[00:00] <cody-somerville> elmo, Have a minute?
[00:01] <elmo> cody-somerville: kind of - it's pretty late here
[00:02] <cody-somerville> elmo, Okay. Maybe another time would be better. I just wanted to follow up with you regarding the discussion we had in Prague.
[00:02] <danshearer> 'evening
[00:03] <elmo> cody-somerville: ah, ok - yeah, happy to talk about that another time (just earlier in the day :)
[00:03] <danshearer> Question about blueprints: let's say you write a stunning blueprint, then come back 5 minutes later 
[00:03] <cody-somerville> elmo, understood :]
[00:03] <danshearer> and realise that actually it is irrelvant/stupid/wrong window typing/whatever. What next?
[00:04] <danshearer> I just marked a blueprint with name, description and url as 'discarded'
[00:04] <danshearer> I know. It's up to me to come up with something worthwhile and put there instead :-)
[01:12] <jamesh> doctormo: if you want to do identifier select, just do consumer.begin('https://login.launchpad.net')
[01:13] <jamesh> if you want beginWithoutDiscovery, it needs to be OPENID_IDP_2_0_TYPE
[01:13] <doctormo> jamesh: working on it
[01:13] <doctormo> That's what I do do
[01:13] <jamesh> but note that the openid code is still in beta: things are subject to change
[01:14] <doctormo> Indeed, there is a big push to use the launchpad accounts with out LoCo websites
[01:15] <jamesh> I'm just saying this because you might need to do a migration at some point.
[01:46] <poolie> oh wow the new search is just amazing
[01:47] <lifeless> :)
[01:48] <poolie> ie it actually works
[01:49] <mwhudson> i think this is an area where new launchpad users have a clear advantage over us oldies:
[01:49] <mwhudson> i don't even _see_ the search box any more
[01:50] <poolie> me too
[01:51] <jml> oh launchpad has a search engine
[01:52] <mwhudson> jml: it's almost like it's 1999 again
[01:52] <lifeless> we should dance
[01:52] <mwhudson> my word, altavista still exists
[01:53] <mwhudson> lifeless: i think that joke has been illegal in most jurisdictions for at least a decade
[01:53] <lifeless> mwhudson: /bow
[02:21] <jamesh> says AltaVista is owned by Overture, who are in turn owned by Yahoo
[02:21] <jamesh> who might get owned by Microsoft in future
[03:25] <SpookyET> hi
[03:25] <SpookyET> I keep getting this email http://pastie.org/217736
[03:26] <wgrant> SpookyET: It means precisely what is said. what is unclear?
[03:27] <SpookyET> md5sums are right
[03:27] <SpookyET> I don't know what it's smoking
[03:29] <wgrant> No, they're not.
[03:29] <wgrant> Not one of those files matches your previous ~ppa3 upload: http://launchpadlibrarian.net/15240870/nano-syntax_20080502-0ubuntu1~ppa3_source.changes
[03:29] <wgrant> Urgh.
[03:29] <wgrant> Wrong link.
[03:29] <SpookyET> I md5sum manually and compared
[03:29] <wgrant> But you uploaded ~ppa3 before.
[03:29] <wgrant> With different md5sums.
[03:30] <SpookyET> not of rdup
[03:30] <SpookyET> you're looking at nano-syntax
[03:30] <wgrant> Or you altered the .orig.tar.gz.
[03:30] <wgrant> Those are the two reasons it would happen, although I was looking at the wrong one before.
[03:31] <SpookyET> I deleted rdup ppa2 to get around this. 
[03:31] <SpookyET> I have modified .orig
[03:31] <wgrant> That's a very very bad idea.
[03:31] <wgrant> Not guaranteed to work.
[03:31] <wgrant> You should never modify the .orig, both because it's semantically and technically wrong.
[03:31] <SpookyET> wgrant: it came with a bad /debian
[03:32] <wgrant> And it was only deleted 5 hours ago - it won't have been physically deleted yet.
[03:32] <wgrant> SpookyET: Even so, you must increment the version.
[03:32] <kiko> SpookyET, you always increment the version. no matter what. :)
[03:33] <wgrant> SpookyET: You should always note in the .orig's version that it has been modified from the upstream tarball, if it has indeed been modified.
[03:33] <wgrant> Or you confuse people who work on your package afterwards, and break Debian syncs if you do it in primary.
[03:33] <wgrant> Saying that something is the original 0.6.0 tarball when it is in fact not is somewhat misleading.
[03:33] <SpookyET> rename rdup_0.6.0.orig.tar.gz to what
[03:34] <wgrant> SpookyET: rdup_0.6.0+spookyet1, I would suggest.
[03:34] <SpookyET> the best way would be for debuild to patch /debian automatically before using it
[03:35] <wgrant> The best way is actually for upstream to avoid putting distro-specific stuff in their source distributions.
[03:35] <wgrant> Makes it much easier for everyone
[03:35] <SpookyET> I don't think it works properly on debian
[03:35] <wgrant> kiko: Is there a way to see the full deletion comment somewhere?
[03:36] <wgrant> It has caused some confusion in Ubuntu, as it has nicely trimmed off a package name so it's ambiguous.
[03:36] <SpookyET> So, it was deleted 5 hours ago, but it's still on disk. How long should I wait?
[03:36] <wgrant> That I cannot tell you. kiko might know.
[03:36] <wgrant> But you don't want to do what you're trying to do.
[03:38] <SpookyET> rdup_0.6.0+spookyet1.orig.tar.gz? the tools seem to want .orig.tar.gz somewher
[03:38] <wgrant> SpookyET: Yes, sorry.
[03:38] <kiko> wgrant, I'm not sure of either, but I'm asleep and as such pretty lamed out
[03:39] <SpookyET> why +? rdup_0.6.0-0spookyet1.orig.tar.gz is similar to -0ubuntu1 format
[03:40] <wgrant> SpookyET: Because we want to change the version of the orig.tar.gz, and that needs to be before any -
[03:40] <wgrant> (well, not exactly, but it's bad practice to do anything else unless upstream does it too)
[03:41] <wgrant> kiko: You speak remarkably well for being asleep, I must say.
[03:41] <wgrant> I thought it was 24-hourly at one point, but it's never good to rely on it.
[03:42] <kiko> I think it's actually less than that now -- but cprov will know for sure
[03:42] <wgrant> kiko: Less frequently, or with less time in between?
[03:42] <SpookyET> well, let's try it
[03:42] <SpookyET> bzr-buildpackage
[03:44] <SpookyET> well, if I rename it, it ain't working
[03:44] <SpookyET> Looking for ../tarballs/rdup_0.6.0.orig.tar.gz to use as upstream source
[03:44] <SpookyET> There is no debian/watch file, so can't use that to retrieve upstream tarball
[03:44] <wgrant> SpookyET: You'll need to alter the version in your changelog as well.
[03:44] <wgrant> TO something like 0.6.0+spookyet1-0ubuntu1~ppa3
[03:49] <SpookyET> wgrant: done. let's see if i get another email
[03:50] <wgrant> SpookyET: You should have it by now.
[03:50] <SpookyET> Is there a dch config so I won't have to -D hardy all the time? it increments to intrepid
[03:50] <wgrant> SpookyET: You're using Intrepid? Brave soul.
[03:51] <SpookyET> wgrant: I'm not
[03:51] <SpookyET> hardy
[03:51] <wgrant> It's not a common use-case to upload to the release pocket of an old distroseries, so I'm not sure there's a way to force it.
[03:51] <wgrant> SpookyET: It looks like it worked, as it's trying to be built.
[03:52] <SpookyET> wgrant: it's a very good backup solution
[03:52] <SpookyET> i'm sure you've heard of hdup
[03:53] <wgrant> I gathered it must be rather good, as Tollef has it in his PPA.
[03:53] <SpookyET> It is. same person who made hdup
[03:53] <SpookyET> http://miek.nl/projects/rdup/
[03:54] <wgrant> I use rdiff-backup for most things at present, but am always on the lookout for something better.
[03:55] <SpookyET> The main page needs some work. he fails to differentiate between rdup the project and rdup the executable. It makes it seem that it only gives you a list of files and you have to figure out what to do with them.
[03:56] <wgrant> It does, yes.
[03:56] <wgrant> But then it mentions things about making gzipped hardlinked backups, even though it says it just prints out lists of files.
[03:57] <SpookyET> rdup /bla | rdup-gz | rdup-gpg | ssh 
[03:57] <SpookyET> spit files, zip'em | encrypt them | ssh them
[03:57] <SpookyET> rdup-simple is similar to rdiff-backup.
[03:58] <SpookyET> I've used rdiff-backup. I can't stand it. It's so bloody slow.
[03:58] <wgrant> It is slow, and eats a ridiculous amount of CPU power, but the end result is nice.
[03:59] <SpookyET> wgrant: an incremental that uploads 5 files can take 2 hours to complete.
[03:59] <wgrant> How big are those files?
[03:59] <SpookyET> kb in size
[03:59] <SpookyET> it goes over the entire tree
[03:59] <wgrant> Never seen anything like that happen.
[03:59] <SpookyET> that's why it's slow
[03:59] <wgrant> Ah, right, yes.
[04:00] <SpookyET> i think it hashes your entire drive
[04:00] <wgrant> Probably.
[04:00] <SpookyET> transfer speed is also slow
[04:00] <wgrant> That's the only sure way to do it.
[04:00] <SpookyET> wgrant: ctime is good
[04:00] <SpookyET> dar uses it. 
[04:01] <wgrant> You have to rely on ctime and mtime being correct.
[04:02] <wgrant> Night kiko-zzz.
[04:03] <SpookyET> it also uses rsync algorithm over ssh which is ungodly slow. I get 2MiB/s with rsync over ssh. I get 8MiB/s with rsync and  rsync-server or scp
[04:05] <SpookyET> There is a really cool backup script called link-backup. It's written in python. It's very fast and it hashes. It creates a catalog. Puts everything in there. Then it hardlinks snapshot tree to that catalog. So if you rename movies isos and other big files, they will not be transfered again
[04:06] <SpookyET> http://www.scottlu.com/Content/Link-Backup.html
[04:09] <SpookyET> It built. You're welcome to try it. man rdup-simple
[04:09] <kiko-zzz> night wgrant 
[04:10] <SpookyET> kiko-zzz: night
[04:20] <SpookyET> wgrant: thank you
[04:21] <wgrant> SpookyET: No problem.
[04:23] <SpookyET> wgrant: I'm used to simpler things --> pacman.
[04:26] <wgrant> SpookyET: How is that simpler?
[04:26] <wgrant> (I've not used it)
[04:29] <SpookyET> wgrant: packages are rdup-0.6.0.pkg.tar.gz . simple .tar.gz. files. Making packages for it is very simple.
[04:29] <SpookyET> One file called PKGBUILD. and one tool called makepkg
[04:30] <SpookyET> Here's a simple one http://aur.archlinux.org/packages/link-backup/link-backup/PKGBUILD
[04:31] <SpookyET> a little more complicated firefox 3: http://aur.archlinux.org/packages/firefox-spookyet/firefox-spookyet/PKGBUILD
[04:32] <wgrant> SpookyET: Looks basically like they've merged all Debian control files into one file.
[04:32] <SpookyET> just shell commands. makes things very simple. no tools upon tools upon tools.
[04:32] <wgrant> Can one do multiple binaries?
[04:33] <SpookyET> wgrant: yeah, but it's shell commands as opposed to a makefile, making "rules" even easier
[04:33] <SpookyET> wgrant: yeah
[04:33] <wgrant> Makefiles can be treated similarly unless you're doing rather strange stuff...
[04:35] <SpookyET> Here's rdup http://aur.archlinux.org/packages/rdup/rdup/PKGBUILD
[04:36] <SpookyET> I like that 99% of the time you can copy the ./configure; make; makeinstall instructions from the developer's site with minimal modification and paste it in the build function
[04:44] <SpookyET> wgrant: Do you like it?
[04:52] <persia> Is there an interface to get to the next page of the package list when copying files in a PPA?  Alternately, is there a way to show more than 20 packages?
[04:53] <wgrant> SpookyET: It's not bad, I must say.
[04:53] <wgrant> persia: You have to do a search.
[04:54] <persia> wgrant: Right.  So binary copy of a large PPA to a new distroseries is tricky.  Thanks.
[04:54] <wgrant> persia: You might be able to specify &batch=somelargenumber on a copy page.
[04:55] <SpookyET> wgrant: There is another one called conary written in python very similar to pacman. It has a recipe file which is like a PKGBUILD, but with python instead of bash instructions
[04:55] <SpookyET> It was created by the rpm author
[04:55] <SpookyET> It's very revolutionary 
[04:56] <SpookyET> http://www.linux.com/articles/60500
[04:56] <wgrant> Oh dear.
[04:56] <wgrant> Why does the search POST it?
[04:56] <wgrant> I almost did something very bad there.
[04:57] <SpookyET> ?
[04:58] <wgrant> SpookyET: Referring to +copy-packages on Launchpad
[04:58] <persia> wgrant: ?batch=75 doesn't do it.  I'll find another way (or not do the binary copy).
[04:58] <SpookyET> url hacking?
[04:59] <wgrant> persia: It doesn't work, right. I suppose it wants POST or something similarly crazy which it shouldn't.
[05:02] <persia> wgrant: Yeah.  I don't think I want to hack it too much with my use case, nor populate my own PPA to be sufficient volume to cause the issue.
[05:06] <SpookyET> Does launchpad not accept binaries for security reasons?
[05:09] <wgrant> SpookyET: That and sanity reasons.
[05:10] <wgrant> It can be very nasty to troubleshoot issues with binaries if you can't guarantee that they were actually built from that source.
[05:10] <wgrant> As well as being illegal in the case of the GPL.
[05:10] <wgrant> Allowing binary uploads removes reproducability.
[05:11] <wgrant> persia: You could hack it up on staging or dogfood - they're nice and safe.
[05:12] <wgrant> It might be a good feature to have to copy all binaries from one DistoSeries to the next, rather like inherit-from-parent.py
[05:13] <wgrant> Everybody will have to do it every 6 months anyway.
[05:14] <SpookyET> The build servers must be quite busy if people upload firefox oo2.org, etc
[05:14] <wgrant> SpookyET: There are three for each arch. Not many people upload OOo, and Firefox isn't much of a problem.
[05:14] <wgrant> Particularly not on the nice fast lpia machines.
[05:55] <NMR_Techie> heeeeeeeeeeeeeeeeeeelp.  lol  i cant sign my code o conduct
[05:57] <NMR_Techie> gpg --clearsign UbuntuCodeOfConduct-1.0.1.txt yields " no such file or directory" after i enter my passphrase
[06:00] <spiv> NMR_Techie: do you have UbuntuCodeOfConduct-1.0.1.txt in your current working directory?
[06:00] <spiv> i.e. does it show up in "ls" output?
[06:18] <Ekushey> persia, do you know if it's possible to change the ubuntu email address or not?
[06:19] <persia> Ekushey: Why me?  I know it's possible, but couldn't tell you how.
[06:22] <Ekushey> persia, i asked you cause i don't know anyone on this channel
[06:22] <wgrant> Ekushey: Your @ubuntu.com email address is generated from your Launchpad username.
[06:22] <wgrant> Altering your username should update the alias within a few days.
[06:23] <Ekushey> wget, it's not possible to get a non-LP username email alias?
[06:23] <wgrant> Ekushey: For some reason, no.
[06:23] <wgrant> It seems rather limiting, but this is how it is.
[06:24] <Ekushey> oh ok :(
[06:31] <NMR_Techie> spiv , yes i do
[08:00] <mpt> Goooooooooooooooooooood morning Launchpadders!
[08:02] <Hobbsee> mpt!
[08:09] <thumper> mpt: I have more questions for you
[08:09] <thumper> mpt: I'll grab you in an hour or so
[08:10] <mpt> ok
[09:21] <wgrant> mpt: Well, that wasn't the solution I was expecting (removal of the help panel). But I guess it works.
[09:21] <mpt> hhe
[09:21]  * mpt cnat tpye
[09:25] <wgrant> mpt: Note that there is a page or two that needs the help tab...
[09:25] <wgrant> (there's a bug on one of them)
[09:26] <mpt> There are pages that need help
[09:27] <mpt> We have a better way to provide that help now
[09:27] <wgrant> Oh?
[09:41] <mpt> thumper?
[09:51] <thumper> mpt: sorry, had to get rid of the in-laws
[09:51] <thumper> mpt: call?
[09:52] <mpt> thumper, yep, shall I call you or vice versa?
[09:52] <thumper> mpt: I'll call you again
[09:52] <mpt> ok
[11:09] <huno> hi there, got a couple of projects registered with LP that I'd like to remove, anyone up to help me? :)
[11:10] <kiko-zzz> huno, yeah, just ask a question (see topic)
[11:10] <huno> kiko-zzz: alright thx
[12:29]  * Se7h need beta testers
[14:27] <bigon> is it possible that a buildd admin have a look at bug #217432
[14:27] <bigon> https://bugs.launchpad.net/bugs/217432
[14:29] <Hobbsee> cprov: ^^ ?
[14:29] <Hobbsee> cprov: build records never created?
[14:32] <cprov> Hobbsee: well, IIRC, this is a arch-indep that doesn't build on i386 (go figure) and it's forbidden in P-a-s.
[14:32] <bigon> yeah
[14:33] <bigon> infinity said he will build it but never do it
[14:33] <persia> cprov: There are about 10 packages like that in Debian.  Some ppc, some sparc
[14:33] <cprov> Hobbsee: soyuz don't support building arch-indep anywhere else than i386
[14:33] <Hobbsee> true, yes.
[14:33]  * Hobbsee wonders why it doesn't get blacklisted or something then, until infinity does it
[14:35]  * bigon wonder why infinity never do it :p
[14:35] <cprov> yes, you have to *deal* with it for now ... building arch-indep in other archs it in our list, but very near the end :(
[14:36] <bigon> :(
[14:36] <cprov> bigon: keep pinging, it's the only think I can suggest, sorry.
[14:37] <cprov> bigon: well, there is also "Fix the source", but it sounds pretty vague at this point :)
[14:37] <persia> cprov: For some sources, it's simply not possible.
[14:38] <cprov> I know.
[15:08] <idle_> Hi there 
[15:08] <idle_> I need help with registering new country for Translation of Ubuntu
[15:09] <idle_> How I can add register new country / language which is not on list 
[15:17] <idle_> How I can add new country / language which is not on list on launchpad translation ?
[15:18] <matsubara> danilos: ^
[15:18] <persia> idle_: People tend to idle here.  No need to repeat yourself so often.
[15:18] <idle_> sorry
[15:19] <matsubara> idle_: you can ask a question in http://answers.launchpad.net/rosetta as well.
[15:20] <idle_> ok , I will check it 
[15:41] <danilos> idle_: you need an ISO 3166 language code for the country, or ISO 639 code for language first
[15:46] <idle_> Hi Danilos , I just found that for my langugae ( Montenegrian ) I need to use something like this sr@jekavian
[15:46] <idle_> Its "jekavica" we use in Montenegro 
[15:47] <idle_> But donno how to be sure that translation will be implemented in the script
[15:49] <idle_> and I made request for adding Montenegro code for ISO , like 1 month ago , but still nothing 
[15:53] <danilos> zdravo idle_ :)
[15:54] <idle_> Pozdrav :)
[15:54] <danilos> idle_: anyway, at the moment we don't support language variants (unfortunately)
[15:55] <danilos> idle_: GNOME Serbian team (which I am a leader of :) has started Serbian Jekavian translations earlier in GNOME directly, though
[15:55] <idle_> ok , but still I can register ME ? 
[15:55] <idle_> aha
[15:55] <danilos> idle_: ME should be among the countries
[15:55] <idle_> Not yet :(
[15:55] <danilos> idle_: I've added both Serbia and Montenegro once they split up
[15:56] <idle_> I'm still not able to see Montenegro
[15:57] <danilos> idle_: where are you looking? (i.e. we have it internally in Launchpad country table, and it's shown that Serbian is spoken in Montenegro: https://translations.launchpad.net/+languages/sr)
[15:58] <danilos> idle_: you can also see locale (one I created a few years back as Jekavian one) on your system as /usr/share/i18n/locales/sr_ME
[16:00] <idle_> I understand , but for example there is Bosnian ( but still register under Serbian ) 
[16:10] <thekorn> hi, is it just me, or is the codebrowse (bazaar.launchpad.net) down today?
[16:11] <Peng> code.lp seems slow too.
[16:11] <Peng> ish
[16:12] <Peng> And yeah, codebrowse seems down.
[16:13] <beuno> abentley, can you restart Loggerhead on LP?
[16:14] <abentley> beuno: No, I'm not empowered to do that.
[16:14] <beuno> argh, mwhudson_ won't be up for a few hours...  who else can do that?
[16:15] <abentley> I could ask mthaddon for you, I guess.
[16:15] <beuno> well, for me and for the rest of the people trying to get in...  :)
[16:16] <mthaddon> will restart codebrowse (just got an alert about it)
[16:16] <beuno> thanks mthaddon 
[16:16] <beuno> I hope the latest changes done to it will make this less frequent...
[16:17] <mthaddon> me too :)
[16:17] <abentley> mwhudson_ seems very positive about those changes.
[16:17] <mthaddon> ok, should be back now
[16:18] <beuno> the new branch uses 1/6th of the RAM, so that should help
[16:55] <thekorn> hmm, Loggerhead seems to be down for me again
[18:19] <tseliot> kiko: Is what I asked here doable? https://answers.edge.launchpad.net/launchpad/+question/35794
[19:05] <mrevell> Meeting time in #launchpad-meeting!
[20:12] <nDuff> Re the initial/provisional OpenID support -- is that provider-only or consumer as well?
[20:13] <nDuff> ...hrm, provider-only.
[20:15]  * nDuff (with his 2-factor authentication token and browser plugin tracking OpenID status and preventing spoofing) is disappointed.
[20:17] <beuno> nDuff, AFAIK, LP doesn't accept OpenID yet
[20:17] <beuno> so only provider
[22:11] <doctormo> I'm having serious issues with openid and launchpad, it seems that every https request just brings back the "Launchpad Login Service" welcome page, even when I follow to openid 2.0 proc
[22:24] <Falken> Hi, quick question, host can I link to another bug in a comment , in a classy way ?
[22:25] <Falken> something like [[http://launchpad.net/stuff|bug #3333]] should work ?
[22:25] <Falken> arf
[22:27] <mwhudson> Falken: 'bug 3333' will become a link to the bug
[22:27] <mwhudson> ubottu: shut it, you
[22:27] <mwhudson> :)
[22:28] <doctormo> let me know when jamesh comes in
[22:28] <Falken> oh great ! as simple as this
[22:28] <Falken> thank you mwhudson
[22:28] <mwhudson> :)
[22:28] <mwhudson> there's no other markup done to speak of
[22:29] <mwhudson> (well, i think cves might get linkified too)
[23:03] <Peng> Why doesn't bazaar.lp offer https?
[23:03] <Peng> Sure, there's bzr+ssh, but when the rest of the site uses https, why not?
[23:08] <lifeless> Peng: its a performance bug that the entire lp site uses https
[23:08] <Peng> Heh.
[23:08] <lifeless> Peng: we'd like to make it https for security related content, and http for other things, to be faster
[23:08] <persia> lifeless: Why is that hard?
[23:08] <lifeless> persia: I didn't say anything about difficulty :)
[23:08] <Peng> I like it erring on the side of security.
[23:09] <beuno> viewing people's LP user page in https is useful to get their SSH and GPG keys
[23:09] <lifeless> Peng: indeed, and so do I. However, let me hit a bug - 1 minute
[23:10] <lifeless> 15 seconds to render https://bugs.edge.launchpad.net/loggerhead/+bug/240504, from an existing browser session (so I have css and javascript already in memory)
[23:10] <Peng> Loggerhead's slowness has to do with https?
[23:11] <lifeless> Peng: I'm talking about launchpad, not lh
[23:11] <lifeless> Peng: it was just a lh bug I used to demo
[23:11] <Peng> Oh.
[23:11] <lifeless> beuno: right, not all things are candidates to be http
[23:11] <Peng> Still, why no https for bazaar.launchpad.net?
[23:11] <lifeless> beuno: but should the projects listing really be https?
[23:12] <lifeless> Peng: well, I guess we could, but why? 
[23:12] <mwhudson> Peng: all branches served by LH are publically available, what would be the point?
[23:12] <mwhudson> apart from annoying all of us with 300ms ping to london
[23:13] <Peng> The main point would be "everything else uses https".
[23:13] <elmo> mwhudson: authentication for free
[23:13] <beuno> lifeless, not if it consumes more resources, no. It's nice to know there aren't any man-in-the-middle attacks on LP though
[23:13] <Peng> Also, a mitm against a bzr repo would be a problem.
[23:13] <elmo> well, for very dubious/broken values of 'free'
[23:14] <Peng> (Though bzr doesn't verify ssl certs anyway, ideally it would.)
[23:14] <lifeless> elmo: and dubious values of authentication
[23:14] <mwhudson> elmo: errrrr
[23:14] <mwhudson> L*H* not L*P*
[23:14] <elmo> oh, LH
[23:14] <elmo> meh
[23:14] <mwhudson> but well yes, not free
[23:15] <mwhudson> we should make all the foundations team move to .au or .nz for a while
[23:15] <elmo> or hackberry
[23:15] <mwhudson> a 10000 km data centre move can't be that hard, can it?
[23:20] <lifeless> its probably easier to buy a single machine here
[23:20] <lifeless> but its unfair to some users
[23:20] <lifeless> so - I propose MoonCentre
[23:20] <lifeless> to be approximately as hard for everyone.
[23:20] <lifeless> might be a bit of a commute for elmo