[00:00] GaryvdM: the index layer does have a cache of its own, and the majority of the cost of traversing is disk io and parsing, yes. [00:00] Ok [00:00] GaryvdM: but with a big enough data set, the cache can be exhausted, which would lead to duplicate IO [00:01] good morning [00:01] So should I send a patch? [00:01] GaryvdM: if you've written the code, sure - we can examine it [00:01] ok [00:03] spiv, lifeless, call in a sec [00:05] ya but that's at the cost of higher latency [00:05] whops [00:05] *whoops [00:05] wrong channel [00:06] Spaz: that would be an appropriate comment in here /quite/ a lot of the time :) [00:07] haha [00:08] mwhudson, it was related to OS dev [00:08] not sure you would be interested :p [00:16] poolie: sure, call is fine [00:20] calling you... [00:54] hi, which is the best way to: If I delete a file, to get it back? [00:55] Macarse, bzr revert FILENAME [00:58] poolie: thanx [03:07] poolie, spiv: how is the work on the smart server working with stacked branches? [03:10] thumper: poolie knows the most about that I think [03:10] spiv: ok, thanks [03:10] afaik the branch is approved [03:10] * mwhudson pokes around in bb [03:10] http://bundlebuggy.aaronbentley.com/project/bzr/request/%3Ce01316480809050416r72b3f857h4a664a2cc4e157b0%40mail.gmail.com%3E [04:18] thumper, not sure if you saw that [04:18] my net connection seems to be flapping again... [04:18] thumper, bug 261315 is in review, i'll merge it for 1.7 [04:18] Launchpad bug 261315 in bzr "getting a stacked branch over the smart protocol fails with "Could not install revisions"" [Critical,Fix committed] https://launchpad.net/bugs/261315 [04:45] poolie: great [04:48] * igc lunch [05:52] Hello fellow Bzr users. I have a strange problem, and would need some assistance. [05:54] I setup a shared repository with the 1.6.1-rich-root format. [05:54] My partner created a local branch on his machine with just bzr init, and then pushed to that repository. [05:54] Doing a bzr info on this remote branch (inside that shared repository) indicates the branch format is "unnamed" [05:55] they're using 1.6? [05:55] Trying to branch that remote branch on my machine then yields a ERROR: KnitPackRepository (url) is not compatible with RemoteRepository(url) error. [05:55] Yes, they should be. [05:55] They downloaded the latest Bzr for Windows binary. [05:56] So, after that first failure, I bzr upgrade --format 1.6.1-rich-root inside the shared repository branch [05:56] It seemed to succeed, and a bzr info on it now reports the format is 1.6.1-rich-root [05:56] however, trying to branch from that yields the same error, with the addition of the error message "different rich-root support" [05:58] you get that error when trying to branch using 1.6.1? [05:58] correct [05:59] Is there a shared repository locally? Why are you using 1.6.1-rich-root? [05:59] The shared repository is on my machine, so it is "local" in that sense [05:59] I have no idea :p [06:00] It seems it is required for bzr-svn support, but I am quite new at Bzr. Should I just leave it at the default format? [06:00] THe "different rich-root support" message is saying that the local repository is in a format that's incompatible with the remote format. [06:00] hum [06:01] Ah, if you're using branches from bzr-svn, then yes you do need a format with rich-root support such as 1.6.1-rich-root. [06:01] if there is no local repository, will the branch command create one? [06:01] And if so, it would likely pick the default format [06:01] leading to that error [06:01] Right. (Even a standalone branch has a repository, it just isn't shared with any other branches.) [06:01] And my current situation is a pure Bzr scenario; in that case, which format is preferable? [06:02] If you're not using any plugins, then the default format created by "bzr init" is fine. [06:02] yeah, so I would have to bzr init --format... then branch from inside there, provide the --format to the branch command [06:02] noted, thank you [06:03] Choosing a non-default format is really a "you should understand why you are doing this" situation :) [06:03] haha, or a RTFM :p [06:49] goood morning bazaar ! [06:50] hola [06:50] hello vila! [06:50] lifeless, i finished your book, it was pretty good [06:51] cool [06:51] poolie: that's the second time you said so, now I'm curious :) What is this book about, what's the title ? [06:52] "Leadership and Self Deception" [06:52] http://www.amazon.com/Leadership-Self-Deception-Getting-Out/dp/1576751740 [06:55] thks :) [06:57] i can give you a synopsis but i think robert is reluctant to disclose any spoilers [06:58] the narrative form perhaps makes it a bit hard to express other than in examples but i think the key message is [06:58] consistent with writers about mindfulness [06:59] hehe, just read the abstract and bookmarked it for later, if you both enjoyed it, it's quite enough reviews to try it :) [06:59] that if you notice you're doing something unpleasant, (making yourself) feel bad, or frustrated with other people [06:59] it may be a clue that you're twisting yourself up to justify something that's not actually true, or a decision you don't really believe in [07:00] 'doing something unpleasant" ? Because *you* feel that unpleasant or others feel that unpleasant ? [07:00] well, either, but i had in mind [07:01] the former [07:02] I see [07:03] for instance, sometimes i keep working late on something in a bloody-minded way, without really enjoying it or necessarily being more productive [07:04] if you asked me at this moment i'd say it's because i really want to fix this bug or whatever it is [07:04] they suggest that if i thought about it more clearly, my motivation might really be to prove to myself that i work hard, or something like that [07:08] lifeless, would you agree with that summary? [07:09] quality vs quantity, a life's work ;) [07:12] poolie: its an interesting take on it, its not the take I have [07:12] vila: I suggest grabbing it and reading it :) [07:13] * vila stretch his arm across earth but fails to grab :D [07:43] so what does one do when bzr says that it is unable to obtain lock (for 65 hours, 53 minutes) and the bzr break-lock command doesn't work? [07:43] kwwii: break-lock should work [07:44] Does it give an error, or just silently fail to break the lock? [07:44] spiv: it says bzr ERROR: Unsupported protocol for url ... [07:44] kwwii: ah, let me guess, this is on Launchpad? [07:45] spiv: yes :-) [07:45] kwwii: Don't use the "lp-1234://..." URL with bzr break-lock, just use the regular URL you use to access the branch. [07:46] e.g. "lp:foo" or "bzr+ssh://bazaar.launchpad.net/...". [07:46] spiv: going to need a bzr dev's help to fix that bug :) [07:46] The error message that suggests break-lock is a bit misleading, it's a known bug. [07:46] mwhudson may even have the bug number handy ;) [07:47] (The error message that suggests *the wrong URL* for break-lock, that is. Suggesting break-lock is entirely appropriate.) [07:47] https://bugs.edge.launchpad.net/bzr/+bug/250451 [07:47] Launchpad bug 250451 in bzr "bzr suggests wrong URL for break-lock with a LP hosted branch" [Undecided,Confirmed] [07:48] spiv: wonderful! that worked, thanks :-) [07:48] you made my morning [07:48] Who needs search APIs when you have lp devs to find bugs for you? :) [07:48] kwwii: glad I could help. [07:48] spiv: i'm sure you know this, but the programmatic api for locks in bzr is *terrible* [07:49] Yeah. [07:49] "Patches welcomed." [07:50] unfortunately it smells like an area which risks "patches argued about for weeks" [07:51] That's why I didn't say "Patches accepted" ;) [07:51] meh [07:51] I don't think it would be argued over [07:51] Actually, I don't think it's so terrible. [07:51] Oh boy, I'm totally going to use that in my next big public keynote rant :) [07:51] (also, no time, as usual) [07:51] (But I wouldn't want reality to get in the way of a joke) [07:52] spiv: (no fear :)) [07:59] night everyone [08:02] mwhudson, what kind of thing are you talking about with the lock api? [08:11] poolie: breaking a lock requires installing a ui factory and using a specific lock id requires setting an environment variable are the ones i remember [08:12] mwhudson: environment variable?! [08:12] lifeless: BZR_EMAIL iirc [08:12] lock ids should be random always [08:14] well not lock id perhaps [08:14] the in "this branch was locked at 8:54pm by " [08:16] oh sure [08:16] biab [08:25] heh - 'bzr co -v' doesn't appear very verbose - it is still silent if the checkout succeeds. [08:28] if a working tree falls in the woods etc etc [08:33] jml: its a bear ? [08:41] * jml could do with a beer [08:41] not sure what I'd do with a mute, narcoleptic forest bear though. [08:42] lifeless: want to talk about testresources? [08:42] jml, want a quick call? [08:43] poolie: sure, why not. [08:44] jml: if we're quick, raid is at 6 [08:44] i prefer raid 1... [08:45] sunwell time :) [08:45] >>>>> raid 1 [08:47] lifeless: poolie got me first. some other time. [08:47] hokay === Stellaris_ is now known as Stellaris__ === Stellaris_ is now known as Stellaris === jakobb is now known as jakob [09:35] Hi to all! [09:36] I don't know if this is the right place, but I have a very strange error using bazaar and I didn't found anything on google or forum.... [09:37] This is the error: [09:37] darkstar:/opt/ezechiele# bzr check [09:37] Checking working tree at 'file:///opt/ezechiele/'. [09:37] bzr: ERROR: not well-formed (invalid token): line 10892, column 74 [09:38] Any suggestion? [09:38] d3ko: please file a bug on launchpad.net including the traceback from ~/.bzr.log, and post the number here? [09:40] ok, I do it asap and I post the number. Thank you very much. [09:51] hi, I tried to use "bzr shell" on windows but invoking this command returns with "ERROR: No module named readline". I did some research and it turns out that the python installation for windows does not ship with such a library (http://pypi.python.org/pypi/readline/2.5.1): "[...]f you are using Windows, which also ships without GNU readline, you might want to consider using the pyreadline module instead, which is a readline replace [09:51] jonnydee: You got cut off at 'readline replace'. [09:51] But that sounds like something worth filing a bug for. [09:51] Assuming, of course, that one isn't already there. [09:52] Odd_Bloke: what do you mean with "cut off at 'readline replace'"? Is it the python line in Bazaar which produces this error? [09:53] jonnydee: each line in irc can only be so long [09:53] jonnydee: so your long line got truncated [09:54] ok, I see. So here is my complete text again: [09:54] hi, I tried to use "bzr shell" on windows but invoking this command returns with "ERROR: No module named readline" [09:54] I did some research and it turns out that the python installation for windows does not ship with such a library (http://pypi.python.org/pypi/readline/2.5.1) [09:54] "[...]f you are using Windows, which also ships without GNU readline, you might want to consider using the pyreadline module [09:54] instead, which is a readline replacement written in pure Python that interacts with the Windows clipboard." [09:55] that's it :) [09:55] Right. [09:55] File a bug. :) [09:55] ok, so I will look for a corresponding bug and file one if no one does exist [09:56] thanks for your help :) [10:01] poolie: ok, I posted the bug. Bug number is 267670, link: https://bugs.launchpad.net/bzr/+bug/267670 [10:01] Launchpad bug 267670 in bzr "bzr check and revert fails with "ERROR: not well-formed (invalid token)"" [Undecided,New] [10:08] Odd_Bloke: I've posted the bug now: https://bugs.launchpad.net/bzr/+bug/267674 [10:08] Launchpad bug 267674 in bzr "Invoking "bzr shell" on Windows returns "bzr: ERROR: No module named readline"" [Undecided,New] [10:08] markh: ^ [10:11] Anyone know of a way of making urllib NOT use a proxy on windows? [10:12] jonnydee: You need to install pyreadline then it works. [10:12] so pyreadline is compatible with gnu readline? [10:13] jonnydee: I can't recall, I did this a while ago [10:13] ok, I will try it... thanks a lot for your help [10:14] http://ipython.scipy.org/moin/PyReadline/Intro [10:14] But possibly not windows... [10:14] Oh, wait, PyReadline is for Windows... [10:14] hey, wow!! it works!!! :) :) :) [10:15] So what should I do with the bug report? [10:15] Maybe put a comment on it suggesting that Windows users should install PyReadline? [10:16] (the error message could say this too) [10:16] d3ko: what is in the directory in the non-working example?? [10:16] Maybe change it such that Bazaar should somehow tell windows users to download and install this library in order to mak the 'shell' command work [10:16] ok [10:16] d3ko: since the bzr add command gives no feedback, it seems that you're committing an empty branch, which shouldn't work to begin with [10:17] d3ko: at least, I get bzr: ERROR: no changes to commit. use --unchanged to commit anyhow [10:17] jakob: no, the add command works well and gives me complete output [10:17] so what is the output?? [10:17] jakob: in the log there is the complete output, I put the [...] because the project is quite big [10:18] :P aah, went too fast :P [10:18] :) [10:21] d3ko: could it have something to do with keymapper/priva.####OD# [10:22] where the #'s are some strange characters that my browser can't print?? [10:23] jakob: yes, maybe [10:23] i try to delete that file, is not fundamental [10:24] d3ko, jakob: yes, I'd say that filename is likely to be the problem [10:25] jakob, spiv: YES! check now is working! [10:25] I try some change and then revert [10:25] I don't think 0x08 in ASCII can be represented in XML files, and bzr uses XML for recording the inventory. [10:26] XML? Urrrgh. [10:26] bzr ought to give a better error in this case, and give it sooner. [10:26] And use less XML!!!! [10:26] awilkins: +1! [10:26] awilkins: yeah, there's a new inventory format that gets worked on from time to time. [10:27] spiv, jakob: yes, all seems working! [10:27] If you ask on the list poolie or someone can probably post an update on where it's at. [10:28] I know that SVN went from XML wc metadata to K/V pairs and gained lots of performance [10:28] That was a spurious file, I think too that bzr ought give an error earlier [10:30] Thank you all!! So fast :) [10:30] I think SVN used XML alot more heavily than bzr is using it. [10:31] Maybe file names should encoded in some way, like Base64? [10:31] spiv: yeah, it was littered all through-out the working copy. It's still used between mod_dav and the client too. [10:31] IIRC it's certainly visible in the profiles of certain operations, but perhaps surprisingly not as dominant as one's "yuck XML yuck" intuition might lead you to expect :) [10:32] I haven't looked at that data recently, though. I know it's something we plan to replace, it just hasn't got to the top of the list of things to do. [10:33] Oh, I believe you. I just have a personal disgust for XML... seems to be everyone's hammer in the corporate environment. :-) [10:34] night all [10:34] jszakmeister: yeah. Our working copy state is in mostly in a single binary file, rather than scattered in lots of little files all through the tree. [10:34] spiv, jszakmeister, jakob: what I have to do with bug report? [10:34] I can give the solution [10:34] Have i to do something else? [10:35] d3ko: please do explain your workaround, i.e. that it was just that one file name that was the problem. [10:35] night, poolie! Here is 11:35 AM :) [10:35] d3ko: that's all you need to do [10:35] poolie: g'night [10:36] spiv: Which is a great idea. :-) I've wanted to switch since SVK showed how keep it more central could be a boon. But the WC code is more than I can take on right now. [10:41] Maybe we should do a protocl buffers version... [10:42] I've been meaning to experiment with proto for one of my own concerns - it's presently large, enterprisey, messages in XML [10:42] But I'm not sure the underlying technology deserves to have it's life prolonged by being given a good serialization [10:43] As far as I'm concerned, the more it sucks, the sooner we can move onto something better. [10:43]  [10:45] spiv: ok, I've added a comment with the explanation, I cannot change the status of the bug, maybe you can [10:47] I don't know if you are also developers, but... i love bazaar :) - great software! [10:48] Lots of us here are (including me). [10:48] So, thanks! :) [10:50] d3ko: I've updated the bug status. Thanks for the report. [10:52] Thanks to you all :) - for support and software. Hope this helps other people. [10:53] hi! any bzr-trac devs inhere? [10:53] Now I have to go (I have to fix some bugs in my own software!). Bye! [10:57] i've run into a strange bug and I don't know how to make a proper bug report with it. [10:58] short: I can't see changesets in branches, but in trunk it's ok. [11:13] does anyone have some experience with setting up authentication.conf files? [11:14] i'm trying for http, but bzr doesn't seem to do anything with it [11:16] vila: ^ [11:16] * spiv -> food [11:48] does anyone know what the state of the bzr-git plugin is? === mario_ is now known as pygi [12:01] jakob: what are you trying to do and what doesn't work ? [12:07] Pilky: Very much alpha still, IIUC. [12:07] jelmer would know more. [12:07] ok, thanks [12:15] vila: from the manual I suspected that if I do eg bzr co http://jakob@host/bzr_repos I wouldn't have to type a password, if I have an authentication.conf file with user=jakob,password=pass [12:15] jakob: that's the idea yes [12:16] well..... it doesn't work; maybe I need to set somewhere where my authentication.conf resides? [12:16] *what* doesn't work ? :) and authentication.conf is in ~/.bazaar directory [12:17] aha, the command works, but I still have to type the password [12:17] (actually.... twice) [12:17] that means bzr couldn't find the right section in your file, how does it look ? [12:18] !paste [12:18] pastebin is a service to post multiple-lined texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu.com (make sure you give us the URL for your paste - see also the channel topic) [12:18] [bzr-repos] [12:18] user=jakob [12:18] password= [12:19] where I left out the actual pass.... that's all; I removed all the other options, to make things default [12:20] actually.... I found the answer already [12:20] And it is ? [12:21] I was about to advice using -Dauth so that the section used is shown, that helps debug the authentication.conf file [12:22] well... I'm not sure yet what the precize answer is... [12:23] I went on to try to let the name of the section in the auth file exactly match the name of the domain as set in the .htaccess file [12:23] then i suddenly started working [12:23] but now I changed things back and it still works :P [12:23] the name in the file is just a comment really [12:24] may be you didn't save your file ? (I hate me when I do that :) [12:24] Does using -Dauth shows you the relevant section name ? [12:24] try bzr info or bzr revno [12:25] with the right location added of course [12:25] -Dauth doesn't give any extra info.... [12:26] sorry, it's in the .bzr.log file [12:26] not on the screen [12:27] both info and revno work like a charm [12:27] (I tried them on the remote location; assuming that was what you meant) [12:27] every command should work the same, if revno works and not another, that's a bug [12:28] s/work the same/& regarding authentication :)/ [12:28] well, indeed the [bzr-repos] section is used [12:29] weird... but what the hey.... things work as expected, so I'm happy now [12:29] even though it remains a bit of a mystery what was going on previously [12:29] good, so your file is correct, be aware though, that your section is really a catch-all and will apply to all schemes [12:30] I know.... now that things work again, I'll put in details :) [12:30] ok, let us know how it turns out :) [12:33] ooh... one thing... can I use wildcards in the authentication.conf file [12:33] like eg: host=www.*astro.rug.nl to allow catching www.astro.rug.nl as well as www.intra.astro.rug.nl? [12:35] hmm... apparently not :( [12:36] jakob: hmm, no, but you can use .astro.rug.nl (note the leadind dot) to match the domain [12:36] vila: aah.. thanks! works like a charm :) [12:37] happy to help ((c) jam) === ajaksu is now known as ajaksu_away [12:50] how long should i wait for deletion of a branch? seems like everytime i push it goes to rev. 2 after deleting branch and im trying to make this rev 1 [12:50] i trie --overwrite but still made it rev 2 [13:03] gnomefreak: When you push, it pushes your local revisions ; if your local branch is at r2, a new branch you push will be at r2 [13:03] ah thats right [13:03] thanks [13:03] gnomefreak: If you want it to be r1, make an empty branch and merge your r2 branch into ti [13:04] no r2 now that i think since i want upstream r1 and r2 for this comit [13:04] Or since it's pretty young, why not trash the .bzr folder and re-init/add/commit [13:04] commi [13:06] vila: I see you're one of the main developers of the web-dav plugin? [13:06] yes [13:07] vila: any idea what the following means: [13:07] bzr: ERROR: http://www.intra.astro.rug.nl/%7Ejakobb/bzr_repos/repos1/.bzr/branch is permanently redirected to http://www.intra.astro.rug.nl/~jakobb/bzr_repos/repos1/.bzr/branch [13:07] apache2 humor [13:07] that's what I was afraid of :/ [13:07] you asked for 'blah blah', that's a directory, please use 'blah blah/' [13:08] let me first ask if I use the plugin right: this is what I'm trying: [13:09] bzr push http+webdav://jakob@www.intra.astro.rug.nl/~jakobb/bzr_repos/repos1/ [13:09] if you control the apache2 server you can use 'DirectorySlash Off' to tell him, you're working not joking :) [13:09] nope, I'm just a user.... that's the whole reason for going bzr [13:10] I want my repos accessible over http, such that I can let others access them as well, but still have some control over who accesses them through .htaccess [13:10] it shouldn't error then, can you file a bug with the traceback and the relevant part of the log while running bzr push -Dhttp [13:10] that's very much impossible with svn [13:10] file it against bzr-webdav, not bzr [13:11] what error message should I get if the webserver is not webdav-enabled? [13:11] another one? [13:11] errr, I don't remember exactly, certainly something like 501 [13:11] awilkins: you still here? [13:12] but I don't understand why a redirected leads to an error, that's why I ask for the traceback and the log [13:12] s/redirected/redirection/ [13:12] I'll file a bug asap [13:13] well.... as far as I can tell it's not even a redirection by the way :P [13:13] jakob: one thing you can try is to define a section in your .htaccess (if your server allows it) [13:13] it's just the ~ that is replaced with %7E [13:13] take a look at the NOTES file in the webdav plugin [13:15] hmm, where is that %7e coming from ? Did you type it or does it come from a remembered location ? [13:18] no, it's something that bzr does all over the place [13:19] Seems to me that bzr URL-escapes ~ at the drop of a hat. [13:19] markh: Yes [13:20] vila: what do the Alias /bzr /src/DAV precisely do? Do they need to be changed for my particular installation or is it some standard stuff? [13:21] fullermd, jakob : pedantically speaking '~' in an url should be escaped, that's what bzr does silently (mostly silently) . I'm still wondering why the http server can return an invalid url... [13:21] jakob: forget the alias, it's just a short cut, you don't need it in your case [13:28] vila: before I actually didn't check the NOTES..... very bad, always RTFM :P [13:29] Well, the fine manual is not that good for that plugin :-) Patches welcome :) [13:29] vila: but apparently the 'DAV on' is needed? that actually gives me back a 500 [13:30] I'll have to check with system management first I guess :/ [13:30] DAV on is what *activates* dav, if the server doesn't support it... yes, you need to check with system management [13:31] but if you do, tell them to use DirectorySlash Off at the minimum, DavDepthInfinity on is also needed to support packs [13:33] jakob: rug == royal university groningen ? [13:34] ha, no, rijksuniversiteit groningen :) Should translate as above though... [13:36] vila: well, actually we don't have much to do with the university [13:37] the astro-domain is for astronomy and we have our seperate system [13:37] linux instead of win***&@#$!#@ [13:38] jakob: well, the question was more: *can* you get in touch easily the sys-admins to allow such modifications [13:38] jakob: you connection is noisy, strange chars at the end of your previous message [13:38] and our own web-server....... which should support dav I was promised some time ago [13:38] apache2 ? [13:38] vila: aaaah, yes sure, the're right accross the hall :) [13:38] :D [13:40] but I'm figuring out some details about at least 2 other bugs right now.... sambaserver stuff :P [13:40] Check with them then, and if they are affraid about the DavDepthInfinity, try to find a place where you can store all bzr related stuff, bzr itself uses a depth of max 3 or 4 [13:40] vila: isn't there some nice PHP-function that checks for webdav?? [13:41] the guy that actually admins the webserver is not at the institute today :( [13:41] * vila knows nothing about PHP (not even the meaning of the TLA :) [13:41] oke... I'll find out.... [13:41] and let you know of course ;) [13:42] but looking at the headers returned by the server should tell you (for which you just have to look at .bzr.log after issuing a bzr command with -Dhttp :) [13:42] except if the server is hiding itself... [13:44] in the last 2 years the system has been breached 3 times... so it just might do that [13:52] vila: DAV/2 it says [13:53] full header is ? [13:53] Server: Apache/2.0.52 (Red Hat) mod_ssl/2.0.52 OpenSSL/0.9.7a PHP/4.3.9 mod_python/3.1.3 Python/2.3.4 DAV/2 SVN/1.1.4 [13:55] I don't understand the '500' then :-( [13:55] Probably means it doesn't allow fiddling with that setting in .htaccess. [13:56] I just found a website that also has some tests on it... I'll work through them first: http://www.howtoforge.com/setting-up-webdav-with-apache2-on-debian-etch [13:56] fullermd: was wondering about that, but you know apache better than me :D [13:57] Oh, I dunno about that. But I know how I've provoked 500's before ;) [13:58] http://httpd.apache.org/docs/2.0/mod/mod_dav.html says 'Context: directory' for the 'Dav on' directive. Does that forbids .htaccess ? [13:58] Not necessarily, but it's probably under Options (seems to be the catchall), and that isn't AllowOverride'd for him there. [13:59] (would be my guess, at any rate) [13:59] (which is to say; it's not an Apache limitation, so much as a that-Apache-config limit) [14:00] http://httpd.apache.org/docs/2.0/mod/core.html#directory says 'Context: server config, virtual host' [14:00] I read that at forbidden in .htacess (which doesn't contradict your point) [14:02] On the other hand the Dav Directive example is inside a 8-/ [14:04] vila,fullermd: this cadaver-program says: 405 Method Not Allowed [14:04] (that's without the DAV on, just to see that it's maybe automatically enabled) [14:05] with the DAV on, I immediately get an 500 [14:05] For what request ? (-Dhttp... ) [14:08] ooh, that's in another program, called cadaver [14:08] that's apparently some sort of webdav- [14:08] interface [14:08] sort of ftp for http-connections [14:09] Is there a way to get bzr-svn to only pull the head revision but keep things in true local .bzr (with no need to reaccess the svn repo) and then to push changes back up? [14:11] i.e. without pulling the whole history [14:12] jakob: ok, but what are you trying to do ? 500 is 'internal server error' I doubt you can debug that without looking at server logs === ajaksu_away is now known as ajaksu [14:22] vila: that's the problem; server logs are not accessible :( [14:22] jakob: :-/ Even by asking kindly ? :) [14:26] vila: well, maybe if he is back at the institute... I usually can get a lot done with the admins, because I find many bugs for them, and often I also manage to find a solution, saving them time ;) [14:27] jakob: ok, keep us informed anyway :) [14:28] and thanks for trying that hard :) Hopefully it will helps others once we understand the limitations. [14:31] vila: I'll keep y'all posted ....... [14:48] If I bzr uncommit, should I be able to find the revision uncommitted in the shared repository? [14:48] yes [14:48] What is the best way to find it? bzr heads doesn't seem to be showing it. [14:49] Ooh, heads --all might help... [15:12] Hi, I'm trying to push my repository to another machine with ssh: bzr push --remember bzr+ssh://... but I get the following error: bzr: ERROR: Generic bzr smart protocol error: bad protocol marker "error\x01Generic bzr smart protocol error: bad request 'bzr request 2'\n" [15:12] What's wrong? Version mismatch? [15:13] Got bzr 0.15 on the target and 1.3.1 on the client :P [15:13] Why doesn't bzr+ssh://somehost/~/someproject work? [15:13] the ~ is not expanded, and I need to provide a complete path to the branch :s [15:23] awilkins: in case you happen to *still* be up - I'd love your review on my patch to solve the transport clone issues I send to the list on (my) Saturday. But for now I must sleep... [15:41] uws: It's a known problem, certainly. Perhaps '~/...'? [15:43] Odd_Bloke: that's just as bad ;) === kiko__ is now known as kiko [16:52] Hey, is it possible to pull all revisions of all branches in some remote repository? [16:54] PeakerWork: there is a multi-pull command in bzrtools, i dont know if that works for your needs or not [16:54] Jc2k: I am just wondering, in the context of comparing "git fetch" with bzr's pull [16:54] yes, im watching that convo :) [16:56] PeakerWork: i think multi-pull will only pull branches you are interested in [16:57] PeakerWork: so i could get branches from all overthe place (not just the mainline repository) [16:57] PeakerWork: and bzr multi-pull will keep them all up to date === ajaksu is now known as ajaksu_away [16:57] Jc2k: but it won't see new branches/etc [16:58] PeakerWork: i think so, but i dont normally care about other peoples branches myself [16:59] of course, when i set up bzr-mirror.gnome.org i did look at what would be needed to let people clone a full repository [16:59] its not much effort, from what i recall [17:02] I much prefer the bzr approach, but I've only used git for a day.. [17:03] when starting with bzr, it was after trying arch, monotone, and a couple of others. I loved bzr because it was the first time I tried things, and they worked exactly as I'd expected [17:03] now with git, I'm puzzled by the CLI again [17:03] i set up bzr-mirror.gnome.org, git-mirror.gnome.org and hg-mirror.gnome.org [17:03] thats a lot of repositories [17:03] git caused me the most pain [17:03] and it still isnt set up right [17:03] maybe Pythoneers minds are wired differently :) [17:04] e.g. http://blogs.gnome.org/johncarr/2008/08/22/git-mirrror-making-it-suck-less/ [17:14] PeakerWork: that entire conversation is annoying. does the entire conversation condense to "we think git is the best because everything is in one folder" ? [17:18] Jc2k: well, there is the pre-fetching of remote branches that might be useful to you later, which is useful.. [17:18] PeakerWork: well, tags would be pre-fetched obviously [17:19] PeakerWork: generally the number of branches to number i care about means that im pulling a lot of noise [17:20] noise is cheap :) [17:20] you can merge the interesting stuff very quickly which is nice [17:21] i suppose [17:21] im not very familiar with the bzr code, but it looks fairly easy to code this up [17:21] i think the reason it hasnt is that the bzr guys have thought about it and just dont agree with the git model [17:22] the switch working tree model is also nice for very large working trees [17:22] performance-wise [17:25] hi [17:25] is there a prepare-ChangeLog.pl with bzr support out there? [17:32] PeakerWork: swithcing working trees is a different problem from "clone a repository" [17:32] LarstiQ: yeah, it is also useful, though [17:32] PeakerWork: it is, and bzr does it as well. [17:34] LarstiQ: that's great, I love bzr. Its still said that git scales better, so I have to put up with it for this large project, instead of bzr :( [17:40] PeakerWork: is that your decision, or collaboration with others? [17:40] PeakerWork: just curious [17:41] LarstiQ: well, a bunch of coworkers here (including me) want to move from perforce into something reasonable (read, dvcs) [17:41] PeakerWork: and the next question is then of course, did you try bzr and find it too slow, or was the decision to use git made based on expecting scaling problems? [17:41] PeakerWork: aha :) [17:41] I can try to convince them of a different decision, but the project is pretty large, and they're very afraid of having it be too slow with bzr, and being committed to a tool that might be too slow [17:41] well, what if it grows too slow as the history grows, rapidly? [17:42] we can't "try" to know that? [17:42] * LarstiQ nods [17:43] PeakerWork: very understandable [17:43] PeakerWork: is it already sizeable? Or do you foresee it growing very fast? [17:44] its pretty large already, file-wise. History-wise, its in perforce and would be hard to extract [17:44] p4 keeps history on a per-file basis [17:44] PeakerWork: there used to be a p4 plugin for bzr, no clue if that is maintained === kiko is now known as kiko-fud [18:03] PeakerWork: if you have some spare time, I'd be interested to know if that still works (and gives satisfactory results) [18:03] I'll try to remember to try that, I'll be off work here for a couple of weeks though === ajaksu_away is now known as ajaksu === mw|out is now known as mw [19:59] I have create a authentication.conf in ~/.bazaar for sftp host, but bzr keep asking me for password, any ideas ? [20:00] evge: I think ssh/sftp allways prompts for user input [20:01] evge: if you have pub/priv keys, you can use a ssh-agent [20:03] yeah, this would be a solution for the future, the initial problem was that olive-gtk poped up an error when I tried to update that branch :/ === Guest21335 is now known as jelmer === kiko-fud is now known as kiko [20:43] jelmer: do you plan on improving subversion error reporting in bzr-svn? I just received SubversionException: (None, 170001) and I spent quite some time to figure it was just Authorization error. ;) [20:44] Snaury__, it should already convert a 170001 error into a regular "Permission Denied" error [20:44] Snaury__, if you get a traceback for 170001, please file a bug about it (and include the traceback) [20:45] Hmm. Then it could be a bug: http://pastebin.ubuntu.com/44642/ [20:45] I was trying to svn-push into a repository where I forgot to edit svnserve.conf and add write rights. [20:46] Snaury__: Looks like we simply need to catch SubversionExceptions there and raise convert_error(ex) [20:49] So, do you want me to file a bug report or will you just fix it without one? [20:49] jelmer ^^ [20:49] Snaury__, please file a bug report [20:50] Snaury__, I don't have time to look into it atm, and I would otherwise probably forget about it [20:50] jelmer: ok, doing it now [20:59] bug #267899 [20:59] Launchpad bug 267899 in bzr-svn "svn-push has uncaught SubversionException 170001" [Undecided,New] https://launchpad.net/bugs/267899 [20:59] Snaury__, thanks! === Snaury__ is now known as Snaury[away] [21:10] Snaury[away], should be fixed in 0.4.13 with a bit of luck === beuno_ is now known as beuno [22:07] hi everyone - I'm just wondering why I cannot a bug concerning the bzr-eclipse-plugin (when entering a comment, eclipse crushes regulary) - does anyone know this one? [22:07] a7p, sorry, you cannot what? [22:08] a7p: just going to ask that ^ [22:08] hi beuno [22:08] hiya Verterok :) how's your first day of vacation? [22:08] ;) I cannot enter a Commit Comment [22:09] a7p, "cannot a bug" -- you mean cannot file a bug? cannot find any references to a previously-filed report on that bug? [22:09] arg, [22:09] sry [22:09] cannot find a bug [22:09] *someone* is always first [22:10] beuno: it's ok...trying to get up to date with bzr-related things :) [22:10] might be you this time. [22:10] Verterok, that may take you a while ;) [22:10] I had this behavior a few month ago, when I tried bazaar-eclipse and is still there, absolutly reproducible - It is impossible to use bzr with eclipse... I cannot believe no one else stubled upon this one. [22:11] a7p: please file a bug. also if you have a stacktrace, that would be great [22:11] a7p: I'll take a look ATM [22:11] beuno: indeed, tons of mails to read [22:11] but if there is no obvious known reason (which I just did not see), I'll gladly file a bug. [22:12] Verterok: thx. [22:12] a7p: I use bzr-eclipse to work on bzr-eclipse and never get into that bug :p [22:13] Verterok: okay, now I feel lonley *g* [22:13] a7p: but if you can reproduce it, attach /.metadata/.log and I'll try to reproduce it [22:14] I am running 1.7dev with eclipse-gandymede on osx - may be a rare setup. [22:14] a7p: not at all. I'm with the exact setup :) [22:14] a7p: OS X 10.4, bzr.dev and eclipse 3.4 [22:15] same, but osx 10.5.4 [22:16] Verterok: dcc, mail or should I attach it to a bug? [22:17] a7p: please attach it to the bug [22:17] okay, l'll send you the LP-id in a minute [22:19] a7p: great!, thanks [22:25] LP: 267924 [22:25] bug #267924 [22:25] Launchpad bug 267924 in bzr-eclipse "Eclipse crushes when a Commit Comment is entered" [Undecided,New] https://launchpad.net/bugs/267924 [22:25] you have to know how to tickle ubottu the right way [22:25] :) [22:25] @ Verterok - hope it helps ... if you have any further questions about my setup, feel free to contact me. [22:26] beuno: been lazy the last months... all work no play ... *g* [22:27] beuno: just out of interest, has the breadcrumbs stuff been looked at? I'm fully prepared to be told that it's naff and needs work. :) [22:27] a7p: ok, I'll take a look as soon I finish a new the merge dialog. thanks [22:27] a7p, you can solve that by getting a job on what you play with [22:28] pickscrape, ah, I suck. I'll get to it this week, I super-promise. [22:28] I've been slacking off on Loggerhead stuff [22:28] no problem :) [22:28] I do foresee there being problems with it. The template stuff is new to me, especially the macro side of things. [22:29] I'm sure we'll work it out pretty quickly, it was shaping up pretty well last time I peaked at it [22:29] tomorrow seems like a good day for me to do some LH work [22:46] Anyone have any idea how I might track down a segfault when running bzr qlog? [22:46] beuno: if you are hireing for anything interessting, feel free to tell me *g* - currently I am doing "flash-shit", but the next paid project is going to be zope3-based ... and I am really happily looking forward to that. [22:48] actionscript and python have a lot in common [22:58] lifeless: but not enough .. [22:58] I also hate the flash-player and the adobe authoring enviroment [22:58] environment [23:59] Good morning.