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

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

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

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2010-05-06

00:00:00 Quit domonoky (Read error: Connection reset by peer)
00:01:59Luca_Suhm... it's stuck at the rockbox logo (or the delays are really long...)
00:02:45FlynDicewith or without uSd inserted?
00:03:14Luca_Swith. without it boots fine
00:03:15 Quit CGL (Quit: Soy spammero ---- \m/ d(>.<)b \m/)
00:03:36FlynDicewhat happens when you insert it now
00:05:48Luca_Sthree attempts: 2 freezes, 1 microSD init failed -3
00:06:53Luca_S4th attempt: it works :/
00:08:34 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
00:11:44 Quit loveless (Quit: loveless)
00:11:53FlynDicehrm, Not sure what's going on there, does it continue to work?
00:13:21FlynDiceI've got to go again, If you could experiment with those delays you may find a combo that works. I'll be back later.
00:13:27Luca_Sprobably the working attempt was a glitch caused by the quick series of reboots.
00:13:44Luca_Sthe most frequent result is freezes
00:13:45Luca_Si'll try changing the delays
00:13:52Luca_Sthank you
00:14:40FlynDiceThank You for taking the time to experiment!
00:19:14 Quit Luca_S (Quit: CGI:IRC (Ping timeout))
00:19:36 Join Luca_s [0] (
00:24:08 Join pixelma [0] (quassel@rockbox/staff/pixelma)
00:28:08 Quit apoelstra (Quit: home-->work)
00:28:54 Join loveless [0] (~loveless@2a01:198:26e:0:216:cbff:feb9:8c8d)
00:29:39 Quit ender` (Quit: Engineers think that equations approximate the real world. Scientists think that the real world approximates equations. Mathematicians are unable to make the connection.)
00:36:44 Quit petur (Quit: Zzzzz)
00:36:49 Quit Luca_s (Quit: CGI:IRC)
00:40:27 Quit stripwax (Ping timeout: 252 seconds)
00:43:02LloreanSo, is there a process yet for taking things in front of the RSB?
00:48:18gevaertsConvince a majority of RSB members?
00:52:10LloreanI was just wondering if there was any process for formally asking for a decision on an issue, or even just bringing up an issue if necessary.
00:52:17*Llorean doesn't have anything at the moment, it just crossed his mind
00:52:19***Saving seen data "./dancer.seen"
00:52:55gevaertsMaybe the RSB should meet and discuss this :)
00:53:39 Quit JohannesSM64 (Ping timeout: 252 seconds)
00:53:42LloreanMaybe I'll add "decide on a process to put something before the RSB" as a devcon thing
00:54:35gevaertsI personally think that the RSB should probably either have a more active role and some procedures, or be scrapped. I'd prefer the former
00:55:32 Quit lpereira (Quit: Leaving.)
00:57:53LloreanHonestly, I always felt that if I could see to it that an issue could be solved without it needing the attention of the RSB, I was doing my job as an RSB member back when I was in it.
00:58:21gevaertsOh, certainly. That's not the problem I see
00:58:22LloreanAt the same time, I think that the RSB itself shouldn't hesitate to step in and say "this disagreement is hurting us, we're going to take this, we'll have an answer for you in a few days"
00:59:22 Quit arun (Read error: Operation timed out)
00:59:36gevaertsThe main problem I see is that if the RSB does nothing / is not needed for a few more years, it will be elected by about three voters. If that happens, it might as well not exist
01:00:24 Quit markun (Ping timeout: 268 seconds)
01:00:30 Join markun [0] (~markun@rockbox/developer/markun)
01:00:31gevaertsProactive RSB action could be a good idea, yes
01:00:48LloreanI think it doesn't really serve a purpose if it's not proactive.
01:01:10 Join arun [0] (~arun@unaffiliated/sindian)
01:02:09gevaertsIt could, if there are clear procedures to call it into action, with reasonable thresholds
01:02:25 Quit leavittx (Read error: Operation timed out)
01:02:45gevaertsI suspect that part of the reason that no issue has been brought to the RSB yet is that nobody has any idea what would happen if it were
01:02:52LloreanMaybe an anonymous way to call it on issues might help. I know there are some things where I've though "I'd like to get a _big_ decision on this, but I don't really want to be the person calling for this decision"
01:02:58 Join leavittx [0] (~leavittx@
01:04:13gevaertsOne thing the RSB could do is to act as an official consensus notary
01:04:32gevaertsi.e. issues are discussed in the usual flamewars, and the RSB decides who wins
01:04:38LloreanThat would be an excellent role for them.
01:04:41 Part toffe82_
01:06:43LloreanA role of saying "this is how this issue is settled" and possibly summarizing *why* it's that way so it's clear when it can be reopened (for example, when one of those points changes / is no longer the case / whatever)
01:06:58LloreanJust to avoid us coming back to the same arguments again and again.
01:08:18*gevaerts makes notes on the devcon agenda
01:09:26 Nick fxb__ is now known as fxb (
01:09:53LloreanOne thing that's also come into my head is the notion of responsibility. A certain forum poster may have been wrong in that we can't assign responsibility to a lot of people
01:10:37LloreanBut we can ask developers to take responsibility for their own code (in such that they'll prioritize fixing bugs they've introduced in previous additions of features to adding even more features) and that seems like something that would need to be either a full community decision or an RSB decision.
01:14:02 Quit Zarggg (Quit: Leaving)
01:14:34LloreanAs a note, I haven't thought this idea through yet. I'm sure there are plenty of negatives against it, but it's just something that's been floating in my head for a couple hours so far.
01:15:00LloreanIt's more of an example of the sort of question that I really don't expect there'll ever be a good consensus on
01:18:49*Llorean 's Beast just crashed with an invalid instruction. =/
01:19:03gevaertsLlorean: I've not put that last thing on the wiki yes. I'll leave that to you :)
01:19:20LloreanI need to think it through a while anyway
01:19:32LloreanI know it's going to be contentious, and I want to try to end up saying it how I mean it first.
01:20:36gevaertsI think I see what you mean, but I suspect that people who agree don't need any ruling, and those who don't will ignore it anyway
01:20:55*gevaerts doesn't think we have a committer who would not prioritise a serious data corruption bu
01:21:37LloreanYeah. And I don't know that it's something that could be agreed upon to be bad enough to require 'enforcement' so unless it is it's pretty useless to even discuss it
01:23:24LloreanjhMikeS: With recent gigabeat S changes, should I be needing a new bootloader? I just got an unidentified instruction crash in SVN Rockbox (couldn't take down the number, and it rebooted shortly). Or could this just relate to me ROLOing into SVN from an older build (i'll try harder to get specific details if it happens again)
01:23:52LloreanOr it could crash again just now, but since the backlight is off I can't see the message.
01:24:13 Join Boldfilter_ [0] (
01:25:46LloreanOkay, just triggered two more in rapid succession, the addresses seem to be rather random
01:26:54 Quit Boldfilter (Ping timeout: 268 seconds)
01:26:56 Nick Boldfilter_ is now known as Boldfilter (
01:27:27 Join bieber [0] (
01:28:37 Join Hillshum [0] (
01:29:10 Part Hillshum
01:29:19 Join Hillshum [0] (
01:34:44 Quit DerPapst (Quit: Leaving.)
01:36:36bieberI've registered on the Wiki, now I need write access, if anyone's around here to grant it
01:36:51 Nick fxb is now known as fxb__ (
01:38:19saratogabieber: done
01:38:31bieberThank you
01:39:19bieberI need to add a page for my SOC project: are there any particular pages I should link to it from once it's done?
01:40:08gevaertsbieber: SummerOfCode2010 I guess
01:42:29 Part Boldfilter
01:43:27 Join anewuser [0] (anewuser@unaffiliated/anewuser)
01:50:23jhMikeSLlorean: No you don't. I'm chasing odd problems right now, to put it nicely.
01:55:34bieberAt risk of sounding like an idiot, how do I add a new page to the wiki?
01:56:51gevaertspretend that it already exists :)
01:57:44bieberOh, okay
01:57:52gevaertsi.e. if you want to create a page called NewHypotheticalPage, you go to
01:59:24 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
02:01:51 Join Rob2223 [0] (
02:02:38 Join Zarggg [0] (
02:05:38 Quit Rob2222 (Ping timeout: 264 seconds)
02:08:45 Quit hebz0rl (Quit: Ex-Chat)
02:09:57joeygare there any players supported with oled screens?
02:10:11joeygi am working on the S9 and it has on OLED, which makes the backlight stuff superfluous
02:10:17joeygsince it has no backlight :)
02:10:28joeygand im not sure about the logistics
02:10:42gevaertsthe sansa clip
02:10:54scorche|sh"backlight stuff"?
02:11:09scorche|share you just after something you can read without a backlight in the sun, etc?
02:11:13joeygturning on the backlight, controlling brightness
02:11:20joeygi have a cowon s9, which has an oled screen
02:11:24joeygwhich produces its own light
02:11:39joeygright now there is no support for it and im am workng on getting the lcd test screen to compile
02:13:36jhMikeSLlorean: seems I might actually have it in order soon. I discovered some things about the DVFS controller that are important.
02:14:06 Quit bieber (Quit: Leaving)
02:15:05saratogajoeyg: the LCD screen chemistry isn't relevant to writing a driver for it
02:15:16saratogaLCD and OLED screens are essentially the same to software
02:15:29joeygyes, but the backlight is controlled separately no?
02:15:53saratogawhat do you mean separately?
02:16:13joeygi mean normally you have a GPIO pin PWM'd to control the brightness of hte screen
02:16:17joeygor am i way off base?
02:16:17 Join bieber [0] (
02:16:37 Join n1s [0] (~n1s@rockbox/developer/n1s)
02:16:38saratogaan OLED screen still has a pin to control brightness
02:17:41 Join CaptainKewl [0] (
02:18:02joeygok, thanks
02:27:46 Quit wincent_balin (Ping timeout: 248 seconds)
02:33:39 Quit MethoS- (Remote host closed the connection)
02:44:17JdGordonLlorean: the problem is even if everyone agrees, how are you going to enforce it (getting owned bugs done quickly).
02:44:45JdGordonthe only thing that could be done is take away commit access which means the bug might never be fixed
02:45:37JdGordonranma: thanks, I was having trouble finding the pins because of the tape covering them :)
02:45:52JdGordonthey are too small for me to work with so im hoping soeone can help me out
02:46:11 Quit krazykit (Quit: Connection reset by beer)
02:48:16 Join krazykit [0] (
02:51:37 Join wincent [0] (
02:52:23***Saving seen data "./dancer.seen"
02:55:08 Join ishottuify [0] (
02:55:50saratogaJdGordon: they're 1mm pitch right?
02:56:09saratogaa small soldering iron, or failing that a larger one with a short copper wire wrapped off the tip should work
02:56:57 Quit ishottuify (Client Quit)
02:59:30 Quit wincent (Ping timeout: 260 seconds)
02:59:37 Join Rob2222 [0] (
03:01:06LloreanJdGordon: Or maybe revert commits of new features. As I said though, enforcement would be the problem. And objectivity, of course. Someone would need to decide whether enough work was being done related to the bugs, etc.
03:01:12LloreanIt's just not really doable
03:01:33LloreanjhMikeS: So basically, for normal use I should stick with my old build for the time being?
03:03:18 Quit Rob2223 (Ping timeout: 276 seconds)
03:05:33jhMikeSLlorean: yeah, for the next few hours or so. I'm testing with a movie right now.
03:07:08LloreanjhMikeS: I could very reliably trigger crashes just by skipping forward, waiting one second (as soon as I heard audio) then skipping forward again, continually. I'd get a crash within about 20 seconds
03:07:51LloreanAlways an invalid instruction, but at a range of addresses starting from 00XXXXXX and ending at FFXXXXXX so I'm guessing they're not terribly useful on their own.
03:18:37JdGordondoes anyone know if the database can do a reverse lookup? so get tags from a filename?
03:22:18*JdGordon tries to remember to ask pamaury when he gets in
03:26:23jhMikeSLlorean: I got alot of garbage addresses. I thought it was all ok last night after playing for awhile but just minutes after committing, I got the same stuff. :\ I was in no mood to bother until the next day. But, it actually led to finding another issue, so it's a blessing in diguise, really.
03:26:37 Quit mikroflops (Ping timeout: 264 seconds)
03:29:14LloreanjhMikeS: I'll just watch for an update that suggests it's safe to come back, then. :)
03:38:55 Nick bgs100 is now known as bgs000 (znc@unaffiliated/bgs100)
03:40:32 Join StealthyXIIGer [0] (~stealthy@
03:40:45 Part StealthyXIIGer
03:43:34jhMikeSLlorean: 45 minutes into the flick and all's well it seems :)
04:07:29JdGordonpixelma: If i really didnt bother with waiting for testers then why isnt the fm patch in svn yet? hrm?!
04:08:14 Quit Barahir (Ping timeout: 260 seconds)
04:08:57ranmaJdGordon: BTW, you don't need nSRST (System reset), but you do need nTRST
04:09:19JdGordonmy jtag cable doesnt have that?
04:09:48ranmaThen you need to pull it high using a resistor (about 1K to 10K should do)
04:09:49 Join Barahir [0] (
04:10:19JdGordonbetween it and vcc?
04:10:35JdGordonvalue doesnt really mater?
04:11:01ranmaIt has an internal pulldown and as long as it's low the JTAG-Circuits are in reset state
04:11:45ranmaThe internal pulldown is more like 100K to 1M, I used 10K to pull it up to vcc.
04:12:22ranmaTheoretically you can connect it directly to VCC, but using a resistor is a bit safer in case you get the wrong pin or a short :)
04:12:50JdGordonhehe ok
04:14:26ranmaHmm, make that 50K for the internal one. IIRC I measured 0.5V when pulling a 3V internal pullup down to 0V with a 10K resistor, which would mean the internal one was 50K
04:14:51 Quit pixelma (Disconnected by services)
04:14:51 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:14:57 Quit amiconn (Disconnected by services)
04:15:00 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:15:10 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:15:19 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
04:15:22 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:24:57 Join Rondom [0] (
04:25:35 Quit Rondom_ (Read error: Operation timed out)
04:31:42 Join CGL [0] (~CGL@
04:46:14 Join JdGordon1 [0] (~jonno@2002:7bf3:8c1f:1234:223:4eff:fe9e:a776)
04:52:26***Saving seen data "./dancer.seen"
05:02:08 Quit dys` (Ping timeout: 276 seconds)
05:02:38 Join dys` [0] (
05:05:38 Quit JdGordon1 (Ping timeout: 276 seconds)
05:19:24 Quit adnyxo (Ping timeout: 252 seconds)
05:23:55CIA-5New commit by 03jethead71 (r25837): i.MX31/Gigabeat S: This should fix stability problems. One problem was to start using the DVFS controller properly so that interrupts will be masked ...
05:26:45 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
05:31:00CIA-5New commit by 03jethead71 (r25838): Changed a macro to a raw in int constant. Didn't mean to do that. Put it back right.
05:32:06LloreanjhMikeS: I'm not able to immediately crash it. So that's good. :)
05:32:32jhMikeSgood deal. seems stable now after running it quite alot here.
05:36:04jhMikeSLlorean: you said "immediately". how about not so immediately?
05:36:39 Quit Horscht (Quit: Verlassend)
05:37:24LloreanWell, by immediately, I meant "the reliable method of skip, listen, skip, listen for a minute or two" didn't work at all. Usually I could get it in less than half a minute before
05:37:41LloreanI'm updating to 25838 then I'll listen normally and let you know if anything bad occurs
05:40:09jhMikeSthe modules seem to be working normally now and it's asking to drop voltage more even at lowf. maybe I'll check that out.
06:08:14*jhMikeS was thinking about using state-retention on the beast, even though rb has no sleep mode, to just rolo the fw image instead of shutting down, which would speed things up and make dual/single boot times irrelevant.
06:10:19n1ssounds nice
06:10:46n1si think noone is against a real sleepmode either (as long as it's optional)
06:12:39jhMikeSI was thinking it could tide it over until there is one. At least it would "boot" faster for both.
06:13:41jhMikeSthe needed cpu support code would be basically the same anyway
06:17:36CIA-5New commit by 03jethead71 (r25839): Gigabeat S: There's no compelling reason to start automatic scaling in the bootloader only stop it milliseconds later. Do init to default frequency ...
06:20:40 Quit saratoga (Quit: Page closed)
06:20:50 Join Boldfilter [0] (
06:42:11 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
06:45:07CIA-5New commit by 03jethead71 (r25840): Gigabeat S bootloader: Fix archaic call placement in dptc_stop which hangs the bootloader before jumping to FW. Let the backlight LED ramp up after ...
06:45:08 Quit n1s (Ping timeout: 246 seconds)
06:48:03 Quit RadicalR (Ping timeout: 265 seconds)
06:49:29 Join RadicalR [0] (
06:50:54 Quit CGL (Quit: Soy spammero ---- \m/ d(>.<)b \m/)
06:52:28***Saving seen data "./dancer.seen"
06:58:09 Quit CaptainKewl (Remote host closed the connection)
07:07:22 Nick dys` is now known as dys (
07:09:48 Join funman [0] (~fun@rockbox/developer/funman)
07:11:44 Quit joeyg (Quit: KVIrc 4.0.0 Insomnia
07:19:19 Quit flatrose (Remote host closed the connection)
07:20:23funmanIf I assume the fuzev2 battery capacity is 550mAh like the v1, backlight with default brightness uses 35mA, not 30mA like the v1
07:21:22 Join esperegu [0] (~quassel@
07:21:56funmanIf I assume it's 465mAh then the difference is closer to 30mA
07:31:54funmanhum there seems to be really big differences: when I added the fuzev1 numbers my runtime was a bit less than 9 hours and in SansaRuntime mc2739 found ~12 hours 3 months before
07:32:37funmanmy interrupted (by a crash) attempt yesterday showed something like more than 14 hours (assuming percentage goes down linearly which i'm not sure)
07:36:09 Quit Rask (Quit: Wood and water, wind and tree, jungle-favor go with thee!)
07:37:32 Join einhirn [0] (
07:44:56topikfunman: that clip+ battery curve you commited is based on playing from internal memory?
07:45:29funmanyes, i also assumed some runtime in the 18 hours range which I can't reproduce now because of frequent crashes
07:45:57topiki keep trying to get it to complete from uSD but without luck so far
07:46:22topiktoo bad it doesn't panic with a message of sorts but just sort of dies
07:46:25funmando you have at least 3 or 4 hours runtime? I could compare the beginning of the curve
07:46:35topiki have 5h32m
07:47:40funmanjust paste it
07:48:50funmanyou resumed without recharging?
07:49:21topiki think that drop is because of the 'refreshing database' nonsense
07:49:30topikalso, i'm not entirely sure that one is from usd
07:49:43topikwill check now
07:51:33topiki think that is internal, sorry
07:51:36funmanI see we didn't implement sd_enable() yet, not sure if it will help
07:51:58topiki have no problems starting my clip+ with the card inserted though
07:52:14 Part Boldfilter
07:52:15 Join fyrestorm [0] (
07:52:39topikwhen the thing is charged again i will restart a battery_bench. should i wait for sd_enabled() to be included?
07:53:01funmanno, i let that up to FlynDice :p
07:53:58topiki'll just compile current svn then while it's charging
07:54:11topikany idea what those 'silent crashes' are yet?
07:54:28topikit happens on internal memory too (like in that bench i just pastebinned)
07:55:14 Quit fyre^OS (Ping timeout: 265 seconds)
07:55:55funmanno, it seems it also happens when CPU is boosted?
07:56:21topiki haven't tried that yet
07:56:32topikboosting must drain power though?
07:57:43funmanyes, reduces runtime by ~ 4 hours
08:00:07topikhmm, stability over runtime
08:02:00*S_a_i_n_t votes for stability everytime
08:02:25S_a_i_n_tI very rarely listen to my DAPs for ~6/7 hours straight
08:03:22topikit would be nice to have both. i'm assuming the crashes happen in internal/uSD access but that's just guessing
08:04:41S_a_i_n_toh,'d be nice to have both.
08:05:04S_a_i_n_tbut, if I had to choose one, I'd go for stability
08:07:59topikright now my fuze v1 is having primary playback duties and clip+ is only for testing, but when it gets stable the clip+ has some advantages
08:08:38funmantopik: you have a recent build on fuzev1? kugel mentioned frequent crashes
08:08:46 Join saratoga [0] (~9803c20d@gateway/web/freenode/x-gfyfjktijkiknezy)
08:10:15topikonly had 1 crash if quite a long time
08:10:25funmanlooks recent
08:10:32topikapril 22nd
08:11:30topiki read his remarks about his crashing fuze but i don't have that experience at all. i use it 2-3 hours each day so it's well tested
08:11:38topiknever for more than an hour at a time though
08:16:26saratogathanks, say that email but didn't click it
08:25:26*S_a_i_n_t wonders if Flyspray is down, or if it's just his machine being weird...
08:25:40 Join ender` [0] (
08:27:24funmanfine here
08:28:07saratogafunman: about those logs, I wonder if trying to get our codec lib to compile as an external library for ffmpeg might be easier
08:29:28 Join S_a_i_n_t_ [0] (S_a_i_n_t@
08:29:35 Quit S_a_i_n_t (Ping timeout: 240 seconds)
08:30:58funmandon't you mean the reverse?
08:31:10funmanmake the codec lib link against libavcodec and use its public API
08:35:59 Join Zagor_muh [0] (~bjst@rockbox/developer/Zagor)
08:36:05 Nick Zagor_muh is now known as Zagor (~bjst@rockbox/developer/Zagor)
08:36:07funmanwell good idea, i hope it doesn't involve too much work making the API work without a libc
08:36:32saratogacouldn't we just add wrapper functions and compile it like normal ffmpeg?
08:36:40 Part Zagor
08:37:08funmanif codec API and ffmpeg API are similar it should be straightforward
08:37:16 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:37:21saratogaffmpeg codecs and parsers are basically just tables of function pointers that init and decode right?
08:37:47saratogawe're not much different
08:38:17saratogathough i've never looked at ffmpeg's file parsers
08:38:28saratogathings like seeking might be trickier
08:38:42funmani can have a look (so i learn about rockbox codecs, AND do something else than f***ing sansas)
08:39:05saratogaour codecs are just called from the files in the root /apps/codecs
08:39:20saratogain there is a loop that calls the parser function, and then the codec decode function
08:39:32saratogathe parser and the decoder are compiled together
08:40:34saratogawma, cook, atrac and flac are all adapted from ffmpeg so they're pretty similar
08:41:21funmanatrac3 in oma looks the simplest, and most of it is seeking code
08:43:11funmani had 7h10 runtime with clipv1, a little bit lower than last bench (which was 7h30)
08:43:38 Join flydutch [0] (
08:43:39saratogamaybe I made things worse?
08:44:08funmani would need to redo the bench with the same album and just this revision reverted to check
08:44:44funmannow i'm running with the change in memory controller on clipv1 and clip+ as i have recent data to compare with
08:45:21 Join Long [0] (
08:46:25saratogai wonder how hard it would be to get our codecs to work with ffmpeg's file parsers
08:47:09funmanwe would need to remove unused parsers from each codec when linking if we use a single avcodec.a
08:47:20saratogathe codec interface is generally really simple, just init, decode and close
08:47:41saratogaand more or less the same for most rockbox codecs
08:47:56saratogawe actually used the ffmpeg interface unmodified for a little while in wma
08:48:41saratogai guess the hard part would be getting the codeclib functions (mdct, etc) working with ffmepg
08:49:51saratogahmm though maybe we don't need a lot of it
08:50:02saratogawe could probably use their bitstream functions unmodified
08:50:11saratogaso really it'd just be the mdct
08:52:31***Saving seen data "./dancer.seen"
08:53:27topikstubborn as i am, here goes another clip+ uSD battery_bench attempt
08:55:33 Join stripwax [0] (
08:55:48 Quit stripwax (Client Quit)
08:58:57funmanthe hard part would be to compile ffmpeg without libc
08:59:32saratogafunman: why do we need that?
08:59:42saratogayou want to run ffmpeg in rockbox or?
08:59:54funmanwell only libavcodec
09:00:15saratogai thought we just wanted to let other people use rockbox's codecs outside of rockbox
09:00:30funmanthat too :)
09:01:04saratogaIMO trying to compile their stuff for rockbox doesn't help us
09:01:11funmanwhat i'm looking at is: a rockbox.codec which uses libavcodec.a so adding new codecs is dead easy
09:01:56 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:01:59saratogabut ffmpeg codecs are floating point so they still have to be convered anyway
09:02:27saratogaplus we have almost all the audio ones already :)
09:03:16funmani wasn't trying to decode video :P
09:03:37saratogaof the codecs people actually use, I think we're only missing WMA Pro, Voice, and Lossless
09:03:45saratogaand thats being generous with Voice
09:03:54saratogaand theres no ffmpeg decoder for lossless
09:04:22saratogaafter that theres just obscure codecs like TTA, ATRAC1, VQF
09:04:50topiknone of the current rockbox targets have floating point?
09:05:07saratogai think the gigabeat S does, but we don't use it
09:05:17funmanhm our binutils is too old for the assembler syntax they use
09:05:40saratogaone more reason to update to eabi
09:07:05funmanah right i have the eabi compiler too but too late i deleted the ffmpeg folder
09:07:05 Quit Long (Quit: CGI:IRC (EOF))
09:07:40funmanBagder proposed creating a mailing list for rockbox/ffmpeg collaboration but IMO the question is: do you, saratoga, need a new mailing list ? ;)
09:08:01saratogai don't really mind however its done
09:08:35saratogai'm so busy this summer, and I want to help mt, so i'll probably be pretty useless
09:08:45funmanat least i think the ffmpeg devs will be more responsive than monty with tremor patch
09:09:01saratogai bugged them on IRC the other day, they said they're just slow
09:09:45funmani guess they're busy with ogg evangelism especially wrt html5
09:10:07funmanbtw did you see they were working on a new ogg format?
09:10:11saratogayeah i think embedded is not their priority
09:10:14saratogano i didn't
09:10:17saratogawhats it for?
09:10:43funmanmonty mentioned it in his reply to mans, not sure if it's useful outside video
09:11:29saratogai assumed that if google really open sources vp8, they'll just work on that
09:11:50 Join B4gder [0] (~daniel@rockbox/developer/bagder)
09:11:57saratogamaybe i should look at getting FS #10243 - Codec executor merged into rockbox
09:12:11saratogaif it was folded into the configure script it would be a pretty nice tool
09:12:54funmanwhat's left with this patch: only integration in buildsystem?
09:14:51funmantransOgg seems only vaporware at the moment, google only turns up to monty blog entry
09:15:51 Join petur [0] (~petur@rockbox/developer/petur)
09:18:05saratogafunman: i'm not even sure it works :)
09:19:27 Join DerPapst [0] (
09:23:44 Join aberet [0] (
09:24:34 Quit tmzt_ (Ping timeout: 264 seconds)
09:24:51 Quit esperegu (Remote host closed the connection)
09:24:56 Join tmzt [0] (
09:25:27 Join esperegu [0] (~quassel@
09:31:29markundoes keylock (soft hold button) only work in the WPS?
09:32:11markunI can't get my clip+ to lock in the file tree
09:33:59markunfunman: is there a reason this is done or shall I enable it for the other screens as well?
09:37:31 Quit saratoga (Quit: Page closed)
09:38:35 Join solexx_ [0] (
09:38:37funmanno idea
09:38:51funmanpixelma mentioned she liked soft hold this way
09:39:38pixelmait's working like this for ages on the Archos targets
09:39:56 Quit solexx (Read error: Operation timed out)
09:40:11Zagormarkun: isn't the reason simply a lack of buttons?
09:40:13JdGordonmarkun: iirc it should be simple enough to add for the file tree
09:40:14funmando the archos fit in a pants pocket?
09:40:27pixelmathe Ondio does
09:40:28Zagorfunman: the ondios do
09:40:41funmanZagor: finding a combination of 2 buttons which work in every screen shouldn't be hard
09:40:53markunmaybe my problem is that I don't know how to quickly jump to the WPS
09:40:55pixelmaon the Ondio?
09:41:02markun(without navigating the menu)
09:41:04Zagorfunman: brave words! :-)
09:41:44 Join efyx [0] (
09:41:50funmanmarkun: i think there's no jump to WPS shortcut on clip
09:41:52JdGordonselect+home isnt it?
09:42:01funmanyeah it's softhold in the wps
09:42:28pixelmamarkun: it's been noted that this doesn't work on the Clip+ I think (thought I read it i the forums)
09:42:46pixelmathe "resume" shortcut I mean
09:43:08markunmaybe "menu + play (up)"?
09:44:43 Quit esperegu (Ping timeout: 264 seconds)
09:45:19markunand what's HOTKEY? (I've been away for too long..)
09:45:51 Join esperegu [0] (~quassel@
09:46:09markunjumps to a user defined item?
09:47:21topikthe buttons on the clip+ take quite a bit of effort to push. is (soft-)hold even that useful?
09:47:40*S_a_i_n_t_ points markun to the manual re: Hotkey ;)
09:47:47S_a_i_n_t_but, you're on the right track.
09:47:52markunI don't always put it on hold
09:48:14markunbut I've had it for 2 days, so "always" is a bit strong ;)
09:48:22pixelmaI'd suggest "long Right" on both Clips then for resume like on the c200s if it's not taken (it might take away the second way to access the context menu but on the Clips and c200s you still have "long Select" for it). Clip and c200 have the same "problem" with up and down in the list being on buttoms that are also labelled something else
09:48:23markunS_a_i_n_t_: thanks :)
09:49:05S_a_i_n_t_markun: Hotkey lets you assign a context menu selection (one for the WPS, one for the Filetree) to a two-key combo.
09:49:16S_a_i_n_t_So you don't have to drill down into the context menu
09:49:37markunpixelma: sounds good
09:49:46S_a_i_n_t_by default, the WPS hotkey is set to playlistviewer, and the filetree hotkey remains undefined.
09:50:24markunS_a_i_n_t_: in that case "menu + down (playlist)" sounds like a good candidate for this hotkey
09:50:37S_a_i_n_t_I find it useful to set the filetree hotkey to either delete of open-with, I haven't decoded which one I use more yet.
09:50:48S_a_i_n_t_probably I find open-with the most usefull.
09:51:34S_a_i_n_t_markun: On the iPods at least it is select+play, I have no idea about other targets
09:51:39S_a_i_n_t_but I assume it is similar
09:52:51S_a_i_n_t_The same key combo is used to set/call the hotkeys
09:53:22S_a_i_n_t_the only thing that is lacking is that not all context items may be set to hotkey, and the player has no way of showing you what can/can't be set
09:53:42S_a_i_n_t_so, it's just trial & error to begin with until you figure out what's what.
09:53:53S_a_i_n_t_The manual details what can/can't be set though
09:54:31 Join lpereira [0] (
09:56:06ThomasAHfunman: Maybe you saw it in the channel log: Clip+ with r25831 crashes often when voice clips are enabled and browsing through directories or database. Otherwise everything seems to work fine
09:56:23funmanboosting CPU doesn't improve things?
09:56:32ThomasAHfunman: boosted and unboosted
09:56:56ThomasAHfunman: (if debug menu shows 240000... it is boosted, right?)
09:57:35markunok, just found out that home+select jumps to the WPS, so just pressing it 2 times enables hold
09:57:57funman240000 is 240kHz, quite low
09:58:30funmanif you don't check the number of 0s it can be either 24MHz (unboosted) or 240MHz (boosted). Just check that the counter is above 1
09:58:49ThomasAHfunman: I don't know how many zeroes there were ... it was yesterday. Oh, if both starts with 24 I would have to look closer :)
09:58:49S_a_i_n_t_This is going to sound like a "king n00b" question I know...but, I have to satisfy my curiousity. why is there more values that boosted/unboosted for the boost settings?
09:59:01S_a_i_n_t_I mean, what is 1, 2, 3, 4, 5 etc?
09:59:13S_a_i_n_t_howcome it's not just 0, 1?
09:59:36funmanS_a_i_n_t_: the counter shows how much threads require boosting
09:59:49topikclip+ died after 1 hour running battery_bench from uSD
10:00:01funmanso if one of them declares it doesn't need it anymore, the counter goes down, and CPU is unboosted only if the counter is 0 (which means no other threads needs boosting)
10:00:38S_a_i_n_t_I thought if it was just ran everything at max clock?
10:01:24funmanyes, that's what boosting means. What makes you think it would be otherwise ?
10:01:28S_a_i_n_t_I haven't played with it too much...but the boost value seems like it can go far, far into the 100s
10:01:45funmanThink of it like a road bus
10:01:50S_a_i_n_t_I just don't get why it isn't just on/off
10:01:53funmanIf there are 100 persons in the bus it keeps moving
10:02:08funmanIf one person leaves the bus, the 99 persons still in it do not want the bus to stop
10:02:32markunS_a_i_n_t_: if one thread decides it no longer need to boost but another one still requires it, boost should still be on.
10:02:38funmanBut the bus goes at the same speed if there is 1 person or 100, it just stops when it's empty
10:02:50S_a_i_n_t_so, the higher I set it, the more threads run boosted without it unboosting?
10:03:31funmansetting it higher than 1 has no effect
10:03:57topikthen i'm with S_a_i_n_t_ that it is confusing that it can be set to 2 and higher
10:04:03S_a_i_n_t_well, that's what I thought...which is why I can't understand the point for values other than 0 and 1
10:04:11funmanuser confusion isn't taken in account in the debug menu ;)
10:04:24S_a_i_n_t_no, that is fairly obvious ;)
10:04:25topikit's built-in though
10:04:52S_a_i_n_t_hence the (Keep Out!) I guess, but keep out means "please fiddle with thses setting" to most users
10:05:38funmani guess setting it higher than 1 is useful when tracking down boosting problems (number of unboost() not being equal to the number of boost())
10:09:17 Join S_a_i_n_t [0] (S_a_i_n_t@
10:09:48 Quit S_a_i_n_t_ (Ping timeout: 276 seconds)
10:12:18 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:14:18JdGordonpamaury: hey, do you know if you can get the id3 info out of the db with only the filename?
10:14:24 Join kaniini [0] (
10:15:02pamauryJdGordon: yes you can iirc
10:15:14 Quit funman (Quit: free(random());)
10:16:07JdGordonits not tagcache_fill_tags() is it?
10:17:12JdGordonif it is really simple to use then it might be nice to replace the playlist viewer (and the viewer in the wps) with this ability instead of just relying on filenames
10:17:27JdGordonmaybe only if ramcache is enabled though?
10:17:34pamauryif tagcache is loaded to ram, I think tagcache_fill_tags does the job. If it's not, see my MTP code, the tagcache_copy_tag function, I think it works too
10:18:49pamauryIf it's not loaded, basically you do a a search for the tag, adding a matching clause for the filename, and then call tagcache_get_numeric/tagcache_retrieve to get the value. You can't get the whole id3 at once
10:19:53pamauryJdGordon: I don't understand what do you mean by replacing the playlist viewer by this ability but I know nothing the wps
10:20:07 Nick HellDragon_ is now known as jd (~jd@Wikipedia/HellDragon)
10:24:08 Quit Kitar|st (Ping timeout: 260 seconds)
10:25:44JdGordonthere was a request in the forums to make the playlist viewer show more than just the filename... it clicked that tagcache could maybe be useful here
10:25:59JdGordonand in the WPS it could mean you always have the unbuffered tracks info avilaable
10:26:02LloreanWasn't there already a patch for that?
10:26:24JdGordonwas there?
10:26:35LloreanSpeaking of the WPS, I've noticed that if you change playlists the WPS will display next track info from the previous playlist for a while (maybe until buffering finishes)
10:26:39JdGordonthe one to get the playtime you mean?
10:26:51LloreanI thought there was one that showed metadata in the playlist viewer.
10:26:57LloreanTrack name tag, specifically
10:28:05ThomasAHHmm, which button do I have to press on the clip+ to boost? I thought it was "right", but ...
10:28:46JdGordonLlorean: care to find it? I cant see it
10:29:12ThomasAHah, I can do it directly in the debug menu with "up"
10:29:40 Join Kitar|st [0] (
10:30:08LloreanJdGordon: I can't find it either. Maybe it was one of those unsupported build things.
10:30:17LloreanOr maybe I'm just crazy. This is also very possible, it's late
10:30:17JdGordonmeh :/
10:33:02JdGordondo I need to reboot for ramcache to be enabled?
10:33:21pamauryI thinks o
10:35:27JdGordonnuts... 2line patch didnt work
10:36:25 Quit rvvs89 (Ping timeout: 246 seconds)
10:37:17*JdGordon forgot to reinit db after adding folders
10:40:50JdGordonfind_entry_ram() keeps returning -1
10:41:42JdGordonit needs dircache enabled?
10:41:53JdGordondircache_get_entry_ptr() is returning NULL
10:46:48pamauryhum, I think tcram needs dircache
10:52:30 Quit S_a_i_n_t (Ping timeout: 264 seconds)
10:52:33***Saving seen data "./dancer.seen"
10:53:06 Join S_a_i_n_t [0] (S_a_i_n_t@
10:58:13 Quit kramer3d (Quit: Leaving)
11:03:44JdGordonyay! enabling dircache made it almost work
11:04:16JdGordononly the track name is showing... i want the track number also?
11:04:43pamauryshowing where ?
11:04:53JdGordonon the wps
11:05:12JdGordonjust displaying info from the id3 struct
11:05:20S_a_i_n_tthere is a tag for
11:05:29JdGordonI'm using the tag
11:05:36pamauryare you sure the tag works ?
11:05:42S_a_i_n_tand it isn't working? weird.
11:06:42JdGordonsomething is no worky :(
11:09:09JdGordonloveley... tagcache_fill_tags() works fine... and I can printf the values after it returns, but when I try to get the tag value it breaks
11:10:24S_a_i_n_tHmmm, actually *is* there a tag to display the track number in the WPS?
11:10:31S_a_i_n_tI can't seem to find it in the manual
11:11:25S_a_i_n_tOooops, %in
11:11:34S_a_i_n_tyeppers, there is indeedy, sorry ;P
11:13:51pixelmatrack numbers from ID3 can be different to playlist position
11:14:35JdGordonthey usually are...
11:15:58pixelmaonly when you have one album in the playlist
11:16:27pixelmait's already wrong when you insert another in advance
11:16:28JdGordonpamaury: any objections to 0-ing out the mp3entry struct passed to tagcache_fill_tags?
11:16:41JdGordonnot doing that was the reason the tag was failing
11:17:01JdGordon(it looked for id3->track_string before id3->tracknum)
11:17:09pamauryno reason, to be honest I though it would already zero out everything which is not valid/present
11:17:39*S_a_i_n_t wonders what it is exactly that JdGordon is working on now...
11:22:57CIA-5New commit by 03jdgordon (r25841): If dircache and database "load to ram" is enabled then get the id3 info from the database in the WPS playlist viewer for non buffered tracks
11:25:17S_a_i_n_taha.....nice ;)
11:25:58S_a_i_n_tJdGordon: I had an idea for a tag that while I'm not sure how it could be implemented, I know I could make use of.
11:26:10JdGordonadd it to the wishlist
11:26:19S_a_i_n_tbasically "if 'font' == X, then Y"
11:26:33S_a_i_n_tI have already added a few there. ;)
11:27:16S_a_i_n_tI figure that the font tag I'm thinking of could help themers to accomadate different sized viewports/exclude wps info when the font starts to get too big.
11:27:48S_a_i_n_tie. increase the viewport size conditially according to the font.
11:27:55JdGordonwell, themes dont usually work unless a specific size font is used, so I'm not sure of the point
11:28:16S_a_i_n_twell, the reverse of that would be the point I guess
11:28:21 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:28:25S_a_i_n_tit *should* remove that restriction
11:28:59S_a_i_n_tmeh...just an idea, but I think it could be pretty handy if it was done well.
11:31:27S_a_i_n_tsomething like that, or %?FS<5|6|7|8|9|etc.>
11:31:38S_a_i_n_twould provide the most diversity
11:31:40Torneif themers go to the trouble of doing it that way it could be useful..
11:32:30S_a_i_n_tI figure that way, a theme could contain both a normal version, and a "high visability/large font" version all in the same wps/sbs
11:32:54JdGordonI dont think that is a good idea
11:33:01S_a_i_n_tTorne: yes, my thoughts exactly...whether or not they *will* is another thing altogether
11:33:01JdGordonskin files are already painful to read
11:33:22S_a_i_n_tits something I've been wishing I could do for a while, but I needed to think of a way to do it.
11:33:45S_a_i_n_tconditionally, by font name, or font height would be tyhe go IMO
11:33:55S_a_i_n_tJdGordon: this is true...
11:34:05S_a_i_n_tIt was just a thought.
11:35:07S_a_i_n_tWell, I think its a good idea...I actually think its a great idea. but it would be hard to implement, and probably create a nightmare for sysntax
11:35:28S_a_i_n_tbut dynamically resizable viewports would be awesome IMO
11:35:43JdGordonand quite impossible :)
11:35:46S_a_i_n_tyou made me think more on that topic with conditional UI viewport
11:35:56JdGordonwell nothing is impossible... but not realistically doable
11:45:59***Started Dancer V4.16
11:45:59***Connected to on port 6667
11:45:59***Logfile for #rockbox started
11:46:00Mode"logbot :+i" by logbot
11:46:02CtcpVersion from frigg!~frigg@freenode/utility-bot/frigg
11:46:02***Server message 501: 'logbot :Unknown MODE flag'
11:46:02 Join logbot [0] (
11:46:02 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:46:02 Join S_a_i_n_t [0] (S_a_i_n_t@
11:46:02 Join Kitar|st [0] (
11:46:02 Join kaniini [0] (
11:46:02 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
11:46:02 Join lpereira [0] (
11:46:02 Join esperegu [0] (~quassel@
11:46:02 Join efyx [0] (
11:46:02 Join solexx_ [0] (
11:46:02 Join tmzt [0] (
11:46:02 Join aberet [0] (
11:46:02 Join DerPapst [0] (
11:46:02 Join petur [0] (~petur@rockbox/developer/petur)
11:46:02 Join B4gder [0] (~daniel@rockbox/developer/bagder)
11:46:02 Join flydutch [0] (
11:46:02 Join ender` [0] (
11:46:02 Join fyrestorm [0] (
11:46:02 Join einhirn [0] (
11:46:02 Join RadicalR [0] (
11:46:02 Join dys [0] (
11:46:02 Join Rondom [0] (
11:46:02 Join amiconn [0] (quassel@rockbox/developer/amiconn)
11:46:02 Join pixelma [0] (quassel@rockbox/staff/pixelma)
11:46:02 Join Barahir [0] (
11:46:02 Join Rob2222 [0] (
11:46:02 Join krazykit [0] (
11:46:02 Join bieber [0] (
11:46:02 Join Zarggg [0] (
11:46:02 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
11:46:02 Join Hillshum [0] (
11:46:02 Join leavittx [0] (~leavittx@
11:46:02 Join arun [0] (~arun@unaffiliated/sindian)
11:46:02 Join markun [0] (~markun@rockbox/developer/markun)
11:46:02 Join loveless [0] (~loveless@2a01:198:26e:0:216:cbff:feb9:8c8d)
11:46:02 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
11:46:02 Join freddy_ [0] (
11:46:02 Join arbingordon [0] (~w@unaffiliated/arbingordon)
11:46:02 Join shai [0] (
11:46:02 Join matthewbpt [0] (
11:46:02 Join liar [0] (
11:46:02 Join AlexP [0] (~ap@rockbox/staff/AlexP)
11:46:02 Join shaggy-h [0] (
11:46:02 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
11:46:02 Join FOAD [0] (~dok@
11:46:02 Join preglow [0] (
11:46:02 Join topik [0] (
11:46:02 Join @ChanServ [0] (ChanServ@services.)
11:46:02 Join ThomasAH [0] (
11:46:02 Join fxb__ [0] (
11:46:02 Join ranma [0] (
11:46:02 Join Beta2K [0] (
11:46:02 Join Guest20471 [0] (
11:46:02 Join Kohlrabi [0] (
11:46:02 Join Bagder [0] (~daniel@rockbox/developer/bagder)
11:46:02 Join soap [0] (~soap@rockbox/staff/soap)
11:46:02 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
11:46:02 Join slck [0] (Venci@Slackware.SlackPix.Com)
11:46:02 Join parafin [0] (
11:46:02 Join sevard [0] (sev@
11:46:02 Join Farthen [0] (
11:46:02 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
11:46:02 Join Torne [0] (torne@rockbox/developer/Torne)
11:46:02 Join avacore [0] (
11:46:02 Join Hadaka [0] (
11:46:02 Join bgs000 [0] (znc@unaffiliated/bgs100)
11:46:02 Join Curtman [0] (
11:46:02 Join ved [0] (
11:46:02 Join crwl [0] (
11:46:02 Join kenguest [0] (
11:46:02 Join Kamyk_ [0] (
11:46:02 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
11:46:02 Join Battousai [0] (~bryan@gentoo/developer/battousai)
11:46:02 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
11:46:02 Join Utchybann [0] (
11:46:02 Join elcan [0] (
11:46:02 Join nima [0] (
11:46:02 Join FlynDice [0] (
11:46:02 Join Necos [0] (
11:46:02 Join yosafbridge [0] (
11:46:02 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
11:46:02 Join MuscleNerd [0] (
11:46:02 Join xavieran [0] (
11:46:02 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
11:46:02 Join wombat23 [0] (
11:46:02 Join rhodan [0] (~quassel@2001:1608:12:2::38)
11:46:02 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
11:46:02 Join Zambezi [0] (Zulu@
11:46:02 Join linuxguy3 [0] (
11:46:02 Join bzed [0] (
11:46:02 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
11:46:02 Join daurnimator [0] (daurnimato@freenode/staff/daurnimator)
11:46:02 Join pjm0616 [0] (~user@
11:46:02 Join stavrob [0] (
11:46:02 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
11:46:02 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
11:46:02 Join kisak [0] (
11:46:02 Join logiclost [0] (
11:46:02 Join guymann [0] (~charlie@
11:46:02 Join jordan` [0] (
11:46:02 Join mapi [0] (
11:46:02 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
11:46:02 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
11:46:02 Join scorche [0] (~scorche@rockbox/administrator/scorche)
11:46:02 Join simabeis [0] (
11:46:02 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
11:46:02 Join antil33t [0] (
11:46:02 Join powell14ski_ [0] (
11:46:02 Join Topy44 [0] (
11:46:02 Join jae [0] (
11:46:02 Join MagusG [0] (
11:46:02 Join grndslm [0] (
11:46:02 Join sharp [0] (
11:46:02 Join blithe [0] (~blithe@
11:46:02 Join CIA-5 [0] (cia@
11:46:02 Join Galois [0] (
11:46:02 Join maraz [0] (
11:46:02 Join aevin [0] (eivindsy@unaffiliated/aevin)
11:46:02 Join n17ikh [0] (
11:46:02 Join jd [0] (~jd@Wikipedia/HellDragon)
11:46:02 Join YPSY [0] (
11:46:02 Join jfc [0] (
11:46:02 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
11:46:36 Join Zagor [0] (
11:46:36 Quit Zagor (Changing host)
11:46:36 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
11:45:47 Join S_a_i_n_t_ [0] (S_a_i_n_t@
11:45:52 Join LinusN [0] (
11:45:52 Quit LinusN (Changing host)
11:45:52 Join LinusN [0] (~linus@rockbox/developer/LinusN)
11:48:13 Quit S_a_i_n_t (Ping timeout: 246 seconds)
11:49:07 Nick S_a_i_n_t_ is now known as S_a_i_n_t (S_a_i_n_t@
11:52:05 Join rvvs89 [0] (
11:52:43 Quit esperegu (Read error: Operation timed out)
11:54:47 Join esperegu [0] (~quassel@
11:59:28*S_a_i_n_t is getting "svn: Unknown hostname ''" tryiny to svn
12:01:37*JdGordon has a possible fix/workaround for the playlist viewer bug
12:02:12pamauryis there a way to wait to a specified (small) amount of time in rockbox (on all targets) ?
12:02:27 Join watto [0] (~watto@
12:03:31pamauryhow long does it sleeps ?
12:03:52pamaurysleep(HZ) sleeps one second ?
12:04:06pamauryok thanks
12:04:45 Join S_a_i_n_t_ [0] (S_a_i_n_t@
12:04:48 Quit S_a_i_n_t (Ping timeout: 264 seconds)
12:05:24 Nick S_a_i_n_t_ is now known as S_a_i_n_t (S_a_i_n_t@
12:09:20JdGordonpamaury: what did we decide? are we trying the menu?
12:10:32pamauryIIRC We decided to have to settings: one for normal mode, one for hold mode. Each setting can have several values (like UMS, MTP, HID, ...) plus a special "Display Menu" option that displays a menu on plug
12:10:46pamauryagree ?
12:11:09JdGordonexcept isnt is UMS or MTP, HID being seperate?
12:11:44pamauryyes indeed
12:16:14JdGordonwhere does the mode choice happen?
12:16:52pamauryin the settings you mean ?
12:17:31JdGordonwhere it chooses to go mtp or ums
12:19:39pamauryI don't know :) In the System sub menu of the settings I guess
12:25:05JdGordonim not entirely sure the git checkout got the right stuff... any added files which would make it obvious if they are missing?
12:27:04pamaurythere should be something like 6 or 7 files in firmware/usbstack/mtp/ subdir
12:27:32JdGordonno such folder
12:28:48JdGordonchanged to mtp-new and all good
12:31:42 Join wodz [0] (
12:34:32jhMikeSfor beast lovers only:
12:35:06*JdGordon wishes his sexy beast wasnt in peices
12:35:12markunjhMikeS: is this '3D effect' any good?
12:36:11jhMikeSmarkun: not too bad. I'm getting sort of fond of it on 6 or 7...
12:36:46jhMikeSseems a bit like a hardware crossfeed
12:37:35wodznice - WM8750 has also 3D thing and I was thinking where to put it in rockbox settings
12:38:38wodzjhMikeS: at least in WM8750 3D thing is more like stereo width with setting over 100%
12:42:07jhMikeSwodz: it sounds like this has some mixing and widening, depending on frequency and such
12:43:12jhMikeSprobably some delay lines
12:44:18wodzanyway I still think that introduceing hardware DSP submenu would be cleaner than mixing sw DSP and hw DSP in one place
12:44:23 Part aberet ("Leaving.")
12:45:18jhMikeSwodz: that's why it's a patch in the tracker for now. it's been rather controversial on just what to with all the settings.
12:46:06wodzI know, there was some disscusion here on channel without any final conclusion
12:47:11JdGordonshouldnt software DSP be mostly disabled if the hw can do it?
12:47:51Tornethe hardware one is less powerful
12:48:30Tornewell, less flexible
12:49:17jhMikeShmmm...don't think I have the cutoff setting doing it right though :\
12:50:09wodzTorne: It was You who have a friend interested in ui redisign in this area?
12:51:02JdGordonpamaury: to enable mtp you said to toggle it in the debug screen and hold down the button on insert right?
12:51:43JdGordon... which makes sense now that I wrote that out... makes sure the msc doesnt attach
12:52:02Tornewodz: Yeah, he was interested in working on this.. I've linked him to the patch and he's gonna try it out
12:52:26wodzit would be nice if he join us on channel
12:54:01wodzregarding hw DSP - WM8750 has quite nice signal processing features when recording. This is used in OF and gives superb results compared to other DAPs I had at hand.
12:59:10pamauryJdGordon: yes
13:01:39JdGordonso what really needs to happen is all the reesuted drivers get enabled in the USB_INSERTED event?
13:02:04JdGordonshould be easy enough to put a nice UI on it to choose which drivers and HID keymaps to enable
13:03:40pamauryJdGordon: yes but you'll also need to tweak usb.c because it does some manual driver enable/disable on insertion. I think I already modified it but you'll need to do a real rewrite, or I'll dot it
13:04:00JdGordonyeah, thats where I'm looking
13:04:29 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:05:03JdGordonI was thinking it would be nice to get the screenshot code out of usb.c also
13:05:35JdGordoneven if it is 5 lines to move out of there
13:07:17wodzside note regarding USB: It would be nice to have somehow animated "USB connected" screen. I always think if my dap is hanging or operate correctly after insertion
13:07:23pamauryyes, but that's not so easy, where do you want to put it ?
13:07:59pamaurywodz: you want a repeated animation or do you expect something that reflect usb activity ?
13:08:47wodzrepeated animation is fine
13:09:11JdGordonWhat I'm tihnking of is adding a function to apps/usb_screen.c to handle the insert (which is called by the usb thread), put the screenshot at the start of that (it would have to still be enabled in debug or somewhere though), then display the menu
13:09:20JdGordonexcept that might not work too well
13:09:27JdGordonor at all, damn multithreads
13:10:34pamauryit will not work if you are not in msc mode
13:10:52JdGordonif it all happens before usb_attach() it will
13:11:32pamauryI mean, if you don't plug in msc mode, other threads are not aware of the usb connection
13:12:36JdGordonyeah, ok, need to send a event to the button queue to get the main thread to call the usb ui thing. and have the usb thread ait for that to return
13:12:41 Join WaveFox [0] (
13:12:44JdGordonqueue_send() i tihnk is the blocking call?
13:12:55jhMikeSJdGordon: yes
13:13:11JdGordonin the usb-ui we enable the desired drivers before returning
13:13:45pamauryyes, seems ok
13:18:50wodzanyone with coldfire based dap at hand wiling to run test_mem plugin and share his results?
13:19:21*JdGordon can on h300
13:20:50JdGordononce it gets some charge though
13:23:48wodzdebugging asm optimized writes is going to kill me :-/
13:24:56 Join lpereira1 [0] (
13:25:01 Quit lpereira (Quit: Leaving.)
13:26:53n1swodz: what is different with ata on the mpio compared to other cf targets?
13:28:06wodzbyte swap in word
13:42:54JdGordonpamaury: no mtp if the usb screen is shown?
13:43:08JdGordonwhich means no mtp+hid?
13:44:02pamaurymtp is independent of the usb screen and hid. You can have usb screen+hid+mtp. It's just that the usb screen is shown only when there is exclusive disk access needed, which only msc requires
13:44:28pamauryI don't know if it's the answer you want
13:45:21pamauryYou can always use the plugin for hid commands anyway
13:45:54JdGordonim not getting any mtp with my changes (which just moves all the driver enabling into the default event handler (which is getting called))
13:46:00JdGordonand the usb screen is being shown
13:46:03***Saving seen data "./dancer.seen"
13:46:39JdGordonunless that thread is blocking it
13:47:19pamaurythen it means the msc driver is enabled. Can you paste the interesting part in usb.c (the USB_INSERTED case) ?
13:49:48JdGordonyeah, neither comp (ubuntu or win7) is picking up my ipod
13:50:44JdGordon first bit is usb.c, second the the default event handler in the main thread
13:51:01 Join mikroflops [0] (
13:52:23*JdGordon makes sure mtp is actually working without my changes
13:53:07pamaurythe SYS_USB_ATTACHED part is not good: it will still enable msc by default (and mtp will not show up because all endpoints are used). Furthermore, you removed all the code about exclusive access ?
13:53:31JdGordonI disabled all the drivers
13:53:34JdGordonexcept mtp
13:54:10JdGordonqueue_send() blocks the usb thread untill that returns
13:54:32pamauryah sorry :) Thne it's because you removed the parts that launch usb. In non exclusive mode this is usb_attach();
13:54:51JdGordonhehe doh :p
13:54:57pamauryIn exclusive mode, it gets call by SYS_USB_CONNECTED_ACK and it's usb_slave_mode(true);
13:56:14JdGordongrr... crashed rythmbox
13:56:42pamauryrythmbox is not very good at handling mtp
13:57:19JdGordonwin7 cant find the driver for it either (this is without my changes
13:57:47 Join Schmogel [0] (
13:58:02 Quit bgs000 (Ping timeout: 248 seconds)
13:58:13pamaurystrange, it worked well one week ago I think and I didn't modify anything important iirc
13:58:32pamauryDo you linux ?
13:59:16 Join bzed_ [0] (
14:00:31 Quit bzed (Ping timeout: 246 seconds)
14:00:36JdGordonignore that, i mus have disabled mtp instead of enabling it
14:00:41JdGordonall good in win7 now...
14:01:03pamauryNow you must have you screen do the same
14:02:59JdGordoncan I add a var to the usb drivers saying if it wants the usb screen enabled? then if any enabled drivers have that use it, otherwise dont?
14:03:29 Join bzed [0] (
14:03:39pamaurywhy not. Perhaps gevaerts has an opinion on that ?
14:04:12 Quit simabeis (Ping timeout: 240 seconds)
14:04:29JdGordonunless I'm reading wrong, if msc is disabled then hid wont work because it needs the usb screen
14:04:40pamauryno it doesn't
14:04:47pamauryYou can use the remote control plugin
14:04:58 Join _Farthen [0] (
14:05:03 Quit bzed_ (Read error: Connection reset by peer)
14:05:16 Quit Farthen (Read error: Connection reset by peer)
14:06:05JdGordonok, so I'll keep the current behaviour
14:07:37 Quit _Farthen (Client Quit)
14:07:40 Join MethoS- [0] (~clemens@
14:07:49 Join hebz0rl [0] (
14:07:58 Quit arbingordon (Quit: `)
14:08:06 Join Farthen [0] (
14:10:20 Part matthewbpt
14:12:55JdGordonbloody hell
14:13:33 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
14:13:38 Join Tuplis [0] (
14:17:10 Quit dfkt (Ping timeout: 245 seconds)
14:17:15 Join dfkt [0] (dfkt@unaffiliated/dfkt)
14:18:00 Quit dfkt_ (Ping timeout: 246 seconds)
14:18:42 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
14:18:49JdGordonyay /me is finally getting somewhere
14:18:56JdGordontoo many damn #ifdefs causing problems
14:20:18JdGordonpamaury: if mtp can do file access also it should add send an event or something if the boot file was changed
14:20:55pamauryyou're right, I haven't checked but it could be necessary, I'll have a look at it
14:21:36 Quit dfkt (Ping timeout: 240 seconds)
14:21:47 Nick dfkt_ is now known as dfkt (~dfkt@unaffiliated/dfkt)
14:22:12 Join simabeis [0] (
14:24:37 Quit n1s (Quit: Lmnar)
14:27:32 Quit hebz0rl (Quit: Ex-Chat)
14:28:34 Quit bzed (Read error: Connection reset by peer)
14:28:43 Join bzed [0] (
14:32:18 Join hebz0rl [0] (
14:37:35 Quit Farthen (Ping timeout: 245 seconds)
14:37:56 Join Farthen [0] (
14:44:48 Join arbingordon [0] (~w@unaffiliated/arbingordon)
14:48:47 Join bgs000 [0] (
14:51:47 Join JohannesSM64 [0] (
14:57:43 Part WaveFox
14:58:56 Quit B4gder (Quit: It is time to say moo)
15:02:15 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:02:47 Quit efyx (Ping timeout: 276 seconds)
15:03:34 Join efyx [0] (
15:04:44 Quit shai (Quit: Leaving)
15:04:58 Join apoelstra [0] (
15:06:13wodzJdGordon: what about test_mem?
15:08:23JdGordonmy h300 is apparently dead
15:08:32JdGordonits been charging since you asked and it still wont turn on
15:09:05*JdGordon 's crude usb connection menu isnt working :(
15:11:00 Join wincent [0] (
15:11:53 Join shai [0] (
15:15:21 Quit tchan (Read error: Connection reset by peer)
15:15:32pamauryJdGordon: you're right, my mtp code should check if the boot file was modified when exiting
15:16:12 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
15:18:37 Join fleebailey33 [0] (~fleebaile@unaffiliated/fleebailey33)
15:23:22 Quit wincent (Changing host)
15:23:22 Join wincent [0] (~wincent@rockbox/developer/wincent)
15:24:01markunpamaury: do you see a normal file tree with mtp where you can update rockbox?
15:24:29JdGordonpamaury: does anything there look obviously broken?
15:24:34JdGordoni mean, more than it not working
15:24:46 Part LinusN
15:24:51JdGordontrying to connect with msc +hid
15:24:58pamaurymarkun: yes, you see the complete file system normally
15:26:07pamauryJdGordon: can I see the code in usc.c ? You have to keep the ack mechanism for exclusive disk access
15:26:29JdGordonI am
15:26:53pamauryI don't really know what to say about the pastebin, what do you want me to search for ?
15:27:08JdGordonI dont know... I'm doing something obviously wrong
15:27:55JdGordon is the menu code.. the important bit is at the end it enables only the drivers it wants
15:28:10JdGordon case SYS_USB_ATTACHED:
15:28:11JdGordon usb_connection_gui();
15:28:11JdGordon return 1; is the middle step
15:28:46JdGordonthe splashes show the drivres i choose are enabled, but then nothing connects and i never get the usb screen
15:29:43pamauryshouldn't you have a #ifdef USB_ENABLE_MTP also ? Just a thought... I'm looking at the code
15:30:26pamauryThe SYS_USB_ATTACHED is a creation of your own ?
15:30:37JdGordonyes to both
15:30:59markunpamaury: what's the advantage of mtp over msc?
15:31:10pamauryCan I see the code which sends the SYS_USB_ATTACHED ? I have the impression that you never really call usb_attach or usb_slave
15:32:04pamaurymarkun: you have access to metadata, you can update the database at the same time you upload files, theorically you can also play some files remotely I think. The focus is on metadata however
15:32:34JdGordonqueue_send() does that magic, blocking the usb thread untill it returns, that is jhMikeS's stuff...
15:32:49pamauryYes but what do you do *after* the queue_send ?
15:33:09 Join DerPapst1 [0] (
15:33:29JdGordonbutton thread gets the event, calls the default handler which then calls usb_connection_gui() (see the 3 lines above)
15:33:59pamaurycan when do you call usb_attach exactly ?
15:34:03pamaury*and when
15:34:10 Quit loveless (Quit: loveless)
15:34:21JdGordonnext time the usb thread is worken up
15:35:13pamauryI don't understand. First, why do you have a return 1; ? If you are in the usb thread, returning will kill it no ?
15:35:28 Quit DerPapst (Ping timeout: 276 seconds)
15:35:54 Quit RadicalR (Ping timeout: 245 seconds)
15:35:55 Join RadicalR2 [0] (
15:36:01JdGordonno, the main thread calls that
15:36:20pamauryAh no sorry, it's in the default handler... Then what the usb.c code does, appart from queueing a SYS_USB_ATTACH ?
15:36:41 Quit apoelstra (Quit: WeeChat 0.3.0)
15:37:02JdGordonsends that off, then waits for it to return before doing the attach just like before, except it expects the required drivers setup already
15:37:04 Quit esperegu (Read error: Operation timed out)
15:37:23pamauryyou only do usb_attach ?
15:38:20JdGordonunless i removed too much code.. it does the whole lot
15:40:02JdGordon without the menu works
15:40:04pamaurySo you don't get to the usb screen. Hum. Change the DEBUG(...) to _logf(...) and have a look at the code to see if threads answer to the ack
15:40:22 Join DerPapst [0] (
15:41:42 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
15:42:07 Quit DerPapst1 (Ping timeout: 240 seconds)
15:44:12 Join esperegu [0] (~quassel@
15:44:23JdGordonlog is empty when only msc is enabled
15:45:36pamaurywith logf enable and after changing all the DEBUGF to _logf in usb.c ?
15:46:04***Saving seen data "./dancer.seen"
15:46:18JdGordonjust the one at the end of USB_INSERTED and SYS_USB_CONNECTED_ACK... but yeah ill do all of em
15:48:08JdGordonmeh, sleep time
15:48:36pamaurywith mtp, so say that it works ?
15:48:48JdGordonwith mtp it does work
15:48:52JdGordonwith msc it doesnt
15:49:06JdGordonusing that last pastebin
15:49:21JdGordonswapping true/false as wanted obviously
15:52:04pamauryand you have nothing logged ?
15:53:05pamauryAre you sure the mass storage driver get enabled ?
15:58:39 Quit MethoS- (Remote host closed the connection)
16:01:15 Quit JohannesSM64 (Ping timeout: 246 seconds)
16:03:22 Quit lpereira1 (Quit: Leaving.)
16:17:52 Join Jack|Ping [0] (~Jack@
16:20:05 Join JohannesSM64 [0] (
16:26:38 Nick fxb__ is now known as fxb (
16:33:21CIA-5New commit by 03wodz (r25842): HD200 - fix stupid bug in lcd_grey_data
16:34:18wodzha finally squashed it :-)
16:34:34 Quit wodz (Quit: Leaving)
16:39:02 Join moos [0] (moos@rockbox/staff/moos)
16:43:47 Quit hebz0rl (Ping timeout: 264 seconds)
16:46:02 Join toffe82 [0] (~chatzilla@
16:46:36 Join Boldfilter [0] (
16:47:15 Join AlexP_ [0] (~ap@rockbox/staff/AlexP)
16:48:53 Quit AlexP (Ping timeout: 260 seconds)
16:51:56 Join apoelstra [0] (
16:52:03 Nick apoelstra is now known as joeyg (
16:53:11 Join arlaneenalra [0] (~jules@
16:53:16 Part arlaneenalra
16:53:38 Join arlaneenalra [0] (~jules@
16:54:34arlaneenalraIs there a known issue with forum registrations? I could not get the captcha to display more than one letter or acknowledge the letters listed in the audio file . . .
16:55:55TorneThe captcha generation has been observed to be broken before :(
16:56:09Torneif you wait a minute and try again a few more times you may find it produces a typeable string
16:56:37arlaneenalraIs there somewhere to file a bug on this? I tried flyspray but it didn't list anything about the site itself
16:58:04Tornehm, not really. generally a forum post :)
16:58:19HillshumOr the mailing list somewhere?
16:58:39TorneZagor: we know the forum captchas are still broken, right? is someone looking at it?
16:58:55Tornescorche: as above
16:59:17ZagorTorne: I have no fingers in the forum pie
16:59:28TorneZagor: ah. you have admin rights, though :)
16:59:30Tornei just looked on the list
16:59:38Zagoroh, I do? :)
16:59:48Torneyour name's red
17:00:57HillshumI'm pretty sure you do.
17:01:14Tornewell, that's not necessarily enough
17:01:19Torneit was just a general guess as to who to ping
17:01:28Tornei expect it requires whoever has shell access to the box the forums ar eon
17:01:32Tornewhich i presume is scorche
17:02:01Zagoror at least someone with half a clue what is broken and how to fix it :)
17:02:05Torneyueah :)
17:12:45 Nick Jack|Ping is now known as Jackster1 (~Jack@
17:13:13 Join Blue_Dude [0] (~chatzilla@rockbox/developer/Blue-Dude)
17:14:14Jackster1my ipod owns with rockbox >.<
17:14:33Blue_DudeHaving issues with icons. I'm trying to add an icon and something seems off. There are 32 default icons and I'm trying to make it 33. How do I do it?
17:15:25 Quit DerPapst (Quit: Leaving.)
17:15:52Blue_DudeWait, forgot something...
17:18:23scorche|shTorne: odd...
17:18:33Tornescorche|sh: hm?
17:18:37Tornei thought it was a known thing
17:18:40Tornedid you think it had been fixed? :)
17:18:54scorche|shi think i might have heard it before, but pushed it out of my mind
17:19:05 Join lpereira [0] (
17:19:20scorche|shor it wasnt really pointed out specifically
17:19:30Tornewell, it has certainly been a problem for a while
17:19:40arlaneenalraA bug queue of some sort might be a good idea . . . .
17:19:42Tornein fact when i registered on the forums i got one or two bad captchas before a working one
17:19:50Tornerefreshing a couple of times fixed it for me back then :)
17:20:06Tornevarious other people have asked about it on irc since then
17:20:27scorche|shthe listening to the letters thing still works
17:20:37Tornehow frequent it is would be hard to determine, since people may just refresh until they get one that works, or give up and go away entirely :)
17:20:57 Join jgarvey [0] (
17:21:00Tornearlaneenalra: you tried listening as well, no? and that didn't work?
17:21:06arlaneenalrascorche: I wasn't able to get it to accept the letters I heard
17:21:12scorche|shon a refresh clearing the cache, i got an invalid captcha about 3/15 times
17:21:27scorche|shout of curiousity, which letters do you see?
17:21:34arlaneenalraZ or U
17:21:45Tornescorche|sh: well, anyway, i assume you are the one in a position to possibly do something about it?
17:21:48arlaneenalraI heard about 5 or six but it wouldn't accept them
17:22:07scorche|shTorne: if i am not, i dont know who is...
17:22:49scorche|shthat reminds me...i need to go transfer some money into the bank account as it is time to pay for the server again...
17:22:59scorche|shscorche: this ^^
17:23:06Tornescorche|sh: right. well, do we have any official way to track issues with forum/infrastructure/etc?
17:23:13 Quit lpereira (Client Quit)
17:23:42scorche|shTorne: forum - not really...for wiki and such things we created a page for tracking bugs
17:23:56scorche|shthough that was probably due to the switch to a new version
17:24:24scorche|shit would be nice to upgrade the forum software, but that will take some time and i need to figure out if we can still get all of our extensions
17:24:32scorche|shperhaps a thing for me to do at devcon
17:24:49Tornewell, this will probably do for now, i suspect
17:24:52Torneyou have been reminded! :)
17:25:13Torneit seems like a fairly useful thing to fix because not being able to sign up to a website sure puts *me* off from using it ;)
17:26:37scorche|shperhaps that is the reason for the decline in new forum memberships... ;)
17:26:41 Join bluebrother [0] (
17:26:41 Quit bluebrother (Changing host)
17:26:41 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
17:27:02scorche|shpersonally, i think that a secure captcha such as ours is something to be respected!
17:27:03 Quit kaniini (Ping timeout: 252 seconds)
17:30:19 Quit bluebroth3r (Ping timeout: 264 seconds)
17:34:17 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:41:47Blue_DudeHere's an idea I want to throw out there. I'm making a new icon for hotkeyable items. The new icon is essentially the old yellow diamond made red. If hotkeys aren't available, then the yellow diamond appears. This slightly changes current appearance since any custom icons on those items will be lost.
17:43:48 Join MethoS- [0] (~clemens@
17:46:05***Saving seen data "./dancer.seen"
17:46:52 Part Zagor
17:47:26 Nick fxb is now known as fxb__ (
17:48:05 Join loveless [0] (~loveless@2a01:198:3a5:0:216:cbff:feb9:8c8d)
17:50:42 Join einhirn [0] (
17:53:24Blue_DudeOr, I could only replace the appropriate "Icon_NOICON" with "Icon_Hotkey" and leave the custom icons alone...
17:54:32 Part arlaneenalra
17:55:22Blue_DudeI kind of like that idea, since at least one icon, "View Playlist", appears in a menu that's not hotkeyable, and there's no good method short of... Well, that's doable, wait a minute.
18:03:58 Quit pamaury (Quit: Quitte)
18:08:58 Quit petur (Quit: *plop*)
18:12:30Blue_DudeOK, a flag works better, and preserves the custom icon in case we don't have hotkeys. I still don't have a good way of limiting context to... Well, wait a minute...
18:14:54 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:16:37 Quit Schmogel (Quit: Miranda IM! Smaller, Faster, Easier.
18:16:41 Join Schmogel [0] (
18:16:52Blue_DudeFixed. The hotkey icon now appears in appropriate menus on appropriate items, and doesn't appear anywhere else. Anyone have problems with committing this?
18:19:54Blue_DudeBTW, I created the "extra" icon for all default button bitmaps, but only the color and 16 color gray bitmaps differ from the current "yellow" diamond. The color version is red, and the gray version is darker gray.
18:19:57 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:20:26 Join einhirn [0] (
18:20:41Blue_DudeMaybe the 1 and 2 bit depth versions would look OK totally blacked in.
18:21:02 Join hebz0rl [0] (
18:21:26 Join fml [0] (
18:21:36fmlBlue_Dude: where is the patch?
18:22:07Blue_DudeComing up. I haven't created a FS yet.
18:24:12 Join w1ll14m [0] (
18:25:01fmlWouldn't read icons scare people? Red = don't touch. What about yellow icons with a red dot inside? I'm just speculating.
18:25:26Blue_DudeUm, how do I upload a patch that contains binaries?
18:25:57fmlBlue_Dude: I think you just attach them separately
18:26:00Blue_DudeThe recording button is red...
18:26:12fmlthat too
18:26:22*Blue_Dude is out of time. Back soon. Drat!
18:26:47 Part fml
18:27:51Blue_DudeFS #11247. Back in a while.
18:28:09Blue_DudeBitmaps up later.
18:30:32 Quit AlexP_ (Ping timeout: 248 seconds)
18:32:30 Join AlexP [0] (~ap@rockbox/staff/AlexP)
18:32:54 Join towolf [0] (
18:33:10 Join pamaury [0] (~c2c7a50a@rockbox/developer/pamaury)
18:33:54 Quit joeyg (Quit: WeeChat 0.3.0)
18:34:47towolfsalve, is there a technical reason that there no speex_enc.codec available to record audio? (i've trawled lists, logs, forums, google. nothing really pertinent yet)
18:35:24 Join raphi_ [0] (
18:37:33 Quit Jackster1 (Ping timeout: 245 seconds)
18:40:14 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:40:36 Join einhirn [0] (
18:41:01 Join Brewtality [0] (
18:42:38Brewtalityhey, who/where should i ask about figuring out why my iPod (mod'd, 240gb hd) keeps freezing with RockBox?
18:43:09 Join Jackster1 [0] (~Jack@
18:46:05domonokytowolf: i think the technical reason for that is: no-one has done the neccessary work... :-)
18:46:31towolfdomonoky, ok, i wondered about that.
18:46:43 Join wincent_balin [0] (
18:47:15domonokytowolf: if you are interested in that. maybe you could code it ? :-)
18:47:35 Quit wincent (Ping timeout: 240 seconds)
18:48:16domonokyBrewtality: which build do you use ? 30GB or 60/80Gb ?
18:49:02Brewtalitydomonoky: 60/80 i am pretty sure
18:49:24towolfdomonoky, i can have a look. but this type of work would likely be uphill all the way for me. but the feature would make sense. speex is really low bitrate.
18:50:31domonokyBrewtality: the try the 30Gb build. The 60/80GB build expexts 64MB ram and the 30GB build 32MB ram. If you only have 32MB ram in the ipod but a 64MB build, it will freeze often.
18:50:35towolfi asked because someone might have established technical infeasability before.
18:51:40domonokytowolf: i think it should not be impossible, (but i am no expert in this area).
18:51:52Brewtalitydomonoky: so there is no way to tell which you have except to try both?
18:52:24S_a_i_n_tnot unless you know exactly how much RAM you DAP has...
18:52:31domonokyBrewtality: the standard disc size normly tells that. But there has been cases of wrongly refurbished units..
18:53:16domonokyBrewtality: the 30GB build should always be safe to use.. so try it and see if the freezes go away.
18:54:32 Join DerPapst [0] (
18:54:50 Quit DerPapst (Client Quit)
18:55:01 Join DerPapst [0] (
18:55:39 Quit Jackster1 (Ping timeout: 276 seconds)
18:58:51 Join Jackster1 [0] (~Jack@
18:59:43 Quit BHSPitMonkey (Quit: Ex-Chat)
18:59:43 Quit DerPapst (Ping timeout: 264 seconds)
19:01:20Brewtalitydomonoky: so... sorry for the newbishness... how do i change the build?
19:02:14domonokyuse rbuitl ?
19:02:38domonokyor just download the correct zip and unzip it to you player
19:02:58 Quit toffe82 (Ping timeout: 245 seconds)
19:03:28Brewtalitydomonoky: okay
19:04:30 Join toffe82 [0] (~chatzilla@
19:08:57Brewtalitydomonoky: so, just extract it to the ipod and "replace all" for identical files?
19:09:32domonokyjup, or use the "Install Rockbox" button in rockbox utility (after selecting the correct target)
19:09:49*domonoky wonders how Brewtality got rockbox on his ipod the first time :-)
19:11:53Brewtalitygah, quit breakin' balls
19:12:37 Join funman [0] (~fun@rockbox/developer/funman)
19:13:40 Quit guymann (Ping timeout: 276 seconds)
19:14:15 Join guymann [0] (
19:15:43 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:17:12gevaertskugel: do you have plans to push your gsoc-fix-creat and c-lib-rework branches to svn soon?
19:18:00 Quit einhirn (Ping timeout: 260 seconds)
19:18:04kugelI planned to do it today yes, unless something speaks against it
19:18:47gevaertssounds good to me
19:19:03kugelhave you looked at the c-lib one?
19:19:12 Join CGL [0] (~CGL@
19:19:16 Join Luca_S [0] (
19:20:23 Quit towolf (Quit: Ex-Chat)
19:20:29gevaertsnot in very much detail
19:21:40gevaertsIt's a big diff, but it doesn't look very complex or dangerous
19:22:16 Join petur [0] (~petur@rockbox/developer/petur)
19:22:28kugelbtw, I wondered if #define instead of typedef has any advantages?
19:22:49kugelwe seem to #define (u)int*_t instead of typedefing
19:23:36kugel(in fact, I'm not sure if either has an advantage, I always thought they're basically equivalent)
19:24:58*gevaerts has no idea
19:25:23gevaertsI'd expect there not to be much difference for simple types, but I haven't thought it through
19:26:23 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:26:28kugelthe only problem I can imagine is if you do typeof/sizeof on them without parentheses and they expand to a two-word type
19:31:19kugelI would like to have the discussion about the target-tree movement soonish
19:32:32Luca_SFlynDice: I've been able to let my FuzeV2 recognize my uSD card by changing the SD_APP_CMD commands in the widebus selections to return a response (MCI_RESP + &response). no change in delays
19:32:56 Quit avacore (Ping timeout: 248 seconds)
19:33:24FlynDiceLuca_S: Glad to hear that!
19:33:35 Join joeyg [0] (
19:34:42FlynDice Luca_S: I grabbed my daughter's fuzev2 today and I have the same uSD that you do and the combo was working just fine....
19:34:49kugelI assume that we ask for a respond makes the controller handle cards differently, more safely apparently
19:35:06CIA-5New commit by 03kugel (r25843): Make creat() posix compliant API-wise. Shouldn't affect the core as it's wrapped via a static inline.
19:35:24CIA-5New commit by 03kugel (r25844): Make open() posix compliant api-wise. A few calls (those with O_CREAT) need the additional optional mode parameter so add it. Impact for the core is ...
19:35:26Luca_Si'll do one final test from clean svn right now, then i'll post a diff somewhere
19:36:39FlynDiceIt makes the send_cmd take longer to complete at least. This is similar to what funman needed to do to get writes to work I think...
19:39:16funmandon't we need to check the APP_CMD response ?
19:39:45 Quit Hillshum (Quit: Leaving)
19:40:13funmanWRITE/READ MULTIPLE BLOCKS answers with a generic R1 status response, but i'm pretty sure APP_CMD answers with some bit which says if the command has been successfully handled
19:41:00kugelis this done in as3525v1?
19:41:29funmansomewhy we removed the check
19:41:51Luca_Sok, this is the (obvious) diff:
19:41:54Luca_Sgotta go now
19:41:58pamaurykugel: about open, just a though I had is that on device, the FAT file system doesn't handle mode, and on simulator, the mode might need to be changed by the user. did you hardcode 0666 ?
19:42:01kugelthat might be the cause for some of the non-working cards that we fixed by delays on as3525v1 too
19:42:03Luca_Sbye and thank you all
19:42:06 Quit Luca_S (Quit: CGI:IRC)
19:42:46kugelpamaury: we don't handle it in the firmware
19:43:08pamaurymy question is on the sim
19:43:29kugelyes I did hardcode. I proposed some different number but it was generally accepted that 0666 is suitable (remember the number is umask'd by the processes umask for it's applied)
19:43:52 Quit Brewtality (Quit: CGI:IRC (EOF))
19:44:18pamauryok, that was just to check that you had a tought about it, I don't object the 0666 as one can always restrict the umask of the sim
19:45:46funmankugel: true we removed it for µSD cards in r23001
19:46:09***Saving seen data "./dancer.seen"
19:46:46FlynDicefunman: I'm looking now. I took out the APP_CMD response because it's not valid until the card is "ready" after the acmd 41 I think and it was failing. At least I think that's why looking now.
19:48:02pamaurykugel: just a last thought, would it be sensible to use a global define for the mode. it's true that 0666 is sensible but I don't like the idea of have this constant everywhere without a define
19:49:13 Join fml [0] (
19:49:31 Join Strife89 [0] (~Strife89@
19:50:10fmlkugel: shouldn't we define a constant instead of repeating 0666? Or even create a helper function in misc.c? The value is repeated N times.
19:50:37kugelI thought about that too, but I think when it comes to these permission the plain number is most obvious. I immediately know what 0666 means, but not what ROCKBOX_DEFAULT_PRERMISSION_MODE is
19:50:58kugelyou only need to change it where 0666 isn't appropriate, and that should be about nowhere
19:51:51fmlkugel: and exactly out of this rason I think we should have a helper function.
19:52:26kugelwhat should this helper function do?
19:53:26funmanwhy do we need to make the files world-writable ?
19:53:52funmanor even readable
19:55:06pamaurybecause there are mask with the umask of the process which will probably not allow the file to be world writable/readable and that you can change anyway
19:56:13funmanhm ok
19:57:15fmlkugel: it would hide unnecessary parameters
19:57:36kugelI don't understand
19:57:58kugelthe point of the commit was to add the unnecessary parameter to make it posix compliant
19:58:31 Quit MethoS- (Remote host closed the connection)
20:00:33FlynDicere: APP_CMD & ACMD 41 see r23718 The app cmd was giving a continuous crc fail(because the card was busy) in the acm41 loop so In order to get cmd crc checking I needed to disregard the fact that the command seems to fail.
20:02:50fmlkugel: and what did you gain by that? Now you have to write code that's basically not needed. If the mode is needed on the sim then we can hard code it in the sim impl. But in the "interface" function we don't need it. Compliance is not a value of itself IMO.
20:03:43kugelcompile rockbox to link OS libraries (and thus include the OS' headers which clash which our non-posix-compliant variants)
20:04:00kugelit's part of RaaA if you didn't notice
20:04:46kugelfml: what you proposed has been the how it is for the past years
20:05:54FlynDiceAck now rereading I see Luca_S is talking about a different App_CMD....
20:07:23 Quit moos (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
20:07:34fmlkugel: no, I didn't notice it was of RaaA. Now I understand the rationale but I still don't like the repeating of 0666. I get allergic when I see rpeating code.
20:08:05 Join dubcee [0] (
20:10:57kugelI know, I feel the same, but in this case it's a bit different
20:11:02 Quit flydutch (Quit: /* empty */)
20:11:39kugelthe fact that people are using the numbers at all, instead of the defines designed for it, shows that the numbers are just more 'catchy'
20:13:44 Quit bieber (Remote host closed the connection)
20:13:50kugelso it's a matter of improving readability, not so much a matter of repeating something hardcoded
20:13:58Jackster1Is there an irc app for rockbox?
20:14:18 Join Luca_S [0] (
20:14:23dubceerockbox irc! i am in need of some assistance...
20:15:55kugelJackster1: no, it would be pretty useless
20:16:15Jackster1You have a point
20:16:55kugelno current target has internet to begin with, but there's also absolutely zero code for internet access
20:17:08Jackster1i see*
20:18:40Blue_DudeOK, the changed icon bitmaps are posted at FS #11247. It should work completely now so give it a try.
20:19:11 Quit dubcee (Remote host closed the connection)
20:19:39kugelBlue_Dude: doesn't tango offer a suitable icon?
20:19:51Blue_DudeThere shouldn't be any issues with it. I don't see any downsides. So unless someone has an objection, I'll commit it later.
20:20:04Blue_Dudekugel: Which one? I needed something similar but distinctive.
20:20:07kugelI could imagine an icon showing an F key from a pc keyboard
20:20:36kugelwhy does it need to be similar?
20:21:48 Nick fxb__ is now known as fxb (
20:21:55Blue_DudeThey menu items are still menu functions, and should look similar. They just have an added capability, so they ought to be distinguishable without being totally dissimilar.
20:22:15Blue_DudeThe *
20:22:41kugelhm, ok I see your point, but I'm not sure I agree with
20:23:02kugelbut I don't really care anyway
20:23:02Blue_DudeYour icon design skills are needed. Have at it! :)
20:23:18kugelon a related note, we should possibly look into updating our tango icons
20:23:56kugelBlue_Dude: I hoped tango would provide something useful for this
20:24:28kugel(tango refers to in case you didn't know)
20:24:53Blue_DudeOh. I thought you meant something already in the BMP strip.
20:25:18 Join bieber [0] (
20:25:49 Join DataGhost [0] (
20:25:49 Quit DataGhost (Changing host)
20:25:49 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
20:26:43 Join pamaury_ [0] (~c2c7a50a@rockbox/developer/pamaury)
20:27:51 Quit pamaury (Ping timeout: 252 seconds)
20:30:40Blue_DudeI don't see anything there that just screams "hotkey" to me.
20:31:34kugelno, not really, but there are a few that I'd consider more suitable than what you propose
20:32:02kugelbut then again I don't consider similarity important ;)
20:34:40 Join DerPapst [0] (
20:35:13*Blue_Dude shrugs.
20:35:16 Part Boldfilter
20:35:33Blue_DudeI just went with what looked good to me. I'm open to suggestions.
20:41:43CIA-5New commit by 03Buschel (r25845): Use correct spelling for 'dB'.
20:42:30kugel" -24~dB. "?
20:44:01kugelgevaerts: any idea what to do about the mem*.S files in the target tree?
20:44:33funmankugel: FlynDice: did you look at the USB code in AMSv2 OF ?
20:44:38kugelI'm wondering if they should be under libc/ too
20:45:13kugelfunman: no
20:45:16gevaertskugel: they could be under libc, but that's not really needed I think
20:45:25funmanwith pamaury_ we are not 100% sure if the controller used is the one from as3525, the one from as353x, or another
20:45:50gevaertstarget/<cpu> won't be included for RaaA anyway, and I'd hope that the host libc has optimised versions too
20:45:51funmani suspect it's not the as3525 one though, so hopefully it's the as353x one like each other chip which changed
20:46:02Blue_DudeAh yes, 24 decibits.
20:46:38funmanwas the indentation change intended?
20:46:42kugelgevaerts: android defintely does ;)
20:46:42Blue_DudeNot sure what the ~ does though.
20:47:01FlynDicefunman: USB is way beyond me at this point...
20:47:06funmanyeah me too :)
20:47:19funmani just want to help pamaury_
20:47:54kugelgevaerts: ok, I'll post-pone that idea
20:50:16 Join Buschel [0] (
20:51:47Buschelkugel: the ~ is non-breakable space. through this "-24" and "dB" will not be separated by a line break
20:52:10kugelah, gotcha, I didn't know that
20:53:22 Quit wincent_balin (Ping timeout: 265 seconds)
20:53:44CIA-5New commit by 03FlynDice (r25846): ad-as3525v2: Ask for a reponse from SD_APP_CMD to help with timing issues when switching to 4 bit widebus. ...
20:58:22Luca_Swhohoo my name in a commit message! FlynDice, if I had to thank you guys for every brilliant idea you had while reverse engineering those little players, I'd have to continue until the end of the world :D
21:00:04ThomasAHfunman: btw, when boosted I did not get crashes with voice clips ... is there an easy way to make this the default?
21:00:43funmancomment out ADJUSTABLE_CPUFREQ in firmware/export/config/sansaclipplus.h
21:00:58ThomasAHfunman: thanks
21:03:44 Join panni_ [0] (
21:04:16FlynDiceLuca_S: If you want to be listed in credits give me your name!
21:07:52 Quit bmbl (Quit: Bye!)
21:08:11funman12:24 runtime on fuzev1 with FS #11246
21:08:22Luca_SFlynDice: let's wait, maybe my contribution broke something else :D
21:08:26pixelmawodz (for the logs): got some test_mem results from my M5 in case you still need them
21:08:45 Part watto
21:09:56Luca_SFS #11246 applies to all ams sansas?
21:10:16 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
21:10:21Blue_DudeAnyway, please check out FS #11247. I'll be back later. Thanks.
21:10:23 Quit Blue_Dude (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
21:10:27FlynDiceLuca_S: That's up to you but I'm happy to add your name to the credits file.
21:10:46FlynDiceI need a real name though....
21:10:47Luca_SFlynDice: Luca Leonardo Scorcia
21:11:01FlynDicedone in 2 mins
21:11:12 Quit joeyg (Quit: WeeChat 0.3.0)
21:12:57funmanLuca_S: yes
21:14:00Luca_Si'll try to do a fuzev2 battery bench with and without it then as soon as it finishes charging
21:14:06funmanThat would make 45mA for fuzev1, and when I added current estimation in rockbox it ran at 65mA
21:14:19funmanI'll test again with only this commit reverted
21:15:25CIA-5New commit by 03FlynDice (r25847): Add Luca_S to CREDITS. Thanks for your time & effort!
21:18:06 Join dfkt|x [0] (
21:18:34 Nick dfkt|x is now known as Guest85977 (
21:19:58 Quit Guest85977 (Changing host)
21:19:58 Join Guest85977 [0] (~dfkt@unaffiliated/dfkt)
21:20:08 Nick Guest85977 is now known as dfkt| (~dfkt@unaffiliated/dfkt)
21:20:26FlynDicefunman: I've got another v2 power saver I think but I'm waiting for the current battery bench wave to pass. There's some LP bits in MCI_CLKENA that I think mean low power mode. I can enable after the cards init and this is what the OF seems to do.
21:21:09funmanoh btw there is a bit 'low power / high quality' in AUDIOSET2
21:21:20funmanperhaps it's the one which caused low volumes in saratoga patch
21:21:33 Quit JohannesSM64 (Quit: WeeChat 0.3.3-dev)
21:21:33FlynDiceback later
21:23:45bieberFor the WPS library I'm working on, I'm wondering if it would be a good idea to integrate QT's libraries into it, since I'm building it for a QT app. Any thoughts?
21:24:37funmanhmm INIT_ATTR is boring when disassembling binary files :(
21:25:13CIA-5New commit by 03Buschel (r25848): Add available recording sample rates for different iPod types to manual.
21:25:44 Quit dfkt| (Remote host closed the connection)
21:27:28 Quit n17ikh (Read error: Connection reset by peer)
21:27:46 Join M3DLG [0] (~M3DLG@
21:28:14CIA-5New commit by 03mcuelenaere (r25849): Ingenic Jz4740: Use lcd_putsf() instead of lcd_puts() in exception handler
21:29:29kugelheh, lcd_putsf was a great addition ;)
21:31:51 Join robin0800 [0] (
21:32:23 Quit M3DLG (Ping timeout: 248 seconds)
21:33:17funmanoops my fuzev1 bench was unpatched
21:35:25 Join n17ikh [0] (
21:41:56 Quit pamaury_ (Quit: Page closed)
21:42:31 Quit MagusG (Ping timeout: 246 seconds)
21:44:18 Join mcuelenaere [0] (~mcuelenae@rockbox/developer/mcuelenaere)
21:46:11domonokybieber: if it makes live easier, you can surely use Qt Core things in your wps lib.
21:46:13***Saving seen data "./dancer.seen"
21:47:03bieberI'm just finding that QT duplicates a lot of the functionality I've been using in stdlib (strings, files, etc.) and it may end up being simpler if I use their versions from the start
21:47:17 Nick fxb is now known as fxb__ (
21:47:46domonokyalso qt has much more comfort functions then stdlib :-)
21:49:14bieberDefinitely a plus
21:51:54 Quit w1ll14m (Ping timeout: 264 seconds)
21:51:55 Quit loveless (Quit: loveless)
21:53:42 Quit funman (Quit: free(random());)
21:56:36domonokybieber: when does the summer vacation start for you ?
21:57:35bieberI'm free now
21:57:57domonokynice, so you successfully finsihed your exams :-)
21:58:15 Quit raphi_ (Quit: leaving...)
21:58:27domonokyplease remember to create a wiki page for your project...
21:58:42bieberI put one up and linked to it from SummerOfCode2010 yesterday
21:58:54domonokyoh, i missed it. sorry
22:00:01bieberHopefully it's satisfactory, I just copied the format the WMA Codec project page was using
22:00:47domonokylooks good for a start.
22:01:14 Quit Strife89 (Quit: Leaving work.)
22:02:53 Join apoelstra [0] (
22:03:02domonokyif you have any Qt questions, feel free to ask me or bluebrother (we both work on rbutil).
22:03:18 Nick apoelstra is now known as joeyg (
22:04:10bieberI need to download that and give it a try sometime. Last time I updated my Rockbox install it didn't exist yet
22:05:48domonokyso you are in serious need to update your rockbox (with rbutil ofcourse) :-)
22:05:49bluebrotherbieber: nice to see you around :)
22:06:11bluebrotherwhat OS are you working on? Windows or Linux?
22:06:14domonokyyour rockbox install must be several years old :-)
22:06:36bluebrothernice, that makes installation much easier :)
22:06:52bieberI think I first installed it at least a couple of years ago. The utility was probably around last time I updated, but I just missed it
22:07:21bluebrotherRockbox Utility will turn 3 this summer. Well, the Qt version of it that is.
22:07:32bieberOkay, then I definitely just missed it
22:07:46bieberThis thing looks pretty handy
22:07:52bluebrotherI think the wx version has been started at least a year before that. Though that was a pain to build on linux
22:07:55bieberHow difficult is it to build a QT app for Windows, btw?
22:08:22bluebrotherdepends how you want to build. If you have the build environment set up it's not more complicated than on linux
22:08:30domonokynot really difficult.. normally its just the Qt download.. but rbutil needs a few specials :-)
22:08:48bieberDo you have to distribute the QT runtimes, too?
22:09:06domonokywe link statically on linux and win.
22:09:11bluebrotherthe problem is that Rockbox Utility needs gcc plus a few other tools that are distributed with msys
22:09:23domonokymac gets the qt libs via embedded "frameworks" .
22:09:42*bluebrother needs to get around to update his MinGW toolchain to gcc 4
22:10:39bluebrotherthe "stable" MinGW, at least if you are using their installer, still uses gcc 3. But Qt 4.6 now officially supports gcc 4 on MinGW, and the old MinGW installer is outdated since ages now
22:10:41bieberSo you can cross-compile it with MinGW from Linux?
22:10:57bluebrothermight be possible, but I build on windows. It's easier.
22:11:25bluebrotherthe problem is that we would need to cross compile Qt, and Qt doesn't like that too much. But for creating static binaries we need to rebuild Qt.
22:11:50domonokywe should really try to find time to make buildservers for rbutil (win,lin,mac) :-)
22:12:20bluebrotherso a virtual machine is simpler. It's slower, too (building on Windows is slow, unfortunately :)
22:13:29bluebrotherdomonoky: I had buildbot running for it a while ago as a test setup. Was only one OS though, but that shouldn't be a problem. The problem is the time thingy ;-)
22:13:51bieberHow long does the Windows build take?
22:14:14domonokybluebrother: jup, as always.. maybe its a idea for what todo at devcon ?
22:14:18 Quit Jackster1 ()
22:15:16kugeldomonoky: add it if it isn't :)
22:15:20domonokyfor me a few minutes.. not fast, but bearable.
22:15:56bieberDoesn't sound bad, especially if you're just building it every time you want to make a release
22:17:28domonokyor for testing..
22:17:53domonokybut you dont recompile everything for that, so it gets faster.
22:18:40 Quit joeyg (Quit: WeeChat 0.3.0)
22:19:09bluebrotherdomonoky: hmm. I would need to decide going to devcon then
22:19:19 Quit Buschel (Ping timeout: 248 seconds)
22:19:28gevaertsbluebrother: well yes, but you should decide that anyway!
22:19:57domonokybluebrother: i just added it, so you should decide to come :-)
22:21:35bieberI've got to go for now, thanks for the suggestions everyone
22:23:09 Quit bieber (Remote host closed the connection)
22:23:28 Join bieber [0] (
22:23:32 Join Jaykay [0] (
22:27:07 Quit amiconn (Ping timeout: 245 seconds)
22:27:07 Quit pixelma (Ping timeout: 245 seconds)
22:28:18 Join pixelma [0] (quassel@rockbox/staff/pixelma)
22:28:46 Join amiconn [0] (quassel@rockbox/developer/amiconn)
22:29:18 Quit freddy_ (Ping timeout: 276 seconds)
22:29:59 Join wincent [0] (
22:30:46 Join MethoS- [0] (~clemens@
22:35:21 Join MagusG [0] (
22:36:19 Join wodz [0] (
22:37:03wodzdomonoky: I posted usbid of MPIO at FS
22:37:33domonokythanks. i will try to complete that when i find time.
22:37:51domonokydoes the MPIO have mtp mode with a different id ?
22:38:13pixelmawodz: got some test_mem results from my M5
22:38:54wodzdomonoky: no - it has hardware usb-ata bridge
22:39:30wodzpixelma: I am eager to see Your results
22:40:02pixelmabufsize: 514048, write ticks: 15 (267 kB/s), read ticks: 26 (154 kB/s) - no change when boosted or unboosted
22:41:28 Quit mcuelenaere (Quit: No Ping reply in 180 seconds.)
22:41:41 Join mcuelenaere [0] (~mcuelenae@rockbox/developer/mcuelenaere)
22:42:10wodzI have unboosted write 95kB/s, ubusted read 55 kB/s, boosted write 267kB/s and boosted read 148kB/s
22:43:24wodzthis may explain why it runs almost always boosted
22:44:17kugelgevaerts: so, I'm about to commit the c-lib one
22:44:37kugelI'm unsure whether I should "activate" it for the sim already (or if at all, that is)
22:44:49wodzdomonky: should I change error handling in mkmpioboot?
22:44:56gevaertsI'd do that in a separate step if possible
22:45:19kugelthe problem is, I don't really know to what extend the sim already uses the host libraries
22:45:39kugelbut yea, that can be a separate one, right
22:46:01kugelwhat a have a last peek at it?
22:50:40kugelgevaerts: i guess not (?) :)
22:50:54 Join JohannesSM64 [0] (
22:51:07gevaertskugel: sorry, feeling a bit too tired :)
22:51:21kugelI could wait for tomorrow
22:51:32kugelbut I'm fairly confident it's fine
22:51:48 Quit fml (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
22:52:04gevaertsfeel free to go ahead. This is the easy stuff in RaaA :)
22:52:35kugelalthough it wasn't as easy as expected
22:56:24 Join M3DLG [0] (
22:56:30 Quit Schmogel (Read error: Connection reset by peer)
22:57:21 Join Unhelpful_ [0] (~quassel@rockbox/developer/Unhelpful)
22:57:40 Quit Unhelpful (Read error: Connection reset by peer)
22:58:07 Quit bmbl (Ping timeout: 252 seconds)
22:58:26kugelis "RaaA" official enough to use it in commit messages?
22:58:41gevaertsgood question
22:58:48gevaertsLet's discuss this at devcon :)
22:58:54*Bagder prefers just RA
22:59:14scorche|shthe sun god?
23:00:06 Join Boldfilter [0] (
23:00:25*scorche|sh prefers RaaA....less acronym conflicts
23:04:01 Join Schmogel [0] (
23:04:45CIA-5New commit by 03kugel (r25850): Move c/h files implementing/defining standard library stuff into a new libc directory, also standard'ify some parts of the code base (almost entirely ...
23:07:50kugelphew, I feared svn would lose the history of the moved files due to me using git
23:08:27 Quit Unhelpful_ (Read error: Connection reset by peer)
23:08:28 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
23:08:51gevaertsThat's the least of your worries right now :)
23:10:21CIA-5New commit by 03kugel (r25851): Add firmware/include to the include path to fix checkwps.
23:10:57 Join Unhelpful_ [0] (~quassel@rockbox/developer/Unhelpful)
23:10:59 Quit Unhelpful (Read error: Connection reset by peer)
23:13:47kugelhm, I fixed some of the size_t related printf warnings for the fuze build, I'm surprised other targets show more of them
23:14:49kugeloh, it seems system dependant
23:15:16kugelmeh, /me was too lazy about checkwps
23:15:49 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
23:16:43 Quit Unhelpful_ (Read error: Connection reset by peer)
23:17:16CIA-5New commit by 03kugel (r25852): Fix some size_t related warnings (hopefully, they seem system dependant).
23:18:45 Quit mcuelenaere (Quit: Rebooting)
23:20:00 Quit Unhelpful (Read error: Connection reset by peer)
23:20:04 Join Unhelpful_ [0] (~quassel@rockbox/developer/Unhelpful)
23:20:32 Quit petur (Quit: Zzzz)
23:20:50*kugel thinks /usr/include/stdlib.h is broken...
23:21:32kugelif it needs int32_t then it should include stdint.h or inttypes.h, not sys/types.h
23:25:37 Quit domonoky (Read error: Connection reset by peer)
23:26:09*kugel isn't sure what to do about the strict aliasing thing
23:31:58 Quit Boldfilter (Quit: Boldfilter)
23:33:08 Join Tux2 [0] (~Tux2@
23:33:47 Quit esperegu (Read error: Connection reset by peer)
23:34:17kugelI also don't understand the hwcodec warning in skin_tokens.c
23:34:33kugelah no, of course
23:34:49kugelthat seems like stupid code, it was a good idea to add the printf attribute to snprintf
23:36:05 Quit wodz (Quit: Leaving)
23:37:05CIA-5New commit by 03kugel (r25853): Fix some more snprintf related warnings.
23:39:04 Join avacore [0] (
23:42:20 Quit jgarvey (Quit: Leaving)
23:43:13 Quit liar (Ping timeout: 252 seconds)
23:46:14***Saving seen data "./dancer.seen"
23:49:13CIA-5New commit by 03kugel (r25854): Fix checkwps by using host system's file IO (thanks to r25843/r25844) and by including a work around for systems that try to get intN_t via ...
23:49:42 Quit Jaykay (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
23:50:51 Part fleebailey33
23:51:06gevaertskugel: another way would be to investigate if we really need sys/
23:51:18 Quit Farthen (Remote host closed the connection)
23:51:19kugelI did a bit
23:51:24gevaertsI think this workaround is fine though, at least for now
23:51:27 Join Farthen [0] (
23:51:56 Nick Farthen is now known as Guest86922 (
23:51:57kugelsys/ is unix and shouldn't be in libc/. removing it would mean another 200 file commit to fix #includes ;)
23:52:11kugelwe do use off_t and mode_t though which are defined there
23:52:15 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:52:22 Quit DataGhost (Ping timeout: 240 seconds)
23:52:48 Quit Guest86922 (Client Quit)
23:53:01gevaertsand which some things don't include apparently :)
23:54:18gevaertsthe good news is that the delta table still looks OK
23:55:27 Quit Tux2 (Remote host closed the connection)
23:55:38 Join joeyg [0] (
23:55:56 Quit Schmogel (Read error: Connection reset by peer)
23:56:35 Quit efyx (Ping timeout: 276 seconds)
23:57:40kugelthat __off64_t database error seems to be another one of that sort
23:57:50kugeldefined in unistd.h but fctnl.h doesn't include it
23:57:54 Quit JohannesSM64 (Ping timeout: 264 seconds)

Previous day | Next day