**** BEGIN LOGGING AT Mon Feb 21 02:59:57 2011 Feb 21 06:25:20 lbt: seriously man, wth? I've had a package submission in OBS for the last 11 days. I'd like to try and get Orbiter building on OBS sometime this year. :( Feb 21 06:51:03 Who else will be at Southern California Linux Expo next weekend? Feb 21 07:11:08 TSCHAKeee: eh Feb 21 07:12:33 TSCHAKeee: ah.... I see. The policy for Extras isn't sorted yet Feb 21 07:12:45 it certainly isn't "me" Feb 21 07:13:04 it's essentially the maemo QA process ported to meego Feb 21 07:20:24 lbt: so what should I do? Feb 21 07:20:57 TSCHAKeee: help form the process that's supposed to handle it. Can't you build your dependancies in your local repo too? Feb 21 07:20:58 lbt: i have 4-5 dependencies i need working. Feb 21 07:21:02 TSCHAKeee: the whole 'Extras' thing is a community effort Feb 21 07:21:09 i mean, builds pick up packages from same project they're in Feb 21 07:21:50 i keep getting unresolved packages for things I am trying to build..even though they are in my user directory...are they all supposed to be in the same project? Feb 21 07:21:56 you can create have multiple packages in each home project Feb 21 07:22:00 TSCHAKeee: that sounds odd Feb 21 07:22:05 it does Feb 21 07:22:21 TSCHAKeee: what's your "use for build" flag set to? (click advanced, then raw config, and pastebin that) Feb 21 07:22:29 you have to actively enable a special option to .... Feb 21 07:22:37 yeah, what Stskeeps said Feb 21 07:22:52 ok, i'll try it tomorrow Feb 21 07:22:56 it's 2am here. Feb 21 07:22:59 thanks. Feb 21 07:23:30 OK... it's a jetlagged 07:20 / 23:20 here :) Feb 21 09:11:57 i'm poking around in the intel tablet stuff Feb 21 09:12:30 it's all done with flat qml files, apache licensed Feb 21 09:17:07 looks like it would be very easy to add new panels Feb 21 09:17:13 unlike on the netbook ux Feb 21 09:17:57 ali1234, awesome which repository are you looking in? Feb 21 09:18:05 i'm not looking in any repository Feb 21 09:18:07 I assume the table image fs Feb 21 09:18:13 i'm looking in the image itslf Feb 21 09:18:58 in /usr/share/meego-ux-panels is all the qml files Feb 21 09:19:06 Guess it shows some of the strengths of qml. I think the ui are quite responsive Feb 21 09:19:16 yeah, really fast Feb 21 09:19:25 but, i can't see any way to resize the whole thing Feb 21 09:20:06 is it possible to scale the whole outer element Feb 21 09:20:13 so everything is relative to that one? Feb 21 09:20:39 maybe Feb 21 09:21:59 lcuk: What do you mean by resize ? Feb 21 09:22:10 You can always use the scale for top Feb 21 09:22:21 i just want to make the fonts smaller Feb 21 09:22:33 obit_sweden, well since ali1234 just said " but, i can't see any way to resize the whole thing" Feb 21 09:22:35 put scale: 0.5 in top element Feb 21 09:23:18 in which file? Feb 21 09:26:46 obit_sweden: made no difference Feb 21 09:27:03 ali1234: Can you find any top/root qml file ? I haven't looked at the files included in the tablet ui yet will have a look at it tonight. Feb 21 09:27:14 there is one called main.qml Feb 21 09:27:43 Ok in standard qml you just put scale: [factor] in top element Feb 21 09:27:55 oh, ok, changes were not saved over reboot Feb 21 09:28:14 They could however start it from c app and then make it resize to fit screen Feb 21 09:28:18 i guess i need to install this Feb 21 09:30:27 if you try to make the changes without reboot and just restart x instead you could see if that makes what u want Feb 21 09:30:43 Are u running it on a livecd or ? Feb 21 09:31:51 yes Feb 21 09:42:22 obit_sweden: this isn't exactly what i had in mind: http://imagebin.org/139019 Feb 21 09:44:10 ali1234, snap! http://liqbase.net/liq.meego.handset.on.x200s.20110221_001.jpg Feb 21 09:44:20 but from different UX image Feb 21 09:44:27 yes Feb 21 09:44:36 (related to bug 12455) Feb 21 09:44:37 but for a totally different reason Feb 21 09:44:38 Bug https://bugs.meego.com/show_bug.cgi?id=12455 nor, Medium, ---, jesse.barnes, NEW, Displayed UI does not adapt to available screen resolution Feb 21 09:44:54 yep Feb 21 09:45:02 this one adapts to screen size by scaling the whole thing Feb 21 09:45:16 yeah ali1234 I know Feb 21 09:45:55 are there any images of what this is supposed to look like? Feb 21 09:46:16 ali1234, I would guess it should look the same no matter how many pixels it was displayed on Feb 21 09:46:46 then it looks really bad, the way there's a gap at the top of panels, but at the bottom they go off the screen Feb 21 09:46:49 ie exactly how it looked before you started messing!? Feb 21 09:50:16 'ning Feb 21 10:06:28 getting closer... Feb 21 10:10:15 http://imagebin.org/139023 Feb 21 10:16:16 scaling the panelsContainer makes gestures fail to work, because their regions are all done in pixels Feb 21 10:37:17 hmm qml is kind of... horribly broken Feb 21 10:50:48 ali1234: why do you say that? Feb 21 10:51:14 meego of the way things like scale and object size interact Feb 21 10:51:20 *because Feb 21 10:51:33 say you have a object of size 100x100 Feb 21 10:51:54 and an object inside that object with scale: 0.5 and anchor.fill: parent Feb 21 10:52:06 it won't be the same size as the parent, it will be 50x50 Feb 21 10:52:40 or say you have an object 100x100 and inside it an object with scale 0.5 and height: parent.height and width: parent.width. what size do you think that object will be? Feb 21 10:53:01 it ends up 50x50 Feb 21 10:53:17 it takes the parentsize and since you put a scale there it scales it down Feb 21 10:53:25 yes Feb 21 10:53:33 makes sense to me Feb 21 10:53:41 but sub objects inherit both the scale 0.5 and the scaled size of the parent Feb 21 10:54:06 so if you nest a third object like that it ends up being 25x25 Feb 21 10:54:39 essentially the scaling is inherited twice Feb 21 10:54:48 yeah, its half the size of its parent... still making sense here :o Feb 21 10:54:49 so you can't just scale the top level object Feb 21 10:54:57 everything is relative to the parent Feb 21 10:55:06 you then have to go and add scale: 1.0 to all the children otherwise everything explodes Feb 21 10:55:20 But use anchor layouts instead of sizes in most cases Feb 21 10:56:05 yeah... anchor layouts suffer from the same problems Feb 21 10:56:35 I don't see the problem Feb 21 10:56:38 you can't use anchor.fill: parent if the parent is scaled, because it won't do what you expect Feb 21 10:57:15 * ahiemstra does not see the problem either Feb 21 10:57:22 Might want to show code Feb 21 10:57:44 then again, I come from a game/3d development direction where this behaviour is precisely what you want Feb 21 10:58:39 i have never seen a scene grapth where scaling the top level scene node by 0.5 scales all the children by 0.5*0.5 and all the children's children by 0.5*0.5&0.5 etc Feb 21 11:00:06 well, that's a point, in most scene graphs the scaling is not inherited in that sense Feb 21 11:00:52 if that's the only issue though... pointing the trolls at it might help Feb 21 11:00:59 hello slaine Feb 21 11:01:19 timeless: hello there Feb 21 11:03:03 what's the quickest way to restart X on meego? Feb 21 11:03:39 "initctl restart xsession" probably Feb 21 11:04:05 initctl: command not found Feb 21 11:04:39 ali1234: what meego are you running? Feb 21 11:04:56 meego-tablet-ia32-pinetrail-1.1.90.2.20110209.4 Feb 21 11:05:28 I didn't think they had switched away from upstart already.. Feb 21 11:05:37 you want to use systemctl then, most likely Feb 21 11:05:49 Don't know what the appropriate service name is though Feb 21 11:06:03 upstart uses "service" not "initctl" Feb 21 11:06:47 there's no xsession service however Feb 21 11:07:15 no "systemctl" either Feb 21 11:08:14 "killall uxlaunch" seems to do the job Feb 21 11:08:39 although it spams all the consoles with annoying messages Feb 21 11:12:34 we're not using upstart, we're using fastinit atm Feb 21 11:34:55 hi Feb 21 11:35:07 RST38h: ping Feb 21 11:35:10 vgrade: ping Feb 21 11:35:40 why are you pinging people all the time? :P Feb 21 11:36:50 because I have questions for them :p Feb 21 11:37:25 I see vgrade has managed to get meego 1.1 working on joggler Feb 21 11:37:46 I would like to know if he tweaked xorg to do that Feb 21 11:38:15 why would you? Feb 21 11:38:16 and I have an issue in my hdr algorithm...would like to ask people of libquill for their opinion on the subject Feb 21 11:38:29 because I have a gma 500 powered netbook Feb 21 11:38:39 meego 1.1 = xorg 1.9, i believe Feb 21 11:38:50 and the best I had with meego is to launch twm on it.... Feb 21 11:39:39 Stskeeps: there is a revamped edition of emgd drivers Feb 21 11:40:04 i know Feb 21 11:41:25 it supports xorg 1.9 Feb 21 11:43:19 vlj, have you looked at vgrade's ks? it shoudl answer your questions Feb 21 11:43:38 he use his own rpm for kernel :/ Feb 21 11:43:57 yes and you can look at his spec, he builds on pub meego obs Feb 21 11:44:52 anyways you can't 'tweak xorg' from the kernel ;-) Feb 21 12:06:21 hy all Feb 21 12:07:25 I have a local OBS set-up where I've imported the MeeGo 1.1 release. I can build just fine for i586, but I'm having some problems for armv7el Feb 21 12:08:57 I get Feb 21 12:08:59 """ Feb 21 12:09:01 running aaa-meego-accelerator postinstall script Feb 21 12:09:03 chroot: cannot run command `sh': Exec format error Feb 21 12:09:05 running bash-x86-arm postinstall script Feb 21 12:09:07 chroot: cannot run command`sh': Exec format error Feb 21 12:09:09 initializing rpm db... Feb 21 12:09:11 chroot: cannot run command `rpm': Exec format error Feb 21 12:09:13 """ Feb 21 12:09:56 andreizro: you need to install qemu too Feb 21 12:10:43 ok I thought it looks like the qemu-user / qemu-arm issue, but I was unsure Feb 21 12:10:50 thanks Feb 21 12:26:46 Folks, this is confusing. I am sure my UMTS-Dongle, Huawei E220/E270, was already recognized on a MeeGo system I am trying to polish. Now I got the data I need to enter in connman, but the dongle disappeared. And, more frustrating, I cant find the Howto I followed to make it work... Any guesses where to start looking? Or even a solution? Feb 21 12:27:26 Someone somewhere wrote it has to be attached while booting, been there, tried that Feb 21 12:28:01 I even tried with all usb ports, no one will ever know why. Feb 21 12:30:20 If someone wants me to go through any testing, I got a second netbook running MeeGo and a second dongle, no combination works AFAIS. Feb 21 12:31:39 styxnsoon, have you tried lsusb? Feb 21 12:31:53 arfoll, the stick is being recognized properly Feb 21 12:32:59 styxnsoon, have you looked at the logs? Feb 21 12:33:17 Bus 001 Device 005: ID 12d1:somenumber Huawei Technologies CO., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem Feb 21 12:33:38 arfoll, which log do you want me to have a look at? Feb 21 12:33:48 arfoll, i didnt Feb 21 12:33:53 /var/log/syslog Feb 21 12:34:06 that's where connmand should log things Feb 21 12:35:11 arfoll, no such file Feb 21 12:36:35 arfoll, it should be there, shouldnt it? Feb 21 12:36:47 arfoll, can i turn logging on somewhere? Feb 21 12:36:50 wait just booted my netbook Feb 21 12:36:56 thx Feb 21 12:39:41 styxnsoon, kill connmand then restart with -d Feb 21 12:40:07 debug should be in /var/log/messages Feb 21 12:42:49 arfoll, stupid question, how do i print the whole file to the console? Feb 21 12:43:00 arfoll, couldnt see anything suspicious in tail Feb 21 12:43:20 styxnsoon, i recommend tail -n Feb 21 12:43:34 but you can use cat to print the entire thing Feb 21 12:44:03 cat was the command im looking for Feb 21 12:44:58 cat with grep connmand is still too much Feb 21 12:45:15 can i view that page per page Feb 21 12:45:32 sorry this is becoming a linux-noob briefing Feb 21 12:45:40 pipe your grep output to less Feb 21 12:46:00 a charm Feb 21 12:46:14 btw you dont need cat if you use grep ;-) Feb 21 12:46:31 grep /var/log/messages connmand | less Feb 21 12:46:52 damned otherway round for connmand & messages Feb 21 12:49:36 arfoll, there seems to be nothing regarding an UMTS-connection Feb 21 12:51:32 styxnsoon, try killing and starting ofono with -d Feb 21 12:52:16 arfoll, the only line suspicious is: "dbus-daemon: [system] Rejected send message, 4 matched rules; [...] (uid=0 pid=413 comm="/usr/sbin/connmand)) Feb 21 12:53:46 styxnsoon, try the ofonod restart and look at the log entries for it, maybe it's just a permission error Feb 21 12:54:42 arfoll, im doing all this as root, should I? Feb 21 12:56:05 styxnsoon, yes connmand runs as root Feb 21 12:56:18 and so does ofonod Feb 21 12:57:45 arfoll, nice pick, I think we found some error: "klogd: [ 63.0201393] ofonod[484]: segfault at 10 ip 080c0dba sp bfea3ae0 error 4 in ofonod[8048000+bb000]] Feb 21 12:58:05 yeah that doesnt look too good Feb 21 12:58:18 it is oFono 0.26 Feb 21 12:59:10 http://repo.meego.com/MeeGo/builds/1.1.90/1.1.90.1.20110201.1/core/repos/ia32/packages/i586/ Feb 21 12:59:21 styxnsoon, maybe try 0.39 which is the latest Feb 21 12:59:41 might be worth reinstalling a full meego 1.1.90 instead of just one package (or connmand as well) Feb 21 13:00:20 make sure to file a bug report especially if the problem dissapears with a later of ofono Feb 21 13:00:35 so it can be fixed in later versions of 1.1.x Feb 21 13:01:19 arfoll, where can i fetch a pre-built installer for 1.1.90? Feb 21 13:01:58 arfoll, this will run stable, wont it? because i am trying to get this to work for my grandmother, her typewriter is not buggy ;) Feb 21 13:02:12 styxnsoon, here is the latest image - http://repo.meego.com/MeeGo/builds/1.1.90/1.1.90.3.20110215.10/netbook/images/meego-netbook-ia32/ Feb 21 13:02:43 styxnsoon, ah... you want stable? then maybe try intalling the ofono rpm first, see if that works if not install the connman rpm as well Feb 21 13:03:21 your a brave man if you are giving meego to your grandmother Feb 21 13:03:41 http://imagebin.org/139043 Feb 21 13:04:04 arfoll, had a look at winxp? i am going to have much less to explain :) Feb 21 13:04:05 ali1234: nice :) Feb 21 13:04:24 ali1234, just read your post very nice! Feb 21 13:04:53 you'll have to diff against the original to see how it works, i forgot to back mine up Feb 21 13:05:29 i'm not sure if this was so difficult because QML sucks, or because this particular example of using QML sucks Feb 21 13:05:44 arfoll, how do i upgrade a specific package in meego? I am glad to have found out the package-manager is called zypper... Feb 21 13:06:00 styxnsoon, download the rpm and install with rpm -ivh Feb 21 13:06:12 so wget and rpm Feb 21 13:06:18 yep Feb 21 13:06:19 the whole thing reminded me of css position:absolute problems, where you keep having to add a pixel here, remove a pixel there, to get it to look right Feb 21 13:07:01 ali1234, stop trying to put me off making my first qml app! Feb 21 13:09:22 also, like i said, this won't affect anything but the panels, the browser will still have huge text for example Feb 21 13:09:57 news: http://www.liewcf.com/fujitsu-lifebook-mh330-meego-netbook-6938/ Feb 21 13:11:23 arfoll, i have to remove the old ofono first, right? rpm throws up some conflicts Feb 21 13:11:47 arfoll, zypper remove ofono? Feb 21 13:13:19 arfoll, that was a bad idea of me :/ this command would remove 92 packages, freeing 132.3MiB Feb 21 13:20:16 arfoll, i decided to install MeeGo from the image you provided Feb 21 13:33:10 If i got no large microSD, im out of luck if I want to install MeeGo on my N900, right? Feb 21 13:33:44 2Gb+ Feb 21 13:33:48 is not overly large Feb 21 13:33:49 I could live WITHOUT multi-boot... Feb 21 13:34:08 lardman, the only card i got is 256MB :( Feb 21 13:34:18 old-skool ;) Feb 21 13:34:30 you should be able to pick up a 4Gb card for ~£5 Feb 21 13:34:33 and if i wanted to buy one, I would go for the fast cards, and they arent that cheap... Feb 21 13:34:34 styxnsoon: it's less the size, it's much, much more important about the speed :) Feb 21 13:34:42 though you'd be better going for the fastest card possible Feb 21 13:34:47 leinir, thats what i thought :) Feb 21 13:34:52 anything that cheap is going to be class 2 and thus sort of dog slow ;) Feb 21 13:35:08 i need class 6 or 8?! Feb 21 13:35:13 cant remember, sry Feb 21 13:35:20 Well, faster is better, really :) Feb 21 13:35:35 though... i wonder if the bus is capable of pulling more than 6MB/s Feb 21 13:35:38 Any way to install MeeGo on N900 NAND Feb 21 13:35:40 ? Feb 21 13:35:57 on the eMMC yes, though not sure whether there are instructions about Feb 21 13:36:45 that was my problem... I would try though. I can dd my current filesystem, so nothing catastrophal should happen Feb 21 13:37:41 lardman, isnt the N900 divided in 2 storage media? One for OS, flying fast, and one for my music and pictures ASOASF Feb 21 13:38:13 there is flash and an internal mmc card, yes Feb 21 13:38:50 I'm sure it would be possible to install to the internal flash, but there are no instructions, it's not supported as the Meego image is not day-to-day usable (that's my understanding for why) Feb 21 13:39:28 meego won't fit on the internal flash iirc Feb 21 13:39:37 it has to be emmc Feb 21 13:42:01 Would it be good or bad to install MeeGo on eMMC? Meaning, will it be faster/slower? Feb 21 13:42:18 styxnsoon, than what? Feb 21 13:42:40 andre__, than a microSD install Feb 21 13:43:26 on a reasonably fast microSD. Feb 21 14:02:26 i am just installing meego 1.1.90 on a netbook, i got another one that is running MeeGo 1.1. Can I tell MeeGo, if yes, how, to update to 1.1.90? I got some configuration done and would rather upgrade than reinstall. Feb 21 14:03:40 styxnsoon: http://wiki.meego.com/Upgrading_MeeGo_Versions Feb 21 14:05:29 jonnor, thanks Feb 21 14:08:25 MeeGo takes quite long to check my account password. If i don't encrypt the data, there's an option in installer, will it be faster? The password still wont lay around somewhere in plaintext, will it? Feb 21 14:12:59 styxnsoon: no, passwords for users in linux is never stored in plaintext Feb 21 14:13:37 styxnsoon: it _might_ be that the thing that is taking long is the unlocking of the luks volume, in which case it would be faster, yes Feb 21 14:15:13 luks unlocking usually takes about 2 seconds (the key is checked X times to make the time spent constant, for sequrity purposes) Feb 21 14:16:22 jonnor, thats my experience. 2s are quite long in comparison to the overall boot performance... Its like 1/8 of boot time. Feb 21 14:21:48 styxnsoon: yes, I know. The maintainers has been prodded about the issue, but were as I understand not very responsive to idea of reducing it Feb 21 14:25:56 jonnor: If its a trade-off with security, I keep waiting. But maybe they just cant be bothered... Feb 21 14:27:40 MeeGo 1.1.90 fails at recognizing my Huaweii-UMTS-Stick as well :/ I even had it plugged while installing... Feb 21 14:27:42 jonnor, from what I understand the entire security of luks rests around the fact your key is hashed and process takes a very long time, therefore brute forcing is very hard Feb 21 14:28:07 theres a very good paper or TKS1 security (which LUKS is) Feb 21 14:28:21 http://www.google.com/url?sa=t&source=web&cd=1&ved=0CBIQFjAA&url=http%3A%2F%2Fclemens.endorphin.org%2FTKS1-draft.pdf&ei=XHZiTeiUBseu8QP-wa3xCA&usg=AFQjCNGjKOgq2O-PV_JxSimHCLYRQtPBPw Feb 21 14:29:57 I dont like it, now 1.1.90 is of the opinion that booting is lame. Screen stays black... Feb 21 14:30:24 But it did boot, once. Feb 21 14:30:39 styxnsoon, work around, close the lid and reopen it (ie suspend/resume) Feb 21 14:30:54 afaik its noted on the bug chain filed about it, hold on I will get a link Feb 21 14:30:59 lcuk, affirmative :) Feb 21 14:32:30 Client fails to connect to D-BUS all the time btw... But thats probably filed. Feb 21 15:27:06 * CosmoHill sets fire to matlab Feb 21 15:28:36 poor matlab, what did its 1-indexed vectors do to you? Feb 21 15:29:21 are you speaking about meego here? Feb 21 15:29:39 I hope meego does not included non-free matlab stuff :) Feb 21 15:33:21 on the other hand, it might include Octave Feb 21 15:34:31 it would seem my router runs on matlab Feb 21 15:36:13 * CosmoHill looks a phoronix and praises them Feb 21 15:36:22 one less article I have to reveiw Feb 21 16:32:19 anyone in here know anything about HPC clusters? Feb 21 17:15:20 Hi Feb 21 17:15:27 Is it a holiday in the US today or something? Feb 21 17:16:33 dneary, yes it is a holiday in the US. President's day Feb 21 17:17:12 dneary: Yes, president's day (a minor holiday). Most businesses are open, most banks are closed, a lot of kids are out of school. Feb 21 17:17:41 Ah Feb 21 17:17:44 That explains it Feb 21 17:18:40 it's half term in the UK Feb 21 17:20:02 its "Family Day" in parts of Canada, wanted a holiday but couldn't come up with a reason Feb 21 17:21:56 Mobile Portland has just informed me that the talk they solicited from me next Monday is cancelled. I was almost done preparing it! Meanwhile, a cascade of emails from resigning coworkers is coming in. I'm not often depressed but . . . Feb 21 17:23:18 If the BoF I submitted to Embedded Linux Conference is rejected, then I'm just going to bed and staying there for a week. Feb 21 17:24:07 Alison_Chaiken: nokia coworkers? Feb 21 17:24:22 * CosmoHill hugs Alison_Chaiken (if approiate) Feb 21 17:24:22 gabrbedd, you guessed it. Feb 21 17:25:10 Alison_Chaiken: I'm so sorry. Heavy day. :-( Feb 21 17:25:30 CosmoHill: Just make sure it's a side-hug. Feb 21 17:33:44 Alison_Chaiken: make a counter-conference with your talk :) Feb 21 17:34:57 berndhs, I volunteered to give the same talk elsewhere. Why not? Work is mostly done! Feel enthusiasm about topic (meego-ivi) still. Thanks for (virtual) support! Feb 21 17:59:24 My main anxiety about changing jobs is that I will have to give back my IdeaPad and ExoPC and spare N900. But TI gave the Pandaboard to me personally, bwahaha! Feb 21 18:00:46 * TSCHAKeee still has A-test boards from when he volunteered time on OLPC Feb 21 18:05:53 Alison_Chaiken, Any good Linux sysadmin-type folks in your part of the company? Anyone want to move to Berlin or Boston? ;-) Feb 21 18:06:53 rmt2, I don't know sysadmins, but I have heard that some N**** office in Boston will be closed down. Feb 21 18:07:05 hu Feb 21 18:07:22 RST38h: hi Feb 21 18:10:17 vlj: hello Feb 21 18:10:39 RST38h: I have an issue with my hdr algorithm Feb 21 18:10:53 and? Feb 21 18:10:56 I managed to get hdr map for red, blue and green component Feb 21 18:11:03 but I cannot merge them properly Feb 21 18:11:22 is there some "magic" formula that define level of red vs blue vs green ? Feb 21 18:13:25 I have not done HDR algorithms Feb 21 18:13:45 but you worked on filters ? Feb 21 18:13:45 But the HDR software seems to indicate that there is no magic formula, it is user configurable Feb 21 18:14:03 You are asking a wrong question. Yes, I have done some image processing before. Feb 21 18:15:14 and processing different color component and merging them back should be "trivial" or ? Feb 21 18:15:56 Ok, let us see... Feb 21 18:16:48 So, you get three components, R, G, B. The problem with HDR is that these components are too long, longer than your color space (which I presume is 24 bits) Feb 21 18:16:58 yup Feb 21 18:17:10 Most HDR programs will let you save high-color HDR images in their private formats Feb 21 18:17:21 But you want the 24bpp image, right? Feb 21 18:17:30 yup Feb 21 18:18:35 HDR apps usually offers you a flexible user-controlled tool to compress images Feb 21 18:18:46 but how can I "clamp" the value without looking like a oversatured image ? Feb 21 18:19:09 Clamping it will cause blown highlights and empty dark areas Feb 21 18:19:22 Which HDR seeks to avoid to begin with Feb 21 18:19:52 Now, treating each component separately probably does not make sense Feb 21 18:20:03 I tried to normalize each component Feb 21 18:20:09 No, no Feb 21 18:20:27 By normalizing you do not blow the ends of the range but you dampen the whole range Feb 21 18:20:45 Most HDR apps will use exponential curves for compression Feb 21 18:20:51 Letting you control the curbe shape Feb 21 18:20:59 Consider this: Feb 21 18:21:48 CompressedColor = A * log(B * SourceColor) Feb 21 18:22:15 you compress with log first, then normalize Feb 21 18:22:50 ok Feb 21 18:23:15 SourceColor is the color in my hdr map ? Feb 21 18:23:54 RST38h: wouldn't A-law/µ-lawbe more appropriate ? Feb 21 18:26:25 vlj: Yes, each component separate Feb 21 18:26:38 pinchart: One of these is linear another is a log Feb 21 18:27:06 RST38h: I would be surprised about that :-) Feb 21 18:28:24 but regardless of your companding algorithm, the result will likely not look good if displayed directly on the screen. I suppose it depends on what vlj is trying to achieve Feb 21 18:31:09 pinchartl: I'm making a "HDR" algorithm that takes several pictures (with distinct exposure time) and "merge" them into a good contrasted image Feb 21 18:31:45 for now I got the set of pictures -> luminance map part of the algorithm Feb 21 18:32:09 now I need to create the 24bits image from that Feb 21 18:32:14 pinchaertl: vlj would like to compress HDR data into 24bpp space Feb 21 18:32:33 pinchartl: HDR apps usually achieve that with logariphmic compression Feb 21 18:32:48 whether the end result looks good or ugly is up to the beholder Feb 21 18:33:32 RST38h: for optimal results an adaptive companding algorithm would be better Feb 21 18:33:52 what do you mean by "adaptative companding algorithm" ? Feb 21 18:34:40 pinchartl: this will most likely look even weirder Feb 21 18:35:25 Alison_Chaiken: Re: the ideapad... yeah, the prices haven't been dropping on those and it's hard to find used ones. I got lucky and snagged one from Wal-mart last year for $400 (with Win7 Starter). But since Win7 Starter is a joke, that offering didn't last long. Feb 21 18:35:32 vlj: making the algorithm parameters vary depending on the area in the image Feb 21 18:36:26 Ideapad with MeeGo is a sweet device once you get it tuned up. Besides trying to create demos with with it for work, I find it nice just to use. Feb 21 18:36:42 but yes, it is a nice experiment to conduct :) Feb 21 18:37:10 Does anyone know which repo the Ideapad (that's the multi-touch thing from Dublin?) uses? Feb 21 18:38:19 * lardman would like to take a look at the source/spec file for Qt, which presumably supports multi-touch on that device Feb 21 18:38:35 lardman: the stock one didn't iirc, the tablet ux builds do.. Feb 21 18:38:48 thanks javispedro Feb 21 18:38:57 not sure if there's a later version on the repo Feb 21 18:39:26 Alison_Chaiken: Agreed. I love my ideapad. I won't give it up! Feb 21 18:41:09 I give up, searching for "Tablet UX" doesn't bring up much, does anyone have a link to the release? Feb 21 18:41:20 lardman: believe it or not, there's not a simple answer. I'll start with the rambling, confusing answer.... First, single-touch has always worked. Then, with 1.1 there was a trick you could do to get multi-touch to work with the Netbook UX, but none of the apps supported it. Feb 21 18:41:40 gabrbedd: I'm listening :) Feb 21 18:42:13 lardman: However, the 1.1 *handset* release *did* have stuff utilizing the ideapad's multitouch -- out of the box. No extra tweaks required. Feb 21 18:42:34 lardman: These are the pinetrain-mtf builds. Feb 21 18:43:08 ok, as I need to build for ARM, where can I find the repo (or more specifically the Qt source + spec file)? Feb 21 18:43:12 lardman: However, last time I checked a daily build on handset, this is broken again. :-) Feb 21 18:43:21 oh Feb 21 18:43:31 hi guys..can someone tell me how i can avoid that dhcp overwrite my routing table on meego 1.1.9...I try changing the default.profile on /var/lib/connman/.. but when I reboot the box is again there Feb 21 18:43:36 lardman: you were asking about ideapad at first, weren't you? Feb 21 18:43:43 gabrbedd: ah, 1.1. worked out of the box, ok Feb 21 18:44:01 gabrbedd: only because it's the only multitouch device that runs Meego that I know of Feb 21 18:44:17 but yes I was asking about Ideapad first :) Feb 21 18:44:28 btw, about netbooks. is there a tegra base netbook available with touchscreen? Feb 21 18:45:46 lardman: The current "Intel MeeGo Tablet UX" has a lot in common with the Handset UX, and MT works out of the box on the ideapad. Feb 21 18:45:53 MeeGoBot: Tablet UX? Feb 21 18:45:55 gabrbedd: tablet ux is a closed binary Feb 21 18:45:59 haha! Feb 21 18:46:28 one without touch they sell here is the toshiba ac100 which sounds otherwise quite good Feb 21 18:46:54 * lardman does some Googling to see if he can find how the Tablet UX is built Feb 21 18:46:59 food first though Feb 21 18:47:15 mihero: none afaik Feb 21 18:47:22 lardman: There's a link to a pre-alpha preview on this page: http://appdeveloper.intel.com/en-us/meego Feb 21 18:47:28 but tegra 2 does not support meego currently Feb 21 18:48:06 lardman: As for sources... there's two levels. There's a kernel level API that has been changing between 1.0, 1.1, and 1.2. Feb 21 18:48:14 yep, but i understood that it is under work Feb 21 18:49:20 lardman: Then there's the way that Qt interprets the `input` drivers... and the sources are http://repo.meego.com/MeeGo/releases/1.1/core/repos/source/qt-4.7.0-3.1.src.rpm Feb 21 18:49:23 * javispedro has lost all four rubber feet on his ideapad. build quality is clearly not great :S Feb 21 18:49:40 * gabrbedd only lost one so far :-p Feb 21 18:51:23 let's see... MeeGoBot: Tablet UX URL? Feb 21 18:51:40 MeeGoBot: Tablet UX URL? Feb 21 18:51:42 gabrbedd: Sorry, I've no idea what 'Tablet UX URL' might be. Feb 21 18:51:49 oh well. Feb 21 18:58:19 thanks gabrbedd Feb 21 18:59:33 lardman: you're welcome! Feb 21 19:00:50 I'll try a ks build as is and add mtev and see what happens with the version of Qt in the 1.1.80 repo I'm building from Feb 21 19:01:02 then I'll do some Qt building if required Feb 21 19:02:52 lardman: mtev was used on 1.1. They have changed it for 1.2 to use the new kernel MT interface that was introduced (IIRC) in 2.6.36. Feb 21 19:03:14 I'm on 2.6.35 iirc Feb 21 19:03:28 so I should be ok with that for the time being at least Feb 21 19:03:40 hopefully anyway :) Feb 21 19:03:59 * javispedro counts days before someone at adobe.com writes post about "the sad state about multitouch input support in Linux" Feb 21 19:04:33 * lardman thinks about the sad state of multi-card multi-head support Feb 21 19:05:09 i think about the sad state of the rpms in the adobo.com repo Feb 21 19:06:10 javispedro: Well, the end is in sight for linux MT woes... with XInput2 and a stable kernel protocol. Feb 21 19:06:55 javispedro: One problem I've been noticing in the kernel drivers is that they keep having to add support for ad-hoc specs (like Microsoft's MT event spec). Feb 21 19:07:16 javispedro: But I think that's stabilizing. Feb 21 19:14:50 lardman, that's one thing we've long been spoiled by on OS X. Feb 21 19:17:04 GAN900: Yes, only supporting one device, one kernel branch, and one windowing system... does have its advantages. :-) Feb 21 19:21:14 i for one am completely unspoiled by OS-X :) Feb 21 19:26:05 GAN900: works well under Windows too it must be said Feb 21 19:26:26 lardman, not in my experience. Feb 21 19:26:52 GAN900: oh, I've never had troubles with 3 heads across 2 different cards (from different manufacturers too) Feb 21 19:27:30 so, does one need to setup the mtev driver? Feb 21 19:27:32 * GAN900 recalls trying to get a projector working on XP sometime around 2002. Feb 21 19:27:51 that was a few years ago ;) Feb 21 19:47:05 gabrbedd: i got f14/planetccrma/rpmfusion/google repos integrated onto the hybrid netbook platform... Feb 21 19:47:25 i was able to launch jack and qjackctl... Feb 21 19:47:47 and hydrogen Feb 21 19:50:04 ha! sounds like fun! Feb 21 19:50:20 npm: ha! sounds like fun! Feb 21 19:53:24 indeed, it's very distracting, esp. when it's not what i should be doing Feb 21 19:53:34 gabrbedd: ^^^ Feb 21 19:54:39 npm: :-) BTW, do you DJ? Feb 21 19:54:46 yes Feb 21 19:55:10 i thought so... i was just curious. :-) Feb 21 19:55:53 or at least i have in the past, and hope to again in the future Feb 21 19:56:49 given that everybody and their dog is a dj these days, i've told myself that i should only DJ using MIDI... Feb 21 19:59:16 but i have 0 time for music these days... Feb 21 20:02:23 npm: true. Feb 21 20:10:09 i'm just happy over the little things, like being able to use fedora's 'roxterm' in meego just by doing 'zypper in roxterm' :-) Feb 21 20:18:53 Any finns or people living here got invitation to Ahjo for tomorrow? Feb 21 20:19:27 ? Feb 21 20:20:31 looks it up: ahjo=the core of the fire where the sampo (in finnish national epic Kalevala) was made. it brought power, happiness and richess to it's owner Feb 21 20:20:47 ahjo == club in helsinki central .. Feb 21 20:21:12 anyone here familiar with the meego vms? Feb 21 20:21:25 is jori hulkonnen playing?? :-) Feb 21 20:21:45 i just got sms saying that there's "free drinks on Intel, register at xxx@gmail.com etc etc .." Feb 21 20:22:08 hrm, looks like it's debian squeeze Feb 21 20:22:23 * timeless tries to remember how to abuse debian Feb 21 20:25:50 "clubahjo.fi, free drinks on intel, tuesday 21.feb at 8PM, Please forward this to 3 people you most like working with. RSVP FIFO. Register Asap by sending your name to and receive confirmatnion mail".. its either a hoax or some really agressive head hunting .. Feb 21 20:31:38 sounds good. Hopefully Intel will pick up the some more devs and get meego going faster forward Feb 21 20:36:43 vlj, delayed ping back, see joggler images on http://bug10738.openaos.org/, take the red pill. EMGD included. MeeGoR&D Feb 21 20:36:45 Bug https://bugs.meego.com/show_bug.cgi?id=10738 nor, Undecided, ---, tero.kojo, NEED, Meego community ftp server or similar needed Feb 21 20:48:17 vgrade1: How up to date, how useful and what's the UX? Feb 21 21:12:55 Jaffa, bang up to date, both netbook and tablet UX images, also arfolls xbmc build. Good for tinkering. MeeGo-R&D Feb 21 21:18:18 Jaffa, http://plixi.com/p/77686885 Feb 21 21:19:49 Jaffa, http://www.youtube.com/watch?v=DjTWQZgsuuQ Feb 21 21:20:01 whoa Feb 21 21:20:13 did intel publish the sources to their tablet UI? Feb 21 21:33:34 wmarone, no, just rehosted the pinetrail image onto Joggler Feb 21 21:35:49 ah Feb 21 21:48:08 vgrade1: Apparently changing the resolution should scale the QML UI on Tablet UX well Feb 21 21:48:52 Jaffa, multiple people have not had that experience Feb 21 21:49:01 and obviously you have seen vgrade1's results now Feb 21 21:52:05 hey lcuk Feb 21 21:52:11 hi CosmoHill \o Feb 21 21:52:27 lcuk, sec while I boot joggler Feb 21 21:52:31 just go back from uni, been working on my cluster after hours Feb 21 21:52:33 damn thing is loud Feb 21 21:53:07 does it also dry your hair? :D Feb 21 21:53:32 no but it will give you static shocks Feb 21 21:53:38 at least I hope their static Feb 21 21:53:43 o_O Feb 21 21:54:05 everything is grounded apart from the metal frame on the trolly Feb 21 21:54:18 that is normally you and shoes and carpet, if its coming from the machines and you don't normally get static nearby then get them checked by uni technician Feb 21 21:54:42 * lcuk does not mess near electric Feb 21 21:54:53 Jaffa: ping Feb 21 21:55:03 I did 14 nodes in 30 mins Feb 21 21:55:06 Jaffa: how would you want an index set of repositories to work? Feb 21 21:55:16 i'm looking at slurping the meego git repos Feb 21 21:55:27 and there's a conflict between a flat list and a tiered list :( Feb 21 21:55:46 kinda weird, you add the nodes one by one in the order you want, giving each an IP. When you're done you unplug them all from the mains then turn them all on at once Feb 21 21:56:09 sounds stupid but it gets the ones that have booted into a OS Feb 21 21:56:31 CosmoHill: What OS? (just curious) Feb 21 21:56:39 Rocks v5.5 Feb 21 21:57:41 cool! :-) Feb 21 21:57:53 1.2Kw peak :o Feb 21 21:58:32 CosmoHill, the network of n900s is similar, but each is assigned a name Feb 21 22:01:23 "IBM announces Globus Toolkit for N900 clusters" Feb 21 22:01:39 they're automatically named but I'm specifying the IP so that they have the same IP under both windows and linux Feb 21 22:05:07 CosmoHill, is that necessary? Feb 21 22:05:16 * lcuk has been trying to get rid of fixed IP for a while Feb 21 22:06:06 the linux nodes IP are static, all I'm doing is making sure they match up with the windows DHCP Feb 21 22:06:26 I think rocks stores the IP in it's dhcp table and on the node itself Feb 21 22:14:11 http://cross-lfs.org/~cosmo/cosmo_cluster/genglia.png Feb 21 22:14:49 nice CosmoHill Feb 21 22:14:57 but what do you computer on the cluster? Feb 21 22:15:04 compute * (damn muscle memory) Feb 21 22:15:35 not gotten that far yet, I'll be comparing windows and linux computing nodes Feb 21 22:16:02 the last cpu spike was me running my goldbach conjecture algorthim from the first semester Feb 21 22:17:23 nice Feb 21 22:20:35 look at the to power pictures in the directory, one is mid goldbach, the other is a min after Feb 21 22:23:08 my theory is correct tho, if you can use a kettle (on that power socket) you can use a cluster Feb 21 22:28:21 lol CosmoHill Feb 21 22:28:34 but will your cluster make a coffee Feb 21 22:29:16 physically? no, it can give you a nice render will realistic fluids tho Feb 21 22:29:22 with* Feb 21 22:29:51 ahh so you can use it for raytracing? Feb 21 22:29:53 * lcuk perks Feb 21 22:30:15 anything that I can benchmark on windows and linux Feb 21 22:30:48 that reminds me Feb 21 22:30:56 someone asked for a liq* tool for benchmarking Feb 21 22:31:30 wasn't me but I'll have one Feb 21 22:31:48 well I found out graphics are already topped outish Feb 21 22:32:06 lcuk: are you still testing MeeGo images? Feb 21 22:32:08 on n900 and ideapad they both get ~58fps for most tasks Feb 21 22:32:12 Venemo, yes Feb 21 22:32:26 lcuk: could you please gimme some info about how telephony works on N900? Feb 21 22:32:48 you speak into the mic bit and listen through whats known as the "earpiece" Feb 21 22:33:10 lcuk, http://www.youtube.com/watch?v=WuJbWMsY3bE, http://www.youtube.com/watch?v=VTV-V-7Zo7Q Feb 21 22:35:37 nice vgrade1 Feb 21 22:37:06 >> sleep Feb 21 23:26:53 * timeless sighs Feb 21 23:26:58 anyone here want to hold my hand? Feb 21 23:27:10 * timeless is trying to remember how to poke rpmmacros Feb 21 23:31:40 poke? Feb 21 23:36:46 thing to do is define your own rpm macros and declare it a standard Feb 21 23:37:37 yeah, i know Feb 21 23:38:12 i need to convince rpm to use %{name} as part of %_sourcedir ,.. i think Feb 21 23:41:09 you can't just do _sourcedir: this%{name}that ? Feb 21 23:41:46 maybe Feb 21 23:41:50 * timeless is running into pain Feb 21 23:41:57 or define a replacement for %_sourcedir Feb 21 23:42:13 * timeless had a script which vaguely worked and now it doesn't Feb 21 23:59:06 Hi, quick question, is there a qmake macro for meego platform? Feb 21 23:59:45 compile time macro which is provided by Qt qmake tool or even something I can hack in code? Feb 22 00:03:16 makulkar: what do you need it for? Feb 22 00:03:44 alterego, in webkit code i need to detect if its building for right platform and build some code Feb 22 00:03:57 what code exactly? Feb 22 00:04:22 alterego, disable some rendering feature which doesnt really work well Feb 22 00:04:32 Hrm, Feb 22 00:04:58 Well, at runtime there should be a /etc/meego-version or something similar. Feb 22 00:05:14 right now we have this, Q_WS_MAEMO_6 macro which is passed as command line while building webkit. Feb 22 00:05:16 As long as it compiled I can't think of a better way to disable something like that. Feb 22 00:05:23 Yeah Feb 22 00:05:48 Unfortunately there is nothing on MeeGo like that :( Feb 22 00:07:47 Actually, maybe check for QT_OPENGL_ES ? Feb 22 00:08:05 Then you know your Qt is compiled against embedded code .. Feb 22 00:08:21 (though this may not always be the case) Feb 22 00:08:43 Q_WS_MEEGO returns a few google hits ;) Feb 22 00:08:50 alterego, right! atleast not something that can get in to webkit code :) Feb 22 00:09:17 javispedro: it doesn't exist unfortunately. Feb 22 00:09:29 can't you define your own macro depending on what you're building for ? Feb 22 00:09:31 right.. mobility should be using it for some reason.. will search there Feb 22 00:09:35 At least not with my compiler setup and I doubt with obs Feb 22 00:10:07 berndhs: yeah, your .spec file can tell qmake extra defines I guess. Feb 22 00:10:54 it seems better to explicitly say "i'm disabling ThisFeature", rather than doing it throught a back door that doesnt explain why Feb 22 00:12:01 makulkar: actually most mobility stuff is just generic GNU/Linux Feb 22 00:12:40 There's some MAEMO_5 dependant stuff (I think) but with MeeGo no. Feb 22 00:13:04 I suppose you could also check for meego touch headers ^_^ Feb 22 00:13:28 but that'll only be available for a few more releases I guess ^_^ Feb 22 00:13:55 My point being you're more interested in a MEEGO_HANDSET_ARM flag I'd imagine Feb 22 00:14:18 You don't want to screw up code for netbook, ivi, or atom handhelds ... Feb 22 00:14:47 i would define my own macro RENDERING_FOO_IS_BROKEN, and then turn off code accordinly Feb 22 00:15:13 alterego, hi, I can't make the sprint this week, as I have customers in at the day job. Feb 22 00:15:24 rather than saying its meego, because people will forget why stuff is turned off Feb 22 00:15:53 vgrade1: :( I don't think I can make it either as I'm in Leicester today, tomorrow and wed, then london on thurs and fri :( Feb 22 00:15:54 alterego, right.. berndhs suggests the same thing as well, this makes more sense. Just send that extra flag while building qtwebkit Feb 22 00:16:22 altergo, I'll ping an email to the guys, ps ds2 is a wind up Feb 22 00:17:34 vgrade1: okay, thanks. Feb 22 00:47:32 ok, so that almost worked Feb 22 00:47:38 except %{name} sucks, it has a version Feb 22 00:47:48 * timeless goes to hunt a better variable Feb 22 00:54:40 what variable do you want? Feb 22 00:57:41 so, i've actually gotten the source side happy Feb 22 00:58:00 what i want to influence is BUILD/Package-Version Feb 22 00:58:07 so that it uses BUILD/Package instead Feb 22 00:58:44 that would be something like BuildRoot Feb 22 00:59:14 i don't think so.. Feb 22 00:59:21 so far BUILDROOT is totally empty Feb 22 00:59:28 i'm only doing rpmbuild -bp Feb 22 00:59:29 ah Feb 22 00:59:36 %_buildroot Feb 22 00:59:47 it's not BUILD/ that i need to influence, it's the part after it Feb 22 01:00:18 %_builddir ? Feb 22 01:00:24 https://www.redhat.com/archives/redhat-list/2003-January/msg02015.html Feb 22 01:00:26 i suspect buildsubdir Feb 22 01:01:16 nope :( Feb 22 01:02:55 hrm, maybe i can't influence that one Feb 22 01:02:59 ok, let's assume i can't Feb 22 01:14:08 * timeless wonders about the best way to build 1400 packages Feb 22 01:15:34 quickly? Feb 22 01:15:48 well, let's go with "efficiently" Feb 22 01:15:54 assume i have say 5 cores available Feb 22 01:16:07 the plan I have for my project is make one, make sure it works the way you expect Feb 22 01:16:10 note that build here is -bp Feb 22 01:16:13 then deploy / make the others Feb 22 01:16:25 build prepair? Feb 22 01:16:31 yeah, my one (ConsoleKit) "builds" Feb 22 01:16:40 so i'm not particularly expecting failures Feb 22 01:16:42 good luck with that, I'm off to bed Feb 22 01:16:43 cyas Feb 22 01:19:20 timeless: before using more cores, check if you are spending CPU time or loading .h files Feb 22 01:19:33 berndhs: this is only preps Feb 22 01:19:41 not actual builds Feb 22 01:21:05 * timeless settles on hrm, ~36 concurrent tasks Feb 22 01:23:08 01:21:55 up 6 days, 23:58, 5 users, load average: 916.30, 436.41, 169.64 Feb 22 01:23:14 ok, that wasn't very nice of me :) Feb 22 01:23:46 depends on who the other users are or if its just you 5 times :P Feb 22 01:23:55 i sure hope it's just me Feb 22 01:24:05 i suppose there might be an apache user Feb 22 01:25:12 * timeless should have used 'nice' Feb 22 01:25:25 * timeless goes to figure out how to use renice Feb 22 01:29:14 /bin/tar: groff-1.18.1.4/doc/gnu.eps: Cannot utime: No such file or directory Feb 22 01:29:20 * timeless wonders what kind of verb 'utime' is Feb 22 01:33:24 + cd /data/mxr-data/repo.meego.com/BUILD Feb 22 01:33:25 sh: %{__id_u}: not found Feb 22 01:33:30 * timeless wonders what that means Feb 22 01:34:05 hrm, i need to go shopping, i'm missing 'unzip' :( Feb 22 01:34:25 oh, i bet 'cannot utime' means 'you don't have utime' Feb 22 01:34:49 utime is a system call I think, changes the time stamps on the file Feb 22 01:35:16 yeah, but in context here, it either means or Feb 22 01:35:19 i'm hoping the latter Feb 22 01:35:24 so if the file is not there, it cant do it Feb 22 01:35:32 /bin/tar: Mesa-7.9.1/src: Directory renamed before its status could be extracted Feb 22 01:35:53 could have something to do with it, that stuff moved or got renamed Feb 22 01:36:10 tar or something made a list of files to deal with, and then they are gone Feb 22 01:36:22 sh: ruby: not found Feb 22 01:36:23 * timeless sighs Feb 22 01:36:41 wtf does something need ruby as part of -bp Feb 22 01:36:45 or unmounted Feb 22 01:37:03 error: Bad owner/group: /data/mxr-data/repo.meego.com/SOURCES/matchbox-window-mnager/matchbox-window-manager-1.2.tar.gz Feb 22 01:41:38 grr Feb 22 01:41:48 renice isn't pamd (?) friendly Feb 22 01:43:33 * timeless adds 'sed' to the shopping list Feb 22 01:45:03 and doxygen? Feb 22 01:45:22 cant live without doxygen these days Feb 22 01:45:29 for buildprep? Feb 22 01:45:32 that seems odd Feb 22 01:45:55 is does yes, unless the docs are part of the distribution Feb 22 01:46:24 01:44:27 up 7 days, 21 min, 5 users, load average: 9347.96, 8051.81, 4856.82 Feb 22 01:46:50 i say nuke 'em, nuke 'em now Feb 22 01:46:56 i'm not sure how Feb 22 01:47:20 kill -9 ? Feb 22 01:47:58 * timeless should try killall -STOP -u timeless Feb 22 01:48:08 but there are a log ot them i guess Feb 22 01:48:10 note that renice +15 -u timeless failed Feb 22 01:48:14 it couldn't find timeless Feb 22 01:48:39 /etc/passwd isn't actually driving this box, and i think renice (and probably others) rely on it Feb 22 01:49:19 kill -9 `ps -u timeless | awk '{print $1}' ` Feb 22 01:49:28 except that will kill itself too :) Feb 22 01:49:36 it wouldn't Feb 22 01:49:42 kill is being run as root Feb 22 01:49:54 ah Feb 22 01:50:00 i mean, it would to the extent that root is a sudo by timeless Feb 22 01:50:02 but.. Feb 22 01:50:11 i should do it using pts/n Feb 22 01:50:15 but if you run as root off an su, it will kill its parent Feb 22 01:50:16 but i'm not sure what /n is here Feb 22 01:50:35 Cannot find user timeless Feb 22 01:51:18 shutdown -r now Feb 22 01:51:55 ps aux|grep rpmbuild|awk '{print $1}'|xargs kill -STOP Feb 22 01:51:58 * timeless tries that Feb 22 01:52:26 shold get some of them Feb 22 01:54:30 hi marienz Feb 22 01:55:08 oddly, the system is vaguely useful Feb 22 01:57:01 hrm, oh, and i need to ignore Prereq Feb 22 01:57:13 /bin/tar: FUR-0.4.6/registry_utils.c: Cannot open: File exists Feb 22 01:57:15 * timeless chuckles Feb 22 02:03:01 bad router, spontaneously reboot less (preferably not at all) Feb 22 02:17:57 heh Feb 22 02:18:46 wow, my vi request from probably an hour or two ago was completed Feb 22 02:21:09 sudo kill -9 vi # die! die! Feb 22 02:22:09 * gabrbedd knows it's not valid bash... but it makes good copy. Feb 22 02:22:14 might be better off rebooting now :) Feb 22 02:27:48 how do you feel the future of meego? Feb 22 02:30:36 / Feb 22 02:30:37 \/ Feb 22 02:33:36 deas: channel is usually dead about this time. Feb 22 02:33:51 Excepth that NishanthMenon just showed up. How awesome is that? Feb 22 02:34:04 :-) Feb 22 02:34:07 deas: Feb 22 02:34:24 gabrbedd, hi Feb 22 02:34:51 deas: I think meego still has a bright future. I think it's future on phones is more questionable. However, Nokia *is* releasing one and it *will* preceed any WP7 phone. Feb 22 02:34:59 Intel is making one, too. Feb 22 02:35:15 * NishanthMenon detects touchy topic Feb 22 02:35:41 NishanthMenon: long time no see! Feb 22 02:35:50 gabrbedd, just hibernating ;) Feb 22 02:35:53 I say Elop will see the light after MeeGo phones embarrass N**** in the market :) Feb 22 02:38:05 berndhs: Don't pretend that Microsoft is a joke. They're not. They'll kick our @$$3z if we don't have our A-game on. Feb 22 02:38:41 its just my guess, depends on what manufacturers offer Feb 22 02:39:18 if the phones are done better, Microsoft has no chance, their market dominance doesn't transfer to phones Feb 22 02:40:30 Microsoft only benefits if the Nokia + WP7 phones are very good, and nobody knows that yet Feb 22 02:42:06 Well, WP7 is shipping, MeeGo is not -- so they're already ahead. Feb 22 02:42:24 no, Nokia has zero WP7 phones right now Feb 22 02:42:24 Anyway, enough talk... I need to go work on MeeGo. :-) Feb 22 02:42:28 they have to make them first Feb 22 02:42:44 yes, dont' slow down, there is no time to lose :) Feb 22 02:42:46 berndhs: Nokia has zero WP7 phones, true.... Feb 22 02:43:02 berndhs: But WP7 is shipping through other mfg's. Feb 22 02:43:33 sure, but the N+M phone has to be better, otherwise it won't help MS any, nor will it help N **** ENDING LOGGING AT Tue Feb 22 02:59:57 2011