**** BEGIN LOGGING AT Thu Dec 06 02:59:59 2012 Dec 06 03:09:26 hey DocScrutinizer05. I guess I dropped it directly on the USB plug while charging. Seems I ripped the D+ and D- pads off. Dec 06 03:09:38 It's beyond my ability to re-solder. Dec 06 03:10:25 I didn't notice until the end of the day. Dec 06 03:13:57 I have seen that thread. Just... after breaking it. Dec 06 03:46:45 ManoftheSea: well, you might give it a try and ask Nokia Care if they could fix it, just tell them you got it 3 years ago and there's NO MORE warranty - otherwise they will keep it and give you some crappy windows phone or sth Dec 06 03:51:22 recently users reorted they got their USB receptacle fixed by Nokia Care. There even been a report about main board swap been done by Nokia - this one on warranty though Dec 06 03:52:07 Oh really? neat. I'm surprised they'd do anything if it's not under warrenty though. Dec 06 03:52:26 allegedly Nokia got some new spare mainboards manufactured Dec 06 03:52:53 well, if they can fix it and charge you for it, why not? Dec 06 03:53:16 I think the user claimed it cost hin 40 GBP or sth Dec 06 03:53:23 him* Dec 06 03:54:05 which sounds like a good deal for >60min of repair work Dec 06 03:54:12 oh yeah, that seems great then. Dec 06 03:55:21 I'll look into it, thanks Doc. Dec 06 03:56:45 yw Dec 06 03:58:05 Also, did I see you're on the Board again? Dec 06 03:58:54 hmm? Dec 06 03:59:29 ManoftheSea: if Nokia says they can't fix, you still might find a hacker that does hardcore fix like in http://talk.maemo.org/showpost.php?p=870017&postcount=27 Dec 06 04:02:00 eBay's got used models ~ $150. Might be cheaper than the repair. Dec 06 04:04:16 dang Dec 06 04:04:21 ManoftheSea: if Nokia says they can't fix, you still might find a hacker that does hardcore fix like in http://talk.maemo.org/showpost.php?p=870017&postcount=27 Dec 06 04:04:23 alas there: http://talk.maemo.org/showpost.php?p=693493&postcount=12 the photos (external links) are gone Dec 06 04:04:26 shame on you, tmo Dec 06 04:10:58 ManoftheSea: http://talk.maemo.org/showthread.php?p=1062790#post1062790 Dec 06 04:11:22 doc, yeah, I've seen these threads. Dec 06 04:12:02 If Nokia can't do it for cheap enough, I'll have to get a USB cable and solder to the pins under teh battery. Dec 06 04:12:14 I was intending to turn my phone into a home server at end of life anyway. Dec 06 04:12:17 http://talk.maemo.org/showthread.php?p=1070122#post1070122 Dec 06 08:10:36 DocScrutinizer05: Dec 06 08:10:40 DocScrutinizer05: ping Dec 06 08:11:55 DocScrutinizer05: a link has been mentioned to stuff on your server on TMO. Dec 06 08:11:57 http://talk.maemo.org/showpost.php?p=1300085&postcount=25 Dec 06 08:48:03 hmm, so what? Dec 06 09:03:09 the stuff on this server is meant for community, so it's normal somebody refers to it. Just *some* few files are in 'hidden' locations so they won't show up in search engines Dec 06 09:04:09 Hi guys, How can i make my handset to automatically disable WLAN interface if it has no connection for 5 min (for powersaving in case i forget to disable it) Dec 06 09:04:10 please don't post complete URLs to those files, neither to hidden path components above them (like "docs") Dec 06 09:04:46 errr Dec 06 09:05:16 XATRIX: simply disable the "connect to WLAN if available" Dec 06 09:05:17 Found a wiki Autodisconnect Dec 06 09:05:34 don't use auto-disconnect! it's a real botch Dec 06 09:05:37 DocScrutinizer05, Nope...it's not a solution... I need it to shut down the interface Dec 06 09:05:49 gives WAY more trouble than benefit Dec 06 09:06:19 Don't think it does... Cause it supposed to disable the iface. Only in this case you will have more powersave Dec 06 09:06:30 XATRIX: I don't understand what you're trying to do then Dec 06 09:06:36 If it doesn't automatically connect, it doesn't mean it won't scan your area Dec 06 09:06:52 nonsense Dec 06 09:06:59 It's fully operational , but just don't connect to Dec 06 09:07:07 that's *exactly* what this means Dec 06 09:07:18 So, i need something to manually disable wlan Dec 06 09:07:32 if you say so *shrug* Dec 06 09:07:33 Oops...i mean to disable wlan if it doesn't use network Dec 06 09:07:57 Such as , if there's no active wlan connection -> unload kernel modules Dec 06 09:08:01 for 5 min Dec 06 09:08:10 * DocScrutinizer05 shakes head Dec 06 09:08:16 If it's idle Dec 06 09:08:31 Also it shout monitor Btooh Dec 06 09:08:40 now that's what you learnt during the 5 years you're using N900? Dec 06 09:09:17 Hm...i'm using N900 for 2days Dec 06 09:09:27 http://wiki.maemo.org/AutoDisconnect - that's what i found out Dec 06 09:09:40 so why do you think you need to teach me about powersaving? Dec 06 09:09:53 But it doesn't mean i'm not used to linux in general Dec 06 09:10:07 we have this saying in Poland, "don't teach your father how to make babies" ;) Dec 06 09:10:10 when I tell you WLAN *is* shut down when not in use... Dec 06 09:10:13 I'm not teaching you... I'm telling you how do i see the subj... Dec 06 09:10:45 tadzik, in Soviet Russia babies made fathers ;) Dec 06 09:10:47 you don't need to disable anything, ICD does that for you Dec 06 09:11:07 We say, that egges teaches the turkey :) Dec 06 09:11:21 ICD ? Dec 06 09:11:34 Internet Connection Daemon 2 Dec 06 09:11:38 Oh Dec 06 09:11:52 What do you say, if i check your point in powertop ?:) Dec 06 09:12:31 I say powetop doesn't give a means to check wlan Dec 06 09:12:34 workring kernel module - is a driver , drivers cause interrupts, system ticks, and processor attention... No driver - no power drain Dec 06 09:12:55 suuuuure Dec 06 09:13:16 and iwconfig, rfkill and whatnot else are for my ass Dec 06 09:13:16 Yea, usually i tune up my laptop in such a way... Dec 06 09:13:34 Also, rfkill could help. But i like a way to unload driver Dec 06 09:13:36 you may as well assume that every inactive process is actually doing busy-waiting Dec 06 09:13:55 Yes, it's correct Dec 06 09:14:07 ...noitsnot Dec 06 09:14:13 yeah, maybe sigstop each inactive process, just to be sure Dec 06 09:14:36 ;-P Dec 06 09:14:37 why do you think htop says "260 threads, 1 running?" Dec 06 09:14:48 because 259 are _not_ running Dec 06 09:15:11 They are, but the system sheduler isn't currently working with them Dec 06 09:15:13 ever peeked at the system's CPU scheduler? Dec 06 09:15:23 actually your thinking been correct.... for PR1.0 WLAN drivers Dec 06 09:15:24 It's not applied for the kernel and kernel modules aswell Dec 06 09:15:27 that's the point of not running Dec 06 09:15:33 and PR1.0 ICD2 Dec 06 09:15:48 tadzik, it's running ! But is in idle state... Dec 06 09:15:57 If there's a process, it's running! Dec 06 09:16:02 since then it's not needed anymore to unload WLAN modules "manually" Dec 06 09:16:07 okay, we have to agree that we don't agree Dec 06 09:16:21 I may have a different definition of "running" Dec 06 09:16:24 That's only my point Dec 06 09:16:36 DocScrutinizer05, Ok..in this case, maybe things changed Dec 06 09:17:42 As for example, if i do unload my Ralink wireless modules on my laptop, it will stay alive on 24m more Dec 06 09:17:50 So, it does sense Dec 06 09:18:08 XATRIX: actually look at internet settings - you'll find not only "connect to WLAN when available" but also a period time for WLAN scans. All invented to keep WLAN inactive and shut down most of the time Dec 06 09:18:10 If i use ipw2200 WLAN intface. It makes no differece Dec 06 09:19:09 DocScrutinizer05, Sorry, what do you mean ? If it keep scanning all the time, it wil drain power Dec 06 09:19:16 yes Dec 06 09:19:28 So, the point of mine was to manually enable WLAN when i actually need it Dec 06 09:20:04 and that period (5min, 10min, 30min) is the time WLAN is powered down, since it eats power like candy *especially* during scans Dec 06 09:20:32 so PR1.1 (iirc) invented that auto-shutdown of WLAN Dec 06 09:20:42 Yes, you right as a part... It means it works like in a passive mode during scans Dec 06 09:21:08 It measn that it drains power much more less, in a scanning mode, because it doesn't use TX Dec 06 09:21:33 it works like in PSM=0 at very least, during scans. Means it doesn't sleep ever Dec 06 09:21:53 Hm Dec 06 09:22:39 while when connected to an AP it has a partner to do PSM and go to sleep until next relevant beacon is scheduled by AP Dec 06 09:23:15 that's done because RX eats lots of power on WLAN Dec 06 09:23:43 and during scan RX duty cycle is 100% Dec 06 09:24:23 so since PR1.1 ICD2 does all that module-unloading resp something similar for you Dec 06 09:24:25 Maybe you have a point ... Dec 06 09:24:42 How can i check my current PR version ? Dec 06 09:25:37 and auto-disconnect is messing up your complete internet connectivity setup to a point where it eventually breaks completely and nobody so far found out how to fix the mess except with a full reflash Dec 06 09:25:43 Also what about 3G/2G network modes? Dec 06 09:26:19 current PR version is to be found in settings->"about product" Dec 06 09:26:56 21.2011.38... is PR1.3.1 Dec 06 09:27:14 20.2010.36 is PR1.3 Dec 06 09:27:25 version: 21.2011.38-1 Dec 06 09:28:25 Is it ok ? Dec 06 09:28:28 2.2009.51-1 is PR1.1 Dec 06 09:28:42 yes, that's most recent version Dec 06 09:28:51 PR1.3.1 Dec 06 09:29:00 So it has everything that i need Dec 06 09:29:03 I hope Dec 06 09:29:06 yes Dec 06 09:30:12 for 2G/3G switching I use the 2G/3G switcher applet Dec 06 09:30:31 I have it also. But want to automate it a bit Dec 06 09:30:50 switching manually to 3G when I use browser and find I need to download huge amounts of data Dec 06 09:31:47 XATRIX: automating it isn't really easy, since modem takes ~10s to switch from one mode to the other. during that time there's no connectivity at all Dec 06 09:32:24 Yes Dec 06 09:32:56 you could do *some* automation by switching to 3G as soon as you start the browser, for example. Some simple dbus msg sent in a browserd/microB wrapper should do Dec 06 09:34:12 or maybe a gconftool invocation, dunno Dec 06 09:37:05 Yea, will think about it Dec 06 09:37:09 Thanks anyway! Dec 06 09:42:17 * jonwil wishes he could think of something useful to do for the n900 Dec 06 09:42:27 * jonwil has run out of things worth reverse engineering Dec 06 09:42:32 there's also dual-band, but it's greedy, as in it prefers 3G to 2G. It would be nice if it could be non-greedy Dec 06 09:43:02 Although I will say this: If I had the ARM version of HexRays (or a similar ARM decompiler) I could make BIG progress in reverse engineering a number of interesting things :P Dec 06 09:44:22 tadzik: indeed I never use 3G setting, only 2G and dual Dec 06 09:44:41 3G is pointless Dec 06 09:45:26 it would be useful if dual would prefer 2G Dec 06 09:45:27 But unfortunatly I dont have a spare 3k to spend on ARM HexRays :( Dec 06 09:46:03 jonwil: maybe find some warez? ;-) Dec 06 09:46:51 trust me, if it was available in the places I get crap like that (e.g. where I got my x86 HexRays from) I would have found it by now :) Dec 06 09:48:12 I also would appreciate a Lauterbach ICE and tracebuffer Dec 06 09:49:14 but that stuff is even 5 times more expensive Dec 06 09:51:52 jonwil: if you don't find good new stuff to RE, maybe you want to learn how to exploit ARM's ETM/ETB then ;-) Dec 06 09:52:48 I think you can create a program that reads out the embedded trace buffer internally (not via JTAG) Dec 06 09:53:31 would be MAD useful, for all sorts of real debugging, and even for RE purposes, once it works Dec 06 09:56:29 unfortunately digging that deep into the inner workings of the ARM CPU is not something I would find interesting :P Dec 06 09:56:57 check ... :-/ Dec 06 09:57:23 how could you say http://search.arm.com/search?site=Site-Search&entqr=0&output=xml_no_dtd&sort=date%3AD%3AL%3Ad1&client=ARM_Site_Refresh&ud=1&oe=UTF-8&ie=UTF-8&getfields=Description&proxystylesheet=ARM_Site_Refresh&q=embedded+trace+macrocell is not interesting? Dec 06 10:03:41 bwaahaaahaaaa NOKIA again! what else? http://lxr.free-electrons.com/source/arch/arm/include/asm/hardware/coresight.h?a=microblaze Dec 06 10:06:44 http://www.arm.com/files/pdf/Better_Trace_for_Better_Software_-_CoreSight_STM_with_LTTng_-_19th_October_2010.pdf Dec 06 10:12:11 http://lttng.org/ Dec 06 10:12:24 ^^^ tracing in linux Dec 06 10:16:05 though, it seems it doesn't support ETM :-S Dec 06 10:17:12 however: >>New whitepaper to learn how CoreSight STM can be used by software developers with open source trace framework as Linux Trace Toolkit (LTTng)<< ( http://www.arm.com/products/system-ip/debug-trace/index.php ) Dec 06 10:27:02 hah, >>ARM, Nokia, Texas Instruments and Lauterbach discussed at IP-ESC 09 how scalable on-chip system visibility can be delivered cost effectively.<< Dec 06 10:29:00 on-chip system visibility_ Dec 06 10:29:01 ? Dec 06 10:34:13 jonwil: yeah, it is a pity we still don;t have access to ARM HexRays Dec 06 10:37:57 If I had ARM HexRays, I recon I could figure out the 2D parts of PowerVR drivers no problems... Dec 06 10:38:28 at least based on my analysis at this point Dec 06 10:49:49 >>Case Study – GStreamer - FFmpeg execution under Android - Target: Mistral EVM OMAP35xx  Based on Cortex-A8 processor<< http://www.arm.com/files/pdf/CoreSight-On-Chip-Visibility-Webinar-March-2010.pdf pp.32 Dec 06 10:58:42 DocScrutinizer05, ping Dec 06 10:58:47 freemangordon, ping Dec 06 10:59:01 DocScrutinizer05, Preliminary results are in! Dec 06 10:59:01 With the n900 set to 250-600MHz, playing a 27minute .mp4 movie file using mplayer and the screen off and locked.# Dec 06 10:59:01 The N900 uses 5% LESS power with smart reflex enabled. Dec 06 10:59:01 Interestingly, with the 'DSP' profile (250-805MHz, SR on) the n900 used 12% more power than 250-600. Dec 06 10:59:39 "Well from what i have researched on Maemo, it is more of a full on linux kernel than Android." is there some sort of a longer explanation of that somewhere? Dec 06 10:59:52 and 7% more power than 250-600 without SR. Dec 06 11:00:00 So that seems pretty conclusive to me. Dec 06 11:00:05 SR DOES save power. Dec 06 11:00:19 Aoyagi, yes. Dec 06 11:00:24 maemo==linux Dec 06 11:00:32 android != linux Dec 06 11:00:43 That's only rephrasing of that quote :P Dec 06 11:00:48 that... doesn't sound correct Dec 06 11:01:10 The qoute is nonsense. Dec 06 11:01:26 Maemo IS linux. Dec 06 11:01:50 Android uses a broken linux kernel and that is where the similarity ends. Dec 06 11:02:14 Maemo users are full on nerds. Android users like to think they are. Dec 06 11:02:15 yep Dec 06 11:02:23 hah Dec 06 11:02:39 DocScrutinizer05, good morning! Dec 06 11:02:43 did you read the above? Dec 06 11:02:49 sure Dec 06 11:02:55 are you a believer? Dec 06 11:03:22 no. But I think those results are conclusive Dec 06 11:03:37 P=CfV^2 Dec 06 11:04:19 From what I've read so far, " Android is a more mass marketed smart phone with bits of linux running it." I guess I need to learn what "only kernel" means :D Dec 06 11:04:41 Remarkably the only mention of the above formula WRT to processors on wikipedia is on the P4 speedstep page. Dec 06 11:05:05 Aoyagi, The kernel is the 'layer' between the hardware and the UI. Dec 06 11:05:39 ((((hardware)kernel)UI)you) Dec 06 11:06:49 DocScrutinizer05, We all knew SR would save power. The interesting proof is that the optimum freqs for battery life are 250-600MHz. Dec 06 11:06:50 So based on that, Maemo uses also Linux UI (along with a kernel that's not (somehow) broken) :S Dec 06 11:07:29 no surprise Dec 06 11:08:29 Linux is a universal operating system, the same linux that runs on your PC is the same linux that runs on your n900. Dec 06 11:08:46 android isn't Dec 06 11:08:52 Simply compiled for a different architecture. Dec 06 11:09:34 Android is a derivative operating system that is not so...flexable. Dec 06 11:09:54 It breaks some of the ideas that underpin linux, so we call it broken. Dec 06 11:10:16 Makes sense. Dec 06 11:10:25 Still tho, I thought flexibility was its sellpoint. Dec 06 11:10:43 Basically you can call it a completely different operating system. Dec 06 11:10:54 Along with tons of social "apps" bloatware and market place full of junk. Dec 06 11:10:59 Remember they are selling it to pseudo nerds. Dec 06 11:13:19 Oh and MeeGo? Dec 06 11:13:36 I haven't heard that much about it... Dec 06 11:13:42 The notion of flexibility is created by the hundreds of 'work-around' apps that people like to install to 'fix' different aspects of their superior OS. Dec 06 11:13:57 Aoyagi, Meego is dead. Forget about it. Dec 06 11:14:02 Yeah, that's what I figured :D Dec 06 11:14:08 About the workarounds. Dec 06 11:14:36 So I take it it was inferior to Maemo at some point. Dec 06 11:14:40 Where as under maemo you do have true flexibility. Dec 06 11:14:51 That is why it is so easy to break your n900. Dec 06 11:15:06 What was inferior to maemo? Dec 06 11:15:15 Meego Dec 06 11:15:23 the partnership was a hell obviously, and eventually resulted in a nasty divorce Dec 06 11:15:27 meego was not inferior. Just alternative. Dec 06 11:15:32 Ah. Dec 06 11:16:22 I have NFC what Nokia were smoking when they decided a maemo/moblin merger was a good idea. Dec 06 11:16:46 Oh yes, this "easy to break" part reminded me - is it difficult (/possible) to get a fresh clean system on N900? I imagine it shouldn't be that hard. Dec 06 11:16:50 meEgo Dec 06 11:17:20 Aoyagi: huh? Dec 06 11:17:24 ~flash Dec 06 11:17:25 somebody said maemo-flashing was http://wiki.maemo.org/Updating_the_tablet_firmware Dec 06 11:19:22 Aoyagi, if you have a linux computer flashing is piss easy. Dec 06 11:19:22 Sounds easy enough. Dec 06 11:19:25 1 command Dec 06 11:19:36 And no I don't have one. Dec 06 11:19:58 you can even perhaps use the opensource flasher, 0xFFFF to flash 1 n900 from another. Dec 06 11:31:57 * jonwil is bored :( Dec 06 11:35:41 Aoyagi: you don't have one what? Dec 06 11:36:13 DocScrutinizer05: Linux computer. Dec 06 11:36:26 hmm, do you have a PC? Dec 06 11:37:04 PC and a laptop. PC is ridden by countless worms and laptop is work machine... Dec 06 11:38:26 then I'd suggest to download and burn a linux live CD, and download the flasher tool for linux and the fiasco-image.bin to C: and access it from Linux booted from CD there Dec 06 11:39:01 I plan to make one of these machines dualboot at some point... or maybe get a new laptop, dunno. Dec 06 11:39:42 well, booting from CD is poor man's dualboot Dec 06 11:40:04 xD Dec 06 11:40:40 Yeah, I might as well get some USB stick and throw some distro on it. Dec 06 11:40:54 should work as well Dec 06 11:41:12 nice detail is you only need console to run flasher Dec 06 11:41:28 so no fancy X11 and KDE/gnome needed Dec 06 11:41:59 Yeah, I like fancy... :S Dec 06 11:42:09 get DamnSmallLinux Dec 06 11:42:39 or, if you like fancy, get Knoppix Dec 06 11:43:08 Anyway, I was asking about flashing because I'll prolly get a used N900, at least for now. And I don't want to have some previous user's horrors swimming it it. Dec 06 11:45:21 g'nite Dec 06 11:46:13 And yeah, I heard about Knoppix somewhere. Dec 06 12:28:01 ..knoppix, fancy? has it even been updated in the recent 7 years or something Dec 06 12:28:39 Does it need to? :-P Dec 06 12:28:57 It works, it always has been, so why bother? :-P Dec 06 12:29:05 oh, apparently it has, I thought it had been essentially abandoned already Dec 06 12:29:18 Ah. Dec 06 12:29:39 Haven't used it for about those 7 years, frankly. :-D Dec 06 12:29:48 dunno why I thought that, maybe because the website lacks any dates Dec 06 12:30:09 But back in the days, when I was using it a lot, it has always been working for me. Dec 06 12:30:48 yeah no problems there, though I never liked the default kde desktop Dec 06 12:31:31 someone ought to make some debian derivative that by default boots you into a root shell and no X or such unless you specifically say so Dec 06 12:32:19 for those times when all you want is to run lspci or something on a system :) Dec 06 12:45:38 yea Dec 06 12:46:21 Or edit your disk partition table and fix it. Dec 06 13:02:15 Is thomasvs always like that? Dec 06 13:48:06 vi_: pong Dec 06 13:48:56 freemangordon, Hello. Read the backscroll. Dec 06 13:49:10 With the n900 set to 250-600MHz, playing a 27minute .mp4 movie file using mplayer and the screen off and locked.# Dec 06 13:49:10 The N900 uses 5% LESS power with smart reflex enabled. Dec 06 13:49:10 Interestingly, with the 'DSP' profile (250-805MHz, SR on) the n900 used 12% more power than 250-600. Dec 06 13:49:21 and 7% more power than 250-600 without SR. Dec 06 13:49:22 So that seems pretty conclusive to me. Dec 06 13:49:22 SR DOES save power. Dec 06 13:49:57 freemangordon, Please by all means, feel vindicated. Dec 06 13:50:08 vi_: aah, yeah. Though OC to 805 using more power even with SR enabled is not unexpected Dec 06 13:50:16 freemangordon, OFC. Dec 06 13:50:26 vi_: also have in mind that SR is device dependent ;) Dec 06 13:50:31 However there is now a quantifiable value. Dec 06 13:50:34 true. Dec 06 13:50:49 vi_: is that -thumb? Dec 06 13:50:56 freemangordon, Yes. Dec 06 13:51:04 do you have non-thumb device? Dec 06 13:51:06 I cannot comment on stock maemo. Dec 06 13:51:11 ...no. Dec 06 13:51:19 I useually go strait to thumb. Dec 06 13:51:30 as IMO -thumb changes the picture a lot Dec 06 13:51:58 well, this was using old mplayer. Dec 06 13:52:03 I'd expect larger savings on stock system Dec 06 13:52:25 freemangordon, One day, I will get to check it. Dec 06 13:52:58 vi_: my thought is that rest of the system is faster when on thumb, which reduces SR effect compared to stock Dec 06 13:53:15 CPU goes to idle faster on -thumb than on stock Dec 06 13:53:23 perhaps. Dec 06 13:53:31 mught b wrong of course, this is just how i feel it Dec 06 13:53:39 I so not have any data to dis/agree. Dec 06 13:53:48 I do not have any data to dis/agree. Dec 06 13:53:50 * ShadowJK wonders what is the optimal frequency from a "race to idle" point of view Dec 06 13:54:03 ShadowJK: 600? Dec 06 13:54:09 ShadowJK, I reckon it is 600MHz all along! Dec 06 13:54:22 Intel has always claimed that the fastest possible frequency will always win for power use on their cpus Dec 06 13:54:31 it depends. Dec 06 13:54:42 and they seemed surprised it'd be different for arm Dec 06 13:54:46 But power use increase exponentially. Dec 06 13:54:53 if there is a requirement to service lots of interrupts, it may not be true Dec 06 13:55:55 there is an android governer that behaves like ondemand upto a freq (600MHz) then like conservative thereafter. Dec 06 13:56:03 That would be ideal. Dec 06 13:57:40 freemangordon, You said there was a maemo build script for mplayer? Dec 06 13:58:23 * ShadowJK worked on getting MPlayer built with neon on n900 in 2009? Dec 06 13:58:34 But it's most likely not working anymore :) Dec 06 13:58:50 ShadowJK, FMG did it! Dec 06 14:00:19 hm? Dec 06 14:01:19 ShadowJK, FMG built the latest mplayer with neon support. Dec 06 14:01:45 ah Dec 06 14:01:57 Except it has some issues with window focus and the keyboard. Dec 06 14:02:20 I was wondering if the maemo build script might reveal some magic. Dec 06 14:03:15 iirc the one in extras has some patches for focus and sound Dec 06 14:03:25 ShadowJK, That will be it then. Dec 06 14:03:57 What would be sweet is a statically compiled hardfp version. Dec 06 14:04:26 thumb would be interesting Dec 06 14:04:36 the one FMG built IS thumb. Dec 06 14:04:51 Thumb, gcc 4.7.2 (linaro), neon. Dec 06 14:04:53 there shouldn't be too much floating point use in mplayer normally Dec 06 14:05:04 unless audio filter chain runs amok Dec 06 14:05:07 ShadowJK, Really? Video decoding. Dec 06 14:05:28 yes, really. Dec 06 14:05:35 Ok. Dec 06 14:06:15 It'd be more common in audio, but afaik the major audio codecs in libav are integer Dec 06 14:06:25 ShadowJK: the one i built is upstream october version Dec 06 14:06:34 2.1 or something Dec 06 14:06:50 freemangordon, What packages did you have to import to build it? Dec 06 14:06:56 nothing Dec 06 14:07:05 Just the mplayer source? Dec 06 14:07:05 I mean a simple configure/make Dec 06 14:07:08 yep Dec 06 14:07:10 aaah. Dec 06 14:07:35 well, I ahve a lot of -dev packages instaled, after all this is the same SB I use for CSSU-thumb Dec 06 14:07:36 So adding the maemo magic focus patches *should* be trivial. Dec 06 14:07:42 Version numbers are pretty meaningless, or were meaningless, I remember they didnt put out a new release for 5 years and had a "SVN is always stable" policy instead Dec 06 14:07:55 vi_: might be Dec 06 14:08:02 ShadowJK, just like extras-devel! Dec 06 14:08:24 freemangordon, mmm. I will look into it. (as best as I can). Dec 06 14:08:33 vi_: the problem is there are no separate debian/patches IIRC Dec 06 14:08:51 * ShadowJK finds maemo-screenblankprevention.patch Dec 06 14:09:15 ShadowJK, where? Dec 06 14:09:20 on my harddrive Dec 06 14:09:25 oh! Dec 06 14:09:36 Did you build the original mplayer? Dec 06 14:10:04 care to send me this patch? Dec 06 14:11:11 http://enivax.net/jk/mplayer-n900/02/ Dec 06 14:11:28 iirc I only did some build system hacking? Dec 06 14:12:00 Revision 30080 :) Dec 06 14:12:07 i wonder if that's the one in extras Dec 06 14:12:50 30099 Dec 06 14:13:01 ^svn version in devel. Dec 06 14:13:17 at least 19 better. Dec 06 14:13:54 15 of those are fixing indentation Dec 06 14:14:16 (because MPlayer policy doesn't allow mixing "cosmetic" changes with code changes) Dec 06 14:16:30 I also found this http://enivax.net/jk/mplayer-focus-fix.patch.txt Dec 06 14:16:37 But I don't know what it does Dec 06 14:16:41 and looks kinda old Dec 06 14:16:43 BTW I had to do a one line patch to be able to build it for thumb. I found the patch on the mailinglidt, it was no upstreamed by that time Dec 06 14:17:29 ShadowJK: but seems related to what vi_ reports Dec 06 14:18:03 anyway, if you can extract the source used for extras mplayer Dec 06 14:18:10 I'm clueless about debian stuff Dec 06 14:18:32 and diff it with a checkout of rev 30099 from MPlayer's svn... Dec 06 14:18:32 ShadowJK: debian stuff is easy Dec 06 14:19:05 I mean there is nothing special in mplayer debian/ Dec 06 14:19:41 I mean the process of obtainining sourcecode from extras Dec 06 14:20:07 aah. apt-get source mplayer ;) Dec 06 14:20:30 would that make my n900 blow up? :) Dec 06 14:20:37 * ShadowJK doesn't have svn there anyway Dec 06 14:38:04 ShadowJK, apt-get has a cow power! Dec 06 14:50:52 Is there any company which doesn't hire technical support that cannot do technical support? Dec 06 14:51:48 Or is there a website which lists per company how to talk to people rather than drones? Dec 06 14:53:24 I wonder what management was thinking when they put those people there. I can imagine it goes something like "we will just put people there who can take a call, not who can answer a call, since other companies do it just as badly". Dec 06 14:55:36 Well, the former support of Electronic Arts was pretty good. Dec 06 14:56:13 I can't think of any other large company that would provide sensible support. Dec 06 14:58:47 Greetings, guys. I'm trying to rebind BackSpace to Control_R by editing rx-51. No success at all. Dec 06 15:00:28 Is it possible? Or may be control can't be rebind on n900 keyboard Dec 06 15:03:13 Also similar problem with Up, Down keys. I tried to make shift+Up acting like PgUp, but instead it gave me something like ;5C Dec 06 15:44:04 rooster_rus: one thing is what the keyboard sends, another is how osso-xterm parses it Dec 06 15:44:15 try it on the notes program Dec 06 17:03:04 I've this new straighttalk APN proxy... There is no "apn proxy" setting that I can find. Just ones specific for http and https... Dec 06 17:04:55 DocScrutinizer05: i need someone to discourage me from uninstalling HAM Dec 06 17:52:29 kerio, uninstall that shit. Dec 06 17:52:33 And report back. Dec 06 17:56:55 vi_: k Dec 06 17:57:15 vi_: nope.jpg Dec 06 17:57:31 vi_: removing hildon-application-manager brings down everything Dec 06 18:11:23 Pali: why the fuck does kernel-power-flasher depend on hildon-application-manager? Dec 06 18:11:43 kerio, becuase of that dialog which show warning information Dec 06 18:12:09 kerio, maemo-confirm-text (or similar name) Dec 06 18:12:11 that's about as sensible as having the battery applet icon depend on mediaplayer-l10n Dec 06 18:13:01