=== yofel_ is now known as yofel [12:59] Sarvatt: re bug #733321 - the drm-intel-next driver does not crash when running phoronix-test-suite [12:59] Launchpad bug 733321 in plymouth (Ubuntu) (and 1 other project) "[sandybridge] GPU lockups and other crashes (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/733321 [20:17] ricotz, will the gnome-shell from the ppa in natty be ubuntu-ized with appindicators and so forth? [20:28] bjsnider, hmm, i dont know, gnome-shell has its own implementation of that [20:28] gnome-shell has its own appindicators code? [20:32] yeah, i think so, functional combined with the message tray [20:33] havent really followed the changes lately [20:47] seems like you don't follow the planets :) [20:49] ricotz, i thought you were a part of the team involved in that ppa? [20:53] tjaalton, quite the saga [20:54] bryceh_: yep :) [20:55] tjaalton, how do you think things will end up? [20:55] seems like people aren't finding common ground, I wonder if they'll all just go on getting madder and madder. [20:55] bryceh_: "in fire" [20:55] :P [20:56] (Kosh in B5) [20:56] heh [20:56] actually [20:56] bryceh_, are you talking about this: http://www.markshuttleworth.com/archives/654? [20:57] I did like owen taylor's post [20:57] http://blog.fishsoup.net/2011/03/11/what-does-the-user-see/ [20:58] bjsnider, yeah plus the ones by dneary, aaron, and now this lecture by jeff [20:58] makes my brain hurt :-) [20:59] i would like to know if this agreement between jon mccann and ted gould in 2008 over appindicators really happened [21:00] problem is only a fraction of the discussions are archived, and people tend to remember different things [21:00] yep [21:01] it sounds like the kind of thing that should have been written down in a contract if it did take place. would mccann really have said "you guys go off and solve this problem and we'll accept the code when you're done"? [21:02] well, I know ted well. And I know he is anything but shy about talking about ideas, even to people who disagree. Indeed he *loves* arguing with people. [21:02] so I can totally believe that they *did* talk [21:03] with no witnesses, nothing on paper, only handshakes [21:03] otoh, ted's not really a 'lets make a contract' type... he'll more often than not go off and do things as he thinks they should [21:04] anyway, I don't know [21:04] bjsnider: The thesis that they talked in 2008 and nothing happened for 2 years is at odds with the facts. [21:05] The work didn't start in earnest until ~ Lucid's UDS in Nov 2009. [21:05] ScottK, canonical went off and really did develop the code, i understand that [21:05] bjsnider: Yes, but not without trying collaboration first. [21:05] There were Gnome people at the UDS discussion and it was decided to discuss the spec at FDO. [21:06] Reading that discussion, which is public, I don't get the feeling the Gnome side of the discussion was particularly open or serious. [21:06] ScottK, I think I was at that meeting too, and seem to recall it that way too [21:06] did gnome participate in that fdo discussion? [21:06] Yes. [21:06] But not very seriously, IMO. [21:07] The fact that Canonical's Gnome and KDE had compatible implementations appear at the same time makes a good case for collaboration was done where there was collaboration to be had. [21:08] I'm not holding Canonical up as perfect, they aren't, but it's certainly not the way certain people are trying to portray it. [21:09] ScottK, now you need to write a 5k word blog post about it [21:10] No. I commented on Waugh's post. That's as far as I care to get dragged into it. [21:10] i haven't gotten that far down yet [21:11] did waugh get fired by canonical or did he leave on his own? [21:11] Many of the criticisms of (the Ayatana part of) Canonical are very valid. [21:11] No idea. [21:11] i mean does he have an ax to grind? [21:11] I think he's definitely got a slanted perspective. No idea if that's why. [21:13] The fact that they require copyright assignment and insist all work be one on LP makes collaboration very difficult. [21:14] Of course Gnome insists all work be done in Gnome Git, so they're equally difficult in that regard. [21:14] and didn't they insist that all discussion take place on their ml? [21:15] (Look into the history of Zeitgeist's rejection from Gnome) [21:15] That's a non-Canonical example where many of the same introspective issues in Gnome seem to be at work. [21:15] Gnome has lots of rather strict requirements for stuff like that, which is why we never had Inkscape join Gnome [21:16] yes,, why bother with their strict requirements when we have strict requirements of our own to worry about [21:16] heh [21:17] well, at the time the requirement was that you must use cvs and bugzilla, and we were wanting svn and some better bug tracker [21:17] Canonical has strict requirements for Canonical funded work. Very different than requirements for getting into Ubuntu. [21:18] yep [21:19] so maybe that was the reason to keep it out of gtk+.. [21:19] it= the lib..indicator [21:20] well, it wouldn't surprise me if it came down to a control issue at some level or other. If it were merely a communications problem or technical issue, those seem much more straightforward to solve. [21:22] maybe the problem is canonical doesn't have enough core gnome devs working for it directly then. because then it would be in their control. subject to shuttleworth's decisions [21:24] but that wouldn't be collaboration :) [21:26] maybe the real question here is does the tangible value of collaboration outweigh the perceived value of control? [21:28] If you care about user experience, I don't think collaboration is optional. [21:29] I agree, but that appears not to be a universally held view... [21:29] Users should be able to use applications in their DE of choice without having to spend a lot of time thinking about what toolkit or open source project it was developed in. [21:29] bryceh_: Agreed. [21:29] Although to be fair to Canonical, I think they care about this in the context of Ubuntu users. [21:30] I think it's an open question though how much they really care about the broader free software ecosystem. [21:32] * bryceh_ nods [21:38] otaylor's post seems more open and inviting to collaborators, while shuttleworth's sound like "you've failed me for the last time, admiral". [21:39] There are varying degrees of willingness to collaborate in all camps. [21:42] but otaylor's pointing out that the programming resources that went into unity could have been used to improve g-s is also applicable everywhere around linux: why is it that every time someone has a problem with rhythmbox he goes off and creates a new media player? why not make r-b better instead of dummying up something from scratch? do we have to have 47 media players in linux? [21:43] now we'll have a confusing number of shells too i guess [21:44] bah, I don't buy into the competition == inefficient use of resources [21:45] Avoiding competition and planning everything out worked out really well for the USSR. [21:46] fact is, when you have a single thingee that everyone contributes to, it inevitably builds up layers of bureaucracy (it has to), and it can get to the point that the activation energy to get into that is way above what it takes to JFDI [21:46] bryceh_, so you'd like to see 47 xorgs [21:46] and endless flame wars about how "the one i use is the bestest" [21:46] cough Wayland cough [21:46] bjsnider, ahem wayland? [21:46] ok, xorg sucks [21:46] bjsnider, Xorg is sort of proof of the premise ;-) [21:47] one competitor is fine [21:47] nah, it's an evolution [21:47] but i could sit here and list all of the media players on linux for the next 5 hours and not finish [21:47] early on you might have 47 different ideas, and that's fine [21:47] but over time those should die off, and the best survive [21:47] bjsnider: If you want to avoid a competitor project then you have to be open enough to what others want that they think it's worthwhile to play together. [21:47] (this doesn't always happen) [21:48] if it does, and it boils down to 2-3 good options, that's ok [21:48] Media players are a bad example because they are so easy to do (at least badly) [21:48] when you get down to just 1, then you get complacent, bureaucratic, and out of touch. So it's time to have a bunch of fresh new ideas [21:48] http://apachelog.wordpress.com/2011/03/02/how-to-create-a-media-player-in-30-seconds/ [21:48] the important thing is to be able to *let go* and let less popular ideas die off [21:49] that's really hard though, especially when it's your own baby that has to die [21:49] ScottK, that post should be destroyed for the good of mankind [21:49] bryceh_, right, so ego is heavily involved in this [21:49] I don't care how many media players there are long as at least one doesn't suck. [21:49] bjsnider, in fact ego is an important part [21:50] what else is going to drive someone to spend countless evenings voluntarily creating something? [21:50] money, for one thing [21:50] they have a cool idea and want to see it exist, and have people find value in it [21:51] bjsnider, sure but I'm talking within the context of open source [21:51] and even with money, there's a lot better ways to make money than writing software ;-) [21:52] And there are better ways to make money writing software than writing free software, even when you're funded. [21:53] ScottK, like taking money from one of those aforementioned 47 media players? [21:53] No. I think that was just idiotic. [21:54] Canonical will lose more money from the bad PR and loss of goodwill than that will ever make the. [21:54] the/them [21:55] I think that's a fair observation [21:55] sort of goes back to what I was saying about value of control vs. value of collaboration [21:56] http://abock.org/2011/03/10/opensuse-11-4-and-banshee-amazon-mp3 <-- Exhibit A [21:56] Gotta run. Chat with you later. [21:56] there seems to be a rather wide set of opinions on where that balance should lay