/srv/irclogs.ubuntu.com/2013/03/08/#ubuntu-desktop.txt

=== Ursinha- is now known as Ursinha
=== Ursinha_ is now known as Ursinha
=== Ursinha_ is now known as Ursinha
=== Ursinha_ is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
pittiGood morning06:08
BigWhale☠ morning ...06:17
BigWhale:>06:17
RAOFStill 4 hours sleep?06:19
BigWhaleRAOF, yes... I'll start hating Steam pretty soon. "Ooooh, a new humble bundle ... " "Oh, look, sunrise."06:22
chrisccoulson4 hours sleep?06:24
chrisccoulsonlucky you ;)06:24
=== asac` is now known as asac
RAOFchrisccoulson: How goes Operation: Two Young Girls06:25
RAOF? :)06:25
chrisccoulson1.5 hours ;)06:25
chrisccoulsonalthough, i did go to bed quite late06:26
chrisccoulsonso it's my own fault06:26
BigWhaleI am so glad I'm past this ... In 15 minutes I'll just scream: "put on your shoes and go to school!" :)06:26
didrocksgood morning!06:27
pittihey didrocks07:02
didrocksbonjour pitti, ça va?07:03
pittididrocks: ça va bien, merci!07:05
SweetsharkMoin a tous!08:32
pittiSweetshark: bonjour Monsieur, comment vas-tu?08:33
pitti$ sudo dpkg -P consolekit08:34
pittihah!08:34
Sweetsharkpitti: je vas bien, merci.08:38
seb128hey desktopers08:38
seb128pitti, victory? ;-)08:38
pittiseb128: well, milestone08:38
pittihttps://plus.google.com/107564545827215425270/posts/15va4vhty9T08:38
Sweetsharkpitti: how can it be that you push the agenda of the french mafia? Das ist nicht fair.08:38
pittiit's faaar from perfect yet08:38
pitti/msg Sweetshark I must, seb128 knows where I live!08:39
pittioops :)08:39
seb128hehe08:39
pittiVive la France^WUbuntu^Wlogind^Wour tech lead^Wla biére^Wnotre seb128!08:40
* seb128 hugs pitti08:40
* pitti tu donne une accolade en retour08:40
pittioh, "la bière", non?08:41
seb128oui08:41
* Sweetshark imagines pitti with a french flags on some barricades ...08:42
pittiSweetshark: nah, moane Fahne isch jetzt bayerisch08:47
* pitti va aller en grève au lieu08:48
Sweetsharkpitti: I wull di wat. Wi schnacken nu missingsch, und mit die andere Sprachens is dann daddeldu!08:51
pittihaha08:51
Laneyguten morgen09:12
seb128Laney, guten tag!09:13
seb128Laney, wie gehts?09:14
Laneygut, danke! es ist Freitag ;-)09:16
Laneywhat's the verb "to ride" (as in a bike)?09:17
Laneyjust fahren?09:18
seb128zyklus?09:20
pittiLaney: yes, "Fahrrad fahren"09:20
Laneyso ... früher fuhr ich mit meinem neuen Fahrrad zum Fluss :-)09:23
SweetsharkLaney: looks good to me.09:31
SweetsharkLaney: now translate for me: Wenn ist das Nunstück git und Slotermeyer? Ja! Beiherhund das Oder die Flipperwaldt gersput!09:31
chrisccoulsonhey seb128, pitti, Laney09:32
chrisccoulsonhappy fry-day!09:32
* Laney dies from laughter09:33
Laneyhey chrisccoulson09:33
chrisccoulsonheh :)09:33
SweetsharkLaney: consider yourself successful then ;)09:33
pittihey chrisccoulson09:34
chrisccoulsonhi pitti, how are you?09:35
pittichrisccoulson: quite fine, thanks! how about yourself?09:35
chrisccoulsonpitti, quite tired, after an intense week of debugging v8 :/09:35
pittiLaney: FWIW, I can't translate that either09:35
* Sweetshark imagines the impact the funniest joke in the world could have nowadays with twitter etc.09:35
Laneypitti: it's a joke from monty python ;-)09:35
pittiSweetshark: OMGNOO09:35
Laneyhttps://en.wikipedia.org/wiki/The_Funniest_Joke_in_the_World09:35
Laneys/en/de/ ;-)09:36
pittiLaney: oh, I did see that one, I just couldn't remember it any more09:36
pitti(probably for the better)09:36
Laneyprobably harmful for your health09:36
seb128chrisccoulson, hey, happy friday to you too!09:39
chrisccoulsonhi seb12809:40
seb128chrisccoulson, did you get at the bottom of your debugging? or is that going to keep you pulling hairs until the W.E?09:40
chrisccoulsonseb128, oh, i've got a pretty good idea what's going on now :)09:40
chrisccoulsoni just need to understand how v8 applies relocations to deserialized jit code, which is what is failing ;)09:41
chrisccoulson(chrome has a snapshot of v8 builtins linked in, to avoid the jit creating them on every startup. but references to C functions from the deserialized jit code aren't being relocated)09:41
* RAOF was going to ask why v8 would serialise its jit output :)09:42
chrisccoulsonit's been an interesting week ;)09:42
chrisccoulsonRAOF, the builtins are serialized during the build process09:42
chrisccoulsonand then linked in to the binary09:42
chrisccoulsoni didn't know this until this week, and spent some time trying to figure out why i wasn't hitting various breakpoints in the jit compiler)09:42
RAOFThey should cache all jit output!09:42
chrisccoulsonlol09:42
* RAOF is only partially joking there.09:43
=== vrruiz_ is now known as rvr_
czajkowskithis has been a very long week10:01
bryceczajkowski, no kidding10:01
seb128is there any short week here?10:02
seb128;-)10:02
czajkowskipicked the wrong week to do so much and run a conference on saturday, I can haz sleep right!10:02
=== vrruiz_ is now known as rvr_
rvr_czajkowski: Some energy http://www.youtube.com/watch?v=KlyXNRrsk4A ;)10:28
Laneywhen's unity 7 planned to be released?11:13
Laneydidrocks: ^?11:13
* Laney is reviewing his first FFe of the cycle :-)11:13
didrocksLaney: if we do have unity 7, it will just be a rename of current unity11:14
didrocksLaney: so not a new "version"11:14
didrocksLaney: we do release everyday, you know :p11:14
LaneyI thought that at some point that would be disabled in favour of a normal one11:14
Laneyso I can assume that FFes with approved branches will land the next day then11:14
* Laney was confused by the "milestone" that the bug has11:15
didrocksLaney: should be, not sure what you are referring to (I guess shutdown dialog?)11:15
Laneycorrect11:15
didrocksLaney: but it's something to ensure with mterry/cyphermox/ken/robru11:15
didrocksLaney: they are the one looking at the MP11:16
didrocksand stopping if upstream is going crazy11:16
didrockslike mterry did with shutdown dialog :p11:16
Laneybut he didn't vote against it, so is the merge going to happen?11:16
Laneyanyway I will approve it (just looking for future info)11:17
didrocksLaney: no, but they shouldn't merge it11:18
didrocksLaney: if they do, we'll revert11:18
didrocksbefore the FFe is acked11:18
LaneyI thought it was done by a bot once the approve votes were there11:18
didrocksLaney: the bot only care about the global state11:19
Laneyah, and who changes that? I've never done it for these projects11:20
didrocksLaney: people approving generally11:20
didrocksah yeah, sometimes, they forget :)11:20
Laneyso feel free to fix it for that one ;-)11:21
didrocksLaney: I prefer the stackholder dealing with it11:23
didrocks(mterry)11:23
didrocksdon't want to jump on anybody shoulder :p11:23
didrocksalready fixing the daily release is enough when jenkins is going crazy like this morning ;)11:23
Laneyfair enough11:26
Laneyseems like it goes crazy quite often ;-)11:26
didrocksLaney: unfortunately, I just got one bug since november on my code11:28
didrocksLaney: all the rest is autopilot tests/jenkins/utah11:28
didrocks(utah is way better now though)11:28
Laneyhehe11:28
chrisccoulsondoes anyone have any autopkgtests that require root access?11:41
Laneychrisccoulson: gvfs at least did have that11:42
chrisccoulsonLaney, ah, thanks11:43
Laneyyeah, still does11:43
Sweetsharkbdrung: dd0b8b521ea9543158b152cf80666b904027af49 trades a package-contains-broken-symlink against a doc-package-depends-on-main-package, which one do you like better?12:14
Sweetsharkbdrung: I would go for reverting dd0b8b521ea9543158b152cf80666b904027af49 -- no need to have delta to have a different warning12:14
=== MacSlow is now known as MacSlow|lunch
bdrungSweetshark: let me check12:27
Sweetsharkbdrung: btw running lintian is now part of the jenkins job ;)12:29
bdrungSweetshark: good. you can revert dd0b8b521ea9543158b152cf80666b904027af4912:36
bdrungSweetshark: do you have the debs from your latest build?12:38
Sweetsharkbdrung: yes, but without the revert or the patches you send to rene12:44
bdrungSweetshark: okay. can i have them?12:44
bdrungSweetshark: btw, how are the source tarballs generated?12:45
bdrungSweetshark: i have another patch: http://paste.debian.net/240472/12:46
chrisccoulsonis libdbus threadsafe?12:58
bdrungSweetshark: can i have your debs?13:07
seb128chrisccoulson, not sure, libdbus-glib is not, gdbus is13:08
seb128chrisccoulson, you might want to grab desrt to ask him when he's around, he probably knows13:08
Laneypretty sure it's not13:09
Laneyhyperair has had recent pain with this IIRC13:09
hyperairhmm?13:09
hyperairwhat's that?13:09
Laneylibdbus thread safety13:09
chrisccoulsonseb128, Laney, yeah, it's ok. i found my answer already13:09
chrisccoulsonsee https://bugzilla.mozilla.org/show_bug.cgi?id=84920413:09
hyperairoh gconf.13:09
ubot2`Mozilla bug 849204 in Widget: Gtk "crash in _dbus_watch_invalidate" [Critical,New]13:09
hyperairheh13:09
hyperairthere we go, same function.13:09
hyperairchrisccoulson: make sure dbus-glib is initialized with threads.13:10
hyperairthere's a recent gconf# upload in debian that fixed it13:10
hyperairand an ubuntu-only gconf patch written by me that does something with it13:10
Sweetsharkbdrung: sources are genned with the get-orig-sources from ./debian/rules13:14
Sweetsharkbdrung: lintian-override for image warning looks good, applied13:15
Sweetsharkbdrung: I will start the upload now and go for lunch ;)13:20
bdrungthanks13:20
kenvandineseb128, can you do a NEW review for me? accounts-qml-module13:21
seb128kenvandine, hey, sure13:21
kenvandinethx13:21
kenvandineseb128, good morning :)13:21
* Laney whispers something about feature freeze13:22
bdrungSweetshark: will you push to git?13:22
Sweetsharkbdrung: can do, would have waited for the new build to finish over lunch otherwise.13:24
bdrungokay13:24
seb128kenvandine, hum, Laney suggested I shouldn't NEW your upload because of FF...13:26
kenvandine:(13:26
kenvandineit's just for universe... nothing using it in the distro13:26
kenvandine:-D13:26
LaneyI think it should be fine following slangasek's relaxed FFe policy for the first few days13:27
Laneyis there anything else coming?13:27
kenvandinenot from me13:27
chrisccoulsonw00t, I OWN ALL THE BUILDERS NOW13:27
chrisccoulson(well, nearly all)13:27
smartboyhwchrisccoulson, :)13:27
kenvandinei tried getting it in yesterday... but was blocked on getting a branch merged13:27
Laneygo for it13:28
seb128Laney, thanks13:28
kenvandineLaney, thanks13:28
kenvandinei only care to get it in to make it easier for the guys working on the core apps project13:29
kenvandinenothing on the desktop uses it yet13:29
kenvandinei guess i didn't get gwibber replaced in time :/13:30
kenvandinethat's sad!13:30
seb128kenvandine, copyright says "either version 3 of the License." (minor detail) ;-)13:50
seb128kenvandine, looks good otherwise13:50
Sweetsharkbdrung: http://people.canonical.com/~bjoern/libreoffice4/debs/13:55
smartboyhwjbicha, would it be possible for you to package blender 2.66a-2 into the repos13:58
smartboyhw?13:58
=== MacSlow|lunch is now known as MacSlow
bdrungSweetshark: best way to download all these files?14:02
kenvandineseb128, grrr... i copied that out of another package :)14:27
kenvandineand can't remember which... so it's weird in more than one :-p14:27
didrocksbah, I'm tired of thunderbird hanging14:34
* didrocks will switch to mutt14:34
smartboyhwlol14:34
Sweetsharkbdrung: hmm, dunno. "wget -r"?14:36
Sweetsharkdidrocks: \o/14:36
didrocksSweetshark: I prefered to use our default though, but it's slowing me down now :p14:36
bdrungSweetshark: i have done that, but it tries to download more than i wanted14:36
* popey hugs mutt14:36
Sweetsharkbdrung: _rene_ is exploding right now because of the lintian changes.14:43
bdrungSweetshark: he talked with me on IRC14:44
bdrungSweetshark: it doesn't need much to ignite him14:46
=== m_conley_away is now known as m_conley
Sweetsharkbdrung: he has a point though on quite a few things14:47
Sweetsharkbdrung: but the root cause of this is that lintian isnt really the most appropriate tools for such a package14:48
bdrungSweetshark: $ grep '^E' lintian reveals libreoffice-dbg: wrong-file-owner-uid-or-gid usr/share/libreoffice/gdb/libreoffice/svl.py 1000/1000 and others. can we fix this by adjusting the permissions in the source tarball?14:52
bdrungi assume that a cp is used to install the files14:52
bdrungSweetshark: i get these with your debs (but my debs had no wrong-file-owner-uid-or-gid errors)14:54
attentedoes anyone have experience setting up the serial test harness in automake?14:59
bdrungSweetshark: have you updated the source package in between?15:03
seb128bdrung, can we spend a bit less time on lintian warnings? seems not the best use of Sweetshark's time...15:11
bdrungseb128: wrong-file-owner-uid-or-gid is a lintian error that should be addressed. i am not sure what causes it, because my test build didn't lead to it.15:13
seb128bdrung, well, seems like those lintian fixes are time consuming, created some bugs and are creating some issues with the Debian maintainer15:14
seb128bdrung, can I suggest that updating libreoffice to a non beta version could be a better use of time?15:14
jbichasmartboyhw: done, blender still needs someone to fix the powerpc build for it to migrate out of -proposed15:18
smartboyhwjbicha, I know15:18
* smartboyhw will be looking at it hopefully15:18
jbichasmartboyhw: good luck :)15:19
jbichado you have access to a powerpc machine?15:19
seb128jbicha, hey, any news about the gcalctool revert freeze exception? did you email the gnome release list?15:20
smartboyhwjbicha, no:P15:20
smartboyhwBut then we can normally see what happened I think15:20
smartboyhwjbicha, hmm I should probably work on getting blender maintained to catch up with debian.15:21
bdrungseb128: the issues with the Debian maintainer are irrelevant on getting the version out for ubuntu.15:28
seb128bdrung, well, time that is spent dealing with lintian is not spent getting the new version uploaded...15:29
bdrungSweetshark: can you push your changes to git? has the source package changed since i grabbed it from  http://people.canonical.com/~bjoern/libreoffice4/ two days ago?15:30
ricotzbdrung, hi, the packages were updated yesterday and most likely reflect "22 hours ago Bjoern Michaelsenonly sed complete lines"15:32
ricotz(bdrung, the source tarballs are identical of course)15:33
Sweetsharkbdrung: I just updating the source package _right_ _now_, so better not sync from there in this moment.15:37
bdrungSweetshark: okay15:37
bdrungSweetshark: can you push your changes to git?15:38
Sweetsharkbdrung: source package uploaded, git pushed. binaries are build locally on git head and add no further lintian warnings, but not uploaded to p.c.c (the binaries are thus at the state of today in the morning? IIRC)15:42
bdrungSweetshark: okay. i will grab the source, will do a final test build, and do the upload (unless i get the wrong-file-owner-uid-or-gid error on my test build)15:45
smartboyhwjbicha, don't know how to fix powerpc build for blender....15:46
kenvandineseb128, can you please binNEW accounts-qml-module?15:59
seb128kenvandine, you are demanding those days :p15:59
kenvandineseb128, i know :-D16:00
kenvandinei am trying to keep people from using the PPA :)16:00
seb128kenvandine, done16:00
kenvandinethx!16:00
seb128kenvandine, the powerpc is depwait on qtdeclarative, I guess that's not going to be sorted today16:00
kenvandineno...16:03
Sweetsharkbdrung: wrong-file-owner-uid-or-gid isnt fixed, neither are the non-standard-file-perm -- both will be fixed in the upstream installer (and as this only affects the gdb debugger hooks, this is noncritial vs. the 280 upstream bug fixes over beta2 that the users actaully care about)16:04
czajkowskihyperair: ello :)16:07
=== Ursinha-afk is now known as Ursinha
seb128ogra_,  hey, btw did you try to intall the nexus image today?16:19
ogra_seb128, the livefs builder broke (once again, sigh) ... seems the tarball is corrupt, i tested the 20130304 image which seems fine and copied it to the last-good-image dir16:21
seb128ogra_, thanks16:21
ogra_IS is on the machine and works on fixing it16:21
seb128ogra_, I will grab it from last-good-image then16:21
ogra_apart from the horribly crashy compiz in ubiquity that image seems fine16:22
seb128k16:23
seb128well if it's only ubiquity it's ok16:23
xnoxogra_: shall we revert to metacity for now? compiz nonetheless uses libmetacity for the gtk-decorators.....16:23
xnoxogra_: compiz will not be getting better with shift to mir.16:24
ogra_xnox, well, it would be nice to at least get some chrash dumps16:24
xnoxogra_: true.16:24
ogra_and the desktop wont shift to mir yet16:24
xnoxI see.16:24
ogra_phablet will16:24
ogra_we might start playing with Mir on the nx7 image around 13.10 i'D say16:25
ogra_ahead of x86 desktop16:25
ogra_(since it will be easy to use the android GLES drivers here)16:26
seb128ogra_, we plan to land Mir on desktop for 13.1016:30
bdrungSweetshark: wrong-file-owner-uid-or-gid is a regression from 4.0.016:31
ogra_seb128, then "during" instead of "around" :)16:31
ogra_but not for 13.0416:31
ogra_(which was the point i wanted to make)16:31
seb128ogra_, well for 13.10 release if there is a 13.10 on 6 months cadence :p16:31
seb128right16:31
Sweetsharkbdrung: are you sure? I dont think so.16:37
bdrungSweetshark: i haven't seen this error before. i will do a local build test and see if i can reproduce it16:38
Sweetsharkbdrung: Unless you prove to me that your locally build 4.0.0beta2 do not have the error, I wont believe its a regression.16:39
Sweetsharkbdrung: because a locally build one has the error, while https://launchpad.net/~libreoffice/+archive/ppa/+files/ure-dbg_4.0.1%7Erc2-0ubuntu1%7Eraring1%7Eppa1_amd64.deb build on a buildd does not16:39
Sweetsharkbdrung: and there were no relevant changes wrt after that.16:40
bdrungSweetshark: your deb came from a local build? a was building with pbuilder16:41
Sweetsharkbdrung: you believe a local pbuilder is no different from a real one? :) In theory thats true. In theory, theory and reality also match. In reality they dont.16:43
bdrungSweetshark: the archive uses sbuild IIRC. so there are sometimes small differences.16:44
Sweetsharkthe package came from a local pbuilder and that is different from a buildd.16:44
bdrungSweetshark: my last builds do not expose wrong-file-owner-uid-or-gid. so i will do another build and that should not expose this error.16:45
Sweetsharkbdrung: also, building arch-dep or -indep can make a difference etc.16:45
bdrungyes16:45
Sweetsharkbdrung: never assume the stuff to be the same, for a package the size of libreoffice there always might be something lurking there.16:46
Sweetsharkbdrung: learned that from painful experience.16:46
Sweetshark(more than once unfortunately)16:46
seb128ogra_, shrug, just put your last-good-image on my nexus and I get a "mount: mounting /dev/mcbllk0p9 on /root failed: Invalid argument"16:46
* seb128 wonders what's wrong with his nexus16:47
ogra_seb128, when did you charge it on thw wallcharger last ?16:47
ogra_*the16:47
seb128yesterday16:47
ogra_it doesnt charge much from a PC USB port16:47
seb128yeah, I never charge it from the PC16:47
ogra_and behaves really odd on low battery16:47
seb128I've the charger on the wall on my desktop16:48
ogra_but if you are sure its full ....16:48
seb128full minus what it lost since yesterday while being off16:48
ogra_yeah, that shouldnt matter16:48
ogra_hwo do you flash it exactly ?16:49
Sweetsharkbdrung: so _if_ wrong-file-owner-uid-or-gid is a regression it is since ~rc2-0ubuntu1~raring1~ppa1 OR it is a heisenbug. In the first case, I dont see what we touched there at all, in the second case -- its not a regression.16:49
seb128ogra_, I'm an idiot, ignore me16:49
ogra_fastboot erase boot; fastboot erase userdata; fastboot flash boot raring-preinstalled-desktop-armhf+nexus7.bootimg; fastboot flash userdata raring-preinstalled-desktop-armhf+nexus7.img16:49
ogra_(let me guess, you didnt unzip)16:50
seb128(correct)16:50
ogra_heh, that just struck me when i pasted the above :)16:50
seb128;-)16:50
bdrungSweetshark: i am checking if there is a difference between the source tarball that i am using and the one that you propose. it could be a heisenbug16:51
seb128ogra_, better with the unzip img :p16:53
Sweetsharkbdrung: I see the wrong-file-owner-uid-or-gid here on my local *.debs, but not on the ones that the ppa buildd did produce and I am pretty sure they are the same tarballs (just renamed them)16:54
ogra_seb128, phew :)16:54
bdrungSweetshark: okay. so you can expect me to do the upload later this day16:55
Sweetsharkbdrung: thx. FWIW, my local the md5sums and the one from the ppa dsc match.16:57
seb128ogra_, installer running \o/17:01
ogra_\o/17:01
ogra_The following packages have unmet dependencies:17:01
ogra_ foomatic-db-compressed-ppds : Conflicts: foomatic-db-engine but 4.0.8-3 is to be installed17:01
ogra_                               Breaks: foomatic-db-engine but 4.0.8-3 is to be installed17:01
* ogra_ scratches head17:01
seb128ogra_, seems like a tkamppeter's issue17:02
ogra_i guess there is a seed change ourstanding or some such ?17:02
ogra_*out17:02
* ogra_ gets that when manually triggereing a nexus7 build ... 17:03
tkamppeterogra_, does foomatic-db-compressed-ppds not simply uninstall foomatic-db-engine? for me it did so.17:05
czajkowskihyperair: ping17:06
ogra_tkamppeter, thats during image build, seems something still wants to pull foomatic-db-engine in (ubuntu-desktop ?)17:06
czajkowskic17:06
tkamppeterogra_, OdyX from Debian simply had put a Breaks: foomatic-db-engine without Conflicts: and Replaces: leading the installation to error out, so I added Conflicts:/Replaces:. ubuntu-desktop is installed on my system.17:07
hyperairczajkowski: pong17:09
czajkowskihyperair: mind a pm ?17:10
hyperairsure17:10
Laneyogra_: tkamppeter: seems like foomatic-filters Recommends it17:11
ogra_ah17:11
ogra_tkamppeter, ^^^ can you make sure that is fixed with the next upload ?17:16
=== rickspencer3_ is now known as rickspencer3
tkamppeterogra_, thanks, then I will remove the Recommends: from foomatic-filters.17:27
=== alan_g is now known as alan_g|afk
tkamppeterogra_, fixed foomatic-filters uploaded.18:57
xnoxcan gnome-control-center have plugins written in our preferred Qml/Qt+ 5 toolkit?20:27
desrtxnox: no.20:28
xnoxfail.20:28
desrtunless you get funky with xembed20:28
desrtand i don't know anyone to have tried that yet20:28
xnoxdesrt: well gtkplug works nice.20:29
desrtwould be a gtksocket in this case...20:29
desrtand a qtplug (or whatever)20:29
xnoxyeah that.20:29
xnoxdesrt: i once embedded network-manager into ubiquity installer window, it did work. was ugly though =)20:30
xnoxso i know it works with gtk+ to gtk+ ;-)20:30
dobeyxnox: well, you can embed windows apps in gnome-panel, so surely it works :)20:52
dobeyand skype and mumble, too20:52
=== matzipan is now known as matzipanini
bdrungSweetshark: my test build didn't have the wrong-file-owner-uid-or-gid lintian error. therefore i am going to upload the package.23:02
bdrungSweetshark: thanks for your work. please tag the latest git commit.23:02

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!