**** BEGIN LOGGING AT Thu Jan 03 02:59:59 2013 Jan 03 03:36:43 Are the latest linux-nexus7 kernel sources that the Ubuntu 13.04 images here -> https://launchpad.net/~ubuntu-nexus7/+archive/ppa/+packages Jan 03 03:36:53 or are there newer debian sources for that package hiding somewhere? Jan 03 03:41:53 found it -> https://launchpad.net/ubuntu/+source/linux-nexus7/3.1.10-8.19 Jan 03 04:03:42 Has there been much activity here in development for the Samsung Chromebook? Jan 03 04:04:32 What sort of Dec? I thought it was working well on the chrimebooks Jan 03 04:04:43 Dev* Jan 03 04:06:12 It's working decently well out of the box. Still needs some work as far as touchpad drivers go, and I don't think hardware video acceleration is working at all yet. Jan 03 04:06:13 is Ubuntu Phone going to install like an android rom? will there be individual roms per phone, or will there be one universal installer? lol Jan 03 04:06:47 We'll know in a few weeks when they release the first builds for the Galaxy Nexus Jan 03 04:07:16 i dont have any devices that will be suported, unfortunetly. Jan 03 04:07:23 Lack of hardware acceleration makes Unity a bit of a cramp. Jan 03 04:07:47 If theyd include the Nexus 7, i'd be good. lol Jan 03 04:08:07 its pretty much my only device that meets hardware requirements. Jan 03 04:08:46 Well I know they used the Galaxy Nexus as the demo phone. Jan 03 04:08:52 Luckily I had the foresight to buy one last year. Jan 03 04:09:52 i wanted to Jan 03 04:09:53 Assuming source is released reasonably quickly after the first build is released, I'm assuming that 3rd party development for Ubuntu Phone will take off pretty quick. Jan 03 04:09:56 Hoping anyway Jan 03 04:09:57 But i was broke at the time Jan 03 04:10:01 and now its not worth buying anymore Jan 03 04:10:26 It'd be awesome to have nightly builds up by a bunch of people just like Cyanogenmod & AOKP etc. Jan 03 04:10:39 Yeah it's pretty much pointless to buy one now. Jan 03 04:10:45 The Nexus 4 is superior in every way Jan 03 04:10:52 The GNex? Jan 03 04:11:00 Cheaper and Samsung. Jan 03 04:11:19 I prefer Samsung to LG, but that's me Jan 03 04:11:22 True enough. Jan 03 04:11:41 But specs... duh spezs Jan 03 04:12:16 A lot of people who bought the Retina Macbooks prefer Samsung to LG too, judging by the amount of complaints of display burn-in on the LG models. Jan 03 04:12:31 Heh Jan 03 04:12:49 There's some burn-in on the GNex that has been with it since the first week I've had it but I'm not sure about the N4. I think it's had its fair share of issues that way too. Jan 03 04:12:56 Did you hear about the iTheft in France? Jan 03 04:13:49 Saw something about it but didn't click any links or read more. Jan 03 04:14:31 1.3 million in product stolen. Which is what...2 macbooks and an iPhone? Jan 03 04:14:46 I LoLd despite how bad it is Jan 03 04:15:05 logicbuffer: my samsung works well with the trackpad Jan 03 04:15:13 after setting somethings in the xorg settings Jan 03 04:15:48 I'll look into that then, I don't remember seeing anything about xorg settings in any of the threads I've read but who knows Jan 03 04:16:24 logicbuffer: let me find it Jan 03 04:16:33 Xorg is nice... very nice for thing like macros etc too right? Jan 03 04:16:37 logicbuffer: -> http://forums.bodhilinux.com/index.php?/topic/7330-fix-touchpad-on-alpha-image/ Jan 03 04:16:39 Did you use the script that downloads all the tarball pieces and unpacks them to the stateful partition? Jan 03 04:16:48 Or some other way of getting Ubuntu on? Jan 03 04:17:02 logicbuffer: I use a debian based system on my device. Jan 03 04:17:44 Could you not pick at the script and install any ARM Linux? Jan 03 04:18:09 And did you upgrade your ssd jef91 Jan 03 04:18:15 Probably could, I didn't bother to even look into it Jan 03 04:18:18 I did not SmallFry Jan 03 04:18:21 still using the stock one Jan 03 04:18:26 9GB for my OS is more than enough Jan 03 04:18:29 I thought the ssd was soldered to the motherboard? Jan 03 04:18:35 Mar Jan 03 04:18:38 Mar Jan 03 04:18:41 Bah Jan 03 04:18:45 Nay Jan 03 04:19:08 Its an MSATA iirc Jan 03 04:19:23 Hmm Jan 03 04:19:47 Maybe though. Idk Jan 03 04:20:04 Ifixit has a teardown most likely Jan 03 04:20:21 I'm sick and don't want to crawl to my rig though Jan 03 04:23:54 Newegg has a 32gb Crucial M4 mSATA drive for $55 w/ free shipping. Not sure if want... Jan 03 04:25:02 More than a dollar a gig. Ugh Jan 03 04:25:25 Check before you order it though, I could be delusional Jan 03 04:25:33 Yeah I'm not going to spring for it just yet. Jan 03 04:26:11 Having the flu messes with your mind, I tell ya Jan 03 04:29:09 hhmm Jan 03 04:29:48 Jef91: how's the memory usage with bodhi? i haven't used enlightenment since the early elive days Jan 03 04:30:09 generally sub 100MB logicbuffer Jan 03 04:30:15 Gnite logicbuffer and jef91, hope to chat again. Jan 03 04:30:17 under 150MB with compositing Jan 03 04:31:00 SmallFry: take care, i'll be here much more often now Jan 03 04:31:45 Cool:) most activity I've seen in here Jan 03 04:31:49 Jef91: if so that's incredible, this is using 1.34gb at the moment Jan 03 04:35:18 ha unity just eats memory logicbuffer Jan 03 04:35:27 truth Jan 03 04:35:41 i think i'll go run that script instead and then come back here Jan 03 04:42:36 Rockbox? Jan 03 04:42:53 I haven't seen devs from them in ages Jan 03 04:50:55 SmallFry: So much for sleep Jan 03 05:01:26 logicbuffer: beware how you measure memory usage Jan 03 05:01:33 1.3gb of actual memory used Jan 03 05:01:39 or that INCLUDEs caches/buffers? Jan 03 05:11:13 raster: http://i.imgur.com/7rPKa.png d Jan 03 05:12:02 i'm not very familiar with top but 1.3g used should mean total ram used, no? Jan 03 05:12:08 logicbuffer: ok - so its more like 500m used Jan 03 05:12:20 almost 700k is caches and another 75 buffers Jan 03 05:12:36 logicbuffer: -> http://www.linuxatemyram.com/ Jan 03 05:12:39 so under 600 Jan 03 05:12:56 think of it this way Jan 03 05:13:06 cache is where linux stored data you load from disk Jan 03 05:13:12 if you opened a 200m video file Jan 03 05:13:14 and plqyed it Jan 03 05:13:19 200m of ram will be used Jan 03 05:13:29 even if the process that played the video has exited and gone Jan 03 05:13:37 as the kernel will have loaded and CACHED that data into memory Jan 03 05:13:43 just in case its needed again Jan 03 05:13:50 even if the app onlyuses 10mb at any time Jan 03 05:13:59 and is walking thru the video stream decoding some of it into ram/buffers Jan 03 05:14:28 (yes - i'm ignoring the kernel fadvize calls etc. - i'm keeping it simple) Jan 03 05:14:33 isn't caching to disk a horrible idea on ssds though? Jan 03 05:14:43 if u copy 500m of photos from your sd card to your ~/Photos folder Jan 03 05:14:46 especially relatively large caches on disks that are already pretty small? Jan 03 05:14:52 those 500m of photos will live in cache Jan 03 05:15:03 umm Jan 03 05:15:07 caching is never a horrible idea Jan 03 05:15:12 ram is faster than an ssd Jan 03 05:15:46 (i know of no cases where it is not - and even though i can imagine some bizarre ones... they are truly bizarre) Jan 03 05:15:55 so having daat sit in ram that is currently unused is good Jan 03 05:15:58 its using ram Jan 03 05:16:04 that would otherwise go unused Jan 03 05:16:10 unused ram is the tool of the devil Jan 03 05:16:15 unused ram is useless ram Jan 03 05:17:00 (ok - i simplified again - yes. ram costs power for self-refresh and being able to turn it off and turn off self-refresh can save you some power drain. as such no linux setup "does this" (in any normal case)) Jan 03 05:17:49 i'm sorry, i was confusing it with swap Jan 03 05:18:31 actually it says swap ~700m cached Jan 03 05:19:01 does that not mean it's holding 700m of something in a file on the SSD and not in ram? Jan 03 05:20:31 thats not swap Jan 03 05:20:32 thats cache Jan 03 05:20:37 u are using no swap Jan 03 05:20:54 its misleading that tis onthe line with "swap" Jan 03 05:21:01 yeah alright i understand now Jan 03 05:21:10 so ubuntu isnt AS BAD as u are thinking Jan 03 05:21:15 u're being a bit too harsh on it Jan 03 05:21:17 so i'm truly only using about 500m but it's holding another 700ish in memory in case it needs to use it Jan 03 05:21:25 what Jef91 os statitng is the actual ram usage Jan 03 05:21:36 ie memory excluding disk cache/buffers Jan 03 05:21:47 measuring memory is a "fine art" Jan 03 05:21:56 ina virtual memory os setup its hard Jan 03 05:22:06 one of the best measurements u can get is "free" Jan 03 05:22:10 do this: Jan 03 05:22:11 free -m Jan 03 05:22:39 Yea. Jan 03 05:22:47 Unity is normally just over 500MB at startup. Jan 03 05:22:54 So 3-5 times as much as E17 Jan 03 05:23:15 2:22PM ~ > free -m Jan 03 05:23:15 total used free shared buffers cached Jan 03 05:23:15 Mem: 2939 2228 711 0 119 1858 Jan 03 05:23:15 -/+ buffers/cache: 250 2689 Jan 03 05:23:15 Swap: 1905 61 1844 Jan 03 05:23:30 the numebr that matters is that "250" number Jan 03 05:23:39 thats your memory used on the os/syst4em as a whole Jan 03 05:23:50 excluding what is swapped out, what is cached or in buffers Jan 03 05:24:07 that is about the most accurate number you have to measure system meory usage Jan 03 05:24:11 okay i see Jan 03 05:24:18 sounds good Jan 03 05:24:45 so the best way to measure gnome vs kde vs unity vs e17 for example Jan 03 05:24:59 is boot a system and have them all boot to the same point Jan 03 05:25:02 eg a bare xserver Jan 03 05:25:04 no wm or anything Jan 03 05:25:08 no login manager Jan 03 05:25:22 now measure "free" Jan 03 05:25:42 that "used -/+ buffers/cache" is the memory you MUST HAVE to run what u have running now Jan 03 05:26:09 any memory in addition to that is luxury "caching" memory or memory for added apps or utilities on top Jan 03 05:26:20 NOW u run the desktop of choice Jan 03 05:26:29 unity/gnome/kde/e17/xfce/whatever Jan 03 05:26:34 then meausure again Jan 03 05:26:38 free Jan 03 05:26:45 that number will have gone up Jan 03 05:26:56 the mount it has gone up by is the "cost" of that desktop in addition to your base os Jan 03 05:27:16 thasts the fairest way to measure desktops/wm's against eachother Jan 03 05:27:21 and of coruse then run specific apps Jan 03 05:27:21 seems like they could've made it slightly more intuitive in top Jan 03 05:27:25 or perform specific tasks Jan 03 05:27:33 yeah Jan 03 05:27:34 eg open browser, go to a fixed webpage Jan 03 05:27:35 etc. Jan 03 05:27:38 and then measure as u go Jan 03 05:27:45 and see where that number goes Jan 03 05:27:59 that's how you are "fair" in measuring Jan 03 05:28:16 top/ps are useless for a regular user in terms of telling u actual mem usage Jan 03 05:28:29 there is 1 tool i know of that makes a "good attempt" to give you a fair memory rating Jan 03 05:28:34 its called "smem" Jan 03 05:28:38 therss a pkg for it Jan 03 05:28:43 apt-get install smem Jan 03 05:28:46 its a python tool Jan 03 05:28:52 if u do it as root u can get full details Jan 03 05:28:55 sudo smem -t Jan 03 05:29:11 and it'll list processes and TRY and allocate a FAIR share of memory to them Jan 03 05:29:26 it accounts for sharing of resources - eg libc, code pages etc. Jan 03 05:29:32 every process shares the same libc Jan 03 05:29:43 so memory used by libc CODE is in memory only once Jan 03 05:29:50 across ALL processes u run Jan 03 05:30:03 so it figures out the reference counts on these pages and divides by how many processes use them Jan 03 05:30:04 etc. Jan 03 05:30:09 so they are allocated their fair share Jan 03 05:30:25 the PSS column is the "fair share" amount Jan 03 05:30:32 its not perfectly accurate Jan 03 05:30:35 but as close as u'll get Jan 03 05:30:51 again - measuring the os as a whole with free is probably the most fair and best way Jan 03 05:31:19 so when Jef91 is saynig 100-150m Jan 03 05:31:24 he's saying the "free" number Jan 03 05:31:31 minus buffers and cache Jan 03 05:31:37 note Jan 03 05:31:52 that the amount of memory needed can be wildly different depending on gpu and gl driver stack Jan 03 05:31:56 good to know Jan 03 05:32:06 its an artifact of the driver itself Jan 03 05:32:31 so hsi numbers and yours may not match if u have different gpus/drivers/versions of driver or even different resolutions of screen Jan 03 05:32:45 yeah since data is moved from ram to kernel to gpu and back and then again to framebuffer, no? Jan 03 05:33:00 depends on architecture Jan 03 05:33:09 integrated gpus have no framebuffer Jan 03 05:33:15 well no dedicated one Jan 03 05:33:20 its part of main memory Jan 03 05:33:25 they dont need to move memory around Jan 03 05:33:29 discrete gpus do Jan 03 05:33:51 they may come with 256m or 512m or 1gb or more of dedicated memory on the gfx card in addition to system memory Jan 03 05:34:44 anyway. i've done my "measuring memory" rant for the month... :) hope it helps - spend some time reading up on it and experimenting and seeing Jan 03 05:35:05 and then when u see the usual misinformation spread as people quote the wrong numbers... spread the word. Jan 03 05:35:15 interesting Jan 03 05:35:30 looks like the mali t604 renders graphics into memory and then passes it from there to another gpu core that handles display Jan 03 05:35:46 its an integrated gpu Jan 03 05:35:57 its texture data, frameubuffer etc live in system ram Jan 03 05:36:05 the gpu has a CACHE on board Jan 03 05:36:06 like the cpu Jan 03 05:36:10 l1/l2/l3 etc. Jan 03 05:36:47 so it'll read data form ram just like your cpu and keep it in a cache Jan 03 05:37:05 if it needs the same data it pulls it from its cache, not from ram - just like your cpu. Jan 03 05:41:29 well this was enlightening Jan 03 05:41:41 thanks haha Jan 03 05:42:29 enlightenment is what i do Jan 03 05:42:30 :) Jan 03 05:42:39 pun intended :) Jan 03 05:44:58 should've seen that coming Jan 03 05:45:41 ;] Jan 03 10:22:25 happy new year everyone ! Jan 03 10:22:53 hny :) Jan 03 10:23:50 * ogra_ is wading through backlogs for the last 3 weeks and through >3000 mails Jan 03 10:24:04 oh ugh Jan 03 10:24:06 holidays? Jan 03 10:24:15 yeah, gotta love them :P Jan 03 10:26:02 i'm starting to be of the opinion that email is overrated Jan 03 10:26:11 and all discussion/work should be done via irc Jan 03 10:26:21 if u arent there - u miss it Jan 03 10:26:23 but thats cool Jan 03 10:26:32 u dont end up with backlogs out the wazoo Jan 03 10:26:33 :) Jan 03 12:11:38 Can Pandaboards be PXE booted? Jan 03 12:15:17 tnelson: you can boot panda from sd or usb Jan 03 12:32:54 tnelson, yeas, our u-boot supports PXE since 12.04 Jan 03 12:33:16 (it defaults to it if there is no configuration found) Jan 03 12:38:32 but first you have to provide uboot... Jan 03 12:40:25 right, you need an SD with u-boot or set up usbboot via OTG Jan 03 12:57:14 * ogra_ hugs infinity ... thx for that flash-kernel SRU, somehow i missed it Jan 03 13:15:56 Ah, so, 'Das U-Boot' is to embedded devices what grub is to x86? Jan 03 13:17:50 tnelson, well, its a bootloader, thats what grub and u=boot have in common :) Jan 03 13:18:08 I wouldn't go much further than that though :p Jan 03 13:23:59 So, I'm looking at http://omappedia.org/wiki/Minimal-FS_SD_Configuration versus https://wiki.ubuntu.com/ARM/Server/Install?action=show&redirect=ARM%2FOMAPHeadlessInstall... Jan 03 13:24:47 The zcat | dd approach would leave me with one pre-sized partition, no? Jan 03 13:24:59 (I've got a 32GB SD card.) Jan 03 13:31:40 Although I've got a 2GB usb stick, presumably I can dd onto that, then install onto the sd card? Jan 03 13:59:28 tnelson, nope Jan 03 13:59:43 SD needs to be the source device Jan 03 16:59:22 So I'm trying to compile to nexus 7 kernel found here -> https://launchpad.net/ubuntu/+source/linux-nexus7/3.1.10-8.19 , but towards the end of the compile the build always fails for me with this error message -> http://paste.debian.net/221410/ Jan 03 16:59:25 What am I doing wrong? Jan 03 17:09:35 Jef91, how do you build it ? Jan 03 17:14:41 ogra_ untared the sources, CDed into the directory and ran: Jan 03 17:14:41 dpkg-buildpackage -b Jan 03 17:15:02 on an arm machine ? Jan 03 17:16:30 Yes ogra_ Jan 03 17:16:52 did you install the build-deps ? Jan 03 17:17:49 Yes ogra_ ... If it was missing build deps dpkg-buildpackage would not have let it start compiling Jan 03 17:18:05 hmm, strange Jan 03 17:23:33 The error hints to me at a syntax error with the control file Jan 03 17:23:46 but I honestly don't know exact what that might be after looking at the file Jan 03 17:25:11 the control file is created during the build process, so its a bit weird, i would bet you have another error further above in your build Jan 03 17:25:56 (the guys in #ubuntu-kernel might know more btw) Jan 03 17:27:09 good call, thanks ogra_ Jan 03 17:27:22 thanks ogra_ Jan 03 17:27:31 np :) Jan 04 01:34:41 infinity: I just build eglibc current-raring with DEB_BUILD_PROFILE=stage1 and only got a libc-dev:arm64 package - all the other packages gave a dpkg-genchanges: warning: package libc6 in control file but not in files list Jan 04 01:35:33 ah yes. I see - that is supposed to happen Jan 04 01:35:56 ah of course. I want stage2, not stage1. doh Jan 04 01:58:20 hmm, seems there isn't a stage2 defined. Jan 04 01:58:53 have hacked one in to do a normal built without selinux- lets see if that works... Jan 04 02:22:54 hmm, fails due to missing libc6-prof files. need to nobble that for arm64. I think. Jan 04 02:48:47 Good evening gentlemen. I'm attempting to compile the BCM4330 driver from source on Ubuntu and running into undefined variable warnings. This is a known issue (https://code.google.com/p/bcmon/issues/detail?id=8) that I'm attempting to fix. I'm new to compiling source code, however from tracing the variables they should be defined through the file Kconfig. My first question (apologies, I did Google): does Makefile automatically load Kconfig? It has Jan 04 02:48:47 the calls (CONFIG_BCM4330), but the variables are left undefined. Any help or education is greatly appreciated :) **** ENDING LOGGING AT Fri Jan 04 02:59:58 2013