[06:19] kamstrup: good morning [06:20] morning kvalo [06:23] hello you two [06:27] morning spikeb [06:40] hi spikeb [06:42] kamstrup: I have to apologise in advance for a really long merge request I'm about to send later today. can you read the intermediate commits or should I try to split it some other way? [06:56] kvalo: I'll probably 'bzr viz' it anyway, as I am sick and tired of LPs review interface [06:59] kamstrup: oh, that's a really nice app. thanks for the tip! [06:59] kamstrup: is there any way to get the diff to the same window as the commits? ala gitk [07:00] hehe, too easy. view > show diffs [07:00] kvalo: I'm not sure what you mean... [07:01] kvalo: ah, the setting [07:01] kvalo: I thought Show Diffs was on by default... [07:02] it wasn't for me [07:03] but maybe bzr noticed that user is a git fan boy, and decided to play tricks on me ;) [07:05] kvalo: yeah, it checks if ~/.gitconfig has mtime newer than 12 hours and if so enables the Emacs style "butterfly mode"... [07:06] in fact I added that code to bzr [07:08] hehe [07:14] kamstrup: but really good that you use bzr viz for review. makes also my life easier. [07:59] good morning [08:02] good morning didrocks [08:02] hey spikeb [08:05] greetings everybody [08:05] greetings MacSlow [08:06] yo spikeb [08:17] didrocks: morning. I want to temporarily include debug symbols to indicator-network. is there a switch for that? [08:18] kvalo: morning, yeah there is :) [08:19] didrocks: ie. I want to disable strip [08:19] DEB_BUILD_OPTIONS="debug nostrip noopt" [08:19] didrocks: perfect, thanks! [08:19] DBO: Clearly you and I both hate sleep [08:19] that will keep the debug symbols in the libs/bin [08:19] and remove all optimization [08:19] DBO: eg, it's for the weak, etc [08:19] and then, don't strip in the package :) [08:19] smspillaz, it was college that did this to me [08:20] smspillaz, after 5 years of uni... I simply stopped sleeping. My brain wont shut off at night, I just lay there thinking for hours if I try. So I figure, what the hell might as well be productive [08:20] DBO: it was law school that did this to me [08:20] DBO: college: good grades, social life, enough sleep: pick 2 [08:21] smspillaz, yeah but I only got good grades of the 3 [08:21] whats wrong with me? [08:22] clearly you must have had a social life [08:22] IRC counts [08:22] didrocks: works just as I wanted, thank you very much! [08:23] kvalo: you're really welcome, I always keep it as my favorite post-it :) [08:23] smspillaz, worst social life EVAR [08:23] kvalo: what I want to do for every dx package is to get those option on odd number (==unstable) and not on even number [08:23] that can be cool [08:23] DBO: hahaahah [08:24] smspillaz, in dallas, we'll go out on the town... well the hotel they take is too is more out in the middle of a highway garden... but we'll take a bus into town, then go out on it! [08:24] woooo! [08:25] hit up the night clubs oh wait fml the drinking age changes [08:25] didrocks: I agree, that would be really handy [08:25] smspillaz, are you not of legal age? [08:25] DBO: your stupid united states and it's 21 year old drinking age. why can't it be more like australia (18) [08:25] smspillaz, because then we'd have drunk children everywhere! [08:25] duh! [08:26] Be thankful they're not still prohibiting all forms of alcholol! [08:26] :P [08:26] because we don't already have drunk children everywhere ... [08:27] not being able to drink = meh not being able to go to the clubs = NOOOOOOOOOOOOOOOOOOOOO [08:27] we just banned caffeinated alcoholic beverages here in Michigan [08:27] so yeah, we are heading back to the all out ban, no worries there [08:27] smspillaz, you can go to most clubs, you just cant drink [08:27] DBO: we had a 200% tax on "alchopops" (whatever the hell they are) [08:27] DBO: oh really? So the rules regarding licenced venues are different then [08:27] smspillaz, yes [08:28] in australia you can't go to any place that serves alcohol without fodo until you're 18 [08:28] smspillaz, if a venue is 21+ it tends to be because its easier to make sure you dont serve minors [08:28] hm [08:28] heh [08:28] actually [08:28] the proof of age cards here don't actually say your DOB [08:29] they just say "proof of age card" [08:29] I wonder if I'd pass as 21 [08:29] (probably not) [08:29] wow, this is wayyyy offtopic [08:29] indeed [08:29] to go a bit ontopic [08:29] it's what happens when you don't sleep [08:29] why does recordmydesktop have tons of artifacts in it [08:30] when I use it with compiz + unity [08:30] DBO: nvidia? [08:30] yes [08:30] DBO: yeah that's why [08:30] thats a bad answer [08:30] DBO: nvidia can't do vsync properly on full screen repaints [08:30] wat? [08:30] not even their windows drivers can get it right [08:31] RAOF ^^ confirm? [08:31] that's at least my experience [08:31] Not that specific issue, but nouveau *does* do video vsync significantly better than nvidia. [08:31] It wouldn't amaze me to find that nvidia doesn't vsync particularly well. [08:32] RAOF, does it do power management yet? [08:32] DBO: you're on a macbook pro right [08:32] yes [08:32] DBO: I'd approach noveau with caution, I think there could be some EFI issues still [08:32] Only in the nouveau kernel tree; some power management support should get merged for 2.6.38, and hence make its way into Natty. [08:32] is compiz+unity the default on the daily isos yet? [08:33] RAOF, only some? [08:33] spikeb, no [08:34] DBO: My understanding is that the developer working on it has dynamic reclocking and similar powermanagement stuff pretty much solid for his hardware. [08:34] ok [08:34] RAOF: G80? [08:34] RAOF, thats about the scariest sentence I ever read [08:34] smspillaz: Yeah; only nv5x+ [08:34] I love having the best supported card [08:35] Its just like using linux and not being screwed [08:35] DBO: He's been reclocking his card thousands of times a second overnight. That's moderately solid. [08:35] RAOF, *his card* is the bit that scares me [08:35] Yeah. :) [08:35] as you know, hardware generalizes in the same way that rocks float [08:35] heh [08:35] Would it reassure you to know that fglrx has more power-management code than the radeon kernel driver + mesa driver + DDX combined? :) [08:36] DBO: do you mind if I can clean up the unity plugin code? [08:36] DBO: you wrap a whole bunch of functions you don't need [08:36] smspillaz, oh yeah the spare functions we never used? [08:36] smspillaz, yeah go ahead [08:36] yeah that's inefficien [08:36] i know [08:36] i just stole your example code [08:36] I saw [08:37] then from there its just been "but I might need that some day... might as well leave it until later" [08:37] it seems to be the new rage round the ayatana circles [08:37] though that's not very good considering that it's poorly formatted [08:37] you remember when you play an RPG, and you get ammo for the Super KillMax 5000, but you never use it because you dont know whats around the corner? [08:37] same mentality [08:37] fair enough [08:38] Nice analogy! [08:38] thanks [08:38] I rather like that analogy too [08:38] * smspillaz was probably just going to go with the whole "you save money in the bank for a rainy day) [08:38] I dont think todays youth would get that really [08:38] you're right. SPEND IT ALL ON GOON [08:39] first off, half of them are going to ask you what rain is, and the other half will ask what a bank is [08:39] DBO: hahaha [08:39] haha [08:39] either I am funny when I am tired, or you all are quite nice to me [08:40] it really is the former [08:40] not so much tired ... but ... delerious [08:40] like that night when I started writing code inside of my law essay [08:40] I write code in my sleep all the time [08:40] and then almost handed the essay in with some code caught in the middle [08:40] then I wake up and wonder where it all went [08:40] its depressing... [08:40] haha smspillaz [08:41] DBO: that must be a massive fml right there [08:41] it is! [08:41] the worst part of it is I have gotten really good at this [08:41] DBO: We need to create a device to hook a comptuer up to your brain [08:41] I'll code in my sleep [08:41] wake up, and just write out what I coded whilst sleeping [08:41] i can sort of code in my sleep, but most certainly not while awake. [08:41] and 90% of the time it will be an effective and creative solution to an earlier problem [08:41] DBO: actually I get that too [08:42] its weird [08:42] DBO: although it's not really code, it's more like, wake up, and "crap i was meant to do it this way all along" [08:42] "quick find a pen find a pen" [08:42] I think in framework design [08:42] so it starts as classes, then methods, then implementations [08:42] and usually a monster pushing me down the stairs [08:43] which somehow does not prevent me from typing [08:44] aaanyhow, night all [08:45] good night DBO [08:45] good night didrocks [08:45] Sleep productively :) [08:45] that sounds like the catch phrase apple would use if they made a sleeping pill [08:46] No; it'd be “Sleep productive” [08:46] yes [08:46] smspillaz: so, it seems from what I see there there is profile support in the gconf backend. I guess the profile switch is handled by libcompizconfig, right? (btw, I don't see your commit there about profile choice by the env variable) [08:46] Adjectival forms are for squares. [08:46] you lost me [08:46] I only speak english, not understand the rules [08:47] “Productively” is an adverb, modifying the verb “to sleep”. So I was wrong, too. *Adverbial* forms :) [08:47] * RAOF restarts to test the new mesa hotness. [08:51] didrocks: oh I need to commit that [08:51] didrocks: also, yes profile suppor tis handed by lcc [08:51] didrocks: basically when you switch profiles it copies the gconf tree over to the master one [08:52] DBO: it looks like hooking the paint of one window seems more and more like a hack when I look at it [08:52] DBO: maybe I should change the way CompWindowPaintList works and change it to something like PaintList [08:52] DBO: so we can make Unity a PaintableObject with it's own draw callbacks [08:52] smspillaz: yeah, that's bad… but we will try to get something better for the gsettings backend if possible :) [08:53] didrocks: the gsettings backend should load directly from the profile [08:53] didrocks: the only reason the gconf backend does that is to maintain compatibility with the old gconf plugin [08:53] smspillaz: ok, makes sense [08:54] smspillaz: so, thinking about people having ini backend and not the gconf backend, we should maybe rather ship the .xml file for the profile (which get automatically imported if you activate the gconf backend, right?) [08:54] didrocks: I think compiz config will handle the change in backends [08:54] smspillaz: btw, if you don't have the time now, we can discuss about it a little bit later, just tell me when you have the time for that + looking at the transition script integrations (nice work on gconf btw!) and make dist [08:55] didrocks: yeah, just gotta finish this paint thing first in nux [08:55] I saw a lot of i/o when switching with a new profile, right [08:55] so I guess it handles that :) [08:55] smspillaz: I'll have a look if we can ship the Profiles in /usr or /etc [08:56] yeah [08:56] didrocks: switching profiles between backends is always going to be expensive unfortunately [08:56] didrocks: IMO if the user has chosen to use the ini backend we should just preserve that [08:56] otherwise gconf as default [08:56] in fact, gconf already is the default [08:57] smspillaz: right, the thing is that I want to profiles to be there for both [08:58] so, you mean, if I create a "bar" profile in gconf [08:58] and switch to ini [08:58] it will create the file for me? [08:58] * didrocks tries [09:00] smspillaz: no, it doesn't create a profile, so we have to ship both :/ [09:09] didrocks: I assume we have glib 2.27/2.28 in Natty? [09:12] kamstrup: 2.27.3 [09:12] didrocks: cool [09:39] kamstrup: do I always need to call g_signal_handler_disconnect() during object destroy for signals connected with g_signal_connect()? [09:41] kvalo: basically yes [09:42] kamstrup: damn, I have missed this altogether [09:43] kamstrup: thanks again [09:43] kvalo: if an object connects to its own signals it's less of an issue because the signals wont be emitted once the object is finalized [09:43] yeah [09:43] kvalo: but if the signals are on another object, the other object will still emit signals, triggering the callbacks, when the first object is dwead [09:44] I learned this the hard way :) [09:44] * kamstrup should alias make=make -j2 [09:44] kvalo: so did I :-) [09:44] no no, -j8! ;) [09:45] * kvalo taps his i7 [09:45] kvalo: if you're hungry I can serve you a knuckle sandwich! [09:46] :9 [09:46] it's not good custom to belittle people who don't have an i7 ;-) [09:50] kamstrup: just to give you some motivation to buy a faster machine ;) [09:51] kvalo: rest absolutely assured that I have plenty motivation to do that :-) [10:01] MacSlow: btw, do not forget the "QL/tooltip empty" for the release (kind reminder :)) [10:12] MacSlow: oh, fix commited? (just received the email), awesome! [10:15] didrocks, that is fixed in trunk for some days now [10:15] MacSlow: ok, but I don't have the time to run trunk. Awesome news :) [10:16] didrocks, I always forget we ride different parts of the same wave :) [10:16] MacSlow: heh :) === MacSlow is now known as MacSlow|lunch === MacSlow|lunch is now known as MacSlow [13:13] hmm, is it a holiday in UK? seems no one from Canonical towers in online.. [13:14] s/in/is [13:45] vish: I think it's not a holiday today [13:46] kamstrup: oh man, this one is huge: https://code.launchpad.net/~kvalo/indicator-network/libconnman-backend/+merge/41734 [13:47] kamstrup: take your time, it's not urgent in any way [13:47] kvalo: probably tomorrow before lunch, ok? [13:47] kamstrup: I tried to keep commits small so that it would be easier to review [13:47] kamstrup: even friday is ok [14:13] njpatel, what do you mean for "parenthesis on new line" [14:13] woops [14:13] brackets [14:13] haha ok [14:13] braces [14:13] even [14:13] fudge [14:14] that would be weird, eh? [14:14] if [14:14] ( [14:14] x == y [14:14] ) [14:14] { [14:14] that's what I was thinking [14:14] and i was ready to quit [14:14] return false; [14:14] } [14:14] lamalex, lol [14:15] njpatel, is that true for loops also? [14:15] yeah [14:15] basically everyone likes something different so I made the choice [14:16] yeah, that's fine as long as it's brackets and not parens :) [14:16] * lamalex wants to be technical lead so he can make these kinds of decisions for people [14:16] .. someday [14:16] My usual approach to these problems is choose something I like or everyone hates, but I'd have had to invent a new way of doing braces to do that [14:16] haha [14:17] so I tried to comb my hair by cutting it this morning [14:17] didn't turn out well [14:18] pics or it didn't happen [14:18] njpatel, I'll bring the bag of hair to QA sprint [14:18] njpatel, and re: AddChild vs addChild- should all methods be pascal cased then? [14:18] I found an interesting bug on natty....when my computer attempts to sleep the graphics decides I only need 16bit color when it wakes up [14:19] lamalex, yeah, I think in unity.h/cpp and Launcher* there is some fooBar, but it all should be FooBar [14:19] yeah, unity.h was kind of the basis I used for headers/style [14:19] apparently not the right choice :{ [14:19] lamalex, no, no, never use DBO's files for code style ;) [14:20] ha- yeah I seem to remember giving him lots of shit when I would review merges for Do [14:21] i'm going to make a precommit astyle hook when I'm bored [14:22] okay, my eyes are screwed thanks to this driver bug, going to reboot so I can code [14:25] hah, I forgot to remove that, "EAT IT, BITCH" line.. [14:25] thanks kamstrup [14:26] lamalex: lol, :-) [14:28] the astyle man page is full of /very/ nerdy references [14:30] DeeModel is going to be ridiculously awesome when ported to GVariant [14:31] so, it seems my monitor is broken :( [14:32] <3 dee [14:32] I want to start a new project just to use dee.. [14:32] I'm going to be (1440*900)/(1920*1200) * 100 less productive [14:34] njpatel: we can still read you, even if you can't see us (are you typing blind?) [14:34] dbarth, no, no, I rebooted, But thanks for making me now think that when I type stuff it makes about as much sense if someone couldn't see what they are typing :) [14:35] :-D [14:35] Okay, time to make glib feel foolish [14:36] njpatel, could you look at kamstrup's comments? He had a question or two directed at you I'd like answered [14:36] sure [14:38] lamalex, just the dbus related one, right? [14:39] yeah [14:39] kamstrup, what is simpler about using a stack allocated GVariantBuilder? [14:39] done, basically leave it how it is until I add a proper object that will own the name [14:41] kamstrup, DeeModel is going to be ridiculously awesome when i can use it in python :-D [14:41] * kenvandine ducks [14:42] kamstrup, kenvandine DeeModel is going to be ridiculously awesome when i can use it in C# [14:42] hehe [14:42] i suspect python will come first :) [14:42] not if RAOF ever writes his gobject-introspection parser/generator for mono... [14:43] Dee is close to working with GIR in python [14:45] oh njpatel there were two things :P [14:45] StateIntrospection vs. Introspection [14:46] Introspection is nicer I guess, but it's not a big deal for me [14:46] kenvandine, please implement autocomplete for gwibber [14:46] lamalex, we will [14:46] <3 da best [14:47] it will be in the new vala client [14:47] i hope :) [14:47] if not 3.0, 3.2 :) [14:47] just getting a usable pure gtk client has to come first [14:49] wow new vala client eh? [14:51] yeah [14:52] depends on getting DeeModel working in python though, gwibber-service will provide the model [14:52] so gwibber-service will remain in python? [14:52] yeah [14:53] not going to re-write all that just to get Dee :) [15:00] kamstrup, https://pastebin.canonical.com/40119/ is why DBusMethodCall was public. I need to put it in one of the GDBus structs- and if it's not public I get that error. Is there a way to define those structs in the header? [15:08] lamalex, can it just be a standard static C call inside the file? Seeing as it's a static C++ call anyway, that should work [15:08] njpatel, as in don't declare it in the header? [15:09] right [15:18] njpatel, how do I prototype those in the cpp file? [15:18] or do I just have to get the order right.. [15:19] lamalex, the normal C way of prototyping at the top of the file [15:19] lamalex, basically it just a C function [15:19] yeah, that's what I thought [15:19] see IndicatorObjectFactoryRemove.cpp for instance [15:20] I was getting a weird compiler error [15:21] njpatel, that file are all normal class methods [15:22] /home/alex/Projects/unity/unity.si/src/IntrospectionDBusInterface.cpp:30: error: invalid conversion from ‘void (*)(GDBusConnection*)’ to ‘void (*)(GDBusConnection*, const gchar*, const gchar*, const gchar*, const gchar*, GVariant*, GDBusMethodInvocation*, void*)’ [15:22] void DBusMethodCall (GDBusConnection*); looks like a fine prototype, right? [15:22] lamalex, IndicatorObjectFactoryRemote::on_proxy_ready_cb is just a normal C callback that links up to the C++ object after some validation checks [15:23] lamalex, the gpointer data ? [15:23] user_data, even [15:23] nevermind-- I'm missing so many arguments in that prototype [15:23] I'm a moron [15:25] DBO: yeah, review my work plz :) [15:26] on it you sexy sexy law student [15:26] smspillaz is a law student? [15:26] fancy [15:26] he studies law? [15:26] DBO: also, after I finish the migration stuff with didrocks I'm going to take scissors to the unity plugin [15:26] ahahahahahaahahahahahahaaha [15:26] njpatel, no thats just my fantasy [15:26] lamalex: njpatel: didn't you see when I got a law case submission caught in the compiz planet feet? [15:26] DBO: I'll keep him busy, do not worry :) [15:26] *feed [15:27] * smspillaz will lawyer your face [15:27] DBO, awww, I have a bunch of "keep 100ft away" orders that I was going to get his help with [15:27] (IANAL!!! (yet)) [15:27] njpatel, yeah, in reality he studies Australian Law [15:28] which is like bizarro canada law [15:28] yeah, I'll section 279 you! [15:28] smspillaz, whenever I read that I read: I ANAL [15:28] sorry [15:28] DBO, oh, that's no use to anyone [15:28] exactly [15:28] actually, western australian criminal law is very different to english law actually [15:28] njpatel: BAHAHAHAHAHAHAHHAHAHHAHAAHHAAHAH [15:28] I am actually laughing to hard right now [15:29] njpatel, australian law is kinda a weird concept to me [15:29] it seems we just cant get away from the innuendo today [15:29] I mean they have already been convicted once [15:29] in-your-endo [15:29] DBO: I saw that one coming [15:29] how I met your mother [15:29] * njpatel refrains [15:29] I can't help it [15:29] every time someone says that word [15:29] I *have* to quote The Todd [15:30] *sigh* [15:30] gee and I think the conversations I have with my friends at uni are bad ... [15:30] also its from Scrubs not How I Met Your Mother [15:30] right, dumb youngsters getting it wrong [15:31] Unity has 11 active reviews? [15:31] yeah, but most are for 0.2 [15:31] I wonder if I can mark them as needs-fixing? [15:32] njpatel, in the cpp files there's a strong mix of void Method and void\nMethod [15:32] which is right [15:32] lamalex, void\nFoo::Method [15:32] ok [15:33] but in the header files it's void Method [15:33] for the sake of not having to have screen-size terminals [15:33] lamalex, yes, we're special like that [15:33] ha [15:33] * lamalex keeps his terminal screen-sized.. [15:33] you don't have the Foo:: in header mostly, that's why [15:33] smspillaz, why does your branch work? [15:33] right [15:34] smspillaz, I really honestly dont understand how it works... [15:34] PanelIndicatorObjectEntryView::PanelIndicatorObjectEntryView () [15:34] we really need to use namespaces [15:34] DBO: because I'm a sexy sexy law student? [15:34] smspillaz, no I want a real answer, then I'll humor your funny bone (set up) [15:34] DBO: probably a lack of stack overflows [15:35] I did see one or two calls to empty boost functiosn while mucking around with the clock on this machine though [15:35] smspillaz, no seriously, to me it looks like your find would hit every single glPaint call [15:35] DBO: oh right, you mean my paint thing [15:35] DBO: basically it hooks every single GL window's glDraw call [15:35] wait, window is a list? [15:36] screen->windows () is [15:36] basicaly it hooks every window's call to glDraw, and if the window is above the nux windows it draws the nux windows underneath [15:36] otherwise it draws them on top [15:36] you can walk screen->windows () by doing window->prev??? [15:36] yes [15:36] THATS AWESOME [15:36] doubly linked lists bitches? [15:37] DBO: there is also screen->reverseWindows () [15:37]