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

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

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

#rockbox log for 2010-07-26

00:01:23CIA-8r27565 build result: All green
00:05:30Buschelwmapro does not use any ams for fixmul or vector operation (e.g. windowing, scalar product). introducing asm via copying from other codecs should be quite easy.
00:05:39***Saving seen data "./dancer.seen"
00:06:10n1syeah, that would probably help a lot
00:08:14Buschel3 fixmul's should be used -> fixmul16, fixmul31 and fixmulshift -> fixmul16/32 can be copied from atrac3 decoder, fixmulshift from musepack
00:08:39Buschelalways for both coldfire and arm
00:09:28n1sif you use the coldfire inline asm functions *don't forget to set the macsr*
00:09:57Buscheldon't worry, I'll keep my hands off this ;)
00:10:12*maraz looks at and sighs
00:10:22n1sBuschel: i don't mean you specifically but that bug has popped up in several codecs
00:11:26BuschelI know, and I forgot it in my mpc update when swithcing to sv8 as well :/
00:14:21Buschelhave to get some sleep now, see you later!
00:16:54 Quit domonoky1 (Read error: Connection reset by peer)
00:17:02 Join milz [0] (
00:19:17 Quit Buschel (Ping timeout: 276 seconds)
00:20:11 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
00:23:03 Quit stripwax (Quit:
00:24:55CIA-8New commit by nls (r27566): codecs: mark some local variables with 'static'
00:25:01 Join fdinel [0] (
00:26:23CIA-8r27566 build result: All green
00:32:31 Quit mt (Read error: Connection reset by peer)
00:33:06 Join mt [0] (~mtee@rockbox/developer/mt)
00:33:06n1sindeed, that fixmul_window function uses 32x32=64 multiplies and that is done with support routines on coldfire, urg
00:36:04 Quit bertrik (Ping timeout: 265 seconds)
00:36:48n1sif emac is used, it will be much faster if we don't get the lsb, is that likely to matter?
00:38:24 Quit mt (Read error: Connection reset by peer)
00:39:00 Join mt [0] (~mtee@rockbox/developer/mt)
00:44:47 Join Drise [0] (
00:45:58CIA-8New commit by kugel (r27567): Rewrite set_file() to be smaller and better readable, and a comment describing what it does.
00:48:01CIA-8r27567 build result: All green
00:48:44 Quit Drise (Client Quit)
00:56:07 Quit JdGordon (Ping timeout: 260 seconds)
00:56:08 Quit S_a_i_n_t (Ping timeout: 246 seconds)
00:59:00 Quit simonrvn (Quit: .)
00:59:52 Join simonrvn [0] (
01:04:13 Quit Jerom (Read error: Connection reset by peer)
01:04:35 Quit mt (Read error: Connection reset by peer)
01:05:11 Join mt [0] (~mtee@rockbox/developer/mt)
01:09:09 Quit lestatar (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
01:09:27mtn1s: Still around ?
01:13:24n1smt: a little while
01:14:49mtWhen you have time, could you test putting quant table in iram ? it's small and accessed once per block per channel
01:16:03n1sok, i'll test that now
01:16:12mtThanks :)
01:22:34n1s63.91% realtime :/
01:23:28kugel0.02% speed up seems strange if it's accessed so often
01:31:02n1sif i find time and motivation tomorrow i'll try to hack up some cf inline asm for the multiplication functions, should help a lot since they use support routines now on cf
01:35:37mtn1s: ok great.
01:35:56mtI'll go sleep now
01:37:39n1smee too
01:37:41 Quit n1s (Quit: Lämnar)
01:38:49 Quit bmbl (Quit: Bye!)
01:47:29 Join JdGordon| [0] (
01:47:30 Quit JdGordon| (Changing host)
01:47:30 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
01:54:06 Quit DerPapst (Quit: Leaving.)
01:57:49 Join Strife89 [0] (
02:05:41***Saving seen data "./dancer.seen"
02:06:09CIA-8New commit by kugel (r27568): Fix potential buffer overflow.
02:08:02CIA-8r27568 build result: All green
02:34:48 Quit fdinel (Read error: Connection reset by peer)
02:39:44 Quit JdGordon| (Ping timeout: 246 seconds)
02:43:51 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
02:44:30 Join bunnyboi [0] (
02:53:04 Join S_a_i_n_t [0] (S_a_i_n_t@
03:08:28 Join Zacman2k [0] (
03:09:17 Quit Zacman2k (Client Quit)
03:12:56 Quit MethoS- (Remote host closed the connection)
03:14:06 Quit kugel (Remote host closed the connection)
03:23:26 Quit kisak (Ping timeout: 260 seconds)
03:23:54 Quit krazykit (Ping timeout: 260 seconds)
03:24:15 Join kisak [0] (
03:24:30 Join krazykit [0] (~kkit@
03:25:56 Quit hebz0rl (Ping timeout: 240 seconds)
03:29:28 Join webguest56 [0] (
03:31:14 Quit webguest56 (Client Quit)
03:40:18 Join hebz0rl [0] (
03:40:38 Join z35 [0] (
03:40:55 Join Drise [0] (
03:43:00 Quit sudoman (Quit: Leaving)
03:47:10marazS_a_i_n_t: heh, haven't heard a single skip or seen a single pixel of font corruption
03:47:38marazi wonder how the OF handles DMA
03:47:44marazif it does it at all?
03:53:33 Join shadoxx_ [0] (
03:57:03 Quit shadoxx (Ping timeout: 260 seconds)
04:00:52 Quit scorche|sh (Remote host closed the connection)
04:05:42***Saving seen data "./dancer.seen"
04:10:20S_a_i_n_tmaraz: That's all very well and good for you, seriously...but, I'm more concerned with why your specific unit can't handle it.
04:11:13S_a_i_n_tAFAIK there are know unknown hardware types in the Nano1g, so, there is a part of yours that isn't performimng.
04:12:01marazwell, what do you suggest we do to find out what that part is?
04:12:23S_a_i_n_tThat's what we have been doing ;)
04:12:46marazare you sure the dma timings aren't to blame?
04:12:47S_a_i_n_tI'm not entirely convinced that the problem lies in the area where you have found a solution.
04:13:15S_a_i_n_tIf they do you account for my 4 units, ad countless others that have no problem with them>
04:13:47S_a_i_n_tI mean, my Nano1Gs are *solid*.
04:13:54S_a_i_n_tlike, "SOLID".
04:13:58marazmine is too, now that it's running on PIO.
04:14:16S_a_i_n_tAnd I bet you have considerably slower transfer too.
04:14:31 Quit TheSeven (Ping timeout: 276 seconds)
04:14:34marazi usually use disk mode >_>
04:14:48marazwould it be greatly faster to use the rockbox firmware?
04:15:21S_a_i_n_tIt wouldn;t be faster at all even if DMA were enabled.
04:15:27S_a_i_n_t(on your unit)
04:16:00S_a_i_n_tThe OF outperforms RB on transfer still fir the Nano1G
04:16:16marazso i reckoned
04:17:18marazperhaps there is a hardware difference of which the OF is aware but rockbox is not
04:17:47S_a_i_n_tHArdware was analyzed by taking units, I doubt this.
04:18:06S_a_i_n_tyou don't guess what the hardware is by looking at the firmware.
04:18:10S_a_i_n_t(most times anyway)
04:18:16marazheh, well, obviously we need to take a look at the insides of my unit, then?
04:18:38 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:18:53marazi'm thinking of swapping the battery for a fresh one, so when i'm doing that i might as well document what's inside
04:19:12S_a_i_n_tGet a Nano2G battery.
04:19:17S_a_i_n_tyou won't regret it.
04:19:22marazit fits?
04:20:09marazso, 400mAh instead of 330mAh?
04:20:57S_a_i_n_t300mAh AFAIK
04:21:07S_a_i_n_t(for the Nano1G
04:21:16marazum, it's 34x56mm, how is it supposed to fit in the 22x39mm the nano1g battery is :\
04:21:29marazor is there extra space?
04:22:05S_a_i_n_tThat doesn't seem right to me...the one I got out of my "spare parts" 2g fits perfectly.
04:22:13S_a_i_n_tPerhaps there was substitution.
04:22:47S_a_i_n_tThe 1G battery is pretty snug in the case.
04:23:32S_a_i_n_tyeah...I tend not to trust "replacement battery" sites.
04:24:08marazfor their bad quality or for their dubious mercantilism?
04:24:27S_a_i_n_tAdd blatantly lying to the list too :P
04:24:41S_a_i_n_tit's too hard to filter out a reputable merchant.
04:25:03marazthis one seems respected among coin collectors
04:25:17marazwhich is, of course, their main business
04:25:19S_a_i_n_tAnd we all know coin collecters ae DAP experts :P
04:25:25marazsure thing.
04:26:22 Quit pixelma (Disconnected by services)
04:26:24 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:26:39 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:26:40DriseJust curious, does anyone know of any updates on the Fuze V2 USB since like a week and a half ago?
04:26:53S_a_i_n_tSVN probably does ;)
04:27:40 Quit amiconn (Disconnected by services)
04:27:42 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:27:49 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:30:05DriseHm... alright, thanks for that page.
04:30:49DriseI've been watching closely for it, and i'm subscribed to the svn commits by email, but I haven't seen anything. Thanks again
04:31:00 Quit bunnyboi (Quit: +++ OK ATH OK)
04:32:56mc2739Drise: when something happens, it will be noted by the commits, or on the SansaAMS wiki page here:
04:33:27DriseRight, I was just seeing if anything had been done but not commited for various reasons.
04:36:45 Quit Drise (Quit: Leaving)
04:42:56 Quit seani (Ping timeout: 258 seconds)
04:43:15 Quit z35 (Ping timeout: 265 seconds)
04:51:49JdGordon|mc2739: e200 cabbie doesnt use viewports (yet) so I'm not too worried if it goes funny when aa turns on
04:51:56JdGordon|although I thought I had that working a while ago :p
04:53:41mc2739JdGordon|: I know S_a_i_n_t is working on converting it to viewports, but I just wanted to let you know it looked strange
04:54:23JdGordon|if that is the last of the issues then awesome :)
04:54:27 Join Barahir_ [0] (
04:54:49JdGordon|I'm doing a full set of zips now so hopefully more testers help out.. otherwise meh
04:54:56JdGordon|commiting in a few days hopefully :)
04:56:53mc2739did you see that I opened a new bug on the FMS? Just another oddity I noticed while testing your patch.
04:57:06JdGordon|yeah, I think that is a dupe though
04:57:18JdGordon|I'm blaming S_a_i_n_t for that one
04:57:44 Quit Barahir (Ping timeout: 240 seconds)
04:58:10mc2739Thats right, he is the one that pushed for the playbar in the fms
04:58:19JdGordon|oh that one
04:58:27 Join seani [0] (~seani@
04:58:29JdGordon|I'm getting confused, yah i saw that one also
04:58:39JdGordon|should be an easy fix
04:58:46mc2739blame him for both
04:58:53JdGordon|ok :)
04:59:33 Join z35 [0] (
05:25:55 Quit hebz0rl (Quit: Ex-Chat)
05:26:37S_a_i_n_tblame me for what?
05:26:40*S_a_i_n_t muses
05:27:29S_a_i_n_tbtw...I realise it isn't finished yet, but, can you commit the Cabbie Viewport conversion so far?
05:27:34JdGordon|the fms, poverty, global warming
05:27:35S_a_i_n_tJdGordon|: ^
05:27:59S_a_i_n_tI figure the ones that are converted may as well go in.
05:28:04JdGordon|I'll do it tonight if I remmeber or am reminded
05:28:11JdGordon|shouldnt really do it from work :)
05:28:30S_a_i_n_tThe remaining ones are gonna piss me off as I need to screenshot them all, but, I'll get them done tonight.
05:28:34S_a_i_n_tOr, most of them.
05:28:41S_a_i_n_tSome, at the least ;)
05:29:16S_a_i_n_tI need to screenshot how each Cabbie looks at the moment, and figure out the viewport values from there.
05:29:23 Quit bieber (Ping timeout: 276 seconds)
05:29:32S_a_i_n_tIt wont be a tradegedy if they aren't ABSOLUTELY perfect...right?
05:29:40 Join bieber [0] (
05:29:44S_a_i_n_tI mean ~1 or 2 px off.
05:30:37JdGordon|of course it bloody would!
05:30:37S_a_i_n_tDue to the overuse of %?C I figure I'm the first to touch these WPSes in a LOOOOONG time.
05:30:53JdGordon|pixel perfect my friend... pixel perfect
05:30:56*S_a_i_n_t senses some sarcasm, or, hopes he does.
05:31:23S_a_i_n_tFor the smaller screens...sure, that would be a problem. I thinkI only have the big bastards to go.
05:32:04S_a_i_n_tmrobe, mini2440 and friends.
05:34:31 Join Skaag [0] (
05:35:28SkaagIs it safe to assume that if the Sansa View doesn't appear in the list of ports, that it's unsupported?
05:35:54JdGordon|that isnt an assumption
05:36:00JdGordon|that is general comprehension
05:36:44Skaagmy brother got that player, and its software totally sucks… he tried putting some mp3's on it, and it doesn't even recognize them, although it attempts to rebuild its database...
05:36:56Skaagso I was hoping there would be a better firmware for it ;-)
05:37:04Skaaglooks like an awesome project
05:39:14 Quit panni__ (Quit: ( :: NoNameScript 3.81 :: ))
05:40:52 Quit Horscht (Quit: Verlassend)
05:41:58JdGordon|there was a GSoC project i tihnk last year to get the view going
05:42:07JdGordon|unfortunately it didnt get far i tinhk
05:43:43S_a_i_n_tThe View can most certainly process .mp3
05:43:56S_a_i_n_tso, something is wrng with the file format methinks.
05:44:14S_a_i_n_tThough...that is more than slightly offtopic here.
05:48:23 Join binaryhermit [0] (
05:48:29 Quit Skaag (Ping timeout: 260 seconds)
05:55:18 Join Skaag [0] (
06:01:18 Join n17ikh [0] (
06:05:43***Saving seen data "./dancer.seen"
06:10:09 Part Skaag
06:13:44*S_a_i_n_t actually reads what he was blamed for...
06:14:11S_a_i_n_tI didn't push for the progressbar in the FMS, it was already there...I just wanted it to *work* ;)
06:14:34S_a_i_n_tfwiw, it works fine just seems to have broken updating everywhere else in the FMS :P
06:15:19S_a_i_n_tmc2739: Can you please check if that theme of mine works on your DAP now that kugel fixed up the %tx tag?
06:15:38S_a_i_n_ttry uncommenting the two lines that use it, and it *should* work.
06:16:51S_a_i_n_tWhen I finally get around to finishing this theme, I will post a version for that model also.
06:17:11S_a_i_n_tThough, an iPod OF replica on a Sansa seems a little weird :P
06:32:42mc2739S_a_i_n_t: your theme works now that the %tx tag has been fixed.
06:36:38S_a_i_n_tmc2739; Excellent, thankyou for confirming that.
06:36:53S_a_i_n_tRefresh my memory...what device are you testing on?
06:37:11mc2739e200 (176x220)
06:37:23S_a_i_n_t(I'll make a new set of backdrops/alignment x&y's for it)
06:37:26 Join Jack [0] (
06:37:52S_a_i_n_tJack: If you have a question, shoot.
06:37:52 Nick Jack is now known as Guest28708 (
06:40:24Guest28708I'm going to be getting a Sansa Fuze (V1) in a couple of days and I was thinking about installing RockBox on it, but first I was wanting to know if Rockbox can be completely uninstalled if I don't like it.
06:41:51mc2739Guest28708: Yes, it can.
06:42:30S_a_i_n_tGuest28708: Have a read through the manual
06:42:49S_a_i_n_t(linked above)
06:43:04Guest28708Ok,, how do you uninstall it? Do you just delete the files off of the device?
06:43:18Guest28708I'll have to check that manual out
06:45:13mc2739Guest28708: please do - it answers many questions
06:45:41Guest28708Ok thanks. I didn't see the manuals section before.
06:46:28 Quit AndyI (Ping timeout: 248 seconds)
06:46:50 Quit Guest28708 (Quit: ChatZilla 0.9.86 [Firefox 3.6.4/20100611143157])
06:53:02 Join AndyI [0] (~pasha_int@
07:06:33 Join BHSPitMini [0] (
07:10:27 Join einhirn [0] (
07:14:05 Quit CaptainKwel (Quit: Ex-Chat)
07:15:29 Quit binaryhermit (Quit: Leaving)
07:16:55 Quit bieber (Ping timeout: 276 seconds)
07:17:11 Join bieber [0] (
07:20:31 Quit BHSPitMini (Read error: Connection reset by peer)
07:23:16 Quit AndyI (Ping timeout: 248 seconds)
07:24:16 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
07:37:41 Quit anewuser (Quit: Sixth edition of 25 samples + 7days + Remix Phase July 19th - July 26th)
07:37:45 Quit bieber (Ping timeout: 240 seconds)
07:37:59 Join bieber [0] (
07:40:40 Join Buschel [0] (
07:44:37 Join shai_ [0] (
07:47:12 Quit shai (Ping timeout: 240 seconds)
07:48:06Topic"append Help test:" by ChanServ (ChanServ@services.)
07:48:18*JdGordon| fails
07:49:09Strife89I'll install the builds to my appropriate DAPs in the morning.
07:49:21Strife89(i.e., in about 9 hours. :/ )
07:49:27JdGordon|what was the old topic?
07:49:43Strife89One sec.
07:49:56Strife89Rockbox 3.6 now released! | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community | This channel is logged at
07:50:07Strife89Yay for logs. :3
07:50:18Topic"Rockbox 3.6 now released! | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community | This channel is logged at Help test:" by ChanServ (ChanServ@services.)
07:50:33JdGordon|close enough :)
07:50:47Strife89Maybe someone will stick the pipe in later. :P
07:51:07Strife89Were I opped, I would.
07:51:14Topic"Rockbox 3.6 now released! | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community | This channel is logged at | Help test:" by ChanServ (ChanServ@services.)
08:05:45***Saving seen data "./dancer.seen"
08:10:28 Join NTKM [0] (
08:18:08 Join jfc^3 [0] (
08:21:39 Quit jfc (Ping timeout: 276 seconds)
08:22:53 Quit NTKM (Quit: NTKM)
08:28:27Buscheln1s/mt: fs#11498 replaces wmapro's multiplication functions with coldfire and arm asm macros. gives a nice speed up on PP5022. but needs to be tested on coldfire target, of course
08:29:07 Quit bmbl (Ping timeout: 258 seconds)
08:34:49 Join ender` [0] (
08:39:49 Quit Buschel (Ping timeout: 260 seconds)
09:18:32 Join funman [0] (~fun@rockbox/developer/funman)
09:22:40 Join GodEater [0] (
09:22:41 Quit GodEater (Changing host)
09:22:41 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
09:23:18 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
09:27:05 Join swilde [0] (
09:34:39 Join Buschel [0] (
09:36:23 Join mitk [0] (~mitk@
09:44:34 Quit JdGordon| (Quit: leaving)
09:45:46pixelmamaraz, S_a_i_n_t: doesn't the Apple OF provide some debug menu that shows quite detailed hardware info? If it does, maybe it's worth checking for tiny differences
09:56:25 Join Rob2223 [0] (
10:00:33 Quit Rob2222 (Ping timeout: 265 seconds)
10:00:41S_a_i_n_tpixelma: It does yes, but, it is not terribly verbose.
10:01:03S_a_i_n_tMore self diagnostics, rather than detailed hardware info.
10:01:22 Quit TheSeven (Ping timeout: 276 seconds)
10:01:52S_a_i_n_tFrom what I know, apart from NAND types between the different capacities with a Nano1G you get what you get.
10:02:11S_a_i_n_tNot like Sansa and their tricky substitution for compatible hardware.
10:02:14pixelmahmm, I thought it was (e.g. it was possible to find out which type of the two different LCD panels a 4G Ipod had IIRC) but I don't have an Ipod so it's all hearsay
10:02:57S_a_i_n_tYou can find out some things...I'll check now. I don't remember offhand.
10:03:25 Join NTKM [0] (
10:05:46***Saving seen data "./dancer.seen"
10:07:12 Join fyrestorm [0] (
10:07:16S_a_i_n_tRTC type: Not specific (just some number, always 41 from what I have seen), NAND ID, capacity, a keytest, a flash memory scan, a power cycle test, a clickwheel test, clickwheel ID, playback test, line in test, general ststus, HD (NAND) spec, a pattern test for the LCD, a sleep test, and a hard (harder than hard reset apparently) reset.
10:07:49S_a_i_n_tIIRC the HDD iPods also have S.M.A.R.T info available.
10:09:37funmanwhich thread are the metadata parsers running in ?
10:09:46pixelmawhat does the NAND ID and spec say? (or maybe the flash memory scan but it possibly only scans for filesystem errors)
10:11:48S_a_i_n_tNAND ID is included in the "five in one" test, and is just a number (model number I assume), HDD (NAND in this case) Spec has model number, firmware rev: partition ifo, LBA and total capacity.
10:12:00S_a_i_n_tThe NAND scan is just that, a scan for FS errors.
10:12:45S_a_i_n_tApart from the NAND, there is no specific chip/component information.
10:13:12S_a_i_n_tApart from LCD type, but that is just a number (either 7 or 11 AFAIK)
10:13:32S_a_i_n_t11 seems to be the better LCD type.
10:16:37pixelmawell, I thought maraz had problems with the flash chip so I thought the NAND model might be interesting, Does the model number mark a specific "line" or is this always a unique number if you can tell from your different Nanos?
10:18:15S_a_i_n_tI believe the model number only varies in the capacity, from my 4, the two 4GB's have the same model number, as the do two 2GBs.
10:19:07S_a_i_n_tIf there is indeed some totally different NAND, it may be interesting...
10:19:47S_a_i_n_tthough, I don't see any reason for it to cause trouble. I can't imagine how many Nano1G Rockbox users there are, and maraz seems to be the only one with this problem.
10:20:35S_a_i_n_tpixelma: Do you know much/anything about convbdf?
10:21:09S_a_i_n_t(I want to add 1px of blank space top and bottom to a this possible?
10:21:13 Quit Buschel (Ping timeout: 240 seconds)
10:22:09pixelmaI don't but I think I saw a forum thread where fml (fml2 in the forums) explains a bit about it
10:22:20*S_a_i_n_t looks, thanks.
10:23:33S_a_i_n_tI have given up trying to recreate a complete unifont charset by hand...
10:23:55S_a_i_n_tI knew it would be hard, but, I didn't really know *how* hard.
10:24:04S_a_i_n_ts/hard/time consuming/ doesn't *quite* do what I want it to. It only modifies the ascent/descent if the font conversion will clip the glyphs as I understand :=/
10:26:23S_a_i_n_tI really just want to add 1px top and bottom.
10:28:49pixelmaif you load the font into a font editor, doesn't it let you set some of these parameters?
10:32:28S_a_i_n_tHmmm...probably. I guess I need to set up a font editor again.
10:32:44 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
10:34:09pixelmaor google for convbdf options or possible parameters
10:34:50S_a_i_n_tHmmm, good point. I didn't know if convbdf was implemented from elsewhere, or created for RB specifically.
10:35:54JdGordonmc2739: I dont see the line artifacts on e200 cabbie (I'm using %?mh instead of %?C to force it)
10:37:11 Join DerPapst [0] (
10:41:19JdGordonhow do I commit directly from git?
10:41:45funmanJdGordon: RTFM
10:41:54JdGordonmeany :)
10:42:56 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:43:04CIA-8New commit by jdgordon (r27569): Fix FS #11497 - switching from a .cue to fm would show the cuesheet bars on the frequency bar thingy
10:43:18JdGordonoh shit...
10:43:41JdGordonhmm... does it do a rebase after?
10:44:00*JdGordon hopes he didnt commit more than he wanted
10:44:20funmanJdGordon: 'git diff trunk' / 'git log trunk..' will show you the difference with svn
10:44:36 Quit NTKM (Remote host closed the connection)
10:44:45funmanalso you can use −−dry-run with git-svn
10:45:00funman <- your commit
10:45:13CIA-8r27569 build result: All green
10:46:35JdGordonyeah, all good :)
10:46:54JdGordonit must have started a rebase straight away which scared me.... looked like it was commiting more files
10:47:44funmangit svn dcommit rebases before committing
10:53:09pixelmaJdGordon: did you have time to look into the crash in some cuesheet function with auto-change directory on hwcodec or does it eventually become a bit easier to fix with the new parser? (and ye, there is a FS# for it)
10:54:44bieberJdGordon: I got your message from earlier about squeezing down the parse tree...if you're really scrounging for bytes, switching the parser up to use arrays instead of linked lists would save you a pointer for each skin_element
10:54:54JdGordonmy ajbr power brick needs 110v so havnt been able to debug at all
10:55:30JdGordonbieber: doesnt that require we know how many elements there are?
10:55:54bieberYes, you'd have to have your containing elements (Viewport, line, etc.) scan ahead to figure out how much space to allocate
10:56:35JdGordonhardly worth it actually
10:56:54 Quit mikroflops (Ping timeout: 265 seconds)
10:56:58JdGordonI got a pretty big decrease by not allocing new params for each tag but I tihnk that is all
10:57:37bieberYou mean by storing a pointer to the tag table? That was a good move :)
10:57:38pixelmayou can charge the NiMHs in an external charger ;)
10:57:51bieberThe pointer thing would be a real act of desperation, but I figured I'd throw it out there :P
10:58:09 Join mikroflops [0] (
10:58:40JdGordonbieber: na, I just have a static array of 12 params, if any tag uses less than that I set its param pointer to that array and prey the callback does what it needs to and throws away the rest
10:58:55JdGordonpixelma: I could if I had one
10:59:04JdGordonI tihnk the strings could save some space also
10:59:06bieberOh, okay
10:59:19CIA-8New commit by funman (r27570): aiff metadata: skip chunks larger than buffer ...
10:59:23JdGordonthis is only for ROCKBOX builds so you dont need to worry
10:59:25CIA-8New commit by funman (r27571): aiff codec: small clean up ...
10:59:39S_a_i_n_tsend the batteries to me, and I'll charge them and post them back ;)
10:59:52 Quit bmbl (Quit: Bye!)
11:00:02JdGordonthey are bassically dead anyway so need to buy new batteries
11:00:19JdGordonalso the ajbr is back in melbourne anyway :)
11:00:44S_a_i_n_tDid you donate it to a museum? :P
11:01:18CIA-8r27570 build result: All green
11:01:52JdGordonpixelma: I dont remember the details, does it only crash if the last track was a .cue?
11:03:20CIA-8r27571 build result: 0 errors, 2 warnings (funman committed)
11:03:32*JdGordon blinks at the gap in the revision numbers in that report
11:03:37pixelmathe crash in the cuesheet function puzzles me a bit, but it is reproducible (and the bug was introduced in the cuesheet rework). It doesn't matter if cuesheet is enabled or not
11:04:14pixelmait just shows that not many use auto-change directory...
11:04:33JdGordonI cant do anything more than having another look at the code for the next (long) while
11:04:39CIA-8New commit by funman (r27572): aiff metadata: fix format string (%zd for size_t)
11:05:46S_a_i_n_tI do use autochange directory, but I fond it doesn't always seem to work. Sometimes after finishing an album/directory it will just stop playback.
11:05:54S_a_i_n_tI don't use cuesheets however
11:06:00JdGordonumm... WTF? is that a NULL/random pointer access?
11:06:25CIA-8r27572 build result: All green
11:07:26JdGordonwait, why the heck is cuesheet still an option?
11:07:41JdGordonIIRC the point of that change was to remove the buf_alloc() requirement
11:07:49JdGordontucker time.. ill investigate after
11:09:16 Join MethoS- [0] (~clemens@
11:09:42pixelmabecause it needs quite a lot of RAM
11:09:47pixelmaI thought
11:10:30JdGordonno, i moved that onto MoB so it wouldnt
11:10:44JdGordonnom nom nom.. bbs
11:10:48pixelmathere is no MoB on hwcodec...
11:11:12JdGordonok, so hwcoddc should stay as an option
11:11:35 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:11:38 Join shai [0] (
11:12:51 Quit shai_ (Ping timeout: 245 seconds)
11:18:44 Join Jaykay [0] (
11:29:01 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:30:09 Join glaucous [0] (
11:30:41glaucousIs it possible to edit the <All Tracks> menu? I'd like it to go by "fmt_title" instead of just title.
11:30:53pixelmaso that change was basically wrong for hwcodec at all?
11:31:03pixelmaJdGordon ^
11:35:17pixelmabut the comit message mentions differences between hwcodec and swcodec, no idea.
11:40:13JdGordonpixelma: dunno.. need to have a careful look at it
11:42:46JdGordonhmm no, there is still an alloced cuesheet
11:47:06JdGordonamiconn: do you know when hwcodec reads the metadata from the file?
11:49:18JdGordonah found it...
11:49:25JdGordonpixelma: test a one line patch?
11:50:17glaucousWhy isn't this working? %format "fmt_title""%s, %4d": %s, %02d. %s" artist year album tracknum title ? tracknum > "0" & year > "0". All is working until I add the "year" part.
11:51:04 Join Strife1989 [0] (
11:51:31pixelmaI'm at work currently (and on a Mac, so installing builds manually already is a PITA), otherwise I would
11:52:48JdGordonok well try the patch when you get a chance and commit if it works
11:52:54 Quit S_a_i_n_t (Ping timeout: 260 seconds)
11:52:55JdGordonAFAIC im done with it
11:53:37pixelmaok, if you can post it somewhere (with a short explanation on what it does)
11:53:58 Join S_a_i_n_t [0] (S_a_i_n_t@
11:54:12 Quit Strife89 (Ping timeout: 240 seconds)
11:54:14JdGordonthe tracker
11:54:45pixelmathanks, I'll try in the evening and report back (or commit if you want)
11:55:15pixelmanice patch name ;)
12:00:46 Quit JdGordon (Ping timeout: 245 seconds)
12:01:31 Quit funman (Quit: free(random());)
12:05:48***Saving seen data "./dancer.seen"
12:09:45 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:11:07 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:11:21 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
12:15:19 Quit glaucous (Remote host closed the connection)
12:15:46JdGordonpixelma: if it works feel free to commit... here is a good message to use:
12:16:40JdGordon"If this isnt proof that we can drop hwcodec support then I dont know what is... 4000 revs between cause and report, another 2000 to fix..."
12:34:13marazS_a_i_n_t: is there anything specific you'd like to know from the OF diagnostic mode?
12:35:41S_a_i_n_tmaybe the NAND info might be interesting, to compare between my Nano1G 2GBs
12:37:32 Quit swilde (Read error: Connection reset by peer)
12:38:29marazS_a_i_n_t: would that be HD_Specs?
12:39:01marazthe model number is WE-M-C45K-19D05LT5SS (firmware revision KB)
12:39:27maraztotal LBA 3999744, if it matters
12:41:28S_a_i_n_tSame NAND as my two Nano1g 2GBs, and my 4GBs also.
12:42:14Tornemaraz: it's interesting that it's DMA that causes it ;)
12:42:19Tornehave you tracked down which actual chip is overheating?
12:42:34Tornei would *hope* the controller
12:42:36marazthat would require opening the device, which i have no tools for
12:42:48Tornesince the actual speed of accessing the NAND is more or less the same whether you use DMA or not
12:42:54S_a_i_n_tThe only difference between the 2 and 4GB models is that the 2GB is missing one NAND chip.
12:42:56Tornewell, a bit faster for reads
12:43:16Tornei say "hope"
12:43:20marazand 25% slower for writes, yes?
12:43:48Torneer, no idea. don't think anyone tested that on the flash based devices
12:44:02Torneit was slower on several disk based models, so i didn't bother turning it on
12:44:40TorneI have no idea if the OF uses DMA on the flash devices
12:44:56Torneit certainly does use DMA on the disk ones, since that's where the code to do it came from
12:45:41S_a_i_n_tthing I can't figure out is why maraz's Nano doesn;t like it.
12:45:47S_a_i_n_tand mine are all fine.
12:45:56Tornewell, i have no idea how many nano1g users there are
12:46:08Tornebut the fact that absolutely nobody else has reported this suggests his hardware is just marginal/broken
12:46:27Tornesorry :)
12:46:33S_a_i_n_tI've been saying that for a while ;)
12:46:55Tornei'm generally not willing to disable a feature based on *one* bug report that can easily be a hardware issue, i'm afraid, maraz
12:47:24marazdon't worry, i'm not requesting *that*
12:47:43Tornebut.. there's not a great deal we can do
12:47:51Torneyou tried lower modes, yes?
12:48:03marazyeah, i tried UDMA1
12:48:06marazto no avail
12:48:25TorneYou could try MDMA
12:48:34Tornei forget how you make it do that though
12:48:40Torneand i don't know if that was ever actually tested ;)
12:48:45marazwhat's the difference between MA099NF/A, DK/A and B/A?
12:49:17marazall are 2GB black 1G nanos
12:49:29marazmine is a DK model
12:49:47 Join scorche [0] (~scorche@rockbox/administrator/scorche)
12:50:02Tornei doubt we have any idea, i'm afraid :)
12:50:13 Quit scorche` (Ping timeout: 248 seconds)
12:50:27Tornethe model numbers change quite a bit and there's rarely any detectable hardware difference that affects us
12:55:25 Quit GodEater (Read error: Operation timed out)
12:55:34 Join [Saint] [0] (S_a_i_n_t@
12:56:00 Quit S_a_i_n_t (Ping timeout: 264 seconds)
12:56:05Tornemaraz: i'm afraid i'm a bit short of ideas
12:56:10 Quit kugel (Ping timeout: 240 seconds)
12:57:06marazTorne: well, it's fixed for now, i guess time will take care of the rest
12:57:15 Nick fxb__ is now known as fxb (
12:57:34[Saint]maraz: Want to test a theme?
12:57:37Tornemaraz: I'm not sure what benefit DMA actually is on the nano1g, as i implied
12:58:15[Saint](hopefully finished, fingers crossed)
12:58:19TorneDMA mostly just reduces the CPU usage while disk access is going on, rather than helping performance a great deal; it helps a lot in certain cases on some models, but mostly not ;)
12:58:32*n1s can't remember if the emac shifts the result in fractional mode
12:58:37Torneit's a good thing in general which is why I turned it on, but you probably aren't losing a great deal by having it off
12:59:05[Saint]mc2739: Another version of that theme:
13:00:28maraz[Saint]: ew, white ;)
13:00:35*maraz much prefers white-on-black
13:00:56[Saint]maraz: You recognise it
13:01:15[Saint]if you like white on black, use my "Symmetry" theme.
13:01:20*maraz looks it up
13:01:48xlargecrap... i've wrecked a perfectly fine e260 :(
13:02:03maraz[Saint]: i'm more for cabbiev3 although it's a bit cluttered too
13:02:36[Saint]a bit cluttered "too"? :P
13:02:37[Saint]Is that saying something about Symmetry? ;)
13:03:02[Saint]bah! :p
13:03:06marazbut don't worry, the way i see it cabbiev3 is basically the only even remotely usable theme :p
13:03:33marazwell i guess cabbiev2 is not so bad either
13:03:44marazi just dislike using the rockbox logo on the wps
13:03:55marazcluttery, cluttery, cluttery!
13:04:05[Saint]I guess I could make a white on black version of the iLike theme, but, it would look *weird*...
13:05:12[Saint]The WPS is about as close to the OF as I can get it, so I think its time to stop.
13:05:45[Saint]One slight addition for the rec-screen...then, done. Finally.
13:07:46*maraz waits for the graphical theme editor
13:10:52 Join voRia [0] (
13:11:06 Part voRia
13:11:47*[Saint] has a thought and wonders if bieber knows/is aware that viewports can hange negative x/y values
13:12:38[Saint](occasionally I think of small things that may be tricky if forgotten about)
13:15:27CIA-8New commit by nls (r27573): libwmapro: coldfire asm for vector_fixmul_window, gives a speedup of ~13%, drop the add_bias argument for the vector_fixmul_window function, since it ...
13:16:16 Join S_a_i_n_t [0] (S_a_i_n_t@
13:17:01 Quit [Saint] (Ping timeout: 258 seconds)
13:17:19CIA-8r27573 build result: All green
13:26:47 Quit powell14ski_ (Ping timeout: 276 seconds)
13:27:43 Quit n17ikh (Ping timeout: 252 seconds)
13:28:44 Join n17ikh [0] (
13:28:51JdGordonS_a_i_n_t: pretty sure we've brought that up before :)
13:31:14 Join GodEater [0] (
13:31:14 Quit GodEater (Changing host)
13:31:14 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
13:36:40 Join funman [0] (~fun@rockbox/developer/funman)
13:39:41*JdGordon thinks he might be finished with this mammoth skin patch!!
13:39:55JdGordonmitk: can you try the latest patch?
13:42:26JdGordonyou dont need the full zip :)
13:42:31mitkJdGordon: This one is really monster :)
13:42:55JdGordonyeah, cant wait to see the back of it :)
13:43:36mitkIt's fuze, I think
13:44:09JdGordonyes v1
13:57:35mitkJdGordon: It's looking like this patch has solved problems with cabbiev2. I can't see any :) Progress bar is not displayed, but as you sad it is temporary and by intention.
13:58:28JdGordonthe reason AA wasnt drawn is the same as progressbar... I was being generous and adding that back in though
14:00:10JdGordonmitk: ah bugfger... forgot to rebase which is why the patch is huge
14:01:28mitkJdGordon: What about binary you made. Proper or bugger? Can test another one if you like
14:01:45JdGordonproper... I manually did the change on the build box
14:02:08JdGordonS_a_i_n_t: ping
14:05:50***Saving seen data "./dancer.seen"
14:07:50CIA-8New commit by jdgordon (r27574): FS #11490 - Viewportify most of the shipped cabbiev2 wps' By Hayden Pearce. Spot a mistake? mention in the tracker page
14:09:15JdGordonWho owns a archos player?
14:09:34CIA-8r27574 build result: All green
14:11:40pixelma9 of 18... who says "most" ;)
14:12:19JdGordonsomething lost in the translation? :confused:
14:13:44pixelmaok, maybe some of the other cabbiev2s already had viewports
14:14:11JdGordonI thought you were replying to my question :p
14:14:58mitkipod color/photo is good also for fuzev1 and fuzev2 so you have 11 of 18 ?
14:16:55pixelmaJdGordon: do you look for someone with an Archos at all or just teh Player?
14:17:05 Quit bluebrother (Disconnected by services)
14:17:06JdGordonthe player
14:17:09 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
14:19:44pixelmalooked it up on RockboxTesting?
14:20:56JdGordondont need to anymore :)
14:24:42pixelmahas the patch been tested on monochrome or greyscale targets (or in the sim) - maybe targets with remote or some slightly unusual themes/screens (e.g. something with conditional peakmeters like the built-in FM screen on hwcodec targets)? I'd like to help out testing but I'll be quite busy with real life this week,
14:26:39JdGordonI've been running it on my mini2g all week, and the sim
14:27:45 Quit antil33t (Read error: Connection reset by peer)
14:32:41mc2739S_a_i_n_t: I tried your new version and noticed the change in font size in ths FMS. It looks much better now that it does not cut off, but it will probably still cut off for users in an FM region that displays 2 decimal positions.
14:34:53 Join antil33t [0] (
14:37:19 Quit Beta2K (Ping timeout: 240 seconds)
14:37:23 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
14:46:38 Quit JdGordon (Read error: Connection reset by peer)
14:48:05 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
14:50:28mrkikoHi all! Since I've been away forso long - has anything happened regarding the rockbox's speech system?^
14:54:12 Join Beta2K [0] (
14:54:12 Quit JdGordon (Read error: Connection reset by peer)
14:54:12 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
14:54:40 Quit JdGordon (Read error: No route to host)
14:55:16 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
15:14:02 Join kugel [0] (
15:14:04 Quit kugel (Changing host)
15:14:04 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:14:08 Quit JdGordon (Quit: Leaving.)
15:17:34 Quit krazykit (Ping timeout: 240 seconds)
15:18:49 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
15:25:25 Quit Beta2K (Ping timeout: 248 seconds)
15:25:48 Join panni_ [0] (
15:36:36 Join krazykit [0] (~kkit@
15:41:49CIA-8New commit by kugel (r27575): Factor out opening and removing DIRCACHE_FILE into separate functions.
15:42:30 Quit t0rc (Remote host closed the connection)
15:43:17 Join jgarvey [0] (
15:43:48CIA-8r27575 build result: All green
15:47:11kugelS_a_i_n_t: are you bored?
15:47:35kugelif yes, you could make a 320x240 cabbiev2 for touchscreens :)
15:47:50kugelI'm going to need a 320x480 version too :p
15:51:50tmztwhat hardware is hvga?
15:53:01kugelmy phone
15:53:57 Quit mitk (Quit: Leaving)
16:00:11 Join thefirstM [0] (
16:03:27thefirstMI am having a problem with writing rockbox.ipodx to the OSOS partition on my Ipod Nano2G with ipodpatcher. I am following the directions on the ipodpatcher wiki page, but sometimes it says that it couldn't read the header from rockbox.ipodx and fails to work. Other times, it works just fine. Does anyone know what might be up with this?
16:03:50balintxwhy limited bookmarks ? :/
16:03:59thefirstMI filed a bug report with a file that won't work. It is bug #11499.
16:05:52***Saving seen data "./dancer.seen"
16:06:28pixelmabalintx: "limited" how?
16:06:51balintxit says I can't add more bookmarks
16:06:57balintxin text_viewer
16:07:07 Join hebz0rl [0] (
16:07:53balintxit is very fustrating while reading ebook :/ okay dont read ebook on an mp3 player.. but if there is textviewer.. Why can't I add more bookmarks? Why not unlimited?
16:12:02kugelbalintx: because memory is not unlimited on these devices
16:12:45balintxokay, thanks..
16:13:04kugelI'm sure you can delete older bookmarks
16:18:07 Quit bieber (Ping timeout: 276 seconds)
16:18:24 Join bieber [0] (
16:22:30 Join Beta2K [0] (
16:24:13 Quit thefirstM (Remote host closed the connection)
16:24:49 Join thefirstM [0] (
16:27:10 Quit Beta2K (Ping timeout: 258 seconds)
16:27:51 Join Beta2K [0] (
16:30:19 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
16:37:00 Quit Beta2K (Ping timeout: 260 seconds)
16:44:18 Quit thefirstM (Remote host closed the connection)
16:53:29kugelgevaerts: I think I converted almost everything now, but that I have potentially overlooked some places. but I would rather not search for these places now and fix them when needed instead. do you want to have a final look at the patch?
16:53:58kugelor, I better announce it in my today's weekly report and wait a few more days
16:54:41kugel(luckily git allows me to get new stuff done even if it depends on non-yet-in-svn stuff separately)
16:57:24 Join domonoky1 [0] (
16:58:36 Join cornflac [0] (
16:58:52 Nick cornflac is now known as cornflaku (
16:59:32 Quit domonoky (Ping timeout: 245 seconds)
17:01:10gevaertskugel: I don't know if I'll have much time the next week or so
17:12:24 Join stooo [0] (
17:12:32 Part stooo
17:15:53 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:18:57 Quit BlakeJohnson86 (Ping timeout: 265 seconds)
17:19:54 Join Beta2K [0] (
17:20:23CIA-8New commit by kugel (r27576): Fix cabbiev2.220x176x16.wps
17:21:59 Quit pamaury (Remote host closed the connection)
17:22:06CIA-8r27576 build result: All green
17:32:20 Quit mt (Ping timeout: 258 seconds)
17:33:27 Join mt [0] (~mtee@rockbox/developer/mt)
17:35:46S_a_i_n_tmc2739: I was unaware that the decimal positions in the .fms were regional... :-/
17:35:49S_a_i_n_tgood to know.
17:36:39S_a_i_n_tDid you notice the new font in the main menu?
17:36:54S_a_i_n_tIts about as close as I can get to the OF now.
17:38:37 Join sudoman [0] (d8ecfceb@gateway/web/freenode/ip.
17:38:51 Quit mc2739 (Ping timeout: 258 seconds)
17:40:36 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
17:42:09 Join fdinel [0] (
17:42:13 Quit antil33t (Read error: Connection reset by peer)
17:42:20 Join antil33t [0] (
17:45:37 Quit fdinel (Client Quit)
17:49:26S_a_i_n_tkugel: ping?
17:49:46kugelS_a_i_n_t: pong
17:50:01S_a_i_n_tDo you have any specific requirements for this theme?
17:50:38kugelit should have some touchregions so that it works with absolute mode
17:50:44 Quit funman (Quit: free(random());)
17:50:47S_a_i_n_tI really need to get the CabbieV2 viewportifying done, but, this is easier since I'll be starting from a decent base and not working from scratch ;)
17:51:06kugelI think the ondas have a touchscreen wps (but IIRC it's doesn't work really well)
17:51:43S_a_i_n_tthe ones I have left to do are slightly more difficult, and if I wait until I have finished those to do your theme it may be some time, so, I'll try whipping something up now.
17:52:21S_a_i_n_tI'll blame you for distracting me though :P
17:53:20kugelno problem :)
17:54:24S_a_i_n_tWhat's the screen resolution of your device? I'll assume you'll want that one first.
17:55:13kugel"<kugel> if yes, you could make a 320x240 cabbiev2 for touchscreens :)<kugel> I'm going to need a 320x480 version too :p"
17:55:49 Nick fxb is now known as fxb__ (
17:55:53kugelI think 320x240 is more important because there's actually a target with touchscreen and that resolution (cowond2). I'm amazed there's still no proper cabbie port for it
17:56:01*S_a_i_n_t wonders what the two sizes are for.
17:57:40kugel320x240 is the current RaaA, and 320x480 will be the first android one
18:02:52 Join Barahir [0] (
18:03:04 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:03:34 Quit Barahir_ (Read error: Operation timed out)
18:03:53 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:04:37*S_a_i_n_t wonders if the new touchscreem syntax is documented anywhere.
18:04:55S_a_i_n_tI assume it isn't still in pipes as customWPS suggests.
18:05:05S_a_i_n_tand, the manual neglects touch altogether
18:05:07 Join powell14ski_ [0] (
18:05:54***Saving seen data "./dancer.seen"
18:12:12 Quit TheSeven (Ping timeout: 258 seconds)
18:15:24 Quit Xerion (Ping timeout: 276 seconds)
18:23:05S_a_i_n_tHmmm, I was thinking I could be a cheap bastard and just simply add touchscreen areas...but, no.
18:23:11S_a_i_n_tIt needs a lot more than that.
18:23:21S_a_i_n_tThe whole layout needs to be rethought...
18:25:04kugelS_a_i_n_t: I think the onda cabbie just pop ups some icons once you touch the screen
18:25:37kugelwith %Ti
18:26:05S_a_i_n_tYes, that's what I'm going to have to do here.
18:26:20S_a_i_n_tAlso, I'm probably going to need to make some new icons.
18:27:42 Quit mrkiko (Quit: leaving)
18:28:41S_a_i_n_tI think I'll look more closely into this in an more appropriate time tomorrow...The coding I can do, but, the creativity fails me.
18:28:55pixelmatouchscreen tags are only explained in manuals of touchscreen targets, currently only the D2 one available on the site
18:28:56S_a_i_n_tI'm wanting to give it an awesome conditional UI.
18:29:10kugelS_a_i_n_t: there are some icons in wps/cabbiev2 (in svn)
18:29:10S_a_i_n_tSo, the screen doesn't needlessly display icons.
18:29:49S_a_i_n_tkugel: yeah, I may or may not need new ones yet...not too sure.'
18:30:04kugelS_a_i_n_t: ohh wait a second, I just see 'cabbiev2.320x240x16.mrobe500.wps'
18:30:14kugelI'm assuming that should work
18:30:24S_a_i_n_tit's monochrome no?
18:30:30S_a_i_n_teasy to fix if it is.
18:30:57pixelmathe MRobe500 isn't. the MRobe100 is monochrome
18:31:13S_a_i_n_tkugel: Well, that's one of the ones I have left to, I'll put it at the top of my list for the morning.
18:31:19pixelmathe 500 has a 640x480 coloured touchscreen
18:31:43kugelI think it was once (and can still be) configured for 320x240, that's probably why it has a cabbie
18:31:57pixelma(and was set to use as "320x240" once)...
18:32:24S_a_i_n_tkugel: As you may see...that is a "pre-viewport" theme.
18:32:28 Quit krazykit (Ping timeout: 260 seconds)
18:32:36S_a_i_n_tAnd the use of %?C for everything is *terrible*
18:32:47S_a_i_n_tOnce I fix that tomorrow...she'll be all yours.
18:33:19S_a_i_n_tThough...I'll still be wanting to see if I can't add some "cool" to it.
18:33:23kugelit has touchregions, enough for me right now :)
18:33:32S_a_i_n_theh ;)
18:33:47 Join krazykit [0] (~kkit@
18:33:52kugelbut feel free to improve it!
18:33:53S_a_i_n_tI assume RaaA will use the new parser when it goes in?
18:34:01S_a_i_n_tWhen it does go in, that theme will break.
18:34:16S_a_i_n_t(which is where I come in)
18:34:29 Join funman [0] (~fun@rockbox/developer/funman)
18:36:43kugelS_a_i_n_t: why will it break?
18:36:53S_a_i_n_tNo viewports
18:37:02kugelis that a problem?
18:37:14S_a_i_n_tEverything will display all weird, or not at all.
18:37:15funmani'm making AMS bootloaders & mkamsboot binaries
18:37:29kugelI don't think we should drop no-viewport-themes support
18:37:38S_a_i_n_tThis is the reason I need to finish "viewportifying" the themes.
18:38:06S_a_i_n_tIt's not dropping it, but, it is limiting what can be done.
18:38:32kugelI think we'll have a talk about that
18:38:53S_a_i_n_tPrimarily I think it affects conditionals
18:39:01kugelviewports make themes terribly target depending, something that's undesirable for some people
18:39:24S_a_i_n_tfrom what I've seenon the tracker for the new skin parser things wont necessarily display correctly if not in a viewport.
18:39:51S_a_i_n_tText is fine as I understand it.
18:40:27kugeltext only/no backdrop is still a perfectly valid approach
18:40:29S_a_i_n_tStatic images can still be drawn in the default viewport without hassle
18:40:43S_a_i_n_tIt's when they are conditional that its foolish to not have viewports.
18:40:55S_a_i_n_tEvery theme should use them IMO
18:41:06kugelhow's that foolish?
18:41:20S_a_i_n_tNot clearing properly.
18:41:48kugelthat's easy to solve
18:41:55S_a_i_n_tapparently not.
18:42:02kugelsvn does it
18:42:19S_a_i_n_tIt actually doesn't.
18:42:35S_a_i_n_tA "false" viewport isn;t cleared in SVN
18:42:47S_a_i_n_t(for text)
18:43:32S_a_i_n_tPerhaps I'm missing something vital though... jd could explain it a lot better.
18:43:43kugelwell, it certainly did at some point
18:44:45S_a_i_n_tThe FM screen is a pretty good example of viewports not clearing at the present in SVN
18:45:05S_a_i_n_tthe poor update makes it easy to see.
18:45:29kugelthat's a different problem though
18:46:38kugelit basically only ever refreshes dynamic content, not static one when it needs to (e.g. when changing presets=
18:48:04S_a_i_n_tMy basic rule of thumb is "if it is dynamic in any way, stick it in a viewport", I'm pretty sure that the tracker for the new skin parser lists some of the areas in which not using a viewport can lead to odd display.
18:48:15kugelthe images in the touchregions in that mr500 theme are properly cleared after the %Tl timeout wears off, only the albumart draws over
18:49:40 Join Jerom [0] (~jerome@
18:51:14kugelthat wps is crap though :)
18:52:23 Join com_kieffer [0] (
18:53:05S_a_i_n_twhat did I break on 220x176?
18:53:54S_a_i_n_thah, again... :/
18:55:14S_a_i_n_tI uploaded a similar fix to for the Video, I seem to like messing up hold icons.
18:55:47S_a_i_n_tif someone is in a committing mood, I'm sure Video owners want their hold icon back.
18:56:05S_a_i_n_tAnd, ....sorry.
18:57:42CIA-8New commit by kugel (r27577): FS #11490 by Hayden Pearce ...
18:57:52*S_a_i_n_t nods.
18:59:29CIA-8r27577 build result: All green
19:04:05funmanfunman/sandisk-sansa/"> <- mkamsboot 1.4 , AMSv1 2.0 bootloaders, AMSv2 3.0 bootloaders
19:04:20funmani tested all the AMSv2, fuzev1 and clipv1
19:04:43funmanAMSv1 bootloaders include USB
19:05:12funmanmkamsboot include not-yet-committed-because-bootloaders-with-usb-weren't-released diff which disables dualboot on USB plug
19:05:40funmanAMSv2 bootloaders work with the new hardware revisions (I also made a new one for the clipv2 while i was at it)
19:05:49funmanplease test m200v4/c200v2/e200v2
19:09:28kugelfunman: works great on my fuzev1
19:13:24funman(in case i forget) i used r27576
19:13:31gevaertskugel: I suspect that this possible RaaA GUI/daemon split might help the n900 too
19:19:32 Quit sudoman (Quit: Page closed)
19:20:25 Quit com_kieffer (Read error: No route to host)
19:21:12 Join com_kieffer [0] (
19:22:37CIA-8New commit by nls (r27578): libatrac: Coldfire asm for atrac3_iqmf_dewindowing, speeds up decoding of my atrac3 rm file by 45%. Based on preliminary patch by Andree Buschmann
19:23:05 Join captainewkl [0] (2669ecc2@gateway/web/freenode/ip.
19:23:20*S_a_i_n_t wonders why the sim doesn't save it to the config when you select a .kbd file
19:23:39S_a_i_n_teven if your .cfg includes said .kbd file, it is not laoded.
19:24:13CIA-8r27578 build result: All green
19:24:40kugelS_a_i_n_t: the file needs to be in /.rockbox IIRC
19:24:55S_a_i_n_tkugel: It is.
19:25:26 Quit jhMikeS (Ping timeout: 258 seconds)
19:25:36S_a_i_n_tIf I want to test my .kbd in the sim I need to load it manually each time.
19:25:58 Join jhMikeS [0] (
19:25:58 Quit jhMikeS (Changing host)
19:25:58 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
19:26:51 Join Buschel [0] (
19:26:58S_a_i_n_tUes, it id definitely saving it to the config on selection, but the sim fails to load it for some reason.
19:27:14funmanjhMikeS: do you have a m200v4 to test a bootloader?
19:27:35Buscheln1s: wow!
19:27:44Buschel45% faster? ;)
19:32:36 Quit juju2143 (Ping timeout: 240 seconds)
19:33:18 Join halmi [0] (
19:34:56 Quit simabeis (Ping timeout: 260 seconds)
19:35:28 Join simabeis [0] (
19:35:56 Quit slck (Ping timeout: 260 seconds)
19:39:58pixelmafunman: I believe domonoky1 has an m200 too but don't know which version it is
19:40:10 Join FlynDice [0] (~FlynDice@
19:40:28funmanyes he has a v4, i saw llorean & jhMikeS also have one on RockboxTesting
19:41:00 Join fdinel [0] (
19:41:37 Join stripwax [0] (
19:41:40FlynDicefunman: Your bootloader works fine for e200v2 happy hacing!
19:41:44 Quit Jaykay (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
19:41:49FlynDiceer hacking even
19:41:59funmanthanks ;)
19:42:22funman(not hacking anything atm)
19:42:49FlynDicehah, me neither and not for a long time... ;-(
19:44:14 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:44:38FlynDiceI gave you all of my spare 15 mins for today ;-) cya!
19:44:45 Quit FlynDice (Remote host closed the connection)
19:46:14funmantoo bad i can't test on the c200v2 i have .. i can't upgrade bootloader anymore because storage is broken
19:47:01 Join juju2143 [0] (juju2143@gateway/shell/
19:54:05funmanranma: can you test the bootloader on your c200v2? (see logs at 19:04)
19:56:59 Quit fyrestorm (Read error: Connection reset by peer)
19:57:18 Join fyrestorm [0] (
19:57:18 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
19:57:56 Quit bieber (Ping timeout: 240 seconds)
19:58:12 Join bieber [0] (
19:58:20 Join BlakeJohnson86 [0] (
20:05:51 Quit com_kieffer (Quit: Lost terminal)
20:05:56***Saving seen data "./dancer.seen"
20:06:39 Quit antil33t (Read error: Connection reset by peer)
20:06:42 Join lestatar [0] (
20:06:45 Join antil33t [0] (
20:07:53ranmafunman: Also got my used fuzev1 today
20:08:25funmannice, did you reproduce the bug that borka seen?
20:09:11 Join stooo [0] (
20:10:03 Part stooo
20:10:29ranmaNot yet, I haven't set up the build environment on this machine yet
20:11:09ranmaI'm also seeing some problem building the eabi toolchain on amd64
20:11:28ranmaIt fails with some truncated/zero-size .o file
20:11:37funmanyou have enough disk space?
20:11:39ranmaOld toolchain just finished building
20:12:00ranmaYes, I'm pretty sure. Only about 400MB of 4G in /tmp were used
20:14:49funmani can pack my binaries if you want
20:16:42ranmaI'll figure out whats going wrong with the toolchain later, for now I started configure with −−no-eabi
20:19:46kugelranma: just as a warning that toolchain is possible not going to work anymore at some point in the future
20:20:11 Join lestatar_ [0] (
20:21:40 Quit lestatar (Ping timeout: 260 seconds)
20:21:49 Nick lestatar_ is now known as lestatar (
20:26:23 Join Topy [0] (
20:29:13 Quit funman (Quit: free(random());)
20:30:03ranmac200v2 works fine with the new bootloader and mkamsboot
20:30:21 Quit T44 (Ping timeout: 265 seconds)
20:31:48 Quit DerPapst (Quit: Leaving.)
20:35:56 Join saratoga [0] (482b333e@gateway/web/freenode/ip.
20:36:26*Buschel measures a nice speed up on ARM for wma pro
20:36:38Buschelsvn: 49.6 MHz, patched: 37.5 MHz
20:36:42saratogai wondered why windowing on WMA Pro was somehow almost as slow as the MDCT :)
20:37:59Buschelneed to test compilation on some other target to test the iram usage...
20:39:57ranmaBootloader 2.0 doesn't seem to work on my Fuze v1 though, it just hangs at the bootloader screen, without any error message...
20:40:25 Nick Strife1989 is now known as Strife89 (
20:41:45bieberS_a_i_n_t: It should render negative coordinates correctly. Did you find a problem, or just thinking?
20:42:08 Join sasquatch [0] (
20:43:03 Join dockimble [0] (~dockimble@
20:43:05 Join anewuser [0] (anewuser@unaffiliated/anewuser)
20:43:47 Quit hebz0rl (Remote host closed the connection)
20:47:38saratogado i just completely not understand what that guy on the forums is saying about AA fonts?
20:48:10Buscheln1s: you there?
20:49:06ranmaOk, more a problem with the newly compiled rockbox hanging, (because of old toolchain maybe) bootloader 2.0 works fine
20:49:06saratogadoesn't it work by just computing what you want to blur in advanced, and then filtering those pixels when you render the font?
20:50:33 Join mt_ [0] (~mtee@
20:50:43Buscheln1s/mt: how much MHz does wmapro need on mcf5249/5250 for realtime?
20:51:56 Quit mt (Ping timeout: 240 seconds)
20:51:59 Nick mt_ is now known as mt (~mtee@
20:58:02 Join Xerion [0] (~xerion@
20:58:56pixelmasaratoga: IIUC he proposes precalculating the alpha channel value of a pixel and then combining e.g. 10% opacity of foreground colour with the background, it's somewhat a step between computing everything on target and fully precalculating against a fixed background (a bit of my interpretation)
21:03:12pixelmaI believe it is what the antialiasing patch does
21:04:40mtBuschel: I don't have a coldfire target :(
21:04:49Buscheldo you have numbers?
21:05:12mtNope, just those on the CodecPerformanceComparison page
21:05:24 Quit sasquatch (Ping timeout: 240 seconds)
21:05:55Buschelok, then let's ask for help :)
21:06:34mtDo you have a patch that needs to be tested on a CF target ?
21:06:45Buschelanyone with a coldfire target (mcf 5249 or mcf5250) volunteering for some wmapro tests?
21:06:46mtor is that 11498 ?
21:06:55Buschelyes, it is 11498 v03
21:07:55 Join DerPapst [0] (
21:08:23 Quit stripwax (Quit:
21:14:17mtBuschel: Nice work :)
21:14:33Buschelmt: thanks
21:14:43 Join hebz0rl [0] (
21:15:12Buschelmt: not sure though whether it is faester to move tmp to IRAM or move some of the sine_windows to IRAM instead...
21:15:28 Quit Xerion (Quit: )
21:15:47 Join Xerion [0] (~xerion@
21:18:36 Quit mt (Ping timeout: 240 seconds)
21:20:48 Join sasquatch [0] (
21:22:58 Quit Jerom (Quit: Leaving.)
21:23:24*bertrik wants to commit
21:23:50 Join Jerom [0] (~heidi@
21:26:01 Join Kitr88 [0] (
21:27:13S_a_i_n_tbieber: Just thinking outloud.
21:27:21bieberOkay, cool
21:27:25bieberI'm pretty sure it works right
21:27:55S_a_i_n_tI'm not sure the documentation actually mtntions negative coords in, just checking ;)
21:29:23 Quit Kitar|st (Ping timeout: 276 seconds)
21:29:31bertrikWhat are negative coordinates good for?
21:30:26 Quit Kitr88 (Ping timeout: 260 seconds)
21:31:21bieberNegative coordinates start from the right/bottom edges instead of top/left
21:31:55bieberSo %V(0, 0, -50, -50, -) would start at the top left corner, and extend out to 50 pixels short of the edges of the screen, for example
21:32:39 Quit halmi (Read error: Connection reset by peer)
21:33:59 Join mt [0] (~mtee@rockbox/developer/mt)
21:34:09*S_a_i_n_t has a feeling ti works slightly differently...but could be wrong.
21:34:41mtstupid connection
21:34:50mtBuschel: I think putting tmp in iram is better
21:35:10Buscheljust measuring for PP5022
21:35:27mtoh sorry I was thinking of sincos_lookup not the windows
21:36:14 Join Kitar|st [0] (
21:36:53mthmm .. I still think tmp might be faster
21:36:58pixelmaS_a_i_n_t, bieber: I *believe* you set the x- and y-coordinate to something negative and width and height stay a positive number but I never used it
21:36:59Buscheltmp mt: difference is minor
21:37:26S_a_i_n_tpixelma: Yes, you're correct there.
21:37:34bieberpixelma: According to CustomWPS at least, either one could be negative, no?
21:38:02bieberThe x and y values may be negative which means "pixels from the far edge".. i.e x == -50 means 50 pixels from the right edge
21:38:02bieberThe width and height values may also be negative which means almost the same as above...
21:38:18bieberUnless we've changed things since then
21:38:29pixelmahmm, it says so
21:38:42S_a_i_n_tAs I understand it, -43 for example extends from the top right corner, 43px from the dig?
21:38:56Buschelmt: win -> 38.1 MHz, tmp -> 37.5 MHz
21:39:20Buschelmt: will now keep tmp in iram and move as much win to iram as possible as well
21:40:08S_a_i_n_t-14x -40y is the same as x= width -40 and y=height -40 as I understand it.
21:40:20S_a_i_n_tIts a weird way of doing things, but, usefull sometimes.
21:40:26mtBuschel: Great ! I'm still working on producing non interleaved output, there's something I'm missing and not sure what it is yet.
21:41:13*mt really wants to see how much speedup Buschel's changes achieve on ColdFire
21:41:53 Quit BlakeJohnson86 (Ping timeout: 265 seconds)
21:43:45Buschelreached a minor additional speed up: 37.3 MHz now
21:44:27mtWhat's that in percentage now ?
21:44:49mt(% realtime I mean)
21:44:54 Join lulinha [0] (~dummy@
21:44:54 Part lulinha
21:48:18Buschelmt: on PP5022 it is 214%
21:48:39mtBuschel: Impressive ! :)
21:48:57 Join Lear [0] (chatzilla@rockbox/developer/lear)
21:49:25mtsvn is ~150-160% iirc
21:50:16Buschelyes, 161%
21:51:32 Join bunnyboi [0] (
21:56:38kugelbertrik: go for it!
21:56:56kugelI had the same on my todo list too
21:58:42bertrikI haven't tested a target with TEA5767 yet, but I'm pretty sure I did it right
22:00:09Buschelamiconn/pixelma: wasn't it one of you who owns an m5/x5? willing to support a short test regarding decoding speed of libwmapro? FS #11498
22:00:28saratogacan the wma pro windows be shared with the mdctlib trig tables?
22:00:41saratogaIIRC they can be, at least for the smaller sizes
22:01:11Buschelsaratoga: good point.
22:01:11saratogaalthough you have to stride them a few entries at a time, which is slower on arm7
22:01:22saratogaso maybe its not worth it, at least on PP
22:03:52saratogai told mt via email that i thought we could save 4MHz or so on PP by using the mdct lib, but i guess the bigger savings is just reducing memory usage by merging all these trig tables
22:05:57***Saving seen data "./dancer.seen"
22:07:35 Quit mt (Ping timeout: 258 seconds)
22:09:01 Join wodz [0] (
22:09:31merbananwmapro and wmastd should use the same windows
22:09:48 Join mt [0] (~mtee@rockbox/developer/mt)
22:10:26wodzBuschel: I can test on mcf5249. Are there some test wmapro files?
22:10:58Buschelwodz: perfect!
22:12:46wodzBuschel: which file do You prefer? I have rather slow connection so I would avoid downloading all of them
22:13:12BuschelI have used wmapro_173k.wma from the rockbox site
22:15:19wodzshould I patch some specific revision or just svn head?
22:15:21CIA-8New commit by bertrik (r27579): FM tuner region code cleanup - FS #11492 by me. ...
22:17:01CIA-8r27579 build result: All green
22:17:56 Quit anewuser (Quit: Sixth edition of 25 samples + 7days + Remix Phase July 19th - July 26th)
22:21:22lestatarhi all...sorry to interrupt - am having issues with themes not working on gigabeat most recent build
22:21:45 Quit amiconn (Quit: - Chat comfortably. Anywhere.)
22:21:59lestatarupgraded to latest build breaks the theme i'd like
22:22:16lestataractually, all other themes i have installed do not work as well.
22:22:42lestatarWPS screen is always the standard black/white text only rb screen, the one with the meters on the bottom
22:22:46lestatarany ideas?
22:23:54 Join amiconn [0] (quassel@rockbox/developer/amiconn)
22:25:40pixelmathere was a change to the WPS syntax not long ago and older themes need to be updated - many of the ones on the theme site were
22:25:55 Join balintx_ [0] (
22:26:20lestatari see
22:26:22 Quit balintx (Ping timeout: 258 seconds)
22:26:26lestatarthanks pixelma
22:26:38lestatarhow can i determine if the theme i like is compatible?
22:27:48pixelmawhere did you get it from? You can also update them yourself
22:27:48Buschelwodz: svn head against patched v03 FS #11498
22:27:48lestatarfrom rb site/themes
22:27:48lestatarit indicates it works with current build
22:27:48 Join drhodes [0] (
22:27:50lestataris the only one 1 am interested in using...
22:28:25wodzBuschel: ok this will take me a while to sync with svn.
22:29:36pixelmalestatar: looks like this one was updated, just reinstall and select it again
22:29:43Buschelwodz: ok
22:29:55Buschelmt: how is your progress with the non-interleaving?
22:30:00lestatarhuh, ok...will do that pixelma...thanks a ton :-D
22:30:10 Join BlakeJohnson86 [0] (
22:30:39wodzBuschel: v03 or v04 from FS #11498 ?
22:30:49Buscheloops, v04 of course
22:31:36 Quit n17ikh (Ping timeout: 240 seconds)
22:31:42mtBuschel: It's working now but it's not much different semantically .. instead of copying sample-by-sample, I do a memcpy for the whole output. Not sure if this will be that much faster.
22:31:52mtBut I'm working on another idea now.
22:34:14wodzmemcpy is quite efficient on CF because of line transfers
22:35:13Buschelmt: ok, just did the same out of interest. the speedup through non-interleaving is huge btw, another 1.0 MHz on PP5022
22:37:24 Quit seani (Ping timeout: 240 seconds)
22:37:55 Quit bmbl (Quit: Bye!)
22:39:24 Quit BlakeJohnson86 (Ping timeout: 265 seconds)
22:50:09 Join ascheel [0] (~ascheel@ampache/staff/ascheel)
22:50:53 Join seani [0] (~seani@
22:51:11ascheelThis is entirely off the wall, but how do I find out if there's a way to help out further progression on Sansa Fuze compatibility with Rockbox? I see the Sansa line is almost complete minus USB.
22:58:07AlexPIt's easy
22:58:10AlexPget coding :)
22:58:28AlexPFind something missing you want to add, look at the code, get to work :)
22:58:51 Join n17ikh [0] (
22:58:53Buschelmt: you have a non-interleaving change that produces an identical crc32 checksum?
23:00:42mt Buschel : I haven't checked .. but I think I wouldn't have too, my idea is to work directly on the "channel[i].out" buffers without copying anything. I think I'm almost there.
23:01:04Buschelmt: ok, I'll drop my changes and wait for yours then
23:01:08mtAh wait
23:01:25mtI think I know why you're not getting identical checksums
23:02:06mtBuschel: when copying sample by sample, I scaled the samples up by one bit, which of course wouldn't be done when memcpy'ing.
23:02:13bertrikascheel, USB should be working for fuze v1, although the bootloader still boots into the original firmware when it detects a USB connection
23:03:00mtBuschel: this was a hack, I'll look into the code better to see where the samples should be scaled up. But I think you could ignore the non identical checksums now :)
23:03:43wodzBuschel: mcf5249 wmapro_173k.wma; svn: 179.40MHz 69.22% realtime; wmapro_v04.patch: 117.63MHz 105.57% realtime
23:04:02Buschelahhh, you wanted to avoid DSP_SET_SAMPLE_DEPTH = 16
23:05:03Buschelwodz: great, we're there! realtime ;)
23:06:14*mt high fives Buschel !
23:06:38*Buschel high fives mt !
23:06:43wodznice work
23:06:51*mt just got rid of a 192k buffer from wma pro
23:07:05Buscheland this is for the slow mcf5249, the mcf5250 will be faster
23:07:55Buscheland with mt's non-interleaving change it will faster as well
23:08:32wodz5249 is the slowest swcodec target isn't it?
23:08:37Buschelwodz: crc32 is the same for svn and patched?
23:09:00wodzBuschel: didn't checked crc
23:09:03AlexPwodz: Depends
23:09:16AlexPwodz: Faster clock that most pp, but not dual core
23:09:21AlexPiram differences too
23:10:59 Quit Jerom (Read error: Connection reset by peer)
23:11:50 Join Jerom [0] (~heidi@
23:18:49CIA-8New commit by bieber (r27580): Theme Editor: Changed binary name to rbthemeeditor
23:19:37mtwodz: Could you test my patch ? (last one in #11498)
23:20:29wodzmt: sure, just after I check crc for Buschel :-)
23:20:29CIA-8r27580 build result: All green
23:20:59mtwodz: Alright, thanks ! :)
23:21:31wodzmt: should I apply patch on top of svn or on top of 11498 v04 ?
23:21:44mtsvn head
23:21:54mtwodz ^
23:23:48 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
23:24:54mtBuschel: Which file runs 161% on arm (svn) ? the 171k one ?
23:27:11mtBuschel: em .. there are 173k and 141k files, which one of those ? :)
23:27:48mt(and 271k too, so there's a mistake in some digit there :P )
23:27:52Buscheloops, just saw "17x" -> the 173k file
23:28:02mtalright :)
23:32:10S_a_i_n_tIs anyone familiar with the syntax for conditional UI viewports?
23:32:26mtBuschel: 170% - 47MHz on arm
23:32:45mtnot bad I think
23:33:13Buschelmt: indeed, what CPU do you have?
23:33:13wodzBuschel: crc32 is the same for svn and v04 patch
23:33:41mtBuschel: pp5024 iirc (sansa e200)
23:33:46Buschelwodz: good :o)
23:34:43BuschelBuschel: should be the same than my PP5022. in this case your patch saves another 2.6 MHz, quite a lot. We should submit all of our patches :o)
23:35:26wodzmt: Your patch does not apply
23:35:54Buschelmt: would you mind if I would submit my (bigger) patch first?
23:36:10mtwodz : What's the error ?
23:36:20mtBuschel: Not at all
23:37:28CIA-8New commit by bieber (r27581): Theme Editor: Added ditor: Added a version number to the project file
23:37:41wodzmt: this is the .rej content
23:38:58CIA-8r27581 build result: All green
23:39:06mtwodz : Weird, this is the only failed hunk ? all the rest applied ?
23:39:29wodzyes. weird in did
23:39:41wodzI don't see why it failed
23:41:39mtwodz: (quick and dirty solution) : here's my wmapro.c file
23:41:48wodzI see now - in your patch there is int32_t decoded[BUFSIZE]; removed but in svn it is static int32_t
23:42:00mtah that's why
23:42:13mtI didn't update after nils's changes
23:43:10CIA-8New commit by Buschel (r27582): Submit part of FS #11498. Major speedup for WMA Professional on ARM and Coldfire CPUs. Introduce asm routines for multiplications, move arrays with ...
23:44:12*Buschel crosses fingers
23:44:45CIA-8r27582 build result: All green
23:44:58mtBuschel scares another codec into the real time zone \o/
23:45:33Buschelthat's what I like :)
23:46:35Buschelwould still be interesting to see the nubmers for mcf5250
23:46:50mtyeah me too.
23:46:56Buschelmt: now it's your turn ;)
23:46:59wodzBuschel: maybe it's time for mp4 in rockbox :-)
23:47:20 Quit ender` (Quit: It is difficult to produce a television documentary that is both incisive and probing when every twelve minutes one is interrupted by dancing rabbits singing about toilet paper.)
23:47:31Buschelwodz: if you would take a look at libfaad... it
23:47:42Buschel*it is not much fun to deal with it...
23:48:20*mt faints on the thought of working with libfaad
23:48:53wodzBuschel: I was thinking about divix and company
23:49:17wodzmore and more players have big screens and quite powerfull cpus
23:49:33Buschelwodz: aah, ok. never worked on this though
23:49:49*Buschel has a less powerful cpu in his dap
23:51:39wodzmt: 70.71% realtime 175.62MHz (svn: 179.40MHz 69.22% realtime)
23:52:25mtonly 1.5% :(
23:52:49wodz3.8MHz still
23:53:05mtyeah maybe not that bad.
23:53:13mtwodz: Thanks for testing :)
23:53:21wodzyou welcome

Previous day | Next day