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 2009-08-19

00:00:02mcuelenaere(or vice versa)
00:00:08JdGordon|yeah, i have
00:00:09JdGordon|works fine
00:00:16mcuelenaereit doesn't here ?
00:00:27 Join webguest [0] (n=04f04edd@gateway/web/cgi-irc/
00:00:31 Join truthtaco__ [0] (
00:01:01JdGordon|there is a very minor bug going from 3 to 32 and vice versa... but if thats what your talking about i wouldnt worry
00:01:47mcuelenaerethat's what I'm talking about
00:01:53mcuelenaereand there's a flyspray report about it ;)
00:02:00JdGordon|AH! thats not the menu at all...
00:02:06JdGordon|thats option_select()
00:02:50webguesthey could anyone tell me how to uninstall rockbox off my sansa e250r
00:03:28JdGordon|the wiki and manual and rbutil all make it very simple
00:04:15CIA-61New commit by rmenes (r22412): A few non-translation fixes for Tagalog.
00:04:23webguestso i can uninstall it the same way you would uninstall it on a e200
00:04:30kugelmcuelenaere: LCD_WIDTH/10 still is a bit low, btw (imo)
00:04:56JdGordon|webguest: yes
00:05:03mcuelenaerekugel: you want the scrollbar to fill up whole your screen? ;)
00:05:20mcuelenaerebut sure, go ahead and increase it
00:06:07kugelno, but it's only 17 on a e200. and still only 6 on my clip
00:06:11JdGordon|mcuelenaere: can you change the left/right side setting to not apply on the fly? its almost never displayed in that setting and feels really wierd having the space move
00:06:36JdGordon|kugel: do MIN(30,LCD_WIDTH/10)
00:07:12JdGordon|I mean MAX
00:07:12mcuelenaereJdGordon|: how? use 2 variables and only update the actual one when leaving the screen?
00:07:28JdGordon|mcuelenaere: just give it the F_TEMPVAR flag in settings_list
00:08:50JdGordon|*may not be named that actually*
00:08:54JdGordon|something along those lines
00:09:10mcuelenaereperhaps I should also use that for scrollbar width to fix that..
00:09:31JdGordon|no... that live updating is great
00:09:54JdGordon|its probalby a 2 line change in optino_select.c
00:09:55mcuelenaereyeah I think so too, but the bug isn't :)
00:10:23BagderZagor: broken build client?
00:10:43ZagorBagder: yes, I'm testing some changes. they didn't work too well...
00:10:46freqmod_qukugel: Is a bug report? ( i.e should i have posted it in the test build thread? ) i need to put it in manufacture mode
00:12:26***Saving seen data "./dancer.seen"
00:12:50mcuelenaereJdGordon|: redrawing the menu after the item changes works, but you can see the scrollbar rapidly switching from 24 to 3 pixels
00:13:00mcuelenaerealso I'm not sure if this breaks anything..
00:13:17JdGordon|you're not in the menu :)
00:13:31mcuelenaereredrawing the list* :)
00:13:51JdGordon|and no, I dont tinhk there should be an issue redrawing the list *after* the setting changes.. in fact I would think that would be the better place for it
00:14:12mcuelenaereI suppose list_do_action() does a list redraw?
00:14:23JdGordon|that just clicked
00:14:46JdGordon|selection change -> redraw -> callback -> redraw(?)
00:15:02JdGordon|bugger it.. close it as wont fix
00:15:04mcuelenaerethat's what I meant
00:15:06JdGordon|its not a big deal
00:15:15mcuelenaereor just commit it with the glitch?
00:15:30mcuelenaereI think correct behaviour with a small glitch is better over incorrect behaviour
00:15:37JdGordon|no.. redraws are pointless
00:16:28JdGordon|I havnt looked at that code in ages so cant give any other ideas
00:16:47 Quit truthtaco_ (Read error: 110 (Connection timed out))
00:17:24 Quit webguest ("CGI:IRC (EOF)")
00:17:28 Join balug_ [0] (
00:17:52kugelfreqmod_qu: I deleted 2 or 3 posts between us
00:18:01freqmod_qukugel: ok
00:19:03 Join funman [0] (n=fun@rockbox/developer/funman)
00:19:19kugelmcuelenaere: what glitch?
00:20:05mcuelenaerekugel: add gui_synclist_draw(&lists); to apps/gui/option_select.c after line 517 and go from biggest to smallest scrollbar width
00:20:18 Join bertrik [0] (
00:21:59kugelI can't see that glitch
00:22:13kugeloh I did try something else
00:22:37kugelI think going with F_TEMPVAR is the way to go
00:22:50mcuelenaerethen you don't have live updating anymore
00:22:53 Join webguest [0] (n=04f04edd@gateway/web/cgi-irc/
00:23:20webguesti still can't uninstall rockbox
00:24:31kugelmcuelenaere: in the if ()?
00:24:42JdGordon|mcuelenaere: I thinkk leave it as is... I'll try to remember to have a looktonight... there might be a trick we could use here
00:25:16JdGordon|also, all these status/scroll bar settings should get moved to the themes menu
00:25:21kugelmcuelenaere: ah after
00:25:24 Join fdinel [0] (
00:25:28kugelthere's no glitch here
00:25:49kugelit rapidly gets small, but isn't that what you wanted?
00:25:55JdGordon|go from 3 to max
00:26:02mcuelenaereno, the other way around
00:26:10webguestmy e200r isnt working with the e200 uninstall instructions
00:26:11 Join ehntoo [0] (
00:26:13mcuelenaereyou can't see it from 3 to max as it'll get bigger
00:26:46kugelworks here (e200 sim)
00:26:57mcuelenaereif you go from max to 3, it'll first draw the wide scrollbar with the indicator-thingy pointing below, and then it will shrink to the smaller version
00:27:13*mcuelenaere can't really explain :)
00:27:55kugelit totally looks fine here
00:28:11mcuelenaereit probably redraws faster then you can see on your CPU
00:28:25mcuelenaereanyway, JdGordon| was against that fix
00:28:38kugelif that works, why not?
00:28:52mcuelenaereit's an extra redraw for all option_list's
00:28:52JdGordon|redraw that list twice EVERY selection change
00:29:11kugelremove the other redraw :p
00:29:21JdGordon|its the list handling
00:30:16webguestcant i just delete it from my computer
00:30:47JdGordon|oh for goodness sake.. read the manual.... its as easy as start rbutil and press uninstall
00:31:10kugelJdGordon|: e200r doesn't work with rbutil
00:31:35JdGordon|I thhuoght it was the same once you did the first bit manually?
00:31:38kugelthe manual has the manual instructions though (IIRC)
00:31:58kugelhe wants to uninstall, and that's part of the bits that are different
00:32:27webguestso how do i uninstall the stuff off my e200r
00:32:29mcuelenaereyes==false and no==true in BOOL_SETTING in settings_list.c ?
00:33:08 Quit suom1 ("Irssi = Total pwnage.")
00:33:33kugelmcuelenaere: BOOL_SETTING can be anything, do you mean ONOFF_SETTING?
00:33:57mcuelenaereno, I mean BOOL_SETTING
00:34:13kugelwell, you choose what true and false is in BOOL_SETTING
00:34:42mcuelenaerethe default order when displaying a BOOL_SETTING is first the 'no' option and second the 'yes' option?
00:34:49mcuelenaere(yes and no according to the macro syntax)
00:35:08kugeldo you mean litterally "yes" and "no"?
00:35:13 Nick truthtaco__ is now known as TruthTaco (
00:35:42mcuelenaereI mean the "yes" and "no" in BOOL_SETTING(flags,var,lang_id,default,name,cfgvals,yes,no,cb)
00:36:25kugelno idea
00:38:53webguesti cant find any intructions on the e200r page
00:39:20 Quit roolku ()
00:39:20CIA-61New commit by alle (r22413): Add comment to the BOOL_SETTING macro; use that macro for OFFON_SETTING (FS #10544)
00:41:24kugelwebguest: which page?
00:42:00kugelehh, there's really no uninstallation instructins
00:42:15kugelwhich leads to the question: webguest, why do you want to uninstall? :p
00:42:18webguestso how do i do it
00:42:33JdGordon|mcuelenaere: you maybe should have done the scrollbar setting like ths statusbar... off/left/right instead of 2 settings
00:42:40CIA-61New commit by mcuelenaere (r22414): Scrollbar position: only update it when exiting the menu + FS #10539 by Alexander Levin (change the order of Left/Right)
00:42:41kugelno idea really, have you tried updating with an unpatched original firmware?
00:43:00JdGordon|boooo!!! the live updating was great
00:43:18mcuelenaereJdGordon|: this is the position, not the width ;)
00:43:22kugelthen fix it :)
00:43:23webguesti have to run it in rhapsody mode, which won't allow me to play some songs unless its in playsforsure mode
00:43:32JdGordon|mcuelenaere: ah sweet :)
00:43:37kugelcan't we have a F_LIFEUPDATE flag?
00:43:39*JdGordon| learns to read
00:43:44webguesti dont think i have any unpatched original firmware
00:43:48kugelonly do the redraw in that case?
00:43:49JdGordon|kugel: thats what F_TEMPVAR is
00:43:53mcuelenaerekugel: life update? :)
00:44:07kugelJdGordon|: TEMPVAR does the opposite
00:44:39kugelif (flags & F_LIVEUPDATE) gui_synclist_draw();
00:45:10mcuelenaereis Alexander Levin in here?
00:45:39webguestso how should i uninstall it?
00:45:57JdGordon|no. redrawing twice sucks
00:46:13kugelfor settings that want it (only this one) it's ok imo
00:46:34mcuelenaereJdGordon|: off/left/right sounds great, go for it! ;)
00:46:58JdGordon|I'm at work... all yours :)
00:47:55pamauryI have a question about MTP: did someone ever tried to implement it on rockbox (or is doing so) ?
00:48:19funmanpamaury: tomers did project to do it, not sure how far he went
00:48:33funmanwebguest: looks like nobody present knows, try writing to the mailing list
00:48:58 Join fml [0] (n=4fd3c348@gateway/web/cgi-irc/
00:49:12webguestmailing list?
00:49:28mcuelenaerefunman: I don't think tomers was going to do MTP, just HID + dynamic USB modules
00:49:31fmlmcuelenaere: hello! How about different widths for scrollbars on the main and remote?
00:49:44mcuelenaerefml: hi, could you commit your manual patch?
00:49:44kugelwebguest: download a OF from sandisk and update your e200r with it
00:49:51kugelthat should work
00:50:01webguestokay i'll try it
00:50:14fmlmcuelenaere: I just wanted to ask you whether you have looked into it. Is everything there correct?
00:50:16pamauryAnd is MTP a really wanted feature ? (question for both users and dev)
00:50:23mcuelenaerefml: looks great
00:50:37fmlmcuelenaere: ok, I'll commit it then.
00:50:40 Join suom1 [0] (
00:50:40mcuelenaerefml: how's remote scrollbar width handled now? the same variable width for both of the screens?
00:50:43JdGordon|and yeah, any UI settings shuold be seperate for the remote and main lcd
00:50:52funmando the e200r use the same OF than e200v1 ?
00:51:33mcuelenaerepamaury: there was once a report on the forums of someone looking into MTP for Rockbox, but haven't heard anything about it
00:52:04kugelalso, statusbar and scrollbar really ought to be in theme settings IMO
00:52:33CIA-61New commit by alle (r22415): update for the recently added 'Scrollbar Width/Position' settings (FS #10538)
00:53:11mcuelenaereI just put it there because the original scrollbar en-/disable was there
00:54:19fmlmcuelenaere, JdGordon: yes, some settings should be split to have separate ones for the main and the remote. And maybe grouped differently in the menu.
00:55:07 Quit webguest ("CGI:IRC (Ping timeout)")
00:55:16JdGordon|not some... all
00:55:40 Join polobricolo [0] (n=polobric@
00:55:55fmlJdGordon: scrollbar position could be used for both, no need to split it IMO
00:56:22JdGordon|for consistancy I would sya it should be split
00:56:25kugeljust add theme settings and remote theme settings and split all
00:56:34fmlBut to be a purist...
00:56:44fmlHe-he, you're faster! :-)
00:56:58pamaurygevaerts: are you there ?
00:57:06JdGordon|thats what happens when you dont care about spelling and grammer :)
00:57:16fmlWe could even introduce a struct for that. The question is how this would plug into the settings saving engine
00:57:25JdGordon|kugel: yeah sounds good.. want to do it? :)
00:57:39JdGordon|not easily
00:57:49JdGordon|but yes, doubling up everywhere is annoying
00:57:54JdGordon|and wasteful
00:58:05gevaertspamaury: depends. I'm somewhere anyway :)
00:58:19fmlNot sleeping at this time is also wasteful. Bye!
00:58:20 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
00:59:27 Quit Zagor ("Clint excited")
00:59:52pamaurygevaerts: Shouldn't the prototype of usb_serial_send have a const ?
01:02:57kugelhmm, the cabbie statusbar doesn't look as nice on 220x176 (compared too 176x220) due to widescreen
01:03:29kugelcabbiev2 wastes a lot of space due to the giant rockbox logo in the wps :(
01:04:56pamauryIs it worth implementing MTP on rockbox or is it quite useless/fancy ? It doesn't seem too hard to implement.
01:05:40JdGordon|I tohught it was on the nodo list?
01:05:43funmantomers work should make this easier (dynamic usb drivers)
01:06:01gevaertspamaury: I agree about usb_serial_send. I'll commit your patch soon
01:06:12gevaertsJdGordon|: is it? I wouldn't know why...
01:06:16LambdaCalculus37JdGordon|: I thought that some people weren't really too against MTP.
01:06:19pamaurygevaerts: I submit on FS, anyway it's a two line long patch :)
01:06:25funmanJdGordon|: no it isn't
01:06:30JdGordon|ok ok.. i was wrong...
01:06:38gevaertspamaury: I'd like to see an mtp implementation some day, so if you want to have a go, feel free
01:06:49LambdaCalculus37JdGordon|: IIRC it was discussed that MTP would make it easier for some people to sync with certain media apps.
01:06:59tmztfunman: r has different usb settings, not sure what you mean by same OF, it could be the same build/image or not
01:07:29funmantmzt: sandisk only offer a "e200v1 firmware", no "e200r firmware"
01:07:58JdGordon|they dont have a rasposdy firmware?
01:08:02pamaurygevaerts: I looked at it and did some preliminary test, the protocol simple to implement. The hard work is to support the huge number of operations/properties :)
01:08:16JdGordon|otherwise the e200r and e200 are technically the same
01:10:03 Quit ender` (" The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch a")
01:10:58pamaurygevaerts: Don't know if Itold you: I made tests on usb_drv_arc with the STREAM bit in USBCMD but it didn't seem to change anything. Anyway the RM explanation is quite unclear to me :)
01:12:31gevaertsah, I hadn'e seen that
01:13:16mcuelenaereMTP would make it able to add/remove files while doing something else on the device, right?
01:13:30 Quit jgarvey ("Leaving")
01:14:24gevaertsit should, yes
01:14:29pamauryTechnically yes
01:15:09gevaertsit should also allow us to do realtime database updates
01:15:11kugele200r has the same hardware (same rockbox build), but a different bootloader and installation
01:15:36mcuelenaeredoesn't MTP also have time sync?
01:15:50*mcuelenaere knows it has battery status output
01:17:58pamaurybasically, MTP allows you to read/write/reset nearly every property on any objet (object=file generally). This includes lots of things like date, time, songs properties, ...
01:18:35pamaurythe same applies for the device, which includes of course battery level and playback properties
01:19:27 Join tvelocity [0] (
01:19:40 Quit polobricolo ("Good bye")
01:19:43JdGordon|you can set battery level from mtp?
01:19:48JdGordon|from the computer i mean?
01:20:19tmztmtp should be able to update the secure clock
01:20:34pamauryno you can't set battery level :) But you can read it
01:20:46tmztbut maybe not main rtc
01:21:03gevaertsJdGordon|: yes you can. It's called "charging" :)
01:21:28pamaurywhat is the secure clock ? I don't think I saw this in the MTP spec
01:22:16tmztto do with drm
01:22:49tmztfor subscription content
01:23:29mcuelenaeredoesn't that use the 'normal' RTC on some targets?
01:23:51pamauryThere is a property about DRM: you can check whether an objet is protected or not but that's all iirc
01:27:06funmani'm fed up with this ams guy
01:27:29pamauryDid you get what you want from him ?
01:27:29funmani'm not even wanting to contact him again
01:27:31tmzt“Secure Clock” means a hardware real time clock that has been secured from unauthorized access.
01:27:39funmanno i didn't even get to talk with him
01:27:52pamaurytmzt: why a different one ?
01:27:52funmanhe simply forgot all the meetings we had made (the 3 of them)
01:28:18pamauryHe probably doesn't want to speak with you :)
01:28:22funmanif someone want to take over the communication i'll give you his contact information
01:28:39funmanyou could just say i'm doing something else now
01:29:02tmztanother result of the sesrch suggests as35xx has one
01:29:43mcuelenaeretmzt: that doesn't imply that the time in the OF isn't based on the same rtc
01:30:18funmanthat implies that you wouldn't be able to modify the OF time?
01:30:30tmztI understand it to mean that the device owner can't change the time, on the MSDRM software on a host
01:31:01tmztonly the ...
01:31:05*mcuelenaere is confused
01:31:11pamaurytmzt: I don't think a thing like that in the MTP spec. But there a property called "online subscription id"
01:31:13mcuelenaerenever mind what I said
01:32:38tmztmsybe it's in here: “WMDRM-PD MTP Extensions Technical Documentation”
01:32:54tmztPD is portable devices
01:32:58pamauryAh could be :)
01:34:43pamauryI also have a question about storage in rockbox. When multivolume is enabled, you can get the number of storage medium but is there a simple way to map between disk and volumes ? Ie to get the number of disk and which volume belongs to which disk ?
01:37:38 Quit Thundercloud (Remote closed the connection)
01:40:52pamauryOk, I'll try to find the answer in the code :) Bye
01:41:35 Quit pamaury ("exit(*(int *)0 / 0);")
01:47:34 Quit gevaerts (Nick collision from services.)
01:47:43 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
01:48:01 Join kushalone [0] (n=kushal@
01:50:23 Quit LambdaCalculus37 ()
01:52:41 Quit funman ("free(random());")
01:55:04 Quit balug_ ("Ex-Chat")
01:56:53 Join Person [0] (n=4c412d6c@gateway/web/cgi-irc/
02:00:43PersonUm, excuse me, I put a folder of Ogg Vorbis files on my iPod (Video, 30 GB) and they don't seem to be showing up when I shuffle all. How can I get them to show up in the playlist?
02:01:01 Quit Rondom (Nick collision from services.)
02:01:12 Join Rondom [0] (
02:01:15 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
02:02:13 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-wwscknztftzrcfcf)
02:03:32UnhelpfulJdGordon: did you ever figure out your missing sim AA?
02:05:09 Join Strife89 [0] (
02:08:21 Join JdGordon_ [0] (i=836b004a@gateway/web/freenode/x-bgqozaxdkjzkyyuv)
02:12:30***Saving seen data "./dancer.seen"
02:13:06 Quit JdGordon_ (Ping timeout: 180 seconds)
02:18:39 Join JdGordon_ [0] (i=63cc90a5@gateway/web/freenode/x-qftsicrdnsmndnbk)
02:21:51 Join CaptainKewl [0] (
02:26:13CIA-61New commit by mcuelenaere (r22416): Onda's: add touchscreen boot menu for targets without/not much physical buttons (VX777)
02:35:35 Join froggyman [0] (
02:39:34 Join JucaBlues [0] (n=felipe@
02:42:23 Quit bertrik (Read error: 113 (No route to host))
02:42:47 Quit JdGordon_ (Ping timeout: 180 seconds)
02:43:30 Quit mcuelenaere ()
02:47:09JucaBluesRecently I've been annoyed by gNewSense policies related to mp3 encoder availability and software patents plus my dad owning an ipod model not supported by rockbox which put me in a dead-end.
02:47:15JucaBluesI have read the irc channel guidelines and I hope this is not condidered off-topic. I'd like to discuss rockbox role in this debate.
02:48:40JucaBlueswhen considering my personal choices, I reach the conclusion that I'd like to know which portable music players support ogg natively and are also supported by rockbox
02:48:43kugelwell, you come at a bad time for that
02:49:09kugelit's usually almost completely silent these hours
02:49:14JucaBluesah, ok
02:49:48kugelto your question, the sansa fuze plays ogg in the OF, but is also (almost) supported by rockbox
02:50:53JucaBlueswhat do you mean by "almost" ?
02:51:21kugelthere are a few issues (mainly with some rare microsd), but it runs fine mostly and there's tools and instructions for installing
02:52:44JucaBluesnice. Then this is probably a good ethical choice for me. I want to buy a device from a manufacturer that supports ogg. But at the same time I dont want to use proprietary firmware.
02:54:07kugelalthough there are v2 fuzes yet, which do not run rockbox at all
02:54:51JucaBluesnow... my dad's ipod is not supported at all. Do you know why? It is a nano 2nd gen
02:55:23saratogawork started on that a couple weeks ago
02:55:37saratogasomeone just recently figured out how to run code on it
02:56:06JucaBluesI thought I could eventually buy him a sansa fuze and then try to help nano 2ng gen development
02:56:15UnhelpfulJucaBlues: rockbox is entirely a volunteer effort. for a long time nobody had a means to inject code into nano2g... now that that's happened work has *started* on RB for it.
02:56:56kugelIIRC it played sound already?
02:57:10kugelor was that on the meizu only (which has nearly identical hardware)?
02:58:15JucaBluesbecause my dad's mp3-only ipod forces me to install mp3-encoder in my gNewSense (which is OK legally, because Brazil does not allow software patents, but is ethically controversial)
02:59:51UnhelpfulJucaBlues: i don't see any ethical problems with using LAME, really
02:59:52PersonUm, excuse me, I put a folder of Ogg Vorbis files on my iPod (Video, 30 GB) and they don't seem to be showing up when I shuffle all. How can I get them to show up in the playlist?
03:01:10JucaBluesUnhelpful: I dont wat to support a patent-encumbered file format, even though it is not a local issue in my country, I'd like to think globally and avoid patented tecnologies at all
03:02:24kugelFOSS isn't free of patens entirely either (though rockbox is)
03:04:21saratogalots of rockbox code is patented
03:04:33saratogawe just don't pay any attention to it
03:05:22kugeloh I didn't know that. what code is patented?
03:05:36JdGordonmpeg definetly
03:06:31saratogafat, most of the codecs as well
03:06:56kugelthere's no patents on code in the EU, so...
03:07:19kugelbut what do I know ;)
03:07:38*JdGordon knows the skin display code is a bit convoluted :p
03:08:04 Join Blue_Dude [0] (
03:09:08kugeldoesn't matter, the few external functions are sufficient so far :)
03:09:38kugelI would like to know why the my custtom sb shows the rewind icon initially
03:09:57kugelit doesn't get it from get_token_value()
03:11:11JdGordonsomething == 0
03:16:22 Quit martian67 (Read error: 104 (Connection reset by peer))
03:17:25 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
03:20:19JdGordondid we ever allow viewports to share id's ?
03:22:14JdGordonapparently yes... time to fix the code then
03:25:32kugelJdGordon: you advertised that when you committed it
03:25:51JdGordonyeah, apparently...
03:26:01JdGordonminor oversight in the latest changes...
03:26:06kugelnot a bug
03:30:18JdGordonwhats not a bug?
03:31:34kugelsharing ids
03:31:43JdGordonI know..
03:31:58JdGordonthe bug is forgetting they are allowed to be shared
03:40:10 Quit Strife89 ("Bed.")
03:47:22 Join wincent [0] (
03:48:45*kugel thinks he fixed all id3 == NULL problems
03:49:12JdGordongood good
03:49:18JdGordonmade the tags safe?
03:49:50kugelreturn "n/a" or "0" if !id3
03:50:14JdGordonexcellent :)
03:51:33JdGordondoes the playback status tag retun anything for stopped?
03:52:20 Part JucaBlues
03:52:55CIA-61New commit by jdgordon (r22417): conditional viewports are allowed to share labels.. so make sure they are all toggled with %xd<label>
03:53:24 Quit martian67 (Read error: 113 (No route to host))
03:55:33kugelyes, stopped
03:55:56JdGordonok cool
03:56:05JdGordonso people will have to get used to using it :)
03:56:05 Quit BlakeJohnson86 (Connection reset by peer)
03:57:35 Join BlakeJohnson86 [0] (
03:57:55JdGordonI love how builds are on average faster now but horibble efficiency :)
03:58:10 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
04:00:50 Quit martian67 (SendQ exceeded)
04:02:40*kugel would prefer setting sane defaults for id3, other parts might rely on it also
04:02:44 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
04:04:19JdGordonI'm not sure... I cant imagine that anything else presumes audio is playing
04:05:17pixelmareally id3 as in metadata or just info about the file (I'm wondering about how things like %?ia is treated if it was the former)?
04:06:34kugelwell, both
04:06:37JdGordonthe id3 we are talking about is the track data
04:06:52JdGordontucker time :)
04:12:34***Saving seen data "./dancer.seen"
04:12:47UnhelpfulJdGordon: did we have a clear measure of efficiency before new build system? :)
04:13:42 Quit Person ("CGI:IRC")
04:13:53 Quit kushalone (Client Quit)
04:14:34 Quit Lss (Read error: 54 (Connection reset by peer))
04:16:59 Join Lss [0] (
04:44:06 Join n17ikh [0] (
04:53:26 Quit TruthTaco (Read error: 113 (No route to host))
04:55:57 Join TruthTaco [0] (
04:57:10 Join dmb [0] (n=Dmb@unaffiliated/dmb)
04:59:50 Quit tvelocity (Remote closed the connection)
05:04:36 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:08:28JdGordonI'm doing memory allocation "backwards" in my buffer... starting at the end and moving towards buf[0]... how do I get the 32bit aligned pointer after decreasing the pointer by size?
05:08:38JdGordonlike buffer_pos = (void *)(((unsigned long)buffer_pos + 3) & ~3); but not
05:11:10JdGordonor do I just add an extra 4 to the requested size and then do the above?
05:14:42*JdGordon prods Unhelpful to prove his nick is wrong!
05:16:13 Join polobricolo [0] (n=polobric@
05:17:18UnhelpfulJdGordon: we, um, have a macro to align the start and end of a buffer? ;)
05:17:30JdGordonwe do?
05:17:48JdGordondoes what I asked make sense?
05:22:52JdGordonits not as stupid/simple as just changing that + to a - is it?
05:27:05UnhelpfulJdGordon: you want to align an address down, w/ 4B alignment?
05:27:15Blue_DudeHey, anybody tried to use ccache with cygwin? Does it help?
05:27:22Unhelpfulit's, um, *more* simple than changing the + to a -. ;)
05:27:42JdGordonits just &~3 isnt it? :)
05:28:18Unhelpfulindeed... &~3 clears the bottom two bits, the +3 is needed to change that from a round-down to a round-up :)
05:28:25JdGordonBlue_Dude: amiconn is our resident windows dev dude thingy... I'm pretty sure he at one point did a bunch of cygwin tests
05:28:36JdGordonUnhelpful: thanks
05:28:48 Quit dmb (Read error: 113 (No route to host))
05:28:49Blue_DudeJdGordon: and for sure it didn't work?
05:28:56JdGordonI dont remember
05:29:02JdGordonI think i remember it being slower
05:29:16JdGordonUnhelpful: I'm turning this skin buffer into something quite scary :)
05:29:36Blue_DudeWell, it's out there in corporate wisdom somewhere...
05:30:00UnhelpfulJdGordon: excellent... just make sure that it's not nearly the most clever you can make it. ;)
05:30:09JdGordonhehe I know :)
05:30:13JdGordonI'll try not to
05:30:31JdGordonadding lots of comments just in case
05:40:21 Quit togetic ("WeeChat 0.3.0-rc2")
05:40:39 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
05:47:31 Quit togetic ("WeeChat 0.3.0-rc2")
05:47:43 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
05:51:45 Quit Horscht ("Verlassend")
05:54:57JdGordonUnhelpful: got a min to have a look at my patch to make sure I didnt do something stupid?
05:55:32Unhelpfuli'll regret it if i say "yes", won't i?
05:55:42JdGordonprobably not... tiny patch
05:56:14Unhelpfuljust so it doesn't have any rename-and-then-edit :/
05:56:28JdGordonno, its very clean
05:56:48JdGordonit replaces the tokens staticly sized array with one that bassically grows as needed
05:58:06Unhelpfuls/bare in mind/bear in mind/ ;)
05:58:29JdGordonhave a bear in mind?
05:58:31 Quit togetic ("WeeChat 0.3.0-rc2")
05:58:34Unhelpfulalso begining->beginning :)
05:58:35JdGordonbeer :)
05:58:43JdGordoncode comments... not typos :)
05:58:46 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
05:59:44 Quit kugel (Read error: 60 (Operation timed out))
05:59:55Unhelpfulnow, you load all of the tokens first, and then all of the images?
06:00:33JdGordonas long as that doesnt change (and it shouldnt) that end is safe
06:01:07Unhelpfuland you should then be able to allocate images up, starting from right after the tokens... allocate-down tricks are best avoided *if* you can. :)
06:01:09JdGordonI tihnk the danger is having the buffer_back grow and overlap buffer_start+"the data which hasnt been technically allocated yet"
06:01:21JdGordonyes images go up
06:01:34JdGordononyl stuff which we know the size of can be on the end growing backwards
06:01:53Unhelpfuland really it's better if you don't do that at all :)
06:02:42JdGordonyeah, but we cant
06:02:49JdGordonwithout wasting ram
06:02:59JdGordonor moving ram around which isnt going to happen
06:03:03JdGordondebug.. its gone now
06:03:52Unhelpfulwhat exactly do we need grow-down for?
06:04:15JdGordontokens that need data, the viewports, progressbars, even image structs
06:05:21Unhelpfulfor image structs, can't you grow those up as you load the images? allocate sizeof(struct bitmap) then make the bitmap data follow immediately?
06:06:03Unhelpfulactually, you could probably do that for all of them... load tokens in one pass, then allocate other data they need in a second pass?
06:06:33Unhelpfulwhat's this? static unsigned char buffer[SKIN_BUFFER_SIZE];
06:06:52 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
06:07:27JdGordonno, the stuff which is allocated at the back end is data which tokens point to, its all statically sized so I dont think its bad growing backwards
06:07:48JdGordonthe image is set to the front to grow "up" when the bmp is loaded
06:08:01JdGordonthat is the actual RAM allocated for the skins
06:08:26Unhelpfuland i assume there are no gaps left, in the end, between the grow-up and grow-down parts of the buffer?
06:08:50Unhelpfulthe bits that i understand right now seem sane... mind you, that's mostly the comments. :)
06:08:53JdGordonthere will be untill it fills up
06:09:11JdGordonand we all know that comments always line up with the code... so all is good :)
06:12:18*JdGordon needs to work on the wps debug info
06:12:35***Saving seen data "./dancer.seen"
06:13:58*JdGordon really loves it when stuff works first go!
06:16:06*JdGordon goes for it
06:19:11CIA-61New commit by jdgordon (r22418): move the statically allocated tokens array into the skin buffer. this is done with a little bit of magic with the buffer... explanation in the ...
06:19:26*JdGordon expects a fairly impressive delta on that one :)
06:37:47 Quit BHSPitLappy ("Ex-Chat")
06:41:35CIA-61New commit by jdgordon (r22419): show the skin buffer size as well as the used amount in the rockbox info screen
06:43:46 Quit martian67 (Remote closed the connection)
06:44:18 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
06:46:50 Quit martian67 (SendQ exceeded)
06:48:46 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
07:06:21 Quit CaptainKewl (Remote closed the connection)
07:07:16CIA-61New commit by jdgordon (r22420): get the skin debug output working again, add a line to say how much of the buffer is being used to aid themers... let me know if this gets close to ...
07:09:15 Join webguest49 [0] (n=47ffaf1e@gateway/web/cgi-irc/
07:09:52webguest49COWON D2 - hi guys, do yo think Rockbox would be fully functional on Cowon D2 by 2010?
07:11:38JdGordonwe are presently look at a spring 2013 release date, we wont bow to marketing pressure
07:13:37martian67JdGordon, might get pushed back to 1015
07:13:40martian67JdGordon, might get pushed back to 2015
07:13:44martian67funding delays :/
07:13:50pixelmawebguest49: no-one can predict that. Rockbox is a volunteer effort and you can't make people work on something, so can't and won't give time estimates. The D2 port is more or less a one person project so even harder to predict... it would help if you find interested coders with some skills
07:13:51martian67you know how head office is
07:15:01CIA-61New commit by jdgordon (r22421): hopefully fix the yellow
07:17:07webguest49What is the best way to fund Rockbox transfer on D2?
07:17:50 Quit Lss (Read error: 104 (Connection reset by peer))
07:19:11pixelmathere is no funding for specific ports -as I said your best bet would be to gather interest and find other coders that are really able to help
07:20:16pixelmain case you find an interested one and he doesn't have a D2 yet, you could fund him by giving him a D2
07:20:22 Quit webguest49 ("CGI:IRC")
07:28:13 Part wincent ("Kopete 0.12.7 :")
07:30:43CIA-61New commit by gevaerts (r22422): Add const to usb_serial_send() buffer. ...
07:32:25 Join stoffel [0] (
07:33:27 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:38:51pixelmaJdGordon: can't find something obvious at a first glance with my c200 WPS, Skin RAM usage is 5.35kB/23.2kB if you are curious (not many images, the few it has are monochrome, but some 20 viewports)
07:40:04JdGordonI think we can make a case for shrinking the buffer down a bit, but I dunno if its worth it or not
07:40:49JdGordon20kb isnt exactly a huge amount of compressed audio...
07:40:56*Unhelpful thought the point was to make the buffer dynamic...
07:41:20JdGordonyes and no
07:41:30JdGordonthe main point was to use the ram better
07:42:33JdGordonI am thinking about adding a user setting for how much to waste, but probaly will get rejected.. "as small as possible" "tiny", "small", "medium", "huge" etc...
07:43:29*Unhelpful thinks that if dynamic can be done without onerous requirements, that's the best use ;)
07:44:04JdGordonwe can only go smaller between reboots, so some argue there is no point making it user choosable
07:44:42Unhelpfulbuffer_alloc space can't be returned without reboot?
07:44:57JdGordonit will get very interesting when the statusbar and fm are skinned though
07:45:49JdGordonwell, in theory the last one alloced could be shrunk. but no way will we do that
07:51:43Unhelpfulbecause it could just end up being broken when something else comes along to use buffer_alloc after it?
07:53:11*Unhelpful will have to think about incorporating resize of buffer_alloc'd buffers into his Grand Not-A-Malloc Buffer Scheme ;)
07:53:41Unhelpfulremoving the reboot requirement for settings that change buffer_alloc use would be a nice-to-have, also. :)
07:54:20JdGordonyeah, I doubt it will ever happen though... there is just too much that uses it
07:54:54JdGordonI thought about using an event which would get everything to realloc what it needs, but thats not safe
07:54:59JdGordonor good
08:01:13Unhelpfulso... i've used 1.3KB of... 168KB total?
08:01:53JdGordontext only wps?
08:03:50Unhelpfuli believe i've a bitmap progressbar... but it's only 2px high or so, and mono as well.
08:05:10JdGordonso 1.3K is expected
08:05:27 Join ender` [0] (
08:05:47JdGordonhmm... is it a .bmp progressbar? or the inbuilt one?
08:05:53pixelmaUnhelpful: don't you have big album art? Is that counted somewhere
08:06:04JdGordonthats not part of the skin count
08:06:23UnhelpfulAA bmps are stored as buffer objects :)
08:06:50pixelmayou couldn't see progress on a 2px high inbuilt progressbar as that could only display its frame
08:07:18*Unhelpful points up re: UnheplfulAA's progressbar
08:08:09JdGordonoh right
08:08:15JdGordonyeah 1.3K sounds correct
08:08:19*pixelma thinks it would be nice if "make reconf" could deal with advanced build options (better)
08:09:10Unhelpfuland the 168K comes from where?
08:11:01JdGordonenough to store 1 full screen 24bit image, and 1 half screen mono i tinhk
08:11:27JdGordonno, 1 24bit and 2 mono images
08:11:50JdGordonskin_buffer.c if you want to manually change that
08:11:52Unhelpfulheheheh.... pb.bmp is the only theme bitmap, and it's a whopping 186B
08:12:38***Saving seen data "./dancer.seen"
08:12:53 Join moos [0] (i=mustapha@rockbox/staff/moos)
08:38:21 Join Grahack [0] (
08:38:35 Join JucaBlues [0] (n=felipe@
08:39:02JucaBlueswhat are the tech requirements for running rockbox?
08:39:41GodEaterJucaBlues: what do you mean ?
08:40:06GodEateryou need a supported player, and you need a PC that is running Windows / OSX or some flavour of linux to do the install
08:40:07JucaBluesI mean... if I have a portable music player that is not officially supported yet, how should I proceed in order to figure out if it is even possible to port RB ?
08:40:09GodEaterthat's about it
08:40:16GodEaterah I see
08:40:29JucaBlueswhat are the common code injection procedures?
08:40:34JucaBluesused to install RB
08:40:40GodEaterthat varies entirely from player to player
08:40:47GodEaterhave a read of that
08:42:17GodEaterwhich player are you looking at out of interest ?
08:42:36 Quit kachna (Read error: 60 (Operation timed out))
08:43:59 Join kachna [0] (n=kachna@
08:45:34 Join petur [50] (n=petur@rockbox/developer/petur)
08:52:18 Join Rob2222 [0] (
09:01:54 Quit Stephen__ ("Leaving")
09:11:14 Quit Rob2223 (Read error: 110 (Connection timed out))
09:15:42 Join lasser [0] (
09:24:47 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:27:19 Join Thundercloud [0] (
09:29:14 Part JucaBlues
09:39:50 Quit lasser ("ChatZilla 0.9.85 [Iceweasel 3.0.6/2009072221]")
09:45:01 Quit Grahack (Read error: 113 (No route to host))
09:45:03 Join mt [0] (n=MTee@rockbox/developer/mt)
09:45:05 Quit Thundercloud (Remote closed the connection)
09:48:14 Join Grahack [0] (
09:59:45 Join flydutch [0] (
10:08:34 Quit bubsy ("I'll be back somewhere in time...")
10:12:40***Saving seen data "./dancer.seen"
10:17:04 Join bubsy [0] (i=Bubsy@
10:35:49 Join decayedcell [0] (
10:42:52 Join daggett [0] (
10:51:05 Quit Jester (Read error: 104 (Connection reset by peer))
10:56:19 Join pamaury [0] (
10:56:35 Quit froggyman (Read error: 113 (No route to host))
11:01:08 Join einhirn [0] (
11:11:00 Join bertrik [0] (
11:13:23 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
11:13:26 Quit daggett (Read error: 113 (No route to host))
11:25:25 Quit polobricolo (Read error: 110 (Connection timed out))
11:54:01CIA-61New commit by alle (r22423): More comments in the settings macros
12:07:38 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
12:12:43***Saving seen data "./dancer.seen"
12:20:53 Join langzeitstudent_ [0] (
12:57:23 Join kugel [0] (n=kugel@rockbox/developer/kugel)
12:58:38 Join pyro_maniac [0] (
13:07:47pyro_maniaci was trying to compile sansapatcher with this result: make: *** No rule to make target `firmware.mi4', needed by `bootimg_c200.c'.
13:08:41CIA-61New commit by alle (r22424): Fix typo in the comment
13:11:55kugelpyro_maniac: you need the c200 bootloader image
13:12:23pyro_maniackugel: but i want to compile it for a e200
13:12:38kugelsansapatcher needs both
13:13:47 Join fml [0] (n=4fd3ff95@gateway/web/cgi-irc/
13:14:02 Join andrewRB [0] (n=d4dbf7d2@gateway/web/cgi-irc/
13:14:11 Quit andrewRB (Client Quit)
13:14:20fmlIs it possible to switch the scroll bar on/off on non bitmap LCDs?
13:16:59pamauryhello, I have a new question about storage: does rockbox support multiple partition per drive ?
13:18:45 Quit fml ("CGI:IRC (EOF)")
13:23:24n1spamaury: you might need to enable it witha a define
13:24:06GrahackHi, some of my playlists have a file name beginning with an underscore and used to be sorted at the end of the catalog. Is it me or this behaviour changed recently (say within a month) ? Is there a mean to sort them at the end back ?
13:25:00 Quit ehntoo (Read error: 60 (Operation timed out))
13:27:16 Quit YpsyZNC (Read error: 60 (Operation timed out))
13:27:34pamauryn1s: so it supports it ? I mean, for example, if on disk has two partitions, is it possible to access the two filesystems with rockbox ?
13:28:49 Join tvelocity [0] (
13:29:38 Join l403 [0] (n=l@
13:30:17Grahackfrom You can partition your player fine, just make sure the first one is FAT32 and then make the other ones whatever file system you want. Just don't expect Rockbox to understand them.
13:31:19pamauryYes but if other partitions are FAT32 ?
13:31:39 Join YpsyZNC [0] (
13:32:53GrahackGood question, I was just realizing that your question may be about FAT32 only. I have no answer, but nls talked about a define...
13:33:33kugelpamaury: rockbox should support that
13:35:06 Join ehntoo [0] (
13:35:12pamaurykugel: And how exactly it works ? I saw that there is the "storage" api in RB but I belive a storage=disk, no ? So how can you get the number of partitions per drive and how do you access them ?
13:39:09pamauryLooking at the code, I begin to understand (I hope !): there are volume which are logical drives and contains filesystem and each drive can coutains several volumes, am I correct ? Them is there a way to know which volume belong to which drive and conversely ?
13:41:34kugelpamaury: ask gevaerts about the specifics, but i had this commit in my mind: r21933:
13:41:48kugelthere are volumes and drives, yes
13:48:55pamauryok ,thanks. I'll ask gevaerts later
13:50:07 Join AndyIL [0] (i=AndyI@
14:00:45 Part decayedcell
14:00:59 Quit AndyI (Read error: 110 (Connection timed out))
14:06:53 Quit pamaury ("exit(*(int *)0 / 0);")
14:06:57 Join pamaury [0] (
14:07:00 Join decayedcell [0] (n=decayed_@
14:07:13 Part decayedcell
14:10:54l403looking at I see the settings_list task has been already done so another task can be added for someone to mark it as done
14:12:44***Saving seen data "./dancer.seen"
14:14:02 Join decayedcell [0] (
14:17:09 Part decayedcell
14:20:53 Quit l403 ("Leaving channel")
14:21:54 Quit ehntoo (Read error: 110 (Connection timed out))
14:24:37 Join funman [0] (n=fun@rockbox/developer/funman)
14:36:43CIA-61New commit by alle (r22425): Combine the settings for scrollbar on/off and its position into one with three possible values
14:38:30CIA-61New commit by alle (r22426): Delete the trivial comment
14:40:35 Join aaron424 [0] (
14:42:32 Join teru [0] (
14:44:01 Join robin0800 [0] (
14:44:48langzeitstudent_Hmm my player (cowon D2) does not have a "play" key. Now I can not say "yes" to any question (e.g. "erease dynamic playlist")... is there any way to get around this without messing with the source code...? :/
14:46:55bertrikIIRC, some targets use a PLAY button to confirm, others use SELECT
14:47:27bertrikSELECT being the middle button in some kind of up/down/left/right button layout
14:48:16bertrikchanging between the play button and the select button for confirmation actions can be done quite easily in the language definition files
14:48:53 Quit funman ("free(random());")
14:49:14langzeitstudent_ah, thought it'd be like the stuff mentioned in the UI simulator button stuff (
14:50:50 Quit flydutch ("/* empty */")
14:52:08 Quit pamaury (Read error: 113 (No route to host))
14:52:10bertrikyeah, it's not just the language file
14:52:18langzeitstudent_Hrm.. the d2 is not mentioned in the list.
14:53:02langzeitstudent_which means "*: "PLAY = Yes"" is used :/
14:53:16langzeitstudent_or is that just the text to display? :P
14:55:49 Join daggett [0] (
14:57:22langzeitstudent_ah found it, i guess. but that means i really do have to get my crossdev toolchain working.. bah
14:58:19 Join dfkt_ [0] (
15:00:51 Quit Bagder (Read error: 110 (Connection timed out))
15:02:19teruthe button to say yes seems to be TOPRIGHT for touchscreen targets.
15:04:29langzeitstudent_the current array for "static const struct button_mapping button_context_yesno[]" is empty (at least in my "trunk" checkout); i just added "{ ACTION_YESNO_ACCEPT, BUTTON_MENU, BUTTON_NONE }" to the list. i wonder if I can check it with the ui simulator?
15:06:22 Join ZincAlloy [0] (n=d9eef3a4@gateway/web/cgi-irc/
15:08:58kugelteru: hi
15:09:21kugelthanks for the fixup of the anti-aliased fonts patch. it seems a lot slower than previous versions though
15:10:14ZincAlloykugel: did you use a smaller font for the new version of "Slick" on purpose?
15:10:42kugelZincAlloy: yea, I noticed the wps has info that isn't shown with the bigger font
15:11:30ZincAlloy:D I left that in because it kinda worked with a smaller font
15:12:39ZincAlloybut most of the information will need some nudging..
15:12:55 Join decayedcell [0] (
15:13:12kugelmy wps screenshot is a bit messed up :(
15:13:20 Quit daggett ("Ex-Chat")
15:14:33kugelZincAlloy: have you seen my screenshot with the custom statusbar at ?
15:14:49ZincAlloyyes, very nice
15:15:47kugelcabbie isn't so well prepared for that sort of stuff :/
15:15:54 Part decayedcell
15:16:00ZincAlloyit doesn't need that sort of stuff
15:16:22 Quit aaron424 ("ChatZilla 0.9.85 [Firefox 3.0.13/2009080315]")
15:16:24GodEatercabbiev2 you mean :)
15:16:30ZincAlloyright :)
15:16:46ZincAlloybut the placement of viewports leaves a lot to be desired on many targets
15:16:47 Quit tvelocity ("Αποχώρησε")
15:16:51 Quit dfkt (Read error: 110 (Connection timed out))
15:17:23kugelsure it needs that, every theme does :)
15:18:09ZincAlloyI guess this new feature will be overused in no time :)
15:18:22langzeitstudent_bertrik, it works now, at least in the ui simulator. is that enough to create a bug & append patch?
15:18:57kugelthe fine thing is, you can use every wps tag, so you can make it exxagerated or minimal :p
15:20:18ZincAlloybtw, I believe there's some themes with licensing issues on the theme site
15:20:26 Join daggett [0] (
15:20:52kugelIMO, the default theme should make use of that
15:20:53ZincAlloythat series of win xp/media player themes for nano...
15:27:11 Quit antil33t (Read error: 104 (Connection reset by peer))
15:27:25 Join antil33t [0] (n=Mudkips@
15:28:23CIA-61New commit by blue_dude (r22427): Resort, upnumber plugin API
15:29:41 Quit pyro_maniac (Read error: 60 (Operation timed out))
15:30:36terukugel: hmm. I use anti-aliased font with gigabeat x and I don't feel it's slow... so i don't feel like to try to speed up.
15:33:05kugelit's slower than the versions before on my fuze & e200
15:33:18kugelso I suspect your changes
15:33:41linuxstbteru: The gigabeat X is one of the fastest devices...
15:33:54teruthere seems to be some strange changes in my patch compared to original patch. reverting them to original might help?
15:34:37kugelI will try, what exactly is unsupported in the versions before? I've never noticed something that the AAF cant do
15:38:14 Quit faemir (Read error: 104 (Connection reset by peer))
15:39:53 Join faemir [0] (n=faemir@
15:41:21ZincAlloykugel: here's an improved version of the slick wps: changes are subtle, but it looks much better now
15:44:08teruthe only reason i tried to support all drawmode is text in lrcplayer, plugin I wrote, wasn't shown as I expected. i don't remember reason of other changes.
15:47:04CIA-61New commit by teru (r22428): shortcuts_view: fix displaying last path segments so that plugin exactly shows what is said in the manual. ...
15:49:36kugelZincAlloy: your theme, I won't argue :) but maybe base it on my version (so that the version number increases properly) I haven't made any changes to the wps
15:50:50 Part LinusN
15:51:47ZincAlloyI did, I only tweaked the placement of viewports in the wps a bit and updated the version number :)
15:52:20ZincAlloyI kinda like the smaller font ;)
15:52:37kugelZincAlloy: oh, I uploaded the wrong .wps
15:53:15 Quit ZincAlloy ("CGI:IRC")
15:53:20 Join ZincAlloy [0] (n=d9eef3a4@gateway/web/cgi-irc/
15:53:49ZincAlloyI see :)
15:55:09bertriklangzeitstudent_, I'm not _really_ a button expert, but a flyspray patch sounds like a good idea
15:55:18kugelZincAlloy: do you want to reupload?
15:55:37ZincAlloyright, I'll do it..
15:57:47CIA-61New commit by teru (r22429): brickmania: fix position of score in function brickmania_sleep.
15:59:47langzeitstudent_bertrik, well its up:
16:00:45kugelWhy top right?
16:01:01langzeitstudent_first time i ever created a patch, hope it works xD
16:01:29langzeitstudent_kugel, dunno. the keymap-touchscreen uses it
16:01:44kugelisn't center basically select?
16:03:05kugellangzeitstudent_: the keymap-cowond2.c isn't meant for BUTTON_TOPRIGHT etc (i.e. the grid mode buttons)
16:04:15kugelfor the grid mode, we have keymap-touchscreen.c which already has yesno_accept (also top right).
16:04:33langzeitstudent_ah. well the touchscreen keymap wasnt used, though. well i'll remove the topright thingy then
16:05:28kugelare you sure? you need to turn the grid mode (in the settings) on for that
16:05:58langzeitstudent_is there a way to replace attached files?
16:07:10kugelnot for normal people at least :)
16:07:57ZincAlloykugel: ok, new version is up
16:08:49kugelZincAlloy: I can delete the other one, right?
16:09:25ZincAlloysure :)
16:10:21langzeitstudent_kugel, 'kay i'll add another file then ah, just checked again; the topright thingy works. i wonder why it didnt before, had the gridmode on.
16:11:23kugelthe yesno screen ought to be remade for touchscreen targets
16:11:31langzeitstudent_"normal people" cant edit their reports either?
16:11:43kugelthat's correct
16:12:15kugelwith a nice big green button in the middle and red around, so that you touch the actual "I agree" button :>
16:12:47***Saving seen data "./dancer.seen"
16:14:01ZincAlloythanks alot, kugel!
16:15:12kugelyou're welcome :)
16:16:17langzeitstudent_alright.. 2nd version up then
16:17:08kugelnow both top right and menu work (in grid mode)?
16:19:19 Quit ZincAlloy ("CGI:IRC")
16:20:55 Nick YpsyZNC is now known as Ypsy (
16:26:02 Join pamaury [0] (
16:38:12 Quit ender` (" I was in the grocery store. I saw a sign that said "pet supplies." So I did. Then I went outside and saw a sign that said "")
16:38:14 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
16:40:51 Join toffe82 [0] (n=chatzill@
16:41:34 Join Lynx_ [0] (
16:42:36 Join ender` [0] (
16:47:51 Join BobSmith [0] (n=c73d19fe@gateway/web/cgi-irc/
16:50:51 Quit BobSmith (Client Quit)
16:51:59pamauryIs it normal that a recent commit broke the code ? (I get LANG_SKIN_RAM_USAGE undeclared)
16:53:11 Join BBBradley [0] (n=c73d19fe@gateway/web/cgi-irc/
16:53:27 Quit n17ikh (Read error: 104 (Connection reset by peer))
16:53:55 Nick dfkt_ is now known as dfkt (
16:54:19JdGordonpamaury: do make clean
16:54:51 Join froggyman [0] (
16:55:09pamauryarf, ok I'll need a full build
17:01:37 Join n17ikh [0] (
17:02:32kugelhmm, statusbar is 1.4k binsize, 330bytes of that is fixing tokens to be id3==NULL safe
17:04:12 Quit ch4os ("Lost terminal")
17:04:40 Join ch4os [0] (n=ch4os@gentoo/user/ch4os)
17:05:31 Join dberg918 [0] (
17:07:24JdGordonkugel: and thats somehow a problem?
17:07:40JdGordonI just freed up 4K on just about all targets
17:07:47dberg918just noticed that the "customlist" patch got committed, is there a wiki page somewhere with instructions for using it?
17:07:51kugelno, I was just saying
17:08:01dberg918I couldn't find anything
17:08:10kugelthere's another 1.8k that may be freed
17:08:18kugelafter that wps_data is rather tiny :)
17:09:30BBBradleyIf anyone here plays with USB charging, I am trying to disable USB charging on the Ipod Video.I want to plug my Ipod into another portable device and not use any power from it. So I am trying to figure out if I can control the suspend pin on the LTC4066 from a GPIO pin on the PP5021c
17:10:05 Join jgarvey [0] (
17:11:18kugeldberg918: see forums
17:11:40TorneBBBradley: I doubt we know
17:12:56TorneBBBradley: we know which GPIO is the 100/500mA selector. Maybe it's near that? :)
17:13:15dberg918ah, thank you kugel
17:13:27BBBradleywhich GPIO is that?
17:14:57BBBradleyI have been playing with this code and wondered if it could turn off USB Charging, so it just switches between 100/500 mA charging?
17:15:10TorneWhat do you mean?
17:15:26TorneThe patch in FS #8802 only toggles HPWR on the LTC4066
17:15:51Torneoh, wait, no it doesn't
17:16:23TorneActually I think it might already be doing hwat you want, or at least it might have been believed to
17:16:26Tornehave you tried it?
17:17:47 Quit dberg918 (Remote closed the connection)
17:17:59BBBradleyI cut the USB cable and inserted a current monitor and yes have tried the code but it has not dropped to zero.
17:18:30TorneIt's unlikely to drop to zero even if it's *not* charging..
17:18:41Tornethe USB PHY itself might draw a little anyway
17:18:42BBBradleyI see
17:18:44Tornehow much does it draw?
17:18:57BBBradleyabout 31 mA
17:19:11Tornewhat state was the battery in at the time?
17:19:14Tornefully hcarged already/
17:19:17 Quit FOAD ("I'll be back")
17:19:29BBBradleyThe battery is fully charged
17:19:40TorneWell, you can't test it like that ;)
17:19:52bertrik30 mA is about the normal idle current draw of many players
17:19:54Torneonce th ebattery is fully charged it will draw less *anyway*
17:20:02Torne30mA is indeed about right for idle draw for the ipodvideo
17:20:08Tornewhich suggests it is still powering itself off the usb port
17:20:38BBBradleywell is there a way to not draw any from the USB port? Is the SUSP controllable on the 4066?
17:21:07TorneWe have no idea
17:21:22TorneWell, I have no idea and I can't find anything on the wiki that indicates anyone else does
17:21:25BBBradleySo if I figure out do you want to know?
17:21:28Tornedreamlayers might, potentially
17:21:39BBBradleyI think we have to cut one apart to find out.
17:21:59TorneIf you can find definitive answers for which GPIOs are connected where for anything w edon't already know we'd like to know :)
17:22:50 Quit teru ("Quit")
17:23:16 Join FOAD [0] (
17:23:20 Quit mt (No route to host)
17:23:50BBBradleyTorne: is there a document of the known GPIO from the portal player?
17:29:46TorneBBBradley: not sure i'm afraid.
17:33:02linuxstbBBBradley: I think there was a list maintained on the ipodlinux wiki, but that site is down...
17:36:23amiconnJdGordon: You freed ram though, which is good, but binsize increased
17:38:05 Quit petur ("work->home")
17:41:46BBBradleyTorne: I am checking to see if the LTC4066 SUSP is connected to the PP5021c, from the 4066fc.pdf document it shows that it is asserted high and has an internal pull down. So this means it would not need to be connected to anything to use the chip, that power flows by default.
17:42:29Torneyes. this is not surprising :)
17:42:41Torneit is quite possible that it isn't connected, therefore.
17:43:09 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
17:43:50Torneare you really in a position where it's going to hurt that much to have <=100mA being drawn? :)
17:45:40 Join flydutch [0] (
17:52:40BBBradleyTorne: I suspect yes. I don't want to use any of the battery power in the device I'm plugging into.
17:53:58Torneyes. but what i'm saying is it's probably tough, and even if not then it will likely be very difficult to discover how to stop, so maybe you should reexamine how much you care about the <=100mA in the light of that
17:55:43BBBradleyI might have to go with a bigger battery then.
18:07:26BBBradleyTorne: why is the ipodlinux wiki down? will it be up again ever?
18:08:18TorneNo idea, and probably not
18:09:10BBBradleyHow old is Rockbox and are the people who started it still working on it?
18:11:31Torne~8 years, and i think so
18:12:49***Saving seen data "./dancer.seen"
18:19:08 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-aqxzplistfdjcbhz)
18:21:31 Part Grahack
18:22:18linuxstbBBBradley: I believe it was hacked, so they turned the server off and they haven't got round to restoring it yet.
18:25:29 Join white_magic [0] (
18:26:11 Join HBK- [0] (
18:28:07 Quit LambdaCalculus37 ("Fwump")
18:29:47 Join domonoky1 [0] (
18:37:36white_magicdoes Rockbox support xvid or divx?
18:39:37BBBradleylinuxstb thanks
18:42:23 Join daurn| [0] (
18:42:32kkurbjunwhite_magic: not currently
18:46:15 Quit daurn (Read error: 60 (Operation timed out))
18:47:31 Quit KBH (Read error: 110 (Connection timed out))
18:48:13white_magiccan you guys recommend a particular player to use with rockbox? i only care about price and battery lifetime
18:48:18 Quit domonoky (Read error: 110 (Connection timed out))
18:50:29mooswhite_magic: depending on what's you are searching for. Here is an usufull link:
18:50:38 Quit daggett ("Ex-Chat")
18:54:13 Join Mags81 [0] (i=18f2ef5f@gateway/web/freenode/x-ezutpejkmsoewmpi)
18:54:43Mags81yo fellas, any updates on rockbox development for Sansa Fuze v2 devices?
19:00:01saratogacheck the wiki
19:00:09saratogabut probably no unless you haven't looked in ages
19:00:59white_magicthanks, moos
19:07:32 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:08:10 Quit kugel (Read error: 110 (Connection timed out))
19:08:56 Quit Mags81 ("Page closed")
19:13:20 Join tvelocity [0] (
19:15:05white_magicfor the battery lifetime specs in the Buyer's Guide, is it for audio or video?
19:15:23JdGordon|audio even
19:18:06white_magici see :(
19:19:42JdGordon|if you want video you're a bit lost..
19:19:51JdGordon|unless you enjoy converting to mpeg2
19:25:28bertrikgetting the meizu m6 display to work looks doable so far
19:27:06JdGordon|arg, this "nothing to resume" splash on startup is really bloody annoying :p
19:34:30ej0rgeJdGordon|: that's why i use "previous screen" instead of "resume playback"
19:34:44JdGordon|smarty pants :)
19:35:08*JdGordon| forgot about that option
19:39:50 Join mt [0] (n=MTee@rockbox/developer/mt)
19:49:43 Join petur [0] (n=peter@rockbox/developer/petur)
19:50:20 Quit stoffel (Read error: 104 (Connection reset by peer))
19:53:35bertrikwow, the meizu m6 can do quite advanced stuff in the lcd controller already with back and foreground buffers, OSD and alpha blending
19:56:36JdGordon|you found a datasheet?
19:58:10bertrikthe meizu m6 uses the lcd controller inside the s5l8700 and we have a datasheet of that
19:58:11 Join mcuelenaere [0] (
19:59:15JdGordon|ah cool :)
20:02:23 Quit robin0800 (Read error: 110 (Connection timed out))
20:05:51bertrikthe built-in lcd controller is a dma bus master, so lcd updates will probably be very fast (as fast as a memcpy)
20:08:09amiconnThe problem with special lcd controller features is that they aren't portable
20:09:34 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
20:10:24amiconnSo they could only be used on certain targets, and that means either restricting usage to plugins which would run on those targets only, or even worse, introducing lots of ifdefs in the core
20:10:40bertrikyeah, the lcd controller frame buffer is already in system memory IIUC, so we'll both a "rockbox frame buffer" and the lcd controller framebuffer in memory
20:11:07amiconnI'm not referring to special ways to speed up updating, but things like hardware transparency, multiple buffers etc
20:11:27 Join stoffel [0] (
20:11:40JdGordon|wouldnt it make sense to use those extra buffers instead of RAM anyway? even if there is a few extra #ifdefs?
20:11:47JdGordon|and the mr500 can do OSD also iirc
20:12:21amiconnProbably the earliest example was the old 'oscillograph' plugin. Originally written for the archos bitmap lcd, it used a special hardware scrolling feature, making it unportable to the irivers (which were newly added back then)
20:12:52***Saving seen data "./dancer.seen"
20:13:14amiconnInstead of working around, 'oscillograph' was later replaced by the more flexible 'oscilloscope' which uses framebuffer based scrolling, and hence can work on all bitmap targets
20:13:49CIA-61New commit by mcuelenaere (r22430): Dutch translation: be consistent w.r.t. scrollbar
20:17:16mcuelenaereJdGordon|: yes, the DM320 has some 2D acceleration features and the hardware has its framebuffer in main RAM; so there are 2 framebuffers (I remember mapping the lcd_framebuffer[][] as the hardware framebuffer on the ZVM, but that gave the effect like if you would do lcd_update() after every lcd_* operation)
20:20:51 Join Thundercloud [0] (
20:22:25n1samiconn: that's true for any special hardware features and of course the advantage of using the feature should be weighted against the code clutter imo
20:23:24amiconnIf an lcd controller would e.g. offer a yuv overlay, it should of course be used for lcd_blit_yuv
20:24:02amiconnThis wouldn't require any ifdefing, and the advantage of using it would be quite high
20:24:55n1syeah, do yo know what kind of magic the beast' lcd controller could do, i think i remember something about colorspace conversion
20:24:59JdGordon|but wouldnt the ifdef for these features be really low down?
20:25:36saratogaIIRC it can do colorspace conversion and hardware resizing
20:26:23n1sMrSomeone should look into using that :)
20:27:55saratogafirst he should get teh S to be a supported target
20:28:27saratogado we have a TODO list for the S somewhere? I've forgotten where we stand on it
20:29:24*Torne writes himself a note to remind him to actually reverse more of the S bootrom
20:30:02saratogaTorne: were you working on reflashing the S bootrom?
20:30:34Tornei'm working on grokking what it does, for several reasons. reflashing could be one, but the more immediate two were "why do later firmware versions not seem to work with a single boot bootloader" and "why doe sit reformat sometimes"
20:30:55amiconnJdGordon: That very much depends on what kind of feature it is. The mentioned hardware scrolling (or hardware rotation rather) wasn't transparent at all, because it needed a special function in the lcd driver setting the amount of rotation
20:31:40amiconnThe plugin (oscillograph) had to call this, and then take it into account when placing other elements (i.e text) as well
20:32:09amiconnIt also had to reset it on exit, otherwise the whole ui would stay rotated until shutdown
20:35:03saratogaTorne: is reflashing it feasible? is there a way to recover from a bad flash (JTAG?)
20:36:42 Quit stoffel (Remote closed the connection)
20:47:07 Quit kkurbjun (Read error: 110 (Connection timed out))
20:48:15 Join kkurbjun [0] (
20:48:44n1ssaratoga: i don't know what is needed either but i'm guessing installing from windows is pretty involved still?
20:48:44amiconnTorne: Extended question: why does it reformat on some units if it detects a broken firmware, but not on others?
20:49:24amiconnMine complained about a broken firmware once so that I had to restore it, but my data partition was kept intact
20:51:04 Join Strife89 [0] (n=michael@
20:53:08kkurbjunthe MR500 already takes advantage of hardware transparencies and YUV frame buffers. It uses them in the mpeg player to zoom the yuv frame buffer and uses the RGB overlay to show the video progress
20:53:25kkurbjun(and other bits of status)
20:54:43kkurbjunthere is also DMA setup to make lcd updates extremely fast, but unfortunately it can't rotate while updating
20:55:00kkurbjunso that would force it to 480x640 when the natural player orientation is 640x480
20:55:20kkurbjunI get over 150fps when running at 480x640, and around 20 fps at 640x480 :-D
20:55:37 Join Lss [0] (
20:59:09amiconnWould it be deisrable to use a transposed framebuffer on the mr500 then?
20:59:43amiconnIt would mean updates will be >150fps, but native bitmap drawing would be slower (no more simple memcpy()ing)
21:00:21kkurbjunwouldn't that require changes to the way that all of the lcd routines work?
21:00:57amiconnMono bitmaps (i.e. text) need special handling anyway. It would also mean to introduce an new framebuffer-level driver, something like lcd-16bit-t.c
21:01:07JdGordon|480x640 is portaight yeah?
21:01:16 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
21:01:24linuxstbn1s: I _think_ beastpatcher.exe makes installation from windows OK now. I did write a to-do for the Beast a long time ago, probably on the GigabeatSPort page
21:01:32amiconnHmm, in fact native bitmaps could stay fast as well - they would be in a different native format as well in this case
21:01:58kkurbjunoh, true, and I guess loading images would jsut require the transpose at load time
21:02:25kkurbjunJdGordon, yes, it is portrait mode
21:02:26amiconnImage loading needs to process things anyway
21:02:47kkurbjunwhat about a plugin like stars?
21:02:47 Quit white_magic ("Leaving")
21:02:54amiconnWhat about it?
21:03:07linuxstbn1s: It seems that getting beastpatccher to run on OS X is troublesome, but I don't think anyone would say that alone should delay "supported" status.
21:03:11kkurbjundoes it draw direct to the framebuffer?
21:03:15JdGordon|kkurbjun: ok, seems a bit funny that its set up with fast updates in portait when the OF is landscape :)
21:03:36n1slinuxstb: i agree with that
21:03:48amiconnkkurbjun: stars? Why should it draw directly?
21:04:00kkurbjunwell, I don't know if the OF ever took advantage of the dma hardware that way - I havn't committed it to svn yet since it still needs to be cleaned up some
21:04:28amiconnThere are a few plugins drawing directly - they would need a new #ifdef LCD_PIXELFORMAT == ... section
21:05:18amiconnThe question is whether the gain is worh the hassle
21:05:24kkurbjunyeah, doom and rockboy draw direct.. I was thinking of starfield −− I assumed that it was drawing direct, but I might be mistaken
21:05:26amiconnIt's certainly all doable
21:05:49amiconnMaybe also plasma, fire and similar
21:06:10amiconnThe xlcd pluginlib would also need to be extended
21:06:12linuxstbPacbox also draws directly, and I would imagine zxbox as well. i.e. all the emulators
21:06:25kkurbjunI think it would be nice for basic browsing −− right now even scrolling in lists with a rotated screen is choppy.. I think that the rotation code could be better written to make that better though
21:06:56saratogaso the remaining issue figuring out why non-dual boot installs sometimes get corrupted?
21:07:59kkurbjunright now it's not optimized at all...
21:08:34linuxstbsaratoga: I think the main issue is simply doing binary releases of beastpatcher and the bootloader, and making sure the install procedure is correctly documented in the manual.
21:09:21saratogawhats needed to build beastpatcher for windows?
21:09:46linuxstbSome MS compiler I think.
21:10:14saratogai have an unmodded S here, i can try it if someone builds the compiler
21:10:18saratogaor if i find a VS machine somewhere
21:10:24n1sfinally getting the beast supported would be great :)
21:10:47saratogayeah might encourage someone to pick up work on it again
21:10:52saratogaits a great piece of hardware
21:11:11*linuxstb looks around for Swedes to move the pp-bootloaders...
21:11:31saratogai asked last night and the issue was figuring otu what rbutil needs each one to be called
21:12:16linuxstbSurely they simply just need to be named the same as the current versions?
21:12:29amiconnkkurbjun: There's certainly some optimisation potential...
21:12:48linuxstbThat's how all previous versions have been released - the existing files moved into an "old" directory, and the new versions replace them.
21:13:36amiconnIt's currently copying in 16 bit amounts in a way that probably causes frequent cache misses
21:14:18amiconnkkurbjun: How fast are portrait mode updates but without dma?
21:15:47kkurbjunyeah, that's what I was thinking, if I had it doing a large read with multiple registers it would probably speed it up quite a bit- the hardware supports bursting at least on the DMA side in chunks of 32 bytes - I don't know if the cache does a burst but it may.. I'm not sure offhand, let me do a test.
21:16:03saratogawhat about the YH-XXX tagets
21:16:10saratogai think we have everything ready for them now?
21:16:18amiconnYou could do 32 bit or even 64 bit reads and writes, with the right register shuffling magic
21:17:20 Quit Galois (Remote closed the connection)
21:17:35 Join Galois [0] (
21:17:47linuxstbsaratoga: I've no idea about that. Does the latest rbutil release support them?
21:18:27saratogait does with added INI
21:18:33saratogaotherwise you have to recompile from SVN IIRC
21:18:49 Quit flydutch ("/* empty */")
21:19:17linuxstbsaratoga: (reading the logs from 2 days ago), Zagor asked for someone to produce a zip with them in the right locations?
21:19:30saratogafor other PP not YH right?
21:19:48bluebrothersaratoga: I can provide you windows binaries of beastpatcher.
21:19:52linuxstbThe problem with that is that it will overwrite the existing files, which it would be nice to backup.
21:20:10linuxstbs/backup/move to archive directories/
21:20:18saratogabluebrother: give me one and I'll try to test it tonight or tomorrow
21:20:36saratogado we have binaries released for the YH bootloaders?
21:20:44saratogai guess we could do it the same time as the other PP bootloaders
21:21:38bluebrothersaratoga: I _think_ that's the latest from svn:
21:21:51saratogathanks I'll try it when i get a chance
21:22:30kkurbjunamiconn: hmm, it's still only 58 fps with the lcd_copy_buffer_rect function
21:22:33linuxstbI can't see any existing YH bootloaders on the download server.
21:22:37bluebrotherunfortunately it doesn't have support for dual-boot patching. But AFAICS I need to figure some license stuff before I can integrate that.
21:22:43kkurbjunfor a fullscreen update
21:22:50kkurbjunbut it's much better than 20
21:23:31 Join polobricolo [0] (n=polobric@
21:23:31bluebrothersaratoga: YH support was added to rbutil, but no bootloaders have been released yet. Thus it's noted as "preliminary" on the wiki page :)
21:23:57linuxstbbluebrother: What license stuff? If you're talking about mknkboot, I wrote that, so am happy to relicense for beastpatcher. It's pretty trivial anyway.
21:23:58amiconnSo the question is, are (guesstimated) 40..50fps good enough for landscape, or do we need that transposed pixel format?
21:26:46saratogawell lets release a bootloader along with the other PP ones
21:28:24kkurbjunamiconn: it is much smoother even without dma - lists are noticeably faster and games like bubbles are pretty playable. I think if it could be boosted to around 40/50fps rotated it would be smooth enough that it's not noticeably slow- right now it looks choppy when rotated
21:28:42bluebrotherlinuxstb: yes, mknkboot would be needed to be BSD too. OTOH I was wondering if going BSD for beastpatcher was needed at all.
21:28:49ej0rgesaratoga: the beast bootloader also lacks hold switch detection, iirc
21:29:26amiconnThe beast bootloader does hold switch detection, and it's annoying
21:29:42ej0rgeit does? mine doesn't. Do i need to update it?
21:30:10amiconnIf you accidentally power on with the hold switch enabled (i.e. connecting the charger or usb) it boots into the OF, and then you need access to a battery switch tool to get back into rockbox...
21:30:18ej0rgeall the beasts I've rockboxed boot right up with the hold switch enabled
21:30:32amiconnYou install single boot?
21:30:47ej0rgesingle boot, yes, on the ones that are v1.1 OF
21:30:49amiconnThen of course you won't see it boot into the OF
21:31:03ej0rgeI did one on v1.2 OF but i sold it.
21:31:18amiconnThe hold switch doesn't prevent booting, but it selects between rockbox and OF
21:31:42ej0rgeI suppose someone might want the OF for wmp9 videos or usbotg
21:31:57amiconnImo that's a rather bad way to do this selection, but it's done this way on several other targets too
21:32:24ej0rgeI agree. better to use some button if at all possible, and have it shut down if booted with the hold switch enabled
21:32:27amiconnThe hold switch should prevent booting, and the selection should be done using button presses during boot
21:32:41ej0rgeespecially as the beast starts to boot immediately on first press of 'power'
21:33:01ej0rgenot like some targets that require you to hold down the button for a bit
21:33:54amiconnOn some (most?) targets this "hold down for a bit" requirement is implemented in software
21:34:31ej0rgeThat brings me to the only other bootloader question i know of - does anybody know the current state of the x5 bootloader with regard to the dual boot patch?
21:34:34linuxstbbluebrother: I thought that debate had already been had? (switching beastpatcher to.BSD)
21:35:08ej0rgeiirc there was some concern that the way the dual boot patch for the x5 bootloader works, it could be broken by a fw update from cowon. But that doesn't seem to be anywhere on the horizon anymore.
21:35:58ej0rgeI get the impression that the patch was written for the v2 bootloader and is now very outdated.
21:36:11bertrikon a few targets I've seen, the software has to do something to "acknowledge" the power up in order to stay powered on without the powerup button pressed
21:36:31amiconnBasically all...
21:37:29amiconnIt's just that on some target *rockbox* or its bootloader doesn't need to do that, in case it is handled by another loader before it
21:38:01 Quit jgarvey (Read error: 110 (Connection timed out))
21:39:56 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
21:39:56 Quit pixelma (Nick collision from services.)
21:39:57 Quit amiconn (Nick collision from services.)
21:39:59 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
21:40:13 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
21:40:19 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
21:46:18 Join aaron424 [0] (
21:47:43 Join Bagder [241] (n=daniel@rockbox/developer/bagder)
21:51:52Unhelpfulremember, we have pluggable output for the image loaders... and the pictureflow plugin already uses a transposed output plugin, because it renders a column at a time, so it's faster if the cover images are laid out column-first
21:52:35 Quit mt (Read error: 113 (No route to host))
21:53:54saratogawas a procedure for tagging a bootloader as a release build ever determined?
21:54:28kkurbjunhmm, last time I tested portrait DMA updates I thought I was getting 150fps - looks like it is around 190 fps in my latest test.
21:54:56kkurbjunUnhelpful: that's interesting I didn't know that pictureflow did that
21:55:57Unhelpfulthe renderer math is easier by columns... you multiply and divide to get the scaling factor for the column, and then the per-pixel math is only adds and shifts
21:57:28bluebrotherlinuxstb: sorry, had a phone call. Yep, but I'm kinda wondering about the problem linking GPL software against MS libs.
21:57:48saratogaif the libs come with the OS then its fine
21:57:58bluebrotherAFAIK qtscrob links against the same lib and is GPL.
21:58:00 Quit martian67 (Remote closed the connection)
21:58:32bluebrotherwell, the lib comes with WMP10. Not sure if that could be considered OS. Nevertheless, we can't distribute it anyway
21:58:33linuxstbIsn't that lib part of Windows Media Player. I wouldnt' call that part of "the OS".
21:59:34bluebrotherbut even if that isn't an OS lib, is there a problem with the GPL linking against other libs? Did I miss something there?
22:00:20bluebrothernot that it matters much wrt beastpatcher anymore ...
22:02:17 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
22:02:39saratogaWMP is included with the OS and thus you can link against it, but obviously not distribute the DLL
22:04:11amiconnWMP10 is not included with the OS, at least not all OS versions it runs on
22:04:43 Quit martian67 (SendQ exceeded)
22:05:48 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
22:06:19saratogaWMP10 can be included on any version of Windows install disk that it can run on (since it has slipstream support), if the OEM decides to include it doesn't matter, its still an OS library
22:07:08 Quit polobricolo (Read error: 110 (Connection timed out))
22:07:29saratogamore importantly, its a vender supplied driver for the OS, so you're within both the letter and spirit of the GPL here
22:07:49saratogathe intention isn't to prevent people who closed source drivers from running GPLed software on their PCs
22:12:56***Saving seen data "./dancer.seen"
22:20:30bertrikdoes rockbox support a framebuffer pixel format where each pixels takes 32-bits as follows: 0x00RRGGBB ?
22:22:44bertrikI guess I should check firmware/export/lcd.h
22:23:52 Join robin0800 [0] (
22:27:21n1sbertrik: i don't think it uses such a format on any target
22:27:37n1swhy not go 24bit though?
22:27:48bertrikargh, I hope I'm mistaken then
22:28:01bertrikn1s, because the LCD controller wants it like that
22:33:24n1shmm, it doesn't have a 16 bit mode?
22:33:30bertrikthis will complicate the Meizu M6 LCD driver quite a bit, the OF seems to use the 24-bit mode
22:33:37n1sbtw all our color targets are 16bit
22:33:53bertrikthe LCD controller does indeed support 16-bit mode, but that means I cannot just copy the OF settings then
22:34:45 Quit bmbl ("Bye!")
22:35:05 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
22:35:17Tornesaratoga: reflashing is feasible but i'm not clear if recovery is possible yet. quite prob ably not.
22:35:39Torneamiconn: it formats just one partition if it can. it only reformats the whole disk if the MBR is damaged or both partitions have faults
22:36:21 Quit bmbl (Client Quit)
22:36:41 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
22:36:57Torneamiconn: i've not yet found all the cases that cause a format, but I have documented some of them on the wiki:
22:39:11Tornesaratoga: i've not looked at the actual board, so not sure about JTAG. There might be a viable sequence of oeprations for flashing which leaves it able to recover itself, though. It detects a number of flash errors quite early on and resorts to jumping to the recovery image to get a new firmware
22:42:41saratogaTorne: how much does the firmware actually do? i know on the e200v1, the firmware actually handles so little that it was quite easy to remove it entirely (although some PP code still runs on the internal ROM)
22:44:59 Join captainchris [0] (
22:45:02TorneIt does rather a lot
22:45:10Torneit's the standard CE bootlaoder
22:45:15captainchrishi everybody
22:45:22Torneloads the CE kernel up in protected mode and all that jazz
22:45:29captainchrisi have a ipod nano 4th generation green 8g
22:45:43captainchrisand rockbox doesn't want to loadit
22:46:19saratogarockbox doesn't run on your player
22:46:28saratogaTorne: does it do much we need though?
22:46:45TorneIt does SDRAM setup, iomux config, clock programming, etc
22:46:59saratogahmm and that may not be implemented in rockbox alraedy?
22:47:02Tornewhich it would be easier if we didn't have to rip off
22:47:12JdGordon|captainchris: shock horror! probably because its not supposed to!
22:47:23Tornesaratoga: i doubt we re-init most of that stuff, no
22:47:51Tornesaratoga: we would probably want to jump in at the point the MMU is turned on
22:48:02Torneget their code to copy us to ram and map us then run us :)
22:48:04captainchrissaratoga i can't install it because rockbox say th eipod model it doesn't supported
22:48:25Tornecaptainchris: that's beacuse that ipod model is not supported. it says that, no?
22:48:59saratoga"rockbox doesn't run on your player"
22:49:04Torneso what are you asking?
22:49:20Unhelpfulthere is a very clear list of supported players on the front page. :P
22:49:31captainchrisi'm under linux , and i can't put my music into
22:49:38saratogaTorne: ideally we'd setup the MMU too, having that code for ARMv6 would be nice
22:49:51Tornesaratoga: well, yah. I mean we *could* do it all
22:49:54saratogawe'll likely see more of those devices in the future
22:49:59Tornesaratoga: but we may have to cleanroom some of it
22:50:07Tornesaratoga: as not all the setup is going to be entirely obvious from the manual
22:50:10captainchrisunder rythmnbox and amarock music appear and on my ipod she doesn't appear
22:50:22saratogathis isn't an ipod support channel, go ask somewhere else
22:50:35Tornesaratoga: eg how to set up all the iomux crap for the particular bus arrangement
22:51:23Tornesaratoga: anyway i think it's dependant on there being a *vagely* safe way to do it :0
22:51:55Tornesaratoga: my priority is working out how to reliably be able to boot our loader off disk without it ever reformatting or requiring the dualboot loader that we can't make
22:52:07saratogayeah that makes sense
22:52:28saratogai thought we can make the dual bootloader now?
22:52:35TorneOh, i mean, we can
22:52:39Tornebut it requires that users do it
22:52:50Tornewhereas a single boot bootloader we can distribute as a binary
22:53:09Tornes'perhaps preferable for people who don't care about booting the OF
22:55:34Unhelpfuls/make/distribute/ perhaps?
22:55:58TorneIt has to contain the OF bootloader, so installing without being copyright-violatey is slightly fiddly
22:57:15 Join MG_Man [0] (n=MGMan@
22:57:52 Quit aaron424 ("ChatZilla 0.9.85 [Firefox 3.0.13/2009080315]")
22:57:56Unhelpfulwe could get around this if we patched the flash to test and decide what to boot, and it loaded rbboot.bin or such if told to boot rockbox...
22:58:21saratogaTorne: we require a patched OF for some targets already, so IMO thats not a serious limitation
22:59:00Unhelpfuland afaik there's no (reasonable) way to get at the flash, or to install that rbboot.bin, without having already overwritten the OF
23:00:58Tornesaratoga: Well it's not uup to me
23:01:03TorneI'm basically reversing it *anyway*
23:01:10Tornethe dual boot hting was just something i was told ;)
23:01:23Tornethe original reason i started was to see if the reason it reformats sometimes is tbecause of TFAT
23:01:39Tornebecause i inadvertantly gave godeater the impression that i might know about transactional filesystems ;)
23:02:20*pixelma sees some weirdness when resuming playlists created with "insert shuffled" again :|
23:03:01saratogaUnhelpful: i wrote a tool a while ago that dumps the OF from the beast during installation
23:03:34MG_Man...Right well, sorry to interrupt, but I have a question that I think is too smal lto bother with on the forums
23:03:49MG_ManIs there a setting to change the general volume of .SPC files on RockBox?
23:04:06MG_ManSo far the codec for it ignores the "Amplification" tag of the ID666 for it
23:04:19MG_ManSo it ends up being quieter than other actual music
23:04:27n1snope, no such setting
23:04:36Tornesaratoga: anyway, yah. i'm not claiming that it's a showstopper or even professing that I care; i don't use a beast :)
23:04:49MG_ManDoes anyone know if there is work to add support for the amplification tag?
23:04:50Tornesaratoga: it's just something i expect to find he answer to in the process of other stuff
23:04:52n1ssounds like that tag could be used for replay gain or something
23:05:04MG_ManThat's essentially what it is
23:05:13MG_ManIt basically makes the file play louder
23:05:14pixelmathe correct track is repeated and tracks are still in the same "order" but the playlist is shifted so that the resumed track is either the first in the playlist (and the ones you already heard are put at the end) and one time the resumed track was suddenly the last in the list
23:05:37n1sbut no, i've not heard anyone touching that codec in a long time
23:05:41TorneMG_Man: that sounds like we should support it a la replaygain, then :)
23:06:00MG_ManIt's in an "extended" version of the tag, so that probably explains why it isnt there, but this tag is in most SPCs
23:06:46MG_ManThe problem wit hthat though is that the tag is not a decimal scale
23:06:53MG_ManSo 2.0 is not twice the volume
23:07:02pixelmadoesn't the recently committed DSP effect help there?
23:07:13MG_ManI think that affects ALL files
23:07:18pixelmaI no nothing about SPC files though
23:07:27TorneMG_Man: we can deal with the scale, but it still requires someone to do the work ;)
23:07:35Tornewho knows about spc, ideally
23:07:59MG_ManIt's a chiptune
23:08:06MG_ManBasically it's a tracker
23:08:08MG_ManSNES music
23:08:17 Join BdN3504 [0] (n=5ce22724@gateway/web/cgi-irc/
23:08:17Tornei know *what* it is. :)
23:08:23MG_ManThink midi but with patches in the file
23:08:25TorneBut most people probably know nothing about its internals
23:09:17pixelmaMG_Man: which revision of Rockbox are you running exactly?
23:09:20MG_ManThe plugin for Rockbox was based off of SNESapu which is an emulator for the SNES chip that plays the music
23:09:22MG_ManHold on..
23:09:33BdN3504for purpose write manual entry "ui viewport" first want import section from wiki but in doubt if everything there said right
23:09:54MG_ManFor the iriver h340
23:09:56BdN3504say "If any viewports are defined, you cannot display text in the default viewport (use it for images only)"
23:10:13MG_ManThe codec was added in Feb 07, so that explains the datedness
23:10:35n1sMG_Man: has the upstream changed since then?
23:10:38MG_ManSomeone would just need to update it to the latest SnesAPU version,but I'm ntot sure if whoever made the codec before is still around for that
23:11:19MG_ManI'm not sure what you mean by that, still somewhat new to this
23:11:25MG_ManBy upstream
23:12:06Torneupstream is the project we got the code from
23:12:08Tornei.e. snesapu
23:12:29BdN3504this no true, i have written many a theme containing tags before first vp definiton
23:12:48MG_ManI'm pretty sure it has, let me go see
23:13:10pixelmaMG_Man: I'm not sure if it does what you want but there was an effect added in r22394 which if I understand correctly "evens" volume (makes silent passages louder, lowers loud passages' volume), maybe it could help there?
23:13:20MG_ManI'll try it
23:13:33MG_ManSince SPC isn't an actual wavetable format, I'm not sure if it does anything
23:13:36pixelmasomething for every codec though
23:14:02pixelmaBdN3504: yes, but I think it was supposed to be true
23:14:24*pixelma played by that rule in own WPSs
23:14:49BdN3504pixelma: so parser might give error some time soon?
23:15:34 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
23:15:52 Quit mcuelenaere ()
23:16:45pixelmathat's what I always expected because otherwise you have something like overlapping viewports which is a nono and as I heard gives you update problems. Better ask JdGordon| though... ;)
23:17:21pixelmaand it would be nice if you talk in more complete sentences, I know you are able to
23:17:44JdGordon|what am I being blamed for?
23:18:56pixelmadrawing in the default vp is still possible even if you use other viewports, I thought that wasn't possible or shouldn't be
23:19:03MG_ManWell I figured out the amplification is bu dB
23:19:23JdGordon|pixelma: ah shoot... did I mess up the check? you should only be able to draw images in it...
23:19:33MG_ManThere's a general pre-amp and then one for every song
23:19:56 Quit AB3JU (
23:20:29JdGordon|I wonder if there would be a huge outcry if we forced the skin to specify its vewport instead of assuming 0,0,width,height? that would nicely avoid the issue
23:20:54MG_ManIt would also break some skins too, but it might be easy to fix
23:21:04pixelmaJdGordon|: hmm, I didn't even do that... lot of other themers do though, I didn't follow whether they only use it for images or not
23:21:13MG_ManAs long as someone fixes them, I don't think it'll be a big problem
23:22:16JdGordon|I didnt on purpose change any of the exsisting drawing behaviour
23:22:24MG_ManAnyway, I can't find a simulator more up to date that I have now
23:22:26JdGordon|IIRC really only text wasnt allowed in the default vp
23:23:27pixelmaJdGordon|: BdN3504 asked to be precise in the description of viewports in the manual. I could imagine that forcing them to define a viewport now would break quite a few WPSs
23:23:32MG_ManAnyway, I'm pretty sure there's an spc to tracker converter out there, which would allow me to change the volume for the song myself
23:24:00BdN3504pixelma: JdGordon|: can't we drop that rule? what harm does it do if we have some text before the vPs?
23:24:09NJoinAB3JU [0] (
23:24:11 Quit captainchris (Client Quit)
23:24:26pixelmaoverlapping viewports?
23:24:31JdGordon|pixelma: it would break heaps... but maybe a warning would be good
23:24:35BdN3504pixelma: JdGordon|: and also, i don't see why viewports can't overlap, if they can
23:24:49JdGordon|i tinhk the comment in the code blames conditionals for it actually
23:25:00BdN3504not, then there is no use in introducing the conditionals
23:25:06BdN3504conditional vps i mean
23:25:20JdGordon|not conditional vps... conditional tokens
23:25:32 Join aaron424 [0] (
23:25:34JdGordon|I dont remember the specifics... and i cant look at the code now to check
23:25:51JdGordon|apps/gui/skin_engine/skin_display.c in skin_update()
23:25:53BdN3504but pixelma said viewports aren't allowed to overlap, or am i wrong?
23:25:56JdGordon|maybe skin_refresh()
23:26:06pixelmaBdN3504: you can't draw two (or even more times) at the same time and spot (which overlapping could potentially cause)
23:26:18JdGordon|technically they can, but the top one might erase the bottom one
23:26:23 Nick Ypsy is now known as YpsyZNC (
23:27:17BdN3504ok, so conditional vps overlapping each other is allowed?
23:27:37BdN3504thank you.
23:27:44pixelmasorry that wasn't clear
23:28:50pixelmajust make sure you have no overlapping ones that are displayed at the same time
23:30:03BdN3504pixelma: JdGordon|: so i'll just include the line that it's maculate coding if you put text before the first vp definition?
23:31:10BdN3504erm opposite of immaculate
23:31:24 Quit aaron424 ("ChatZilla 0.9.85 [Firefox 3.0.13/2009080315]")
23:31:31MG_ManThe one who seems to have brought SPC to Rockbox is either Adam or blargg
23:31:39MG_ManDo any of you know if any of them are stil lactive?
23:32:48BdN3504JdGordon|: What i meant to say is the integrity of the code will not be guaranteed, once you put text infront of...
23:34:41BdN3504pixelma: JdGordon|: my initial question still remains though. Is it adequate to just copy and adapt the wiki entry for the manual? or do you think i have to write it from scratch?
23:37:22Unhelpfulsaratoga: oh, yeah.... that one that assumes the blocks containing the OF were not overwritten?
23:37:43MG_ManOkay, I found an SPC to IT converter, so I can just play it as a tracker and be fine
23:37:51MG_ManI guess it's not an issue
23:39:24pixelmaBdN3504: the plan was to get more of these things into the manual and clean up the wiki a bit instead, although I think the CustomWPS page is still important though (also because things can change easier and the wiki is updated a bit quicker than the manual usually). I can't see what's wrong with taking bits and pieces from there... I even copied a lot from the wiki for the ZXBox chapter myself... ;)
23:40:00amiconnbertrik: We do not have a 24bit or 32bit lcd driver yet, because until now there was no target needing it
23:40:14amiconnBut writing one should be quite simple
23:40:41BdN3504pixelma: thanks again. i'll start now, then. (no pun intended)
23:40:50MG_ManAnd anotherthing
23:40:55MG_ManThe MIDI patchset download is down
23:42:11MG_ManOh, it seems I was wrong
23:42:18pixelmayeah, isn't it now hosted somewhere else?
23:42:19*Unhelpful would think that this amplification tag thing would only need changes in a the metadata reader, and would therefore not require deep knowledge of SPC?
23:42:27MG_ManThe SPC plugin is based off blargg's Game_Music_Emu SPC player, which is based on OpenSPC
23:42:34Unhelpfulwe could just treat it like a replaygain tag
23:42:39MG_ManIt really is
23:42:42MG_ManA preamp
23:42:42BdN3504JdGordon|: one thing, especially for you: these two opts "Maximum file size used is \opt{lcd_bitmap}{1600}\opt{player}{400} bytes." are now obsolete, right?
23:42:44MG_ManBy DB
23:43:17MG_ManAnyway, if the patchset host is moved, it isn't updated on the site
23:43:30MG_ManThat is if the audiocodecs page is where I'm supposed to get it
23:43:35 Join IuDeX [0] (n=c2bb48fd@gateway/web/cgi-irc/
23:43:37JdGordon|BdN3504: behaviour is guarenteed... drawing in the default vp isnt allowed and wont work, and yes, that is very obsolete
23:43:44 Quit IuDeX (Client Quit)
23:44:42JdGordon|that has been wrong for a very long time :)
23:44:44pixelmascorche|sh: didn't you "take" the midi patchset? I could remember wrong though
23:46:59 Quit BHSPitLappy (Remote closed the connection)
23:48:23CIA-61New commit by nls (r22431): Slight cleanup in lcd.h: kill fixme and remove useless include, add some includes elsewhere and add a stub for sims instead of ifdef
23:48:37JdGordon|BdN3504: umm... some things which might eb good to make sure are in the manual... %V| (without the l) will always be displayed, you *should* only do things like %xl %X and other "setup" things before the first %V lines (or %vd)
23:49:25MG_ManI suppose the SPC plugin is fine, just the ID666 wrapper for it needs to be updated to support ID666 Extended
23:50:07BdN3504JdGordon|: and %wd also HAS to be in front of the first vp declaration, right?
23:50:10JdGordon|BdN3504: also, remove any mention of limits... the only limit anywhere is one progressbar per viewport
23:50:23 Quit BBBradley ("CGI:IRC")
23:50:26JdGordon|no, %wd can be anywhere
23:50:32JdGordon|thats statusbar disable?
23:50:43JdGordon|yeah, it can go anywhere
23:50:54BdN3504sorry, didn't know that :(
23:50:57MG_Mananyway, I can't seem to find a site with details on ID666Extended, so I guess I can just say the fields here, but I dont think that'll ehlp wuth utilizing them
23:51:28JdGordon|at least it *should* work anywhere... I cant imagine people would put it anywhere except at the start, but iirc the code should be fine
23:51:46 Join jgarvey [0] (
23:51:55 Quit Strife89 (Read error: 113 (No route to host))
23:53:50MG_ManI found a site with useful specs on ID666 extended
23:53:57MG_ManLook at Extended ID666 Items
23:54:07MG_ManAt the bottom
23:54:14BdN3504pixelma: JdGordon|: where'd you put the vp part: as the first subsection of "building your own", the last or somewhere in the middle?
23:54:22bluebrothern1s: already spotted the red?
23:54:27BdN3504i'd prefer right on top
23:55:06MG_ManSo tags for 01h-07h are already in, but no support for 10h-36h which are the extended tag fields
23:55:18n1sbluebrother: have now ;)

Previous day | Next day