=== asac_ is now known as asac [08:45] good morning everyone [08:46] hello chrisccoulson [08:46] morning [08:46] hey seb128 [08:46] hey chrisccoulson and seb128 [08:46] hey huats! [08:46] lut huats [08:46] seb128 - it's your last day in dublin today isn't it? [08:47] right [08:50] seb128 - i started working on porting gconf to polkit-1 this week with the patch from the bugzilla, but it needs quite a bit of refactoring. the patch doesn't apply to our version [08:50] what changes do we have there? [08:50] there's loads of whitespace changes which makes it a real pain :-/ [08:51] they probably just stripped trailing spaces [08:51] should be easy enough to do in the changeset too [08:51] seb128 - i'm not sure what changes we have. the patch was probably targetted at a slightly different version [08:52] i shall hopefully finish that today though [08:54] Good morning [08:54] hey pitti [08:55] pitti, hey long time not seen! [08:56] * pitti pokes seb128 into the side [08:56] utch [08:56] the fun thing is that I can actually do this! :-) [08:56] * seb128 moves chair a bit further away [09:39] pitti: hi... latest apport update works but shows this error.> /usr/share/apport/apport-gtk:204: GtkWarning: Theme directory devices/22 of theme humanity has no size field self.w('window_information_collection').show() [09:39] but i'm not using humanity theme! [09:48] mac_v: hmm [09:48] jcastro, hey, pitivi should be fixed in karmic [09:48] jcastro, can you try and let me know if it works for you too? [09:48] mac_v: doesn't sound like an apport bug, though [09:49] pitti: this started only after the recent update , hence it thought it might be apport bug... [09:49] mac_v: do you have the humanity theme installed? [09:49] yup [09:50] i have it installed , but not using it [09:51] pitti: oh..wait a sec... humanity is actually an icon theme , i dont think there is a humanity gtk theme [09:53] hi all [09:55] oh, another andreas [09:55] oops ... sorry :) [09:55] why sorry? it's a great name! :) [09:55] but that's my name ;) [09:56] yes [09:57] can you help me with a problem? [09:57] pitti: nevermind... i found the cause , HUmanity is one of the Inherits in the icons theme i'm using... thanks for clearing that up [10:02] seb128, it works, thanks! [10:02] jcastro, cool [10:02] you're welcome [10:04] andreas_, don't ask if you can ask, just ask [10:08] sorry, I was on the phone ... ok, here my problem: [10:09] I am very new to ubuntu and I have some strange effects ... graphics problems I think [10:10] for example: when I play a game with cards and I start new, then it dosn't clear the area, where you play [10:11] or I play a game battleship, and I played one round and start new, then it doesn't clear the area, where I place my ships ... [10:11] it leaves fragments of the game before [10:12] another effect: when I play a game (like openttd) in fullscreen mode, it sometimes switches back to windowed mode, without me doing anything [10:13] could it be that I have the wrong hardware driver for my nvidia card? [10:25] seb128 - re bug 409621 - the symbol isn't there because it's in a module that isn't loaded yet (i think) [10:25] Launchpad bug 409621 in gnome-settings-daemon "The program 'gnome-settings-daemon' received an X Window System error. During on a FreeNX server suring a session." [Low,Incomplete] https://launchpad.net/bugs/409621 [10:25] not sure how to work around that one [10:27] chrisccoulson, usually gdb figures that just fine when loading the code which has the symbol [10:28] tseliot, why do we have the option to turn the touchpad off? Is it just to cater for people who have a mouse connected, or is there another reason? [10:29] seb128 - ah, i didn't know that [10:29] mpt didn't the option to turn the touchpad off just get removed? [10:29] it was removed from gnome-settings-daemon in the last version [10:29] mpt, some people don't like touchpads because they touch those by accidents for example [10:30] chrisccoulson, if that was in Karmic, I don't know yet [10:30] seb128, then how do they move the pointer? [10:30] andreas_, hard to tell. Try searching launchpad and see if anyone had similar problems [10:30] mpt - yes, that's in karmic. the option to turn the touchpad off is gone from g-c-c and is not supported in gnome-settings-daemon either [10:30] mpt: are you referring to the Synaptics Off property? [10:30] andreas_, or ask in #ubuntu [10:30] mpt, some laptop have this small red or blue round in the middle of the keyboard too [10:30] tseliot, yes [10:30] mpt, not sure how you call it [10:30] seb128, ah, good point [10:31] tseliot, can we tell whether the hardware has an alternative pointing device built in? [10:31] mpt: that can be used to enable/disable the touchpad or to disable only tapping and scrolling [10:31] ok, thanks [10:31] tseliot, yes, I was more interested in why we started exposing it [10:32] (iirc someone quit Ubuntu development over us exposing it, so we must have done it for some important reason:-) [10:32] mpt: yes, I have access to the list of input devices [10:33] mpt: some people prefer to plug in a mouse. However, when the put their hands on their laptop, it can happen that they accidentally touch the touchpad and move the cursor or tap [10:34] only some touchpads have a physical button to disable them [10:34] tseliot, the reason I ask is that I was wondering if we could replace the option for turning off the touchpad with an option for turning it off whenever (and only whenever) a mouse is connected [10:35] mpt: ok, I see your point now [10:35] but TrackPoints etc make the issue a bit more complicated [10:35] mpt: I was about to point that out ;) [10:40] mpt: furthermore I'm not sure we can identify the actual mouse. For example on my system I have: [10:40] "Macintosh mouse button emulation" id=3 [XExtensionPointer] [10:40] "Razer Razer Salmosa" id=4 [XExtensionPointer] [10:41] where the former is not a real device [10:41] and the latter is a mouse [11:14] asac: the prerm script in flashplugin-nonfree in karmic is busted. I'll see how to workaround it. I would like to show you my changes to make sure you're okay with them. [11:16] mdeslaur: remove the --remove-all part of the alternatives [11:16] asac: I added --quiet to it instead [11:16] mdeslaur: someone should really redo all the maintainer scripts [11:16] at some point [11:16] asac: and I need to add the failed-upgrade argument i think [11:17] mdeslaur: so the features we need is the xulrunner-addons alternative, but also we need the plugin to be installed in /usr/share/ubufox/plugins/ [11:17] mdeslaur: yes. add failed-upgrade for recovery ... but only for lower versions [11:18] asac: oh, there's a way to specify failed-upgrade only for specific versions? [11:18] * mdeslaur goes to look it up [11:19] hey asac, on the subject of --remove-all - you asked me a while ago to see if it was still needed when i did the change to mozilla-plugins-gnash [11:19] i had a play around a while ago, and i haven't found any issues without the --remove-all [11:19] chrisccoulson: yeah. i double checked twice and we should really remove it everywhere [11:19] Laney, hey, thanks for the goocanvas upload [11:20] asac - ok, thats good then :) [11:21] asac - i'll close the gnash bug then. is there any way of finding out which existing scripts call --remove-all? [11:21] seb128: no worries [11:23] Laney, want to do some REVUing too? ;-) [11:23] seb128: maybe later [11:23] remind me about 6? [11:23] Laney, ok [11:23] 7 your time... [11:23] Laney, I might be not online after work [11:23] my ubuntu time has been sucked up by haskell lately [11:23] alright [11:23] still in Dublin for sprint [11:23] Laney, http://revu.ubuntuwire.com/details.py?package=mago [11:23] you can email me the list if you want [11:23] i'll probably forget otherwise [11:24] is Ireland any less wet than England is currently? [11:25] Laney, it has been sunny since wednesday [11:25] nice [11:25] we got some sun yesterday morning [11:25] but since about 3pm it has rained constantly [11:25] Laney - did you not get any sun this morning? [11:25] nope [11:25] i had to wear my shades driving to work this morning:) [11:25] :O [11:25] you're only down the road too! [11:25] no fair [11:25] but i suspect it will start raining by the time i leave work at lunchtime ;) [11:26] if it's not already raining - it might be now, but i wouldn't know as i can't see outside [11:26] Laney, that one is the only one to review if you want to note it [11:26] ok [11:27] chrisccoulson, gnome-session is using gnome-power-manager for suspend and hibernate right? [11:27] seb128 - gnome-session uses dk-power directly now [11:27] chrisccoulson, ie if suspend is not available there the button should not be listed [11:28] oh ok [11:28] no wonder than tweaking the can_suspend doesn't work [11:28] AFAIK, g-p-m does not expose any methods for suspend/hibernate anymore [11:28] ok good, thanks [11:28] I didn't look into this new world order much yet [11:29] seb128 - i'm seeing some people reporting that their screen is not locked after resuming from suspend - do you know about that? i'm wondering if thats a side-effect of the new architecture [11:29] chrisccoulson: good question. most likely a zgrep on an archive mirror or so [11:29] I get the bug there [11:30] but dunno why it's happening yet, I didn't debug [11:30] g-p-m used to lock the screen when gnome-session used that, but now it is bypassed [11:30] i can't test it here because my machine will never resume from suspend or hibernate [11:30] well it works for some people [11:30] so that's weird [11:30] yeah, that is a bit strange [11:30] i wonder how the screen is supposed to get locked now? [11:31] do you know which part is responsive for screen locking now? [11:31] ie gnome-session? [11:31] ok [11:31] same question ;-) [11:31] vuntz, ^ ? [11:31] seb128 - gnome-session doesn't do it. and i suspect dk-power doesn't do it either [11:32] so what does it? [11:32] seb128: ask hughsie [11:32] hm, it still locks the screen for me.. [11:32] vuntz, ok thanks [11:32] I'll test killing g-p-m if it's that [11:32] vuntz, do you know if GNOME will switch to webkit everywhere before 2.28? [11:32] the screen locking part must be magic:) [11:32] vuntz, it has been accepted but recent tarballs are still using gecko [11:33] chrisccoulson: right, I use the lid [11:33] seb128: the goal is to do that, yes [11:33] so g-p-m locks the screen [11:33] seb128: except evo, I guess [11:33] vuntz, ok thanks [11:34] with usign the power button, g-p-m isn't involved [11:34] seb128: there's yelp. Anything else? [11:34] and nothing locks the screen [11:34] vuntz, not that I know that's the one asac was asking about [11:34] so shaun is okay as long as there's a patch and I believe that kov was working on that [11:35] I though there was a webkit version in svn for a while [11:35] yeah, but it was not perfect, I guess [11:35] ok thanks [11:35] asac, ^ [11:35] ArneGoetje: jaunty ppa->proposed copying in progress [11:36] pitti - thanks. i suspected that might be the case now [11:37] tseliot, I'm having trouble understanding the difference between the "Synaptics Tap Action" options (TapButtonX) and the "Synaptics Click Action" options (ClickFingerX). [11:37] tseliot, can synaptics tell how many fingers you use to press the actual button, as opposed to the touchpad itself? Otherwise, what's the difference? [11:39] pitti - what handles your power button now then? i thought that was still g-p-m? [11:39] (i really should learn how all this new stuff works) [11:39] chrisccoulson: yes, it should actually [11:39] seb128: thanks. i uploaded a 1.9.1 port for now ... so its not really urgent. i can just POSTPONE yelp to webkit work item [11:40] chrisccoulson: after the recent migration of a lot of stuff to gnome-session I don't quite know either :/ [11:40] so, the power button case should still work, but i would expect the "suspend from the session dialog" case to not work anymore [11:40] asac, POSTPONE it we don't track it in an ubuntu specific way [11:40] asac, if upstream gets it good otherwise that's ok too [11:40] right, and system -> switch off probably won't either [11:40] pitti, right that's what I use [11:40] pitti - that's what i meant [11:41] so gnome-session probably needs to tell the screensaver to lock now, which is what g-p-m used to do before [11:41] seb128: yes. its in the ffox35 transition spec. done [11:41] chrisccoulson: please ask hughsie :-) [11:42] vuntz - ok, i'll do that [11:42] do you know where i can find him? [11:42] chrisccoulson, #gnome-hackers [11:42] on the GNOME irc [11:42] seb128 - thanks:) [11:43] i'll ask him in a bit - i really should be doing work right now :-/ [11:49] mpt: "Synaptics Tap Action" affects the way tapping works (e.g. what happens if I tap with one finger?) while "Synaptics Click Action" only affects left clicks [11:49] mpt: and left click events can be triggered by taps too [11:50] mpt: "Synaptics Click Action" is not about physical button (and the documentation should be clearer about this) [11:50] s/button/buttons/ [11:51] tseliot, so when ClickFinger1 is defined as "Which mouse button is reported when left-clicking with one finger", what exactly does it mean by "left" and "clicking"? [11:53] mpt: ok, good question. I'll give you a concrete example [11:53] mpt: let's assume that my touchpad doesn't have physical button and that therefore I can only tap on the touchpad [11:54] Like a 2009 MacBook. ok. [11:54] mpt: and I want to be able to right click and middle click [11:55] mpt: so I can tell the driver that it should interpret a tap performed with one finger as a left click (ClickFinger1 = 1) [11:56] mpt: a tap performed with two fingers as a right click (ClickFinger2 = 3) [11:56] (that's what I want to be able to turn on:-) [11:56] mpt: and a tap performed with three fingers as a middle click (ClickFinger3 = 2) [11:57] ok [11:57] :-) [11:57] So how do the TapButtonN options differ then? [11:58] If I set TapButton2 = 3 instead of ClickFinger2 = 3, what happens? [12:05] mpt: I think it would be the same in that case [12:05] tseliot: ^^ [12:05] tseliot, so that brings me back to the question of what's the difference between the options :-) [12:06] mpt: they overlap a bit, I know [12:07] hm, I guess that can be up to you to figure out [12:08] I'm currently thinking of showing an option to let you secondary click by tapping with two fingers, so you can work out which parameter you need to twiddle to make that work [12:10] mpt: ok, maybe add that option and I'll deal with that [12:12] mpt: the implementation will be my problem, not yours ;) [12:12] I just wanted to make sure that they weren't importantly different things that needed to be presented separately [12:13] mac_v, no the indicator applet can't be removed from session start [12:13] mac_v, it's a gnome-panel object in the config there [12:13] mac_v, itg's not a notification area icon [12:14] seb128: ??how come i removed it while in jaunty! , let me test it again [12:15] mpt: oh, wait, it looks like you can lay two fingers on the touchpad and then press the left physical button (if you have one) and have a different behaviour (thanks to ClickFinger) [12:15] mpt: have a look at this thread: http://www.uluga.ubuntuforums.org/showthread.php?p=6821807 [12:15] didrocks: if you want to discuss the mime stuff with the openSUSE guy, I can introduce you [12:15] * tseliot -> lunch [12:15] ahhh [12:16] mpt: see comment 4 [12:16] mpt: talk you later [12:16] mac_v, the item in the list is the script to add the panel on upgrade not the actual applet [12:16] tseliot, so by "left-clicking with one finger" etc, the man page really means "clicking on the left button with one finger resting on the touchpad" etc [12:17] mpt: mac users should have this, I guess [12:18] * tseliot -> lunch (seriously now) [12:18] asac: I've put flash for karmic on chinstrap in ~mdeslaur/flash-karmic. Could you take a look at the debdiff and sponsor it for upload, please? [12:18] asac: I added ubufox, but xulrunner-addons was already there [12:21] mdeslaur: let me check [12:24] seb128: i just did... removed it from the startup app [12:38] Laney - it's still sunny here ;) [12:51] awe: http://people.canonical.com/~asac/tmp/NMAppletScreenshot.png [12:56] mdeslaur: i am not sure we need to rmeove the alternatives on failed upgrade. have you checked what happens? also we need to explicitly exit 0 in failed-upgrade afaik [12:58] mdeslaur: the ubufox thing explicitly must not be an alternative [12:58] mdeslaur: background: ubufox provides a way to switch between multiple plugins oyu have [12:58] so the idea is keep the xulrunner-addons alternative for folks that dont use ubufox, but have variants in the ubufox/plugins directory [13:06] pitti - just looking at bug 410241. i already removed /usr/share/gnome-pkg-tools/1/rules/patch-translations.mk from debian/rules in gnome-panel, but it seems that debian/po-up/patches.pot is still included in the translations [13:06] Launchpad bug 410241 in gnome-session "Merge debian/po-up/patches.pot templates into the main template" [Low,In progress] https://launchpad.net/bugs/410241 [13:06] should i just remove debian/po-up from the source? [13:06] (and obviously make sure that any additional translatable files are in po/POTFILES.in) [13:07] asac: oh! I thought ubufox was an unbranded browser..let me check it out [13:10] chrisccoulson: sure, soudns good [13:11] pitti - cool, i'll do that then [13:18] asac: hey. You might have some opinion on http://bugzilla.gnome.org/show_bug.cgi?id=580903 ? [13:19] Gnome bug 580903 in Backend:Webkit "No way to change user agent for distributors" [Normal,Unconfirmed] [13:25] mpt: I'm back. If you have other questions, I'm here [14:37] mpt: hi... about keeping the indicator-applet icon always visible> it seems the user testing showed users always kept clicking the icon expecting to view their mail, but isnt the fault because the icon is a envelope even when an IM client is open? wouldnt it confuse users even more when the icon is always present even when no app is run?! also the panel will now have 2 envelopes always present! 1 from the evolution mail client and 1 from indica [14:37] tor applet! [14:37] mpt: instead of keeping it always visible , how about using the icon of a speech bubble for when IM clients are alone open and an envelope when mail client is opened in addition ? or a better common icon for the indicator applet? [14:45] mac_v, I think it would be better to use the same icon regardless of message type, though I agree it's a problem to use an envelope for IM. [14:45] Anyway, the current icon is under the direction of sabdfl. [14:46] If you have alternative suggestions you could propose them to djsiegel [14:46] hehe... what can we say then! ;p [14:46] who has been discussing with sabdfl the design of the menu. [14:46] mpt: alternative icons? or ? [14:47] Alternative icon design, or behavior, yes [14:47] brb [14:47] oh ok... thanx [14:51] vuntz: how does the user agent string look like after that vendor change? [14:51] ffox has: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.0.13) Gecko/2009080311 Ubuntu/9.10 (karmic) Firefox/3.0.13 [14:52] ffox uses the following prefs: [14:53] general.useragent.vendor=Ubuntu [14:53] general.useragent.vendorComment=karmic [14:53] general.useragent.vendorSub=9.10 [14:56] asac: Mozilla/5.0 (X11; U; Linux i686; fr-fr) AppleWebKit/531.2+ (KHTML, like Gecko) Safari/531.2+ Epiphany/2.27.90 SUSE/2.27.90 [14:56] where "SUSE/2.27.90" comes from the file [14:56] "Mozilla/5.0 (X11; U; Linux i686; fr-fr) AppleWebKit/531.2+ (KHTML, like Gecko) Safari/531.2+" comes from webkit [14:56] and "Epiphany/2.27.90" comes from epiphany [14:57] we can certainly change the syntax in the file to put vendor, vendorSub and vendorComment [15:05] vuntz: i would suggest to use the same grammar as ffox ... e.g. dont appent SUSE/... ... but put it in between Safari/ and Epiphany [15:05] also allow a codename like the karmic thing [15:06] the specification is kind of vague and we already have lots of issues with websites not parsing things properly [15:06] Riddell: bug 402878 [15:06] Launchpad bug 402878 in kubuntu-netbook-default-settings "Folderview (with install icon) missing for live desktop" [High,Triaged] https://launchpad.net/bugs/402878 [15:15] mac_v: i don't want category or type icons [15:15] i.e. no envelopes and chat signs [15:15] the only icons should denote real content, specifically *users* [15:16] sabdfl: how about > http://dl.getdropbox.com/u/1325768/icon.gif < something like this but a fuller shape? such an icon seems lively , not sure if its OK for you [15:16] mac_v: for? [15:16] for the indicator applet [15:17] sabdfl: or a simple human bust as the icon? [15:18] asac: so, where do you select/enable/modify the plugins in ubufox? [15:21] mdeslaur: go to video.google.com ... then the tools -> manage content plugins is enabled [15:21] mdeslaur: if you have mozilla-plugin-gnash installed and flashplugin-installer you should be able to choose among them [15:21] you can also search for more there [15:21] problem previously was that if all alternatives wer pointing to gnash it wouldnt see flashplugin-installer anymore [15:22] so you can test if it works by updating all flashplugin alternatives to not point to flashplugin-installer [15:22] i looked at it once and it feeled a bit tricky to do that because of the different code paths for nspluginwrapper vs. not-nspluginwrapper in postinst [15:22] felt like one should refactor the script to do that better [15:23] mdeslaur: the manage content plugins is only enabled if you have a flash thing on the current site. also it doesnt work with flashblock [15:23] but i have fixing that for karmic on my list [15:25] today image has the sound muted and can't be unmuted [15:31] asac, the protected-network icons look ugly scattered horizontally like that [15:31] asac, perhaps they could go to the left of the network name instead [15:35] mpt, wouldn't that create a uneven left line? especially in the case where only unencrypted networks show up [15:36] or, well, whatever that icon symbolizes [15:36] andreasn, it would, but I don't think that would be quite as bad :-) [15:37] indeed, the only other option I can think of would be to align all the other with the most rightmost secure icon [15:38] or it could use text [15:38] * andreasn hides [15:38] One way of avoiding the uneven left line would be to have a (less prominent) icon for open networks [15:39] a smiley? for people driving around your house and roaming for open networks [15:40] that they can do nasty things to [15:40] Driving around my house? :-) [15:40] maybe it's just my friends that does that [15:41] * mpt widens his hallways to give the cars room, and points andreasn to http://www.schneier.com/blog/archives/2008/01/my_open_wireles.html [15:43] oh [15:43] still, that icon is quite cryptic [15:44] yes, it is [15:44] Robert Ancell and I were surprised yesterday that there doesn't seem to be a stock pair of open and closed padlock icons in Gnome [15:45] it used to have that, but then we changed the metaphor to a shield (for reasons I can't really remember) [15:46] and the epiphany devs didn't like that and started to ship their own padlocks icons [15:46] dobey, do you remember why we did that? [15:47] andreasn, yeah, since about 1995 Web browsers haven't been able to get away with using anything other than a padlock for encrypted sites [15:47] andreasn: because lock != secure [15:48] "lock_open" and "lock_closed" are meaningless [15:48] So even if another icon made more sense, using it would be incompatible with the Web. [15:49] firefox use some more hints these day, and well, a lock too [15:49] because the web is so compatibile with everything in the desktop... [15:49] firefox uses a lock because it's always used a lock [15:49] the lock itself is pretty much entirely useless and without any real meaning [15:50] * mpt goes back to work, not interested in conversations that are 15 years out of date :-P [15:54] asac: okay, updated source in ~mdeslaur/flash-karmic [15:54] asac: upgrade tested, and ubufox integration works [16:06] mdeslaur: upgrade with failed-upgrade tested? nice. [16:06] asac: yep [16:16] mpt: fight club + internet == http://identi.ca/notice/7674880 [16:17] dobey, sure would be nice if every Web site was encrypted as a matter of course [16:18] mpt: yeah. of course, a bunch of web sites put lock icons inside the page, and they tell you to look for that, and not the browser's lock icon [16:18] which only adds to the fun of trying to tell the user wtf is going on with their data :) [16:19] Usually they don't do that intentionally (because phishing sites could use in-page padlocks just as easily) [16:20] except for the daft in-page Visa/Verisign/etc verification pages, which are just begging to be faked [16:20] bank of america does it [16:20] pages->badges [16:20] i got a pretty awesome phishing mail the other day [16:21] Encrypting by default would also reuire a caching change in browsers to avoid slowing down the entire Web too much [16:24] mdeslaur: so what does happen for failed-upstream ... e.g. it removes the new alternative stuff and then it runs the old postinstall and adds the right stuff again? [16:24] err failed-upgrade ;) [16:25] mpt: yeah, though it probably wouldn't actually solve the problems [16:25] hmmmm policykit is screwing around with my suspend =\ [16:25] mpt: encrypted phishing page is still a phishing page [16:26] does suspend work for anyone here? [16:26] data:text/html,

This%20is%20a%20fake%20site

[16:27] asac: you told me to remove the alternative stuff from failed-upgrade... [16:27] mdeslaur: i wasnt sure ... so what happend if you trigered failed-upgrade? [16:28] asac: it now doesn't remove the existing alternatives, and installs new one [16:28] chrisccoulson, hey, want to do the g-c-c update? [16:28] asac: let me check if it just overwrites, or if it installes them in double [16:28] seb128 - yeah, i can do that [16:28] chrisccoulson, thanks [16:28] more GtkBuilder porting ;) [16:29] mdeslaur: so at least i wanted the exit 0 only to happen for the version bound we had ... otherwise we still want to fail [16:29] mdeslaur: i thought that we dont want to do anything else, but exit 0 for that case [16:29] e.g. we just catch any eventual error and continue as planned [16:30] mdeslaur: i thought it wouldnt add anything ... but i could be wrong - if it calls the new postinst twice or something ... let me check the debian reference [16:35] seb128 / pitti - i just asked hughsie about the screensaver locking issue on suspend, and he thinks gnome-session should lock the screensaver now [16:35] yeah that's what I was assuming too [16:35] that makes sense anyway [16:35] it's all vuntz's fault ;-) [16:35] heh [16:36] asac: not sure why you think it would call the new postinst twice. Basically, it tries old-prerm, that fails, so it calls new-prerm failed-upgrade and then continues [16:36] mdeslaur: ok so i think really just use exit 0 if version < X ... that should work (none of the whole removal stuff is needed= ... but [16:36] asac: okay, you're right...the rm stuff will still get executed by the old-prerm before failing [16:36] mdeslaur: right. so we can still do it if we are not sure if there are other error cases [16:37] asac: is the ubufox link creation stuff okay? [16:37] let me check ;) [16:40] seb128: it's all hughsie's fault, he wrote the patch [16:40] mdeslaur: i think the mkdir -p fails if the dir already exists? [16:40] vuntz, yeahyeah [16:41] mdeslaur: at best put a .dirs file in the package with /usr/lib/ubufox/plugins ... that should prevent the problems === pedro__ is now known as pedro_ [16:41] asac: no, it shouldn't fail [16:41] asac: I already have a .dirs file with it in [16:42] asac: I guess I can remove the mkdir than [16:42] seb128: well, of course, I could also blame you :-) [16:42] mdeslaur: yes the mkdir would always fail then i would think ... unless we dont have set -e [16:42] mkdir /tmp/ && echo yes [16:42] mkdir: cannot create directory `/tmp/': File exists [16:42] or is -p different? [16:42] -p is differentr [16:42] ok [16:42] but remove it [16:43] mdeslaur: btw its /usr/share/ubufox/plugins [16:43] not lib [16:43] asac: d'oh [16:44] mdeslaur: so while testing you probably didnt change _all_ alternatives that firefox looks at [16:44] asac: yes I did [16:44] thats at least xulrunner-addons firefox-addons mozilla [16:44] ;) [16:44]