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 2010-11-02

00:01:48 Quit wodz (Quit: Leaving)
00:03:45 Quit bimbel (Quit: Verlassend)
00:03:46 Quit ender` (Quit: When I was a kid I used to pray every night for a new bicycle. Then I realised that the Lord doesn't work that way so I stole one and asked Him to forgive me.)
00:05:32 Quit linuxguy3 (Ping timeout: 276 seconds)
00:05:52 Join linuxguy3 [0] (
00:07:39 Join soap_ [0] (~soap@rockbox/staff/soap)
00:08:06 Quit soap (Disconnected by services)
00:08:09 Nick soap_ is now known as soap (~soap@rockbox/staff/soap)
00:08:21 Quit hebz0rl (Quit: Leaving)
00:08:38 Join soap_ [0] (~soap@rockbox/staff/soap)
00:09:35 Join user890104 [0] (
00:10:24soapskimming the logs for 3.7.1 and mpc I didn't see discussion, but I presume the mpc on coldfile issue is alone enough for 3.7.1?
00:11:09n1sthe fix was committed to the branch too, i dunno if that bug alone is worth a pointrelease
00:11:35n1smaybe we wait at least a week to see if any more good fixes pop up
00:12:02 Quit linuxguy3 (Read error: Operation timed out)
00:12:58 Join LeonTrollski [0] (~LeonTroll@
00:13:08 Quit LeonTrollski (Client Quit)
00:13:24soapIf it is a regression from 3.6 it is a black eye not to push out the fix. IMHO.
00:13:52 Join linuxguy3 [0] (
00:15:38n1safaict the bug was introduced in r25056, may 7 so it likely was in 3.6 too
00:16:05pixelmait sounded like quite a major bug (even if mpc doesn't seem to be used widely but playback is still the main part of Rockbox, I hope)
00:16:38n1swell it crashes, but only on some apparently very rare files
00:16:52kugelsoap: IMO we should wait at least a month until the point release, I bet there's more bugs to be revealed
00:17:13n1si'm not against a point release but having to do 3 of them within a week seems less than desirable i think
00:18:21soapI think two timeframe should be discussed and agreed to. One timeframe for "critical bugs" and one for "non critical bugs".
00:18:39 Quit Judas_PhD (Ping timeout: 240 seconds)
00:18:54 Quit n1s (Quit: Lämnar)
00:19:18soapThis being, as pixelma so rightfully said major, a crash on Rockbox's core functionality would (I assume) fall into catagory one.
00:19:50JdGordon|is there any real reason we couldnt just rerelease it as 3.7?
00:20:22pixelmaI think I remember 3.5.1 was only done for the Archos targets for a critical fix
00:20:24soap"What version are you using" "3.7" "Update to 3.7 to fix that crash"
00:20:58JdGordon|hehe ok, good point
00:21:27 Quit linuxguy3 (Ping timeout: 252 seconds)
00:21:42pixelmathis only affected Coldfire targets... on the other hand different point releases for arch or whatever thing in common is probably confusing too
00:21:45kugelpixelma: it was a different situation back then. we didn't plan to do point releases until now
00:22:18 Join linuxguy3 [0] (
00:22:47JdGordon|well, the point of the point release plan was to keep a much slower moving stable branch. We really don't need to release a enw point for every bugfix, if people are being hit with this particular bug they can always use a build from the stable branch untill 3.7.1 is released
00:23:07JdGordon|we shouldnt be releasing a new point branch every week which is probably going to happen if we do one fr each non-trivial bug
00:23:51 Part toffe82
00:23:55soappixelma, if one were to only fix one bug per target per release cycle out-of-sync point releases would not be so confusing (like in the 3.5.1 example) - but if multiple releases are envisioned one almost has to bump the version number (be there changes for a specific target or not) for all targets.
00:25:19soapjdGordon - why not issue major fixes within a short time period, and minor fixes as part of a monthy (or quarterly, or anually, whatever) roundup release?
00:25:27S_a_i_n_tI noticed part of my patch FS11578 has been committed....but now I *totally* don't know what I'm doing to get variable backlight brightness working again. If anyone could have a look at re-syncing FS #11578 for me I would really appreciate it. I personally believed it was in a committable state prior to it being's something that I would quite like to see get committed.
00:26:39JdGordon|soap: sure, but that gets into a big discussion of what the bar is. I personally dont see this fix as massive enough for an imediate re-relase
00:27:01JdGordon|obviously I don't use mpc, and mpc on coldfire
00:27:32soapI don't wish to argue the point, but I am curious (and will give you last word). If a playback crash isn't a major issue what is?
00:28:11soapI can't imagine anything more massive slipping through the current system.
00:28:28soapAs it will be only edge cases which aren't caught in the freeze week.
00:28:32kugelIMO no bug (unless it's "doesn't boot") if we manage timely point releases
00:28:39kugel+is major enough
00:29:22pixelmacrashes on H100/H300 aren't nice though as you need a paperclip to get out of them
00:29:45kugel3.6 is still downloadable in the meantime
00:29:47pixelmathe Iaudios have a hard power-off at least
00:29:57 Quit GeekShadow (Quit: The cake is a lie !)
00:30:02pixelmaerr... if you are on the road?
00:30:13 Quit linuxguy3 (Ping timeout: 252 seconds)
00:31:14 Join linuxguy3 [0] (
00:31:17JdGordon|soap: this was apparently brken since july, with only one bug report. I tihnk that shows it is in a little used codec on an even less used target. If there was lots of noise about it sure, but I don't think there is much point going through the hassle of doing a point release just for it
00:36:00***Saving seen data "./dancer.seen"
00:39:06 Join Rob2222 [0] (
00:41:15 Quit linuxguy3 (Ping timeout: 252 seconds)
00:42:13 Join linuxguy3 [0] (
00:49:20 Join Judas_PhD [0] (
00:51:27 Quit linuxguy3 (Read error: Operation timed out)
00:51:57 Part dutchie
00:55:11 Join linuxguy3 [0] (
01:02:18CIA-8New commit by 03pixelma (r28441): Manual, lamp plugin button table: fix typo. Simplify and rearrange the code while at it to improve readability, especially with regards to the ...
01:04:10CIA-8r28441 build result: All green
01:06:01 Part stacker55
01:06:31 Quit linuxguy3 (Ping timeout: 252 seconds)
01:07:10 Join linuxguy3 [0] (
01:10:59 Join binaryhermit [0] (
01:12:24 Quit binaryhermit (Client Quit)
01:14:18 Quit krazykit (Ping timeout: 276 seconds)
01:15:01 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
01:15:47 Join krazykit [0] (
01:15:58 Quit linuxguy3 (Ping timeout: 240 seconds)
01:17:08 Join linuxguy3 [0] (
01:19:56 Quit krabador (Quit: Sto andando via)
01:25:13 Quit linuxguy3 (Ping timeout: 252 seconds)
01:26:05 Join linuxguy3 [0] (
01:34:04 Quit linuxguy3 (Ping timeout: 252 seconds)
01:35:04 Join linuxguy3 [0] (
01:41:03 Join evilnick_ [0] (
01:41:59 Quit linuxguy3 (Ping timeout: 240 seconds)
01:42:59 Join linuxguy3 [0] (
01:43:47 Quit evilnick__ (Ping timeout: 255 seconds)
01:44:14 Quit S_a_i_n_t (Ping timeout: 255 seconds)
01:44:27 Join Horscht [0] (~Horscht@xbmc/user/horscht)
01:45:37 Quit Horschti (Ping timeout: 252 seconds)
01:46:12 Join S_a_i_n_t [0] (S_a_i_n_t@
01:48:37 Quit DerPapst (Quit: Leaving.)
01:51:26 Quit alexbobP (Ping timeout: 245 seconds)
01:56:13 Quit kugel (Remote host closed the connection)
01:58:17 Join alexbobP [0] (~alex@
02:20:29 Join LeonTrollski [0] (~LeonTroll@
02:26:10Topic"Rockbox 3.7 released! | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community | This channel is logged at" by ChanServ (ChanServ@services.)
02:29:40LeonTrollskii uninstalled everything and then did a fresh install of 3.7 using rockbox utility on my Sansa v1. im still having the problem where it cycles in and out of USB mode when its plugged into the USB.
02:30:12LeonTrollskialso tried 'current build' same problem.
02:31:18LeonTrollskiany fix for this?
02:31:28JdGordon|which sansa?
02:32:22saratogayou can use the OF for USB mode if you have to
02:32:44JdGordon|I thouhgt th fuzev1 usb was expected to be stable?
02:33:43LeonTrollskiwell, transfering files onto the microSD card with the originol is only letting me put 2GB on there
02:34:28JdGordon|do you mean it keeps disconnecting and reconnecting in rockbox when usb is plugged in?
02:34:34JdGordon|or its doing boot cycles?
02:34:42LeonTrollskii was hoping that rockbox would fix that.
02:34:55LeonTrollskithe USB symbol pops up
02:35:05LeonTrollskiand the PC detects it as a device
02:35:16LeonTrollskibut then it switches back to regular mode
02:35:24LeonTrollskiand the PC says it disconnected
02:35:29LeonTrollskiover and over
02:36:02***Saving seen data "./dancer.seen"
02:36:07saratogaLeonTrollski: are you sure you have a fuze?
02:36:11LeonTrollskisometimes staying in one mode for as much as 30 or so seconds but usually more like one or two
02:36:16saratogathe fuze OF will let you move more then 2GB
02:36:51saratogaso if yours doesn't, its either not really a fuze, or else something is very wrong with your computer or player
02:37:41LeonTrollskithey move on there but any folders added after 2GB show as empty when i acess them after the transfer
02:38:27LeonTrollskitransfering songs 1 by one does the same thing
02:38:50LeonTrollskiafter navigating away from the folder i put them in and back to it they are gone
02:38:58JdGordon|saratoga: is the e200 the only ones without sdhc support?
02:39:19LeonTrollskithe PC still shows the card as having the space filled though
02:39:27saratogaLeonTrollski: ah that makes more sense
02:39:33saratogawhere did you get the card? ebay?
02:39:50saratogaJdGordon|: the e200v1 and c200v1 I think
02:39:55saratogathird party seller?
02:40:40LeonTrollskiit could be, but it was still in original packaging.
02:47:12 Quit factor (Read error: Connection reset by peer)
02:59:02saratogathen its probably a fake card
03:00:23 Join factor [0] (
03:02:02 Quit michael_cheah (Remote host closed the connection)
03:18:18 Quit JdGordon (Ping timeout: 255 seconds)
03:27:46 Join kisak_ [0] (
03:31:02 Quit kisak (Ping timeout: 272 seconds)
03:32:26 Quit kisak_ (Ping timeout: 240 seconds)
03:34:28 Join kisak [0] (
03:40:15 Quit edboyer93 ()
03:54:06 Quit bluebrother (Read error: Connection reset by peer)
03:54:27 Join bluebrother [0] (
03:54:27 Quit bluebrother (Changing host)
03:54:27 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
03:59:41 Quit MethoS- (Remote host closed the connection)
04:11:21 Quit TheSeven (Ping timeout: 240 seconds)
04:15:18 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:21:19 Quit amiconn (Disconnected by services)
04:21:19 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:21:21 Quit pixelma (Disconnected by services)
04:21:23 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:21:25 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:21:39 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:29:21CIA-8New commit by 03rmenes (r28442): Begin committing keymap plugins for the Philips GoGear HDD6330. ...
04:31:06CIA-8r28442 build result: All green
04:36:03***Saving seen data "./dancer.seen"
04:39:28 Join evilnick [0] (
04:39:46 Quit Barahir_ (Ping timeout: 265 seconds)
04:41:03 Join Barahir [0] (
04:42:01 Quit evilnick_ (Ping timeout: 264 seconds)
04:56:02 Quit anewuser ()
04:57:21 Quit LambdaCalculus37 (Quit: bedtime)
05:16:39 Quit kisak (Read error: Operation timed out)
05:21:43 Join krazykit` [0] (
05:21:49 Quit krazykit (Ping timeout: 265 seconds)
05:22:54 Join RoC_MasterMind [0] (
05:33:23 Join mt [0] (29e999cc@gateway/web/freenode/ip.
05:33:35 Quit ps-auxw (Ping timeout: 240 seconds)
05:33:46 Quit mt (Changing host)
05:33:46 Join mt [0] (29e999cc@rockbox/developer/mt)
05:34:21 Part mt
05:34:30 Quit Horscht (Quit: Verlassend)
05:38:58 Join maffe [0] (
05:39:07 Quit maffe (Client Quit)
05:44:38 Quit Judas_PhD (Quit: This is a quitting message)
05:45:55 Join ps-auxw [0] (
06:07:33 Quit RoC_MasterMind (Quit: Leaving)
06:13:27 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
06:21:55 Quit antil33t (Ping timeout: 240 seconds)
06:30:23 Join Judas_PhD [0] (
06:36:07***Saving seen data "./dancer.seen"
06:38:01 Quit saratoga (Quit: Page closed)
06:45:55S_a_i_n_tliar: ping?
07:00:00 Nick krazykit` is now known as krazykit (
07:05:44 Quit Ramsey[LC] (Remote host closed the connection)
07:09:34S_a_i_n_tfyrestorm: Are you able to compile a build yourself?
07:09:53fyrestormits been awhile
07:10:34S_a_i_n_tFS #11664 and a small change to your targets config and you'll have USB.
07:13:12S_a_i_n_tfrom the looks of the config file...the OF should be handling USB
07:13:19fyrestormit is
07:13:31S_a_i_n_tie. it *should* boot to the OF for USB transfer as I understand it.
07:13:43fyrestormneed to turn it off
07:13:53fyrestormif you plug it in while off it does boot to OF
07:13:56*S_a_i_n_t shrugs
07:15:55S_a_i_n_twell, as I understand it, if you apply FS #1164 and comment #define USB_HANDLED_BY_OF out, and uncomment #define USE_ROCKBOX_USB from rockbox\firmware\export\config\sansafuzev2.h you should have working USB
07:16:06S_a_i_n_tthough, it isn't 100% yet.
07:16:11fyrestormill keep it as is
07:16:14S_a_i_n_tIf it was, it would be in core already.
07:16:20fyrestormim playing it safe
07:16:57S_a_i_n_tbut a Fuzev2 user reports it as 100% working, it appears the problems are from the Clip+ mostly
07:17:10S_a_i_n_tI don't blame you for waiting though.
07:18:58 Join n1s [0] (~n1s@rockbox/developer/n1s)
07:19:30S_a_i_n_tI could compile a build and mail it to you if you like, but I'm not at my build machine for a few hours.
07:19:37fyrestormits ok
07:22:29 Join Buschel [0] (
07:24:34 Join stoffel [0] (
07:48:57 Quit n1s (Quit: Lämnar)
07:50:26 Quit CIA-8 (Ping timeout: 264 seconds)
07:50:28Buschelregarding last nights discussion regarding the mpc bug. the misbehaviour itself was in the code since the very first sv8 port. the bug did not appear before r28198 (Oct, 2nd).
07:51:25Buschelthis new bitstream handling is very fishy... I loved the good old implementation which was simple and robust.
08:01:42 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
08:13:53 Quit stoffel (Remote host closed the connection)
08:23:40 Join antil33t [0] (
08:26:34 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:29:26 Quit elcan (Read error: Connection reset by peer)
08:35:04 Join hebz0rl [0] (
08:35:42 Join elcan [0] (
08:35:47 Join wodz [0] (
08:36:08***Saving seen data "./dancer.seen"
08:36:54 Join ender` [0] (
08:37:35 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
09:01:24 Join petur [0] (d408b802@rockbox/developer/petur)
09:05:56wodzS_a_i_n_t: how work on 2bit graphic is going?
09:07:55pixelmaIIRC, the automatic reboot to the OF on USB plug was disabled on the AMS Sansas because of the freeze issues some were experiencing which you could only get out of by letting the battery drain completely
09:08:02pixelmaS_a_i_n_t: ^
09:08:38wodzpixelma: As you seem to be expert in remote drawings would you be so kind to prepare one for mpio remote?
09:09:39pixelmaI first need to do the Gigabeat one as it is already used in the Gigabeat S manual
09:09:59pixelmawodz: do you have a picture or a scan?
09:10:29wodzpixelma: I can take pictures of mine + there is drawing in mpio manual
09:12:09pixelmaI can have a look but as I said the Gigabeat one is first :)
09:12:39wodzsure, no hurry
09:12:57wodzwhere can I send you 'base materials'?
09:13:35 Join xxcv [0] (
09:13:46 Quit bertrik (Read error: Operation timed out)
09:13:51pixelmaeither upload somewhere or send me a mail, I'll PM you my address
09:14:21 Quit LeonTrollski (Ping timeout: 240 seconds)
09:16:25 Join CIA-7 [0] (~CIA@
09:20:55 Join b0hoon [0] (~quassel@
09:22:14 Quit sasquatch (Ping timeout: 264 seconds)
09:22:43b0hoonpixelma: ping
09:23:05 Join sasquatch [0] (~username@
09:23:20pixelmayou make me wake up quicker today ;)
09:23:39b0hoonsorry :/
09:24:04pixelmano problem :) What's the matter?
09:24:28b0hoonpixelma: could you look at FS #11705 please? Do i have a talent ? :P
09:24:49b0hoonpixelma: Serious - is it good enough?
09:26:10liarS_a_i_n_t: pong
09:35:17pixelmab0hoon: looks quite accurately drawn, I like that. Only the arrows and labels could be a bit more "ordered" but I know that it's not always easy if there are a lot of them. One thing that can cause problems when doing the PDF is transparency which is why I always try to avoid it. You can easily use an opque gradient to achieve the same effect
09:36:28 Quit bluefoxx (Quit: The problems are all in your head.)
09:36:49pixelmaopaque too
09:38:30AlexPgevaerts: Do you think r28429 qualifies for backport?
09:39:41JdGordon|AlexP: really no harm either way, but thats really stretching the bar
09:39:51AlexPyes, I was unsure
09:39:56AlexPBut someone siggested it last night
09:43:42 Join Rob2223 [0] (
09:44:10b0hoonpixelma: thanks a lot i will check the conversion
09:44:43pixelmab0hoon: a minor thing last. It would be good if arrows and text would be on a "text" layer. That can be easily done in the end though. Good work so far :)
09:45:13 Quit wodz (Quit: Leaving)
09:45:27b0hoonpixelma: ok i'll try to do it
09:47:53 Quit Rob2222 (Ping timeout: 276 seconds)
09:59:14 Join LinusN [0] (~linus@rockbox/developer/LinusN)
10:01:37 Quit Gabe_G23 (Ping timeout: 252 seconds)
10:14:00 Join Luca_S [0] (
10:15:06Bagderis a Rockbox competitor on Android?
10:16:05JdGordon|it and winamp... we need to release soon to be in the running!
10:16:25BagderI'd love that
10:16:34 Join pamaury [0] (
10:16:34 Quit pamaury (Changing host)
10:16:34 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:17:03Luca_SJdGordon|: do you have a fuzev2 target? I'd like to talk to you about a problem I have with the WPS before submitting a bug on FS
10:17:12JdGordon|the fact we will be free and no ads probably means we have a good bet at being popular :p
10:17:18Zagorin my opinion we still need more polish to release. otherwise we're not a competitor anyway.
10:22:05 Quit Rob2223 (Read error: Connection reset by peer)
10:22:05 Join Rob2222 [0] (
10:22:06Bagderwell, we also need to release for people to become aware that we exist
10:22:06JdGordon|Luca_S: I do
10:22:06Bagderthat might also bring some new contributors
10:22:06JdGordon|I think we are good enough to do a "beta" release
10:22:06JdGordon|would anyone mind me adding a "report bug" menu item for android?
10:22:06Zagorthat we do, but I think android is a slightly different target from hardware players. when you own a hardware player and would like rockbox on it, you generally don't mind waiting for the port to mature.
10:22:06DBUGEnqueued KICK JdGordon|
10:22:06JdGordon|which would open either email with a template or the browser?
10:22:06BagderZagor: perhaps, but we can also assume that the general public don't know about rockbox, not even potential contributors on android
10:22:06Zagorin the case of android, we are releasing an app saying "this solves the problem better than other apps out there". that's quite different.
10:22:06 Join b0hoon_ [0] (~quassel@
10:22:06DBUGEnqueued KICK Bagder
10:22:06Bagderwe don't have to say that
10:22:06JdGordon|IMO the only real blocker atm is database
10:22:06 Join einhirn [0] (
10:22:06ZagorBagder: how do we avoid saying that?
10:22:06 Quit b0hoon (Read error: Connection reset by peer)
10:22:06Luca_Ssorry, CGI lagging.
10:22:06BagderZagor: by using the word "BETA" all over or similar
10:22:06Luca_Sthe problem I'm experiencing is as follows (default cabbiev2 theme)
10:22:06Luca_Sif I skip three, four songs, the next track indicator shows the currently playing song and stays like that even ten seconds later, when backlight goes off. when I move the wheel to turn on backlight again, the display is updated with the correct next track
10:22:06 Nick b0hoon_ is now known as b0hoon (~quassel@
10:22:30Luca_SI feel quite safe to exclude buffering issues because I tried (with Saint's help) to change the WPS definition to using the simple file name instead of ID3 data, and had the same problem
10:23:22Luca_Sit's easily reproducible by skipping forward, sometimes even a single skip is enough.
10:25:45ZagorBagder: even with a beta tag, I think we are likely to do more damage than good by releasing in the current state. even a beta must feel like something you want to try the next release of.
10:27:28Zagoron the other hand we can of course choose to say "screw first impressions. release early, release often."
10:27:49LloreanJdGordon| mentioned database. What else is problematical on it?
10:28:27Zagorthe default theme icons, the panic screen, missing "back" button support
10:29:04 Join Gabe_G23 [0] (~gabe@bzflag/player/GabrielG)
10:29:28Zagorand we haven't even started on the manual, have we?
10:31:16b0hoondxmmichelle - spamer on the forum
10:31:21LloreanHow hard are these to address in the near term? What's wrong with the icons, what's the "panic screen" (are you just referring to panic errors, and do those happen often on the Android?) and is there something preventing back button support?
10:32:50Zagorthe icons are microscropic in 800x480. the panic screen can't be exited, busy-loops and potentially hangs the phone (did for me). I don't know what needs to be done for back button support.
10:33:23LloreanIcons are just tango right, so grabbing larger versions should be (relatively) trivial. That panic thing sounds really nasty though.
10:34:22 Join DerPapst [0] (~Alexander@
10:36:10***Saving seen data "./dancer.seen"
10:38:36 Join Staphylo [0] (
10:41:49JdGordon|Zagor: back button works now
10:42:00JdGordon|or, how would you expect it to work?
10:42:59JdGordon|Luca_S: put it on FS or wait a few hours when i get home
10:43:18Zagorit always exits back to android, doesn't it? I would expect it to go back in the menus until the top level and then exit to android.
10:44:24CtcpIgnored 2 channel CTCP requests in 0 seconds at the last flood
10:44:24*mystica555_ pipes up for once: indeed, i would love the 'back' button to go 'back' in the app; theres a 'home' button to go to the os
10:44:46 Quit Luca_S (Quit: CGI:IRC (Ping timeout))
10:44:58JdGordon|that was fixed last week. it works like a normal back button on other targets.. we have no way (yet) of knwoing if we are in the main menu so it cant go back->main menu-> os hom
10:45:14JdGordon|but the home button works so its workable for now
10:45:34mystica555_i should try a newer build then
10:45:34 Join wodz [0] (
10:45:50ZagorJdGordon|: ok, good
10:46:03JdGordon|database can be fixed with a patch that is on the tracker to start searching at a user defined dir (/sdcard default)
10:46:35JdGordon|if exit() works then panic can be fixed cleanly with a java ui
10:47:41mystica555_as i've not been entirely active in the forums (there forums, right?) or the mailinglist (i lurk) or here (joined today) i'm not sure if these answers exist, but i do have 2 questions that are related: 1 have there been any bugreports of random audio lagginess/freezing such as when the phone (tmobile g1 htc dream) checks gmail or something? and 2: if you do know of them are fixes in the works/done already? i tried a build from about a month ago
10:47:41mystica555_ and itd just periodically stop outputting for ~100-300ms during playback
10:47:50mystica555_(wow sorry about the long line)
10:48:13mystica555_*there are
10:48:54 Join Strife89TX [0] (~cstrife89@
10:49:08wodzmystica555_: there is known bug in htc version of android which doesn't allow to grab much CPU power when application runs in background
10:49:47mystica555_this is foreground; i also am running a kernel with BFS on it which no other audio app seems to have issues with.. even those that seem to use non-hardware codecs
10:50:09mystica555_or is it considered backgrounded if the screen is off?
10:50:41mystica555_i'd go grab my phone now, but its currently on the end of a short usb cable to my router at the moment...
10:50:59wodzmystica555_: dunno - I don't have android phone. Ask kugel I would suggest.
10:51:47JdGordon|try the recent build, alot has changed in a month
10:52:38mystica555_will do the next time im out and about. until then, usb cord stays plugged in so i have internet :)
10:59:11Zagorhmm, latest build refuses to install
11:00:05JdGordon|whats the error?
11:02:28Zagorno error. android just says "Application not installed" when I try to install it.
11:02:54JdGordon|was it already installed? uninstall it first
11:03:36Zagorit does a replace. or tries to, anyway.
11:04:29Zagoryup, works if uninstalled first
11:04:54mystica555_lots of android apps seem to need that..
11:05:10ZagorI've never had to do it before.
11:05:41 Join Luca_S [0] (
11:06:28ZagorJdGordon|: the new loading dialog is nice.
11:06:50Luca_Ssorry, CGI disconnected without me noticing. I'll open a FS bug ASAP. To that regard, is it normal that it doesn't show the Fuze player type? and that it allows only 3.6 release?
11:06:57ZagorI think we need a bigger default font for 480x800
11:08:34Zagorthe wps volume icon doesn't update
11:09:32Zagorbut overall, a lot has happened in a week. this is more ready than I thought. good job!
11:13:23JdGordon|checkout the yes/no screen and the keyboard :)
11:14:03mystica555_regarding the screen-sizes for the android port, some devices are now coming on the market with 240x400 wqvga screens. would the 320x240 res app work with say, blank space ?
11:15:42JdGordon|yes, and it looks crap... we need to do at least run time screen size checking
11:15:59JdGordon|it should be simple if mcueulernes(?) patch works
11:16:40 Join Staphylo` [0] (
11:16:52 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
11:18:35LambdaCalculus37Zagor: Can you upload the GoGear HDD1630 and HDD6330 bootloaders to the site?
11:18:35mystica555_good friend of mine whos been rockboxing since his ipod 4g i think just got the new samsung intercept wqvga phone..thus i bring it up
11:19:37 Quit Staphylo (Ping timeout: 250 seconds)
11:21:14 Quit liar (Ping timeout: 255 seconds)
11:21:53ZagorLambdaCalculus37: I have
11:22:16Zagorthey just haven't been mirrored yet
11:22:27 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:23:01 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:25:19 Join Rob2223 [0] (
11:25:30LambdaCalculus37Zagor: Ahh, okay. Thanks. :)
11:28:12 Part LinusN
11:28:37 Join liar [0] (
11:28:53 Quit Rob2222 (Ping timeout: 255 seconds)
11:30:54ZagorJdGordon|: back never exits. is that intentional?
11:31:05JdGordon|known limitation
11:31:17JdGordon|the java doesnt know if c is in the main menu
11:31:21pixelmamystica555_: there should be some things for 240x400 already though, if I remember the Ondas have the same resolution
11:31:58JdGordon|that should be very fixable though
11:32:37Zagorsometimes the tap doesn't "take". it clears the selection from the last menu line, but doesn't draw a new selection nor go into the tapped item.
11:32:39JdGordon|but then if you are in the wps and press a touch region to go to the menu, should back go to OS home or wps?
11:33:17kugelZagor: that error comes when the signature of the apk chaned
11:33:27kugelyou need to uninstall
11:33:32Zagorkugel: aha
11:34:04kugeldid you download my apk from the wiki? the latest ones are compiled from my desktop (because my laptop is in RMA) and have a different signature
11:34:14Zagoryes I did
11:34:27 Quit Luca_S (Quit: CGI:IRC)
11:35:26wodzkuge: can you look at FS #11712 again?
11:35:44 Quit preglow (Ping timeout: 240 seconds)
11:35:53kugelmystica555_: you can compile an apk for 240x400 without problems
11:36:05 Quit aevin (Ping timeout: 250 seconds)
11:36:06mystica555_kugel: ok. i've not gone into the whole development process for ityet
11:36:13Zagoranother minor thing: long press gives context menu, but only after you release.
11:36:24kugelthere's even a cabbiev2 port, but it's older (because made for the ondas) and different from the 320x480 and 480x800 ports
11:36:38Zagor(in file tree)
11:36:47JdGordon|Zagor: yeah, thats been driving me mad for a while... its the same behaviour in the os on long home though
11:36:56mystica555_ive just been using current builds or releases, whatever i figure is 'new enough' on my iriver and my G1. friend of mine has the odd-device-out
11:37:18ZagorJdGordon|: not for me. long home gives me "running tasks" without releasing the button
11:37:34 Join preglow [0] (
11:37:36mystica555_how usable would the cabbiev2 240x400 be in comparison go the more standard sizes?
11:37:37kugelZagor: that should be a relatively simple fix
11:37:43 Join aevin [0] (eivindsy@unaffiliated/aevin)
11:37:44kugelmystica555_: no idea :)
11:37:56mystica555_ah hehe
11:38:38Zagorkugel: have you seen my "lost tap" issue?
11:40:05JdGordon|Zagor: ah! so it does :p
11:41:28kugelZagor: you probably exceeded the scroll threshold with the touch, meaning it thinks you want to scroll (and thus hides the selection bar)
11:41:59kugelthe threshold is ok for me, but apparently too small for most other people. I think we'll increase it or make it configurable (or both)
11:42:16Zagorok, that treshold needs to be increased then. I get it all the time.
11:42:23Zagorhow many pixels is it? 4?
11:42:29b0hoonLambdaCalculus37: i know about the memory detection but is it possible to disasm your hdd6330 to check it for sure?
11:42:38Zagorthat's waaay too small
11:43:00kugelit's likely worse on 480x800 displays with higher dpi
11:43:06Zagordoesn't android have a standard value for this?
11:43:23kugelI think font size/2 (with a fixed minimum) would work better
11:43:28b0hoonLambdaCalculus37: btw: it was you who wrote on the wiki that it has K4S51163PF ;)
11:43:36kugelI don't know about standard value
11:43:42ZagorI don't think such basic UI behaviour should change with font size
11:44:11JdGordon|it shoud come from the touch pressure
11:44:39CIA-7New commit by 03wodz (r28443): MPIO HD200: rename button defines to adhere how they are labeled on the device.
11:44:40kugelthe font size takes dpi better into account than fixed pixel, because you usually use a much larger font on higher dpi screens
11:44:43mystica555_i don't think android exposes 'tap width' or 'pressure' .. just an exact point or 2.
11:45:17Zagorkugel: yes, but I don't want the scroll treshold to change just because I choose a different theme
11:45:18mystica555_this, because it can be either capacitive or resistive/stylus operated
11:45:35JdGordon|there is a touchscreen debug thing in "dev tools" which shows it, so i assume it does
11:45:45mystica555_must be newer than the hardware ive got then
11:45:47kugelfor android it also seems to be roughly half of the row height
11:46:17Zagorkugel: how many font sizes did you try that with?
11:46:35CIA-7r28443 build result: 15 errors, 0 warnings (wodz committed)
11:46:52kugelnow 3
11:47:31kugelwell, 2 different row heights, one of it with 2 different font sizes
11:47:34wodzI missed bootloader :/
11:49:24kugelI think changing the theme doesn't mean the font size changed by a noticeable amount. commonly all themes to use have the font size in the same magnitude (if not even the same)
11:49:43LambdaCalculus37b0hoon: I didn't write that in the wiki. I think it was either lowlight or toffe82 that may have written that.
11:49:49Zagorkugel: font size can change a lot with different themes
11:49:56kugelso the scroll threshold wouldn't change much, if at all. not noticeably anyway
11:50:13kugelI'm looking at the common case, not edge cases
11:50:33JdGordon|excellant, basing design descisions on false premises always works well!
11:50:38kugelyou wouldn't chane from 35-adobe to sysfont. that makes little sense
11:50:54Zagorkugel: no but you can surely change from a 40 to a 20-point font
11:51:01AlexPyou can never tell what people will do :)
11:51:06CIA-7New commit by 03wodz (r28444): MPIO HD200: fix red (I missed bootloader and simulator when renaming buttons)
11:51:07JdGordon|thats assuming you even know what font the new theme uses
11:51:13Zagorthat shouldn't cause a drastic change in basic touch behaviour
11:51:15kugelso, what do you propose?
11:51:29AlexPcan you get screen dpi?
11:51:57kugelnot the exact value, only high, medium or low
11:52:15AlexPis it good enough to base it on that?
11:52:17pixelmaactual pixel size can make a huge difference there
11:52:28AlexPpixelma: Right, but you also have size
11:52:47CIA-7r28444 build result: All green
11:53:00JdGordon|add a setting with a default of 15 or something
11:53:23Bagderor make the default depend on the high, medium or low values
11:53:26kugelZagor: from 40 to 20 is also unrelealistic if you ask me. from 40 to 30 is more releastic and for this case the scroll thresshold wouldn't change much
11:54:07wodzBut the change of the font may be made by accident.
11:54:10Zagorwebkit uses 25 pixels
11:54:12AlexPwr have themes for other targets were the range in font sizes is more than two
11:54:34pixelmakugel: I doubt your "unrealistic"
11:54:41Bagderme too
11:54:56Bagdersome people will use very small fonts, some will use very large ones
11:55:00Bagderand some will test both ends
11:55:14kugelthat's true
11:55:31pixelmaand additional to that you run on a lot of different displays which may have the same resolution but not the same size
11:55:33kugelbut the same person usually doesn't change the font size drastically
11:55:44Bagderunless they're trying out stuff
11:55:52Bagderwhich many do before they've decided
11:56:11Zagorthere's a TOUCH_SLOP constant in android
11:56:11 Quit ender` (Read error: Connection reset by peer)
11:56:17JdGordon|you have no way of knowing what users do... bloody stop saying things like you do... unless of course you've actually done a proper user stufy
11:56:23AlexPI just think it is bad in principle to have one setting change a seemingly unreleated one
11:56:34 Join ender` [0] (
11:57:15 Quit evilnick (Ping timeout: 276 seconds)
11:58:11CIA-7New commit by 03rmenes (r28445): Commit a few more GoGear HDD6330 keymaps.
11:58:46Zagorthe function to use is getScaledTouchSlop(). "Distance a touch can wander before we think the user is scrolling"
11:59:01 Join LinusN [0] (~linus@rockbox/developer/LinusN)
11:59:30 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
11:59:52CIA-7r28445 build result: All green
12:00:45kugelZagor: interesting, someone should look into it :)
12:00:54 Quit LambdaCalculus37 (Quit: Fwump)
12:01:44 Quit Staphylo` (Quit: Bye les gens =))
12:02:49kugelhm, it appears the exact dpi can be retrieved
12:02:55 Part LinusN
12:04:03AlexPcool, so some sort of dpi + height formula :)
12:04:54kugelandroid does dpi * TOUCH_SLOP * 0.5
12:05:23kugelhm not that can't be right
12:05:44Zagorkugel: anything wrong with just using the function provided for this purpose? :)
12:06:40Bagderkugel: it depends what TOUCH_SLOP is... ?
12:06:56kugelI think it would be more useful to have a get_dpi() from the target tree and do our own calculation based on that. the android function only works on android
12:07:24Zagorkugel: this is an android port. it is supposed to use the android environment.
12:07:30Bagderbtw, how high dpi do a typical target have?
12:08:02kugelwell, the scrolling works on all touchscreens, so we should get the threshold right on all of them
12:08:04Zagorit is supposed to try and blend in. that's another way this port is different from others. normally we do it our own way and screw the OF way. that doesn't work here.
12:08:29Zagorkugel: sure, but we can still use the android-supplied value on android.
12:08:42AlexPI think I'd expect scrolling to work the same way everywhere
12:08:43kugelBagder: from ~120 to 200+
12:08:48Bagderok, thanks
12:09:03Bagdermy 24" at home says 96
12:09:14AlexP(everywhere being between apps on android)
12:10:11kugelwe can of course base the calculation off the androids calculation (so that the outcome is the same) but I think the other touchscreens should also get this reasonable thresshold
12:10:40kugelthe rockbox ui is the same on all of them
12:10:49AlexPwhy not just use the android function on android, and do our own thing on other targets?
12:11:07Zagorno, we use the android value on android. and we do a calculation on other targets that don't have such data themselves.
12:11:43Zagorhaving the android app work like other android apps is *much* more important than having the exact same treshold across different touchscreen ports
12:12:14mystica555_Zagor: this is true: i've had a few apps that just refuse to scroll 'right' ; opera mini being the biggest offender
12:12:30JdGordon|especially when the threshold is also going to be different for resistive vs capacative screens
12:12:56kugelI thought we consider the same experience across all our ports as very important (we've put off a lot users with keymap changes for that reason)
12:13:17JdGordon|for keymaps - yes
12:13:21Zagorkugel: I just told you: raaa is different. we're guests in another system. normally we are the masters.
12:13:27kugeland again, our calculation can be the same as androids one
12:13:42Zagorno it can't. android's calculation might change in the next version.
12:14:46Bagderit might even vary between versions while we have a single app
12:17:14 Quit xxcv (Ping timeout: 245 seconds)
12:18:47AlexPI honestly don't understand why using the android value is bad?
12:19:10AlexPWe can make the other targets use a calculation to give approximately the same thing for consistency
12:20:04kugelit's not bad
12:20:49kugeland the value of android most likely would've been taken anyway, either with our calculation or by using the api
12:24:25kugelit doesn't seem possible. the android function isn't based on the exact dpi
12:24:41 Join Staphylo [0] (
12:24:56pixelmaAlexP: was there something else than the "ohter" you wanted fixed in lamp.c?
12:25:07AlexPnot that I remember
12:25:12AlexPjust that minor typo
12:25:23kugelI'll try to come up with something on the weekend
12:27:21 Join LinusN [0] (~linus@rockbox/developer/LinusN)
12:27:37pixelmaok, did that and rearranged a bit but didn't look for more typos
12:27:58AlexPI didn't look hard, I just happened to notice that one
12:29:35 Join MethoS- [0] (~clemens@
12:29:36pixelmaerr... meant lamp.tex of course
12:33:42 Join Alchimysta [0] (
12:33:43S_a_i_n_tJdGordon|: Did you test %Tp ?
12:34:01JdGordon|probably not
12:34:11S_a_i_n_tseems to be broken.
12:35:01AlchimystaHow often the CPU has the Sansa Fuze?
12:35:27S_a_i_n_tthat doesn't make any sense, could you try again?
12:36:11AlchimystaThe google translator is not good, I rephrase the question
12:36:12***Saving seen data "./dancer.seen"
12:36:36AlchimystaHow many mhz the cpu has the Fuze?
12:36:46S_a_i_n_task google
12:37:41soapgoogle will not easily know what frequency Rockbox runs the fuze CPU at.
12:38:15S_a_i_n_tthat wasn't the question though, as far as I understood it.
12:38:17 Join teru [0] (
12:38:26AlchimystaI have tried .. but to no avail .. you can see from the Rockbox mhz cpu?
12:38:34wodz <- why ACTION_FM_RECORD and ACTION_FM_MODE doesn't work at all?
12:38:58wodzAlchimysta: the freq depends on how you configure PLL
12:39:34AlchimystaBut the standard rate?
12:40:04wodzthere is not something like standard - you can set any freq within the limits
12:41:01AlchimystaHow it works?
12:41:07pixelmawodz: swcodec targets don't record FM from the FM sreen (but from the recording screen with source set to "FM") only hwcodec does so I wouldn't expect ACTION_FM_RECORD to work
12:41:08kugelwodz: is DEBUG_CANCEL used anywhere except in the target tree debug files? If not I'd move the defines into the C files
12:41:31pixelmawodz: assuming this is for an swcodec target
12:41:38wodzkugel: it is in PCF debug screen I think
12:41:54 Quit soap (Quit: Leaving)
12:42:13wodzpixelma: I see
12:42:18Alchimystawodz: "you can set any freq within the limits" How it do?
12:42:35kugelwhat did you else change beside the get_action()=
12:42:50kugelI guess your patch is fine (assuming it works)
12:43:43wodzkugel: I synced with svn dbg_ports() for player
12:43:57JdGordon|wodz: the fm screen code is still a mess.. you need to look at the #defines at the top of radio.c to make sure the code for that action is compiled in
12:44:16pixelmawodz: can't tell you about the FM_MODE from that paste
12:44:18wodzAlchimysta: Please use correct english because it is VERY hard to follow
12:44:21mystica555_is the question perhaps, "when the player is operating normally, what processor frequencies are used"? my sansa e200 was 30/80-boost if memory serves.
12:44:59wodze200 is PP he asked about Fuze
12:45:14mystica555_but i am providing a reference for what he may be actually asking
12:45:22mystica555_my h320 i think is 40/120
12:45:33mystica555_so perhaps he wants those mhz values
12:45:45wodz124 actually
12:45:52mystica555_versus 'anything you want to set it to', specifically 'what does rockbox set it to during normal runtime'
12:46:09wodzbeast is over 500 when boosted
12:46:43S_a_i_n_tJdGordon|: the odd thing is, the WPS has no problem with %Tp...but the .sbs doesn't seem to like it at all.
12:46:56JdGordon|FS it
12:47:13pixelmamystica555_: there are two hardware variants of the e200s and CPU is a main difference, what you said is for the e200v1 which is portalplayer
12:47:28mystica555_pixelma: yes, i understand its the portalplayer vs AMS cpu
12:47:39mystica555_i am clarifying his question so someone can give the values for the Fuze cpu
12:48:13mystica555_which i haven't owned yet
12:48:19wodzJdGordon|: what a mess (radio.c)
12:48:19AlchimystaWods: I'm italian and Use the google translate and is not good "that's why I speak bad English"
12:48:43*mystica555_ seems to be rather good at decyphering google translate stuff...considering i read a lot of things processed via it...
12:49:07JdGordon|wodz: Tell me about it :p it and recording screens are really the last of the origional gui code from the pre-ipod days
12:49:13JdGordon|its had almost no attention
12:49:17AlchimystaDo you know any good translator?
12:50:02wodzsure - lots of it costs something like 10euro per hour
12:50:33mystica555_so does anyone actually know what frequencies standard/boost are on the Fuze? or is this just going to go into a 'how to translate yourself' discussion
12:50:51 Join TheLemonMan [0] (~lemonboy@
12:53:16AlchimystaDoes anyone speak Italian?
12:54:55TheLemonMani do
12:56:05wodzclock derivation for as SOC looks quite complicated
12:57:45AlchimystaI tried to compile the nes emulator but without success, you could briefly explain it to me?
12:58:16TheLemonManhows this related with italian ?
12:59:02 Join evilnick [0] (
12:59:06AlchimystaI did not understand :(
12:59:12wodzok sansa fuzev2 uses 40MHz as default
13:01:05pixelmaTheLemonMan: I think he needs someone to explain the process of patching and compiling in Italian (if you do, please take it to a PM :) )
13:01:17wodzand 248MHz when boosted if I understand sources correctly
13:01:32TheLemonManooh got it
13:02:25Alchimysta<pixelma> : Yes :D
13:03:03TheLemonManAlchimysta, i PM'd you
13:05:00rasherMan, when someone comes up with an easily reproducible codec crash, you guys can be quick! (FS #11710)
13:07:38JdGordon|anything easily reproducable is usually killed pretty quickly
13:09:17pixelmaexcept database hang with no audio files
13:12:11JdGordon|thats a special case of noone wanting to go near tagcache.c
13:12:31JdGordon|and a mostly ridiculous repro
13:13:01wodzJdGordon|: defines at top of radio.c did the trick :-)
13:15:41CIA-7New commit by 03wodz (r28446): MPIO HD200: adjust FM keymap
13:17:30CIA-7r28446 build result: All green
13:21:36 Join Ramsey[LC] [0] (~RamseyLC]
13:31:07 Quit DerPapst (Read error: Connection reset by peer)
13:31:21 Part LinusN
13:32:59 Quit Strife89TX (Quit: Leaving)
13:37:26 Join DerPapst [0] (~Alexander@
13:37:53wodzthis arm eabi compiler is painfully slow :/
13:42:38 Join JdGord [0] (~jd@
13:44:56 Quit dfkt (Ping timeout: 252 seconds)
13:45:16S_a_i_n_tanyone here have an iPod video I can test something on?
13:45:45S_a_i_n_tlack of debug output in the sim is driving me FUCKING MENTAL!!!
13:52:41 Quit DerPapst (Read error: Connection reset by peer)
13:54:00 Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
13:54:30LambdaCalculus37S_a_i_n_t: Just read the logs. I have an iPod video, so let me know what you want to try.
13:55:00 Quit MethoS- (Remote host closed the connection)
13:55:30S_a_i_n_taha...awesome. I'll upload a theme, and if you could update to a current build and test it would be great.
13:55:47S_a_i_n_tit would be *really* great if you had FM hardware too.
13:55:48LambdaCalculus37S_a_i_n_t: Sure thing. Let me know when you're ready.
13:56:47 Join DerPapst [0] (~Alexander@
13:57:28 Join kazaik [0] (
13:58:08 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
13:58:08LambdaCalculus37S_a_i_n_t: I don't have FM hardware for my iPod.
13:58:25 Quit JdGord (Quit: Bye)
13:59:41kugelJdGordon: does the skin update code yield?
14:00:32kugelI'm wondering how the next track info event can be missed (FS #11713)
14:01:09S_a_i_n_tLambdaCalculus37: Ok, I figured you wouldn't have the FM hardware...that's no big loss, no one seems to. here goes though:
14:01:56 Join DerPapst1 [0] (~Alexander@
14:02:21S_a_i_n_tLambdaCalculus37: I'm mainly concerned about the menu screen .sbs
14:02:28S_a_i_n_tif it doesn't look like:
14:02:36S_a_i_n_tI'll be pissed ;)
14:02:57S_a_i_n_t(sorry...I don't have Video screenshots of how it should look)
14:03:21 Quit DerPapst (Ping timeout: 240 seconds)
14:03:56wodzok I tested FS #11712 on MPIO HD200 (CF) and mini 1G (PP). It would be awesome if someone could test this on other PP, CF and SH targets
14:05:39TheSevenwodz: what is that code doing? what kind of debugging functionality does it offer?
14:05:43 Join komputes [0] (~komputes@ubuntu/member/komputes)
14:06:05LambdaCalculus37S_a_i_n_t: I hate to be a pest, but can you repackage the theme as a zip file? I can't open .7z files on my work PC.
14:06:29S_a_i_n_tLambdaCalculus37: Sure, sorry about that.
14:07:28wodzTheSeven: It doesn't offer any new functionality - this simply moves dbg_ports() function to target tree for PP, SH and CF (for other targets it is done this way)
14:08:05TheSevenah, dbg_ports() was that thing showing GPIO info if boot fails in an early stage?
14:08:11S_a_i_n_tLambdaCalculus37: Here you go, take 2 ;)
14:08:26LambdaCalculus37wodz: Can you roll me a build for the iPod video and one for the HDD6330 with FS #11712 applied? I can test it on those two targets.
14:09:03wodzTheSeven: no, this is gpio debug info under system->debug->view I/O
14:09:21wodzLambdaCalculus37: I can do this later today
14:09:41TheSevenok. IIRC that's also called on boot failures on nano2g...
14:09:59TheSevenor at least it was during early development
14:10:17LambdaCalculus37wodz: Okay.
14:12:39 Join DSStrife89 [0] (~Strife89@
14:12:39*TheSeven wonders what's the best way to do storage driver profiling
14:14:17LambdaCalculus37S_a_i_n_t: The theme works in the WPS, but it needs some more work to make it look nice. The menu butts into the status bar on the very top.
14:17:29S_a_i_n_tI have NO bloody idea why it's doing that.
14:17:51 Quit DSStrife89 (Read error: Connection reset by peer)
14:17:53 Join Strife89DS [0] (~Strife89@
14:17:57S_a_i_n_tthemeeditor displays it fine.
14:19:11AlchimystaIs there any software (windows) to create themes for Rockbox?
14:20:58LambdaCalculus37S_a_i_n_t: How does it look in the UI sim?
14:21:14S_a_i_n_tthe same...but I was hoping the sim was lying. :/
14:21:26S_a_i_n_tit supposed to look just like the Nano pics I sent you.
14:21:31S_a_i_n_t(in the link)
14:21:42S_a_i_n_tBut, it has broken itself somehow.
14:21:47LambdaCalculus37S_a_i_n_t: Can you pastebin the WPS code for the iPod video's version? Maybe someone who knows the WPS code well can help you out.
14:21:48 Quit Amec (Read error: Connection reset by peer)
14:21:48 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
14:22:19S_a_i_n_tLambdaCalculus37: "someone who knows WPS code well" is supposed to be me ;)
14:22:26 Join Amec [0] (
14:22:46 Join n1s [0] (
14:22:46 Quit n1s (Changing host)
14:22:46 Join n1s [0] (~n1s@rockbox/developer/n1s)
14:23:00wodzn1s: ping
14:24:01wodzn1s: do you still have problems to track down bug in CF profile build with new gcc?
14:25:33kugelI wish we can have gcc 4.4.x for all our targets, the we can use #pragma gcc optimize to select optimization in codecs instead of makefile magic (which fails for android)
14:25:58n1swodz: yes, well, i got it to boot by making the profile_thread_stopped call in switch_thread conditional on thread being nonull but i don't think thread being null makes sense, also it seems to hang when running test_codec
14:26:31n1skugel: would also be nicer for selecting different optimizations for the various arm cores
14:26:32wodzn1s: I have CF target with BDM
14:27:02wodzthis may help to see what is going on
14:27:07n1swodz: nice, are you willing to look into it?
14:27:22 Join krazykit` [0] (
14:27:28wodzn1s: I can help but I am unfamiliar with profiling stuff
14:27:41*TheSeven wonders what to do with his post-mortem memory dumper patch
14:27:55S_a_i_n_tWho feels like giving me a lesson in skin coding?
14:28:04n1swodz: it's fairly simple to use when it works :)
14:28:08wodzTheSeven: commit if works
14:28:12S_a_i_n_tShow me what I've done wrong...please, like...serioiusly...please.
14:28:23wodzn1s: I know but I don't know the internals
14:28:41TheSevenS_a_i_n_t: I can only recognize a bunch of %jkd%kj%k%kjH%KJ5 gibberish in that paste :)
14:28:42n1swodz: to get the boot hang i just enabled profiling support in the config, no modified code had a chance to run
14:29:18 Quit Strife89DS (Quit: Probably no point in letting the DS run for now.)
14:29:46S_a_i_n_tanyone familiar with skin code wanna have a looksee at my pastebin?
14:29:53n1sthe only thing that does (or should do) is including the two function calls in switch_thread and including profile.c in the build
14:30:49 Quit krazykit (Ping timeout: 265 seconds)
14:31:25wodzn1s: I have to go now, I can look at this tomorrow I think
14:31:30 Join ajb [0] (
14:31:38 Join anewuser [0] (anewuser@unaffiliated/anewuser)
14:31:39n1sbut even if i commented out the whole body of the profile_thread_stopped function the call still caused a hang so it seems thread is null for some reason but i don't understand why that doesn't hang without profiling
14:31:45n1swodz: that'd be great :)
14:31:58 Quit wodz (Quit: Leaving)
14:35:13 Join Strife89TX [0] (~cstrife89@
14:35:59 Quit Amec (Read error: Connection reset by peer)
14:36:14***Saving seen data "./dancer.seen"
14:38:59S_a_i_n_tOk, this is seems to be that the Video is broken specifically.
14:39:08S_a_i_n_tas the same code works on the Nano
14:39:26Strife89TXS_a_i_n_t: ... You *did* make clean first, right?
14:39:27S_a_i_n_tSomething recently broke the Video..but I have no idea what yet.
14:39:40LambdaCalculus37Instructions for installing Rockbox on the HDD1630 and HDD6330 are finally completed!
14:39:43S_a_i_n_tStrife89TX: what're you on about?
14:40:16TheSevenS_a_i_n_t: did you test my new hwkeyclick patch?
14:40:19TheSevendoes it work?
14:40:36 Join T44 [0] (
14:40:37Strife89TXS_a_i_n_t: Clean your build dir and see if that changes anything.
14:40:47S_a_i_n_tI haven't had time just soon as I sat down I've been trying to figure out why the Video is broken.
14:41:14S_a_i_n_tStrife89TX: I can't imagine how it would...especially as I DL'd the sim.
14:41:14 Join Amec [0] (
14:41:29Strife89TXS_a_i_n_t: Ah.
14:44:20 Quit Topy (Ping timeout: 252 seconds)
14:45:21S_a_i_n_tok....what. the. FUCK!?!?!
14:45:33S_a_i_n_t%Rp and %tp are broen on the Video
14:45:43S_a_i_n_tbut...why just on the Video?
14:45:50Torneare those radio related?
14:45:51S_a_i_n_tworks fine on Nano.
14:46:01S_a_i_n_tRadio and Recording
14:46:07Tornei can imagine nobody ever tries the ipod radio stuff
14:46:21S_a_i_n_twhy broken on just one target though?
14:46:34S_a_i_n_tthat makes *no* sense to me.
14:46:56S_a_i_n_t%Rp == "has recording?" and %tp == "has FM?"
14:47:02*Torne has a look
14:47:43S_a_i_n_tThanks...I was trying to code my way out of it with theme code, driving myself mental...but I've come to the conclusion that my code is fine and the build is fucked.
14:47:58Tornei'm assuming you've tested this on a stock current build?
14:48:12S_a_i_n_tI've tested it on stock sims
14:48:12Torneand what do you mean by "broken"? theme fails to load?
14:48:38S_a_i_n_tjust the elements with those two tags
14:49:15S_a_i_n_tI tested it by commenting everything out, then uncommenting one by one.
14:49:23S_a_i_n_tit's just %Rp and %tp
14:49:46S_a_i_n_tthe same exact code works in the Nano2G sim though, and the Nano1G sim too
14:49:51S_a_i_n_tjust not in the video sim.
14:52:21 Join esperegu [0] (~quassel@
14:52:54S_a_i_n_tTorne: Can you read skin code?
14:52:59Tornecan't see any way %Rp could be broken
14:53:09Torneit's solely conditional on HAVE_RECORDING
14:53:19anewuserhmmm, what's the ircplayer plugin on this new rockbox release
14:53:19Torneand no
14:53:23anewusercongrats! btw.
14:53:29S_a_i_n_tyou can see my changes, that's the only thing that makes it work.
14:53:30n1shmm, thread is not null, since my panic call never triggers
14:54:02S_a_i_n_tthe bloody odd thing is that it works fine in the Nano sim.
14:54:46Torneyeah, dunno
14:55:03Torneunsurprisingly the feature tag code is trivial
14:55:04S_a_i_n_tJdGordon|: JdGordon: around?
14:55:36S_a_i_n_tany idea how %Rp and %tp could be broken...but just on the Video (to my knowledge)?
14:57:51AlchimystaHow do I play midi files on Rockbox?
14:58:30S_a_i_n_tLambdaCalculus37: are you able to try another theme on your iPod please?
14:58:34TorneAlchimysta: read the manual
14:58:48S_a_i_n_tsame theme, just a test to make sure it's broken on device as well as on the sim.
14:59:26AlchimystaHow do I install the musical instruments?
14:59:49LambdaCalculus37S_a_i_n_t: Sure. Linky, please.
15:00:00LambdaCalculus37Alchimysta: Read the manual.
15:00:37AlchimystaOk, do not get mad ... -.- "
15:01:02 Join evilnick_ [0] (
15:02:13LambdaCalculus37Alchimysta: Everything that can be explained is explained in the manual. That's what it's there for.
15:02:37 Quit evilnick (Ping timeout: 240 seconds)
15:02:38AlchimystaI can not find anything in the manual regarding the midi ..
15:03:22S_a_i_n_tAlchimysta: No one is getting mad, just frustrated that you ask a lot of questions you could easily answer yourself by reading the manual.
15:03:40S_a_i_n_tand thanks for testing btw LambdaCalculus37
15:03:46S_a_i_n_tit's a big help.
15:04:21LambdaCalculus37S_a_i_n_t: No problem. :)
15:05:50 Join qwd [0] (~x@unaffiliated/qwd)
15:06:10LambdaCalculus37S_a_i_n_t: MUCH better! MUCH, MUCH better!
15:06:18AlexPanewuser: It is lrc (as in lowercase L), and the manual has a description of it
15:06:56S_a_i_n_tLambdaCalculus37: thanks...I'd rather it was still broken though ;)
15:07:00qwdHey, I'm going to install Rockbox on my Ipod 5.5 80gb white but I want to remove everything that's on it first and put my new music collection on it. How do I do this?
15:07:09S_a_i_n_tAs, this means there's some BLOODY weird bug in the Video code ;p
15:08:01S_a_i_n_tLambdaCalculus37: Now you can impress your friends, fool you co-workers, amaze your family with your "hacked" Apple iPod FW ;)
15:08:12LambdaCalculus37S_a_i_n_t: If i had the FM radio hardware for the iPod video, I would gladly help test out the FM radio skin. Alas, as I can't get the hardware in stores anymore, and I don't have a disposable income to buy the hardware online, I can't help with that.
15:08:23S_a_i_n_tqwd: restore it with iTunes
15:08:45Alchimystavoice files exist in Italian?
15:08:53S_a_i_n_tLambdaCalculus37: No problem, you've been a great help anyway.
15:09:11LambdaCalculus37S_a_i_n_t: And my friends and family know I'm a Rockbox developer already. :P
15:09:14qwdS_a_i_n_t: doh, sorry, should have mentioned I'm on Debian :)
15:09:24AlexPAlchimysta: You need to make them yourself, Rockbox Utility can help
15:11:09qwdCan I just reformat the hard drive or will that break something?
15:11:41S_a_i_n_tif you're keen, you can restore the iPod manually.
15:12:03AlexPAlchimysta: And for midi, search the pdf manual for "midi"
15:12:06AlexPIt is there
15:12:36*AlexP is extra kind today
15:13:03 Join wodz [0] (
15:13:05AlchimystaI had already found it, thanks
15:13:10S_a_i_n_tqwd: but, if you have acces to a windows machine...just restore with iTunes
15:13:18qwdS_a_i_n_t: Thanks!
15:13:32S_a_i_n_tpretty sure iTunes will run in Wine though
15:14:03anewuserAlexP: oh, hadn't noticed when checking the manual, lrc ok
15:14:08anewuserdamn 1 vs l
15:14:14wodzn1s: How did you compile new toolchain to not interfere with current one?
15:14:40S_a_i_n_tqwd: you probably know this, but be very, very, very carefull with dd
15:15:20n1si used −−program-suffix for the config to add a -4.4 suffix, if you do that you also need a −−program-prefix as otherwise the prefix goes missing
15:15:45n1si am not actually sure that i'm using the newer toolchain though...
15:15:54AlchimystaExist other firmware opensource for mp4?
15:15:59qwdS_a_i_n_t: hmm, maybe I'll just manually remove the songs from gtkpod and delete the folders containing miscellaneous files I've saved there. Sounds safer :) Thanks for the help
15:16:16S_a_i_n_tNo problem.
15:16:29n1snot toolchain, not sure im using new binutils
15:16:45n1si set the new suffixed versions in my makefile
15:18:04wodzn1s: So you just compiled new gcc or whole toolchain?
15:18:13kugelwodz: you could simply use a different −−prefix and set $PATH accordingly when executing configure
15:19:03n1swodz: i compiled the whole toolchain, but i'm not sure which binutils one our buildsystem is actually using
15:19:25kugelyou didnt upgrade binutils as well?
15:20:08kugelwe use 2.20.1 for ARM, IIUC using a recent with an old binutils is asking for trouble
15:20:11n1s" i compiled the whole toolchain, but i'm not sure which binutils one our buildsystem is actually using"
15:20:50kugelconfigure prints the toolchain info it's using for compilatoin
15:20:54n1ssince gcc itself calls as and ld i'ts kind of hard to verify
15:21:26n1syes, it's printed correctly and set in the Makefile but i don't know how to *verify*
15:21:37kugelgcc calls the tools that are in the prefix (it makes some kind of sysroot for this)
15:22:03n1sthe prefix is /usr/local for all toolchains now
15:22:12Tornegcc calls whatever its specs say to call
15:23:10kugelI mean prefix as in /usr/local/m68k-elf/
15:24:54kugelwhen it calls as it calls <prefix>/<program prefix>/<program prefix>/bin/as IIUC
15:25:05n1sah, seems the new binutils install overwrote those, nice of it
15:25:10 Quit petur (Quit: reboot)
15:25:20 Quit Alchimysta (Quit: CGI:IRC (Ping timeout))
15:25:54kugelm68k-elf-* are essentially wrappers
15:26:34n1si have both versions in /usr/local/bin though
15:27:04 Quit evilnick_ (Ping timeout: 240 seconds)
15:27:07kugelthen I suspect you messed up your toolchains :)
15:27:39n1swell, how do you suggest i do it with this new IMO stupid scheme of dumping everything in /usr/local/bin ?
15:28:03kugeluse RBDEV_PREFIX to install into somewhere else
15:28:55n1si'm not talking about my setup here, i want to get this committed sometime and i thought it was changed to be this way for a reason
15:29:20 Join Alchimysta [0] (
15:29:21 Part qwd
15:30:05S_a_i_n_tkugel: You know skin code...any idea how %Rp (is target able to record?) and %tp (does target have FM?) could be broken on the iPod Video, but not on the Nano1 or 2Gs?
15:30:10kugelthe binaries in /usr/local/m68k-elf/m68k-elf/bin don't have the -4.4 suffix?
15:30:13S_a_i_n_tIt doesn't seem possible to me.
15:30:54kugelother than wrong code, but I can't look at it right now
15:31:39S_a_i_n_tfair enough.
15:31:42S_a_i_n_tIt just doesn'
15:31:51S_a_i_n_t+t seem possible to me.
15:31:54n1skugel: i don't have that dir, i have /usr/local/m68k-elf/bin and they have no prefix or suffixes at all
15:32:08n1sas, ld, gcc etc
15:32:14S_a_i_n_tso I thought I'd ask...I assumed if it were broken for one, it would break for all.
15:32:22kugeln1s: sorry, I meant that dir
15:33:32kugelthen I suppose there can't be a transition phase cannot happen
15:34:20kugela transition phase as in we have current builds with the old compiler for group a, and new gcc builds for group b
15:34:30gevaertsAlexP: I'd say "partly"
15:35:12gevaertsYes for the manual bits, maybe for SYS_USB_CONNECTED, no for the backlight
15:36:12CIA-7New commit by 03rmenes (r28447): More plugin keymaps for the GoGear HDD6330.
15:37:37 Join evilnick [0] (
15:37:54wodzn1s: so what is the consensus - how should I prepare gcc-4.4 toolchain to not trash current one?
15:37:55CIA-7r28447 build result: All green
15:38:10n1swodz: use a differen −−prefix
15:38:32kugelset the RBDEV_PREFIX environment variable to non-/usr/local
15:39:25wodzn1s: so finally what binutils version should I use?
15:39:47 Quit factor (Quit: Leaving)
15:39:47n1swodz: i use 2.20.1 shouldn't matter much though
15:40:33kugeli would use the latest stable release which is 2.20.1
15:41:59n1sbuilding with gcc3.4 still works for me so i assume that the new binutils works fine with it too but i think it can not build with the new version for some reason
15:42:18n1ss/can not build/can not be built/
15:42:22 Quit Alchimysta (Quit: CGI:IRC (EOF))
15:43:51n1si should retry that as then we could just keep 2 different gcc's with the same binutils
15:44:20 Join factor [0] (
15:44:36wodzLambdaCalculus37: ping
15:45:39 Quit kugel (Remote host closed the connection)
15:46:48*TheSeven just found the reason for the nano2g flash slowness
15:48:49S_a_i_n_tTheSeven: Oh?
15:49:09LambdaCalculus37wodz: Pong
15:49:15TheSevenit's a silly bug that slipped in during some code cleanup
15:49:48TheSevenaveraging ~3.6MB/s write speed right now (copying music)
15:50:25gevaertsTheSeven: could it cause other problems?
15:50:30TheSeventhat used to be 5MB/s after I introduced the parallelization patch, but right now there's some profiling code enabled that migh slow it down
15:50:43TheSevenbefore the fix it was ~1.2MB/s
15:51:15TheSevengevaerts: it was just causing the flash chip capabilities to be misdetected (neither cached nor interleaved writes supported by any chip)
15:51:36gevaertsAh, ok
15:51:56TheSevengevaerts: could it be possible that this in endpoint 0x80 bit issue i pointed out yesterday is what causes our hid problems?
15:51:59gevaertsSo 3.7 is "fine", but the fix will presumably also be small and safe enough to backport?
15:52:20wodzLambdaCalculus37: I am uploading builds with FS #11712. I'll provide links in a minute
15:52:25TheSevenit's just slow, but the fix is a two-line correction of something that's apprently wrong
15:52:36 Quit jepler (Ping timeout: 265 seconds)
15:52:53 Quit factor (Quit: Leaving)
15:52:56LambdaCalculus37wodz: Cool. :)
15:53:11 Join factor [0] (
15:54:30 Quit teru (Quit: Quit)
15:55:03 Join Drise [0] (Drise@
15:55:30gevaertsTheSeven: I doubt that. HID only looks at the direction
15:55:44wodzLambdaCalculus37: and
15:55:53DriseHey guys. Why hasn't usb been enabled for SansaAMSv2 with the patch in flyspray?
15:56:23TheSevengevaerts: might it confuse UMS with HID packets because of this?
15:56:31 Part Drise
15:56:31 Join Drise [0] (Drise@
15:56:37S_a_i_n_tDrise: Read the comments on the task
15:56:44S_a_i_n_tit explains it quite well.
15:56:53gevaertsTheSeven: only if usb_core.c is buggy. That one should never call the wrong transfer_complete() function
15:57:01TheSevenand which way is the correct way actually? with or without 0x80? i'm passing it without right now
15:57:08LambdaCalculus37wodz: Going to try the iPod video build first.
15:57:43wodzLambdaCalculus37: The test is simple - go to system->debug->view I/O. It should look the same as without patch
15:58:47TheSevengevaerts: should I add something to the commit message indicating that this fix will be backported?
15:59:24 Join jepler [0] (~jepler@emc/developer/pdpc.professional.jepler)
15:59:26 Join toffe82 [0] (
15:59:38b0hoonpixelma: how to move the object to the other layer? :D
16:00:09DriseFrom what I've read, and correct me if I'm wrong, that the patch is working, with the occasional connection in usb1.1
16:00:25gevaertsTheSeven: depends. If you plan to do the backporting yourself right away, I don't think any mention is needed. If you don't, a hint for later backporters is definitely useful
16:00:38*gevaerts asks AlexP for his opinion
16:01:14*TheSeven thinks a formalized commit message tag could be helpful to automate things later
16:01:24LambdaCalculus37wodz: Looks to be working on the iPod video.
16:01:24toffe82LambdaCalculus37: the schematics of the 6330 and 6320 mentioned 64mb of RAM
16:01:45toffe82there is nothing about another version with less memory
16:01:46 Quit factor (Quit: Leaving)
16:01:49LambdaCalculus37toffe82: That's rather strange why playback would crash on mine.
16:01:49gevaertsTheSeven: I suspect usb_core_transfer_complete() wants the endpoint without the direction bit
16:01:53 Join factor [0] (
16:02:08TheSevengevaerts: in that case my implementation would be correct. damn.
16:02:11LambdaCalculus37toffe82: I'm going to try 64MB again on mine just to make sure it isn't being goofy with me.
16:02:16gevaertsTheSeven: bad luck :)
16:03:26gevaertsTheSeven: The endpoint number you pass in there is used as an array index in usb_core_handle_transfer_completion(), so leaving the direction bit in would actually be dangerous
16:04:22LambdaCalculus37wodz: Let me get out my USB cable for my GoGear and I'll try that one out.
16:04:49TheSeventhis IMHO deserves a comment, as it's breaking the convention used in other places (e.g. the endpoint allocator)
16:05:42 Quit swilde (Remote host closed the connection)
16:06:03DriseS_a_i_n_t: What is the command to build the tools in cygwin?
16:06:20S_a_i_n_t"the tools"?
16:06:44S_a_i_n_twhat are you trying to build?
16:06:47*wodz is looking for testers with Archos Player and other SH Archos target
16:07:22Drisefor compiling.
16:07:55TheSevengevaerts: should i just commit this to svn://
16:07:57S_a_i_n_tI am not sure what tools you are talking you have the source downloaded, and the toolchain built?
16:08:03S_a_i_n_tDrise: ^
16:08:22gevaertsTheSeven: I think so
16:08:32DriseAnd if I don't respond immediatly, it is becuase i have lag +20 seconds or so. And yes I need to build the tool chain, I think.
16:08:46DriseI do have source downloaded.
16:08:51S_a_i_n_tDrise: Do you have the source downloaded?
16:09:21S_a_i_n_taha..ok, cd to "\rockbox\tools\
16:09:35S_a_i_n_tthen do "./"
16:09:40S_a_i_n_t(without the quotes)
16:10:06S_a_i_n_tthen select "e" to build the eabi toolchain.
16:10:24S_a_i_n_tthen wait 6-to-12 hours for it to compile ;)
16:12:06DriseAnd hope for no errors.
16:12:13DriseLike last time.
16:13:26 Quit Strife89TX (Ping timeout: 252 seconds)
16:14:10LambdaCalculus37wodz: Sorry for the delay, I'm taking a little advantage of having my HDD6330 hooked up to start transfering some music into it. :)
16:17:03 Join Strife89TX [0] (~cstrife89@
16:17:23*Drise is lagging +30 seconds...
16:21:10S_a_i_n_tDrise: If you do get any errors, either PM me or post here and highlight me...pastebin the failed section of build output, and I should be able to tell you what packages you need to install.
16:21:37S_a_i_n_tthere's quite a few packages you don't *need* to build the toolchain...but can speed up building it quite a lot.
16:21:39DriseS_a_i_n_t: Is there a way to set "checkpoints" so to speak with this thing in case of errors? Is there something we (I) could do to implement this?
16:21:54S_a_i_n_tit needs to build start to finish
16:22:01S_a_i_n_tif it fails, you must start again.
16:22:24S_a_i_n_tyou don't have to DL everything again, but the build dir MUST be cleared if it fails.
16:24:07DriseWhich is in tmp correct?
16:24:23 Join MethoS- [0] (~clemens@
16:25:32 Quit factor (Ping timeout: 252 seconds)
16:25:39AlexPgevaerts, TheSeven: If you intend to do it yourslef, then I think nothing special is needed - I plan to just look at the commits to the branch to do e.g. release notes. If you don't intend to do it, then it'd be good to have a marker.
16:26:06CIA-7New commit by 03theseven (r28448): Fix a small bug that prevented the iPod Nano 2G NAND driver from detecting interleaved write-capable chips.
16:26:11CIA-7New commit by 03theseven (r28449): Fix a small bug that prevented the iPod Nano 2G NAND driver from detecting interleaved write-capable chips. (Backported from r28448)
16:26:21 Join factor [0] (
16:26:22AlexPLike that looks ideal to me :)
16:27:33 Quit Strife89TX (Quit: Vamoose.)
16:27:45CIA-7r28448 build result: All green
16:28:22 Join petur [0] (d408b802@rockbox/developer/petur)
16:36:17***Saving seen data "./dancer.seen"
16:36:21*LambdaCalculus37 prepares to promote the GoGear HDD16x0 to Unstable
16:37:37 Quit Drise (Ping timeout: 240 seconds)
16:46:22 Quit _jhMikeS_ (Read error: Connection reset by peer)
16:46:42 Join jhMikeS [0] (
16:46:43 Quit jhMikeS (Changing host)
16:46:43 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
16:52:14 Join kugel [0] (~kugel@rockbox/developer/kugel)
16:52:48 Join Drise [0] (Drise@
16:57:22 Quit Drise (Ping timeout: 265 seconds)
16:58:26 Join Drise [0] (
16:58:33LambdaCalculus37wodz: Works on my HDD6330 as well.
16:58:43 Part b0hoon ("Back to work.")
17:00:42LambdaCalculus37wodz: I don't have my Archos JBR with me today; if I did I'd help test the patch there as well.
17:06:53pSiKOhi everybody, I have a question about Mpeg movie on rockbox
17:07:01pSiKOI follow the
17:07:29pSiKOand use mencoder, on my pc movie is playing fine
17:07:47pSiKObut on RB, the soud is delayed
17:08:01pSiKOsomeone knows why?
17:08:36pSiKOand btw if some can write here the option he use to encode video, that could help me, thx
17:09:09pSiKOI'll try with many video the result is always the same
17:09:25gevaertspSiKO: which player do you have?
17:09:32pSiKOsans fuze v2
17:09:34n1swodz: i'm pretty sure i spotted the bug
17:09:44n1si just don't know how to fix it
17:09:48DriseI've known of audio bitrates can mess with sounds syncing, but that is from Sansa Fuze using the sansa converter
17:10:31DrisepSiKO, give me a minute and i can tell you what to look for
17:10:49pSiKODrise: thx a lot :)
17:11:40DriseWhat is the audio sample rate of your movie? (assuming windows.... right click>properties>details>audio sample rate)
17:12:02pSiKOsource or destination ?
17:12:39pSiKOsource is AAC @64kb
17:12:49pSiKOdestination is mp3@128kb
17:12:59DriseI'm talking about the file itself.
17:13:23DriseThe sample rate is different than the encoding rate and type.
17:13:30gevaertsWhat video bitrate did you use?
17:13:31Driseright click the file
17:13:41pSiKOI'm not under windows
17:14:03pSiKOvide bitrate is mpeg2@600kb/s
17:14:41pSiKODrise: you speak about audio sample on the source video or on converted video ?
17:15:19gevaertspSiKO: Drise is asking for sample rate, not bitrate
17:15:20pSiKObasicly it's a flv from youtube, converted using info on the mpeg wiki
17:16:35DriseI've always had issues with the sample rate causing sync issues.
17:16:43pSiKOgevaerts: sorry but what do you mean by "sample rate" ?
17:16:52gevaertspSiKO: have you tried lower bitrates? 128kbps for audio sounds high if the source is only 64kbps, and I'm not at all convinced that 600kbps video makes sense on 220x176. I'd go for 64kbps and 500kbps
17:17:27pSiKOok thx, I'll try right now
17:17:34gevaertspSiKO: hm, not sure how to explain that. It's the number that's usually 22.050 kHz or 44.1 kHz or so
17:17:51pSiKObtw if someone have a video that play fine on Sansa/RB can he post the file details here ?
17:18:06pSiKOha ok nice sound is re-sampled at 44.1 kHz
17:18:10gevaertspSiKO: the videos at should play fine
17:18:29gevaertsThe first one should be right for the fuze
17:18:58pSiKOhaaa! , I doesnt seen there is sample !!
17:19:08pSiKOthx again guys
17:19:11 Part Zagor
17:19:27 Join edboyer93 [0] (
17:19:36 Join Strife89TX [0] (~cstrife89@
17:19:45pSiKOI take thise one ""
17:20:03gevaertsyes, that should be the right one
17:20:45pSiKOif it's played correctly, I use the same setting for all my video
17:21:32wodzn1s: whats the problem than?
17:22:05wodzLambdaCalculus37: Can you test on SH any time soon?
17:23:40n1swodz: when calling the profile_thread_stopped function, the arg is pushed on stack (and stack pointer decremented), then after that function returns the store_context function stores the stack pointer before it is incremented again so the wrong stackpointer is restored when it switches to that thread the next time
17:24:31n1sfun thing is that i don't know if this can be fixed without writing all of switch_thread in asm :(
17:25:02TheLemonManlol theres a thread switcher in C ?
17:25:30LambdaCalculus37wodz: I can test tonight.
17:25:53wodzLambdaCalculus37: do you want me to provide the build?
17:26:03TheLemonManohmy, the world is near the end then
17:26:14 Join stoffel [0] (
17:26:50 Quit kazaik (Ping timeout: 245 seconds)
17:26:59*n1s wonders if jhMikeS has any ideas about this
17:27:13LambdaCalculus37wodz: I can roll my own build tonight with the patch applied.
17:27:43TheLemonMann1s, any link ?
17:28:21n1sTheLemonMan: it's in firmware/thread.c you can browse the source from the website
17:29:07TheLemonManits machine independent ?
17:29:29n1sthe function isn't long so writing it in asm wouldn't be very much work but it's not so nice if someone want's to change it...
17:30:01n1sTheLemonMan: mostly, the store/restore_context functions are inline asm
17:30:40n1sand various interrupt/core sleep functions are specific for each core
17:31:42TheLemonManoh, i see, looks pretty nice
17:32:09Torneokay, so I kinda have some kind of SMART support working, but I have no idea what it should do if it notices the disk is dying.
17:32:15Tornewe don't have a clickthrough-type screen.. :)
17:32:42n1s"your disk is dying. (yes/no)?" :)
17:33:05Tornen1s: "your disk is dying, are you afraid? (yes/no)" )
17:34:01S_a_i_n_tTorne: "*Panic*: The End Is Nigh!"
17:34:02Tornea splash seems kinda inadequate
17:34:11Tornepanicing could be very annoying though since it may be mistaken
17:34:36Tornei thought maybe a yesno where yes sets a setting so it doesn't tell you again
17:34:43gevaertsPanic every second time?
17:34:44Tornebut that also seems like it could be dismissed too easily
17:35:19S_a_i_n_tsurely you could make it so it can't be dismissed for X seconds?
17:35:33TheLemonManan icon that keeps reminding you that the end is near ?
17:35:34Tornenot with yesno
17:35:47Tornethta would have to be inventing a new UI element type thing
17:35:52gevaertsTorne: I know! Provide a wps tag for it
17:36:22 Quit GodEater (*.net *.split)
17:36:23 Quit bzed (*.net *.split)
17:36:28 Join bzed [0] (
17:36:40 Join GodEater [0] (
17:36:41 Quit GodEater (Changing host)
17:36:41 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
17:37:07TheLemonManformat the hdd to stop the sufference ?
17:37:22Tornegenerally that doesn't mkae threshold-exceeded go away :)
17:37:40Torneanyway enough silly suggestions, i think..
17:38:01Tornei considered this before for the 32/64mb build thing as a way to tell you you had the 32mb build on a 64mb player and couldn't think of anything nice then either
17:38:05Tornebut we solved that one by unifying the builds :0
17:38:17gevaertsNo such solution now though
17:38:34pixelmaquick "poll" about readability: "#if (CONFIG_CODEC == SWCODEC) || !defined(SIMULATOR)". How obvious is this line and can people say what it is good for? I know it, I'm just curious why it was changed to something wrong twice and if there is something to make it better (and prevent a third time)
17:39:36n1spixelma: i agree that one is non obvious but not sure how to make it nicer
17:40:01n1sprobably adding acomment if there's none is easiest
17:40:20n1sTorne: i'd go with putting the info in the debug menu as a first step
17:40:26alexbobPpixelma: that seems obvious to me...
17:40:29pixelmabut you understand what it is good for?
17:40:32Tornen1s: yeah, done that already
17:40:32gevaertspixelma: should that be !defined(SIMULATOR) or (CONFIG_PLATFORM & PLATFORM_NATIVE) ?
17:40:39alexbobPoh, I don't understand what it's fore
17:40:42alexbobPbut I can tell what it does :P
17:40:44Tornen1s: well, not committed it, but implemented it and tried it on my ipod :)
17:40:58n1spixelma: everything but hwcodec sims so i guess everything that actually plays music :)
17:41:05DriseS_a_i_n_t: Are you here?
17:41:10Tornen1s: at the moment it just checks for smart being supported and enabled, and if it's enabled it does SMART RETURN STATUS to ask the drive if thresholds are exceeded
17:41:14Torneit does it once at boot
17:41:18gevaertspixelma: I'm guessing it's about sound playback on the sim?
17:41:25Tornebut if thresholds *are* exceeded then it should probably tell you somehow :)
17:41:27S_a_i_n_tDrise: yes
17:41:28pixelmaand the 100 points go to n1s
17:41:31DriseI got warning about vnsprint, it is assuming it is ok.
17:41:56pixelmagevaerts: exclusion of metronome in hwcodec sims
17:41:59DriseShould I be worried?
17:41:59n1sTorne: cool, i have no good suggestions for the ui stuff either though :/
17:42:05gevaertsI'd say !(CONFIG_CODEC == HWCODEC && defined(SIMULATOR))
17:42:37S_a_i_n_tDrise: Basically, you'll know if it's not ok, because everything will'll only know everything is ok when you get the message telling you it has completed, and to edit <blah> to your path
17:42:49pixelmagevaerts: that's even less readable to me
17:42:50S_a_i_n_t*don't* edit the $PATH don't need to
17:42:55S_a_i_n_tDrise: ^
17:43:03gevaertspixelma: in that case a good comment :)
17:43:33pixelmagot it now, hmm...
17:43:56pixelmasomeone should implement MAS emulation ;)
17:44:17 Quit drizztbsd (Ping timeout: 252 seconds)
17:44:19gevaertsyes, that would fix the entire issue :)
17:44:21n1sTorne: i think a splash and setting could work actually, slightly annouying untill you actively disable it, which can't really be done by accident
17:44:38Tornen1s: a longish splash, you mean? :)
17:44:57Tornealso, i'm not sure what hte text should be
17:44:58pixelmaadding a comment was my idea too
17:45:06Torneit would have to be fairly brief so that it fits on various screens
17:45:13pixelmabut I hoped there could be something nicer
17:45:14Tornesince i'm resaonably sure SMART should work on basically every ATA-based device
17:45:21Torneand that includes the Archoses :)
17:45:26gevaertspixelma: can't one buy a MAS on a parallel port adapter? Maybe we could do that :)
17:48:45wodzgevaerts: It won't work in the USB era :-)
17:50:02tmztTorne: the western digital basic drives had it disabled in firmware
17:50:12tmztthey were only meant to be used through usb so didn't need it
17:50:48 Quit docgnome (Ping timeout: 252 seconds)
17:51:22Tornetmzt: seems odd..
17:51:37tmztthey were sata as well
17:52:00 Quit grndslm_ (Ping timeout: 265 seconds)
17:55:01Torneoh, you're not talking about anything relevant to rockbox
17:55:02Tornei see ;)
17:55:09 Join Topy [0] (
17:56:28 Quit T44 (Read error: Connection reset by peer)
17:59:17Tornen1s: not sure what the splash could say that is reasonably likely to fit on variou sscreens ;)
17:59:36n1snope, that is tricky
18:01:25pixelmathe line above was changed to just have the SWCODEC part and was wrapped in a "#ifndef SIMULATOR" which means - metronome is now unnecessarily excluded from hwcodec builds and swcodec sims
18:02:19pixelmathe whole thing was done to "improve readability" but I don't think it does in maybe 90% of the changes :\
18:02:34pixelmait was a bigger change
18:03:08n1sreadability is usually subjective
18:03:37 Quit Strife89TX (Quit: Heading to the library.)
18:04:09gevaertsreadability is definitely subjective, but it should never win over correctness...
18:04:25n1sno, definitely not :)
18:05:05gevaertspixelma: which revision is this?
18:05:11 Join docgnome [0] (~docgnome@
18:05:45pixelmagevaerts: r26114, I stick to which gives me the complete picture
18:06:33 Join panni_ [0] (
18:06:34pixelmaor for the diff
18:07:06gevaertspixelma: yes, that's definitely wrong
18:07:45gevaertsI guess he was misled by all other !defined(SIMULATOR) conditions being &&
18:07:59pixelmayeah, thought so too
18:08:12pixelmaso I should leave the other changes and move metronome to a more appropriate place (and restore the correct condition)?
18:08:34gevaertsI think that's best, yes
18:09:35pixelmaeven if it's just a plugin, but maybe this could be backported then too
18:10:36gevaertsleaving out a plugin by accident is definitely a regression, yes :)
18:11:07wodzpixelma: is there svg version of your remotes? I would like to borrow 'rocker arrow' for mpio :-)
18:12:08pixelmathere are SVGs of the iriverh100 and iaudiom3 remotes in SVN (manual/rockbox_interface/images)
18:12:33kugel implements the scroll threshold
18:12:40 Quit BlakeJohnson86 (Ping timeout: 245 seconds)
18:12:52kugeland the immediate opening of the context menu
18:12:57pixelmawodz: the iaudiom3-front also has those arrows
18:15:32wodzpixelma: can this one be pushed as well?
18:16:01pixelmagevaerts: (CONFIG_PLATFORM & PLATFORM_NATIVE) replaced the SIMULATOR defines?
18:16:41wodzpixelma: not entirely I think
18:17:07gevaertspixelma: actually, (CONFIG_PLATFORM & PLATFORM_HOSTED). I also changed the "!" bit :). But yes, as wodz says, not always. Maybe in this case SIMULATOR is still the right one...
18:17:24*gevaerts thinks that kugel will probably know
18:18:21pixelmawodz: hmm, not sure but I didn't draw an arrow for the push movement on the remotes either. I didn't like the look and thought placing the button label there is clear enough
18:18:43LambdaCalculus37The HDD6330's LCD refresh appears to be quite slow.
18:18:51kugelgevaerts: what's the issue?
18:18:55 Quit factor (Ping timeout: 245 seconds)
18:18:57pixelmagevaerts: well the only "problem" are hwcodec sims
18:19:32gevaertskugel: metronome needs to be excluded on hwcodec sims. Do we need SIMULATOR or CONFIG_PLATFORM & PLATFORM_HOSTED ?
18:20:16 Quit Drise (Quit: Going to class.... *sigh* Calculus...)
18:22:28kugelgevaerts: it shouldn't matter much if it's with && defined(HWCODEC)
18:23:12 Join BlakeJohnson86 [0] (
18:23:28*kugel looks at the file
18:24:12kugelfor most of the plugins SIMULATOR seems more correct
18:24:22 Quit petur (Quit: later!)
18:25:42pixelmadefined(HWCODEC) doesn't exist IIUC, that's not the point here though, just saying ;)
18:26:32 Join patgodo [0] (
18:27:03pixelmaI'm a bit confused now where to go from here
18:30:07kugelfor metronome it should be "#if CONFIG_CODEC == HWCODEC && !defined(SIMULATOR) || CONFIG_CODEC == SWCODEC"
18:30:11 Quit sasquatch (Ping timeout: 265 seconds)
18:30:24 Join b0hoon [0] (
18:30:27 Join JesusFreak316 [0] (
18:30:34LambdaCalculus37b0hoon: Ping
18:30:56b0hoonLambdaCalculus37: pong
18:31:21pixelmakugel: it was "#if CONFIG_CODEC == SWCODEC || !defined SIMULATOR" before
18:31:37pixelmawhich worked back then
18:31:40LambdaCalculus37b0hoon: You were right about the HDD6330 having 64MB of RAM. Mine was just being goofy on me.
18:31:44 Join Strife89 [0] (a80dbf53@gateway/web/freenode/ip.
18:31:52kugelthat's the same, just shorter and less clear :)
18:32:23 Quit n1s (Quit: Lämnar)
18:32:24pixelmaless clear seems to be subjective and which is why I asked before
18:32:33b0hoonLambdaCalculus37: hmm that's not good if it doesn't work for you :/
18:32:39AlexPgevaerts: is the r28429 diff, if you get 5 minutes could you chop out the uneeded bits? I think I know which they are, but to avoid mistakes :) I'll then do some test builds
18:33:01b0hoonLambdaCalculus37: so something is wrong with the hardware?
18:33:34LambdaCalculus37b0hoon: No, nothing's wrong now. I'm running a 64MB build and it's working just dandy.
18:34:05b0hoonLambdaCalculus37: so why it didn't work before?
18:34:21LambdaCalculus37b0hoon: No clue at all. A Heisenbug, perhaps?
18:35:00b0hoonLambdaCalculus37: yep :)
18:35:30b0hoonLambdaCalculus37: good work with the plugins :)
18:35:46LambdaCalculus37b0hoon: Thanks! I'm going to finish off more tonight.
18:36:01pixelmakugel: gevaerts suggested "#if !(CONFIG_CODEC == HWCODEC && defined(SIMULATOR))" which is a third way to write it and I'd prefer that slightly over yours
18:36:20***Saving seen data "./dancer.seen"
18:36:59 Quit MethoS- (Remote host closed the connection)
18:37:05kugelI don't mind which way
18:37:10LambdaCalculus37b0hoon: Have you noticed LCD refresh to be slow on your HDD6320?
18:38:11pixelmaanyway... you answered the question if SIMULATOR should be used or not. I'll only change that and move it outside the platform block. If someone else thinks that something needs changing about the latter, feel free
18:38:16b0hoonLambdaCalculus37: yes, it's slow... actually it is slower than in case of ipod video
18:38:32 Join soap__ [0] (
18:38:53b0hoonLambdaCalculus37: but Ipod video has a mapped video memory IIRC
18:39:03LambdaCalculus37b0hoon: I can practically see the screen redraw if I look at it from a certain angle.
18:39:38LambdaCalculus37b0hoon: Any clue how the video memory on the HDD63x0 works?
18:39:45wodzpixelma: What should be on the labels for buttons? Should be full word (Play, Volume Up, Forward) or should it adhere to the label on the device itself (which in case of MPIO would mean FF, Func and REW instead of Forward, Function, Rewind)
18:39:45Torneipod video's isn't really mapped, i thought?
18:39:58Tornethe BCM has a framebuffer but the framebuffer is not directly read to drive the display, the BCM has an internal framebuffer too
18:40:01Tornei could be wrong
18:40:27 Join bertrik [0] (
18:40:27 Quit bertrik (Changing host)
18:40:27 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:40:28b0hoonTorne: no i could be wrong totally
18:41:07pixelmawodz: the way you refer to it in the manual, most use the full words and I like it better personally too
18:41:15LambdaCalculus37Torne: IIRC on the iPod video we use very little functionality from the BCM.
18:41:15Torneb0hoon: yeah, it's as i describe more or less
18:41:24TorneLambdaCalculus37: little, yes, but it's still not a direct mapped framebuffer
18:41:37Tornelcd_update and lcd_update_rect poke the BCM into sucking the contents of our video buffer up
18:41:58 Quit Staphylo (Read error: Connection reset by peer)
18:41:59Tornehm, in fact no, we have to write it into it.
18:42:04Torneso yeah, it's not even DMAed
18:42:08Tornethat's why it's so slow :)
18:42:29LambdaCalculus37b0hoon: Do you know anything about the LCD controller in the HDD63x0?
18:42:29 Join Staphylo [0] (
18:42:47pixelmawodz: I mean if the user will read "Press Play to start playback" or so then the picture should also have "Play" on it.
18:42:49 Join sasquatch [0] (~username@
18:42:57b0hoonLambdaCalculus37: in the OF LCD is filled through the LCD2 bridge
18:43:04b0hoonLambdaCalculus37: not much
18:43:50b0hoonLambdaCalculus37: that is the main problem
18:44:32b0hoonb0hoon: there is no any sophisticated code in the OF i'm afraid
18:44:42b0hoonLambdaCalculus37: ^
18:45:07b0hoonLambdaCalculus37: :D but it's time to look at it deeper
18:45:23 Join Drise [0] (Drise@
18:46:00LambdaCalculus37b0hoon: On the wiki page, only the HDD16x0's LCD controller is listed; it's an Leadis Technology LDS176. Is this the same chip in the HDD63x0 or is there a different chip?
18:46:18 Nick soap__ is now known as soap (
18:46:28 Quit soap (Changing host)
18:46:28 Join soap [0] (~soap@rockbox/staff/soap)
18:47:26 Quit patgodo (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
18:47:50b0hoonLambdaCalculus37: it's hard to tell what is it, LCD is not described, LDS was matched to the behavior in the OF
18:48:33soappixelma, I think the choice to use full words in this situation is pretty cut-and-dry. I can't see an issue using full words (as opposed to the minimal words on the device itself) except for in the (currently non-existant case) where the hardware has ambigious abreviations.
18:49:28pixelmaif it's a common abreviation yes
18:49:58soapright. I don't believe there is a case currently where the hardware has "unusual" abreviations silk-screened on it.
18:49:58 Quit preglow (Ping timeout: 272 seconds)
18:50:02 Join preglow [0] (
18:50:34soapIf that were to be encountered I think a stronger case for using the abreviation as found on the hardware could be made.
18:50:36 Quit aevin (Ping timeout: 272 seconds)
18:50:43 Join aevin [0] (eivindsy@unaffiliated/aevin)
18:51:23LambdaCalculus37b0hoon: The service manual doesn't seem to describe an LCD controller in the block diagram.
18:51:34 Join Alchimysta [0] (
18:52:33 Join dt84 [0] (~dror@
18:52:46b0hoonyes, it's not on the main board it's in lcd
18:53:04LambdaCalculus37b0hoon: I see the ADV7179 TV-out controller, but it looks like it draws direct to the LCD.
18:53:41LambdaCalculus37b0hoon: Ahhh, now I see what you and lowlight have been talking about.
18:54:26dt84hi. I'm new to rockbox, installed it 2 days ago on my Sansa Clip+. As I noticed, rockbox sits on the original firmware, not replacing it, right? I noticed a bug: a few times when I opened the player, the original GUI appeared. Is it a known issue?
18:54:28pixelmasoap: there are targets with "A-B" buttons but we refer to them as "A-B" too
18:55:37AlexPdt84: No, it replaces it
18:55:47AlexPdt84: But we keep the OF round so you can dual boot
18:56:17b0hoonLambdaCalculus37: about the radio?
18:56:21AlexPdt84: If you hold certain button on boot, or insert USB you will get the OF
18:56:21dt84AlexP: so what I noticed wasn't a but, but a feature?
18:56:59LambdaCalculus37b0hoon: No, about how the LCD works.
18:57:08AlexPdt84: It depends what you mean by "open the player"
18:57:40dt84AlexP: Turning it on
18:57:57 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:57:58AlexPyes, I got that - I meant how you were turning it on :)
18:58:31dt84oh :)
18:58:50dt84I'm aware of only 1 way of turning it on - the on/off button. I'm not talking about USB or anything like that
18:59:09LambdaCalculus37b0hoon: Oh! Speaking of the radio... mine works.
18:59:27AlexPOK - if you have a look in the manual, there is a certain button that if you press it when turning on it will boot the OF - I guess you might have been pressing that by mistke
18:59:46AlexPIt might even by if you turn on with hold on, I can't remember for the clip+
18:59:59b0hoonLambdaCalculus37: ADV7179 doesn't draw directly to the LCD, the from the bus are to the ADV, it gives the TV out only
19:00:16b0hoonLambdaCalculus37: the arrows from the bus
19:00:25AlexPdt84: hmmm, no - for the clip+ it is left apparently
19:00:36b0hoonLambdaCalculus37: mine not :D
19:01:47 Quit aevin (Read error: Operation timed out)
19:01:52 Join aevin [0] (eivindsy@unaffiliated/aevin)
19:04:54LambdaCalculus37b0hoon: At any rate, do you think we should at least get the HDD16x0 promoted to Unstable? The HDD63x0 should, at the very least, have all of the plugin keymaps committed and plugins enabled before we promote that build.
19:05:23LambdaCalculus37b0hoon: The HDD16x0 works well enough that it can be considered good for everyday use.
19:05:25b0hoonLambdaCalculus37: HDD6320 doesn't have the module with the radio
19:05:34 Quit anewuser ()
19:05:43dt84AlexP: where can you read this? from what I can see, the "home" button also loads the OF
19:05:47 Quit Drise (Read error: Connection reset by peer)
19:05:51b0hoonLambdaCalculus37: yes it should go to unstable IMO
19:06:02 Join Drise [0] (Drise@
19:06:09LambdaCalculus37b0hoon: We should add dynamic checking at startup to look for the FM radio. IIRC the Gigabeast does the same thing since the S30 doesn't have FM radio.
19:06:34LambdaCalculus37b0hoon: Would you like to do the honors? :)
19:07:28AlexPdt84: I was looking at the manual: - if home also starts the OF then the manual is wrong and needs updating
19:07:46AlchimystaDoes Rockbox read ebook?
19:07:48b0hoonLambdaCalculus37: i don't remember how :P
19:07:58LambdaCalculus37b0hoon: :P
19:08:49dt84AlexP: I see. Thanks :)
19:08:51bertrikLambdaCalculus37, the normal mechanism is to use tuner_get(RADIO_PRESENT) to detect presence of the tuner. This is already implemented for all tuner drivers.
19:08:52 Quit kugel (Remote host closed the connection)
19:09:11AlexPAlchimysta: Read the manual
19:09:16LambdaCalculus37bertrik: That's how the beast does it, correct?
19:09:17bertrikIt gets a little more complicated when it's possible to have more than one tuner type for one target
19:09:27AlchimystaYes o no?
19:09:32AlexPRead the manual
19:09:34LambdaCalculus37b0hoon: Did you see my comments on FS #10329?
19:09:42AlexPStop asking questions here that are in the manual
19:09:46AlexPIt is very very annoying
19:09:52LambdaCalculus37Alchimysta: Because the manual will tell you what you need to know.
19:10:05AlexPAlchimysta: We didn't bother writing the manual just so we could repeat it all again for anyone that asks
19:10:20b0hoonLambdaCalculus37, bertrik: it must be detected somehow through some GPIOs probably
19:10:21bertrikLambdaCalculus37, yes, all targets already do it like this AFAIK. The radio menu item is only present when the tuner has been actually detected at runtime.
19:10:30Alchimystaok :(
19:10:37AlexPdt84: Is it just pressing home, or holding home while pressing power?
19:10:44pixelmaAlchimysta: people get tired of answering the same question again and again so a manual was created and people can find out themselves
19:10:54b0hoonLambdaCalculus37: yes.
19:11:34LambdaCalculus37bertrik: Ahh.
19:11:44bertrikb0hoon, I think most drivers just look for some magic id in one of the tuner registers instead of looking at specific GPIOs
19:12:23b0hoonbertrik: for the tea5767 it always return true
19:12:24dt84AlexP: holding it.
19:12:33AlexPdt84: Thanks
19:12:43bertrikb0hoon, oh, that
19:12:48bertrik's not nice!
19:13:22b0hoonbertrik: trunk/firmware/drivers/tuner/tea5767.c
19:13:50LambdaCalculus37b0hoon: That's not good at all if it always returns true.
19:13:58dt84AlexP: btw, I asked you before where did you get that info, cause I read the pdf manual and couldn't find info about that. So I guess the pdf and the online manuals are not synced. Just thought you might want to know.
19:14:08 Join domonoky [0] (
19:14:12 Quit domonoky (Changing host)
19:14:12 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
19:14:21AlexPdt84: Hmmm, they should be - let me check
19:15:02pixelmadownside of the online manual is that it is not searchable
19:15:27pixelmaPDF and online version are build from the same source and are only out of sync when something is broken
19:15:38AlexPdt84: I see that section in both pdf and online
19:15:58dt84AlexP: ?
19:16:32dt84I searched for "Starting the original firmware" in the pdf file, with no results.
19:16:48bertrikb0hoon, LambdaCalculus37, maybe we could at least check if there is a device responding to the i2c address
19:17:03bertrikthere is also a 4-bit chip id
19:17:06b0hoonbertrik: the module with the radio for the HDD6330 is detected on some pins of the Soc (it's from the service manual)
19:17:21b0hoonbertrik: good idea
19:17:26AlexPdt84: Do you see section 3.1.3?
19:18:13pixelmaI linked it... should be at the top of your page
19:18:24dt84AlexP: Yes. How odd. I guess that might be a strange bug in my pdf reader.
19:18:28AlexPpixelma: Yeah, talking about the pdf one though
19:18:37b0hoonbertrik: but the service manual is outdated (it's some first version) and the pins are inappropriate :/
19:18:46pixelmaoh, hmm
19:18:50AlexPdt84: I've just tried and mine doesn't find it either, even when looking at it (okular)
19:18:59AlexPdt84: What pdf reader?
19:19:00bertrikcan anybody with a target containing a tea5767 tuner, check if there's something in the FM debug menu. I expect some 5 byte output, the 4th byte contains the chip id
19:19:16dt84AlexP: Ubuntu's Document Viewer 2.32.0.
19:19:18AlexPdt84: Looks like something odd in the manual then
19:19:25AlexPAnyone have adobe reader?
19:19:45pixelmayep, downloading the PDF as we speak
19:19:49LambdaCalculus37bertrik: Besides the GoGear HDD6330, which other devices have the TEA5767 tuner?
19:19:58AlexPIf I cut and paste the heading here I get "Starting the original rmware"
19:20:06b0hoonLambdaCalculus37: H10 IIRC
19:20:28LambdaCalculus37b0hoon: I haven't got an H10.
19:20:32AlexPdt84, pixelma: There is something odd in that heading
19:20:43AlexPfi !=
19:20:57dt84oh. I guess that explains it...
19:21:00pixelmayes, the "fi" gets interpreted as the fi "character"
19:21:16b0hoonLambdaCalculus37: me neither
19:21:26bertrikThe tea5760 tuner driver does use a chip and manufacturer id to detect presence
19:21:51LambdaCalculus37bertrik: On my HDD6330: Read: 29 FF 36 20 CC; Write: A9 FF 00 5A 40
19:22:02pixelmaAlexP, dt84: some readers. It's been reported as bug before and I thought it was fixed
19:22:15pixelmaapparently it's not
19:22:22AlexPit seems not, no :)
19:22:40 Quit DerPapst1 (Quit: Leaving.)
19:23:36pixelmahmm, Adobe Reader doesn't cope with it either
19:26:25bertrikLambdaCalculus37, that's consistent with the chip id from the datasheet, which says about 4th byte, bit 3..1 : "Chip Identification: these bits have to be set to logic 0"
19:26:55AlchimystaAdobe Reader is sh*t, the most reader of pdf is Slim PDF Reader
19:27:25BagderAlchimysta: that's not helpful
19:27:48AlchimystaExcuse me -.-"
19:27:57bertrikIt would have been nice if they would have chosen something else than simply 0 for the chip id, but I think we can implement at least basic detection with this.
19:29:57 Quit Drise (Ping timeout: 240 seconds)
19:30:11AlexPWhere are the buttons to boot the OF defined (clipplus)?
19:30:52b0hoonbertrik: ok, for me its: 01 01 01 01 2C, sometimes it reads 02 02 02 02 2C, Write: 80 00 00 5A 00
19:30:52AlexPpixelma: Do you remember when before?
19:32:00bertrikb0hoon, hm, that's not consistent with the datasheet, is this on a stable target with working radio?
19:32:43b0hoonbertrik: no it is HDD6320 without the radio module
19:33:01knittlhi. the clock on my sansa e280v2 is too fast. could this be related to rockbox? i didn't notice it before using it
19:33:40b0hoonbertrik: just for comparing
19:34:37gevaertsAlexP: I think should do it. It compiles, but I didn't test
19:34:41bertrikok, thanks. I wonder if perhaps the fmradio_i2c_read call is failing and returning garbage data.
19:34:54AlexPgevaerts: cool, I'll test, cheers
19:35:14 Quit Alchimysta (Quit: CGI:IRC)
19:35:24LambdaCalculus37b0hoon: And IIRC there's also the similar issue with FM radio detection on the HDD16x0 platform as well?
19:35:44bertrikI'm not sure if we have a rockbox-wide convention on the return value of fmradio_i2c_read, I'll have a look
19:36:04b0hoonLambdaCalculus37: i don't know i don't have one yet to test
19:37:45bertrikBTW, it's a bit silly that there are two I2C addresses defined for the TEA5767. The chip only has one. We probably need two in the tuner driver because the various i2c drivers in rockbox have different conventions about I2C addresses (with or without the R/W bit)
19:38:10AlexPgevaerts: What is this meant to be fixing?
19:38:24b0hoonLambdaCalculus37: on my LCD there is a signature of the LCD controller or the model of the LCD: LPH9165-2
19:38:45gevaertsAlexP: connecting USB while the plugin is running.
19:39:06gevaertsTo be honest, I'm not convinced it's worth a backport, but...
19:39:11b0hoonLambdaCalculus37: google finds it but where is the datasheet? :D
19:39:12AlexPgevaerts: I thought that, but I just tried without the patch on e200v1 and it goes to USb fine
19:39:29AlexPah, this time it didn't
19:39:36AlexPIt did work first time :)
19:39:46gevaertsAlexP: I didn't look at it in depth :)
19:40:02 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
19:40:06AlexPI'm not convinced either
19:40:59LambdaCalculus37b0hoon: I have an HDD1630, but I need to repair it.
19:41:34bertrikLambdaCalculus37, b0hoon, shall I create a quick patch to implement tuner detection?
19:42:02LambdaCalculus37bertrik: Sure. :)
19:42:20b0hoonbertrik: yes please.
19:43:51 Join kazaik [0] (
19:47:18 Quit b0hoon (Ping timeout: 240 seconds)
19:48:20 Join b0hoon [0] (~quassel@
19:49:46 Join Jaykay [0] (
19:54:03 Join linuxstb [0] (
19:54:12 Quit linuxstb (Changing host)
19:54:12 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
19:55:18 Join DerPapst [0] (
19:59:03bertrikLambdaCalculus37, b0hoon, can you check this patch for tea5767 detection?
20:00:38*bertrik is a bit surprised this wasn't implemented already
20:01:03bertrikor maybe I'm overlooking something that prevents proper detection
20:03:40bertrikit adds 90 bytes to the hdd1630 binary
20:07:46LambdaCalculus37bertrik: Can you post it to Flyspray? I have to leave.
20:09:38 Quit LambdaCalculus37 (Quit: CGI:IRC 0.5.9 (2006/06/06))
20:10:22pixelmaAlexP: can't find it at the moment, maybe it was only talk about it in the forums
20:11:34pixelmabertrik: haven't looked at the patch at all but on the OndioFM there must be something too to tell the two tuners apart, one of which is also a common TEA the other a Samsung chip
20:11:53bertrikpixelma, I'll have a look at how it's done there
20:11:57 Join webguest09 [0] (
20:12:38 Quit webguest09 (Client Quit)
20:13:23amiconnbertrik, pixelma: On Ondio we don't need to perform detection - just checking a bit in the hardware mask is enough
20:13:42bertrikah ok
20:14:39bertrikI remember seeing something like that, but can't find it anymore
20:14:51amiconnThat's something archos got right (except on the Player) - those targets have a hardware mask byte, and hardware variants can be distinguished by checking the mask bits
20:15:44 Join LeonTrollski [0] (~LeonTroll@
20:15:52 Join Buschel [0] (
20:15:57 Quit LeonTrollski (Client Quit)
20:18:30 Quit toffe82 (Ping timeout: 245 seconds)
20:18:48bertrikOn the clip+, we have to detect either si470x or rda5802 and they are even at the same i2c address
20:19:37pixelmagevaerts: your line doesn't work, there is no such thing as HWCODEC, just MASxxxx
20:20:05 Join toffe82 [0] (
20:20:05pixelmawhich means it would need to be written for all MAS types
20:21:24gevaertshm, better to check for !SWCODEC then
20:22:38pixelmaok, trying
20:23:36*pixelma is glad to have done some testcompiling ;)
20:24:59 Join Luca_S [0] (
20:29:34amiconnbertrik: firmware/export/hwcompat.h
20:30:34 Join piggz [0] (~piggz@
20:31:17piggzhi, is there a 3.7 release for ipodvideo-64mb ?
20:32:34domonokypiggz: the 64mb and 32mb version have been merged. So there is only one ipodvideo version for 3.7 now.
20:32:42soapthe ipod video builds have been unified. Now the build autodetects 32 vs 64
20:33:28piggzok, the rockbox utility tries to download a 64m version which doesnt exist
20:35:35bertrikLamdaCalculus37, b0hoon, the patch is at
20:35:40piggzunless my utility is old
20:36:06pixelmapiggz: just wanted to ask you about the exact Utility version
20:36:21***Saving seen data "./dancer.seen"
20:36:47piggzpixelma: 1.2.6
20:38:17pixelma has 1.2.8 as download version
20:39:13pixelmathe changelog mentions the Ipod Video builds for 1.2.8
20:39:36TheSevenhm, is there a way to circumvent that "ipod patcher does not suport the 2nd generation nano!" bug?
20:39:40piggzpixelma: ok, np, i will update in a few mins
20:40:28gevaertsTheSeven: I guess we just need a new ipodpatcher release..
20:40:50TheSevenhow do i compile that for windows?
20:41:11TheSeveni have a user complaining about it in #freemyipod-support right now
20:41:23gevaertsmingw can do it
20:41:29TheSevendo you remember what triggered it and if it can be bypassed somehow?
20:41:51gevaertsThere was some old leftover code
20:41:55gevaertsI think I removed it
20:44:53 Quit froggyman (Quit: Bye)
20:49:34TheSevenhm, mingw fails
20:49:58TheSevenit's trying to execute "sh" from within a makefile and missing some files apparently
20:50:39 Quit b0hoon (Ping timeout: 252 seconds)
20:54:00 Join b0hoon [0] (~quassel@
20:55:42 Join stacker55 [0] (
20:58:05 Join froggyman [0] (~seth@unaffiliated/froggyman)
20:58:15stacker55On the wiki page for the HDD6330 ( the installation steps are posted for this device. The only thing missing there are the bootloaders. I had one on my machine, and like to help others who want to install RockBox on their HDD6320 / HDD6330 by provding the bootloader i used on my device. Can someone tell me where i can put the firmware.ebn file? Should i attach it to the wiki page, or is it bette
20:58:54 Quit Jaykay (Ping timeout: 252 seconds)
20:59:40 Quit slam__ (Quit: Leaving)
21:01:25 Quit piggz (Remote host closed the connection)
21:02:18 Join Jaykay [0] (
21:08:26wodzstacker55: as there was a talk about promoting HDD6330 to unstable I think there will be official bootloader release soon
21:09:00stacker55ah, well i can atest to it working nicely
21:09:05CIA-7New commit by 03rmenes (r28450): More plugin keymaps for the HDD6330. Quite a lot of commits. ;)
21:10:47CIA-7r28450 build result: All green
21:12:44stacker55it also means there is no need for the firmware file i used, right?
21:13:56 Quit Strife89 (Quit: Vamoose.)
21:14:51 Quit froggyman (Ping timeout: 252 seconds)
21:15:27wodzpixelma: Can you look if this is ok for manual?
21:15:38 Join LambdaCalculus37 [0] (
21:15:38 Quit LambdaCalculus37 (Changing host)
21:15:38 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
21:16:20AlexPwodz: What do you click on to download? I don't unerstand polish :)
21:16:59AlexPah, just the filename
21:17:44wodzAlexP: Or round 'pobierz' button at the bottom
21:19:37 Quit stoffel (Ping timeout: 255 seconds)
21:20:26pixelmalooks good
21:20:43pixelmais the "Volume Up / Down" a slider?
21:21:47b0hoonbertrik: it doesn't work for my HDD6330, HW is detected. :(
21:22:15b0hoonbertrik: for my HDD6320, sorry i'm tired
21:22:21bertrikare you sure it doesn't have a FM tuner chip?
21:22:48b0hoonbertrik: yes i'm sure :)
21:23:16LambdaCalculus37bertrik: Works on my HDD6330; the FM radio hardware is detected.
21:23:21stacker55if there's something i can test, let me know... (HDD6320 here)
21:23:57LambdaCalculus37stacker55: Want to try this patch out?
21:23:58b0hoonbertrik: it has a module without the chip, the module is detachable
21:24:01pixelmawodz: and (I saw it before on pictures too) - the display looks a bit like a portrait one but the used area is square I think. that would mean you have different "borders" on top/bottom and left/right. Is that correct (seems weird to me but if the Mpio designers chose it this way...)
21:24:44 Join froggyman [0] (~seth@unaffiliated/froggyman)
21:25:11stacker55i'll fire up my dev environment
21:26:05LambdaCalculus37stacker55: Basically what we want is if you still see the FM radio on your main menu or not.
21:27:15*bertrik blames the PP i2c driver
21:28:11bertrikit doesn't give a clear indication of i2c error
21:28:37CIA-7New commit by 03rmenes (r28451): It has been confirmed that the Philips GoGear HDD6330 does ...
21:29:37bertrikI could preinitialise the i2c read buffer with all 0xFF, so it won't accidentally detect the chip id in an all 0x00 buffer
21:29:51pixelmaLambdaCalculus37: I thought that would crash for you?
21:30:09CIA-7r28451 build result: All green
21:31:46pixelmaif it does I just see little sense to only set it back and forth all the time and I'd keep it at the low number as long as that works for everyone (even though it's a bit wasteful for others) until a better solution is found
21:32:44LambdaCalculus37pixelma: Actually, it was just a fluke thing. I tried building it again with 64MB and it does indeed work. My first build was just being goofy.
21:34:02LambdaCalculus37pixelma: I built another build with 64MB earlier and managed to get through several albums without a single issue. So I figured that since it's working right on my HDD6330, then I should set it back to 64MB.
21:37:24LambdaCalculus37bertrik: I'm getting different registers for the FM radio with your patch applied.
21:38:17LambdaCalculus37bertrik: Read - 17 FD 00 F0 2C; Write - 80 00 00 5A 00
21:39:02bertrikI expect them to be different depending on the station tuned, etc.
21:39:27bertrikBut the chip id bits still match the datasheet :)
21:39:43LambdaCalculus37bertrik: Good. :)
21:40:53pixelmagevaerts: for some reason I can't make the line work, seems I'm overlooking something but somehow metronome will be included in an OndioFM sim, I'm very close to just get back to the original solution
21:44:33b0hoonpixelma: could you take a look at the new image: FS #11705, please?
21:48:25 Join Drise [0] (
21:48:54LambdaCalculus37b0hoon: They do look pretty good.
21:50:55 Join Drise_ [0] (
21:51:57pixelmayes, I like it too. About the ordering the text a bit - these small tweaks can be done later
21:51:58 Quit soap (Ping timeout: 240 seconds)
21:52:35pixelmathis GoGear 6330 seems to have a completely flat front. Lucky you :)
21:53:12b0hoonpixelma: i have a problem with that... what exactly? i don't have a talent :P
21:53:21 Quit Drise (Ping timeout: 252 seconds)
21:53:33gevaertspixelma: do you have a patch to start with?
21:53:43b0hoonpixelma: yes it's completely flat
21:54:05b0hoonLambdaCalculus37: thanks ;)
21:54:26pixelmab0hoon: it was more meant as "Commit it. If it is in SVN someone else can have a look"
21:55:02b0hoonb0hoon: ok, thanks a lot :)
21:55:22*b0hoon is stupid
21:55:32b0hoonpixelma: ^
21:55:43pixelmagevaerts: one moment, I can only give you one with the line in the old style (I was a bit annoyed and wanted to have something working)
21:56:26 Quit sasquatch (Ping timeout: 276 seconds)
21:56:41Buscheln1s: the beast uses an arm11 cpu?
21:57:13pixelmagevaerts: but it has the "move part outside the simulator if"
21:57:46 Join sasquatch [0] (~username@
21:59:16 Quit Drise_ (Ping timeout: 250 seconds)
22:01:17CIA-7New commit by 03Buschel (r28452): Optimize mpc's synthesis filter by reducing stalls for arm9 and above. Speed up ranges from 4% (nano 2g) to 11% (beast).
22:01:27stacker55i seem to run into the problem that the VMWare image does not have an up to date arm11 compiler, nor will the toos/ script update it (it bugs with the message that it cannot detect the OS i'm using); will aptitude installing arm compilers bork the image ?
22:03:44wodzpixelma: mpio lcd has non square pixels
22:03:48CIA-7r28452 build result: All green
22:03:49 Join Drise_ [0] (Drise@
22:04:51pixelmawodz: interesting. A candidate for aspect ratio correction? :)
22:05:14wodzpixelma: and about volume up and down in mpio - it is kind of rocker you can press it from left side to decrease vol and on the right side to increase
22:06:23CIA-7New commit by 03Buschel (r28453): Using icode attribute for S5L870x is faster on some codecs. Speed up for mpc is ~3%, for aac ~1%.
22:07:41 Join soap [0] (~soap@rockbox/staff/soap)
22:07:45wodzCorrection wouldn't be that big I think
22:08:05CIA-7r28453 build result: All green
22:08:09LambdaCalculus37pixelma, b0hoon: I think that FS #10329 looks good as well.
22:09:13b0hoonb0hoon: ok, i will commit it.
22:09:55pixelmab0hoon: xou are speaking to yourself a lot today ;)
22:10:44b0hoonpixelma: i'm so tired my fingers are faster than my brain :/
22:11:20gevaertspixelma: that seems to work for me. I built a sim and normal build for Player and gigabeat f with it, and I got the metronome plugon for all of those except the Player sim
22:13:01pixelmayes, the patch as is works for me too but I wanted your "don't build if it is - hwcodec (or not swcodec) and sim -" style
22:13:28pixelmathat would make the comment clearer
22:13:41CIA-7New commit by 03Buschel (r28454): Minor optimization of mpc codec via moving dewindowing asm to icode section for S5L870x.
22:14:49 Join n1s [0] (
22:14:49 Quit n1s (Changing host)
22:14:49 Join n1s [0] (~n1s@rockbox/developer/n1s)
22:15:23CIA-7r28454 build result: All green
22:16:02CIA-7New commit by 03b0hoon (r28455): HDD1630: add small image made from SVG. FS #10329, Author: Marko Pahlke.
22:16:05 Quit bmbl (Quit: Verlassend)
22:20:02 Quit Jaykay (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
22:20:30 Join Horscht [0] (~Horscht@xbmc/user/horscht)
22:20:59CIA-7New commit by 03b0hoon (r28456): HDD6330: add small image made from SVG, FS #11705 by me.
22:21:17pixelmagevaerts: so, shall I just commit it as I have it now?
22:21:25gevaertspixelma: sure
22:21:46gevaertsWe can discuss this for ages, but all solutions are basically equivalent
22:22:10pixelmaseeing today's, I guess you are right
22:24:05LambdaCalculus37So no one objects to pushing the GoGear HDD1630 to Unstable?
22:24:37 Quit Drise_ (Read error: Connection reset by peer)
22:25:10 Quit dt84 (Ping timeout: 240 seconds)
22:29:16 Join anewuser [0] (anewuser@unaffiliated/anewuser)
22:30:22CIA-7New commit by 03pixelma (r28457): Enable building of metronome for hwcodec targets and swcodec sims which was accidentally disabled in r26114. Add a short comment.
22:31:17 Join MethoS- [0] (~clemens@
22:31:36 Quit froggyman (Quit: Bye)
22:31:49pixelmab0hoon: was the commit of the images in the release branch?
22:31:58CIA-7r28457 build result: All green
22:32:45pixelmacan't see them there either but I wonder why they don't appear on the front page, maybe I'm blind
22:33:25b0hoonpixelma: no
22:33:51 Quit TheLemonMan (Quit: Help me, i got shot! *DIES*)
22:34:08pixelmaah, www
22:34:21b0hoonpixelma: because these targets are not promoted to unstable yet
22:34:38LambdaCalculus37b0hoon: I'm going to promote the HDD16x0 now.
22:34:51LambdaCalculus37b0hoon: It works well enough to be classified as such.
22:36:04 Quit komputes (Read error: Connection reset by peer)
22:36:10pixelmacommits to the www module don't show on the frontpage either. I only thought you committed SVGs too but those were just the small pictures used in the tables (current build etc.)
22:36:24***Saving seen data "./dancer.seen"
22:37:00b0hoonLambdaCalculus37: go ahead
22:37:10CIA-7New commit by 03rmenes (r28458): Promote the Philips GoGear HDD16x0 series to Unstable.
22:37:20CIA-7New commit by 03pixelma (r28459): Backport r28457 - Enable building of metronome for hwcodec targets and swcodec sims which was accidentally disabled in r26114. Add a short comment.
22:39:23LambdaCalculus37Ladies and gentlemen, we have our first Philips GoGear port promoted to Unstable! :)
22:39:41Luca_S\o/ congrats :D
22:39:58b0hoonpixelma: yes, i wonder now how to add the svg, png and the pdf. pdf and png will be binary but what with the svg (mime-type)?
22:40:38pixelmaplease binary too
22:41:39b0hoonpixelma: ok, thanks.
22:42:41 Quit JesusFreak316 (Ping timeout: 245 seconds)
22:42:45b0hoonLambdaCalculus37: don't forget about tools/ -> status = 2
22:43:19TheSevenBuschel: RE r28453: the 96/80kB core/codec IRAM split isn't really optimal. IIRC you could swap it around to give the codecs 96kB IRAM as the core won't use more anyway
22:43:43TheSevenwith some tuning I managed to compile a build with 128kB codec IRAM like half a year ago
22:43:45pixelmab0hoon: it hasn't been done accurately for all but I tried to use the recommended mime-types
22:45:56LambdaCalculus37b0hoon: Doing that now and adding the HDD6330 to tools/ as well.
22:47:01CIA-7New commit by 03rmenes (r28460): Making sure that the HDD16x0 series is correctly reflected ...
22:48:49CIA-7r28460 build result: All green
22:49:11Luca_S shows Internal Server Error...
22:49:49gevaertsBagder: ^
22:50:19*LambdaCalculus37 whistles innocently
22:50:40*Luca_S didn't see anything.............
22:50:42pixelmaweird, it's the second page I see this on today
22:51:54Bagderany idea when it started?
22:52:14Luca_Sa few minutes ago
22:52:22LambdaCalculus37I hope that wasn't from my commit.
22:55:36 Quit pamaury (Remote host closed the connection)
22:55:47Bagderhehe, it was
22:56:01Bagderme fix
22:57:28Bagdersee line 60 of ...
22:58:17LambdaCalculus37Bagder: Oops.
22:58:22CIA-7New commit by 03bagder (r28461): fix: correct the perl syntax error ...
22:59:21Bagderof course it's not that good that the entire thing is that sensitive
22:59:22*pixelma points finger ;)
22:59:27wodzpixelma: any special way of creating pdf and png out of svg?
23:00:13CIA-7r28461 build result: All green
23:00:24Bagdersite back alive!
23:01:19 Quit GeekShadow (Read error: Connection reset by peer)
23:01:21pixelmawodz: so I don't have to type it all again
23:01:42wodzthx for link
23:01:43*Bagder crawls back into his cave
23:01:54 Quit ender` (Read error: Connection reset by peer)
23:03:36 Join ender` [0] (
23:06:25 Quit elcan (Read error: Connection reset by peer)
23:07:07BuschelTheSeven: why was this 96/80 (S5L8701) or 64/64 (S5L8700) chosen? other cpu's use 48KB for core and the rest for codec
23:07:23 Join fml_ [0] (
23:07:52TheSevenIIRC because no codec would use more than 80kB anyway back then
23:08:02fml_Do all SW_CODEC targets have the timestretch feature? Or do they have to have enough RAM?
23:08:31BuschelTheSeven: so, we could simply do the standard 48/x trick for both cpu's?
23:09:00TheSevenno idea about the 8700, but i guess the 8701 needs more than 48k for the core
23:09:12TheSevenremember that it uses 8k more stack than other targets
23:12:16Buschellet's see :)
23:12:31 Quit antil33t (Read error: Connection reset by peer)
23:12:46 Join elcan [0] (
23:14:24Buschelit linked
23:17:44Buschelhmm, it also linked with 32KB of iram... must be something going wrong :/
23:18:14 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
23:18:19 Quit esperegu (Read error: Connection reset by peer)
23:18:44 Join antil33t [0] (
23:20:41 Quit BlakeJohnson86 (Remote host closed the connection)
23:22:05 Join BlakeJohnson86 [0] (
23:31:14wodzWhat to do with OF package if official vendor (MPIO) site is down for long time? OF is needed to flash rb bootloader.
23:32:11 Join dt84 [0] (~dror@
23:32:55LambdaCalculus37wodz: You don't have a copy of the OF?
23:33:22wodzI do
23:33:46dt84After installing Rockbox on my Clip+, Ubuntu doesn't recogzines the player as "Sansa Clip+" (or whatever it said before), now it just written "1.9 GB Filesystem". Known issue?
23:34:10BuschelTheSeven: Is the IRAM size of the S5L8700 128 or 256KB? I see two different definitions
23:34:32 Quit fml_ (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
23:34:43BuschelTheSeven: btw, 48KB are enough for core on S5L8701
23:34:46TheSeven176kB actually
23:34:51wodzLambdaCalculus37: I am asking because in order to flash rb bootloader you have to have OF upgrade files which are not easy to find now.
23:34:57TheSevenand 256kB on the 8700
23:35:11Buschelthat was my question ;)
23:35:15TheSeven(which only seems to use 128kB for no apparent reason... unusable ports...)
23:36:59Buschel128KB (208KB for S5L8700) is quite a lot :) This should help compensate for the crappy memory speed
23:37:17Buschel...if used for codecs...
23:41:09TheSevenbertrik: is that ok?
23:42:02bertrikI have no idea, it's been too long since I looked at the s5l8700 memory map
23:42:31TheSeveni'm quite certain it's correct, but i'm not sure if there was a reason that you only used 128k on the meizus
23:43:28BuschelI will just add it to the tracker. so you can add any comments or results...
23:43:50TheSevenBuschel: I think you can just commit it if it builds ok
23:44:15TheSeveneven if it breaks the meizus it won't hurt much
23:44:32TheSeveni don't think anyone has looked at that port during the last six months
23:45:34bertrikMy last serious compile was in june for the meizu M6SP just before devcon
23:46:08TheSevenwhat's the state of the s5l8700 targets btw?
23:46:20TheSevendid you figure out the lcd driver?
23:46:22 Quit BlakeJohnson86 (Ping timeout: 240 seconds)
23:47:41bertrikmeizu m3: simple bootloader demo works, meizu m6sp: I can display a logo but then it hangs ,samsung yp-s3: a lot of stuff works (lcd, buttons, radio, rtc)but not file access
23:48:39 Part dt84
23:50:14CIA-7New commit by 03alle (r28462): Manual: make the presence of the timestretch menu description depend on the right condition (pitchscreen instead of just swcodec)
23:50:42 Quit shai (Quit: Leaving)
23:51:11*Buschel compiles meizum3
23:51:41bertrikLinux checks a couple of more things to detect a tea5767: 1) tuner registers should not be all the same 2) also check 5th byte in addition to chipid in 4th byte
23:51:55CIA-7r28462 build result: All green
23:51:56bertrikUnfortunately the PP i2c can only transfer 4 bytes
23:52:28bertrikAnd you can't really tell if the i2c transfer was ACKed or not, as far as I can tell
23:53:35wodzbertrik: doesn't PP i2c transfer any number of bytes but the buffer is 4bytes big?
23:54:00bertrikI don't know, there is no doc as far as I know
23:54:43Buschelbetrik: meizu-m3 does not compile anymore... this is not due to the iram config changes... I'll just submit the patch
23:56:33CIA-7New commit by 03Buschel (r28463): Reconfigure IRAM for S5L870x. S5L8700 has 256 KB of IRAM, use a 48 KB / 208 KB split for core/codec. S5L8701 has 176 KB of IRAM, use a 48 KB / 128 KB ...
23:57:48 Join Horschti [0] (~Horscht@xbmc/user/horscht)
23:58:11CIA-7r28463 build result: All green
23:58:34 Join BlakeJohnson86 [0] (

Previous day | Next day