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 2010-06-17

00:00:21 Join hamish_ [0] (~hamish@
00:00:34 Quit hamish_ (Client Quit)
00:05:04***Saving seen data "./dancer.seen"
00:13:35 Quit petur (Quit: Zzzzz)
00:17:05 Quit efyx (Remote host closed the connection)
00:27:42 Quit pamaury (Read error: Connection reset by peer)
00:30:18 Quit literal (Quit: leaving)
00:30:32 Quit CGL (Quit:
00:32:34 Join literal [0] (
00:33:01 Quit literal (Client Quit)
00:33:04 Join literal [0] (
00:44:32 Join Zigtown [0] (
00:48:29 Quit bertrik (Quit: De groeten)
00:49:18 Quit bieber (Ping timeout: 265 seconds)
01:02:05 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
01:04:22 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
01:09:26 Nick advcomp2019__ is now known as advcomp2019 (~advcomp20@unaffiliated/advcomp2019)
01:12:29 Join bluebrother [0] (
01:12:29 Quit bluebrother (Changing host)
01:12:29 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
01:15:57 Quit bluebroth3r (Ping timeout: 260 seconds)
01:18:38 Quit dfkt|x (Remote host closed the connection)
01:19:12 Join robin0800 [0] (~quassel@
01:19:59 Quit DerPapst (Quit: Leaving.)
01:22:00 Quit voRia (Quit: Leaving.)
01:26:17 Join Transformer [0] (
01:26:29 Quit Transformer (Excess Flood)
01:27:14 Quit robin0800 (Remote host closed the connection)
01:32:53 Join robin0800 [0] (
01:38:24 Quit hebz0rl (Quit: Ex-Chat)
01:40:03 Part toffe82
01:46:33 Join Buckwheat [0] (~chatzilla@
01:47:01 Join ubuntu-nathan [0] (ubuntu-nat@
01:47:03ubuntu-nathanHi all!
01:47:05*ubuntu-nathan is back
01:47:24Buckwheathey ubuntu nathan
01:47:30ubuntu-nathanI've succesfully done the cross-compiler
01:47:33BuckwheatI talked to you yesterday
01:47:40ubuntu-nathanI remember
01:47:47ubuntu-nathandownloaded the sources of rb
01:48:16ubuntu-nathanadded the sources directory to the includes directory
01:48:38ubuntu-nathanbut I want to build only my test source
01:48:44ubuntu-nathannot the entire rockbox
01:48:48ubuntu-nathanhow could I do this?
01:51:13ubuntu-nathanlike, there is any special parater that is needed?
01:55:22 Quit robin0800 (Read error: Connection reset by peer)
02:01:25 Quit GeekShadow (Quit: The cake is a lie !)
02:05:07***Saving seen data "./dancer.seen"
02:08:34 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
02:20:44 Quit MethoS- (Read error: Connection reset by peer)
02:24:13ubuntu-nathanI'm having problems to register at the forums
02:24:20ubuntu-nathanon the image I just see a k
02:24:41ubuntu-nathanand I always get an error because I've typed the image code incorect
02:24:50ubuntu-nathanwhat's wrong?
02:28:56S_a_i_n_tjust try the "re-captcha" button until you get one you can see properly
02:29:16S_a_i_n_tcaptcha's are designed to be a little hard to read.
02:29:47S_a_i_n_tIt stops machines reading them and making spam entries etc.
02:31:26ubuntu-nathannow I've done it
02:40:10CIA-85New commit by bieber (r26874): Theme Editor: Began working on skin preview viewer
02:41:57CIA-85r26874 build result: 36 errors, 0 warnings (bieber committed)
02:52:28 Join Antibuddha [0] (
02:52:41Antibuddhawhat happened to that special bootloader for 240gb ipods?
02:52:46Antibuddhait is no longer on forums
02:53:42saratogait was merged with the main bootloader
02:54:22ubuntu-nathannice to see another dev here
02:54:33Antibuddhado i have to let itunes fix my ipod, or can i format ipod to fat32 and then use new bootloader
02:55:05saratogayou can format without itunes, but its complicated
02:55:09saratogatheres a wiki page with instructions
02:55:10Antibuddhamy ipod says Error reading partition table
02:55:15Antibuddhaok i'll get itunes again
02:55:35Antibuddhai formatted it to fat32 but cant get rockbox to install
02:55:56 Quit BHSPitMonkey (Ping timeout: 265 seconds)
02:56:13Antibuddharestoring now :D
03:14:56 Join fdinel [0] (
03:16:53 Join bieber [0] (
03:19:36 Quit bieber (Remote host closed the connection)
03:19:56 Join bieber [0] (
03:29:29 Quit ubuntu-nathan (Read error: Connection reset by peer)
03:33:48Antibuddhaim transferring files at 3.6MB/sec
03:33:58Antibuddhais there a usb setting for ipod 5g to make it go faster
03:34:09Buckwheatantibuddha: there might be
03:34:22BuckwheatIll look in the forms
03:34:38Antibuddhacool, thank you
03:34:43Buckwheatno :)
03:34:46Antibuddhai get a big headache trying to find anything on forums
03:34:48Antibuddhaokay ty
03:34:55Buckwheatno problem*
03:35:11Buckwheatthats what ctrl-f is fore
03:35:12S_a_i_n_tUm, no...there isn't.
03:35:19S_a_i_n_tUSB speed on RB is what it is.
03:35:53Buckwheattry using the origional firmware to transfer files
03:36:53Buckwheatantibuddha: i didnt find anything sorry
03:37:13Antibuddhawill just leave it on all day to transfer
03:37:37S_a_i_n_tIt can pay to use the OF.
03:37:51S_a_i_n_tfor very large transfers.
03:37:57Buckwheatthats what i do
03:38:01Buckwheatsaves alot of time
03:38:08Antibuddhahow do i run ipod in OF? hold menu?
03:38:14Antibuddhaand navigate button
03:41:20Buckwheatanybody know how to boot the fuze in the of?
03:41:41Buckwheati cant find where to do that anywhere
03:42:22 Join powell14ski_ [0] (
03:42:37 Quit panni_ (Read error: Connection reset by peer)
03:43:47Buckwheatantibuddha: for the sansa you push the left button while turning it on maby its the same with ipod
03:44:29Antibuddhai just read online
03:44:33Antibuddhaits so confusing
03:44:45Buckwheatwhts confusing
03:44:48 Join CGL [0] (~CGL@
03:44:49 Quit steve|m (Ping timeout: 276 seconds)
03:44:52Antibuddhathe steps
03:45:08Buckwheatjust hold the left button when powering
03:45:34 Nick Strife1989 is now known as Strife89 (
03:45:46Antibuddhamore steps for ipod
03:46:22 Quit krazykit (Quit: Connection reset by beer)
03:46:39 Join krazykit [0] (~kkit@
03:49:15 Quit Antibuddha (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
03:53:36 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
03:56:27 Quit Buckwheat (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
03:58:22 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
03:58:25 Join steve|m [0] (
04:01:03 Join Buckwheat [0] (~chatzilla@
04:01:58BuckwheatHas any body else have the data abort error on sasna v2 when trying to read a txt file
04:02:18S_a_i_n_tForums, Flyspray...yes.
04:02:33Buckwheatis it fixable
04:02:54 Quit Zigtown (Remote host closed the connection)
04:04:00S_a_i_n_tIt doesn't affect all targets as far as I know, it's mentioned on the forums, and has a flyspray task.
04:04:14S_a_i_n_tIf there was a "quick fix" for it, it would be in current SVN already.
04:04:20Buckwheatand flyspray is?
04:04:49S_a_i_n_tOne of the things you need to check before asking suck questions?
04:05:09***Saving seen data "./dancer.seen"
04:05:11S_a_i_n_tFlyspray is the bug tracker.
04:05:17S_a_i_n_t(linked above)
04:06:13Buckwheatwhere is it linked?
04:06:47 Join obo [0] (~obo@rockbox/developer/obo)
04:06:48BuckwheatI feel stupid now...
04:11:01 Quit BradC (Ping timeout: 276 seconds)
04:16:06 Join cpn [0] (
04:17:58 Join Barahir [0] (
04:18:24 Quit cpn (Client Quit)
04:18:29 Join cpn [0] (
04:20:19 Quit cpn (Client Quit)
04:21:24 Quit Barahir_ (Ping timeout: 265 seconds)
04:30:04JdGordonbieber: hehe just tried it and it is very wierd :P
04:30:07JdGordonmaybe undo it
04:32:09 Join GRUBaDubDub_ [0] (
04:36:11 Quit GRUBaDubDub_ (Client Quit)
04:36:36 Join Zigtown [0] (
04:36:52 Join anewuser [0] (anewuser@unaffiliated/anewuser)
04:38:05 Quit pixelma (Disconnected by services)
04:38:05 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:38:19 Quit amiconn (Disconnected by services)
04:38:20 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:38:22 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:38:39 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:40:06 Quit leavittx (Read error: Connection timed out)
04:41:18 Join leavittx [0] (~leavittx@
04:48:24 Quit t0rc (Quit: Leaving)
04:58:18 Quit Zigtown (Remote host closed the connection)
04:58:29 Quit Buckwheat (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
04:58:51 Join funman [0] (~fun@rockbox/developer/funman)
05:02:15 Quit TheSeven (Ping timeout: 240 seconds)
05:05:59 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:06:36 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
05:25:41 Quit BHSPitMonkey (Ping timeout: 265 seconds)
05:37:21 Quit Horscht (Quit: Verlassend)
05:39:03 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
05:47:27 Join JdGordon1 [0] (
05:47:38 Quit JdGordon (Ping timeout: 240 seconds)
05:49:35 Quit BHSPitMonkey (Remote host closed the connection)
06:02:12 Nick JdGordon1 is now known as JdGordon (
06:04:57 Quit curio_ (Remote host closed the connection)
06:05:13***Saving seen data "./dancer.seen"
06:06:19 Join Dark_Rak3r [0] (
06:25:28 Join Zigtown [0] (
06:38:32 Join robin0800 [0] (~quassel@
06:42:38bieberJdGordon: Maybe I'll just add a checkbox to turn it on/off
06:43:02JdGordoni tinhk its odd that it selects way too much
06:43:21JdGordonand it puts it into multiselect mode?
06:43:24JdGordonor is that a bug?
06:43:51bieberNo, it puts it into multiselect so it can select everything on the line
06:44:46bieberDo you think it would be better with just the first element on the line selected?
06:45:11JdGordonwouldnt it be possible to actually select the element that the cursor is in?
06:46:00bieberNot without some more parser changes, I'm afraid
06:46:20JdGordondo you know much about .pro files?
06:46:32bieberAbout as much as I need to to make it compile
06:46:50JdGordonI've pulled the parser out into a seperate folder wit its own Makefile, but not having much luck getting your .pro to then call that Makefile
06:47:09JdGordonI've got it so it depends on the lib and links to in, but not to actually build the lib
06:47:26bieberHmm, I'm sure how you'd do that
06:47:34bieberI know it supports linking in libraries, lemme Google it
06:47:56JdGordonLIBS += -lskin_parser -L../../libs/skin_parser
06:48:13JdGordonis pretty much what I added, but that doesnt know how to build skin_parser.a
06:51:41bieberI don't know if it has any capacity to call another Makefile. bluebrother might have done something like that with RBUtil, if it can be done
06:53:27 Quit robin0800 (Remote host closed the connection)
06:54:30*JdGordon has the core stripped of the old display, debug and parse code
06:54:35JdGordonnow for the fun part...
06:54:45 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
06:57:44S_a_i_n_tIf the skin syntax is changing anymore in the near future, perhaps checkwps could get some love?
06:58:26S_a_i_n_tThere are a couple of themes that pass checkwps, but have blatantly wrong syntax, and don;t load on target.
06:58:37S_a_i_n_t(How that works I have no idea)
06:59:10 Quit flatrose (Remote host closed the connection)
07:00:00JdGordonyes checkwps will be fixed
07:00:11JdGordonI dont know how a theme can pass it and fail on target
07:00:58JdGordonit uses the exact same code after all...
07:02:09S_a_i_n_tOne in particular is the "Electric Bars of Colour" Theme.
07:02:57S_a_i_n_tThe original theme added x,y coordinates to the %Xd line (why, I have no it loaded in the old syntax, I have no idea either.)
07:03:20S_a_i_n_tSkinupdater converted the line to "%X(backdrop.bmp)0%|0%|" and it passes checkwps
07:03:25JdGordonoh thats easy
07:03:29S_a_i_n_tyet fails on device.
07:03:43JdGordonthe old parser just eats everything at the end of the %X line
07:03:52JdGordonso you could put any old garbage there
07:04:22S_a_i_n_tbut...why does it pass checkwps as is now?
07:04:23JdGordonthat might have been changed to not do that anymore, and bieber's parser definetly doesnt allow that
07:04:44bieberJdGordon: Do we have a wiki page for all the assorted rendering rules yet, or should I set one up?
07:04:49bieberI'm getting started on rendering now :)
07:05:21S_a_i_n_tI'm going to need to make a list of themes that failed, and painstakingly open them up and see why they failed.
07:05:23JdGordonthat outlines how the current one works, have a look at utils/newparser/skin_render to see how I'm doing it with your parser
07:05:33S_a_i_n_tBut, I'm not really looking forward to doing so.
07:13:35JdGordonbieber: the important rules are, 1) disable viewports are never rendered/scanned/whatever, 2) the default viewport is *always* scanned, but nothing is drawn unless it is the only viewport
07:14:19bieberDisable viewports are named viewports that haven't been loaded?
07:15:02bieberAnd if I have some stuff at the beginning of a file, and then a viewport declaration, the stuff at the beginning won't be drawn?
07:15:02JdGordonyeah, they are created by default disabled, you need to use the %Vd() to enable it
07:15:35bieberHeh, having viewports at the top level is going to help a lot here
07:16:15funmanit seems on AMSv2 the µSD content is refreshed by the OF each time we write to the rtc wakeup register
07:16:18S_a_i_n_tThat's different behaviour than it not?
07:16:41S_a_i_n_tre: default viewport, and not drawing in it if there are other viewports declared.
07:17:33JdGordonno, that is what happens now
07:17:45JdGordonfunman: so fix usb so we dont have to load the OF :D
07:18:06funmanJdGordon: you must scream on pamaury!
07:18:27JdGordondamn namespace collisions :(
07:18:39JdGordonskip_whitespace() is in the parser and misc.c and completly different
07:18:48funmanprefix them
07:18:54JdGordonwhich is why I want to depend on the .a instead of compiling seperatly
07:19:03bieberIs it generally considered a good idea to break the sources files for an application up into separate directories? I'm a little leery of ../'s in my #includes, but my file count is starting to get out of hand
07:19:31 Quit scorche` (Read error: Connection reset by peer)
07:20:04JdGordonyou can fix the include paths to include those directories so you dont need to add the ../'s
07:20:19bieberHow do I do that?
07:20:32funmanin the makefile usually
07:21:19JdGordonthe INCLUDEPATH variab;le in the .pro
07:24:51bieberOkay, so will I just need to add each subdirectory to that? And I assume the project root directory is included by default? Eh, I guess I'm about to find out
07:25:19JdGordonI would assume so
07:25:44bieberAnd will SVN figure out that things have been moved around, or do I need to give it an mv command or something similar?
07:26:01JdGordonsvn mv
07:26:07JdGordonsvn is stupid :)
07:27:05 Quit Kamyk (Read error: Operation timed out)
07:28:24bluebrotherbieber, JdGordon: you need to set QMAKE_EXTRA_TARGETS, PRE_TARGETDEPS and a <targetname>.commands in the project file.
07:28:41CIA-85New commit by funman (r26875): AMSv2 RTC: no need to write to RTC_WAKEUP register ...
07:28:49bluebrothersee and following
07:28:56JdGordoncool thanks
07:29:56 Quit bucko_ (Ping timeout: 276 seconds)
07:30:13 Join bucko [0] (~bucko@
07:30:50 Join Kamyk [0] (
07:31:46 Join LinusN [0] (linus@rockbox/developer/LinusN)
07:31:58funmanof course when i want to test RTC it always works, it's only buggy when i'm going to bed and just want to use it
07:32:56CIA-85r26875 build result: 95 errors, 0 warnings (funman committed)
07:33:30 Join scorche [0] (~scorche@rockbox/administrator/scorche)
07:33:58funmanit's not me!! i swear!
07:35:37S_a_i_n_tJdGordon: What defines "drawing"?
07:35:56JdGordonputting stuff on the display
07:36:02S_a_i_n_tI'm pretty sure my themes call bitmaps in the default viewprt...
07:36:03bluebrotherwho is frederico-dannya? Looks like that machine has problems
07:36:06JdGordonso images and text
07:36:14S_a_i_n_tI'll try find an example.
07:36:21bieberAlright, setting INCLUDEPATH did it
07:36:26bieberNo more morass of files
07:37:03CIA-85New commit by bieber (r26876): Theme Editor: Moved source files into subdirectories
07:38:07bieberThanks everyone
07:38:26S_a_i_n_tYeah...I'm definitly able to draw in the default viewprt before a viewport declaration.
07:38:35S_a_i_n_tUnless I totally misunderstand it.
07:39:03CIA-85r26876 build result: 21 errors, 0 warnings (bieber committed)
07:40:25 Join einhirn [0] (
07:41:26S_a_i_n_tJdGordon: Is this an example of "drawing in the default viewport"?
07:41:28funmancan we force this client out ?
07:41:39*S_a_i_n_t thinks so, but isn't quite sure
07:43:07JdGordonbluebrother: do I need to add the LIBS line to the .pro also?
07:43:11S_a_i_n_tplaybar, volume, animations, repeat, shuffle etc. AA I think. All draw fine, and are before any viewport declaration.
07:44:40JdGordonthat is bad skin code
07:45:20S_a_i_n_tbad code, or bad that it works when it apparently shouldn;t?
07:45:32JdGordonboth, but the first mostly
07:46:03pixelmaI'm confused about the info about drawing in the default viewport I read here - one time it is ok, the other it is not
07:46:13S_a_i_n_tthat's why I was confused about the "not drawing in the default viewport if another viewport is called", as IFAIK it has never been this way.
07:46:34S_a_i_n_tpixelma: Me too...this theme has always worked for me :/
07:46:50S_a_i_n_tAnd I admit, the code *is* nasty, but it works fine.
07:47:47 Join pondlife [0] (~Steve@rockbox/developer/pondlife)
07:47:57 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
07:48:26JdGordonit shouldnt be working....
07:48:45S_a_i_n_tthis is what confuses me, it always has.
07:48:55S_a_i_n_tthe skin is about a year old now too...
07:49:32JdGordonthere is a reason the skin engine is getting a massive rework :)
07:49:38JdGordonnot *just* to annoy everyone
07:49:50S_a_i_n_tI see a great many other skins that do this also, and from the DL/vote counts...I assume they work also.
07:50:04funmanso you mean, there are additional reasons, but it's also a bit to annoy everyone?
07:50:19*funman knew it !
07:50:20pixelmaIIRC there are lots of themes on the theme site that draw in the default viewport (and use viewports). Almost every time someone posts his code in the forums to get help there's that
07:50:38S_a_i_n_tpixelma: Yes, you're quite correct.
07:50:42S_a_i_n_tI see it all the time.
07:51:14S_a_i_n_tOne of the many reasons I thought drawing in the default viewport was allowed.
07:51:17 Quit bmbl (Ping timeout: 258 seconds)
07:51:26S_a_i_n_tprimarily though, as it works fine.
07:53:10S_a_i_n_tIh that "Symmetry" theme, I've only used viewports for text, every image is drawn in the default viewport. Knowing now that this "doesn't work" makes me wonder why it does work, and how long it will be untill 90% of the themes on the themesite break.
07:53:12*JdGordon was going to point funman to but sees he replied to that thread already anyway
07:53:54JdGordonthey shouldnt work... unless I'm looking at the wrong version of the file (which is entirely possible)
07:54:46S_a_i_n_tI remember you mentioning the change re: viewports AGES ago...but I assumed it never happened.
07:54:52pixelmaJdGordon: I think that post would have been better by lots if you had explained the reasoning
07:55:08 Quit funman (Quit: free(random());)
07:55:53JdGordonbluebrother: I'm building my .a into the OBJECT_DIR, but Im getting lots of undefined references so it isnt being linked? am I missing something? LIBS in the Makefile looks like it is correct.. it has the $(SUBLIBS) which I assume qt adds...?
07:56:01 Join Buschel [0] (
07:56:22JdGordonpixelma: I'm 99% sure I have explained my reasons eariler in the thread, I also dont feel i need to explain it to someone who talks like that
07:58:39JdGordonbluebrother: ah, I missed the LIBS += line in
08:05:14***Saving seen data "./dancer.seen"
08:05:15 Quit krazykit (Ping timeout: 240 seconds)
08:12:58 Join krazykit [0] (~kkit@
08:19:57JdGordonbluebrother: sorry, last one, how do I hook up make clean for it?
08:20:32JdGordon is what I've got
08:28:47 Join Zagor [0] (bjst@rockbox/developer/Zagor)
08:31:52 Join mitk [0] (~mitk@
08:35:51S_a_i_n_tJdGordon: Is "-" an acceptable value for FG?BG colour in the new syntax?
08:35:54 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
08:36:46 Join ender` [0] (
08:39:29 Quit TheSeven (Ping timeout: 276 seconds)
08:39:42JdGordonshould be
08:40:40 Join flydutch [0] (
08:41:01 Quit TheSeven|Mobile (Quit: TheSeven|Mobile)
08:41:25 Join Dhraakellian [0] (
08:41:54 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:43:46 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
08:45:28 Quit Zigtown (Remote host closed the connection)
08:52:04CIA-85New commit by jdgordon (r26877): Move the skin parser to a seperate library
08:52:36 Quit TheSeven|Mobile (Read error: Connection reset by peer)
08:53:02JdGordonbieber: svn up :)
08:53:43CIA-85r26877 build result: 20 errors, 0 warnings (jdgordon committed)
08:54:07JdGordonoh shutup CIA-85 :)
08:54:59 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
08:55:09bieberDid you get distclean working?
08:56:10JdGordonno, but that isnt a major problem
08:56:21JdGordonI imagine its a one line fix
08:56:54 Quit Buschel (Ping timeout: 260 seconds)
08:59:48CIA-85New commit by bieber (r26878): Theme Editor: Working on renderer infrastructure
09:00:32bieberSVN will still have kept keywords while moving files around, right?
09:01:27 Join petur [0] (d408b802@rockbox/developer/petur)
09:01:31CIA-85r26878 build result: 24 errors, 0 warnings (bieber committed)
09:02:59 Join Rob2223 [0] (
09:03:32 Quit JdGordon (Ping timeout: 276 seconds)
09:03:46 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:04:07 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
09:05:44amiconnSomeone *please* disable this "frederico-dannya" build box
09:06:30 Quit Rob2222 (Ping timeout: 240 seconds)
09:08:20 Part LinusN
09:10:21 Join B4gder [0] (~daniel@rockbox/developer/bagder)
09:12:37S_a_i_n_tCan someone refresh my menory...sublines can't be put inside a conditional statement can they? ie. "%?Xx<%Xx|%t(1.0)%s%alBlah;%t(1.0)%s%alBlah>"
09:12:41 Join LinusN [0] (linus@rockbox/developer/LinusN)
09:12:58S_a_i_n_tOr can they. it *looks* right, only I don;t remember.
09:14:36JdGordonin the new parser they can be in conditionals, in svn they cant (IIRC)
09:14:37GodEatertry it in the sim and see?
09:15:11S_a_i_n_tGodEater: The sim is pretty bloody useless for theme development ATM
09:15:21S_a_i_n_t(which is why I asked here)
09:15:52 Join mk429_ [0] (
09:16:18JdGordonit depends what you want with the theme... if you want it to work in svn then the sim is fine... if you want it to work in the futuire (hopefully soon) then use the editor
09:16:42JdGordoneither way use the editor to get the syntax correct (except for the few cases like above)
09:17:43S_a_i_n_tIt's alright...viewprtifying this theme reminded my why I took the ugly root in the first place.
09:17:56S_a_i_n_tI knew there was a reason I did it, only I didn;t remember why.
09:18:01 Quit mk429 (Ping timeout: 264 seconds)
09:18:23S_a_i_n_tIt was all a big hack to avoid a huge mess with a bajillion viewports.conditionals.
09:19:29S_a_i_n_tSublines in a conditional is definitely something I'll appreciate when the time comes however.
09:21:34JdGordoncan gcc to 2 way linking with .a's? i.e can the .a have a function call which is in the binary which gets linked against the .a?
09:21:42 Join efyx [0] (
09:23:45JdGordonbieber: I just checked the dependacy logic... if oyu change a file in the lib and do a mae it will rebuild the .a so no real need for make clean to work, so :)
09:23:58bieberOkay, good
09:24:14 Quit rob (Quit: ( :: NoNameScript 4.22 :: ))
09:24:43 Quit bertrik (Quit: De groeten)
09:24:45 Join r0b- [0] (
09:31:35*JdGordon feels like the buffer management code shouldnt be in the libskin_parser.a
09:33:54 Join Teutonick [0] (
09:34:24Teutonickhow about rockbox for clip+ ? when it will be done?
09:34:26CIA-85New commit by bieber (r26879): Theme Editor: Stopped combo boxes in configuration editor from scrolling on mouse-wheel
09:35:24 Quit Kitar|st (Ping timeout: 260 seconds)
09:35:40 Join Kitar|st [0] (
09:35:59CIA-85r26879 build result: 27 errors, 0 warnings (bieber committed)
09:38:36 Join Buschel [0] (
09:39:36 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
09:40:08 Quit Kitar|st (Ping timeout: 252 seconds)
09:40:32 Part Zagor
09:40:41 Join Zagor [0] (bjst@rockbox/developer/Zagor)
09:43:37 Join BradC [0] (
09:45:58 Join Kitar|st [0] (Kitar_st@
09:48:51JdGordondid anything happen with getting the build system to built rbutil at devcon?
09:49:17JdGordonassuming no, can we add utils/themeeditor to the folder which dont tirgger a build?
09:49:59S_a_i_n_tTeutonick: Rockbox is already available for the Clip+
09:50:11S_a_i_n_tit is just not classed as "Stable" yet IIUC
09:51:24S_a_i_n_tWhich means that currently it isn't supported by RButil, but you are free to do a manual installation.
09:51:41 Quit TheSeven|Mobile (Ping timeout: 252 seconds)
09:53:33TeutonickS_a_i_n_t: thanx! but is it still buggy ? or it is relis ?
09:54:11S_a_i_n_tIt is perfectly usable, even "Stable" targets can contain bugs.
09:55:05 Quit liar (Quit: ...)
09:56:55CIA-85New commit by jdgordon (r26880): make the parser slightly more usable for rockbox, move the buffer allocation into the lib (maybe not the best spot?)
09:57:13 Join hamish_ [0] (~hamish@
09:58:06 Part hamish_
09:58:19CIA-85r26880 build result: 15 errors, 0 warnings (jdgordon committed)
09:59:08 Quit CGL (Remote host closed the connection)
10:00:21 Quit togetic (Read error: Operation timed out)
10:00:45 Quit Teutonick (Quit: CGI:IRC)
10:05:18***Saving seen data "./dancer.seen"
10:06:23 Join togetic [0] (~togetic@unaffiliated/ibuffy)
10:13:59 Quit Buschel (Ping timeout: 272 seconds)
10:14:42 Part pondlife
10:15:29 Join swilde [0] (
10:18:24CIA-85New commit by gevaerts (r26881): block frederico-dannya build client
10:20:12 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:28:58 Join dfkt [0] (dfkt@unaffiliated/dfkt)
10:30:40 Join pyro_maniac [0] (
10:32:58JdGordongevaerts: Zagor: busy? make files driving me mad :p
10:34:49n1sJdGordon: did you forget to add the skin_buffer.c file in the last commit?
10:35:01n1ssvn add, even
10:35:02JdGordonvery probably... is it not there?
10:35:17B4gderr26880 build result: 15 errors, 0 warnings (jdgordon committed)
10:35:33 Join DerPapst [0] (
10:35:37*B4gder serves as echo
10:35:45ZagorB4gder: we need a new client blacklist entry
10:35:48*JdGordon assumed they were broken build machine still...
10:36:30CIA-85New commit by jdgordon (r26882): woops, forgot to add these
10:36:54JdGordonwhich they were :) that lib isnt being built anyway
10:37:15JdGordonnot yet... I need help getting its makefile run by non bootloader builds
10:37:32JdGordonI assumed it would be added to apps.make but doesnt look like it?
10:37:48*JdGordon is drowing in root.make
10:37:53CIA-85r26882 build result: 12 errors, 0 warnings (jdgordon committed)
10:38:04gevaertsJdGordon: sorry, yes
10:38:10gevaertsZagor: I added that one
10:38:19JdGordonah crud.. dinner
10:38:19gevaertsor not?
10:42:21amiconnJdGordon: Yes, ld can do that, but you need to put the .a on the command line several times (ld seems to never go backwards in the list of objects to link)
10:42:57Zagorgevaerts: my fault, I had added an entry without committing which caused a merge conflict
10:45:39CIA-85New commit by zagor (r26883): Blocked fnarfbargle-uk-BradC
10:47:45S_a_i_n_trasher: Are the sims here ("> ) "post-skinbreak" sims?
10:54:32ZagorJdGordon: what is your make issue?
10:57:23 Join dfkt|x [0] (
10:57:23 Quit dfkt|x (Changing host)
10:57:23 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
10:59:35JdGordonZagor: I've got the Makefile in lib/skin_parser working now, I just have no idea how to get that into the main build
11:00:19JdGordonamiconn: ok, yuck :p I'll leave the buffer allocing code in the lib for the time being then
11:02:44 Quit dfkt|x (Remote host closed the connection)
11:17:14BradCOh dear..
11:18:43BradCCIA-85 New commit by zagor (r26883): Blocked fnarfbargle-uk-BradC - That one should have been down since its first broken build. Has it been committing sin since ?
11:19:55 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
11:20:10ZagorBradC: no, it was just a local change that I committed now
11:20:24ZagorBradC: have you fixed the client? should I remove the block?
11:21:10BradCI have not fixed the client yet (massive hard disk corruption) but I've completely disabled it and prevented it running, so yes if you can remove the block I'll make sure it's working properly before I try and bring it back up.
11:31:11 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:35:02 Quit dfkt|x (Remote host closed the connection)
11:41:04 Join halmi [0] (
11:41:44 Quit halmi (Client Quit)
11:42:06 Join halmi [0] (
11:43:40JdGordonZagor: gevaerts: ok, do I need to change my Makefile to be more like firmware.make and then just include it in the apps and checkwps build in root.make?
11:44:30ZagorJdGordon: I'll take a look now
11:44:37JdGordonor better to make apps.make include it?
11:45:48JdGordonI assume the CC= and AR= lines in my Makefile are just wrong
11:46:05Zagoryes :)
11:46:23*JdGordon isnt a complete dimwit then :D
11:56:08S_a_i_n_tJdGordon: Are you able to tell me why " %t(5.0)%?mv(0.5)<%ac%pvdB|%?mm<%ac%pvdB|%s%al%Sx(Repeat): %Sx(All)|%s%al%Sx(Repeat): %Sx(One)|%s%al%Sx(Repeat): %Sx(Shuffle)|%s%al%Sx(Repeat): %Sx(A-B)>> " doesn't work?
11:56:41JdGordonnothing obvious... hang on
11:56:47S_a_i_n_twhen %t(5.0)%?mv(0.5)<%ac%pvdB|Blah> does.
11:57:39JdGordonin the sim? or target? or checkwps?
11:57:48S_a_i_n_tliterally, "blah" does. I want it to alternate between volume in dB and Shuffle Mode, but drop to volume if the volume is changed while displaying Shuffle mode.
11:57:57S_a_i_n_t(it sounds more complicated than it is)
11:58:13S_a_i_n_tIn the sim, and on target.
11:58:31S_a_i_n_tIt parses, but it doesn't display the volume when it's being changed.
11:58:36S_a_i_n_tOnly the Shuffle Mode.
11:58:57S_a_i_n_tbut, if I replace the shuffle conditional with "blah" it works fine.
11:59:45JdGordonno clue... it will once the new parser happens :p (bet you're sick of hearing that)
11:59:57S_a_i_n_tIndeed I am.
12:00:02JdGordontough :D
12:00:15*S_a_i_n_t throws a new parser at JdGordon
12:01:24S_a_i_n_tThe only way I seem to be able to do it is with an unnecessary conditional viewport...but I was quite proud of the fact I managed to code all the conditional viewports out of this theme...until now :/ seems that %mv doesn't like being nested with other conditionals.
12:03:24S_a_i_n_tThis seems a fairly recent thing, this used to work.
12:05:11ZagorJdGordon: should I add it to the trunk?
12:05:20***Saving seen data "./dancer.seen"
12:08:28S_a_i_n_tJdGordon: This is *really* weird. A Conditional Viewport fails to parse in the SIM if the identifier is a capital letter.
12:08:31 Join MethoS- [0] (~clemens@
12:09:17S_a_i_n_t"%Vl (A,-,-,-" fails to parse, but "%Vl(a,-,-,-" is fine :/
12:09:56S_a_i_n_tSomething is seriosly wrong with syntax between the sim and what is intended/expected.
12:10:06S_a_i_n_t*seriously even
12:16:55 Quit MethoS- (Remote host closed the connection)
12:19:03 Quit pamaury (Remote host closed the connection)
12:20:54 Join JdGordon| [0] (
12:20:54 Quit JdGordon| (Changing host)
12:20:54 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
12:21:13 Quit ehntoo_ (Ping timeout: 265 seconds)
12:21:32JdGordon|Zagor: yes please
12:23:04JdGordon|S_a_i_n_t: that is expected
12:23:24JdGordon|all labels will be replaced wirh strings once we move also
12:28:02S_a_i_n_tcome again?
12:28:24S_a_i_n_tHow many times will the skins break?!?
12:28:48*S_a_i_n_t imagines that will break themes again...correct?
12:29:02JdGordon|none of this crappy one letter identifiers anymore
12:29:10JdGordon|it wont brrak old yhemes
12:29:28S_a_i_n_tso, what will the alternative be?
12:29:57JdGordon|instead ofvonlyballowing ine letter it will take a whole string
12:30:15S_a_i_n_tI also notice there is no mention anywhere that identifiers will reject capital letters.
12:30:35S_a_i_n_tHmmm, perhaps I'll have to see it in action, but I don;t really see the point.
12:35:19 Quit GeekShadow (Ping timeout: 240 seconds)
12:38:58JdGordon|%xd(shuffle) is mucj nicer than xd(a)
12:39:53S_a_i_n_tI find as long as your %Xl lines are descriptive, there's no problem.
12:40:31S_a_i_n_tie "%xl(D,Symmetry Playmode Status.bmp,0,0,)" etc.
12:41:20JdGordon|im not forcing this. just allowing ot
12:41:46S_a_i_n_tAh, great.
12:44:30 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:50:53ZagorJdGordon|: linked to the main builds?
12:54:36ZagorJdGordon|: I'll fix all the warnings too, then
12:56:02JdGordon|yes pleasd although it isnt striclty needed yet. but will make it easier ro work witth
12:58:00Zagorwhat is skin_debug_tree() supposed to do on target? printf() isn't very useful there...
12:58:31Zagoror should we exclude skin_debug from target altogether?
12:59:59JdGordon|it wont be compiled for target. i think it will cause errors if debug isnt compiled at all
13:00:27JdGordon|a patch would be great if you cant be bitgered fixing those errors
13:00:42Zagorit is compiled for target currently. do you want it #ifdef:ed away?
13:00:54 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
13:01:43JdGordon|pribably. i'm using ifedf ROCKBOX in the rest of the lib
13:02:38 Quit dfkt (Disconnected by services)
13:02:41 Nick dfkt_ is now known as dfkt (~dfkt@unaffiliated/dfkt)
13:04:34CIA-85New commit by zagor (r26884): Added skin_parser library to build system. Fixed some warnings.
13:06:18CIA-85r26884 build result: 15 errors, 0 warnings (zagor committed)
13:06:42Zagorwhy the heck is frederico still getting builds?
13:07:53Zagoroh, I forgot root.make
13:10:03CIA-85New commit by zagor (r26885): Added skin_parser library to build system.
13:11:34CIA-85r26885 build result: All green
13:11:41 Quit elcan (Read error: Operation timed out)
13:15:33 Join elcan [0] (
13:25:00pixelmaS_a_i_n_t: what happens if you remove the %s from the longer line you posted a while ago?
13:25:35JdGordon|Zagor: thanks
13:26:36S_a_i_n_tpixelma: Which? All of them?
13:26:58pixelmayes all
13:28:17S_a_i_n_tHmmm..gimme one minute and I shall try.
13:30:08S_a_i_n_tpixelma: Wow...nice.
13:30:11S_a_i_n_tIt works.
13:30:23S_a_i_n_tThough I kinda need the alignment :/
13:30:32S_a_i_n_tHow did you know to try that?
13:32:14 Quit r0b- (Ping timeout: 272 seconds)
13:32:33 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
13:33:43S_a_i_n_tpixelma: For the moment, I have "fixed" it by making the "true" case of %mv point to a conditional viewport. But this is a very ugly "fix" IMO
13:39:41 Quit dfkt|x (Remote host closed the connection)
13:45:47 Quit ThomasAH (Ping timeout: 240 seconds)
13:45:53 Quit swilde (Ping timeout: 258 seconds)
13:46:06 Join ThomasAH [0] (
13:46:40 Join Highlander [0] (
13:47:11pixelmathere is a problem with continually updating info in scrolling lines like you need for scrolling lines (known issue with old bug report but not easy to fix (at least with the old parser). I just didn't know if %s in sublines that won't get displayed anyway could cause the effect too
13:47:37pixelmaerrr... like you need for volume display
13:47:44JdGordon|%s anywhere in the line should work.. maybe
13:48:02S_a_i_n_tJdGordon: Nope.
13:48:11JdGordon|and enabling a vireport isnt a nasty hack
13:48:15S_a_i_n_tthat line *did* work, prior to the syntax change :/
13:48:26JdGordon|yeah yeah....
13:50:40pixelmawell the line "works" but can't update the content correctly even in the subline which has no %s in it, if I understand correctly
13:50:54S_a_i_n_tIt's nasty if you're trying to avoid conditional viewports. The nastyness comes from being able to visably see the viewport take over.
13:51:12S_a_i_n_tpixelma: without the %s, it updates fine.
13:51:30S_a_i_n_tit's just unfortunate that space requirements mean I need the line to scroll.
13:54:22 Join wodz [0] (
13:55:04wodzpixelma: how are tests on Ondio going?
13:55:07S_a_i_n_tI've simply left the old (yet updated) line in the theme, commented it out, and left a not for myself (or others) that it should be safe to comment out the hacky fix I made, and uncomment the original line when the new parser is finally in.
13:55:42 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
13:57:13pixelmawodz: looked fine so far with all the different things I tried, no "long term" test yet though
14:00:08 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:00:16JdGordon|t t
14:00:44 Quit kramer3d (Ping timeout: 272 seconds)
14:02:06 Join Jaykay [0] (
14:04:52wodzamiconn: I have implemented natural cubic b-spline interpolation (instead of linear) in test_grey. I hope You will like it :-)
14:05:22***Saving seen data "./dancer.seen"
14:12:23 Quit dfkt|x (Remote host closed the connection)
14:25:32 Join MethoS- [0] (~clemens@
14:29:02 Quit MethoS- (Remote host closed the connection)
14:46:43 Quit krazykit (Quit: work)
14:50:41 Join teru [0] (
14:53:08JdGordonZagor: thanks alot, except it looks like it isnt getting built for sims?
14:53:11 Quit Topy44 (Ping timeout: 260 seconds)
14:53:14JdGordonI did a make reconf
14:53:29 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
14:53:30Zagorah, I never tested that
14:54:25 Join Topy44 [0] (
14:55:50JdGordoneasy enough, just added it to the uisimulator.make file
14:55:58 Part LinusN
14:56:02Zagordon't commit that :)
14:56:27JdGordonoh ok :)
14:56:48JdGordonyou're making the sim depend on rockbox.elf or something instead?
14:57:19 Quit dfkt (Ping timeout: 245 seconds)
14:57:56wodzZagor: what is the story with zero wait boost for coldfire?
14:57:56Zagorno, you're right. that is the proper fix. I was distracted :)
14:58:16Zagorwodz: story?
14:59:44CIA-85New commit by jdgordon (r26886): make the sim build the skin lib also
15:00:31JdGordonOK, *now* I can get the old code out :)
15:00:37wodzZagor: FS #9797 - comments ware rather positive
15:01:31CIA-85r26886 build result: All green
15:01:34 Nick dfkt_ is now known as dfkt (dfkt@unaffiliated/dfkt)
15:01:51Zagorwodz: the idea with the zero-wait patches was to make fundamental changes to the way we boost, by making it more fine grained and on-demand in order to get a lower average clock speed. I never managed to produce any significant gains though.
15:03:32Zagorcommitting only the quick boost patches will likely cause very little improvement and risk causing hard-to-pin-down bugs.
15:04:05Zagoror create suspicion in people's minds that strange bugs are caused by it
15:04:27wodzZagor: so in general this abuses frequency change scheme (at least in coldfire) and provide little improvement in battery life - right?
15:05:08Zagorit's not clear-cut that it is abusive, but sure it can be interpreted that way
15:06:03ZagorI think we can safely close them. as I mention in a commit, they will not be committed but were part of a research effort.
15:06:11Zagor*in a comment
15:08:22TorneZagor: a (comparitively) huge number of people use the equivalent for ipod
15:08:52Torneand buschel is still pushing it, i think
15:08:58ZagorTorne: FS #9800 ?
15:09:38Tornehm, not htat one
15:09:41TorneFS #8668
15:09:56 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
15:10:08Torneor maybe that's not quite the same thing?
15:10:10Tornenot sure.
15:11:06Zagorno, it's not the same. he just boosts at other places. my patch changes how boost is done.
15:11:41TorneOh, so it's still using the regular boost method, just a lower unboosted speed?
15:11:50Torne(it seems like you'd want to do both)
15:14:19 Quit Highlander (Quit: Quitte)
15:16:53Tornei mean, i'm not a huge fan of 8668 :)
15:22:35JdGordonZagor: just to be difficult.... is it possible to use the same makefile to build the lib for rockbox and the editor? do I just setup a mostly blank MAkefile and include the one you added?
15:23:36 Quit bieber (Ping timeout: 265 seconds)
15:23:52 Join bieber [0] (
15:23:53JdGordonI guess even just getting the hardcoded file list out of Makefile so it uses SOURCES for both would be good enough
15:24:03ZagorJdGordon: exactly. you just need the parent Makefile to define the necessary rules for compilation
15:24:23ZagorI'd prefer if we can unify the build
15:28:30 Quit halmi (Read error: Connection reset by peer)
15:28:52 Join halmi [0] (
15:29:36 Quit JdGordon| (Quit: Lost terminal)
15:39:08 Join jgarvey [0] (
15:44:24 Join Stephen__ [0] (~S@
15:46:44Stephen__how do i build the skin updater ?
15:46:56JdGordonrun make in that directory
15:47:28Stephen__does it have to run on zips or will the .wps just do ?
15:48:49mc2739Stephen__: just the wps, sbs and fms files
15:48:56Stephen__cool thanks
15:50:06 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
15:51:04 Join krazykit [0] (~kkit@
15:51:54 Quit S_a_i_n_t ()
15:52:16 Join S_a_i_n_t [0] (S_a_i_n_t@
15:53:10 Quit S_a_i_n_t (Client Quit)
15:53:15 Join S_a_i_n_t [0] (S_a_i_n_t@
15:53:18 Quit mitk (Quit: Leaving)
15:54:11CIA-85New commit by jdgordon (r26887): tiny clean up of memory allocation
15:54:47Stephen__whats the syntax for the skinupdater ?
15:55:14S_a_i_n_t./skinupdater.exe oldtheme.wps/sbs newtheme.wps/sbs
15:55:28Stephen__thanks S_a_i_n_t
15:55:35 Nick bimbel is now known as bmbl (~Miranda@unaffiliated/bmbl)
15:55:35S_a_i_n_tno problem
15:55:52CIA-85r26887 build result: All green
15:56:07Stephen__it works!
15:56:15Stephen__now to update my themes in work.
15:56:51JdGordonwell of course it works :D
15:57:06JdGordonStephen__: was it you who was uploading the svn themes to the theme site?
15:57:13Stephen__yeah it was.
15:57:25JdGordonhows that going?
15:58:14Stephen__had to stop to get the breakage out of the way. i have archos and ipod mini i think left to go. have them donew just have to update the wps now, short on time tho due to rl
15:58:46JdGordonit would be nice to be able to remove them from svn
15:59:17Stephen__I have them zipped up in the old syntax, i guess i could run them now and do them i think that'd be them finished then.
16:00:07Stephen__after today dunno when i'll be on agqin to do it.
16:00:28JdGordonthey've all been updated to the new format in svn
16:03:35Stephen__yeah but that would mean redownlaoding and organizing ;o)
16:04:15Stephen__actually that would be better save for any of my errors or conversion mistakes
16:05:23***Saving seen data "./dancer.seen"
16:08:13S_a_i_n_tActually, the themes still need to be checked after runnng ./skinupdater too...
16:08:21S_a_i_n_tit's not without its quirks.
16:09:18Stephen__yeah but well easier than doing it by hand which I had contemplated hell with that ;)
16:10:39JdGordonit should handle those old skins pretty well (why we bother to keep them at all is beyond me)
16:11:05S_a_i_n_twhile you're updating the themes, make sure to take DancePuffDuo out into the woods and shoot it, then burn it, then piss on it :P
16:11:28S_a_i_n_tthen burn it again.
16:12:09JdGordonit's not *that* terrible
16:12:25JdGordoncall it a proof of concept and it is quite good
16:12:40*JdGordon secretly has that as his theme on 8 of his daps!
16:12:53S_a_i_n_ti knew it!
16:13:50 Quit B4gder (Quit: It is time to say moo)
16:13:52S_a_i_n_t"bohes" also needs to be put out of its misery too...
16:14:01S_a_i_n_tdammit, "boxes"
16:14:22S_a_i_n_tOk theme, hideous colour.
16:15:22Stephen__ok erm problem with last archos theme
16:15:27Stephen__site complains of,
16:15:28Stephen__Warning: join() [function.join]: Invalid arguments passed in /home/themes/www/upload.php on line 106
16:15:29JdGordonthat is the origional colour you damn youngin!
16:15:52JdGordonthat colour was all we had at one point!
16:16:16S_a_i_n_tStephen__: I got that seems to be something to do with the split of the themesite?
16:16:48S_a_i_n_tIt didn't affect the theme upload when it spat that at, "meh" ;)
16:16:51Stephen__i got 2 up before that!
16:17:16S_a_i_n_toh, hmmm...I did two in a row and got the same error both times.
16:17:43S_a_i_n_tchecked my email, got the confirmation, confirmed, theme was, dunno what the deal was.
16:18:36Stephen__just uplaoded 2 more now
16:18:42Stephen__it's just one it's failing on
16:19:17S_a_i_n_tChecked it for any weirdness in the syntax?
16:19:41S_a_i_n_tcheckwps has managed to let a few through that don't actually run on device.
16:21:26Stephen__looks ok to me
16:23:12Stephen__even changing the target fails it
16:24:23Stephen__and it was the last one too!
16:24:51 Join komputes [0] (~komputes@ubuntu/member/komputes)
16:24:58Stephen__if anyone wants to do it while i'm away it's rockboxed for the archoses
16:25:10Stephen__failing that i'll try again when i get back online
16:25:34Stephen__AND JdGordon If you can ping me with missing targets i check the logs.
16:26:50 Join halmi_ [0] (
16:27:43 Join toffe82 [0] (~chatzilla@
16:29:35 Quit halmi (Ping timeout: 260 seconds)
16:30:36AlexPStephen__: Are you in CREDITS?
16:30:52Stephen__emm nope, shouldn't be
16:31:10AlexPWell, I think you should which is why I was asking :)
16:31:19AlexPFor all this theme moving stuff
16:32:13Stephen__ah right, that would be great thanks AlexP
16:32:25AlexPWhat is your full name?
16:32:32Stephen__I noticed Rockboxed is missing for most targets, seems to have been skipped for the conversion.
16:32:37Stephen__Stephen Carroll
16:34:45 Join Buckwheat [0] (~chatzilla@
16:35:12CIA-85New commit by alex (r26888): Add Stephen Carroll to CREDITS for all his hard work stripping out the additional themes from SVN and getting them on the themesite.
16:35:21AlexPStephen__: There you go :)
16:35:37Stephen__thanks very much AlexP
16:35:49AlexPno problem, thanks for your work on this
16:35:57Stephen__I t5hink most are done, 1 failing for the archoses
16:36:05Stephen__happy to help
16:36:39CIA-85r26888 build result: All green
16:38:59JdGordonAlexP: slightly premature with the "stripping out from svn" bit :)
16:39:13AlexPSlightly perhaps :)
16:39:36AlexPBut you could read that as extracting but leaving in place (if you were pushing a point) :)
16:45:26 Part Zagor
16:59:20pixelmaStephen__: did you give e.g. the 160x128 themes etc. unique names when needed (there are some that "just work" in different colour depths - and some with slightly different versions for different colour depths like e.g. cabbiev2)
16:59:47Stephen__errm no pixelma
17:00:10Stephen__i can do that tho but it will take time, is there a preference you'd prefer for the naming ?
17:01:24 Quit teru (Quit: Quit)
17:01:31pixelmaI'm curious how the latter would appear on the theme site with e.g. the greyscale and colour version *technically* working on both
17:01:47pixelmanot sure what to advice there
17:02:42Stephen__yeah that's on eproblem, altho if I add colours to the viewport line that would not show it under the greyscale targets would it ?
17:08:12 Join DerPapst1 [0] (
17:09:06 Quit S_a_i_n_t (Ping timeout: 260 seconds)
17:09:16 Join S_a_i_n_t_ [0] (S_a_i_n_t@
17:10:47Stephen__ok got that last theme uploaded.
17:10:51 Quit DerPapst (Ping timeout: 260 seconds)
17:11:02Stephen__pixelma, when there's a format for the naming I can rename the files.
17:13:00 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:15:10AlexPStephen__: So is that all of them for all targets everything ever etc.? :)
17:15:26Stephen__not all targets, just screen sizes.
17:15:35Stephen__i can do all targets if you want,
17:15:41Stephen__just need a naming scheme
17:16:10 Quit krazykit (Quit: Connection reset by beer)
17:16:54AlexPI guess the major differences are grey/colour?
17:17:01Stephen__yeah pretty much.
17:17:34Stephen__that's 160x128 i think is all that matters right ?
17:18:17AlexPmaybe name-c name-g and name-m or something
17:18:25AlexPfor colour, grey and mono
17:18:50Stephen__there's a few missing i think from the conversion, I can redo them ?
17:18:51 Quit t0rc (Quit: Leaving)
17:19:00Stephen__pixelma, what you think of -c -g and -
17:19:03Stephen__-m ?
17:19:08Stephen__I'm happy to go with that ?
17:19:08pixelma128x128 exists in grey and colour too (only cabbiev2 I think)
17:19:25Stephen__cabbie is staying in tho isn't it ?
17:19:41pixelmacurrently yes
17:19:43AlexPI went for -c etc to avoid colour/color and grey/gray arguments :)
17:19:52pixelmaand you forgot the one Player theme ;)
17:20:04*Stephen__ makes mental note
17:23:32pixelmahmm, the Mpio HD200 (it is called I think) has no theme site entry at all yet although I think it could. That's the one with the 128x128 greyscale display
17:24:24CIA-85New commit by nls (r26889): ARMv6 versions of X(N)PROD31 macros and MULT32 macro. Saves about 1MHz or 3% decoding vorbis on gigabeat S.
17:24:33pixelmabluebrother, wodz: do I remember correctly that the Mpio is supported by the Rockbox Utility already?
17:24:53Stephen__fuze v2 has none either
17:25:34pixelmawell, fuze v1 themes will work on it, no? (And H300 ones etc.)
17:26:00pixelmaStephen__: just remembered another thing - what did you do with the remote WPSs?
17:26:09CIA-85r26889 build result: All green
17:26:29Stephen__think i ledft them out.
17:26:41Stephen__should i add them yeah to the config and in the wps folder ?
17:27:44CIA-85New commit by nls (r26890): test_codec: print crc's that start with 0 properly.
17:29:00 Quit bmbl (Ping timeout: 258 seconds)
17:29:14pixelmaI'd actually prefer a different solution but that needs work on the theme site internals before - have the remotes as separate theme "targets"
17:29:20 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
17:29:24CIA-85r26890 build result: All green
17:30:33Stephen__yeah i left em out. will have to redo them to add them in.
17:31:03Stephen__and rename the files
17:41:08 Join Highlander [0] (
17:42:32pixelmaStephen__: that would mean uploading a theme in some cases up to three times (one without rwps, one with Iriver rwps, one with Iaudio rwps - the latter two if existent).Even if you do - could you keep packages of Iriver RWPSs and Iaudio RWPSs? In case someone implements the changes to the themesite I mentioned
17:43:17Stephen__i have the rwps i don't use them and they're in the old syntax, but easily changed
17:44:56Stephen__should i leave them out for the time being ?
17:45:17AlexPI think they should be added really
17:46:06AlexPThey are quite important for remote targets, as without one if you change theme you might change font which won't work with the old rwps and the remote becomes less useable
17:46:24 Join Buschel [0] (
17:48:02Stephen__default ones ?
17:53:05 Quit petur (Quit: disconnect network and go home :))
18:01:22CIA-85New commit by nls (r26891): Delete some duplicate inline asm macros since they now live in the codeclib.
18:03:06CIA-85r26891 build result: All green
18:05:24***Saving seen data "./dancer.seen"
18:05:45 Quit GeekShadow (Quit: The cake is a lie !)
18:07:26 Quit Buschel (Ping timeout: 260 seconds)
18:09:18 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:12:42 Quit pamaury (Remote host closed the connection)
18:15:00Stephen__see yas
18:15:02 Quit Stephen__ (Quit: Leaving)
18:15:03 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:17:22BuckwheatSomething is wrong with the installer the sansa fuze v2 dosen,t show up
18:19:21 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:21:29 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:22:37AlexPWhat is the sansa fusse v22 listed as on the website?
18:23:00Buckwheatsansa fuze v2
18:23:06AlexPno, unstable
18:23:19AlexPAnd have you enabled unstable targets in Rockbox Utility?
18:23:37AlexPThen it isn't supported yet
18:23:44AlexPWhich as it is unstable is no big deal
18:24:04BuckwheatI have rockbox on my sansa
18:24:11Buckwheati installed it form the installer
18:24:15Buckwheati just want to update it
18:24:28domonokywhich rbutil ?
18:24:51Buckwheatthe one from the sight in the releases link
18:25:00domonokyfuze v2 is not included in the normal released rbutil. you need to use a svn version of that.
18:25:09Buckwheatoh where can i get that
18:25:20domonokybluebrother provides some svn binarys...
18:26:05domonokywe should make a new rbutil release somewhere in the next time :-)
18:26:08AlexPSo you must have installed it from a SVN version, but then downloaded the release version afterwards
18:26:38Buckwheatsomeone gave me a link yesterday so thats what got me confused
18:27:08Buckwheatthank you
18:27:47BuckwheatI wonder if the fuze v1 themes work on the v2
18:28:04AlexPThey depend mainly on screen size
18:28:14AlexPSo any with the same size will work
18:28:23Buckwheatok thank you
18:28:33AlexP(pretty much)
18:29:30kugelJdGordon: why are you disallowing drawing into the default viewport
18:29:40kugelwon't that break many, many themes?
18:29:53kugele.g. those which don't use viewports at all?
18:29:59 Join Buschel [0] (
18:32:50bertrikkugel, are you familiar with the clocking setup of the amsv2 sansas?
18:33:10 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:35:47pixelmadomonoky: do you remember our remotes on theme site talk at DevCon?
18:36:40domonokypixelma: yes, i remember a bit, but dont have time or motivation todo something about that :-)
18:37:15 Quit Buschel (Ping timeout: 276 seconds)
18:37:58 Join rob [0] (
18:44:34 Nick YPSY is now known as Ypsy (
18:46:27 Quit BradC (Ping timeout: 265 seconds)
18:46:53 Join funman [0] (~fun@rockbox/developer/funman)
18:48:45bertrikfunman, I think I can make some sense of the ams v2 PLL settings
18:49:13 Join domonoky1 [0] (
18:49:35saratogabertrik: did you find more clock speeds?
18:49:41CIA-85New commit by nls (r26892): ARMv6 vector mutiplication asm, speeds up vorbis decoding about 0.1MHz on gigabeat S.
18:50:09bertrikno, I just looked at two PLL settings: for 160 MHz and 240 MHz
18:50:28 Quit domonoky (Ping timeout: 265 seconds)
18:51:18CIA-85r26892 build result: All green
18:51:23 Join CGL [0] (~CGL@
18:51:50funmanis it similar to as3525 ?
18:52:25 Join Luca_S [0] (
18:52:42bertrikyes, I think so, lowest 8 bits are the multiplier (F-1), bits 9-11 is the divisor (R-1) I think
18:52:55kugelbertrik: not really
18:53:55bertrikSo for 240 MHz we have value value 0x1813, so F = 20 and R = 1 -> 12 MHz * F / R = 240 MHz
18:53:56funmanbertrik: try changing all of them and measure pll (i didn't test the 160MHz setting). I have a script to check if a PLL setting is valid (according to requirements listed in as3525 datasheet)
18:54:34bertrikAnd for 160 MHz we have value 0x154F, so F = 80 and R = 6 -> 12 MHz * 80 / 6 = 160 MHz
18:55:48bertrikI don't know what bit 12 does, bits 13/14 are the output divider I think
18:56:07bertrik(wait, I'm off by 1 bit)
18:56:41bertrikbit 0-7 = F, bit 8-10 = R, bit 11 = ?, bit 12-13 = OD
18:57:04kugelhow can F == 80?
18:57:49*kugel slaps forhead
18:57:58kugelnevermind :p
18:59:58CIA-85New commit by funman (r26893): Make sure files which aren't windows-specific use \n line endings only
19:01:39 Join Tbroflaro [0] (
19:01:42CIA-85r26893 build result: All green
19:01:56 Quit Tbroflaro (Client Quit)
19:02:29 Join Tbroflaro [0] (
19:02:39bertrikI plan to try to switch the audio playback to PLL B and experiment with PLL B settings, that way any wrong settings can only mess up playback and do not cause a hang of the rest of the player
19:02:47 Part Tbroflaro
19:04:24funman% svn propdel svn:eol-style -R svn://
19:04:24funmansvn: Setting property on non-local target 'svn://' needs a base revision
19:04:24bertrikSetting the PLL to 96 MHz (or some multiple) should allow us to get the sample rate very accurate (0.04% error). This error is currently especially pronounced on the ams v2 players (currently 1.1% error).
19:04:27funman[fun@kru ~]% svn propdel svn:eol-style -R −−revprop -r HEAD svn://
19:04:30funmansvn: Repository has not been enabled to accept revision propchanges;
19:04:33funmanask the administrator to create a pre-revprop-change hook
19:04:40 Join BradC [0] (heh29705@
19:04:47funmani could read the property without a checkout but for deleting it seems different?
19:04:49gevaertsfunman: I didn't notice we had a consensus for that
19:04:57saratogacould we use 12MHz*81/4 = 243MHz?
19:05:11funmangevaerts: i sent an email more than 24h ago
19:05:15saratogathat would give us a sample rate of 44,150
19:05:46funmanDate: Wed, 16 Jun 2010 08:35:06 +0200
19:05:47gevaertsfunman: right, so you ignore all mails before that
19:06:40funmanonly you objected, and you didn't tell why, so i sent an email again to be sure it wasn't an heated reply
19:06:42TheSevenis it technically possible to auto-apply an eol-style property to all text files on the server side?
19:06:53gevaertsI *did* tell why
19:07:06 Quit Luca_S (Quit: CGI:IRC)
19:07:07n1sTheSeven: i think it can be doen with a post-commit hook
19:07:19gevaerts*you* on the other hand didn't feel like saying what's wrong with eol-style=LF, you only stated that you can't be bothered setting it
19:07:40bertriksaratoga, aha, that would be nice, so we still have to enable only one PLL and yet get a reasonably accurate playback rate
19:07:44funmanin ?
19:07:46domonoky1are people sure, there are no txt files which needs windows line endings ? (sapi_voice.vbs could need it)
19:07:49TheSevengevaerts: that's why I wouldn't bother setting it, nor delete it, and just let the server take care of it
19:07:58funmangevaerts: if you have arguments irc isn't the place for them
19:08:10saratogabertrik: plus the final clock speed doens't really change, so hopefully nothing breaks ...
19:08:17gevaertsTheSeven: maybe, but then the server side system needs to be in place first
19:08:27funmandomonoky1: perhaps windows drivers files also
19:08:48TheSevendomonoky1: we should check those
19:08:59TheSevenwhich files *are* actually windows-specific?
19:09:02gevaertsfunman: right, I didn't copy bluebrother's email, I only said that it expresses what I meant
19:09:04domonoky1"Mr. Someone" will do it :-)
19:10:01bertriksaratoga, there are some restrictions on the frequency ranges of some signals internal to the PLL (at least there are on as3525), I don't know the restrictions yet for ams v2, so I don't know if/how it is possible to set 12 MHz*81/4
19:10:26TheSevenwhat sounds best to me right now is adding some hook that automatically sets eol-style=LF on everything that doesn't have an eol-style property on commit
19:10:40 Quit funman (Quit: free(random());)
19:10:43TheSeventhis way you can still override it if neccessary, and still don't need to care about it while committing
19:11:03saratogawith all the dividers we have, even if that setting doesn't work, i'm sure we can find another good one
19:11:25TheSevenhowever, if that could work, deleting the property now would be absolutely pointless IMHO
19:11:41n1shmm, apparently setting properties automatically isn't good with svn, it causes problems according to this page they recommend instead rejecting the commit
19:11:45bertrikI thought eol-style:native was supposed to avoid eol problems in the first place, now we want to force only one way?
19:12:05Tornebertrik: it doesn't actually avoid them, because the build environment isn't entirely line ending neutral :)
19:12:19gevaertsbertrik: eol-style:native doesn't actually avoid issues. For our usage, it actually tends to cause some
19:12:22Tornethe 'safe' way is to always use unix line endings, even on windows
19:12:30Torneconfigure cygwin to use unix line endings, if you use it
19:12:37Torneand just don't use an editor that's too thick to handle unix line endings.
19:12:48Tornewhich is, more or less, just notepad at this point
19:14:07n1sof course, rejecting commits with \cr\lf will not work well if we need those for some files which some people seem to think
19:14:23TheSevenDid I understand correctly that eol-style=LF will change all CRLFs to LFs on commit, even if your editor converted the file to CRLF before? (for whatever reason)
19:14:57Tornen1s: we do, but you can *not* reject those if they say eol-style=CRLF
19:15:08Tornei.e. for files that don't want unix line endings, someone must explicitly set eol-style
19:15:09TheSevenok, then eol-style=native doesn't seem to have any advantage for us
19:15:24n1sTorne: ah, right
19:15:29TorneTheSeven: if eol-style is set to *anything* it mangles the line endings for you, basically
19:15:34TorneTheSeven: it only leaves it alone if it's not set at all
19:15:43TheSeventhat's exactly what we need...
19:16:11Tornebut yeah, in older versions at least, setting eol-style on commit doesn't work properly
19:16:12amiconncrlf isn't necessary for anything except (a) silly editors and *maybe* (b) .cmd files which we don't have
19:16:15Tornedunno if that's still true
19:16:29amiconnvbscript doesn't care - either line endings work
19:16:41Torneit would be slightly annoying to be *forced* to set eol-style manually on every file, or else have your commit rejected :(
19:16:46TheSevenamiconn: someone mentioned INF files above, I'm not sure about those
19:16:55n1sdo we have inf files?
19:18:53saratogamatlab spits out a couple PLL settings that are > 99.9% pitch accurate
19:19:09 Quit DerPapst1 (Quit: Leaving.)
19:19:12 Join DerPapst [0] (
19:19:15saratogacouple dozen i mean
19:20:09saratogatheres one thats 99.97% of the correct value
19:22:12kugelI guess sandisk would just use a more suitable PLL setting instead of resampling to 48KHz if it was that easy (unless the OF doesn't actually resample)
19:23:35amiconnIiuc the OF play off-pitch
19:24:03 Quit DerPapst (Ping timeout: 264 seconds)
19:24:06saratogayeah they don't hit the pitch exactly either, but they do a lot better
19:24:09amiconnsaratoga: Is this just taking the ratio into account, or also the pll upper/ lower limits?
19:24:23saratogaamiconn: just the available ratios, I'm not sure what the limits are
19:25:48saratogadoes 24MHz on AMSv2 change the PLL or some other divider?
19:26:12bluebrotherTheSeven: IMO commit hooks should _never_ change the data that's committed. Including properties.
19:26:41bluebrotherif there's a way for git-svn to handle it (haven't found the time to check yet) what's the problem? Just let it do that.
19:27:04*bluebrother goes trying
19:30:17bertriksaratoga, 24 MHz is the crystal frequency I think
19:30:46saratogayeah thats the as3543 datasheet says
19:31:00saratogabut does the PLL actually reclock or does it get divided by the CPU?
19:31:08saratogaprobably the latter?
19:31:40bertrikI don't know, I guess we just switch clock source
19:33:46*amiconn would assume pass-through
19:33:46 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:34:31bluebrotherautoprops does work, at least on my setup. That is with git and svn 1.6.9
19:39:02 Quit n1s (Quit: Lämnar)
19:41:01amiconnHmm. What's the manual rockbox installation procedure for Clip+?
19:41:17 Join Luca_S [0] (
19:42:27bertrikget an OF, patch a bootloader into it with mkamsboot, copy it to the root of the player, unzip a on it and let the clip+ do a firmware update
19:43:04 Join watto [0] (~watto@
19:43:42bertrikYou can set the region by modifying the last character of the OF upgrade file name, like clppt.bin to enable a diagnosis menu in the OF
19:44:29bertrik(where the t in clppt.bin selects region "test" I suppose)
19:44:59amiconnNo single boot yet?
19:46:38bertrikNo, we don't do USB on clip+ yet, so we rely on the OF for that
19:46:46amiconnAh, there was that...
19:46:55amiconnDoes it matter which OF version I choose?
19:47:05saratogaif you lose the OF you brick the player, or at least have to take it apart and try to unbrick it
19:47:31saratogathe patcher has to recognize the firmware file, but otherwise no
19:47:43saratogalatest svn should patch all firmware files out there
19:47:55 Join Jerom [0] (~Jerom@
19:48:03bertrikI used the OF from (from the sandisk forums IIRC) and it works fine for me
19:50:50CIA-85New commit by ranma (r26894): Either the removal of the unneeded flush or the additional epn_info setup ...
19:51:13 Nick Ypsy is now known as YPSY (
19:52:24CIA-85r26894 build result: All green
19:55:50bertrikhm, now I'm a bit confused, I thought 0x1813 was the PLL setting for 240 MHz, but in the code I see 0x113B (R=2, F=60?)
19:59:25ranmaOk, with more logging, it looks like rockbox itself always reads one sector at a time.
19:59:45ranmaI suspect multisector read is broken in sd-as3525.c :)
20:00:22AlexPbluebrother: Should tex be in your list, or is it just the source?
20:03:32ranmaHmm, no that's not it.
20:03:40 Join n1s [0] (~n1s@rockbox/developer/n1s)
20:04:03ranmaEven if I limit the usb_storage buffer to 1 sector, the call (with count=1) doesn't return...
20:05:19 Join DerPapst [0] (
20:05:28***Saving seen data "./dancer.seen"
20:06:06*amiconn found a non-critical problem with the backlight defaults
20:06:41amiconnOn some (many? all?) targets the ordinary backlight timeout was increased, and is now longer than the timeout when plugged
20:06:50amiconnThat *somehow* doesn't make sense
20:12:37bluebrotherAlexP: right, forgot that :o
20:13:13bertrikI think it worked to run the audio off the PLL B on the clip+ too. If it's really so, that would mean we can definitely get very accurate with the sample rate.
20:13:16bluebrotherthough it doesn't matter for tex files. However, making all those files LF wouldn't be a bad thing IMO
20:15:42 Quit flydutch (Quit: /* empty */)
20:18:36 Join captainewkllllll [0] (2669ecc2@gateway/web/freenode/ip.
20:19:26bertrikyay, the PLL change for the clip+ did indeed work. Now the question is how much extra power the PLL takes and if it's worth it for the increased sample rate accuracy.
20:20:21Luca_Sto what frequence did you change it?
20:20:33bertrik192 MHz
20:21:13bertrikThis should result in 0.04% sample rate error instead of the current 1.1%
20:22:06saratogabertrik: what settings did you use?
20:23:18 Join pamaury [0] (
20:23:18 Quit pamaury (Changing host)
20:23:18 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
20:24:46bertriksaratoga, PLLB is set to 0x155F, which should be F=96 and R=6
20:25:30kugelbertrik: what about the 243MHz setting?
20:25:58bertrikkugel, try it! :)
20:25:59kugelthe one saratoga suggested
20:27:30bertrikthe PLL A setting for 240 MHz in the code does not match what funman told me on IRC (as far as I remember)
20:28:00kugelanother question, you cannot really be sure you are at 192MHz are you?
20:28:47 Quit n1s (Quit: Lämnar)
20:29:35bertrikI changed the audio clock to use PLL B as a source and audio speed seems normal, so I'm pretty sure. It slows down / speeds up if I change the PLL B setting
20:30:54bertrikI'm just testing "by ear" now, I haven't set up some setup for accurately measuring playback rate errors.
20:30:55 Join GeekShad0w [0] (~Antoine@
20:32:06 Join notlistening [0] (
20:33:54 Quit GeekShadow (Ping timeout: 260 seconds)
20:34:29bertrikkugel, saratoga: the patch is at
20:34:31 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
20:35:11saratogabertrik: what are the bits in the PLL register ?
20:36:06bertrikMy guess is: bit 0-7 = F-1 (multiplier), bit 8-10 = R-1 (divisor), bit 11 = unknown, bit 12-13=OD (output divider)
20:36:46saratogathe output divisor is because the internal frequency cannot get too high or low?
20:36:58bertrikwith a base frequency of 12 MHz (actually the crystal is 24 MHz I think, seems there's a factor 2 getting lost somewhere in the PLL)
20:37:14 Quit GeekShad0w (Ping timeout: 258 seconds)
20:37:19kugelmaybe that's bit 11?
20:38:08saratogaaccording to the AS3543 datasheet, the oscillator can run at 24 or 12MHz, with a default of 24
20:38:12bertriksaratoga, yes, I think so, it divides the Fvco and Fvco has limits like 200-400MHz
20:39:53ranmaHmm, I'm seeing MCI_RX_OVERRUN when read_sectors is called from usb_storage
20:40:40ranmaWeirdly dma_used is 0, it should be 1, cause it's still blocked on wakeup_wait AFAICS
20:41:22bertriksaratoga, what AS3543 datasheet do you have? Mine only shows the codec and PMU stuff, not clocking stuff (except for I2S PLL)
20:41:46saratoga"AS3543 3v2"
20:41:52saratogaTable 72. Clk_Cntr Register
20:43:06bertrikhm, I have rev 1.10, can you send it to me?
20:43:41saratogai think thats the same one
20:43:51saratogado you have that table on page 73?
20:44:34ranmaYep, definitively some weird dma_retain/release problem.
20:44:57ranmaIf I add dma_retain() to usb_drv_init the reads work fine.
20:47:05CIA-85New commit by bieber (r26895): Theme Editor: Made allll destructors virtual
20:48:02bertriksaratoga, we're talking about completely different things then. I was talking about a setting written to CGU_PLLB
20:48:41CIA-85r26895 build result: All green
20:48:55bertrikThe CPU is clock master for the I2S bus, so the PLL in the codec doesn't do anything
20:51:47bertrikOr could it be that this part contains the actual oscillator for the crystal and pre-divides the clock before it goes to the rest of the SoC (and into the PLLs)?
20:51:48 Quit grndslm (Quit: Leaving)
20:52:21 Join k-rad [0] (~ident@unaffiliated/k-rad)
20:52:37k-radcan you install a rockbox current build through the automated utility ?
20:54:31Jeromk-rad: I don't know if you can do this but updating rockbox to current build is very easy
20:54:34 Quit notlistening (Read error: Connection reset by peer)
20:55:20JeromIt's just a matter of extracting the zip into the root of your player
20:56:29bluebrotherk-rad: yes
20:57:10bluebrotheryou just need to use the "Installation" tab, not the "Quick Start"
21:04:45k-radthank you
21:06:18ranmaAh, found the bug in sd-as3525.c :)
21:08:05ranmaIf transfer() is called on lun1 and the external sd is not present, it does a goto to sd_transfer_error before calling dma_retain()
21:09:02 Join notlistening [0] (
21:13:51k-radmight rockbox give a e200 series sansa support for microsd > 8gb ?
21:14:34bertrikI think we already support up to 32 GB on e200v1
21:14:43CIA-85New commit by ranma (r26896): sd-as3525 was calling dma_release() without calling dma_retain() first, if ...
21:15:38ranmabertrik: usb_storage is working on as3525v1 for me now :)
21:15:51ranma(At least reads, didn't test writes so far)
21:15:53 Join Xerion_ [0] (
21:16:17CIA-85r26896 build result: All green
21:16:36kugelranma: \o/
21:16:47k-radthat was extremely easy to get working. thank you bluebrother
21:16:55bluebrotherRockbox supports SDHC. SDHC is specified up to 32GB (which is a limitation by the specification and not a technical one)
21:17:39bluebrothertheoretically SDHC could support much larger capacities. However, they decided to limit it to 32GB and come up with a new standard for larger cards (SDXC IIRC)
21:18:22 Quit Xerion (Ping timeout: 265 seconds)
21:18:23 Nick Xerion_ is now known as Xerion (
21:18:26 Join petur [0] (~petur@rockbox/developer/petur)
21:19:15Luca_Syay! finally I can test patches again (had to setup a linux machine, the vmware dev env is hopeless :/ but this is much faster :D)
21:20:38 Part watto
21:27:58 Join webguest37 [0] (
21:28:01webguest37Hi there
21:28:11webguest37I got a problem with my sansa clip v1 ;(
21:32:11 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
21:32:12 Quit kugel (Ping timeout: 260 seconds)
21:32:36 Quit Horscht (Quit: Verlassend)
21:32:58pixelmawebguest37: if you want help you need to give more info
21:33:11bluebrotherwebguest37: problems are unfortunate but do happen. However, I think there are quite some people waiting for the actual problem description right now
21:33:27JeromDOEPINT(i) = DOEPINT(i); <= Does that make any sense ?
21:33:57AlexPbluebrother: I'm not, I'm having too much fun inventing possible problems in my head
21:34:24ranmaJerom: clearing pending interrupts maybe?
21:34:28bluebrotherAlexP: haven't thought of that option. At least I was too lazy to invent problems by myself ;-)
21:35:08AlexPI'm currently imagining that it has been nibbled by a rampaging sheep
21:36:21JeromI don't understand how writing then writing the same value has any effect
21:36:54bertrikit's probably a clear-on-write register, very usual for interrupt clear registers
21:37:55JeromThanks, I'm reading through RB's code and find it quite interesting :)
21:38:14 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
21:39:20 Quit dfkt|x (Remote host closed the connection)
21:39:28amiconnPF cache building causes strange noises in the earphones on Clip+
21:40:07Luca_Sdamn.. I wanted to try bertrik's pll patch... and i just noticed that i left my headphones at work
21:40:17 Part k-rad ("Leaving")
21:40:21bertrikthe clearing of in interrupt bit has to be atomic, so instead of reading the value, clearing one bit and writing it back (with the risk that you ignore another interrupt that came in in the meantime), you just write the bit you want to clear
21:40:25pixelmaI find making fun of people who have a problem with their player a bit inappropriate even if it is about the fact that he doesn't tell us currently. I remember an occasion where I thought I had damaged my player and I felt helpless. It's not nice if then someone comes and makes fun of you. Sorry
21:41:35 Quit webguest37 (Quit: CGI:IRC (Ping timeout))
21:42:59amiconnSeems it's actually SD access. Also happens when (re)buffering
21:44:17JeromThanks bertrik, I always enjoy learning something. Btw is this atomic: int intr = USB_DEV_INTR & ~USB_DEV_INTR_MASK; USB_DEV_INTR = intr; ?
21:45:32amiconnYou can't do atomic bit manipulation on ARM without hardware support
21:46:50tmztbluebrother: now reading that SDHC hosts will be forward compatible with SDXC but it requires software support
21:47:40bluebrothertmzt: I've read recently that _some_ (i.e. slow) cards will be compatible. However, SDXC defines a different file system (exFAT) :(
21:47:41bertrikJerom, I'm not sure actually
21:47:47Luca_SI found some crappy in-ear headphones and my fuzev2 seems to play fine with bertrik's patch. of course i can't say anything about sound qality
21:48:03tmztyeah, part of the requirement is that the define be formatted with vfat
21:48:03*amiconn expects that an SDXC can be reformatted as FAT32
21:48:18tmztis there any reason the cards can't operate in a slower mode?
21:48:44tmzthere's an SDXC but it's full size
21:49:14bluebrotherafaiu there will be (at least) two SDXC versions: the slower ones (which are compatible to SDHC) and the faster ones which use a different interface mode or something like that.
21:49:59tmztfrom wiki:
21:50:00tmztSDHC devices will only support the SDXC cards which use UHS104 speeds;[30] SDHC devices will not recognize the SDXC cards which use the faster (SD 4.0), final specification of SDXC.[42]
21:52:02scorche|shtmzt: that really doesnt mean much...people said the same thing about SD -> SDHC, however it is just a software thing - what we really need to see is the actual specification, not just clippings from Wikipedia (i assume you meant wikipedia when you said "wiki" instead of the Rockbox wiki or any of the other thousands of wikis out there)
21:52:32tmztright, wikipedia. the problem is the lack of knowledge in the public domain
21:52:59scorche|shyup...hence why i said we needs to see the actual spec
21:53:01 Join krazykit [0] (~kkit@
21:54:46 Join dfkt|x [0] (
21:54:46 Quit dfkt|x (Changing host)
21:54:46 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
21:56:55dfktbertrik -
21:57:22dfktyour patch is 1001hz, svn is... off
21:59:43bertrikso, already an improvement compared to current svn :)
22:00:20dfktfor the discerning ears that notice it, maybe :)
22:00:41bertrikThe theoretically expected value was 1000.4 Hz
22:00:49Luca_Swhat does the OF show?
22:01:02dfktseems my spectrum analysis only does full hertz
22:01:09dfktmaybe it is a decimal
22:01:23 Join liar [0] (
22:02:23dfktLuca_S, an older version of the OF palyed at 1003hz - i think they made it better in the meantime, though
22:02:25dfkti can check
22:04:56dfktaha! bertrik's patch is exactly congruent with the newest of
22:05:07saratogaseeing better then 1Hz would require more then 44000 points in your fft at 44.1khz, see if you can increase the fft size
22:05:23dfktwell, it's 65k
22:05:33***Saving seen data "./dancer.seen"
22:06:01dfkthighest i can go
22:06:08saratogaso you get .7Hz resolution
22:06:30saratogahmm but i guess the screen only shows to the nearest 1Hz
22:06:34saratogaprobably good enough anyway
22:06:36dfktseems like it
22:06:45dfktany suggestions for a better spectrum analyzer app?
22:06:52dfkton windows
22:07:19bertrikdfkt, interesting find
22:07:48saratogai just use matlab :)
22:08:11dfktseems like it's roughly the same as the OF, and by roughly i mean exactly the same, as i see it :)
22:08:16saratogaso in one day we get the pitch thing figured out, volume range extended and almost working usb?
22:08:28gevaertsCouldn't you improve the resolution by just increasing the base frequency, i.e. like by using a 10kHz tone?
22:08:35*dfkt checks if wavelab is better suited than sound forge
22:11:50BuckwheatDoes anybody know how to disable the media refreshing on the sansa fuze v2? i have checked the FAQ.
22:12:06dfktwavelab can do 256k :) -
22:12:07saratogai don't think theres anyway to do it
22:12:43Luca_SBuckwheat: there is no way, however today's svn fixes a problem when rebooting from RB to OF unnecessarily caused a OF DB refresh
22:13:35saratogawow nice
22:13:45Luca_Syeah, it's a glorious day :D
22:13:58saratogai like the shape of the peak, you can see how the oscillator drifts :)
22:15:14CIA-85New commit by jethead71 (r26897): Switch iPod 3G to use EABI toolchain.
22:15:48dfktso, OF seems farther off, actually
22:16:00CIA-85New commit by jethead71 (r26898): Switch iPod 3G to use EABI toolchain. Make necessary threading changes to avoid use of stack after switching to idle stack.
22:18:04CIA-85r26898 build result: 28 errors, 7 warnings (jethead71 committed)
22:19:18 Join grawity [0] (
22:20:42*grawity looks around.
22:21:25grawityWhich .wps file syntax is the current one? One of the themes I downloaded uses %tag(x,y,z) and others have %tag|x|y|z|
22:21:31amiconnjhMikeS: oh
22:21:43AlexPgrawity: The former is svn, the latter is 3.6
22:21:52AlexPIt changed after the release
22:22:23bertrikmy theory about the assignment of bits to divisor R in the sansa clip+ does not appear to be correct (multiplier F is correct though I think)
22:22:52amiconnjhMikeS: Do you think this will work on G1/G2 as well?
22:22:58 Join Zigtown [0] (
22:23:25amiconnRed and yellow though
22:24:15grawityDamn they change fast.
22:24:27jhMikeSmost likely. gotta figure out what's the deal with that red though
22:24:41wodzamiconn: Please take a look at
22:24:54*grawity just installed r26480-100602 like last week... thought that _was_ the latest build, for nano2g.
22:26:18*jhMikeS thinks he just forgot a keyword
22:26:38AlexPgrawity: It probably was at the time, but there is a new current build on every source code change, of which there are many each day
22:27:48grawityWell, at least I had a fun time porting the WPS to "old" syntax using the ipod's "keyboard" =)
22:28:11AlexPThat seems rather masochistic :)
22:28:30grawityhey, I was bored.
22:29:40BuckwheatYou mean you can make wps right from the player?
22:29:50AlexPThey are only text files
22:29:50grawityBuckwheat: well, there is a text editor in Rockbox
22:30:00CIA-85New commit by jethead71 (r26899): Fix red. Need 'inline' with 'always_inline'.
22:30:06AlexPBut given you need to use the virtual keyboard it isn't something I'd recommend
22:30:30Buckwheati know that but i didnt know you could change the file type
22:30:31 Quit notlistening (Quit: Leaving)
22:30:39grawityWhat file type?
22:30:43AlexPContext menu, open with
22:30:54Buckwheatone second
22:31:58CIA-85r26899 build result: All green
22:32:47*grawity tries to port the same to the opposite direction.
22:33:35Buckwheatevery time i try to open a text file i get, "Data about at 308007F4 fsr 0x8 (domain 0, fault 8) address 0x9689AEFA
22:33:44AlexPThat's a known bug
22:33:53Buckwheatis there a fix
22:33:54AlexPThere is a task on flyspray
22:34:06AlexPIf there was a fix it would have been fixed
22:34:08 Join voRia [0] (~voria@
22:34:21jhMikeSamiconn: besides allocating tons of stack space for no good reason I could see, thread_exit was trying to optimize a common sub-expression by using a stack var (which didn't work after switching). Making the function "noreturn" kept thread_exit from trying to store return contexts.
22:35:14*grawity makes a mental note to report a bug about games putting scores in an inconvienent to backup location.
22:35:17dfktsaratoga, bertrik - all together (svn/1000hz/patch/of):
22:35:42grawityAh, one more thing: What's the difference between .sbs and .rsbs?
22:35:54bluebrotherthe r one is for the remote
22:36:04bluebrothersame for wps vs. rwps
22:36:13grawityHmm, and what does 'remote' mean here?
22:36:16AlexPdfkt: Nice :)
22:36:27AlexPgrawity: Some targets have a remote with a lcd
22:36:36bluebrotherremote is some hardware to control the player with
22:36:43grawityAh, that.
22:36:44bluebrothera remote is a remote :)
22:38:04 Quit Zigtown (Remote host closed the connection)
22:38:54bertrikdfkt, thanks! So the OF doesn't really do quite the same as I did with the patch. You didn't see any evidence of resampling, right?
22:39:17 Quit Jerom (Remote host closed the connection)
22:39:37dfktsaratoga meant, there is no resampling in the OF, iirc - i wouldn't know what to look for
22:41:03saratogausually if theres resampling you'll see odd effects from tones played close to half the sample rate
22:41:21dfktah yes, you could play a sine sweep and listen by ear
22:41:43dfktor just play a 22.050 tone?
22:41:55 Quit wodz (Ping timeout: 245 seconds)
22:42:17saratogalike a 20khz 0dB tone will playback as a 20khz tone plus another attenuated one at some difference of the new and old sample rate
22:42:40saratogaits usually pretty easy to see on an FFT like the one above
22:43:29saratogawith good resampling you usually can't hear it though, since the extra tone will be very weak
22:44:14CIA-85New commit by bieber (r26900): Theme Editor: Continuing work on rendering, skin preview will now show backdrop or background color depending on config file
22:44:36dfkti'll try a 20khz tone, sounds interesting to check
22:45:49 Quit Jaykay (Remote host closed the connection)
22:46:01CIA-85r26900 build result: All green
22:47:57 Join wodz [0] (
22:49:07dfktwhite is RB, red is OF:
22:50:57dfkttone was 21khz
22:52:49grawityQuestion: System -> Rockbox Info. Does the "Skin RAM usage" show the _used_ or _free_ size?
22:53:02bluebrotherthe allocated size afaiu
22:53:20bluebrotherhmm, scrap that
22:54:10 Join fml [0] (
22:56:52 Quit fml (Client Quit)
22:57:18bertrikdfkt, weird, the OF seems clean, RB has artifacts
22:58:09 Quit dfkt|x (Remote host closed the connection)
22:59:19dfktbertrik, yes - RB doesn't actually play 21khz, but lower and higher tones
23:03:15 Quit CGL (Ping timeout: 248 seconds)
23:03:25*grawity wonders if all iPods are making this quiet high-pitched noise...
23:03:51 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:05:20wodzgrawity: I don't hear any noise on my mini 1G
23:05:49grawity(And I mean from something inside, not through the earphones...)
23:06:15wodzgrawity: You mean spinning disk probably than
23:06:26grawityIt has flash memory.
23:07:03dfktthe screen?
23:07:06grawitythough it does seem that the sound changes while reading from disk.
23:07:19Bagderthe original mini 1G has a disk, doesn't it?
23:07:31grawityI have a nano2g.
23:07:31wodzThe only thing which can produce noise is DC-DC converter but they are running at least tenth of kHz frequency
23:07:39Bagderah I misread
23:07:58*Bagder backs back into the shadows
23:08:13 Quit krazykit (Quit: Connection reset by beer)
23:08:31 Quit GeekShadow (Quit: The cake is a lie !)
23:08:36 Quit petur (Quit: Zzzzzz)
23:08:53grawityEh, doesn't matter - I only hear it when holding the ipod to my ear anyway. Was just curious.
23:09:25 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:12:46 Quit evilnick_B (Quit: Page closed)
23:13:40wodzpixelma: HD200 is added to RockboxUtility (and is disabled by default). I didn't tested automatic installation at all.
23:17:31kugeldfkt: you are no fan of logarithmic scale, are you?
23:17:44kugelvery interesting findings nevertheless
23:20:55dfkti don't think it matters much if it's linear or logarithmic for these results
23:21:11 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
23:21:19 Join krazykit [0] (~kkit@
23:22:42wodzI come up again with question - When in FM mode I could in theory turn off large part of codec chip (saving is ~6mA or ~10% of total current consumption). This has one drawback - voices will have no effect (codec stopped). I do know nothing about about voice system in rb. Is there some way to know if voice is about to play so I could enable relevant parts of codec?
23:26:32 Quit krazykit (Quit: ok, going home)
23:27:49 Quit liar (Ping timeout: 258 seconds)
23:32:34 Join liar [0] (~liar@
23:32:36jhMikeSwodz: from the voice thread, you can tell if voice is about to be active (apps/voice_thread.c). you'd need to timeout though, since you can't really tell when it's actually done playing, only done inserting into the pcm buffer.
23:33:41wodzjhMikeS: can't I probe if pcm buffer is empty?
23:35:35jhMikeSsure. don't forget keyclick too. pcm.c could switch the codec on when a play request is about to happen
23:36:06 Quit Buckwheat (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
23:36:59wodzI will explore this a bit I think
23:38:26jhMikeSI think what would work best is to activate the codec if not already active, otherwise setup a powerdown timeout if pcm hasn't played for a certiain number of seconds (could use the backlight thread).
23:39:23 Join Buschel_ [0] (
23:39:49wodzYou mean check if codec is on at the begining of pcm_play_data()?
23:41:55jhMikeSyes, also pcm_play_pause
23:43:33wodzI don't quite understand how to setup timeout with backlight thread
23:44:57wodzYou mean something like for buttonlight?
23:45:07jhMikeStrack it in the SYS_TIMEOUT case, unless it's fading with software fading it has a period of HZ
23:46:23jhMikeSbacklight_timeout_handler looks like the spot
23:46:23 Join krazykit [0] (~kkit@
23:48:26wodzthanks, I'll explore this
23:50:51 Quit Buschel_ (Ping timeout: 260 seconds)
23:51:18 Quit Luca_S (Quit: CGI:IRC)
23:55:54 Quit ender` (Quit: The best way to accelerate a Macintosh is at 9.8m/sec²)
23:58:57wodzinteresting - setting timeout that way will have nice side effect that codec poweroff will happen always if playback is stopped long enough
23:59:00 Quit grawity (Quit: Good night.)

Previous day | Next day