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

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

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

#rockbox log for 2009-11-14

00:04:22 Quit BHSPitMonkey (Read error: 113 (No route to host))
00:12:01 Quit dmb (Read error: 104 (Connection reset by peer))
00:16:15 Join dfkt_ [0] (n=dfkt@unaffiliated/dfkt)
00:19:45 Nick fxb__ is now known as fxb (
00:21:10 Join JdGordon1 [0] (
00:21:32 Quit dfkt (Nick collision from services.)
00:21:37 Nick dfkt_ is now known as dfkt (n=dfkt@unaffiliated/dfkt)
00:21:55 Quit JdGordon1 (Client Quit)
00:23:09 Quit Horscht ("Verlassend")
00:23:45 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
00:33:03 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
00:39:44 Join xilinx [0] (n=chatzill@
00:40:30 Join AleMaxx_ [0] (
00:40:44 Quit AleMaxx (Read error: 110 (Connection timed out))
00:41:18 Quit AleMaxx_ (Client Quit)
00:42:23 Quit domonoky1 (Read error: 104 (Connection reset by peer))
00:50:31 Quit bertrik ("De groeten")
00:51:28 Quit dfkt (Read error: 110 (Connection timed out))
01:09:23 Quit ender` (" I just found out what Alt+F4 does.")
01:11:14 Quit amiconn (Nick collision from services.)
01:11:16 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
01:11:26 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
01:11:26 Quit pixelma (Nick collision from services.)
01:11:36 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
01:11:47 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
01:14:25***Saving seen data "./dancer.seen"
01:20:23 Join AEnima1577 [0] (
01:21:40 Join Lynx0 [0] (
01:28:41 Quit Lynx_ (Read error: 145 (Connection timed out))
01:28:42 Nick Lynx0 is now known as Lynx_ (
01:29:01 Join angerrsdfdsft [0] (
01:29:34angerrsdfdsftis there a way to install rockbox on sansa in msc mode
01:30:40advcomp2019angerrsdfdsft, that is the only way to install rockbox
01:31:47 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
01:33:55angerrsdfdsfti mean mtp
01:34:33advcomp2019nope.. it has to be in msc mode
01:39:57angerrsdfdsfthow to fix a brick sansa
01:40:44 Quit Tomis (Read error: 104 (Connection reset by peer))
01:40:45 Join Tomis2 [0] (n=Tomis@
01:48:10 Part toffe82
01:49:10 Quit angerrsdfdsft ("CGI:IRC (Ping timeout)")
01:50:36 Quit xilinx ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
01:50:50 Quit Lynx_ (" HydraIRC -> <- Wibbly Wobbly IRC")
01:54:20 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
02:00:16 Quit Galois (Read error: 113 (No route to host))
02:09:50 Join Galois [0] (
02:10:56 Join Strife89 [0] (
02:16:34 Quit linuxstb (Read error: 113 (No route to host))
02:23:42 Join ExcruciationX [0] (
02:24:16 Part ExcruciationX ("Leaving")
02:29:48 Nick fxb is now known as fxb__ (
03:13:48 Join midgey [0] (n=tjross@rockbox/developer/midgey)
03:14:28***Saving seen data "./dancer.seen"
03:24:46 Join BlakeJohnson861 [0] (
03:26:08 Quit BlakeJohnson86 (Read error: 110 (Connection timed out))
03:44:59 Quit mc2739 ("leaving")
03:51:24 Quit GeekShadow (Read error: 104 (Connection reset by peer))
03:51:39 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
03:56:12 Join T44 [0] (
03:56:13 Quit killan (Read error: 54 (Connection reset by peer))
04:08:28 Quit Rondom (Nick collision from services.)
04:08:38 Join Rondom [0] (
04:13:46 Quit Topy44 (Read error: 110 (Connection timed out))
04:14:06 Quit LambdaCalculus37 ("Leaving")
04:14:38 Join wincent [0] (
04:19:37 Quit Rob2223 ()
04:25:40 Part wincent ("Kopete 0.12.7 :")
04:26:30 Join Rob2222 [0] (
04:40:09 Quit AEnima1577 ("Leaving.")
04:43:09 Join Zorda_ [0] (
04:56:05 Quit Zorda (Read error: 110 (Connection timed out))
04:57:14 Join gtkspert_ [0] (
05:05:37 Quit gitster` (Read error: 145 (Connection timed out))
05:12:35 Quit gtkspert (Read error: 101 (Network is unreachable))
05:14:32***Saving seen data "./dancer.seen"
05:23:40 Join shai [0] (
05:33:52 Join AndyIL [0] (n=pasha_in@
05:37:21 Quit MethoS- (Remote closed the connection)
05:38:55 Join Tomis [0] (n=Tomis@
05:40:33Unhelpfuli failed @ maths. we can't do 9 levels of blending with SWAR, this will require 7 free bits between components. 5 will work fine, though.
05:45:28 Quit AndyI (Read error: 110 (Connection timed out))
05:45:41 Quit Strife89 ("Bed.")
05:48:15 Quit ps-auxw (Read error: 110 (Connection timed out))
05:56:05 Quit Tomis2 (Read error: 110 (Connection timed out))
06:20:11 Join AEnima1577 [0] (
06:36:46 Quit martian67 (Client Quit)
06:47:32 Join Tomis2 [0] (n=Tomis@
06:49:13 Quit Tomis (Read error: 104 (Connection reset by peer))
06:52:32 Quit gevaerts (Nick collision from services.)
06:52:44 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
07:04:58 Quit Tomis2 (Read error: 110 (Connection timed out))
07:14:36***Saving seen data "./dancer.seen"
07:19:41 Join Tomis [0] (n=Tomis@
07:26:27 Quit phanboy4 (Read error: 60 (Operation timed out))
07:39:10 Join Horschti [0] (
07:41:44 Quit AEnima1577 ("Leaving.")
07:42:59 Join AEnima1577 [0] (
07:46:25 Quit Horscht (Read error: 60 (Operation timed out))
08:04:11 Join Rob2223 [0] (
08:21:51 Quit Rob2222 (Connection timed out)
08:22:18 Join Bagder_ [0] (
08:26:31 Join robin0800 [0] (
08:41:46 Join jus__ [0] (
08:41:55 Quit jus__ (Read error: 54 (Connection reset by peer))
08:46:49Unhelpfulkugel, amiconn: another subpixel-aa font test, now with black-on-white or white-on-black:
08:50:25Unhelpfulthe AA'd samples are two different methods for reducing the 256-level glyphs down to 4 levels. the 4-level values are then mapped to a 5-level range, so that after mixing colors division can be done with a right-shift. i think the second sample, which favors full-on or full-off a bit more, looks slightly better... either is nice enough that i think i'll get to work on trying to actually do this in RB ;)
08:52:15 Join stoffel [0] (n=quassel@
08:52:43 Join Grahack [0] (
08:56:09 Join flydutch [0] (
08:58:05 Join FOAD_ [0] (
09:06:05 Quit FOAD (Read error: 145 (Connection timed out))
09:06:06 Nick FOAD_ is now known as FOAD (
09:14:37***Saving seen data "./dancer.seen"
09:17:14 Quit Grahack ("Leaving.")
09:24:47 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:29:57 Join n1s [0] (n=n1s@rockbox/developer/n1s)
09:33:51 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
09:35:39 Join stoffel_ [0] (
09:35:59 Quit Rob2223 ()
09:38:47 Quit stoffel (Read error: 145 (Connection timed out))
09:38:54 Join dmb [0] (n=Dmb@unaffiliated/dmb)
09:40:22 Join bertrik [0] (
09:43:20shaiI want to help with translations... but is it me, or does this look like a very length process?
09:44:04Bagder_well, you need to provide translations for a fair amount of phrases. that can be lengthy, yes
09:44:08Bagder_how else would it work?
09:46:33shaiWell... I do Gallery2 translations and its been setup in such a way, that translating is just as easy as going to Setting -> Start Translation (approx.) and it'll show you English text on the left, and a space on the right to translate. You put in the word/sentence/code on the right, and click save (or was it update?).
09:47:38pixelmashai: I think rasher's translation page could be a great help for you then (basically a web frontend). Which languag do you want to help translating?
09:48:39shaiI found lots of English text that needs translating and other Hebrew words that don't really say the correct meaning in the context of which they were meant for">
09:49:26shaiI saw that page... and just noticed that link to Rasher's page
09:49:52shaiAs for the "Translations status" saying 0 0 0 for Hebrew, that's way outdated... like it says "As of 26 Aug 2008""> says "4 months ago" for hebrew
09:51:09bertrikAFAIK there has been recent hebrew translation activity by tomers
09:51:55shaiYea, tomers and I know each other... small country.
09:51:58bertrikOctober 24th 2009 was the last hebrew update
09:52:48pixelma0 0 0 for Hebrew is from the wiki and that page isn't updated automatically, someone has to do it manually (and didn't for a while)
09:53:49pixelmaI think that even the status table on rasher's site is a bit outdated (or I missed something)
09:54:54shaiHow much text do you guys think I'll find to be hard-coded and untranslatable?
09:55:41Bagder_not very much
09:55:50Bagder_we have been translating rockbox for years
09:56:24shaiAnd if I do find and report it... would it get fixed quickly (in a reasonable time frame)?
09:57:31shaiI love QA ... don't get me wrong ... and tomers knows me and how much I do .. but if there is something I don't like in QA, is pushing on the gas in nutral.
09:58:31n1sreported bugs are fixed by people that decide they want to fix them, when that happens is anyone's guess
09:59:41Bagder_we have MANY open bugs
09:59:43 Join Rob2222 [0] (
10:01:14Unhelpfulthere's a fair bit in plugins, isn't there?
10:01:48Bagder_I'm sure, I haven't really looked that closely lately
10:02:05*Bagder_ is circling the project at high altitude
10:02:18pixelmasure, and that can't be translated yet
10:02:25Bagder_ah yes
10:02:27*pixelma looks around for midgey
10:06:09Unhelpfulwith 2-bit subpixels, glyph data should cost 6b/pixel... i could pack 5 per word and just waste the other 2, or i could just pack 2-bit subpixels 16 to a word. i think the wasted 2 bits is probably fine if dealing with whole pixels not being word-aligned would hurt performance?
10:07:13n1sUnhelpful: a lot of our open bugs are also hardware specific or hard to reproduce
10:07:41n1swe've had some success with tracker cleanup weeks in the past but the mod
10:07:56liari am trying to debug why rockbox freezes on my nano 2g when i connect an usb cable, it seems like the problem isnt in the usb thread, it seems there is another thread locking up and stopping all other threads from working, how could i debug this?
10:07:58n1ss/mod/most recent attempts have not done so much/
10:09:01 Quit stoffel_ (Read error: 113 (No route to host))
10:09:53Unhelpfulevery time it's suggested i go look for bugs, i find ones specific to platforms i don't have, or buried deep in code i don't know. somehow i doubt i'm the only developer, especially among the newer crop, who has found this a problem. :/
10:11:51shains1, I do suppose I could always bug tomers to push in my fixes ;)
10:12:06shaiOr rather fix requests ...
10:13:26Unhelpfulshould this *really* be a bug at all?
10:19:12Unhelpfulcouldn't we very easily set default foreground/background colors and a solid-fg-on-solid-bg drawmode to fix this? it's not like anything will care what the state of the gfx engine is after the panic is "done".
10:19:50 Quit robin0800 (Remote closed the connection)
10:21:38n1sUnhelpful: you mean 10665?
10:21:49n1sah, right
10:22:16n1s10665 sound seasy
10:22:48Unhelpfulthe hard part is making it panic ;)
10:29:05n1scan anyone with an e200 confirm 10381?
10:30:42Unhelpfulit already sets colors, actually :)
10:31:49Unhelpfulso add a lcd_set_drawmode(DRMODE_SOLID)?
10:35:26n1swho knows stuff sbout the fonts around here?
10:36:11Unhelpfuli've learned a *bit* working on aa-fonts. i might be able to help...
10:36:20n1sFS #9906 seems like a simple fix for a font error but i have nfi about those things
10:37:16Unhelpfulhrm, it looks like panicf disables backdrop, clears the screen, then sets foreground and background colors. it still might be worth adding a lcd_set_drawmode to make sure it's not in DRMODE_INVERVSEVID, but i'd say it looks like this bug is already dead.
10:38:06 Quit dmb ("Leaving")
10:38:36amiconnThere are actually two functions to consider: panicf() and UIE()
10:39:01amiconn10665 looks like an UIE() example
10:40:41Unhelpfulso UIE should do all of the same setup as panicf, essentially?
10:41:22Unhelpfulis there a reason it doesn't just use panicf?
10:52:15amiconnUIE() needs to be as self contained as possible
10:53:05amiconnThings have already gone very wrong when it gets called. The CPU might even be in a mode where not all memory is reachable
10:53:08 Quit elcan (Remote closed the connection)
10:54:44 Join ender` [0] (
10:55:07Unhelpfulah. so it might be a bad idea to go doing more in the way of graphics?
10:55:43Unhelpfulit still does lcd_clear_display, though, at least on ARM... that can hardly be *less* likely to work than setting the colors and drawmode
10:56:34pixelmamaybe clearing the backdrop, as I understood it this is the main problem </guess>
10:58:02shaiWho does documentation updates? I noticed that the iPOD Video (5th Gen) has a current screen capture of The Playlist Submenu ... but still talks about non-existing features such as Insert Last Shuffled and Queue Last Shuffled.
10:58:12shaiThe PDF version.
10:58:40 Join elcan [0] (
10:58:52pixelmanon-existent features? You mean on the screenshot?
10:59:10shaiI don't see them on the screen shot nor on my iPOD Video.
10:59:25shaiLet me double check myself.
10:59:59shaiAck. Yes.
11:00:19Unhelpfulpixelma: the report claims text over WPS, so i'm guessing that the backdrop clear was added after this report. clearing the display is probably fine, but it should, as panicf does, also set colors, and disable backdrop
11:00:20shaiThose two don't exist in the screen capture nor on my iPOD.
11:00:48Unhelpfuland maybe add setting the drawmode to DRMODE_SOLID to both UIE and panicf
11:01:04 Join teru [0] (
11:01:30pixelmashai: when did you last update your build? Those were added fairly recently and the manual is updated daily
11:01:44shaiToday (approx. 1 hour ago).
11:02:11shaiI installed the Current Build using the Rockbox Utility.
11:02:14shaiAs I always do ...
11:02:19terushai: are you selecting playlist file?
11:02:24 Join Tomers [0] (
11:02:29shaiNo. an MP3.
11:02:32shaiTomers, Morning :)
11:02:58Tomersmorning :-)
11:03:19shaiActually... noon ...
11:05:47teruIIUC, they are only shown when folder or pleylist file is selected.
11:05:48 Quit Bagder_ ("It is time to say moo")
11:06:29shaiI selected neither...
11:06:37pixelmadoesn't make much sense for a single file
11:06:40Tomersshai: Just port your new translation at FlySpray, and me or anyone else will commit, if appropriate.
11:07:02TomersI would love to have feedback on Hebrew translations, and will commit promptly.
11:07:24shaiDid you read the chat history?
11:07:47 Join Zarggg_ [0] (
11:08:06 Quit elcan (Remote closed the connection)
11:08:31Tomersshai: Also, working on manual stuff is highly appreciated. If you notice some screenshot changes, upload them to FS, hopefully with the same file name as used in the manual (why don't you pull RB code through SVN using TortoiseSVN in Windows)?
11:08:38shaipixelma, Understood. It does appear when selecting a folder (didn't try a playlist but I assume its the same).
11:08:38Tomersshai: Yes
11:09:22shaiTomers, So does your answer refelect the question about hard-coded text too?
11:09:24 Join Grahack [0] (
11:09:26pixelmaI thought TortoiseSVN can be troublesome
11:09:42n1sit can
11:10:03Tomerspixelma: 'read-only' mode - not for commit.
11:10:03shaiI personally don't like Windows for this... I'd rather use my Linux box for it...
11:10:08n1ssomething with the whole line ending mess
11:10:41Tomersshai: It will be much better to handle this in Linux. Windows is hell for this purpose - I just assumed you are using only Windows
11:11:04pixelmaTomers: I also read about people not able to compile with a TortoiseSVN setup, commit or not
11:11:14shaiTomers, and for my question on hard-codec text ... ?
11:11:30Tomersshai: Lately you've complained about having non-translatable string in WPS ("Next Song:", etc.). Why not bringing it up now? I don't know what are the technical barriers preventing this from being implemented, so we should ask
11:12:05shaiSo, its on the table...
11:12:16Tomerspixelma: Good to know. Thanks. I gave up compiling rbutil in Windows, but maybe this is not related to the fact I fetched the code in TortoiseSVN
11:13:12shaiTomers, I'm not 100% sure how much nor which text is really hard-codec ... but maybe you can fill in that gap.
11:13:25*gevaerts points Tomers and shai to r22837
11:13:34 Join elcan [0] (
11:13:42Tomersshai: Also, we should push the issue of having Hebrew characters being part of some common fonts in RB, such as 12-Adobe-Helvetica, which is the default font on many targets
11:13:45*shai goes to find the URL to add r22837 to ...
11:14:28shaiTomers, Agreed... that's on the table too ... having the need to update two placed before you can read Hebrew is a drag.
11:14:28Tomersshai: Regarding Hebrew in fonts, refer to">
11:14:41***Saving seen data "./dancer.seen"
11:14:53Tomersshai: I think rasher is the address regarding fonts (gevaerts: am I right?)
11:15:05gevaertsTomers: he does stats
11:15:20gevaertsI don't know if he also does fonts as such
11:15:50shaiHebrew is way behind...
11:16:03shaiTomers, 10x
11:16:32Tomersshai: Actually, I think Hebrew is in good condition. At least it is highly usable after some configuration (font, default codepage)
11:17:02shaiThe Log Message in that rev. is way to technical for me...
11:17:12TomersUnless you have critics on the quality of translation :-) I tried to translate as much as I could, but still tried to avoid over-translation
11:17:29Tomersof technical terms that have no proper translations
11:18:13shaiI saw a word there... that I would never put in Hebrew unless I really knew what I was doing ;)
11:18:26shaiThat word, is the main reason I'm here today.
11:18:39Tomersshai: I guess you mean גאייה which is the translation of fade-in ?
11:18:41shaiBut I'm up to an hour and still trying to get my head around the idea ...
11:18:58shaiגומר actually...
11:19:11shaiIt should have been מסיים
11:20:21shaiAnd in that context... it should have really been מעדכן
11:21:12shaiBecause you don't really know if its finishing the job... so the English original should also be altered (IMHO).
11:22:23Tomersshai: Bring the full string (english) here and ask people around - maybe they agree with you about changing this string. I'll fix the Hebrew translation
11:23:04Tomersgevaerts: Could you please save me some energy and briefly explain how can I use r22837 to translate, e.g. "Next Song:" ? AFAIU, what is needed to be done is changing the Cabbie theme to have that new tag instead of hard-coded string, then adding the translation in hebrew.lang?
11:23:58gevaertsTomers: I'd have to go and figure it out myself. I know it's there, but that's about it.
11:24:17Tomersgevaerts: OK. I'll dig it up myself. Thanks
11:24:36TomersNot for a new topic
11:24:39 Quit Zarggg (Read error: 110 (Connection timed out))
11:24:41shaiTomers, Is this the full translation page?">
11:24:53gevaertsTomers: if you have questions about it, try to catch JdGordon
11:25:58pixelmait currently uses strings in Rockbox that are already there (for use in menus etc.), I would not want to add strings to the language file just for a WPS, not sure if the default one should be an exception. I'm just a bit afraid that this will blow up the language files
11:27:17shaipixelma, was that meant to me?
11:27:40pixelmano, Tomers
11:28:12TomersI think it would be useful to have rbutil 'implement' some default setting depending on the language it uses. E.g. if rbutil is used on a computer where Hebrew is set as default language, and thus uses Hebrew GUI, it should install Rockbox on the DAP and (optionally ask the user for confirmation) set the DAP's language to Hebrew, and set some more related configurations, such as setting...
11:28:14Tomers...default codepage to Hebrew (I found Unicode to be somewhat unusable here, as most ID3 tags are non-Unicode, but that's another story). Also might set 12H/24H clock according to the OS setting where rbutil is running on, etc. What do you guys think?
11:30:01Tomerspixelma: I'm +1 for having translation for default theme. But I will understand the reasons why this might be wise. Does the mechanism that will be used to translate plugins could be used for translating themes?
11:30:54*gevaerts thinks that a set of common phrases should be acceptable
11:31:01pixelmaI don't know, it hasn't been committed
11:32:01Tomerspixelma: What hasn't been committed?
11:32:41pixelmathe plugin translation system
11:34:17Tomerspixelma: I now it hasn't. But could it be used to translate themes also? When it will be committed, of course?
11:36:07 Quit AlexP (Remote closed the connection)
11:36:14Tomersshai: I can't find that badly translated sting anywhere (גומר)
11:36:17 Join AlexP [0] (n=alex@rockbox/staff/AlexP)
11:38:58 Join flai [0] (
11:40:14 Quit flai ("")
11:41:13 Quit Tomers ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
11:41:32 Join tomers [0] (
11:43:05Unhelpfulhow does this look?
11:47:10 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
11:51:11*Unhelpful *thinks* he sees a way to implement partitioned multiply-accumulate, provided the multipliers are *very* small... but probably 3 mac instructions is cheaper than 8 adds + 4 ands, plus setup?
11:51:29amiconnUnhelpful: I don't think clearing the backdrop or setting the drawmode in UIE() would be a problem
11:51:31 Quit AaronM (Read error: 60 (Operation timed out))
11:51:40amiconnJust that UIE() shouldn
11:52:27amiconn't call potentially complex functions without reason, because it increases the chance that it breaks
11:54:42Unhelpfulamiconn: and clearing the backdrop is fairly simple, so that should be ok, it's just things like calling into panicf that we want to avoid?
11:56:05 Quit Grahack ("Leaving.")
12:08:36n1shmm, why are we ifdefing for HAVE_LCD_BITMAP in syste
12:08:49n1sm-foo (where foo != sh)?
12:09:32Unhelpfuln1s: in case things ever change? i can remove that if you like :)
12:10:24 Join stoffel [0] (
12:14:01 Quit BHSPitLappy (Read error: 60 (Operation timed out))
12:16:34n1sUnhelpful: less useless #ifdefsa are good imho :)
12:17:41Unhelpfulfine, i'll nuke those and commit, it compiles on every color or greyscale arm or coldfire, and on charcell or bitmap sh :)
12:18:00gevaertsI'm not sure...
12:18:19Unhelpfulare there mono targets aside from sh?
12:18:31pixelmathe M:Robe100
12:18:48gevaertsplease don't go too far with this...
12:18:49amiconnAnd the Clip
12:18:53pixelmaand m200 (port in progress)
12:19:12gevaertsand DAX
12:19:25Unhelpfulwith what?
12:19:50gevaertsremoving support for screen types that happen not to be used right now
12:21:16Unhelpfulthat's a vote for leaving the extra #ifdef?
12:22:03AlexPUnhelpful: logik dax, it has some code in svn
12:22:26pixelmaAlexP: sloow
12:22:51gevaertsI would leave them in, athough I wouldn't complain about the charcell-on-non-sh ones
12:23:03AlexPpixelma: no, I just misread
12:23:33AlexPpixelma: when gevaerts said DAX I thought he replied asking what it was
12:23:34*Unhelpful will remove just those, then :P
12:27:43CIA-5New commit by unhelpful (r23623): Set DRMODE_SOLID, uniform colors, and sysfont before clearing LCD to display panic or exception messages - see FS #10665.
12:31:22 Join kugel [0] (n=kugel@rockbox/developer/kugel)
12:37:17CIA-5New commit by unhelpful (r23624): Fix dangling #endif
12:39:59 Quit yosafbridge ("Coyote finally caught me")
12:40:10 Join yosafbridge [0] (
12:45:08kugelUnhelpful: I saw you're going to work on subpixel anti-aliasing?
12:45:11kugelawesome :)
12:45:25 Quit tomers (Read error: 113 (No route to host))
12:50:12n1scare to look at one of those funny scrolling-lines-keep-scrolling-after-exiting-screen bugs?
12:51:40n1seh that was to kugel, iirc you fixed some of them?
12:51:54 Join kugel_ [0] (
12:52:04 Quit kugel (Nick collision from services.)
12:52:10 Nick kugel_ is now known as kugel (
12:58:07 Join stoffel_ [0] (
12:59:11 Join Grahack [0] (
12:59:36 Quit kugel (Nick collision from services.)
12:59:41 Join kugel [0] (
13:01:21 Quit stoffel (Read error: 110 (Connection timed out))
13:02:05 Quit kugel (Remote closed the connection)
13:06:34n1sOMG, we have both "lcd_scroll_stop" and "lcd_stop_scroll" functions
13:11:46CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
13:11:46CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
13:14:44***Saving seen data "./dancer.seen"
13:19:25 Join bimbel [0] (n=Miranda@unaffiliated/bmbl)
13:27:27 Quit bmbl (Connection timed out)
13:29:07n1syay i fixed the bug but have no idea if it's correct at all
13:38:25n1sJdGordon: ping?
13:43:04 Quit Bob_C_ ("ChatZilla 0.9.85 [Firefox 3.5.4/20091016092926]")
13:50:53n1sif anyone who understands this gui/scrolling code reads the logs i posted a patch that fixes the bugs i fs#10616 but i'm not sure it's correct (it even feels wrong) so i'd be happy if someone else took a look at it
13:51:28n1ss/i fs/in fs/
13:52:30 Join Omlet [0] (i=omlet05@
13:56:56pixelmathere is also the bug that scrolling lines keep scrolling and on screen if you enter the USB screen (I believe there was a similar bug with the FM screen once, not sure how it was fixed but it was)
14:04:37 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
14:07:51 Join MethoS- [0] (n=clemens@
14:13:51 Join kugel [0] (n=kugel@rockbox/developer/kugel)
14:14:21kugeln1s: heh, welcome in to the scrolling confusion :)
14:14:23 Join einhirn [0] (
14:14:59kugelstop_scroll() stops all scrolling. that should be avoided since it also stops scrolling in unrelated parts (read: custom statusbar)
14:15:04 Join ps-auxw [0] (
14:15:55kugelscroll_stop() only stops scrollingin the viewport you pass. However, that must be the same viewport that was used for set_viewport() before putting the text
14:16:53n1skugel: one of those should really be renamed
14:18:09 Quit einhirn (Read error: 104 (Connection reset by peer))
14:18:26n1skugel: did you look at the patch? the thing i think is wrong with it is that i would think the menu drawing code should handle stopping scrolling lines
14:22:22 Join bughunter21 [0] (
14:22:22 Quit bughunter2 (Nick collision from services.)
14:22:35 Nick bughunter21 is now known as bughunter2 (
14:24:25 Join DerPapst [0] (
14:26:39 Join bughunter21 [0] (
14:26:39 Quit bughunter2 (Nick collision from services.)
14:26:50 Nick bughunter21 is now known as bughunter2 (
14:33:15 Quit MethoS- (Remote closed the connection)
14:41:55 Join killan [0] (
14:42:01 Quit AEnima1577 ("Leaving.")
14:44:31 Nick bimbel is now known as bmbl (n=Miranda@unaffiliated/bmbl)
14:45:51 Quit midgey (Read error: 60 (Operation timed out))
14:47:29 Join fdinel [0] (
14:47:46 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
15:05:08 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
15:05:58 Quit stoffel_ (Remote closed the connection)
15:07:34 Quit Omlet ("( :: NoNameScript 4.22 :: )")
15:14:47***Saving seen data "./dancer.seen"
15:20:53 Join midgey [0] (
15:23:10 Quit dfkt (Read error: 110 (Connection timed out))
15:29:14 Join kugel_ [0] (
15:29:28 Quit kugel (Nick collision from services.)
15:29:36 Nick kugel_ is now known as kugel (
15:32:58shaiCan I have the iPOD connected to the PC on one hand via USB so I can manipulate the files there (copy music on and off) while at the same time have it available to play my music?
15:36:55 Join dfkt [0] (
15:39:06 Quit Grahack ("Leaving.")
15:41:35 Join kugel_ [0] (
15:41:47 Quit kugel (Nick collision from services.)
15:41:55 Nick kugel_ is now known as kugel (
15:43:34 Quit antil33t (Read error: 54 (Connection reset by peer))
15:43:39 Join antil33t [0] (
15:44:36 Quit kugel (Nick collision from services.)
15:44:42 Join kugel [0] (
15:47:00 Quit fdinel (Read error: 110 (Connection timed out))
15:47:24shaiAlexP, too bad ...
15:47:32 Join mcuelenaere [0] (
15:52:18 Join Bob_C [0] (
15:52:22 Join funman [0] (n=fun@rockbox/developer/funman)
15:55:17 Quit dfkt_ (Read error: 110 (Connection timed out))
15:56:05 Quit kugel (Nick collision from services.)
15:56:11 Join kugel [0] (
15:58:54 Join Bagder_ [0] (
16:00:30 Quit kugel (Nick collision from services.)
16:00:36 Join kugel [0] (
16:01:55 Quit kugel (Nick collision from services.)
16:02:01 Join kugel [0] (
16:23:37 Quit teru ("Quit")
16:28:06 Join einhirn [0] (
16:32:24 Quit DerPapst ("Leaving.")
16:35:43 Quit bughunter2 ("Leaving.")
16:37:15 Join Highlander [0] (
16:40:01 Quit kugel (Read error: 110 (Connection timed out))
16:49:14 Join Omlet [0] (
16:49:20 Quit Omlet (Remote closed the connection)
17:01:27 Quit Bagder_ (Remote closed the connection)
17:14:51***Saving seen data "./dancer.seen"
17:18:27 Quit einhirn (Read error: 104 (Connection reset by peer))
17:19:06 Join kugel [0] (
17:27:27 Join DerPapst [0] (
17:30:50 Join Crunchie [0] (
17:32:30 Join Strife89 [0] (
17:32:55 Quit Zambezi (Read error: 60 (Operation timed out))
17:36:48 Join Tomis2 [0] (n=Tomis@
17:41:26 Quit Tomis (Read error: 145 (Connection timed out))
17:41:26 Nick Tomis2 is now known as Tomis (n=Tomis@
17:49:54 Join tomers [0] (
18:05:09 Quit Llorean (Read error: 104 (Connection reset by peer))
18:05:22 Quit bluebrother (Nick collision from services.)
18:05:23 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
18:13:24kkurbjunkugel: about those statusbar bugs: are you testing on a 32bit or 64 bit machine?
18:14:10kkurbjunok, I'm on 64 bit too, have you had a chance to try it against the mr500 sim?
18:14:16kugelno, not yet
18:14:30kugeldoesn't gdb give any info?
18:16:13 Join derekja [0] (
18:16:19 Part derekja
18:18:07kkurbjunkugel: strange, I am trying this in a later build and it's not segfaulting
18:18:31kkurbjunmaybe the segfault is not an issue
18:20:43 Quit dfkt (Excess Flood)
18:20:49 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
18:23:11Crunchiedoes video now work on the nano 2g? last time I checked the sound/video was out of sync
18:23:42 Join toffe82 [0] (
18:24:23funmancheck with different samples
18:25:01Tomisworks for me Crunchie
18:25:09Tomisi just watched a full movie on mine last night
18:25:15CrunchieI checked it with the sample elephant dreams
18:25:20Tomisonly took about 7% of my battery actually
18:25:34Tomisi was fully expecting it would take much more
18:25:41Crunchieoh okay will check out a new revision
18:25:51tomersI've noticed that most wps'es contains the string 'Next Track' while only CabbieV2 and DancePuffDuo has 'Next Song'. Is there a differentiation between 'Song' and 'Track' in wps, or is this simply an issue of lack of consistency?
18:31:28kkurbjunkugel: ok, I reproduced it on r: 23572 - I am going to see if I can pin down which revision fixed it
18:33:34 Quit kugel (Remote closed the connection)
18:35:08 Nick YPSY is now known as Ypsy (
18:40:16 Join Bagder_ [0] (
18:41:11Crunchiehmm it's still out of sync with the latest revision
18:41:59 Join Paleskin [0] (
18:42:07 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:42:11Crunchiebut oh well I'm glad I can run things now on my nano 2g :)
18:42:45Paleskinrockbox r23549-091107 on sansa fuze
18:42:59kkurbjunkugel: looks like bug 2 is no longer an issue since r23580
18:43:03Paleskinflac playback autopause during playback
18:43:10Paleskinand to resume need reboot
18:43:16Paleskinis this a known bug ?
18:43:20Tomisi used the 310kbps encode of elephants dream i think Crunchie
18:43:22kkurbjunI updated the tracker entry to reflect that
18:43:23Tomisseamed fine
18:43:34Tomisi didn't watch it all the way through though
18:44:05 Join Grahack [0] (
18:44:22Tomisi encoded a full movie at 150k video and 64k audio, about an hour and a half long
18:44:24kkurbjunI'm not sure aobut your assessment on bug 3 though - we must do some clearing of the screen before going into the wps, otherwise there would be artifacts of the list everywhere
18:44:36CrunchieI've used this
18:44:39Tomisit played fine
18:44:46Tomisyea, that's the one i usedd
18:46:16kkurbjunkugel: With the custom statusbar graphic glitches I have only been able to cleanly use it with the main menu, I have to disable it in the wps and re-produce the portion of the custom sbs in the wps that I would otherwise have done with a conditional in the sbs.
18:46:17kugelkkurbjun: the skin engine does this clear, but not as immediately as you might think
18:46:49kugelall my sbs are re-used in the wps
18:46:56kkurbjunit looks like it does the clear before the wps screen tests true
18:47:30kugelI doubt that
18:47:33kkurbjunby looks I mean just from testing it form a user side
18:48:00kkurbjunif you could try those tests it would be really helpful because I would like to understand how to use it properly
18:48:12kugelyea sure, once I have some time
18:48:27kkurbjunIf I can't get it working right with the conditionals I don't see how anyone else will unless they get lucky
18:50:00kugelwhat glitches are you experiencing exactly?
18:51:09kkurbjunit shows portions of the custom sbs that were supposed to only show in the lists. If those areas are not overdrawn with a viewport in the wps the artifacts show
18:51:30kkurbjunI use this tag: %?cs<%Vda| | | >
18:51:38kkurbjunwhere %vda should only show in the list
18:52:14kkurbjunand then I go into the wps it shows an artifact of the albumart positioning that was only supposed to show in the menu
18:52:56 Join panni_ [0] (
18:53:14Paleskinoh yeah, how do I upgrade from one build to another ? should I erase the rockbox folder on fuze, and let rockbox utilities do the rest ?
18:54:34Tomisyou can just download and extract the latest build to the device, having it replace existing files
18:54:37gevaertsPaleskin: no need to delete things. The new build will just overwrite things
18:55:13gevaertsPaleskin: have you checked the filesystem? Corrupted filesystems have been known to cause really strange behaviour
18:55:38kugelkkurbjun: why do you put a space there, btw?
18:56:33 Quit funman ("free(random());")
18:56:35kkurbjunI dunno, I just had it in tehre in the example - it gives the same glitches if I just do %?cs<%Vda>
18:57:52kugelI'll have a look, that glitches shouldn't happen of course
18:58:09 Join phanboy4 [0] (
18:58:50Paleskin@gevaerts, how do I check my filesystem ?
18:59:02Tomisdisk utility Paleskin
18:59:12kugel(oops) do you have something between %Vi and the folling %V[l]?
18:59:23Tomisoh sorry
18:59:25Tomiswrong room
18:59:42Paleskin@Tomis, control panel −−−−-> adm−−−−-> storage ?
18:59:59kkurbjunthis is the sbs I am seeing the glitches with:
19:00:02Tomisi was thinking i was in one of the mac channels, sorry
19:00:38PaleskinI'm using xp dual boot with 7
19:02:02Paleskinlet say If I deleted the rockbox folder first, then telling rockbox utilities to upgrade, so there aren't any leftover files, will something bad happens ?
19:02:12kkurbjunkugel: with that sbs I see a couple of graphic glitches, the left side of the wps backdrop seems to be from teh list backdrop, I have a small box that also appears to be from the list backdrop and then the album art from the list still shows
19:02:38Tomisi dont see why anything would be wrong with that Paleskin
19:02:45Tomisyou're just deleting files and writing new ones
19:03:14Paleskinshould I reinstall the boot loader ?
19:03:32kugelkkurbjun: basically, the problem is that the sbs can't just clear everything, it's not really proper if the wps does this either
19:04:08kugelif the sbs clears out everything, it could clear list/wps area. and the wps could clear sbs area
19:04:55kugelI'm not sure how feasible/possible doing it right is
19:05:02 Join nawks [0] (
19:05:13kkurbjunkugel, couldn't the sbs be re-drawn after the wps screen is entered and the screen is cleared?
19:05:50kugelIIRC that's being done. maybe it's not
19:08:05kugelyou could try inserting a send_event(GUI_EVENT_REFRESH, NULL) somewhere in the if (restore ...) block
19:08:52kkurbjunkugel: what file are you referring to? I am not familiar with the skin code at all
19:08:55kugelor better, in gui_wps_display() before the skin_redraw call
19:09:11kugelwps.c and skin_display.c
19:10:44kkurbjunhmm, I get compile errors with send_event(GUI_EVENT_REFRESH, NULL);
19:10:55kkurbjunit says that GUI_EVENT_REFRESH’ undeclared
19:11:52kugel#include "appevents.h"
19:12:58kkurbjun:), it gets rid of the sbs glitches, but nothing shows in the wps till I go to the next trach
19:13:35kkurbjunI just added that call to gui_wps_display
19:13:47kkurbjunright above return
19:14:55***Saving seen data "./dancer.seen"
19:17:19kkurbjunkugel, if I add it in the if(restore block it gets rid of the albumart glitch and the wps shows mostly correct, but the left side of the backdrop still appears to have the list backdrop used - it appears to be equal to the scrollbar width
19:17:27 Join MaadMan [0] (
19:17:36 Part Paleskin
19:17:52kugelyea, the wps backdrop is set in gui_wps_display, that's why
19:17:55kkurbjunand I still have an unexplicable box right next to the now playing area
19:18:30*kugel doesn't understand why the wps is be empty
19:24:06kugelkkurbjun: that doesn't happen here
19:24:18kugelah yea, I see why
19:24:28kugelbloody remotes :/
19:29:31Zorda_itunes to restore
19:29:33Zorda_what do i do
19:29:37Zorda_i cant get itunes :<
19:29:54Tomiswhy not
19:30:34Zorda_Because it dosnt run on this computer i alreaty told you tomers >.> isnt there another option?
19:30:59Tomisit only needs windows or a mac
19:31:03 Quit Prot (Read error: 113 (No route to host))
19:31:20Zorda_tomers it dosnt run on my computer i alreaty told you somewhere else
19:31:43Tomisif you need to restore to the factory state, you need itunes, or restore from a backup, if you made one before installing rockbox
19:31:46tomersZorda_: s/tomers/Tomis/ :-)
19:32:37Tomisdo you really want to go back to the factory firmware?
19:32:48Zorda_Tomis Do i have a choice?
19:32:52Tomisor is it just that's what the ipod is displaying on it's screen
19:33:16Zorda_thats what its saying.
19:33:45Tomisbut do you want go back to the factory firmware
19:33:49Tomisif you don't care
19:33:49Zorda_Not really.
19:34:01Tomisthen you can just blow the thing away
19:34:08Zorda_and put on rockbox?
19:34:12Tomisyea, exactly
19:34:24Zorda_cool thanks
19:34:26 Join hillshum [0] (
19:34:32Zorda_thats what i was wondering ifi could do
19:34:48Tomisyea it's just showing that message because something got messed up
19:35:03Tomissomething probably went wrong during your rockbox instal
19:35:23Tomisjust go through the process again
19:39:56kugelkkurbjun: I think I have a good patch soon
19:41:08 Quit phanboy4 (Read error: 104 (Connection reset by peer))
19:43:09 Join vodi [0] (
19:44:07Zorda_hmmm its in disc mode but ipodpatcher isnt identying it..
19:45:46Zorda_did i brick it? lol
19:46:21kugelkkurbjun: should do it
19:48:02 Quit jfc^3 (Read error: 104 (Connection reset by peer))
19:48:15kugelit doesn't seem to remove the list backdrop parts here, which is weird since clear_display() and lcd_update() is called
19:48:58 Part vodi
19:49:39 Join jfc [0] (
19:50:40kugelon a theme which didn't have this problem before :S
19:51:34 Join Zambezi [0] (i=Zulu@
19:52:53 Quit ender` (" First things first, but not necessarily in that order.")
19:56:54Zorda_i dindt backup
19:57:02Zorda_guess thats the end of my ipod
19:59:16Zorda_or i could do the obious thing and take my netbook to where i can get itunes
19:59:37Zorda_it still works as a flashdrive
19:59:38Tomisjust blow it away and resinstal rockbox
19:59:48Zorda_tried didnt work
20:00:16Tomiswhat ipod is it
20:00:28Zorda_nano second gen
20:00:39Tomisdo a manual instal
20:00:46Zorda_i did
20:01:08Tomishere, lemme find the instructions i used
20:01:31Tomisipodpatcher didnt like mine either
20:01:38Tomisi went with iloader and it works fine
20:02:28 Join ender` [0] (
20:02:44Tomisfollwo those isntructions for the bootloader instal
20:03:08Tomisthen follow rockbox's instructions for manual instal of rockbox, ignoring the rockbox bootloader instal
20:09:15liaron my nano 2g(with rockbox) there is a thread which locks up and stops everything from working. how could i find out which one it actually is?
20:17:00gevaertsZorda_: have a look at
20:17:19*gevaerts doesn't know why people say you need itunes for this
20:20:04 Join phanboy4 [0] (
20:20:57 Join domonoky1 [0] (
20:22:29Tomisbecause it's the manufacturer's way and it's easy gevaerts , that'd be my guess
20:23:10Tomisi'm surprised apple hasn't sent a c&d to the guy hosting the firmware files though
20:23:35gevaertsTomis: possibly, but then "if you need to restore to the factory state, you need itunes, or restore from a backup, if you made one before installing rockbox" is just spreading misinformation
20:24:22Tomisyea i forgot about the manual way to restoring it
20:27:49liargevaerts: but what about the FTL? is that saved on the flash where the files are saved or anywhere else?
20:28:26kugelkkurbjun: did you try that patch?
20:30:47Zorda_gevaerts it dosnt have my model
20:30:53 Quit MaadMan ("Leaving")
20:31:03Zorda_could i just go with first gen nano?
20:32:04Tomisit does Zorda_
20:32:23gevaertsliar: the FTL sites between the flash and the filesystem
20:32:44Tomisyou were talkin abiut the list of firmwares at the linked page?
20:33:16gevaertsZorda_: no. They're different. If you're really stuck, I'd recommend waiting aroud for TheSeven (although I really have no idea when he will be online again)
20:37:03liargevaerts: what i mean is: if you put a backup back on your flash the ftl will be restored too?
20:37:46gevaertsliar: I don't think so. I think it needs to be reset explicitely, but I don't really know (no 2nd gen nano here)
20:38:09liargevaerts: yeah so the itunes way is more save
20:38:12 Join stoffel [0] (
20:38:16 Quit domonoky (Read error: 110 (Connection timed out))
20:47:40 Quit stoffel (Remote closed the connection)
20:49:16 Quit thegeek (Read error: 104 (Connection reset by peer))
20:50:08 Quit Tomis ()
20:53:51 Join Tomis [0] (n=Tomis@
21:07:34CIA-5New commit by kugel (r23625): Fix peakmeter from delaying the initial display of the wps. Also change HZ/100 to TIMEOUT_NOBLOCK which is more appropriater in the get_action() call ...
21:07:39 Quit kugel (Remote closed the connection)
21:08:03 Join kugel [0] (n=kugel@rockbox/developer/kugel)
21:08:32tomerskugel: Can you please give me feedback on "FS #10783 - WPS translation" ?
21:09:35kugelwhat kind of feedback? I'm for it, if you mean that :)
21:10:15kugelnot sure if I like %Sr, but that's minor anyway
21:10:20gevaertstomers: definitely commit the "next track" bit!
21:10:42tomerskugel: Thanks. Just want to ask several developers first...
21:11:08tomerskugel, gevaerts: I would rename it to whatever you think appropriate
21:12:31tomersgevaerts: There are two other phrases - 'Next', and 'of'. I think they can also go in - just a total three strings for this purpose is cheap :-)
21:13:03kkurbjunkugel, that patch does appear to fix the conditional viewport problem, the left size backdrop issue still seems to be there
21:13:16tomersRegarding the lang file, I use 'user: core' and also the asterisk for the phrases. Is it OK?
21:13:20gevaertstomers: I mean FS #10782. I don't have an opinion yet on the others
21:13:24kugelkkurbjun: yea, I can't explain that
21:14:02kkurbjunthere's still that little box that is clipped too
21:14:07tomersgevaerts: I will commit FS #10782 now....
21:14:20kugelI suspect a bug somewhere else. a clear/update entire display really shouldn't leave that over
21:14:58***Saving seen data "./dancer.seen"
21:15:00kkurbjunI did find that it doesn't show if I disable the custom statusbar when going into the wps screen with %wd
21:15:13kkurbjunthe left side glitch that is
21:15:28kkurbjunit doesn't show with the built-in statusbar either
21:15:50kugelwell, the patch does that combo
21:16:27kugelyou aren't unfamiliar with the lcd drivers, any idea how that could leave parts of the display untouched?
21:16:57kkurbjunThe onlything I can figure is that a clear was called before the new backdrop is set
21:17:04CIA-5New commit by tomers (r23626): FS #10782 - WPS: Rename 'Next Song' to 'Next Track' ...
21:17:38kugelkkurbjun: I tried explicit calls to display->clear_display() and display->update() too. that changed nothing except giving the expected flicker
21:17:40kkurbjunthe backdrop is just done with a pointer and the lcd routines use that when doing any operations with fonts, transparencies, and clearing
21:18:38 Quit linuxguy3 (Remote closed the connection)
21:18:49 Join linuxguy3 [0] (
21:20:23kkurbjunit would have to be something where that portion of the screen is drawn when the backdrop is set to the list backdrop rather than the wps
21:23:31gevaertstomers: I think I like FS #10783, with two caveats : (a) I have no opinion on WPS tag names (see other people for that), and (b) maybe MrSomeone should do a survey on the theme site to find a list of common strings, so we can actually support more than just cabbie (of course the commit doesn't necessarily have to wait for this)
21:23:53kkurbjunkugel: do you know when lcd_set_backdrop is called with respect to clearing the screen?
21:24:21kkurbjunand is there a chance that it is called pointing to the list backdrop after the wps is entered?
21:25:13kugelah I see what you mean
21:25:21kugelthe backdrop must be set before clearing
21:26:27 Quit Grahack (Read error: 113 (No route to host))
21:26:37kugelkkurbjun: that works for me now
21:28:36kkurbjunkugel, nice that is working properly now
21:30:22 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
21:30:39kkurbjunI think I found what is happening to the touchscreen buttons - it looks like they are being offset when the custom sbs is loaded and enabled on the wps screen
21:30:51kkurbjunthey are shifted down from the positions specified
21:31:24kkurbjunI'm not sure why that is happening though with respect to the code implementation
21:31:56 Join Grahack [0] (
21:32:04kkurbjunoh, I wonder if it has to do with the UI viewport setting
21:32:15kkurbjunit looks like they are slightly offset down too
21:32:30kkurbjunand it seems to be approximately equivalent to the UI viewport setting
21:32:40kugelarent the coordinates absolute?
21:33:01kkurbjunthey should be
21:33:44kkurbjunsorry the second offset note should read right rather than down
21:33:50kugelon a related note, someone should look into changing the touch button handling, so that the sb can have touchregions
21:34:06kkurbjunthey are offset down and to the right about the same amount as the ui viewport offset from the top left of the screen
21:34:33kkurbjunthat would be nice
21:36:00kkurbjunit also looks like the buttons that are defined outside of the ui viewport are not able to be pressed
21:36:53kkurbjunyeah, with the slider if I slide it to the right I can only use the area that overlaps with the ui viewport setting
21:37:57kugelI can't see where the code is wrong
21:38:25kugelah well I can
21:39:06 Quit Bagder_ ("It is time to say moo")
21:39:08kugelJdGordon recently made a change so that the default viewport of the wps is (potentially) a lot smaller
21:39:44kugelthe touchregion code natuarlly checks if the press is within the viewport the touchregion is defined in
21:40:12kugelif that happens to be the default viewport, the touchregion might be capped
21:40:52kugelI don't consider this wrong, you should use a special viewport for the touchregion instead
21:41:04kugel(that's the only thing I can imagine right now)
21:41:18kkurbjunwhich viewport should be used for the touchregion?
21:41:29kkurbjunI mean what do I need to do in the wps/sbs?
21:41:47tomersgevaerts: Thanks for reviewing! I don't know who to ask regarding the name of the skin tag... Any idea?
21:41:49kugeldefine the touchregion after a viewport you have control over
21:42:06kugelor rather "more control"
21:43:26saratogagot the new mdct to at least compile in rockbox, now to make it actually work
21:43:50kugelthat wouldn't explain why it adds an offset though, and I assume the volume bar isfully drawn too (that would be strange since that also exceeds the viewport)
21:44:10 Quit Strife89 ("Leaving")
21:44:27kkurbjunSo right now it doesn't look like the ui viewport is used when you are in the list view with respect to the touchscreen handling, and then when you are in a wps it is used in some strange way
21:44:58tomersgevaerts: Maybe I should commit, and later rename if someone has a better suggestion?
21:45:08tomersbecause I thikn it is ready
21:45:11kkurbjunwell the volume slider is displayed in its own user defined viewport
21:45:18 Quit Tomis ()
21:45:24kugelput the touchregion there then :)
21:45:46*kugel isn't sure anymore whether the coordinates are display or viewport relative, I would assume the latter
21:45:50gevaertstomers: I wouldn't commit if there's a good chance it has to be renamed
21:46:04pixelmatomers: what name of the skin tag? What is it used for?
21:46:11kkurbjunyeah, I try do that as a workaround, but it seems like the logic is broken to me
21:46:19gevaertstomers: on the other hand, only one person has objected so far, so maybe he can propose an alternative :)
21:47:11kugelkkurbjun: maybe it's clearer when you read the dev ml thread
21:47:29kkurbjunI read that thread between you and jdgordon
21:48:10kugelthe default viewport of the wps uses the ui vp, I don't see how that's broken logic (I argued against that but both logics are reasonable)
21:48:51kugelthat one was never guaranteed to be fullscreen actually
21:49:08kkurbjunhuh - what is the thinking for having the wps use the ui vp for the default?
21:49:27kugelask JdGordon :)
21:50:24kugelanyway, this vp can't be fullscreen or it gets in the way of statusbars (classic and custom)
21:50:32 Join maffe [0] (
21:50:46kkurbjunok, I'll ask him about that
21:50:52tomerspixelma: - New conditional skin token 'Sr' which checks whether the currently loaded language is an RTL language (Hebrew/Arabic). It uses lang_is_rtl().
21:51:17kkurbjunso it looks like the graphic glitches are fixed at least on the wps - when I'm in the list view and I hit resume I still get a glitch after the noting to resume message times out
21:51:29kugelif %T is viewport relative, then having it in the default vp was always wrong (no matter of UI vp or sbs)
21:51:49kkurbjunkugel, but the ui viewport might be dynamic in a sense - like I was trying to implement
21:52:04 Join Tomis [0] (n=Tomis@
21:52:09kkurbjunso that you could have a status area while you are in the lists that goes away when you go to the wps
21:52:20kkurbjunI mean the ui viewport doesn't change size
21:52:30kkurbjunbut the area that I would want it to take effect is less
21:52:46kkurbjunactually, in the wps I would not want the ui viewport used at all
21:52:57kugelyou can do that, but you can't rely on the default wps viewport then
21:52:59kkurbjunbut I guess that's a matter of debate
21:53:19kkurbjunyeah, I gotcha it makes more sense now that I understand what it's doing
21:53:26CIA-5New commit by tomers (r23627): Updated Hebrew language
21:53:47kkurbjunI think that the behavior is confusing if you don't know what the underlying code is doing though.
21:54:20kugelI guess you can imagine what bad things happen if the default wps viewport overlaps with sbs (that also happens with specified viewports but that's considered as user fault)
21:55:27kugelJdGordon wanted to get rid of having 2 default setups (the %Vi and the UI vp one)
21:55:43tomersgevaerts,pixelma: I won't commit "FS #10783 - WPS translation" now. I'll let it cook until tomorrow....
21:55:48*tomers Good night
21:56:27kkurbjunkugel: I don't understand having two setups though - doesn't the %Vi tag do the same thing that the cfg one in some sense?
21:56:31 Quit tomers ("ChatZilla 0.9.85 [Firefox 3.5.4/20091028153816]")
21:56:52kkurbjunit seems like allowing the user to define one outside of the theme is a good way to break things as this touchscreen stuff shows
21:57:10kugelthe code didn't quite do what it suggested to do, there was set_fullscreen() which was %Vi or really fullscreen minus the area used by the classic sb. then there was _set_defaults which was UI vp or the result of set_fullscreen()
21:58:39kugelnow there's just set_fullsscreen which is really fullscreen and set_defaults which is the UI vp intersected with %Vi (or fullscreen if neither is set)
21:59:28kkurbjunso %Vi doesn't specify the UI viewport?
21:59:53kugelit's the viewport that doesn't overlap with the sbs
22:00:10kkurbjunI'm not using the UI viewport then I was using the wrong terminology in the above messages
22:00:34kugelso the wps' default viewport is as big as %Vi
22:01:01kugelif you want to overlap that because you know what you're doing, then you need to specify viewports
22:01:04kkurbjunbut that's different from the UI viewport?
22:02:07*kugel thinks the better way of doing is to handle all this stuff within the sbs
22:02:33kugelyou seem to be deactivating parts of the sbs in favor of filling that space from the wps?
22:03:06kkurbjunso consider this case. The wps disables the sbs when you go into it because it wants the default viewprot to be the fullscreen expecting to set touchscreen positions relative to the fullscreen. then the user sets a ui viewport setting with a cfg - what happens to the touchscreen areas?
22:03:22kkurbjunyeah, that's what I was doing
22:04:38kugelwhy don't you do that in sbs?
22:05:19kkurbjundo what?
22:05:38kugelfill that space from the sbs instead of from the wps
22:05:49kkurbjunthe album art is larger in the wps
22:06:03kkurbjunI give a smaller area for album art in the list
22:06:19kkurbjunand then when you go to the wps I only want to show the time section of the sbs
22:06:28kkurbjunwith that conditional setting
22:07:06kkurbjunthen then the expectation I have is that you can use the area that does not have an active viewport in the sbs
22:07:20kkurbjunwhich works now with that patch
22:07:24kugelyea, that's how I did it originally
22:07:29kkurbjunbut the touchscreen stuff gets all offset
22:07:39kkurbjunwhich is fine if you know what your doing
22:07:51kugelyou really shouldn't be having touchregions in the default wps viewport
22:08:00kkurbjunbut the concern I have is with the cfg ui viewport setting
22:08:24kugelhaving them in a specified viewport is not only a work around, it's the only way to get a guaranteed vp
22:09:16kugelif you want a fullscreen one you can always put a "%V|0|0|-|-|-|-|-|" at the very beginning
22:10:17kkurbjunit worked fine before with the standard statusbar, the other concern I have is how do you effectively communicate this to a theme author - after talking to you and jdgordon I have an understanding of how this all fits together, but it's extremely confusing/looks like a bugs if you don't know what the code is doing
22:12:13kkurbjunyeah, the %V|0|0|-|-|-|-|-| tag worked
22:12:19 Join Jaykay [0] (
22:12:20kugelmy initial commit didn't change that behavior, if you had %wd you got it very fullscreen. JdGordon changed that
22:12:46kugelbut, if you don't have a UI vp, and %wd and still don't get fullscreen now, it's a bug
22:13:57kkurbjunkugel, what if the custom sbs doesn't specify a %vi tag though and the wps enables the statusbar and you have a cfg setting for the ui viewport
22:14:30kugelthe UI vp is used for the wps
22:14:32kkurbjunthat would generally be an invalid case I would agree, but there's nothing stopping an author from doing that to my knowledgte
22:15:06kugelnot having %Vi easily leads to problems, I don't disagree here. I think CustomWPS is explicit about that
22:15:11kkurbjunand that breaks the theme if they use the default viewport for the touch offsets
22:15:26kugelthat's bound to be broken, really
22:16:00kugelI consider such a theme badly coded no matter of ui vp/sbs/whatever
22:16:12kkurbjunI bet I could make a pretty usable sbs that doesn't specify the %vi tag
22:16:24kugelkkurbjun: that's why it isn't required ;)
22:17:18kkurbjunand if I didn't know better there would be nothing stopping me, but when you throw the ui viewport cfg in the mix it can start breaking all kinds of things in the wps depending on how it is made
22:17:58kugelwe obviously can't make everybody happy
22:18:32kugelwhat we have now certainly gives the best freedom and good fall back mechanisms
22:18:55kkurbjunI guess my point is it seems that the %vi tag should be the only thing that changes the display - that way the theme/skin author can make sure that everything plays nice
22:19:13kugeland ignore the ui vp?
22:19:25kkurbjunwhen you let the end user that didn't have a hand in the creation of the theme or skin set those things it's setup for confusion
22:19:36kugeldon't confuse %Vi with making things look nice. it's the tag to make an sbs portable
22:23:25kkurbjunI guess I would count myself confused :-D
22:23:49kugelis somewhere mentioned that %Vi is the UI viewport?
22:24:15kkurbjun"This viewport is used as Custom UI Viewport in case the theme doesn't have a ui viewport set in the theme.cfg"
22:24:40kugelyea, that's correct
22:24:51kkurbjunso %vi is the ui viewport?
22:25:05kugelonly " in case the theme doesn't have a ui viewport set in the theme.cfg"
22:25:25*gevaerts thinks that that line is actually meaningless
22:25:39kugeland it's not something you can rely one, since you can interchange wps's and sbs's
22:25:48kkurbjunyeah.. that makes sense so the theme can set it in the cfg or the sbs
22:25:52gevaertsThe actual used viewport is the intersection between UI viewport and %Vi, right?
22:26:00kugelmaybe it should be worded to make clear that it's used as a fallback rather
22:26:04gevaertshm, wait, this is about WPS
22:26:08*gevaerts doesn't know then
22:26:55kkurbjunbut the user could also set it in their cfg independent of the theme, sbs, or wps
22:27:15kugelkkurbjun: %Vi is designed to make SBSs work with themes that aren't designed for it, to not cause redraw problems
22:27:15gevaertsyes, but that doesn't mean it's a good idea
22:27:43kugelif you want to align stuff to the backdrop, then that's unrelated to the sbs, then the ui viewport setting should be used
22:27:57gevaertsoh wait, it's not wps, it's sbs...
22:28:14*gevaerts thinks that the %Vi description is misleading
22:28:38kkurbjunbut %Vi wouldn't accomplish that with a wps that really used viewports to their full extent
22:29:33kugelof course not
22:29:39kugelthe wps needs to make sure itself
22:30:08kugelby using %wd for example (although that doesn't protect against the UI viewport)
22:30:10gevaertskugel: if the UI viewport is not set, the UI behaves as if the UI viewport is set to fullscreen, right? If so, isn't the effect *always* that the used viewport is the intersection of %Vi and UI viewport?
22:30:22kkurbjunso what is the expectation from the user side - it sounds like the intent is for sbs's to be portable, but they, in reality are not if you use viewports in the wps
22:30:37 Join toffe82_ [0] (
22:30:37kugelgevaerts: sounds about right
22:30:55kugelkkurbjun: I can't follow that conclusion
22:31:27*gevaerts thinks that in an sbs world a wps shouldn't enable the status bar
22:32:41kkurbjunso consider all the themes that use the %we tag because of the builtin statusbar and it uses viewports in the wps - then they load up a custom sbs that sets a %vi tag to try and be "portable" which doesn't really accomplish anything
22:32:50pixelmaisn't the biggest problem when you don't use own viewports in the WPS and rely on the default viewport?
22:33:14pixelmai.e. draw stuff in it
22:33:25kkurbjunthe only way that %vi is portable is if you exclusively rely on the default viewport
22:33:41kkurbjunI mean have a wps that exclusively relies on it
22:34:05kkurbjunor have a wps that disables the statusbar
22:34:07kugelkkurbjun: that's the wps' fault, not the sbs one
22:34:11gevaertsnot really, because the wps might also rely on a certain minimum size
22:34:18 Join bomber227 [0] (
22:34:31kugelof course we can't have *everything* just work
22:34:41kkurbjungevaerts: how do you mean?
22:35:30 Quit shai ("Leaving")
22:35:43gevaertskkurbjun: in the old days, the status bar always had the same size so the wps could take that into account. In the %Vi days, the left-over space can easily be too small to show the wps in any sensible way
22:35:57gevaertsand the wps author can't really do anything about that
22:36:04bomber227i finally got my sansa fuze in the mail and now have a clipv2 thats not in use. now i was wondering if there's any way i could help with the clipv2 developement? i can't think of how but id like to help anyways
22:38:23kkurbjunkugel: I see the point of having the flexibility from the developer/skin author standpoint, but it puts alot of burden on the end user when we offer menus for selecting your custom sbs and wps in the same menu screen
22:38:36kugelI disagree
22:39:29kkurbjunand mixing/matching them does not work right - how are they supposed to understand the intricacy of what they just did to their interface
22:39:49saratogabomber227: hows your c programming?
22:40:27bomber2270. i have started learning BASIC, though....
22:40:30*kugel thinks kkurbjun is too user centric ;)
22:40:42kugeldon't forget our ideal to not care about them :)
22:40:44gevaertsbasically the current way can work with independent backdrop+UI viewport (i.e. the "list theme"), sbs, and wps, *as long as* the wps does not show the status bar. I think that that's sufficient to warrant the flexibility, but yes, it would be nice to make wps+random sbs work
22:41:12kkurbjunif we pulled those menu items and only allowed these to be set from a "theme" it would allow all that flexibility while preventing the user from making all kinds of bad combinations
22:41:42kugel"preventing the user from making all kinds of bad combinations" is simply impossible
22:41:49kugeland we don't need that
22:41:56saratogabomber227: well without the ability to program developing isn't going to be possible
22:42:13kkurbjunwell it would at least make it more difficult for a novice user to destroy the usability of their player
22:42:15saratogai suppose if you're not using the player you could see if any developers were interested in working on it
22:42:19gevaertsThe user has been able to pick bad combinations forever, with much worse results than are possible with the sbs+wps mess
22:42:29bomber227yeah. i was thinking maybe testing or soemthing
22:42:32 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
22:43:16 Quit kugel (Remote closed the connection)
22:43:32kkurbjunwhen I say user I'm talking about the person that loads up a bunch of themes on their player and starts messing with the theme menu - I'm not talking about someone developing a new theme
22:44:14kkurbjun /loads up a bunch of themes with rbutil/ after installing rockbox that is
22:44:14gevaertskkurbjun: how can they destroy the usability of their player? I mean, they might have an unreadable wps if they pick a bad combination, but they can still get out, and return to the settings menu. That's *a lot* better than what happens if they pick bad colour combinations
22:44:51kkurbjunif you have a %vi setting and a cfg setting that only overlap with one pixel that would be a touch interface to navigate
22:45:06kkurbjuntough that is
22:45:36gevaertshm, true. That could be solved by having some sort of minimum and refusing to load the sbs
22:46:59 Quit toffe82 (Read error: 110 (Connection timed out))
22:47:12bomber227alright well i guess i cant help
22:47:28kkurbjunI'm sure there are other ways to solve it, I'm just saying that it's a potential nightmare for the end user. only allowing them to set these through the cfg seems like a good way to give the most flexibility while protecting the user to some degree.
22:47:53 Quit Jaykay ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
22:48:32gevaertsit definitely protects them, but should it? There are dozens of other settings that can have bad effects if combined in certain ways, should we block all of those?
22:48:51kkurbjunwhat settings are you referring to?
22:49:51gevaertsthere are some UI things (backlight off, or colour combinations), there's the EQ that can't handle too many bands on slow CPUs, there's the traditional party mode support issue, ...
22:50:06 Quit bomber227 ("CGI:IRC")
22:53:51kkurbjunin the case of the EQ it seems to make sense that the number of bands should be limited if the cpu cannot keep up with it - generally I thought backlight off is disabled on targets that you can't physically read the screen - colour combinations might make sense to be in the theme
22:54:31gevaertsin the EQ you can't limit it, because the amount of available CPU also depends on the codec used. You might be able to do five bands with flac, but not with ape
22:55:03saratogaon PP a better solution would be to move the EQ to the second core
22:55:13saratogathen you'd probably be ok with everything but maybe MP3
22:55:14gevaertsI think it's reasonable to do some things such as a guaranteed minimum viewport (maybe just implemented by just falling back to built-in everything, although fancier ways might be doable), so that the user can always get back to the settings menu, but I really don't like the idea of just dropping nearly all theme related settings
22:55:30gevaertssaratoga: yes, on PP :)
22:55:51saratogaare there other targets too slow for 5 band EQ on most formats (APE excluded)?
22:57:01gevaertsI mean, picking a theme you like and then changing the menu backdrop should be possible. It can cause bad combinations, but I'd expect this to be a reasonably common thing to do for people who can handle computers but don't want to edit theme .cfgs by hand
22:57:03pixelmaI thought colour combinations are checked and not settable but the problem would be colours on backdrops
22:58:39pixelmaonly making settings possible in a package through themes doesn't necessarily protect against errors - it could be that the theme author did a mistake
22:59:50kkurbjunpixelma: I agree, the theme author could have messed up but I am counting them more in the power-user category they at least know about the different file types and know what to change
23:00:26gevaertssaratoga: from looking at CodecPerformanceComparison I suspect that AAC might get in trouble as well, but my point was really that trying to set always safe limits is a bad idea
23:00:42kkurbjungevaerts: I agree that could be a possibility, but I don't know where the target is for reasonable
23:00:43saratogaAAC on what?
23:01:45gevaertssaratoga: on coldfire it seems to be between 115% and 160%. 115% doesn't really sound enough to add five bands of EQ
23:02:50gevaertskkurbjun: "there's enough room to find a menu". That could mean enough room for two lines of 22 characters, and compile with HAVE_LCD_CHARCELLS :)
23:02:53saratogaaccording to that page its about 60MHz, which leaves another 65MHz for EQ
23:03:02saratogai don't think our EQ is that slow
23:03:08saratogathough AAC+ is another story
23:03:14saratogaand one that i hope to get to soonish
23:03:35kkurbjungevaerts: that also depends on the font size..
23:03:53saratogaactually i'm hoping to get some AAC benchmarks from the new IMDCT lib this weekend
23:03:53 Quit Grahack ("Leaving.")
23:04:05kkurbjunand that would fail on the mr500 remote in any case
23:04:09gevaertskkurbjun: true, but the "use built-in everything" should handle that
23:04:16saratogai think coldfire performance will hopefully get a lot better, particularly if I can get someone to give me a few lines of ASM
23:04:30gevaertssaratoga: I was looking at the last set of results on the page
23:04:42saratogathe first set are the most current i think
23:04:51gevaertsah indeed
23:05:07saratogathe last set are a couple days before I put the new MDCT into AAC :)
23:05:21gevaertsanyway, if you can handle AAC with five band EQ, I'll just add timestretch and compression :)
23:05:36saratogaare those features slow?
23:05:45gevaertsI don't know!
23:06:14gevaertsI'm just pointing out that there can be lots of "bad" combinations of settings that we can't easily detect and block
23:06:32saratogai think with some more optimization we could have most/all SWCODEC able to use full DSP effects on all sane formats
23:06:57gevaertsBut there's no law that forces people to use sane formats
23:07:15saratogatrue, but the insane ones probably won't play real time at all
23:08:29saratogai just realized i have no idea how to turn on eq
23:08:30gevaertsAPE tends to have a just-better-than-realtime profile for most targets. It's not always the same one of course
23:09:27saratogai'm benchmarking EQ right now
23:11:34saratogaabout 5MHz per EQ band
23:11:38saratoga(on PP)
23:11:46saratogaso 5 bands is just 25MHz
23:13:12gevaertsnot too bad
23:13:29saratogaso basically PP should be able to use 5 EQ bands for all formats which it can decode in realtime except probably ATRAC3
23:14:16saratogastill putting them on the second core would give a nice battery life boost
23:14:38saratogai should ask Blue_Dude hard that would be to do
23:15:01***Saving seen data "./dancer.seen"
23:15:11gevaertsit would definitely be nice
23:15:35 Join thegeek [0] (
23:17:17gevaertsthere's also the eternal ipod video and its relatively slow LCD update that would really benefit from this
23:21:17*amiconn hands saratoga a 'how'
23:33:36CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
23:35:05 Quit gevaerts (Nick collision from services.)
23:35:14 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
23:36:14 Quit linuxstb (Read error: 60 (Operation timed out))
23:36:49 Quit Highlander ("Quitte")
23:39:44 Quit bmbl ("Bye!")
23:48:06 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
23:58:02 Join Quintasan [0] (n=quintasa@ubuntu/member/quintasan)
23:58:58QuintasanHiho, I'd like to package RBUtil for Ubuntu, on what license is rbutil released, I'd like to know if it uses any other licenses for some files

Previous day | Next day