[05:35] https://dev.launchpad.net/Running here tells to dist-upgrade, what the hell. [05:36] after lp-branches/temp/utilities/rocketfuel-setup, i got there: http://bpaste.net/show/IcMIAb7WlsuTtuyjG3M1/ [05:46] That seems reasonable. It's the simplest way to make sure you're current with the PPA you just added. [05:47] That looks like update-sourcecode hasn't been run. Try utilities/rocketfuel-get. [10:59] cjwatson: i got the idea. it downlaod thing to ~/launchpad, not the working directory of running it. [11:02] Right [11:29] cjwatson: it seems hard-coded... [11:29] cjwatson: i got another problem on start the web-server. it keep report "Adress already in use", even if i stop the apache. [11:37] haobug: I think you need somebody a bit more experienced than I am, sorry [11:37] I trust you're doing all this in an LXC instance [11:37] er... | cjwatson [11:38] It isn't worth the pain of trying to do it in other environments, IME [11:38] LXC instance? what is that ? [11:38] linked from the top of the page you're following [11:38] https://dev.launchpad.net/Running/LXC [11:38] oh sorry/ [11:40] Using an LXC container means that you don't get things like server port clashes with other things on your system. [11:40] i didn't configured the LXC. i left out the `dist-upgrade` line from the guide. [11:43] Downloading ubuntu precise minimal ... [11:43] I: Retrieving Release <-- it says, i noticed it installed qemu, does that means i am going to install another ubuntu within my current one? [11:45] cjwatson: can i use the offline iso image to speed up this process? [11:45] No idea [11:46] And yes, an LXC container involves a nested installation [11:46] It's manageable even on my terrible internet connection so it shouldn't take all that long. [11:47] cjwatson: i am in a intranet with about 50K band-limitation... [11:49] Gosh, even worse than mine. [11:51] cjwatson: it's a bit annoying to install another ubuntu on top an already installed one. [11:51] LXC containers are fantastically convenient once you have it set up. Perhaps persuade your intranet administrator to give you a local mirror [11:51] It's probably possible to mangle /usr/share/lxc/templates/lxc-ubuntu somehow to get packages from an ISO but it might still take more effort than just setting it off and doing something else while it completes [11:52] Assuming by 50K you mean 50Kbytes, my network connection is only about five times faster than yours, typically [11:53] So not like orders of magnitude better or anything [11:53] cjwatson: ..... are there another way or docs i can read through and figure out a way myself to solve the 'already in use' problem? [11:54] I really can't recommend it. [11:54] I tried to avoid the LXC setup for a while for similar reasons, and it was too much effort to be worth it. [11:55] wait a second, you are in this channel, you must be someone develop on it or just use it, okay i got that, you installed the LXC... [11:56] I'm a Launchpad contributor, yes [11:56] Even on an awful connection, in anything more than the extremely short term it's worth it to accept the one-time cost of setting up a container and save painful debugging effort later [11:57] then should should be familiar with some 'internal' or hacks, that you can tell me to fix this ;) [11:57] It might be possible to set MIRROR=file:///path/to/unpacked/iso in the environment when creating the container [11:58] I'm only familiar with what I've needed to do myself, which I've explained to you [11:58] cjwatson: can i mount the iso directly ? [11:58] Yes [11:59] Even if that works for creating the container, though, you'll still need to set /etc/apt/sources.list back to a proper mirror later; the ISO won't have enough packages to actually run Launchpad [11:59] cjwatson: oh shit, my iso is amd64. [11:59] Is that a problem? [11:59] is it ok to change the i386 to amd64? [11:59] Gosh, why does that say i386 [12:00] RAM usage [12:00] But if you're only running one instance it's not usually a problem. [12:00] LXC ftw :) [12:00] Oh, I just changed it to amd64 ... [12:00] Revert it if you think that was wrong :) [12:00] It's not like we test Launchpad on i386. [12:01] All my non-buildd installations are still i386 for that reason, actually. Though my desktop has 24GiB of RAM now so that's possibly less necessary than it once was. [12:01] Do you think I should change the docs back and add an explanatory note, then? [12:02] Meh. [12:02] Most people aren't me, so they won't run several instances simultaneously, so it probably doesn't matter either way. [12:02] wow! the mounted iso works: http://bpaste.net/show/NASyuEoXKIH4iSrGwqHk/ [12:03] Eh, might as well I guess, I've added a note [12:03] Since otherwise I'll probably forget again later [12:03] Hah, turns out my container is i386 too, I'd entirely forgotten that [12:05] wgrant: does the i386 matters? i am using amd64; cjwatson: are you the one maintains that page? [12:05] haobug: The page was basically written by the two of us, I think. [12:05] The architecture doesn't matter for functionality. [12:05] But i386 uses less RAM [12:06] I think I tweaked it a bit but can't claim to have written it. [12:07] wgrant: okay that should be a question to right person ha? [12:07] Well, you have an answer now anyway [12:07] Oh yeah, I thought you made more substantial changes than you actually did, it seems. [12:08] yes. i am perfer to use i386, i am now using amd64 because i got an iso of that from local network in my intranet. [12:10] i have asked question about the Karma mathematics modal, can you give me some explanation? [12:11] haobug: There's not much to it. You can see the algorithm in cronscripts/foaf-update-karma-cache.py [12:12] Each karma-granting action has a value assigned. An action's value decays linearly to 0 over the year after it was performed. [12:13] Karma is categorised by app: bugs, code, translations, etc. The total value of the karma actions in each category is scaled to be roughly equal. So if there are 1000 points of actions in the last year in Code, but 10000 in Bugs, Code points are worth 10x as much as Bugs. [12:13] The karma value displayed on each user's page is just the sum of the scaled, decayed actions. [12:19] the guy ask me to setup the launchpad told me the launchpad has the measurement system all of aspects of contribution. some LUG members and me discussed about a good way for measuring contribution of members. then i run into this :D. [12:19] I hope you aren't planning to actually run LP for real on a system with 50KB network access; that doesn't sound like fun at all [12:21] cjwatson: the internat locally are connected at speed about 100MB(1000Mb? i am not sure). i can download the internal file at speed of 90MB+. [12:23] "scaled to be roughly equal", i don't get that all. 1000 points in code, and the 10000 points in bugs are all contributed by one or the whole system sum up? [12:24] haobug: System-wide. [12:24] oh yeah, that's a currency system as we discussed at LUG meeting. [12:28] wgrant: the points drops as time goes, because there are new contributions included. that's a clever idea. [12:30] haobug: Each action's value decays to 0 over a year because karma is designed to show recent activity. Someone who hasn't participated in the project for a year will have no karma, so karma can to some extent be used to judge how active someone has been lately. [12:32] wgrant: what does that mean, i can't read python by now, the karma can drop without new contribution, just because time goes? [12:34] haobug: Right. If all I've done in Launchpad is file a single bug, that might be worth 10 points. So I'll have 10 karma immediately after filing that bug. Six months later, I'll have 5 karma. Another six months after that, I'll have 0 karma. Each action only has value for a year. [12:37] wgrant: that sound strange. with the contributions grow as time goes, the point of each individual should drop. why do you introduce such decaying? [12:39] If contributions actually grow with time, then they will gain new karma. [12:40] But five-year-old contributions aren't as interesting as recent ones from the point of view of getting a basic idea of who's currently most active. [12:40] wgrant: the whole worth of the launchpad system is dropping too, if the contribution activity is steady. [12:40] Karma isn't intended to be a total measure of all contributions ever. It's meant to be a rough indicator of who's active at the moment. [12:41] oh..... i almost got the idea. it is not a currency system as i thought minutes ago. [12:42] No; or if it is, it's a highly deflationary one. [12:42] But you can't exchange karma for anything so it's not currency by any sensible understanding of the word. [12:43] the whole worth of the system drops, how can that be a currency system? [12:45] >you can't exchange karma for anything [12:45] we are planning to make it can do exchange with the reality world in our LUG, with some sponsors. but it is just a plan in mind by now. [12:46] Deflationary currencies are still currency, even if they're generally pretty awful ones. But not being able to exchange it for anything definitely makes it not currency. [12:47] Karma wasn't designed as a currency, so you'll need to make design-level changes if you want to try to use it as one. [12:50] cjwatson: er... is that changes big? [12:51] I don't know [12:51] That would be your job to determine if you plan to do this :-) [12:51] Since it kind of depends on what changes you want [12:55] i don't know exactly. but i think the total worth of the system should not drop. [12:56] do you know some other alternatives system like launchpad manage all of this and has a ready to use currency system? [12:58] I don't; it's not one of my interests. [13:00] So do I... === Ursinha is now known as Ursinha-afk [17:05] Hi! I'm making login.launchpad.net OOPS via OpenID .. wondering if anybody is up for trying ot fix it? [17:05] OOPS-a74e777c522c4ea69b4f537c946d667a [17:05] caused by === Ursinha-afk is now known as Ursinha [17:05] AUthOpenIDAXRequire email http://axschema.org/contact/email . [17:05] AuthOpenIDAXUsername email [17:07] hm [17:10] SpamapS: That's not an OOPS in Launchpad proper; you want SSO support [17:11] (Can't remember if there's an appropriate IRC channel, sorry; the support link just leads me to https://forms.canonical.com/sso-support/) [17:11] cjwatson: Ah right. [18:30] cjwatson: SpamapS: there was #canonical-isd at one point [18:31] That could be it, thanks, memory didn't serve [18:33] I sent an email [18:33] got an rt response [18:33] I'll set the effort aside to see what they say