delay, but when the song ends naturally there is no delay?
15:26:34Nico_Plinuxstb: FS #7345: Enumerated image sets for WPS
15:26:34Jessicalinuxstb: yes
15:27:17linuxstbJessica: That just sounds like the fact that Rockbox needs to spin up the hard disk to load the next song. When songs end naturally, the next track should have been pre-buffered.
15:27:26*Nico_P can't find a good tag for the track elapsed percentage
15:27:31 Quit mnakiomji (Read error: 104 (Connection reset by peer))
15:27:35Jessicaahhh! ok!
15:27:39linuxstbNico_P: Switch to XML ;)
15:27:44 Join onakiifga [0] (i=0@
15:27:45Jessicatyvm, linuxstb
15:27:58Nico_Pnow that's bloat :)
15:29:10DerPapstdoes this tags always have to be 2 digits?
15:29:26DerPapst*do these
15:30:13DerPapsti shoudl have paid more attention in my english classes ~5 years ago :-/
15:31:24Nico_PDerPapst: you mean two chars ?
15:31:40DerPapstor could there be 3 also?
15:31:42JessicaCan I delete all the langs out of the \.rockbox\langs folder except english to save space? Also, can I delete unwanted plugins?
15:31:49Nico_PDerPapst: it could be anything really
15:32:11linuxstbJessica: Yes, but it's a pain as you'll need to do that each time you update Rockbox. Is your hard disk that full?
15:32:15DerPapsthow about pcp then (last p for percent) :P
15:32:21Nico_PDerPapst: the old WPS code made it hard to have more than two, but now it could be anything
15:32:35Jessicalinuxstb... no, I'm just cheap!
15:33:19Nico_PDerPapst: I like it :)
15:33:33Jessicabtw, the dictionary plugin choked when I tried to run it. Shame, cuz I'm a big scrabble geek
15:33:45DerPapstheh :-P
15:37:32lxxwhat font in rockbox support russian?
15:37:57 Join bicurious [0] (
15:38:16DerPapstlxx: probably the unicode ones
15:38:31amiconnpcp sounds like a chemical substnce...
15:38:39bicuriousHi, I'm trying to install rockbox and the documentation says to use the specific manual but there's no link for 5.5g ipods. Is there a manual I just don't see?
15:39:16linuxstbbicurious: Use the ipod video manual - Rockbox doesn't differentiate between the 5g and 5.5g
15:39:17bicuriousor alternately will the 4g instructions work on 5.5g
15:39:22bicuriousok thanks
15:39:32DerPapstthere is a nice info page on the wiki which lists all the fonts that support different char sets. dunno the link
15:39:36PaulJamlxx: you can find detailed informations about the fonts here:">
15:40:00markunlxx: and
15:40:02markundamn :)
15:40:05*linuxstb wins
15:40:27*markun shouldn't drink beer while at work
15:41:00bicuriousOk thanks. Does anyone know if any stability issues with 5.5g have been adressed since it came out? I was encountering a lot of freezes of the navigation requiring hard resets.
15:41:35linuxstbWhen did you experience those freezes?
15:42:30bicuriousIt was specifically when I'd play a song, set the player on 'hold' and then listen to a few songs
15:42:49bicuriousI believe this was a release within the first week of the 5.5g release
15:43:50DerPapstthere have been made some stability changes to rockbox lately.
15:44:06bicuriousOk cool
15:44:12DerPapstbest would be to try it out again and see if it still happens
15:44:21bicuriousPutting some mp3s on it now
15:44:28bicurious(freshly rma'd 80gb)
15:47:25JdGordonlinuxstb: the hack you just removed, could that be used for what I want?
15:53:49linuxstbJdGordon: I guess so - as long as it doesn't look like a hack...
15:53:51lxxi have flac file with cue file in cue file track names wrote on Russian what i need to do for Rockbox start understand Russian symbols?
15:55:48PaulJamlxx: have you set the proper codepage in the settings?
15:56:07lxxwhere is it?
15:57:07PaulJamsettings => general settings => display => codepage
15:57:49lxxi need set unicode or Russian?
15:58:15PaulJamtry and see what works
15:59:38PaulJamit propably depends on the encoding used in the .cue file.
16:00:11 Quit TiMiD[FD] (Read error: 110 (Connection timed out))
16:00:20markunlxx: it depends on the encoding of your tags
16:02:38 Quit Entasis ("Leaving")
16:04:50Nico_PDerPapst: actually what I said isn't quite true... three letter tags require increasing the "name" field for the token struct, which results in a non-negligible binsize increase
16:05:02lxxi dont but nothing help
16:05:14Nico_Pa two-letter tag would be better
16:05:36lxxand mp3`s files with russian names symbols dont recognized
16:08:05PaulJamNico_P: is it really important that the two letters make sense? one propably has to look up the tags in the wiki anyway.
16:08:24Nico_PPaulJam: not really... maybe a new tag isn't necessary anyway
16:09:12*GodEater_ posts a photo from DevConPub 2 - The Revenge
16:09:39Nico_Pok i'll be %px
16:11:43JdGordonGodEater_: hahahaha
16:12:04GodEater_JdGordon: I let him win. Honest.
16:12:49markunlxx: which font did you select?
16:13:12markunThe file names should always show up correctly with the right font
16:13:48 Quit Soap (Read error: 110 (Connection timed out))
16:14:47 Quit Jessica ("CGI:IRC (EOF)")
16:17:04Nico_Poh crap I forgot to trim down the lines before I committed
16:17:39linuxstbNico_P: "Ooops" is the word to use...
16:18:00JdGordon:'( why is this so simple and wont bloody work for me
16:18:02lxxi try all fonts
16:18:12lxxand no one work correctly
16:19:06markunlxx: which OS do you use on you computer?
16:19:43lxxwindows xp
16:19:52Nico_Plol @ GodEater_'s photo
16:19:56Nico_Pwho won ? :)
16:20:13GodEater_Nico_P: see above
16:20:35markunlxx: if explorer shows the files in russian on your player they should also show up fine with for example unifont
16:20:40 Join roolku [0] (
16:20:46Nico_PGodEater_: isn't that linuxstb on the photo ?
16:21:15markunlxx: what do you see? little squares or something very different?
16:21:22GodEater_no that's definiltely pondlife
16:21:51*Nico_P thought it was pondlife vs. linuxstb
16:21:54DerPapstahahahaa that photo is just cool :D
16:21:57roolkuSlasheri: are you around?
16:22:05GodEater_no it's pondlife vs. me
16:22:18pondlifeI won!
16:22:21markunso nice to see faces :)
16:22:22Nico_PGodEater_: I must've been confused by the photo at
16:22:36Nico_PGodEater_: I confused you with linuxstb, sorry
16:22:41GodEater_hehe np
16:22:56GodEater_pondlife: you must have won - you have a t-shirt and mug right ? :)
16:23:22pondlifeThe mug is only slightly too small for digestive dunking.
16:23:26pondlifeThe T-shirt is fine
16:23:34GodEater_ginger nuts work ok ?
16:23:42pondlifeNot been to get them yet.
16:23:46pondlifeWill post an update
16:23:56GodEater_I shall await it with eafer anticipation
16:24:00GodEater_eager even
16:24:08pondlifeWell, you could test that function.
16:24:16GodEater_I also lack ginger nuts
16:25:10 Join TiMiD[FD] [0] (n=TiMiD[
16:25:18Nico_Pdid you guys discuss any rockbox-related topics ?
16:25:20GodEater_I'm keen to hear whether the issues with playback.c you highlighted last night have been fixed yet pondlife ? ;)
16:25:40GodEater_Nico_P: oh yes - pondlife waxed lyrical about how much he looooooves playback.c
16:25:49Nico_Phehe :)
16:25:55 Quit bicurious ()
16:26:03GodEater_and how "Metadata on Buffer" will never work
16:26:11pondlifeNico_P: Yes, we discussed how fundamentally important your rework of playback is to the future of Rockbox. No pressure, then.
16:26:42pondlifeHelp us Nico P Kenobi. you're our only hope...
16:26:47Nico_Pno pressure at all :)
16:26:51GodEater_Nico_P: pondlife also stated that apprently playback is full of bugs. linuxstb and I failed to believe him, and claimed it always worked for us.
16:27:04GodEater_which it does
16:27:06DerPapstouch... poor Nico_P :P
16:27:07amarcan someone have a look at
16:27:10pondlifeThen stripwax produced a reproducible Codec Failure"!
16:27:12amarit is the only patch that I opened that is still open
16:27:15lxxhow install new wps ?
16:27:29JdGordonthe "reset all devices" the sansa does in system_init(), does that reset the ram also? or just everything else?
16:27:34Slasheriroolku: hi
16:27:39linuxstbpondlife: Only by using obscure options...
16:27:42GodEater_pondlife: yes that was interesting. Who's wokring on fixing it ?
16:27:47roolkuHi Slasheri :)
16:28:06roolkuSlasheri: I am trying to understand your queueing commit
16:28:07Nico_Pbtw, are there any other photos so we can see all of you ?
16:28:16pondlifeHopefully not.
16:28:21GodEater_Nico_P: there are - but they're dreadful
16:28:29GodEater_I'm really not keen on posting them
16:28:37roolkuSlasheri: am I right you can now only change the rating once all tracks have been buffered?
16:28:51GodEater_besides, the only person you've not got a picture of is stripwax
16:28:59GodEater_and he has a pic of himself on his profile in the forums
16:29:01roolkuSlasheri: i.e. once id3->tagcache_idx has been set?
16:29:04GodEater_so you don't need one
16:29:26Slasheriroolku: yes, correct. that is to prevent disappearing of ratings or overwriting of them
16:29:30JdGordonamar: both looks fine, ill commit tomorow unless im beaten to it... too tired now
16:29:45Slasheriso system prevents changing the rating at all unless all preconditions are met
16:29:50pondlifeOK, we have 43 "Music playback" bugs logged at the moment... not all of which are strictly playback related.
16:29:53amarno problem thanks
16:30:10pondlifeOops, long URL - sorry.
16:30:15roolkuSlasheri: but that stops me from skipping through my songs and rating them :(
16:30:36Slasheriroolku: hmm.. then that must be a problem with the playback engine
16:30:41roolkuSlasheri: I never experienced any lost ratings - under what conditions does that happen?
16:31:13Slasheriroolku: in fact that is a bug in playback engine as well, but on end of playlist, it does not create buffer events for first track at least
16:31:28Slasheriand when unbuffer events is received, ratings of that track is lost
16:31:51roolkuSlasheri: hm
16:32:03Slasheriso i really think nothing should be updated in the db unless tagcache_idx is set (id3 entry is valid)
16:32:47*JdGordon slaps Nico_P with his 11bytes of utter bloat!
16:32:51Slasheriand we could set up that earlier indeed, it would just require modifying of the playback code
16:33:15roolkuSlasheri: well, I checked for rating!=0 (i.e. user has set the rating in the id3 tag) before overwriting it from the tagcache info
16:33:36GodEater_pondlife: 5797 looks like the one stripwax demonstrated to us
16:33:52*Nico_P is starting to work on the XML parser for the NextGen WPS system !
16:33:54roolkuSlasheri: this check is still there, but it is not qued to be updated anymore
16:34:04Slasheriroolku: yes, but if rating was never read in (it's 0 in id3 struct), then we will overwrite a valid rating with that 0
16:34:06Nico_Pscrew MoB, that's useless !
16:34:11Slasheriand that happened to me a lot
16:34:17Slasheriin fact i lost all my ratings :D
16:34:59roolkuSlasheri: huh? that sounds bad
16:35:08*roolku quickly backs up his ratings
16:35:16Slasheriit was.. now i can actually use that feature :)
16:35:39Slasheriand no need to skip to next track either if i want to a new rating to be saved
16:35:44 Quit wg_ ("leaving")
16:35:56Slasheriand writing back stats is now a lot faster
16:36:14Slasheribecause there is no need to do tagcache_searches for individual tracks
16:36:37roolkuSlasheri: I see the advantage, but was wondering if set ratings (!=0) could still be saved in the unbuffer event?
16:36:44pondlifeHmm, how did increase the binary size so much??
16:37:00*pondlife thinks that table may be out of sync...
16:37:38pixelmawell TiMiD's commit before decreased it by the same amount before...
16:37:43amiconnpondlife: TiMiD[FD]'s commit (with the reversed versions) decreased it by the same amount
16:37:56roolkuSlasheri: the ones that have been set before the buffer event completed
16:38:00*DerPapst just wanted to mentioni that too
16:38:02pondlifeBut it also shouldn't have made that much difference.
16:38:07DerPapsttoo slow as always :P
16:38:12Slasherineed to go now but back soon ->
16:38:55pixelmaand I even had time to add a second useless "before" to my sentence ;)
16:39:17roolkuSlasheri: no prob - nice talking to you again.
16:40:02bluebrothersong progress percentage? Do we really need such a thing?
16:40:32JdGordonof course!
16:40:50*bluebrother considers this a useless feature
16:40:58DerPapstno way
16:41:02*amiconn would rather like to see a MoB progress percentage
16:41:03JdGordonshock horrror!
16:43:00*amar wonders what MoB progress percentage does
16:44:10JdGordonit raises everyones expectations...
16:45:19 Quit roolku ()
16:45:39Nico_Pbluebrother: I think it can allow a bit more flexibility to WPS creators for a cheap cost
16:47:13JdGordonones the ppl on the forum figure out how to use it, expect some really funky progress bars
16:47:36JdGordonexcept there arnt enough bmps alpwd in the wps to make it smooth enough
16:47:53 Quit midkay (Read error: 110 (Connection timed out))
16:48:49Nico_PJdGordon: there are two patches out there that should help with that
16:48:53GodEater_unless that "cutting bmps up into smaller ones in the wps" patch makes it in there
16:49:03 Join iwantanimac [0] (i=cb3b15aa@gateway/web/cgi-irc/
16:49:09 Join nls [0] (
16:49:10JdGordonNico_P: yeah, but will either of them be commited?
16:49:23 Join pondlife1 [0] (
16:49:37Nico_PI plan on looking at them closely one of these days
16:50:16JdGordonbed time for me
16:50:24 Quit JdGordon ("Konversation terminated!")
16:50:27bluebrotherin that case, a "show partial bmp" would make more sense. Something the progress bar already can do ;-)
16:50:39iwantanimacHi all. Haven't been on in a while. Is there any reason the current SVN build would be stopping playing a video halfway through?
16:53:20 Join toffe82 [0] (
16:54:14 Join SunPower [0] (
16:54:55 Part SunPower
16:56:02 Quit morrijr (Remote closed the connection)
16:56:23 Join Jens [0] (i=Jens@pdpc/supporter/active/Jens)
16:57:07 Quit pondlife (Read error: 110 (Connection timed out))
16:57:25 Join morrijr [0] (
17:02:09 Quit iwantanimac ("CGI:IRC (EOF)")
17:09:46***Saving seen data "./dancer.seen"
17:13:11Nico_Pamiconn: what ?
17:13:28amiconnThe midi player plugin struggles on H10, as the sample rate is set to 44.1kHz now
17:13:53amiconnAnd outputting 'buffer miss' to the screen if it can't keep up makes the effect worse
17:14:24amiconnSome simple midi files do work
17:14:57amiconns/simple/very simple/
17:15:59 Quit lxx ()
17:16:33 Quit pondlife1 ("disconnected has pondlife")
17:17:35Nico_Pamiconn: how come it struggles ?
17:17:58amiconnWell, it can't keep up synthesizing
17:18:16Nico_Pso it's the CPU that's too slow ?
17:18:21amiconnlinuxstb: Does mpegplayer boost all the time, or does it unboost if limit fps kicks in
17:18:23DerPapstsame on the ipod video. esp. the light show you get when backlight fades of :D
17:18:56amiconnVideo is completely smooth now on H10 small (I get >36fps without limit & skip)
17:19:09amiconn(not yet committed the optimisation)
17:19:23*amiconn waits for arathis, or another H10 large user
17:22:07 Join midkay [0] (n=midkay@rockbox/developer/midkay)
17:22:12Nico_Pamiconn: did you find the fix for the crashes ?
17:22:56amiconnlooks like I did
17:23:07Nico_Pcool :)
17:23:20amiconnAlso not committed yet; Arathis wanted to test during the day and report back in the evening
17:24:05Nico_Ptomorrow I'll have an H10 5GB at hand if that can help with the testing
17:27:32 Join webguest67 [0] (i=8e491086@gateway/web/cgi-irc/
17:28:59amiconnlinuxstb: Is it normal that I still get HW version == 0xffffffff on G5.5?
17:29:13 Join saratoga [0] (i=980398fe@gateway/web/cgi-irc/
17:31:50 Quit Siltaar ("CGI:IRC")
17:31:57Nico_Pbluebrother: have you made progress on rbutilQt ?
17:32:03 Join pondlife [0] (
17:37:45 Quit webguest67 ("CGI:IRC")
17:44:27 Quit TTSbox_ ("Leaving")
18:05:46 Quit wrobbie (Read error: 110 (Connection timed out))
18:12:43 Join webguest30 [0] (i=ac8e7de4@gateway/web/cgi-irc/
18:12:47 Join Domonoky [0] (
18:13:40pixelmapatience - and if you have a specific question, just ask
18:13:55webguest30I just came here to ask a quick question.
18:14:40webguest30Sorry about the first bit. My client was lagging and what I typed didn't show up.
18:14:56pixelmait'd be even quicker if you'd just ask away ;)
18:15:56webguest30Ok ^^''
18:16:12webguest30You have probably heard this a million times.
18:16:22webguest30What are the advantages of rockbox?
18:16:48dionoeathe name sounds cool
18:18:36pixelmawebguest30: that's a bit subjective, depends a bit on your player and your expectations... maybe this is a good read for you
18:19:11 Join GodEater [0] (
18:25:15 Quit GodEater ("User pushed the X - because it's Xtra, baby")
18:25:34alienbiker99pixelma, yes i have an h10
18:26:14pixelmaalienbiker99: amiconn was looking around for a tester with a 20GB H10
18:26:33alienbiker99id have to find it, does he still need a tesrer?
18:26:39 Join GodEater [0] (
18:27:05pixelmaI think so, but hopefully mentioning him will get him here
18:27:19pixelmaso he can tell you more
18:28:47 Quit Nibbier (Read error: 113 (No route to host))
18:28:47 Join hannesd_ [0] (
18:29:54 Quit obo ("bye")
18:30:29 Join Nibbier [0] (
18:31:49 Join lxx [0] (n=lexx@
18:33:50webguest30Can somone give me a link on how to install please.
18:34:27amiconnalienbiker99: It would be nice if you could test something for me. It's not yet committed though
18:34:40lxxwhat video formats support rockbox?
18:34:48DerPapstmpeg 1 and 2
18:34:48alienbiker99ok sure, let me find my h10 and cable
18:35:13 Quit amar ("Konversation terminated!")
18:35:14amiconnDo you want a patch so you can build yourself, or should I prepare the build(s)?
18:35:50alienbiker99it would be great if you you could prepare the build
18:36:58amiconnOkay. The first test (only including the crash fix) can be done with the same build I made for Arathis
18:37:17alienbiker99ok, itll be about 25-30 minutes before i can get the h10 though
18:37:21amiconnIt's still up,
18:37:34alienbiker99ill download what you just sent me
18:38:09amiconnBasically I want you to run the test_fps plugin (not included in normal builds) both with the cpu unboosted and boosted, and tell me the results
18:38:40amiconnAfter that I'll prepare another build, where you should do the same.
18:38:57alienbiker99ok, how do i change between unboosted and boosted
18:39:17amiconnJust start test_fps after boot, then the cpu should be unboosted
18:39:35amiconntest_fps will display the cpu clock as well, should be 30MHz
18:40:38amiconnThen you leave the plugin, go to Debug->CPU frequency, increase the boost counter to 1, leave the debug menu, and start the plugin again
18:40:57amiconnIt should now measure a higher speed, and show the CPU clock being 78MHz
18:41:11 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:41:23amiconnEither reboot afterwards, or go to Debug->CPU frequency again and reset the boost counter to 0
18:42:19alienbiker99ok, do i need to send you a log or anything?
18:43:08amiconnNo, just tell me the (few) values the plugin displays for each run
18:44:57 Quit hannesd (Connection timed out)
18:48:40 Join Xerion [0] (
18:53:03 Quit lxx ()
18:54:24bluebrotherNico_P: sure :) Got a new tarball uploaded today
18:54:32 Join Arathis [0] (
18:54:58bluebrotherDomonoky: want to check my latest changes? I think I'm reaching a state where I can put it to svn ;-)
18:55:07Nico_Pbluebrother: I went back in the logs and found your websapce :) you seem to be making nice progress indeed
18:55:18 Join Rondom [0] (
18:55:19Nico_PSVN is good
18:56:25 Join AceNik [0] (n=AceNik@
18:56:54bluebrotherI'm just wondering where the best place to put it is. Maybe a subfolder in the existing rbutil module?
18:57:25bluebrotherthat way we could share common code
18:58:19AceNikhey guys 2 problems with rockbox on the H10: 1.mpegplayer: after quiting from one video & playing the other causes a data abort at 0000IE 2. there are randm shutdowns in the player while playing music, not shutdowns but lockups, & the reset button comes to the rescue
18:58:39amiconnAceNik: Known, and a fix is in the works
18:59:11AceNikamiconn: thanks, is the first one known too ?
18:59:25Nico_Pbluebrother: eventually you want rbutilQt to replace the curent rbutil, don't you ?
18:59:33amiconnI think they might be related
18:59:44bluebrotherI think that was an idea behind it, yes.
18:59:54 Quit pondlife ("disconnected has pondlife")
19:00:18AceNikamiconn: btw there is one thing i wanted to enquire if rockbox has a feature already can anyone else apply for a patent for that feature for their company, cause this is what apple is doing
19:01:32alienbiker99patents are getting way out of hand these days
19:02:07AceNikamiconn: the sncviewer patch is currently doing very well, but apple has applied for a patent in the same for its players, will it still be possible for us to use this if apple succeeds?
19:02:35Nico_Pbluebrother: a subfolder sounds good, but will it be easy to move up one level once it's complete enough to replace the current rbutil ?
19:03:46 Quit midkay ("Leaving")
19:04:22Febsalienbiker99: in theory, one cannot obtain a patent for an invention that was “known or used by others in this country, or patented or described in a printed publication in this or a foreign country, before the invention thereof by the applicant for patent."
19:04:45Febs(the reference to "this country" means the United States)
19:05:23GodEaterFebs: the "in theory" bit is the key point though :)
19:05:25AceNikbut sncviewer is still a patch yet, rockbox hasnt included it yet , can there be a revolt to apple's patents from rockbox, & can they include this patch ?
19:05:39bluebrotherI don't think moving should be a big issue
19:06:10bluebrotherit could even stay in that subfolder (currently all files are in a subfolder too) and just add a project file to build from the top folder
19:06:25 Join midkay [0] (n=midkay@rockbox/developer/midkay)
19:08:17AceNikguys the sncviewer has made loads of progress & is working brilliantly well, can we include it & revolt against apple's patent application
19:09:03Nico_PAceNik: would it need to be included ? can't the person who wrote the patch oppose to apple's patent application ?
19:09:04FebsI can't really comment on this specific patch as I am not familiar with it, but just looking at the language I quoted above, the test appears to be "known OR used." If the sncviewer is working, then it would appear to satisfy both of those criteria, regardless of whether it is officially included in Rockbox.
19:09:47***Saving seen data "./dancer.seen"
19:09:53alienbiker99but how do you go about opposing the patent
19:09:56Nico_PFebs: is there a way of preventing apple from getting the patent ?
19:10:24Nico_Pbluebrother: I'd say go for it then :)
19:10:29AceNikall i mean to point is , rather than an individual making a scene, rockbox as a community should act on it, alienbiker99: if i knew i would have already done that
19:10:51alienbiker99ah ok
19:11:23AceNikhere is the patch :
19:11:48AceNik& here is the patent application apple's applied for:
19:13:06AceNiki think the main authorities at rockbox would have enough strings to pull to stop this patent from going through, but tell me one thing, will we still be able to use this feature officially in rockbox even if apple succeeds?
19:13:48 Part maffe
19:14:18 Join maffe [0] (
19:15:16*amiconn is puzzled by a detail of the ipod system_init() and neither dan_a nor linuxstb seem to be around :(
19:16:05Arathisamiconn: my player ist still running after more than eight hours. I'll just let it run out of battery to see how much the battery runtime improved
19:16:20amiconnNice to hear :)
19:16:51amiconnI have some further tests to do on H10 20GB, but alienbiker99 is also around now...
19:17:44ArathisI will be around for about 3h
19:18:21Nico_PAceNik: any idea why isn't this plugin in SVN ?
19:19:44AceNikamiconn: if u need to test something im there too , but guys tell me can the sncviewer be included, the plugin was recently reworked on, & just needs the right attention & nod from proper people then it can be included, thats what im requesting for, cause its way better than the earlier version
19:20:40AceNik Nico_P: if you look at the kind of patent apple has pplied for , they have gone overboard with providing the blinking text kind of karaoke effect , which can be an extension for this plugin
19:20:58alienbiker99ill test now
19:22:18AceNik Nico_P: i just feel there are so many plugins & patches waited to be included, but either the developers get lazy to talk on IRC for inclusion or , they have some more serious work left to be done on it , once noticed a plugin can be improved in a few days, like the " sokoban plugin patch was done "
19:25:29alienbiker99amiconn, what do you need me t otell you
19:25:55amiconnThe two fps values (for 1/4 and 1/4)
19:26:03amiconnErm, 1/1 and 1/4
19:26:33alienbiker99unboosted: 1/1: 27.5 ; 1/4: 109.0
19:27:35alienbiker99boosted: 1/1: 42.0 ; 1/4: 166.0
19:28:19alienbiker99do you have another build?
19:30:23 Quit Jens ()
19:34:36 Quit AceNik ("bye guys have fun , enjoy !!!!!")
19:34:53 Join AceNik [0] (n=AceNik@
19:35:11FebsDo you have a moment for private chat?
19:35:20AceNikya tell me ?
19:37:52AceNikcant you get my messages?
19:39:05FebsNo. Are you identified to NickServ?
19:39:08Nico_PAceNik: you need to be registred to be able to PM... are you ?
19:39:19AceNikno im not, how do you do that ?
19:39:44amiconnOnly one of you needs to be registered, who could the allow messages from unregistered users
19:40:18AceNikfebs: allow me then , im unregistered, but how do i register then ?
19:40:50 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
19:40:56 Part toffe82
19:41:29Arathisamiconn: what's that buidl for anyway?
19:41:40amiconnLCD speeeeed :)
19:42:08LloreanWith that many e's it's got to be good.
19:42:43 Join lxx [0] (n=lexx@
19:42:59lxxhow enable icon sets in rockbox?
19:43:03 Join hannesd [0] (
19:43:14Arathishow will an LCD speedup be effectifly visible?
19:44:23amiconnArathis: On the H10/6GB it's very obvious if you e.g. try to play jewels while playing music
19:44:48 Join miepchen^schlaf [0] (
19:44:49amiconnWithout the patch, jewels are falling so slow it's no fun to play it
19:44:59Febslxx: (basically, you just need to download an iconset, extract the .zip file to your player, and then browse to and select the .cfg file relating to the iconset that you want.)
19:45:15Arathisamiconn: but how does the LCD effect that? thought this would be something about the cpu
19:45:18amiconnThere will be further speedups (porting from the ipod color lcd driver), this is only a first step and proof of concept
19:45:22alienbiker99unboosted: 1/: 28.5 ; 1/4: 114.0
19:45:44alienbiker99boosted; 1/1; 74.0 ; 1/4: 298.0
19:46:09Arathisbtw: still 16% battery-power :)
19:46:11lxx <Febs> where i need to select cfg file?
19:46:35amiconnalienbiker99: No graphical glitches on the display?
19:46:47amiconnAbout the same speedup as on the H10/6 btw
19:46:55Febslxx, after you have extracted the iconset, you should be able to find the .cfg file by going to Settings −−> Browse themes.
19:47:03amiconnA few percent at 30MHz, and nearly 80% at 78MHz
19:47:03 Join hannesd__ [0] (
19:47:21alienbiker99dont think so
19:47:27amiconnArathis: This patch changes the LCD controller clock (the same way as it is set on ipod color)
19:47:50 Quit hannesd_ (Read error: 110 (Connection timed out))
19:47:58amiconnalienbiker99: This should also improve video playback in case you're interested in that
19:48:09Arathisah, I see
19:48:12lxx <Febs> and what i need to do next?
19:48:25alienbiker99i could test teh elephants dream
19:48:28FebsAceNik: try identifying yourself to NickServ by typing /msg NickServ IDENTIFY [yourpassword]
19:49:06Febslxx: select the iconset you want to use by navigating to its .cfg file and pressing Select.
19:49:30 Join obo [0] (n=obo@rockbox/developer/obo)
19:50:44AceNikwhr do i type this febs?
19:50:50AceNikhere itself in the msg box
19:51:30*bluebrother waves at obo
19:51:39*obo waves back
19:51:54bluebrothercan you help me killing a warning?
19:52:05oboI can try
19:52:14*amiconn summons dan_a
19:52:16 Quit linuxstb (Read error: 110 (Connection timed out))
19:53:29AceNikfebs: it says im not registered
19:53:42AceNikill brb in 30 mins then ill get back to you ok
19:53:59FebsAceNik: yeah, I need to get lunch anyway.
19:54:07 Part AceNik
19:54:34bluebrotherwell, when building rbutilqt I get a warning rbutilqt.cpp:30: warning: unused parameter ‘parent’
19:55:08bluebrotherI'm pretty sure this is related to me setting up the UI with this as argument instead of parent
19:55:43*obo grabs the latest source
19:55:48bluebrotherbut when using parent it doesn't work −− parent is of type QWidget* and ui is QMainWindow*
19:56:12bluebrotherbut as far as I understood I shouldn't initialize with this.
19:56:25bluebrotherand I haven't found out what I'm doing wrong
19:56:37FebsHere is the Apple patent application:
20:01:08 Quit webguest30 ("CGI:IRC")
20:01:53 Quit hannesd (Connection timed out)
20:04:05 Join amar [0] (n=amar@
20:04:22 Quit Nico_P (Remote closed the connection)
20:06:58amiconnjhMikeS: Do you think it would be possible to use dma for pcm data transfer?
20:09:19obobluebrother: (void)parent; works, I don't know if there is a better way
20:09:45*amiconn has a suspicion why battery life in rockbox sucjs on PP
20:10:20bluebrotherhmm. I know about the void cast, but I think I did something else that's not 100% ok
20:10:31*Domonoky is compiling bluebrothers new code..
20:13:08 Join entheh [0] (
20:13:13bluebrotherah. Seems I found something :)
20:13:34oboRbUtilQt::RbUtilQt(QWidget *parent) : QMainWindow( parent )
20:13:43 Quit lxx ()
20:13:48amiconnnah, theory discarded
20:14:53 Nick hannesd__ is now known as hannesd (
20:15:25DerPapstdamn :(
20:15:44Domonokybluebrother: your new version works fine in windows..
20:15:57bluebrotherDomonoky: I checked in windows this time ;-)
20:16:08amiconnBut halving the fiq frequency like on sansa should help
20:16:48Domonokybut you should change the signal on which "updateDevice" is triggerd from highlighted to index_changed
20:18:02bluebrotheryou mean the slot I assume?
20:18:54Domonokythe signal from the combobox...
20:19:00 Join Jens [0] (i=Jens@pdpc/supporter/active/Jens)
20:19:18bluebrotherwell, I can't change that signal easily
20:19:46Domonokyyou should just, use the other signal :-) the signal named "index_changed"
20:20:35 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
20:20:43bluebrotherah, now I see what you mean. Stupid me ;-)
20:21:42bluebrotherthere was some reason why I didn't use that signal in the first place. Seems it wasn't a good reason :P
20:21:57 Quit PaulJam (".")
20:22:45 Quit DerPapst (""need all available ram. brb"")
20:22:46Domonokythe only problem is, that the signal is also emited, when you change the index with a function call..
20:23:02Domonokybut thats not really a problem..
20:23:12bluebrotheris this really a problem? It nicely updated the UI in that case ;-)
20:23:24 Join kaaloo [0] (
20:23:34bluebrotherok, my intention is to get the code into a subfolder of the rbutil module in svn soon.
20:23:37 Part kaaloo
20:23:52Domonokybluebrother: that would be really good :-)
20:24:15Domonokybluebrother: have you done anything about error handling ?
20:24:16bluebrotherso others can fix the errors I made :P
20:24:31bluebrotherI added something about handling http errors.
20:25:10 Quit Rondom (Nick collision from services.)
20:25:11bluebrotherfor installation I'm wondering if it's sufficient to have the list and put errors there. That's the way I'm currently doing it.
20:25:26 Join Rondom [0] (
20:26:56 Join pepie34 [0] (
20:28:21Domonokyhm, this list could work for all installations..
20:28:44bluebrotheryeah, that was one of my thoughts. Haven't checked against the other installs though ;-)
20:29:29Domonokywe dont have a progress for all installations, but we could show this list even if there is nor progressbar..
20:29:40 Join thegeek [0] (
20:30:11Domonokyand the important errors all happen while installing something..
20:30:13bluebrotherideally that list gets replaced with a nicer one which has icons at the beginning
20:30:33 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
20:30:35 Join PaulJam [0] (
20:30:38bluebrotherand if a step finished successfully, display a checkmark. If it fails, add a warning sign
20:30:55 Quit Rondom (Nick collision from services.)
20:31:11 Join Rondom [0] (
20:32:41bluebrotherhmm, still a warning about the unused parameter parent. But it seems at least the left-over temporary files are gone now.
20:32:49linuxstb_amiconn: 1) No, mpegplayer doesn't unboost the CPU - but you're right, it could do. 2) Yes, the hw revision is somewhere else on the 5g - I'll try and find it...
20:33:55linuxstb_amiconn: Also, regarding "buffer miss" in the midi player, there's a patch on flyspray which claims it solves it - but I didn't commit as I don't have any midi files to test the plugin with. Maybe you could try it - it simply increases the size of a #define
20:37:22 Quit Mouser_X (Nick collision from services.)
20:37:33 Join Mouser_X [0] (
20:40:17bluebrotherhmm, there is a similar parent issue in configure.cpp which I overlooked all the time ...
20:42:42 Join advcomp2019 [0] (n=advcomp2@
20:43:25 Join NirIzr [0] (
20:43:44 Join Siltaar [0] (
20:46:12 Join DrMoos [0] (
20:47:09 Quit moos (Read error: 104 (Connection reset by peer))
20:48:25 Join kubiix [0] (
20:48:29Domonokybluebrother: make the constructor look like this: Install::Install(QWidget *parent) : QDialog(parent)
20:48:51Domonokyie. call the constructor of the base class with the parent...
20:49:34 Quit Rondom (Connection timed out)
20:50:48 Quit pepie34 ("Ex-Chat")
20:54:11 Join DerPapst [0] (
20:54:13bluebrotherah, great.
20:54:25DerPapstwow this takes longer than i thought...
20:54:49bluebrotherI'm still not that familiar with cpp right now. Oh well.
20:56:32 Join webguest66 [0] (i=4a77eac8@gateway/web/cgi-irc/
20:56:40 Quit webguest66 (Client Quit)
20:57:08saratogalinuxstb: do you know why the MAD speedup for ARM patch (FS:6705) was never committed?
20:57:12 Join webguest66 [0] (i=4a77eac8@gateway/web/cgi-irc/
20:58:21DerPapstsaratoga: it broke building mpegplayer on 3G iPods at least.
20:58:21 Quit webguest66 (Client Quit)
20:58:30DerPapstdunno if that has been fixed though
20:58:46saratogais that the only reason?
20:59:05DerPapstdon't know that either ;)
20:59:36DerPapsti would like to have that changes for 3G iPods even if it breakes mpegplayer
21:00:46DerPapstvideos on 3G are a pita anyways. the contrast of 1G-3G iPods changes depending on the displayed content :-/
21:02:17saratogaDerPapst: mpegplayer builds for the 3G with that patch
21:02:38saratoganot sure if it actually runs, though theres no errors and just a trival warning about an unused variable
21:03:18DerPapstwell then i wonder why it isn't in svn yet.
21:03:20linuxstb_saratoga: I recall preglow wanted to test the accuracy of the output, but he never got around to it. It's possible the patch may also need a little cosmetic cleaning.
21:03:56*DerPapst doesn't wonder anymore
21:04:50*DerPapst has actually never tested mpegplayer on his 3G iPod
21:05:45DerPapstif i convert a movie to play it on my 3G do i have to reduce it's color depth to 2 bit b&w?
21:06:04saratogalinuxstb: I believe preglow is on vaction this week so i probably can't ask him
21:06:22saratogahowever, I could pretty easily decode a few random files and compute the RMS error verses the current decoder
21:06:28linuxstb_It's pretty much unusable on greyscale PP targets - the greyscale library (including LCD updates) and audio decoding are both on the same core, and it can't keep up.
21:06:35saratogawould that be sufficient?
21:06:54linuxstb_saratoga: I think so, yes. It's a pretty major patch, so we want to be sure it doesn't affect the output.
21:07:15linuxstb_It would be nice to finally commit it.
21:07:41DerPapsti've only used that video player in iPL which plays uncompressed avi.
21:09:04linuxstb_On greyscale targets mpegplayer uses the same greyscale lib that the jpeg viewer uses in order to display about 33 (I think) shades of grey.
21:09:34linuxstb_But no-one has done any work to get it working efficiently...
21:09:49***Saving seen data "./dancer.seen"
21:10:28DerPapstso i won't have to reduce the colourdepth of the movieclip?
21:10:39linuxstb_No. I don't think you can with mpeg anyway.
21:10:57DerPapstheh.. ok
21:11:21linuxstb_The colour info is stores as yuv422, and mpegplayer just uses the y component for display on greyscale targets.
21:11:35 Quit morrijr (Read error: 110 (Connection timed out))
21:11:36*linuxstb_ is infected with DerPapst's typo skills
21:11:42advcomp2019how many people are working on the e200r series to get rockbox on it i am wondering
21:11:59linuxstb_advcomp2019: I believe the current number is zero.
21:13:33DerPapsti know that ~33 shades of grey is possible at least archos targets but i wasn't aware that b&w iPods can display that may shades with rockbox too.
21:13:52DerPapstinsert "on" where it fits
21:13:55linuxstb_amiconn got it working relatively recently.
21:14:32linuxstb_I _think_ the greyscale lib now works on all greyscale targets.
21:15:17DerPapstwill check that when i'm bored again ;)
21:15:57 Quit Arathis ("Bye, bye")
21:18:03 Join Guile`` [0] (n=Guile@
21:18:45 Join Arathis [0] (
21:20:11DerPapstbah.. out of memory after one hour calculating :-/
21:20:14Arathisamiconn: player finally ran out of battery without any crash/freeze
21:22:14linuxstb_Arathis: How long did it last? (or wasn't it fully charged)?
21:23:03 Join AceNik [0] (n=AceNik@
21:23:09Arathiswas fully charged. comparing the results
21:23:26AceNikhey Febs
21:23:58AceNikno how do i do it
21:24:15 Join sarixe [0] (
21:24:34FebsHmm. Apparently you can see my private messages but somehow you're still responding on the public channel.
21:24:59Arathisamiconn: my last runtime-test without crash was about 8h 15m. this time it was 9h 20m
21:25:00AceNikcan you see mine ?
21:25:38AceNikwhere which site do i go to register?
21:26:41DerPapst /ns REGISTER password
21:26:46linuxstb_Arathis: That's a nice improvement, but still nowhere near iriver's quoted 16 hours... ;(
21:26:57Arathissadly true
21:28:40NirIzrI've got iriver H10 (6G) and it goes for like 8 to 10 hours... so theres not that big of a differance...
21:29:15AceNikfebs: i registered
21:29:29AceNikDerPapst: thanks a lot
21:30:05 Quit amar ("Konversation terminated!")
21:30:26DerPapstand next time you do /ns IDENTIFY password
21:30:27ArathisNirIzr: 8-10h with rockbox or OF?
21:32:28linuxstb_I thought iriver quote 16hr for the 20GB and 12hr for the 5/6GB?
21:32:50AceNik Arathis: on rockbox also u can get 8-10 hrs if you have decent pair of headphones, so ul listen at a lowwer voluume since headphones would be good
21:33:40bluebrotherok, I'm going to commit my rbutil work to /rbutil/rbutilqt shortly. Does anyone have objections against this?
21:33:46NirIzrive got good headphones... but i do thuch my iriver while playing (check the song, change etc)
21:37:09 Join spiorf_ [0] (
21:39:34AceNikguys dont you think there need to be days assigned as patching days, like on those days patches can be synced to rockbox, cause otherwise loads of features can go waste as they'll be patented by someone else
21:40:52 Part NirIzr
21:42:14GodEaterAceNik: we don't much care about the patents though - software patents are not valid in the EU
21:42:36 Join moos [0] (
21:43:01linuxstb_And the existence of a patch in the tracker is enough to prove prior art (I would hope...)
21:43:04 Quit DrMoos (Read error: 104 (Connection reset by peer))
21:43:06amiconnAceNik: Power drain by headphones is neglectible compared to the rest of the player, unless one is almost deaf and is listening at really high volume
21:43:13amiconnBut even then it's just a few percent
21:43:37amiconnI do all my runtime tests at line level without headphones connected
21:43:51AceNikGodEater: i dont understand, would the development of rockbox be hampered? , i understand they limited scope but still , cant you guys include them n then under GPL do something, like evert patch added for review also gets a stamp from GPL belonging to you first
21:44:20FebsGPL addresses copyright issues, not patent.
21:44:25 Quit moos (Client Quit)
21:44:41amiconnArathis: Nice to hear :)
21:44:47GodEaterAceNik: as linuxstb says - if we were to start caring about software patents, and I can't see why we would, just the fact that we have them in the patch tracker is enough
21:45:19AceNikamiconn: i had the original iriver senhessier headphones, i used to listem to music at -8, now i bought new ones' something with 32ohms impedance, i listen at -24
21:45:24 Join PaulJam_ [0] (
21:45:26amiconnI hope to commit the fix soon, but I also want to rework the clock setting function so that it looks like it should, and there is one detail in the current setup I don't understand
21:46:33 Join veitchy [0] (i=519d3382@gateway/web/cgi-irc/
21:46:35amiconndan_a could probably tell my why he did it this way, but he isn't around often these days
21:46:50AceNikGodEater: fine but i still feel ultimately chances are stuff would be copied & reused somehow or the other, & you guys need to have a stronger ground
21:46:53amiconnMaybe linuxstb has an explanation though...
21:47:16veitchyhi, am i in the right place to seek some tech help for a newbie??
21:47:46FebsAceNik: patents are not the solution.
21:47:53AceNik veitchy: go on
21:47:57GodEaterAceNik: but this is a GPL project - we don't mind people copying our work - that's sort of the point
21:48:05AceNik veitchy: have you read the manual first
21:48:08GodEaterprovided they credit us - it doesn't matter
21:48:25GodEater(and include the source for whatever they do with our work too)
21:48:46veitchythanks, just installed RB on a ipod nano and manged to set display background colour the same as text....
21:49:00veitchycan't read menus
21:49:08GodEaterveitchy: just reboot it, and as soon as you see the rockbox logo switch the hold switch on
21:49:19GodEaterthat will reset the settings to their defaults
21:49:22AceNikGodEater: the most common stuff is the lyrics display being such a nice feature & you not interested in it since it still inst committed, but the same patent is on its way & do you think apple will credit anyone else for it
21:49:23GodEaterand you'll be able to read again
21:49:27veitchycheers GodEater
21:49:48veitchymuch appreciated
21:49:49GodEaterAceNik: I don't give a toss if apple patent it or not - their patent is not enforceable in the EU
21:49:55 Part veitchy
21:50:14AceNik veitchy: or better go into disk mode & delete the background, ul boot into normal background with same settings
21:50:55 Quit spiorf (Read error: 110 (Connection timed out))
21:50:59AceNikGodEater: but isnt rockbos goin to launch as a software globally ?
21:51:07AceNikhow will it matter then ?
21:51:10FebsRockbox is already global.
21:51:11GodEaterAceNik: define launch
21:51:15AceNikrockbox* so sorry
21:52:08AceNikGodEater: the day a official working version stable build is ready dont you guys think advertising & launching would be an option ?
21:52:20GodEaterAceNik: what on earth for ?
21:52:27AceNiklike you guys are an alternate OS right ?
21:52:38GodEaterAceNik: so ?
21:53:00AceNikGodEater: dont you think this will be way bigger someday ?
21:53:01 Quit PaulJam (Read error: 110 (Connection timed out))
21:53:28linuxstb_amiconn: What's your question about the system init?
21:53:31GodEaterAceNik: whether it is or isn't, is not a matter for advertising and a "launch" whatever that is
21:53:40bluebrotherRockbox has been released as stable. The last release was 2.5 for the archos players
21:53:55GodEaterAceNik: Rockbox is not a commercial venture, and I don't foresee it ever becoming one
21:53:58bluebrotherand as nobody wants to sell Rockbox, what's the point in some "launch"?
21:54:40thegeekI imagine he is talking about some kind of intensive launch to attract new users
21:54:43PaulJam_some people sell rockbox (or rather the installation service)
21:54:47AceNik GodEater: i dont belive in software being sold, but there needs to be a certain brand value which has to be developed
21:55:06LloreanAceNik: Why?
21:55:07GodEaterAceNik: no there doesn't
21:55:31bluebrotherwell, that's the same as linux distros: you can sell the software, but anyone can still get it from the project itself
21:55:44AceNikdoesnt the linux project have a certain brand value
21:55:49*bluebrother smells a "product" issue
21:55:51GodEaterAceNik: No it doesn't
21:56:02GodEaterAceNik: the people who make money with linux have brand value
21:56:08GodEaterAceNik: such as RedHat, IBM
21:56:13GodEaterlinux itself has nothing
21:56:20AceNikthose are individual companies
21:56:28bluebrotherwell, "Linux" is a registered trademark. But that's all.
21:56:32LloreanAceNik: Redhat has brand value. And Ubuntu does. Linux is just Linux, and most of the "brand value" of that term is more misconception than anything.
21:56:54GodEateranyone who thinks "linux" has brand value, has no concept of what it is
21:56:59bluebrothertechnically, linux is just the kernel. You can't do much with it without any userland ...
21:57:10AceNik GodEater: well i dont get the future of RB then , may be you can be generou to tell me the future of RB
21:57:23GodEaterAceNik: we don't think about the future
21:57:26GodEaterwe do this for fun
21:57:43LloreanAceNik: The future of Rockbox: People who are interested in it will continue to improve the parts of it they're interested in.
21:57:46AceNik GodEater: well its loads of fun & learnin im thankful for that fo sure
21:57:47LloreanWhy does there need to be any more than that?
21:58:09bluebrotherRockbox is *not* a product nor intended to become one. It's just fun.
21:58:38AceNikbut people who are interested in it like me the ambitious one's will also long for every other ipod to have this piece of code, cause its so much better
21:58:39GodEaterif at some point a DAP vendor chooses to use Rockbox as the basis of their product then marketing it becomes their problem - not ours
21:58:50 Quit Rob222241 ()
21:59:04bluebrotherif you want to work on something that is intended to become some sort of product you might look at ubuntu or similar.
21:59:17LloreanAceNik: You're more than welcome to long for other people to install it. But it doesn't make *my* iPod any better if Bob down the street installs it, and honestly, if he's happy with what he's got why should he have to learn something new?
21:59:29GodEaterLlorean: hear hear
21:59:38GodEaterI don't give a monkeys what Bob's ipod is running
21:59:47AceNikguys dont you wish you had more men working on RB project , & also most of the DAP's have rockbox on them ?
22:00:09GodEaterAceNik: we welcome volunteers, but we don't want world DAP domination
22:00:13LloreanAceNik: It takes a special sort of person to actually work on the coding. That sort of person is much more likely to find Rockbox as it is.
22:00:44bluebrotherwe don't people to just work on Rockbox. We want people who are interested in it.
22:00:51GodEaterfrankly I could do with *less* people using it - would make the forums a much quieter place ;)
22:01:00LloreanMore popularity can of course increase visibility, but it also greatly increases support overhead. If you try to present it as a product (which is what you do when you advertise) people begin to *expect* an established support base.
22:01:14AceNikLlorean: i dont know , like im confused with ur ideas, but gotta live with them
22:01:14bluebrotherthere's a difference between coding for some random project and being interested in it.
22:01:25*Febs revives the motion to remove Doom.
22:01:29*GodEater is willing to do full time support when someone is willing to pay his mortgage
22:01:35*GodEater seconds Febs
22:01:40LloreanThe best people *for* the project are the people who want more out of their MP3 player.
22:01:51*bluebrother thirds Febs
22:02:01linuxstb_AceNik: As an example, the Rockbox port to the ipods didn't bring a huge influx of developers. In fact, I'm hard-pressed to think of any.
22:02:10AceNikwhy are you guy wnating to remove doom now ?
22:02:23GodEaterAceNik: because none of us like it!
22:02:26FebsWe're (somewhat) joking.
22:02:33bluebrotherremove doom!
22:02:34LloreanAceNik: It's a support nightmare and generally brings in a lot of people who provide a higher than average support load without contributing?
22:02:37FebsBut it brings nothing but support headaches.
22:02:52bluebrotherbecause it's annoying to see people installing Rockbox for the sole purpose of playing doom on their Ipod
22:02:52LloreanAceNik: Basically, it's an example of what makes Rockbox harder to work on.
22:02:56GodEaterand it makes people think Rockbox is about playing games - which it isn't
22:03:14AceNik linuxstb_: trust me there are loads of them who dont know, im in india ,& the amount of guys interested to work on the code itself is a lot , but again we guy are just studyin basics of programming yet
22:03:15*bluebrother didn't want to call it "support nightmare" though it in fact is
22:03:32GodEaterbluebrother should be braver ;)
22:03:47bluebrotherremove doom! :P
22:03:53AceNiki feel you should nt remove doom come on guys, even the iphone has doom in developement
22:03:58GodEaterand rockboy!
22:04:18LloreanAceNik: The point is that games detract from the focus, and attract a lot of people whose only interest is the "shiny" factor
22:04:35*bluebrother didn't heard any complaints and is about being brave enough of commiting his qt work ...
22:04:42AceNikwell the shiny factor just needs another forum probably
22:04:43LloreanAceNik: They aren't seriously talking about removing Doom, only using it to illustrate the kinds of hurdles that develop. Advertising would simply create more of these hurdles.
22:04:53*GodEater is semi-serious
22:05:10LloreanGodEater: And semi-serious is not "serious" so there. :-P It has a few extra characters.
22:05:29GodEaterput it another way - if it happened - I wouldn't cry
22:05:29 Join klbostee [0] (
22:05:35LloreanAnd there will be a "shiny factor" forum over my dead body.
22:05:52GodEatercan't we arrange to get doom / rockboy forked into a seperate project ? :)
22:06:05bluebrotherwe shouldn't ship doom but require user who want to use it to compile themselves ;-)
22:06:09LloreanThere's already a WPS and Customization one, that's more than enough nod to the people whose sole interest is in making their player look cooler.
22:06:24klbosteeI fixed a small bug in apps/playback.c, and I would like to submit a patch for it. How should I create this patch?
22:06:36linuxstb_klbostee: Did you checkout the code from SVN?
22:06:43bluebrotherklbostee: best way is svn diff
22:06:45GodEaterchrist - someone who's brave enough to fiddle with playback.c
22:06:46AceNikok well you know guys i saw what the amount of work the devcon's can do, having to develop extra eyecandy stuff is a days work , but the features need to be revised all year long
22:06:51GodEatersign him up right now!!!!
22:06:56linuxstb_klbostee: Then just type "svn diff apps/playback.c > mypatch.diff"
22:07:08amiconnlinuxstb: If you would take a look at system_init() in system-pp502x.c - do you have an idea why ipod_set_cpu_frequency() is executed on the cop if dualcore is enabled and adjustable cpu frequency is disabled?
22:07:08klbosteeok, thanks
22:07:10AceNikbut guys the devcon west had promised there would be days for patches & bugs dealing
22:07:19 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
22:07:23 Join miepchen^schlaf [0] (
22:07:29amiconnAll the other inits are executed on the cpu....
22:07:34GodEaterAceNik: I doubt they used the word "promise" anywhere
22:07:36LloreanAceNik: I don't know where you think we promised *anything*
22:07:36bluebrotherthere were no promises
22:07:41LloreanI know I made no promises.
22:07:50*GodEater doesn't recall reading any such thing
22:07:55AceNikcome on guys
22:07:57bluebrotherand having a FS cleanup period was discussed at DC, not DCW
22:08:03*amiconn disagrees both with GodEater and Febs
22:08:21GodEateramiconn: about doom ?
22:08:46GodEateramiconn: well you spend no time reading posts about it in the forums - so I guess you're not as fed up with it
22:09:07amiconnI know that many questions arise regarding doom - but if we wouldn't have it, there would be even more questions
22:09:32AceNikLlorean: what happened to the screensaver stuff , i discussed with you last , & also the ability to switch between plugin & wps at the same time when they running?
22:09:32*bluebrother thinks it would only be a feature request rotting in the tracker
22:09:42amiconnBoth why we don't have doom, and why we don't port <insert favourite not too recent 3d game here>
22:09:44*GodEater agress with bluebrother
22:10:01GodEateramiconn: not too recent ?!?! Did you read the request for doom 3 ?
22:10:03LloreanAceNik: I have no idea what you're talking about, with screensaver stuff.
22:10:20Febs"why we don't port <insert favourite not too recent 3d game here>" <−−- This question gets asked repeatedly anyway.
22:10:25AceNik] amiconn: guys seriously let there be a separate forum for core technicalities & whih on;y developers have access to
22:10:38*linuxstb_ doesn't understand why people don't just ignore forum posts that annoy them.
22:10:41GodEaterAceNik: we don't need it - we have this irc channel
22:10:48amiconnGodEater: Nope, but with the real doom (1,2) as an example, it becomes easier to explain why doom 3 isn't possible
22:10:53AceNiklinuxstb: ok how bout ability to switch between plugin & wps at the same time when they running?
22:11:07linuxstb_AceNik: Yes, that would be nice...
22:11:11LloreanAceNik: Why should there be a separate, hidden forum? There's this IRC channel, the developer mailing list, and categories in the forums that are relevant to different parts of the code.
22:11:29bluebrotherirc is *much* nicer than forums anyway
22:11:34*amiconn does what linuxstb wonders ;)
22:11:43LloreanIRC + Mailing List is more than good enough, I think
22:11:49amiconnJust ignoring annoying forum posts I mean
22:11:50AceNikamiconn: who thought of having doom3 on their DAP ......hehe this drives me crazy to think that guys knowledge ont he DAP's power
22:11:54linuxstb_amiconn: That function is a maze of #ifdefs...
22:12:09GodEaterif everyone ignored posts in the forums which annoyed them, then what's the point in having the forums in the first place ?
22:12:10FebsBelieve it or not, I ignore MANY forum posts.
22:12:21*Llorean ignores more now than he should.
22:12:26amiconnlinuxstb: Not as much as many other parts. My goal is to clean up this stuff - completely
22:12:35pixelmaAceNik: ;)
22:12:37AceNiki ignore plenty , but sometime answer new commers, i was one once
22:12:37amiconnBut I like to understand why it's doen on the cop...
22:12:45amiconnIt doesn't make sense to me
22:12:54GodEaterif we provide forums - we sort of have a responsibility to read them and answer them
22:13:10LloreanFebs: That shadow person who was posting on the Zune and Nano 2G apparently thinks you're harassing him.
22:13:16amiconnAceNik: If I'd do that, >90% of my posts would consist of 4 letters (or 2x 4 letters)
22:13:23GodEaterotherwise Llorean, Febs and everyone else should just hand in their badges and let the forums become a complete free for all
22:13:28FebsLlorean: yes, he sent me a PM to that effect. Which I ignored. QED.
22:13:34markunLlorean: he's pretty anoying
22:13:34amiconnrtfm, rtfw
22:13:48 Quit sarixe (Connection timed out)
22:13:56LloreanFebs: I gave him a lecture on how being 15 (or whatever he told me his age was) doesn't make the rules any less valid, and that he'll be expected to follow them just like everyone else.
22:14:00 Join jgarvey [0] (
22:14:20GodEaterFebs: are you still deleting posts from him ?
22:14:31LloreanSo basically, if he gives you any more trouble, I've no objection to ridding ourselves of him for a week or two.
22:14:34amiconnlinuxstb: I wantto move the pll power init to system_init() - there really is no reason to do it on every clock change. On is on...
22:14:38AceNikthat guy is mad
22:15:10FebsGodEater: He hasn't posted since I wrote the fairly stern public post asking him not to post in the New Ports forum. Instead, he has started complaining privately to Llorean and me.
22:15:37GodEaterFebs: you're so mean ;)
22:15:41FebsI did not want to issue that public post, but I PMed him several times and he just ignored me.
22:16:01AceNikok guys seriously can there be a unision between core developers & newbie's someday on an IRC, & that day use to work on bug & patches, cause you never know if someone comes up with a nice feeature
22:16:17GodEaterAceNik: that's why we have flyspray
22:16:35 Join Davide-NYC [0] (
22:16:48AceNikya but i feel flyspary is ignored till someone comes up with a number on the IRC & then people check
22:17:00GodEaterit's not ignored
22:17:04Davide-NYCRelevant feature request! /msg NickServ IDENTIFY
22:17:10GodEatermany of the devs spend more time reading it than reading the forums
22:17:15Davide-NYCRelevant feature request!
22:17:16AceNiklike i did with the snc viewer , but i t being spoken of no
22:17:18markunAceNik: better to give a number in IRC than trying to explain it here
22:17:42LloreanAceNik: Because many developers have plenty of stuff to work on.
22:17:56LloreanThey can't all drop the bugs they're trying to fix just because *you* like a feature.
22:18:02LloreanSo patience is a virtue. Mention it, but don't be a pest.
22:18:03markunand feature request != feature demand
22:18:32AceNikLlorean: i know that but ther has to be a day, like how you have days for laundry
22:18:38AceNik& iorning
22:18:45GodEaterno there doesn't
22:19:12AceNikok guy but atleast comment on the sncviewers's inclusion?
22:19:20LloreanAceNik: There doesn't have to be a day. People get to things when they get to things.
22:19:24markunAceNik: we work on rockbox for fun. If your idea of fun is ironing and laundry then work on that.
22:19:43bluebrotherThings in here work. Why is there a need to change everything?
22:19:48GodEatermarkun: hahha - well said ;)
22:19:54AceNikmarkun : sorry for that it was just an example
22:20:00markunI know
22:20:04LloreanAceNik: For a very very long time the sncviewer author couldn't be bothered with even posting his patch to flyspray
22:20:22*bluebrother sighs
22:20:31AceNikok back to work then , sncviewer is newly reworked
22:20:38AceNikits done by bascule now
22:20:57AceNikit works perfect now, with improvements to UI & stuff
22:21:06LloreanIt's not done by Bascule, he simply posted it.
22:21:12AceNikthe earlier one was on evenon the FS
22:21:15LloreanAs well it sounds like the code needs a lot of cleanup to readability
22:21:19bluebrotherGodEater: I was brave and did a huuuge commit ;-)
22:21:27GodEaterbluebrother: yay!
22:21:28AceNikthen whatever it is its reworked by someone
22:21:44bluebrotherso let's see what people will say about my try ...
22:22:09linuxstb_amiconn: I'm not sure I can help explain it... IIUC, it's using the alternative ipod_set_cpu_frequency() function, which I'm guessing is the CPU init IPL does - the one that's incorrectly commented as !defined(BOOTLOADER)
22:22:14GodEaterAceNik: just because it works does NOT mean it deserves inclusion
22:22:26GodEaterAceNik: if it's completely unfathomable, then we don't want it - because we can't support it
22:22:34AceNikGodEater: ok
22:22:43*bluebrother LOLs at the output of sloccount
22:23:02linuxstb_amiconn: Ignore my wrong remark about !defined(BOOTLOADER)...
22:24:31bluebrotherwhich reminds me ... anyone against rejecting FS #7459?
22:25:17 Join Rondom [0] (
22:26:01LloreanI'd say go ahead
22:26:42linuxstb_AceNik: Speaking personally, I'm more than happy with the features Rockbox has currently (I only use a tiny percentage of them), and don't feel a desire to work on new ones. I would rather see the existing features work better.
22:27:37AceNiklinuxstb_: i agree, which brings me to the point can you tell me how to code around the H10 for rockbox, would like to help
22:28:09AceNiki reliase h10 has less developers, last2 devcons there was no one listed
22:28:10linuxstb_AceNik: 1) Learn C; 2) Checkout the Rockbox source and try to change something; 3) Ask specific questions here.
22:28:13GodEaterspeaking of new features that none of us will use - I'm hoping to have a working beta of my "favourites" plugin working sometime tomorrow
22:28:34GodEaterimplemented as Llorean suggested
22:28:42LloreanAceNik: 95% of work has nothing to do with a specific player anyway. It usually affects a category (all PortalPlayer) or all players.
22:28:43linuxstb_GodEater: I look forward to ignoring it when it's on flyspray.., :)
22:28:51 Join belze_ [0] (
22:28:59AceNiklinuxstb_: AceNik: 1) Learn C : already am ; 2) Checkout the Rockbox source and try to change something −−−− already figured out ; 3) Ask specific questions here.−−−− will surely do that
22:29:04markunGodEater: a mame golden quote I think:
22:29:04GodEaterlinuxstb_: I look forward to absolutely no comments from you about it then ;)
22:29:30belze_is there a possibility, that might become possible on the h110 as well?
22:29:31linuxstb_GodEater: Have you managed to decipher the list widget code now?
22:29:49GodEaterlinuxstb_: thought I'd said that last night - yes I think so ;)
22:30:02linuxstb_GodEater: My memory of last night is a little hazy...
22:30:06AceNiklinuxstb_: things like how the hradware reacts with the software isnt clear to me although following up with rockbox has taught me a little
22:30:07GodEaterhaven't looked at it all day though what with having to go to a funeral
22:30:40GodEatermarkun: that reply is nearly perfect ;)
22:31:07GodEaterAceNik: That's a) Not even a question, and b) very unspecific
22:33:08AceNiklike barry told me i need to know the way the hardware of H10 is , & be able to understand what the codes in the OF are doin to the player , is that necessary ?
22:33:22Davide-NYCGuys, dig this super easy, reasonable adn relevant feature request. Should take very little time to implement and would make the recording contingent happy.
22:33:28 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:34:26GodEaterDavide-NYC: if it's so simple - why not do it yourself ?
22:34:32linuxstb_AceNik: It depends on exactly what you're looking to do. But if you want to work on a low-level feature that depends on the H10's hardware, then you obviously need to work out how to talk to that hardware.
22:34:50LloreanDavide-NYC: Also you pasted that in here a mere 15 minutes ago
22:34:51Davide-NYCGodEater: be careful what you wish for ;-)
22:34:58pixelmaDavide-NYC: and keep an eye on the Archos WRS ;)
22:35:40AceNik linuxstb_: lets c ill start with the " hello world " tomorrow first , then head for the RBUTIL stuff too
22:35:41Davide-NYCpixelma: I have a solution for the tiny LCDs...
22:35:59Davide-NYCthe joystick "UP" botton currently does nothing.
22:36:07AceNikfor now im goin , i wanted to help with the scrollpad on h10 when i had spoked to barry
22:36:28GodEaterAceNik: what has that got to do with rbutil ?
22:36:31belze_would generally work on the h110 in the current form?
22:36:35Davide-NYCI say it should be a momentatry hold redraw of the first two lines and we should hide the filename and filesize
22:36:40belze_or does it have a different flash?
22:37:01AceNikGodEater: i had loads of ideas i wanted t work with rbutil , i discussed it with dominik
22:37:19linuxstb_belze_: You should try and talk to Slasheri - he did the h120/h140 flashing work.
22:37:25belze_ah, ok
22:37:41AceNikGodEater: so i need to start with the UI first he said
22:38:03AceNikc ya tomorrow sorry for the trouble
22:38:22linuxstb_AceNik: The current rbutil appears to be being dropped in favour of a new Qt-based version...
22:38:32pixelmaDavide-NYC: I believe "up" has a function - at least on my Ondio... it's for selecting volume or gain to change them
22:38:45AceNik linuxstb_: whats Qt
22:38:54Davide-NYCYes, but one you are at the topmost item it does nothing
22:39:17bluebrotherQt is something that's really cute :D
22:39:28bluebrotheras the name says
22:39:37bluebrotherDomonoky: I committed!
22:39:49Domonokynice :-)
22:39:51bluebrotherso, let the gaaaaaames begin!
22:40:35 Quit lostlogic (Remote closed the connection)
22:40:51 Join lostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
22:40:54GodEaterit's been ages since I've done any Qt work ;)
22:41:34bluebrotherit's been ages since I wanted to find time to get around Qt ;)
22:42:23GodEaternow all we need is a GTK port, and then we have a mini gnome/kde war going on in the rockbox world :)
22:42:45*bluebrother grabs gvim and runs to the KDE front
22:43:07GodEaterhow ironic, isn't gvim GTK ? ;)
22:43:20bluebrotherit is :o
22:43:31GodEateryou're clearly very confused :)
22:44:01bluebrotherhmm. Not enough alcohol for my brain to work *g*
22:44:11linuxstb_GodEater: The current rbutil is gtk...
22:44:16GodEatersurely you should be using kate or something
22:44:21GodEaterlinuxstb: yeah sort of
22:44:30Domonokynot its wx !! :-)
22:44:35GodEaterbut not on windows it isn't
22:44:48bluebrotherkate is nice, but vim is great.
22:45:21*linuxstb_ prepares to compile qt on his mac...
22:45:35GodEaterhope you have a good cup of coffee handy
22:45:39GodEaterit's not a short compile
22:45:46linuxstb_I'm planning to leave it going overnight...
22:46:22GodEatergood plan
22:46:36 Quit AceNik ("bye guys have fun , enjoy !!!!!")
22:47:06bluebrotherhas anyone qt build statically around? I'm curious how big the binary will get.
22:47:20linuxstb_bluebrother: Are there any special options i need to include when compiling qt?
22:47:46bluebrotherI don't think so, I used the binary package of my linux distro.
22:49:03 Quit jgarvey ("Leaving")
22:49:09linuxstb_I guess I could try -static
22:50:10 Quit Davide-NYC ("ChatZilla [Firefox]")
22:51:21 Quit lostlogic (Remote closed the connection)
22:51:32 Quit amiconn (Read error: 110 (Connection timed out))
22:51:32 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:54:21 Quit Rondom ("Ex-Chat")
22:54:50 Quit GodEater ("User pushed the X - because it's Xtra, baby")
22:56:21 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
22:57:07linuxstb_Bagder: Hi. Thanks for the t-shirts and mugs - they were duely distributed last night.
22:57:50 Quit Domonoky (Read error: 104 (Connection reset by peer))
22:58:07 Part maffe
22:58:17 Join maffe [0] (
23:03:22scorchegood day mister endless vacation =)
23:03:38GodEater_Bagder: check the thread to see how they were fairly distributed
23:04:18Bagdera very good picture indeed
23:06:07 Join jgarvey [0] (
23:09:53***Saving seen data "./dancer.seen"
23:13:16 Join mbr [0] (
23:15:05*bluebrother just uploaded a w32 binary of rbutilqt at
23:15:12bluebrotherin case someone wants to try
23:18:09 Join Davide-NYC [0] (
23:18:21 Quit Jens ("Schlafen.")
23:18:30Davide-NYCSo I'm trying to tackle a bit of this code myself.
23:18:50Davide-NYCAt first I'm just trying to identify where my little bit of disaster should go
23:19:06Davide-NYCI'm trying to adda line to the recording screen
23:19:26 Join _jz [0] (
23:19:30Davide-NYCI figure line 1404 in recording.c is the correct location.
23:19:36_jzhi there !
23:19:39Davide-NYCAm I warm?
23:19:41_jzrockbox looks completely amazing
23:19:56_jzbut i just flashed my sensa E250 and now it doesn't appear anymore as a block device under linux :/
23:19:56Davide-NYCwe like it
23:20:09*Davide-NYC ducks
23:20:21bluebrother_jz: have you read the FAQ and caution pages in the wiki?
23:20:34_jzi just read the installation part of the doc
23:20:45_jzdid i burn something irremediably ??
23:21:26bluebrotherwell, have you read up on how to access usb mode?
23:21:41bluebrotherRockbox doesn't handle usb yet, so you need to boot the OF for usb access
23:21:58_jzOF = ? original firmware ?
23:22:26_jzach! crap! that's now officially the first and only defect i see in rockbox ;))
23:22:33scorcheas a note, rockbox is not flashed onto the sansa devices
23:22:47_jzyeah it's just added, sorry ;)
23:22:51 Quit jgarvey ("Leaving")
23:22:57 Join jgarvey [0] (
23:22:58_jzhmmm there should be a link to that faq in the doc !
23:23:06 Join metalic [0] (i=5b7e2023@gateway/web/cgi-irc/
23:23:10_jz(or maybe there is but i just jumped to the "installation" part :)
23:23:20*scorche points to the link in the topic
23:23:23bluebrotherand to answer the other question, you _can_ seriously brick a sansa, but it requires some work to do so
23:23:38 Quit PaulJam_ (Read error: 110 (Connection timed out))
23:24:05metalichi guys! I didnt see the part that said dont install on mac formated partition and now I have rockbox boot loader and a no partition found, nothing works HELPPP?
23:24:29 Quit miepchen^schlaf ("Verlassend")
23:24:43 Quit jgarvey (Client Quit)
23:24:54bluebrothermetalic: do a restore from itunes
23:25:07_jzscorche: thanx, sorry i didn't look at the topic
23:25:12metalicthe computer dosnt recognise the ipod
23:25:22bluebrotherhave you put it in disc mode?
23:25:24linuxstbmetalic: Put your ipod into disk mode manually (MENU+SELECT to reboot, then immediately press and hold SELECT+PLAY).
23:25:43metalicit accually says that on the screen as well
23:25:46metalicbut nothing happens
23:25:55linuxstbThen try again - it _always_ works.
23:25:57metalicmaybe because of the hsf+ format?'
23:26:05bluebrotherwell, does Itunes detect the Ipod?
23:26:08metalicokey 2 sec...
23:26:10linuxstbMake sure the hold switch is off as well...
23:26:11_jzcongratulations to everyone working on that r0cking project
23:26:22_jzrockbox bears it names very well
23:26:32_jzplus it made me buy a player i would have never ever bought otherwise
23:26:39linuxstbmetalic: You need to hold MENU+SELECT for a few seconds before it reboots.
23:26:45_jzso you should go and try to extort some money from Sandisk :)
23:27:29 Quit Davide-NYC ("ChatZilla [Firefox]")
23:27:55metalicOH THANK GOD!!
23:28:15metalicyou have to kinda press meny + play directly after reboot
23:28:31bluebrotherIpods are almost unbrickable (well, a hammer would always do ;-)
23:28:38metalichaha yeah
23:28:43metalicor droping it :p
23:28:59bluebrotherif you drop it hard enough
23:29:07metalichave you guys tested the jclix theme?'
23:29:17metalicor enough times ;)
23:29:21bluebrotherjclix needs an unsupported build afaik
23:29:37metalicthey have one included on their site
23:29:43metalicIts a really nice theme
23:29:51metalicclean and good looking
23:29:52_jzthanx guys, and sorry for the stoopid question of someone who freaked out before finding out about the FAQ :)
23:30:04scorchewell, just keep in mind that it is unsupported
23:30:14_jzMay the Great GNU be with you :)
23:30:21 Part _jz
23:30:44wishesthats incredibly sad :)
23:30:59metalicthanks for the great help!!!
23:31:09metalicIm gonna try the thing out anyway :D
23:32:12 Quit metalic ("CGI:IRC (EOF)")
23:32:33DerPapstwishes: what do you mean?
23:32:52wishesDerPapst, the Great GNU quote :)
23:33:53DerPapstRNU maybe
23:34:20DerPapst^ isn't as cool as GNU
23:34:25*DerPapst gives up
23:34:57 Quit Lear ("ChatZilla [Firefox]")
23:42:22linuxstb_bluebrother: Hopefully you can reduce the binary size of rbutilqt...
23:42:58linuxstb_(looking at the 12MB of qt .dlls you included)
23:44:01amiconnlinuxstb: The ifndef BOOTLOADER is on purpose afaics, so the bootloader doesn't touch clock setup
23:44:16bluebrotherI really hope this gets better when linking statically
23:44:39linuxstb_amiconn: Yes, I misread that as a comment, rather than a #elif...
23:44:55amiconnAnd I know that it's the alternative setup which is used when HAVE_ADJUSTABLE_CPU_FREQ isn't defined - what irritates me is that it's done on the cop when using dualcore
23:45:25amiconnAnd this can't be just a coincidence - there is another case for when running single core
23:47:04amiconn<unrelated>I wonder how many utils we need, perhaps one per gui toolkit?</unrelated>
23:48:32 Quit mbr ("User disconnected")
23:48:53 Quit davina ("xchat on Ubuntu 7.04")
23:49:22*amiconn rties to trace this back in the irc logs around dualcore commit time
23:50:07 Join fionnghall [0] (i=57f4b213@gateway/web/cgi-irc/
23:51:56 Quit fionnghall (Client Quit)
23:53:29 Quit spiorf_ (Read error: 110 (Connection timed out))
23:55:10linuxstb_amiconn: Could your perfcheck() function be run on the COP to test the COP's speed?
23:55:43amiconnWhy not? But both cores always run at the same speed - unless you figured out clock skipping...
23:56:42amiconnThe perfcheck is very simple - it's basically like udelay, but with irq and fiq disabled and counting cycles until a certain (calibrated) duration is over
23:56:49DerPapstis that even possible to run one core at a different speed?
23:57:07DerPapst*than the other
23:57:13amiconnIt should, if my (our?) interpretation of the pp product brief is correct
23:57:54amiconnDerPapst: Not independent of each other, but the product brief states "independent clock skipping feature" for the cop
23:58:16DerPapsti thought that would give a lot of problems on cashe
23:58:22amiconnI think that means the cop can be programmed to leave out clock cycles
23:58:40 Nick idnar_ is now known as idnar (n=mithrand@unaffiliated/idnar)
23:58:42amiconnI'm not at all sure in what way this feature works
23:58:57DerPapstbasicly like a nop instruction

Previous day | Next day