[07:25] <wgrant> Is there a nice bzr viz-ish tool around that will show me several branches in one revision tree? I've got a set of inter-related branches from a couple of months ago that I'd really like to see the relationships between.
[07:25] <lifeless> bzr viz has some stuff to do that, and I think qlog can to
[07:26] <lifeless> failing that: bzr merge --force them all together, do a single commit, and then use bzr viz :P
[07:26] <wgrant> I considered that last option.
[07:26] <wgrant> i tried qlog, but it seems to have hung.
[07:26] <wgrant> But maybe it's just processing 70000 LP revs.
[07:32] <geartrooper> hello, is it possible to lock a folder in ubuntu for a bzr repo that friends can only see the folder?
[07:36] <cody-somerville> geartrooper, pardon?
[07:36] <geartrooper> ty for responding cody-somerville.  Is it possible to use my machine for a repo and have friends use only the repo folder?
[07:37] <cody-somerville> geartrooper, for read only or read and write operations?
[07:37] <geartrooper> read and write
[07:38] <cody-somerville> Yes its possible but requires setup.
[07:38] <geartrooper> is there a tutorial for such cody-somerville
[07:39] <cody-somerville> geartrooper, Have you considered using launchpad.net to host the branches?
[07:39] <cody-somerville> geartrooper, https://launchpad.net/+tour/branch-hosting-tracking
[07:40] <geartrooper> I have
[07:42] <cody-somerville> If you could share with me why you didn't find it a suitable option it'll help me understand what your needs are.
[07:42] <geartrooper> well, I am using bzr for a secret gaming project.  I would like to keep the assets hidden from potential players.
[07:43] <cody-somerville> Launchpad offers commercial subscriptions but I take it you've already evaluated and ruled that option out for whatever reasons?
[07:44] <geartrooper> that assumption would be correct
[07:45] <cody-somerville> An alternative to setting up a restricted file server on your personal machine is to use a distributed development model.
[07:46] <geartrooper> please elaborate
[07:46] <cody-somerville> Certainly.
[07:47] <cody-somerville> geartrooper, http://doc.bazaar.canonical.com/bzr.2.0/en/user-guide/bazaar_workflows.html
[07:48] <wgrant> lifeless: qlog eventually worked (but hung for a looooong time). thanks.
[07:49] <geartrooper> cody-somerville, does centralized run the risk of a changing ip address on desktop machines?
[07:50] <cody-somerville> geartrooper, depends on your internet connection. Most residential connections have a dynamic (ie. changing) ip address.
[07:50] <geartrooper> crud
[07:50] <lifeless> wgrant: file a bug
[07:51] <geartrooper> well, launchpad will work fine, we'll just have to be smart.  cody-somerville thank you very much for your help and best wishes on the project
[07:51] <lifeless> geartrooper: you can by an lp subscription
[07:51] <lifeless> geartrooper: which will let you make your branches private and only accessible to folk you choose
[07:51] <geartrooper> noted
[07:52] <geartrooper> thankyou guys very much.
[14:06] <cyberix> What is "the gui" for bzr in Ubuntu?
[16:45] <bialix> hello
[16:45] <bialix> anybody here remember old baz?
[16:55] <jelmer_> not really
[16:55] <bialix> evening jelmer
[16:55] <jelmer> hey bialix
[16:56] <jelmer> congrats on the new bzr-explorer
[16:56] <bialix> do you know maybe, mainline idea is specific for bzr or it comes from baz/arch?
[16:56] <jelmer> no idea, sorry
[16:56] <bialix> jelmer: thanks, but all congrat belongs to igc
[16:57]  * bialix writing blog article about 1 branch is 1 directory
[17:15] <mjg> hello.  is it possible to configure loggerhead to allow commits?
[17:15] <mjg> (ie not just be read only)?
[17:16] <jelmer> mjg: Not that I'm aware of
[17:16] <mjg> OK - so how to have a read/write repo accessed over http?
[17:18] <mjg> is it possible?
[17:19] <jelmer> Yeah, I'm pretty sure it is possible
[17:19] <jelmer> Unfortunately I don't have any experience with it myself
[17:20] <mjg> :) thanks
[17:21] <mjg> it's quite frustrating trying to get it set up
[17:26] <bialix> mjg: read about bzt+http
[17:26] <bialix> mjg: also there is webdav plugin
[17:26] <bialix> *bzr+http
[17:26] <mjg> gotcha
[17:33] <mjg> OK: I have run  bzr serve --directory=/home/tapir/bzr/ on my repo server
[17:34] <mjg> on the client machine I can run  bzr branch bzr://tapir:4155/myProject/trunk and get a working copy no problem
[17:34] <mjg> but if I try  bzr branch bzr+http://tapir:4155/myProject/trunk I get errors
[17:35] <mjg> do I need to start the server differently to use the client like this?
[17:36] <NfNitLoop> I'm not sure what bzr+http:// means.   bzr:// is for bzr serves, and http:// is for a bzr directory served via HTTP.
[17:37] <mjg> OK - I am still a bit confused, but if I try  bzr branch http://tapir:4155/myProject/trunk I still get an error
[17:38] <mjg> which is: bzr: ERROR: Invalid http response for http://tapir:4155/myProject/trunk/.bzr/branch-format: Bad status line received
[17:38] <NfNitLoop> well, bzr serve isn't http.
[17:38] <NfNitLoop> It's its own protocol.
[17:38] <mjg> OK - I see
[17:38] <mjg> slowly
[17:38] <NfNitLoop> so you wouldn't expect that to work.
[17:38] <mjg> I thought the "smart" of the "smart server" would allow that to work
[17:39] <NfNitLoop> No, the "smart" of the smart server is that it can do some of the work of picking which revisions to send you.
[17:39] <NfNitLoop> instead of your client having to poke around, say, and HTTP share and find them.
[17:39] <NfNitLoop> an* HTTP
[17:39] <NfNitLoop> and I'm sure there's other stuff too, I'm not one of the developers.
[17:40] <NfNitLoop> But generally with your own custom protocol, you can make things go faster than trying to do it over HTTP. :)
[17:40] <mjg> sure :) but I have a strict firewall to contend with
[17:40] <NfNitLoop> ah.   Does that firewall allow you to ssh?
[17:40] <mjg> so I'd like to be able to checkout & commit over http
[17:40] <mjg> no
[17:41] <NfNitLoop> Hrmm.
[17:41] <mjg> it is port 80 only, so I'd like to run bzr over http and use apache as a reverse proxy
[17:41] <NfNitLoop> You can check-out via HTTP by just putting your project in a directory that's servable via your HTTP server.
[17:41] <mjg> yes, but then I can't check it back in
[17:41] <mjg> :(
[17:42] <NfNitLoop> correct.
[17:42] <NfNitLoop> you could check in locally, though, and then pull up to there via HTTP.
[17:43] <mjg> if I could ssh to the remote server, you mean?
[17:44] <NfNitLoop> Yeah.  I imagine you have *some* way of connecting to it if you're getting files onto it in the first place.
[17:44] <mjg> if I go to another building and use the terminal, yes
[17:45] <NfNitLoop> maybe it's not the best place to store your branches, then.
[17:47] <mjg> perhaps not.  I want to store my repo on my home computer then be able to checkout, work, and checkin again from a pc in my lab at work.  But work has a horrid firewall which doesn't allow me to ssh into home.
[17:47] <mjg> so I thought I could run a read/write bzr server over http
[17:47] <mjg> but it doesn't seem possible, from what I am finding
[17:47] <NfNitLoop> You can only read bzr over HTTP at the moment.
[17:48] <mjg> oh,  then I am going to have to remember to carry my usb-key with me all the time then
[17:48] <NfNitLoop> That works.
[17:48] <mjg> until I lose it :o)
[17:49] <NfNitLoop> put it on your keychain. :)
[17:50] <mjg> well I guess I could run the standard bzr serve on port 80
[17:50] <mjg> but what about security?
[17:50] <mjg> the advantage of running over http is to put t behind an apache proxy that uses http-auth
[17:50] <NfNitLoop> at that point, why not run ssh on port 80? :)
[17:51] <NfNitLoop> http-auth isn't very secure, all your passwords go over plaintext.
[17:51] <mjg> yy
[17:54] <mjg> good points, all
[17:54] <mjg> I'll have to think this through.  thanks for your help
[19:56] <jelmer> jam, lifeless: around?
[20:05] <thumper> morning
[20:07] <jelmer> hi thumper
[20:08] <thumper> jelmer: have the branch stalled for bzr-hg into LP?
[20:09] <jelmer> thumper: it's waiting for some changes to the database schema
[20:33] <jelmer> jam: ping
[20:52] <lifeless> jelmer: ?
[20:55] <lifeless> NfNitLoop: you can write over http, just use the wsgi service
[20:55] <lifeless> mjg: ^
[20:55] <jelmer> lifeless: I'm playing with using indexes for the shamap in bzr-git
[20:56] <jelmer> lifeless: but I was having a hard time finding a helper class to do my directory management for me
[20:56] <jelmer> lifeless: just sent an email to the list about this
[20:56] <lifeless> the simplest example I know of is inbzr-search
[20:57] <lifeless> I haven't factored out a just-help-this class yet
[20:59] <lifeless> I want to write a flexible schema package built on these concepts
[20:59] <lifeless> I've even got a name :P but no substantial code
[21:05] <NfNitLoop> lifeless: Oh, huh.  Sorry, thought I saw someone in here the other day saying that HTTP didn't support writes.
[21:06] <lifeless> plain http with no smart server or webdav doesn't
[21:06] <lifeless> http with webdav
[21:06] <lifeless> or bzr+http (which just needs the server configured)
[21:06] <lifeless> do
[21:06] <jelmer> lifeless: yeah, such a class would be neat - I don't feel like reimplementing this behaviour in bzr-git, bzr-svn *and* bzr-hg :-)
[21:07] <NfNitLoop> guess I've got some reading to do. :)
[21:17] <_habnabit> If I have a file and I'm trying to split it into multiple files, is there any way to keep the  revision history in the new files?
[21:18] <jelmer> _habnabit, at the moment, no
[21:18] <_habnabit> Dang.
[21:19] <_habnabit> It sounds like it's a planned feature, though.
[21:19] <_habnabit> (And this is probably a subset of allowing versioned copying. Is there an open ticket for that?)
[21:20] <jelmer> _habnabit, there is a spec, not much more than that
[21:20] <_habnabit> Aha.
[22:55] <poolie> good morning
[22:57] <jelmer> 'morning poolie
[22:58] <poolie> hi jelmer
[22:58] <poolie> how's soyuz going for you?
[22:59] <jelmer> poolie: fun :-)
[22:59] <jelmer> I've landed 7 branches so far, though almost all for trivial things
[22:59] <poolie> great
[23:00] <poolie> well, you need to get your feet wet
[23:01] <jelmer> poolie: Also, http://jelmer.vernstok.nl/blog/archives/244-My-first-week-as-a-Launchpad-developer-impressions.html
[23:02] <poolie> that's good that reviews and questions are quickly answered
[23:03] <poolie> and re karmic, as you probably know, it's a conscious tradeoff decision
[23:03] <jelmer> poolie: yeah, I realise that - it's still something I ran into though
[23:03] <poolie> iwbni it worked in more places but i think the external version dependencies are quite tight
[23:03] <poolie> still worth mentioning
[23:04] <poolie> we should add you to planet launchpad?
[23:04] <jelmer> that'd be great, I wasn't aware there was such a thing :-)
[23:05] <lifeless> jelmer: you could patch pilot too :P
[23:05] <poolie> ooh
[23:05] <poolie> i was just wondering who should do it next
[23:05] <poolie> also you are on our planet, but it's apprently not updating
[23:07] <jelmer> poolie: it was only posted a few minutes ago, I had forgotten to set it to Published in serendipity
[23:07] <jelmer> lifeless: yeah, I guess I should at some point...
[23:14] <lifeless> jelmer: why delay, pilot today
[23:19] <jelmer> lifeless: lack of spare time at the moment :-(
[23:26] <jelmer> lifeless: Wait, did I just see you quit IRC for a second?
[23:27] <lifeless> bouncing irssi because of freenodes retarded behaviour on channel joins before auth
[23:35] <fagan> Hey all I have a small problem with bzr in lucid anyone around to figure it out with me?
[23:36] <jelmer> hi Shane
[23:36] <jelmer> fagan: what is the problem?
[23:36] <fagan> jelmer: hang on ill get you a pastebin of the message
[23:37] <fagan> http://pastebin.ubuntu.com/340831/
[23:37] <fagan> I have my pub key set and imported to launchpad but I still get the error
[23:37] <Peng> fagan: Is that with Launchpad?
[23:37] <Peng> Ah.
[23:37] <fagan> yep
[23:37] <Peng> Well, apparently you don't. :P
[23:37] <Peng> Are you sure it's the right key? Did you paste it properly?
[23:37] <fagan> Yep
[23:37] <fagan> Synced the key
[23:38] <fagan> then copyed the fingerprint and confirmed it
[23:38] <Peng> The fingerprint? Not the id_rsa.pub file?
[23:39] <fagan> When you sync your key all you have to do is give the fingerprint
[23:39] <fagan> bzr didnt ask for my password though
[23:40] <fagan> Could that be the problem?
[23:40] <Peng> ....
[23:40] <Peng> "give the fingerprint" means "paste the contents of id_rsa.pub in the form"?
[23:41] <fagan> I did that
[23:41] <fagan> the key is definitly linked to my launchpad
[23:41] <Peng> LP would've probably given an error if you did it wrong anyway.
[23:41] <Peng> Sorry, my only other guess is that your SSH client is using the wrong key.
[23:41] <fagan> hmmmm
[23:42] <Peng> Or you're using the wrong username.
[23:42] <spiv> Try sftp -v your_launchpad_username@bazaar.launchpad.net
[23:43] <spiv> That will show you which key it is trying.  (And if it works then that suggests that bzr isn't trying the right username)
[23:43] <spiv> Hmm.
[23:43] <poolie> hello spiv, peng
[23:43] <spiv> I wonder how hard it would be to include the username bzr tries in that error text...
[23:43] <Peng> Good $time_of_day, poolie. :)
[23:44] <fagan> Ah its using the wrong key
[23:44] <Peng> \o/
[23:44] <poolie> spiv, that would be good, though...
[23:44] <fagan> Ill fix it thanks all
[23:44] <poolie> maybe sometimes we let the ssh client default? i don't know
[23:44] <spiv> Hmm, and we could probably add a -Dssh flag that would cause -v to be passed to openssh.
[23:44] <spiv> We sometimes do, but I think in the case of Launchpad after lp-login we pass it.
[23:45] <Peng> fagan: How did it wind up using the wrong key?
[23:45] <poolie> i think i've sometimes seen launchpad sftp default to 'mbp@sourcefrog.net' as the username
[23:45] <fagan> It said the number and I just matched it
[23:45] <poolie> i haven't dug into why
[23:46] <fagan> I just looked at the file and looked at launchpad and went I uploaded the wrong one
[23:48] <fagan> I feel nice and dumb after that