/srv/irclogs.ubuntu.com/2011/03/14/#ubuntu-desktop.txt

kklimondahey, can someone take a look at https://code.launchpad.net/~kklimonda/evolution-data-server/ubuntu/+merge/53196 when you have a moment? It's a cosmetic change, not really worth its own upload (hence I haven't subscribed ubuntu-sponsors) but It's still worth updating when someone makes a new upload.06:25
pittiGood morning07:27
TheMusoHey pitti.07:43
pittihey TheMuso, had a nice weekend?07:43
TheMusopitti: Very nice indeed thanks, although was still warm here. Its a bit cooler today however.07:43
pittispring has just begun here, I for one am happy that it's finally not freezing any more :)07:44
TheMusoI'm somewhat the other way. Looking forward to a bit of cooler weather.07:44
TheMusoWhich of course will change com the end of winter, or sooner.07:45
didrocksgood morning08:03
pittibonjour didrocks!08:15
didrocksGuten Morgen pitti :)08:15
rodrigo_morning08:20
didrockshey rodrigo_08:24
rodrigo_hi didrocks08:25
jbichawill the gnome3 ppa be getting a rebuild soon for 2.91.91?08:27
SweetsharkMorning all08:28
rodrigo_jbicha, yes, I've been on vacation last week, so didn't do any update, but will get it up-to-date in the next few days08:28
jbicharodrigo_: thanks, I was just curious08:29
chrisccoulsongood morning everyone09:04
pittihey chrisccoulson09:07
chrisccoulsonhi pitti, how are you?09:07
pittiI'm great, thanks! Had a Taekwondo camp yesterday again, and helped a friend moving on Saturday, so some muscle strain :)09:08
pittibut it was really good yesterday09:08
pittichrisccoulson: btw, I did some samplings of the existing search plugins, and some are already localized; so I extracted them all from maverick's -base packages and committed them into langpack-o-matic09:17
pittichrisccoulson: writing the necessary code/test cases now09:17
pittichrisccoulson: I propose we only install them into -base and do a full maverick -base update, to clean up the replaces: mess09:17
pittidpm: ^ FYI09:18
pittidpm: (I just requested a full maverick export)09:18
pittithen maverick and natty will be consistent, and both will work again09:18
dpmhey pitti good morning, sounds good. How will this affect the ongoing testing of maverick langpacks? Do you want to call it off, release new full langpacks and re-new the call for testing?09:20
chrisccoulsonpitti - thanks for working on this. i didn't realise that some were already localized :/09:20
=== RAOF_ is now known as RAOF
pittiI'm off for some two hours for a doctor appointment09:26
mptmvo, good morning09:34
mvohey mpt09:36
mptmvo, have you received any contributions this cycle on bits of <https://wiki.ubuntu.com/SoftwareUpdates>?09:37
mvompt: not that I'm aware of09:38
mvompt: there is the branch about renaming update-notifier to update-manager-daemon09:38
mvobut that was prev cycle iirc09:38
mptok09:38
mvompt: haven't look at the spec in detail since a while, but I remember we wanted to push more updates into the background, right?09:39
mvompt: oh, there was another patch to run the "update" automatically on manual invocation via the desktop file09:40
mptmvo, more updates in the background was mdz's goal, but that's more a platform team thing to decide that they're confident enough to make it the default09:40
mvompt: fair enough, some support has landed in the unattended-upgrades branch now (but not enabled by default yet). so we can ask people running natty (once its final) to use it and give us feedback09:41
mptCorrecting the settings probably should happen before changing the defaults, otherwise there would be all sorts of confusion about exactly what had changed09:42
mvoindeed09:44
mvoI think that this is something we should push for next cycle09:44
mvofixing the backend to be proper polkit etc09:44
mptyah09:45
pittismspillaz: since Friday or today (ish) I have huuuge shadows around the focussed window, which spills over a lot in the adjacent terminal windows. is that a bug, or meant to be that way?11:29
smspillazpitti: its meant to be that way11:36
smspillazwe could clip the shadows of windows when they are close to each other but *shudder* that's a hack11:36
pittiugh11:36
pittithe spillover to the adjacent virtual desktop is now also much worse than before11:37
pittismspillaz: e. g. this (http://people.canonical.com/~pitti/tmp/shadow-spillover.png) is workspace 2 with a fullscreen firefox11:39
pittiyou still see the shadow from teh terminal on workspace 111:39
and471can someone upload the images http://i.imgur.com/CJL4C.jpg http://i.imgur.com/CJL4C.jpg to a site such as imagebin.org11:42
and471that way I can see them (blocked by content screening for me)11:43
and471?11:43
pittiand471: http://imagebin.org/14291511:47
and471pitti: thanks very much11:48
=== MacSlow is now known as MacSlow|lunch
Laneyanyone seen asac recently? is he on holiday?12:40
seb128Laney, the canonical calendar says he's on holiday today at least12:41
seb128seems he should be back tomorrow12:41
Laneyok12:46
=== ogra is now known as Guest42034
ari-tczewHello. I noticed that some developers from desktop team often upgrade packages in Ubuntu if package exists in Debian. I was wondering why get upload access in Debian and upgrading there?13:10
=== Guest42034 is now known as ogra_
seb128there is quite a bunch of sources updates in debian and synced but it's not always pratical13:12
seb128it requires having a debian installation and testing those there as well which is extra work13:12
ari-tczewwhy not practical?13:12
seb128debian often doesn't track unstable series as we do13:12
seb128there is also quite some sources who have a maintainer who handle updates in a slower way than debian but doesn't want someone taking over his package13:13
seb128than "ubuntu" rather13:13
ari-tczewseb128: I see not logical flow of work. e.g. Robert Ancell upgrades package then some day later merge with Debian. Why not get sync/merge directly from Debian?13:14
ari-tczewanyway later merges do extra work as well13:15
seb128because we don't know when Debian will update and it makes sense to wait on them13:15
seb128they could not upgrade before natty13:15
ari-tczewseb128: so I'm talking: get upload access in Debian and upgrade there.13:15
seb128it doubles the work and often those have a maintainer and not free to update13:16
seb128Debian updates are usually done by the maintainer not free to be done by whoever wants to do it13:16
ari-tczewseb128: actual model doesn't double the work? maintainer and Robert upgrade packages. I'm counting this 1...2... double work!13:17
seb128right, but not double the work for us13:17
seb128it's two people doing the work once,it's not one doing it twice13:18
ari-tczewI don't understand and it's still uneconomical for me.13:18
seb128well you might need to try for a bit to understand it ;-)13:18
ari-tczewseb128: no, this is uneconomical13:19
ari-tczewand eod.13:19
seb128I've basically stopped maintaining things in Debian when we have an Ubuntu diff13:19
ari-tczewseb128: so let's blacklist all packages related to desktop team.13:19
seb128because it means: having to have a Debian unstable system uptodate to be able to do binaries build there, having the upload the binaries to debian and the source to debian and ubuntu13:19
ari-tczewit doesn't make sense merge with Debian once a year.13:19
seb128having to deal with bugs in debian due to your update13:20
seb128having to ask for a sync request13:20
seb128test your update on both distros13:20
seb128you basically need 2 computers instead of one13:20
seb128or to play with vms, etc13:20
seb128not to mention you have to care about extra architectures and extra bugs if you become maintainer on the debian side13:21
ari-tczewseb128: good point, VM ;-)13:21
seb128which you need to keep updated, which means blocking your downloads for ours every day if you are on a slow internet13:22
seb128ours->hours13:22
seb128not to mention that to get there you need to get upload rights to Debian, which is non trivial to get13:23
=== MacSlow|lunch is now known as MacSlow
ari-tczewseb128_: does Robert test packages merged with Debian before uploading?13:27
seb128_dunno, you would have to ask him but likely yes13:28
seb128_why?13:28
ari-tczewseb128_: if so, he spends a time anyway on testing, whatever whether it's on Debian or natty.13:28
seb128_well testing on 2 distros mean you need to start them both, use them both for a while, see if it's buggy13:29
seb128_so basically it means having 2 computers and switching between those13:29
seb128_it's not because something builds or work on debian than it does on ubuntu13:29
seb128_sometimes the gcc behaviour is different or the xorg stack is not in the same versions13:30
dpm<dpm> hey pitti good morning, sounds good. How will this affect the ongoing testing of maverick langpacks? Do you want to call it off, release new full langpacks and re-new the call for testing?13:31
pittiari-tczew: is apt.VersionCompare obsolete somehow? "from apt import VersionCompare" works in python-apt, but not any more in python3-apt13:31
pittidpm: yes, as the current maverick ones are mostly broken anyway due to bug 73276813:31
ubot2Launchpad bug 732768 in langpack-o-matic "ask.com is the only search provider" [High,Fix released] https://launchpad.net/bugs/73276813:31
ari-tczewpitti: I don't have an idea. why you're asking me?13:32
pittiari-tczew: sorry; brain fart13:32
pittimvo: is apt.VersionCompare obsolete somehow? "from apt import VersionCompare" works in python-apt, but not any more in python3-apt13:32
ari-tczewkk13:32
dpmpitti, ok, I'll call it off then. There's another potential regression in the current maverick-proposed langpacks (bug 544203)  - which is the best way to track it? Just leave it assigned to langpack-pt-br?13:35
ubot2Launchpad bug 544203 in language-pack-pt-base "pt_BR and pt_PT firefox translations not included in Maverick" [Undecided,New] https://launchpad.net/bugs/54420313:35
mvopitti: its the new push for "lower_case". i.e. version_compare() will work13:36
pittimvo: ah, danke!13:37
mvoyw13:37
pittimvo: hmm: AttributeError: 'module' object has no attribute 'version_compare'13:40
pittimvo: (neither in the py2 version)13:40
mvopitti: python -c 'import apt_pkg; apt_pkg.version_compare' works for me, hold on a sec for py313:46
mvopitti: aha, in apt13:46
* mvo looks13:46
=== seb128_ is now known as seb128
pittimvo: oh, apt.apt_pkg.version_compare() apparenlty?13:56
mvoyeah13:56
mvoa bit silly IMO, worth talking to juliank about it13:56
pittimvo: at least that works fine; now on to the replacement of apt_pkg.ParseTagFile().. :)14:00
pitti(got that, too)14:01
=== bcurtiswx is now known as Guest95906
rickspencer3tseliot, hey, can you please check out bug #73398914:24
ubot2Launchpad bug 733989 in xserver-xorg-video-intel "dragging with Dell Mini 10v track pad is unforgiving" [Undecided,New] https://launchpad.net/bugs/73398914:24
rickspencer3note that it's a possible input issue, not a video issue14:24
rickspencer3but it seems that something changed on my Dell Mini 10v14:25
rickspencer3pedro_, can someone take a look at bug #734600 ?14:25
ubot2Launchpad bug 734600 in gst-plugins-bad0.10 "camerabin throws an error in gst_camerabin_capture_start gstcamerabin.c(4036)" [Undecided,New] https://launchpad.net/bugs/73460014:25
* tseliot has a look14:26
=== james_w` is now known as james_w
mterryHeyo!  I've got a FF exception for ya'll.  So, I want to split time-admin into a separate package from gnome-system-tools (so that we can avoid having indicator-datetime-preferences and time-admin both installed by default).  This will require a trivial binary package NEW approval and some seed changes for flavors that don't use indicator-datetime by default.  This seems like maybe a FFe is in order?14:30
tseliotrickspencer3: I see this line in your X log. Maybe the code that handles the option was changed:14:30
tseliotOption "AreaBottomEdge" requires a percent value14:30
mterrys/for ya'll/question for ya'll/14:30
chrisccoulsonmterry, what else are we using from gnome-system-tools? we're pretty much at the point where we can drop it from the CD, right?14:31
chrisccoulsonoh14:31
chrisccoulsonusers-admin too14:31
seb128right, users-admin14:31
seb128someone was supposed to package the new user-account thing previous cycle? ;-)14:31
seb128mterry, hey, check with pitti but seems a trivial ffe to get14:31
chrisccoulsonyeah, it's merged in to the new control-center now though :P14:31
seb128mterry, you probably need a bug for the tracking14:32
mterryseb128, k14:32
pittimterry: sounds fine, go ahead14:32
pittimterry: will indicator-datetime-preferences do all the bits that time-admin provides?14:33
mterrypitti, yeah, both ntp and manual date changes14:33
mterrypitti, and timezone14:33
pittinice14:33
mterry(and not "will", but "does"  :))14:33
pittimterry: replacing t-a with i-d-p will require an FFE, though14:34
seb128pitti, it's not replacing, we have both now14:35
pittiah, we do? ncie14:35
pittican't tell, indicator-time still crashing :/14:35
mterrypitti, https://bugs.launchpad.net/ubuntu/+source/gnome-system-tools/+bug/73275714:35
ubot2Launchpad bug 732757 in gnome-system-tools "FFe: time-admin and indicator-datetime-preferences shouldn't both be installed by default" [Undecided,New]14:35
tseliotrickspencer3: so, yes, upstream changed the value that the option (which I originally implemented) accepts. It should be just a matter of updating a configuration file. I'll write further details in the bug report and fix the issue. Thanks for reporting the problem14:35
mterrypitti, indicator-datetime-preferences crashes or indicator-datetime does?  you can run the prefs dialog manually14:36
rickspencer3thanks tseliot!14:36
pittimterry: the indicator-service does; right, it does start manually14:36
pittimterry: Set the time "AUtomatically from the Internet" is grayed out for me, though14:38
mterrypitti, even if you authenticate?14:38
pittiyes14:38
mterrypitti, that's set insensitive if we asked system-tools daemon if you can use ntp or not.  It apparently said no?14:39
pittimterry: ah, I don't have ntp server installed indeed14:40
mterryhah, feature!14:40
pittisame confusion with 'ntpdate' as before, I guess14:41
pittiperhaps it shouldn't be displayed at all then14:41
mterrypitti, current behavior is part of the spec, so that's mpt's call14:44
seb128tedg, mterry, chrisccoulson: the recent appmenu-gtk update broke the desktop menu, not sure if any of you want to work on that14:44
mterryseb128, desktop menu?14:44
seb128like it doesn't display the nautilus menu when being on an empty workspace14:44
tedgseb128, Hmm, it does for me...14:45
chrisccoulsoni'm still trying to get to the bottom of the unity-panel-service memory leaks ;)14:45
chrisccoulson(of which, i've managed to fix some)14:45
seb128doesn't here and we got a bug report where the user said that downgrading appmenu-gtk fixes it14:45
tedgseb128, Which bug number?14:45
seb128tedg, 73305014:46
tedgbug 73305014:46
ubot2Launchpad bug 733050 in appmenu-gtk "appmenu-gtk 0.1.96-0ubuntu1 prevents desktop global menu from appearing" [Undecided,New] https://launchpad.net/bugs/73305014:46
mptmterry, I didn't realize (though I should have guessed) it was possible to not have the ability to use NTP14:47
mterrympt, huh yeah.  I thought I had done that whole 'insensitive if can't use ntp' bit off a line in the spec, but looking now, I can't see it.  I went rogue (<- pitti, I take it back, it is my fault.  :))14:48
mptmterry, what do you think of prompting to install the NTP software if you click "Automatically from the internet" without it installed?14:48
mterrympt, that's fine.  I can work on that and file an FFe14:49
pittiI think that's what time-admin does, isn't it?14:50
pittiyes, just confirmed14:50
pittimterry, mtp ^14:50
mterryshoot, so we're not feature compatible14:50
pittiTBH I don't really like this14:51
pittiwe already have ntpdate, so time should be good enough for desktops14:51
pittiI really don't see a strong case for installing ntpd locally14:51
pittiif I were the ruler of the world, I'd kill this radio button completely :)14:51
mptpitti, and do what instead?14:52
pittibut I guess that was discussed over and over already14:52
pittimpt: nothing; just offer to correct the time manually14:52
mptpitti, it hasn't as far as I know, I was just pulling it out of my imagination14:52
pittiit already gets set whenever you (re)connect to the internet14:52
pittiset by ntpdate, I mean14:52
kenvandinepitti, it does?14:53
pittikenvandine: /etc/network/if-up.d/ntpdate14:54
mptpitti, so are you saying that if Ubuntu can connect to an NTP server, you sholdn't be able to override it manually at all?14:54
mptshouldn't14:54
pittimpt: there are two ways to get NTP14:54
kenvandinepitti, and that happens by default right?14:54
pittimpt: (1) ntpdate; that's what we install by default, there is no GUI for configuration, and it's done when you connect to the net14:54
kenvandineif so... then i think ntpd shouldn't get installed14:54
pittimpt: (2) a local ntpd server, which has a different configuration file, which you can set up by time-admin (and apparently the indicator prefs now)14:55
pitti(2) is useful if you don't reboot your computer very often (like servers), and never change/reconnect to your network14:55
mptpitti, I didn't know that #2 existed, so I wasn't designing with it in mind :-)14:56
pittimpt: right; it's utterly confusing14:56
mptpitti, what's confusing about it?14:57
pittiit leads people into thinking that there's no internet sync by default, and that they better install that14:57
pittimpt: that there are two independent NTP ways, and that the GUI doesn't point this out14:57
mptBut "Automatically from the Internet" is selected by default, right?14:57
mptSo it is synced by default14:57
pittii. e. that GUI thing will switch from "sync time when you connect to the network" to "permanently sync time with the network"14:57
kenvandinei would say just drop it from the gui then... let it just happen14:57
pittimpt: no, we don't install ntpd by default (for good reasons), so it's grayed out by default14:58
mptpitti, but the interface isn't for ntpd. Again, I didn't even know that ntpd existed.14:58
pittimpt: time-admin offers you to install ntpd when you select the option, in time-admin-prefs it's insensitive14:58
seb128mterry, ^14:58
mptgnargh14:58
kenvandinempt,  i think it is for time-admin14:58
seb128mterry, maybe wait before adding the option ;-)14:58
pittimpt: time-admin's interface is for ntpd; if time-admin-prefs replicates that, then it will be, too (and I think it does according to what mterry said)14:58
mptok, so, next question14:59
pittimpt: if even you and I get confused by it, I wonder what users will think..14:59
pittimpt: don't get me wrong, time-admin isn't any better here14:59
mterryseb128, yeah, I'm watching.  I remember now dealing with this in oem a bit actually.14:59
pittibut maybe that's the right time to remove this confusion14:59
mpt*Why* does time-admin have an interface for ntpd?14:59
pittimpt: good question14:59
kenvandinempt, designed by geeks..14:59
kenvandine:)14:59
pittiit's a rather advanced thing15:00
mptIt doesn't look like it15:00
tedgseb128, So are you getting the no-desktop menus error?15:00
mptpitti, the choices in time-admin are "Manual" and "Keep synchronized with Internet servers". I see no mention of "Keep synchronized with a local network server".15:00
mptor anything like that15:00
glatzorhello mpt, pitti, seb128 and mvo15:01
seb128hey glatzor15:01
pittihey glatzor15:01
seb128tedg, on unity yes, I didn't try in GNOME15:01
glatzormvo: have you already found the time to look at python-apt multiarch?15:01
seb128tedg, well it's not so much an error that you don't get the nautilus menu on empty desktops15:01
pittimpt: if you install ntpd locally, the local server will regularly sync with internet ones, and then your local computer clock to the local ntpd15:01
tedgseb128, Can you see if killing nautilus fixes it?  I'm curious if it's just restarting nautilus with the newer version that fixes it :)15:01
seb128tedg, it does15:02
mvophey glatzor15:02
pittibrb, plumber just arrived15:02
mvoglatzor: no, sorry :(15:02
mptpitti, so the behavior of "Keep synchronized with Internet servers" is different depending on whether ntpd is installed?15:03
tedgseb128, Okay. I bet it's a race, probably not version dependent.15:03
mterrympt, no, I think what pitti's saying is that we don't actually have a real "manual time" option.  That the two choices right now are secretly "keep synchronized all the time" and "synchronize when you connect to the internet"15:04
=== rodrigo__ is now known as rodrigo_
mptmterry, then what are the time and date controls for?15:05
mptFor making temporary changes that get washed away next time you connect to the Internet?15:05
pittimpt: yes15:05
mterrympt, good question.  Yup, sounds like.  I remember dealing with a bug about that in oem now15:05
mptEh.15:05
pittiif you have an offline computer, or wnat to change it on the plane, etc.15:05
mptThat's daft.15:05
pittimpt: note you can disable ntpdate in /etc/default/ntpdate15:06
pittibut we don't have a GUI for it15:06
mterrypitti, but there's no policykit-friendly way to change ntpdate's config, eh?15:06
pittiso perhaps we should devote the time-admin-pref to configuring ntpdate instaead of ntpd?15:06
pittimterry: no, you'd need a backend which changes /etc/default/ntpdate15:06
mterrypitti, yeah, we could switch system-tools-backend to configure that setting15:07
pittimterry: ah, it's actually using s-t-b?15:07
mterrypitti, yeah15:07
pittiI had hoped it has its own thing, as I'd like to drop s-t-b at some point15:07
pitti(and g-s-t)15:07
seb128mterry, didn't we say to uyse g-s-d?15:07
mterrypitti, actually, let me confirm.  i just wrote this, but even I'm confused15:07
mterryseb128, yeah, i think i was wrong about s-t-b15:08
seb128mterry, you backport g-s-d3 code for that iirc15:08
pittimpt: are you concerned about "temporary changes that get washed away next time you connect to the Internet?"?15:08
mterryseb128, pitti: yep, uses new g-s-d code15:08
pittimpt: i. e. would you deliberately want to set a wrong clock?15:08
mterryso same point though, we could modify g-s-d to configure ntpdate instead15:09
mptpitti, yes, I know some people who deliberately set their clock ten minutes early to make them more likely to get to things on time (for example)15:09
=== bjf is now known as bjf[afk]
mterryhah15:09
pittimpt: off-topic, but curious: does that actually work? I mean, your brain would know that the time is off..15:09
mptpitti, yeah, as long as you don't think about it too hard15:10
seb128doesn't work for me, I know how much my clock is off and adapt to that ;-)15:10
tedgpitti, Doesn't matter if it works, many people swear by it.  :-)15:12
* tedg just sets his clock to yesterday so that I can always say "I sent you that e-mail yesterday!"15:13
mptYou could still get the previous behavior of "I'll tell you what the time is now, but update it when I next go online" by setting the time manually then clicking the "Automatically..." radio button15:13
pittimpt: so ideally we would re-devote this radio button between "manual" (completely) and "automatic" which would be the default and mean ntpdate15:13
mptyes15:13
pittimterry: as a poor man's PK backend, you could call pkexec sed -i ... :)15:13
mterrypitti, makes sense to do it the Real way, that way time-admin gets benefit too15:14
mterrypitti, is this an Ubuntu thing, using ntpdate?15:15
mterrypitti, I'm curious if I should bother talking to upstream about such a change15:15
pittimterry: Debian as well at least; not sure about Fedora15:15
pittimterry: certainly worth investigating15:15
mterrypitti, OK.  Sounds like I should file a FFe for this too, eh?15:16
pittimterry: for changing what g-s-d backend configures? I guess might be better, yes15:16
pittiI'm happy to ack it, but at least there's a paper trail15:17
pittiwe'll need a bug report for the thing anyway (for documentation, upstream link, etc.)15:17
mptsooooo, the ntpd case15:19
=== zyga is now known as zyga-ill
mptmterry, pitti: If you do have ntpd installed, is there any GUI-worthy situation in which you'd want to use an Internet NTP server *instead of* the local NTP server?15:21
pittimpt: I don't think so; you want to configure the input sources for the local ntpd of course (that's what time-admin does), but aside from testing scenarios it's much quicker to use the local ntpd then15:22
mptmterry, pitti: ok, how about this:15:23
mptWhenever “Set the time:” is set to “Automatically from the Internet”, the “Time:” and “Date:” controls should be insensitive, and the time and date should be updated from the local NTP server (if there is one) or otherwise Internet NTP servers whenever you go online.15:24
mterrypitti, if ntpd is installed, does ntpdate use it?15:24
pittimpt: that sounds like a great way to fit all three methods15:24
mterrypitti, i.e. does ntpdate become a no-op?15:24
pittimterry: RTFS, hang on15:24
mterrympt, and "manual time" should actually do what it says on the tin, right?  and not reset when you connect to the Internet?15:25
pittimterry: apparently not, when I interpret /usr/sbin/ntpdate-debian correctly15:25
pittintpdate will always run if it is installed15:25
mptmterry, right15:26
mterrypitti, so you'd connect, ntpdate would use Internet servers, then in a sec ntpd would set local time from local server?  so you'd potentially have a time hiccup?15:26
pittimterry: yes, but usually at the ntpdate time15:26
mterrypitti, yeah.  that's a separate, tiny bug15:26
pittimterry: diff between ntpdate and local ntpd should be by and large 0, if you configured your ntpd15:27
mterrympt, ok, I understand what we want, am working on paperwork and code15:27
mptmterry, https://wiki.ubuntu.com/TimeAndDate?action=diff&rev2=24&rev1=2215:29
mterrypitti, /etc/default/ntpdate does not seem to have a conditional to control whether it runs?15:29
pittimterry: you can disable $NTPSERVERS15:30
pittilike, comment it out15:30
mterrypitti, hrm...  ok15:30
seb128mterry, btw not sure if you checked the open bugs but while you are at it there is a bug about the dialog not having a proper name and icon if you want to fix those ;-)15:33
seb128mterry, easy to notice in the unity launcher15:33
mterryseb128, it was missing a .desktop file, already fixed in trunk  ;)15:34
seb128mterry, great ;-)15:34
mptmterry and pitti, thanks for helping me understand all that.15:34
mterrypitti, https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/73489415:35
ubot2Launchpad bug 734894 in gnome-settings-daemon "FFe: DBus time API should control ntpdate, not ntpd" [Undecided,New]15:35
pittimterry: (will look in a bit, me@phone)15:44
pittiseb128, didrocks, mterry: do you have anything for gnome-control-center? otherwise I'll do an upload now15:55
didrockspitti: nothing particular for me15:55
seb128pitti, no15:55
mterrypitti, no15:55
pittigood, thanks15:56
GunnarHjpitti: Hello Martin, could you pls take a look at bug 636693? Currently the design of indicator-session and gdm-guest-session are contradictory in that the indicator menu item "Guest Session" precludes interactivity such as a zenity dialog in /etc/guest-session/prepare.sh. I for one fail to understand why Ted persists. Maybe you and he could talk about it author to author at some time or other?16:00
ubot2Launchpad bug 636693 in indicator-session "Premature lock when launching guest session" [Low,Confirmed] https://launchpad.net/bugs/63669316:00
mterryIs there a robust way to programmatically comment/uncomment lines in bash scripts?  I'd like to be able to run something like "comment VARIABLE /path/to/script".  I can do it with sed, but doesn't seem as fool-proof16:08
=== alecu is now known as alecu-lunch
mterrypitti, what are issues with installing ntpd by default?16:10
pittimterry: nonzero overhead (extra CPU power, battery drainage, network traffic, CD space) for very little benefit16:13
mterryk16:14
pittimterry: and formally it violates the "no open ports in default install" policy16:14
pittimterry: commented and asked skaet to review16:16
pittihey GunnarHj16:20
pittiGunnarHj: ok, tab opened for it; I'll have a look soon16:20
GunnarHjpitti: Good, then I 'rest my case' for now. :)16:22
seb128GunnarHj, still arguing with ted over this one?16:24
seb128GunnarHj, I've to admit I fail to see what you really want to get there16:24
mterrypitti, hrmm.  NTPSERVERS is ignored if /etc/ntp.conf exists and NTPDATE_USE_NTP_CONF=yes (which it does by default).  I may want to add an NTPDATE_ENABLED var16:26
pittimterry: that's correct16:27
GunnarHjseb128: Yes, unfortunately  I am. The use case that made me suggest it was a zenity dialog where you select the guest session language on-the-fly when launching the session.16:28
pittimterry: but /etc/ntp.conf doesn't exist by default, only if you install ntpd16:28
mterrypitti, yeah, but if you do have both installed, i still want to be able to disable ntp.conf16:28
mterryI mean, ntpdate16:28
pittimterry: I think we should do that differently -- if ntpdate-debian sees that ntpd is installed and enabled, it probably shouldn't run at all16:29
pittimterry: that would also avoid the "time jump" at connection time16:29
pitti(ntpd can gradually move the system clock to the right time)16:29
mterrypitti, OK16:30
mterrypitti, and you'd still prefer to (un)comment NTPSERVERS?  Seems so indirect16:32
mterryDoesn't really matter.  Not adding an NTPDATE_ENABLED is easier than adding it16:33
pittimterry: I have no strong opinion there, except that if we add a new var I'd like to see it getting added to Debian as well16:33
pittiother packages might read the file, too16:33
=== smspillaz is now known as smspillazzzz
mterrypitti, since NTPDATE_USE_NTP_CONF=yes is a thing, we probably don't have a "time jump" issue?  the existing ntpdate.ifup seems to explicitly stop ntpd and start it after, which implies there's a usecase.   What would be the usecase for running ntpdate if ntpd is on?16:42
GunnarHjseb128: Btw, there are more details about the background at http://ubuntuforums.org/showthread.php?t=156607816:43
pittimterry: the only thing that this does is to take the $NTPSERVERS configuration from /etc/ntp.conf instead of from /etc/default/ntpdate16:43
pittimterry: but ntpdate still gets called, so you would have a time jump16:43
mterrypitti, because ntpd gradually adjusts rather than all at once?16:44
pittimterry: hm, indeed; I think it's even more wrong to stop/restart ntpd, ntpdate should shut up if it's running16:44
pittimterry: yes16:44
pittimterry: and for the latter, because ntpd updates regularly instead of just at if-up.d time16:44
mterrypitti, ok, will make the patch and throw to debian too16:44
pitticool, thanks!16:45
=== bjf[afk] is now known as bjf
seb128pitti, there is a new gobject-introspection which can be synced from Debian, should we get it?16:52
* pitti checks upstream changelog16:52
pittiseb128: yes, only one bug fix, can be synced16:53
seb128pitti, ok, will do16:53
pittimerci16:53
seb128de rien16:53
bcurtiswxGood day all.16:55
seb128hey bcurtiswx16:56
bcurtiswxHey seb12816:56
bcurtiswxkenvandine: Do you use an colloquy bouncer?16:57
kenvandinebcurtiswx, never heard of it16:57
kenvandineisn't colloquy a mac thing?16:57
bcurtiswxkenvandine: Yes I'm using my iPad.16:57
kenvandinenope, not using it16:58
bcurtiswxDo you have an irc bouncer tho?16:58
kenvandinei haven't owned an apple product in many years16:58
kenvandineznc16:58
bcurtiswxOk may I pm you?16:58
kenvandinesure16:58
=== alecu-lunch is now known as alecu
evilvishpitti: hi, i would like to drop the scour dependency for humanity.. :)17:14
=== tkamppeter_ is now known as tkamppeter
tkamppeterjasoncwarner,17:23
tkamppeterjasoncwarner, hi17:23
seb128tedg, kenvandine, mterry: bug #734616 might be worth investigating17:49
ubot2Launchpad bug 734616 in lernid "lernid hangs closing the "about" window or the preferences dialog window" [Undecided,New] https://launchpad.net/bugs/73461617:49
tedgIs Lernid python?17:50
seb128tedg, yes17:51
tedgHmm, I wonder if that's the same bug as that music player one.17:51
seb128tedg, it's the application jono wrote for UDW sessions17:52
tedgAh, so you're saying people might use it here in a bit? :)17:52
seb128;-)17:53
seb128tedg, feel free to dup it if it's the same issue17:53
tedgI think it might be a dup of bug 717162  -- but I have no idea how to debug them.17:53
ubot2Launchpad bug 717162 in unity-2d "quodlibet UI freeze in Unity after accessing its menu" [High,Confirmed] https://launchpad.net/bugs/71716217:53
tedgI e-mailed barry, as I'm sure he's at PyCon to see if he can help.17:54
tedgIt seems to be where the Python meets the C.  Which is a very scary place.17:54
kenvandinetedg, that does seem similar17:54
seb128hehe17:54
kenvandinetedg, all the other menus in quodlibet are working17:54
evilvishseb128: lernid team is not actively developing it, it has a lot of problems on its own.. so it not sure it is worth spending time on it ;)17:54
tedgkenvandine, I think it must be something with launching the dialog.17:55
kenvandinesomething with the response?17:56
kenvandinein quodlibet it happens with the file chooser17:56
seb128evilvish, oh, it's not about trying to stabilize lernid but rather catching indicator-appmenu issues because they can impact on other applications17:56
evilvishah, OK :)17:56
tedgkenvandine, Yeah, there may be stacked mainloops there.17:56
seb128kenvandine, the guy thinks "this is related to the use of dialog.run() in a signal handler from the indicator-appmenu"17:56
kenvandineseb128, ok, that is basically what happens in quodlibet17:57
seb128not sure what made him wrote that though17:57
tedgThe odd part there is that we're processing those events in the idle loop.  So they should be safer if nothing else.17:59
tedgI wonder if Python does something funky to lock the main loop from GUI events normally.17:59
seb128tedg, could it be an issue if those dialogs were transient from the caller?18:00
mterrypitti, just an update on ntpdate.ifup to make sure I'm not going off the deep end: because of the complexity of determining if ntpd is a thing (could be removed, not purged, so ntp.conf is still there, so NTPSERVERS is ignored), I'm thinking it'd be easiest to just move ifup.d/ntpdate to ifup.d/ntpdate.disabled.  How does that sound?18:00
tedgseb128, I imagine they're created, run, and destroyed.  That's how most GTK programs do it.18:01
pittimterry: ah, I actually thought about "status ntpd", but renaming away the ifup.d script sounds fine as well18:02
pittimterry: ".disabled" works in if-up.d? I. e. anything with a period is ignored?18:03
mterrypitti, yeah, it uses run-parts18:03
pittievilvish: oh, why? (btw, will do another upload of scour with a bunch of bug fixes)18:03
evilvishpitti: i just subscribed you to a bug report..18:04
evilvishi've*18:04
evilvishpitti: in short, it is not stable enough to be used for icon themes. it changes the design when the whole purpose of the icon theme is just that18:05
evilvishpitti: or what we could do is, reduce the fuzzy check to depend on the package, allow icon themes to set the check limit to something lower so that no icon is visibly modified …18:05
pittievilvish: ah, I'll reply in the bug report, there are some confusions there18:06
evilvishpitti: cool.., but 0.5.35 is not a perfect build either ;)18:06
kenvandinetedg, quodlibet did use a .run for the dialog, but it does get past that18:06
evilvisherr, 0.5.3.618:06
kenvandineit gets the response event and moves on18:06
kenvandinethe code after that is executed18:06
kenvandineit is like the existing main loop is hung though18:07
tedgkenvandine, Can you put that on the bug... I think this one is going to take a bunch of people looking at it.18:08
kenvandineyeah18:08
kenvandinewill do18:08
kenvandineit really looked like it did the right thing... just stopped doing it18:08
* kenvandine goes to comment18:08
pittievilvish: replied18:12
seb128mterry, you won bug #733688 (crashes in sanitize_label_text (label=0x0) which seems code you changed recently?)18:13
ubot2Launchpad bug 733688 in libdbusmenu "bluetooth-applet crashed with SIGSEGV in g_cclosure_marshal_VOID__PARAM()" [Medium,New] https://launchpad.net/bugs/73368818:13
mterry:-/  yeah, seems like me18:13
seb128mterry, but feel free to bounce to ted if you are busy on other things18:13
mterryseb128, I'm busy right this second, but I can get to it this week18:13
mterryor is this a frequent crasher?18:13
seb128mterry, ok, so I let it on your list18:13
seb128ted is of 2 days at the end of the week and seems to have quite to do already18:14
seb128mterry, no, just spotted 1 bug about it18:14
seb128but it seems an obvious case of non handling of null value18:14
seb128mterry, no hurry, whenever you have time before beta will do, I will ping you if it starts being a frequent crasher18:16
evilvishpitti: i've not noticed any problems with editing, only 18 files which had been converted properly.. i dint realize its a new option we in inkscape18:29
evilvishhad not* been18:29
pittievilvish: but that's what upstream specifically warns not to do..18:31
evilvishpitti: yea, but that is not our working branch.. just a release branch.. btw, I'm OK with a 0.0% check:)18:31
pittievilvish: ok, I'm about to do a scour upload anyway (probably need to finish tomorrow morning, need to run out for taekwondo)18:32
pittievilvish: I'll just lower the default threshold then18:32
pittigood night everyone!18:33
evilvishpitti: neat! and then we have have a re-build for the branch.. :)18:33
pittievilvish: yep, will do18:33
evilvishnite pitti :)18:33
pittievilvish: I still recommend to revert the committed scour'ed packages18:33
pittithe source package should have the acutal source files IMHO18:33
pittis/packages/images/18:34
pittiand it would also be a nuisance to keep this up to date manually18:34
evilvishpitti: yea.. but reverting is not needed IMO, there really is no problem with the images though ;)18:35
evilvishpitti: just that I dont need to do it manually later ;)18:35
pittiyou checked them all in inkscape and made sure they didn't lose any annotation, etc.? because that's what scour does..18:35
evilvishpitti: yup.. i checked *several* files..18:35
evilvishi did it to compare if the colors and lines and stuff matched18:36
=== lan3y is now known as Laney
mterry_pitti, hello.  So now that I fixed gnome-settings-daemon to configure ntpdate as well, bug 732757 could use an official pat on the back19:44
ubot2Launchpad bug 732757 in gnome-system-tools "FFe: time-admin and indicator-datetime-preferences shouldn't both be installed by default" [Undecided,New] https://launchpad.net/bugs/73275719:44
rickspencer3kenvandine, hey, you mentioned somehting once about getting "streams" from gwibber, like I could get the stream of images21:53
rickspencer3do you remember what I'm talking about, and is there any documentation or samples for that?21:54
rickspencer3robert_ancell, RAOF either of you guyses clueful about Gstreamer, specifically,  camerabin?22:07
robert_ancellrickspencer3, haven't played with camerabin yet22:08
rickspencer3it stopped working for me this weekend, and I don't know who should look at it :?22:08
movican someone using networkmanager and wifi please stand up? i need someone to generate a file for me22:09
rickspencer3robert_ancell, it's bug #734600, if you know someone who can look at it22:10
ubot2Launchpad bug 734600 in gst-plugins-bad0.10 "camerabin throws an error in gst_camerabin_capture_start gstcamerabin.c(4036)" [Undecided,New] https://launchpad.net/bugs/73460022:10
rickspencer3movi, what do you need?22:10
movirickspencer3, i need a keyfile. i'm building a GUI-less server, and want networkmanager to handle the wifi. but the only "right" way to create that file is to let nm-applet create it. the documentation sucks22:11
movirickspencer3, basically go to the nm-applet connection editor, create a connection to a wifi network which is protected by a WPA/WPA2 security scheme, and select "avaible to all users". then hopefully it will show up as a file in /etc/NetwormManager/system-connections22:12
rickspencer3movi, I don't have any protected AP to connect to, sorry22:12
movirickspencer3, that doesn't really matter. you don't need to connect to it at all :)22:12
movijust create it in the connectino manager.22:13
rickspencer3movi, do you want me to pick WPA/WPA2 personal or enterprise?22:17
movipersonal22:19
movican you even tick the "avaible for all users?22:19
rickspencer3yeah22:19
* rickspencer3 peruses resulting file for pii22:19
movi?22:20
rickspencer3Personally Identifying Information22:20
moviah22:20
rickspencer3in other words, making sure I'm not being socially engineered ;)22:20
moviit'll have a mac adress probably22:20
movinope, you're not. but stay vigilant, that's a good habit :)22:20
=== ogra_ is now known as ogra
rickspencer3lol22:21
RAOFrickspencer3: Heh.  That's a most unhelpful error message for the camerabin :)22:21
rickspencer3but of course if I was being socially engineered, that's exactly what I would expect you to say movi :)22:21
rickspencer3movi, http://paste.ubuntu.com/580332/ , I have no idea why they couldn't just document that, seems enough to make22:21
rickspencer3RAOF, uh, yah22:22
moviah, thank you22:22
rickspencer3RAOF, I noticed cheese works22:22
movihopefully this will work22:22
rickspencer3good luck movi22:22
movioh, just one more question22:22
movinetwork-manager-0.8.x ?22:22
rickspencer3ah geez22:23
rickspencer3latest and greatest in natty, I just dist-upgraded22:23
rickspencer3RAOF, any ideas about camerabin, there?22:23
moviwill i get kbanned if i mention i'm doing this for gentoo :> ?22:23
moviand hence the reason why i'm asking22:24
AmaranthWow when you setup a NM system-wide connection it stores your wifi password in plaintext?22:24
AmaranthScrew that, stick with per-user where it gets stored in your keyring22:24
rickspencer3sweet, huh22:24
movirickspencer3, quick question, was ssid= field really plaintext? or was it ascii encoded like 124;112; ..22:24
rickspencer3I dunno man22:24
movihow did you name the AP ?22:25
moviTEST_AP ?22:25
rickspencer3yeah22:25
moviok, thanks22:25
rickspencer3it just looks like a simple text file to me22:25
moviAmaranth, doesn't work like that, no-one logs into the machine locally22:25
movialso, it's not plaintext, it has to be hashed22:25
movihttp://projects.gnome.org/NetworkManager/developers/settings-spec-08.html22:26
Amaranthmovi: err, it's 12345678 in his pastebin22:26
movithis is the documentation for the keyfile. welcoming no?22:26
Amaranthvery clearly plaintext22:26
rickspencer3looks like POT to me22:26
moviAmaranth, i've figured that's what he changed not to be incriminated?22:26
moviPOT ?22:26
rickspencer3Plain Old Text22:27
rickspencer3movi, no, the password I sepcified was 1234567822:27
AmaranthSurely he just put in a dummy psk when creating the connection, no need to change that22:27
rickspencer3nah, that's my normal password22:27
rickspencer3I use it for everything22:27
movi:>22:27
rickspencer3actually, I shouldn't have said that, even jest, now everyone knows one password I don't use, one step closer to hacking me22:27
rickspencer3movi, seriously, what I pastebinned is what NM made for me22:28
Amaranthmovi: networkmanager itself hashes it for auth with the AP but it is stored plaintext22:28
AmaranthOtherwise you couldn't edit the connection afterward and have it show you the password22:28
movimm22:32
moviit still doesn't work22:32
movinope, it works22:35
movijust needed to chmod 600 the keyflie22:35
movithanks rickspencer322:35
rickspencer3sweet22:35
moviand yeah, it may be insecure, but it's the closes way to having an "intelligent" connectino manager22:35
rickspencer3well, it's not really "insecure"22:35
movione which won't fall over when the AP is unavaible for a couple of secconds22:36
rickspencer3I mean, you need su to see it at all22:36
rickspencer3and also, it's a password for a public AP, which has presumedly been passed out to all kinds of people22:36
rickspencer3if someone can read that file, the fact that they can log onto the AP easily is probably the least of yoru concerns ;)22:36
moviumm, what do you mean by public ap?22:37
movii mean it's visible, but i'd like to think it's my own, and nobody exept me hass the PW ;)22:38
movianyway, that's done. now to fix the xbmc/hal mess and finally squish that php+ipv6 bug22:42
movithanks again guys :)22:43

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