Team Chat Logs

2006 10
Mo Tu We Th Fr Sa Su
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30      

November 01, 2006

[00:01:00] <alect> coderanger: you're a bit of a setuptools nut...is there a way to make it recognise existing non-setuptools packages?
[00:01:17] <alect> eg. PIL is installed without setuptools
[00:01:34] <coderanger> Is it an egg?
[00:01:53] <alect> no...
[00:01:59] <alect> thus my question
[00:02:46] <coderanger> Well distutils can install eggs too
[00:03:03] <coderanger> Other than that, not that I know of
[00:03:05] <alect> since when?
[00:03:18] <coderanger> I think it can
[00:04:09] <alect> makes it difficult to use setuptools dependencies
[00:04:26] <alect> a lot of systems will have modules installed via the o/s packaging system
[00:04:41] <alect> and the setuptools dependency checks fail
[00:05:47] <coderanger> Yeah, with 2.5 distutils has rudimentary provides/requires support
[00:06:02] <alect> uncool
[00:06:08] <coderanger> But compared to setuptools its almost useless
[00:06:45] <alect> haven't had any spam recently
[00:06:47] <coderanger> You could do a try: import Image except: in the setup.py
[00:06:53] <coderanger> but thats not nearly as nice
[00:06:53] <alect> unless you've been deleting it all :)
[00:06:59] <alect> no it's not :(
[00:07:24] <coderanger> alect: Nope, and I have been checking very often since I was expecting a flood
[00:07:31] <alect> same here
[00:07:36] <alect> bizarre
[00:07:48] <coderanger> I guess trac-hacks just isn't that high-profile ;-)
[00:08:03] <alect> although two people have complained about a backtrace in the regex filter
[00:08:10] <alect> so perhaps the spambots are hitting that as well
[00:08:13] <alect> one can only hope ;)
[00:08:17] <coderanger> It's also probably not linked to from many places, so less spiders will hit it
[00:08:29] <alect> it has quite a high page rank on google
[00:08:40] <coderanger> (as compared to t.e.o and such)
[00:09:32] <coderanger> link:trac-hacks.org shows ~1k links
[00:09:38] <alect> pr 5, t.e.o has pr 8
[00:09:38] <coderanger> for t.e.o its 164k
[00:09:48] <alect> link: is not reliable
[00:09:55] <alect> "trac-hacks.org" is a better option
[00:10:04] <coderanger> heh, you should go fix that ;-)
[00:10:45] <coderanger> In any event, lets not jinx it :)
[00:10:58] <alect> yeah
[00:12:52] <coderanger> At some point maybe all us uber-nerds should just make our own version us askimet
[00:13:06] <alect> i've thought that myself
[00:13:19] <alect> a collaborative one
[00:13:41] <coderanger> Yeah
[00:15:38] <coderanger> At the point that I try integrating ticketdelete into trac itself (probably after workflow), there really needs to be a "flag as spam" checkbox
[00:15:47] <alect> hell yes
[00:17:12] <alect> though that will also depend on some method of template manipulation
[00:18:04] <alect> iniadmin is my friend
[00:18:15] * alect pats iniadmin
[00:18:18] <coderanger> Well if its part of the core, it can probably part of the template
[00:18:41] <coderanger> I should look at the active-plugins-only patch again and see if theres a better way to do it
[00:19:03] <alect> yeah true
[00:19:20] * row has joined #trac
[00:23:25] * nextime has joined #trac
[00:29:42] * jesterKing is now known as amino
[00:32:47] * otaku42_away is now known as otaku42
[00:34:20] <alect> coderanger: can you edit http://swapoff.org/wiki/SandBox
[00:34:28] <alect> see if the captcha works okay
[00:34:34] <coderanger> I see no edit button
[00:34:47] <alect> hrm
[00:35:04] <coderanger> unless I am supposed to log in
[00:35:15] <alect> okay try now
[00:36:03] <alect> i enabled captcha, and it intercepted iniadmin...whoops
[00:36:05] <alect> worked though
[00:36:17] <coderanger> askimet rejected
[00:36:27] <alect> wtf
[00:36:32] <alect> fucking akismet
[00:36:44] <coderanger> I did get a captcha though
[00:37:02] <alect> awesome
[00:37:15] <alect> one sec
[00:39:24] <alect> okay try again
[00:39:28] <alect> last time i promise
[00:40:08] <coderanger> Worked
[00:40:13] <alect> sweet
[00:40:16] <alect> danke
[00:40:19] <coderanger> Though the captchas are pretty hard to read
[00:40:34] <alect> it's true
[00:40:44] <alect> it keeps refreshing though
[00:40:54] <alect> so eventually you'll get one that doesn't suck
[00:41:52] <coderanger> I maintain that the only good solution to spam is a crowbar, and flame-thrower, and a lot of plane tickets
[00:42:09] <alect> yeah
[00:43:02] * cmlenz has quit IRC
[00:48:10] * kelmo has joined #trac
[00:52:44] <coderanger> Time for me try in vain to sleep
[00:52:50] <coderanger> Gnite y'all
[00:54:42] * alect waves
[01:00:55] * cmlenz has joined #trac
[01:16:28] * tkp has joined #trac
[01:44:54] * cboos has joined #trac
[01:48:17] <alect> hey cboos
[01:48:45] <cboos> hello
[01:48:46] <blinx> hi
[01:48:50] <cboos> good morning!
[02:01:35] <kelmo> trac.edgewall.org ticket spam filter is a bit aggressive ;-)
[02:02:19] <cboos> yeah, but no more than spammers are ;)
[02:02:30] <kelmo> too right
[02:02:33] <cboos> kelmo: the trick is to go in the "Settings"
[02:02:50] <alect> that trick won't last long i suspect
[02:02:56] <cboos> and save your name there (at least the name, possibly also your e-mail)
[02:02:59] <alect> spammers will catch on to that
[02:03:24] <cboos> yeah, that's why we should put the captcha in place too ;)
[02:03:46] <cboos> btw alect, why did you feel that every POST had to be intercepted,
[02:03:47] <cmlenz> the settings thing is only a temporary workaround
[02:03:56] <cmlenz> kelmo: see http://groups.google.com/group/trac-users/browse_thread/thread/934d8d6f3abf0928/3274c97b04f45ec6
[02:03:59] <cboos> as opposed to only protecting the settings?
[02:05:01] <alect> well, requiring user settings is a bit hackish right?
[02:05:25] <alect> but actually it's mainly because i haven't written the code to only intercept certain modules
[02:05:32] <kelmo> cmlenz: hmm, interesting
[02:05:35] <cboos> I mean, if the "Settings" thing would be the only way to earn (+) points, protecting that with a captch should be enough, no?
[02:06:10] <cboos> and requiring to have "Settings" doesn't seem to be hackish for me, it's like a "light" registration, basically "tell us you're not a bot before"
[02:06:22] <cmlenz> cboos: you want to enable people to submit stuff without them having to enter settings
[02:06:45] <kelmo> more humane than captcha imho
[02:06:45] <cboos> no I don't want to ;)
[02:06:49] <cmlenz> so ideally the captcha would kick in on the settings page, or if the user has no session, on the individual submission pages
[02:08:05] <cboos> getting the submission rejecting and telling the people to go identify themselves would also work, IMO, and would probably be simpler...
[02:08:18] <cboos> s/rejecting/rejected/
[02:10:24] <cboos> cmlenz: have you seen the updates on #4043?
[02:10:37] <cboos> ... scary, isn't it? ;)
[02:12:08] <cboos> I think I'll post a mail asking for hints on the psycopg2 ML ... what are the version of PostgreSQL and psycopg2 that we use, btw?
[02:16:56] * bethor has joined #trac
[02:18:23] <bethor> hello, short question: Are all uploaded files stored in the database?
[02:19:54] <coderanger> No
[02:20:29] <coderanger> The files themselves end up in the attachments folder in the trac env
[02:20:38] <coderanger> Though there is some metadata in the DB
[02:22:19] <bethor> thanks coderanger . So my jail for trac has to be a little bit bigger than I thought...
[02:23:00] <coderanger> You can restrict the size of attached files in trac.ini
[02:23:17] <bethor> that is probably a good idea.
[02:25:34] <kelmo> can attachments be disabled alltother?
[02:25:40] <kelmo> +ge
[02:27:10] <coderanger> Set the upload size to 1?
[02:28:00] <coderanger> You could also make a filter plugin to intercept uploads and reject them
[02:29:14] * haakonn has quit IRC
[02:29:20] * amino is now known as jesterKing
[02:35:35] * jesterKing is now known as jesterKinky
[02:36:56] * cboos_ has joined #trac
[02:48:31] * s0undt3ch_ has joined #trac
[02:51:36] * Hydra_ has joined #trac
[02:52:22] * cboos has quit IRC
[02:54:11] * Hydra^ has quit IRC
[03:03:14] * cboos_ is now known as cboos
[03:05:43] * s0undt3ch has quit IRC
[03:05:44] * s0undt3ch_ is now known as s0undt3ch
[03:11:09] <cmlenz> cboos: psycopg2-2.0.5.1 on python2.4
[03:11:31] <cboos> ok, thanks. And PostgreSQL?
[03:12:26] <alect> <cmlenz> so ideally the captcha would kick in on the settings page, or if the user has no session, on the individual submission pages
[03:12:34] <alect> ^^ i could add that behaviour to the captcha plugin
[03:13:10] <cmlenz> yeah
[03:13:28] <alect> actually, question
[03:13:57] <cmlenz> hm?
[03:13:58] <alect> i did some research on POST redirects, and it seems that it's not possible in a 100% reliable fashion. is that true?
[03:14:36] <alect> because at the moment, i'm pickling the req.args, preserving it in the session, then unpickling and manually calling handle_request() after the captcha
[03:14:44] <cmlenz> you mean redirecting so that the requested to the redirected location is a POST? I don't think that's possible
[03:14:45] <alect> which, as you can imagine, is not ideal
[03:14:59] <cmlenz> s/requested/request
[03:15:03] <cmlenz> right
[03:15:05] <alect> yeah that's exactly what i meant
[03:15:08] * idnar has quit IRC
[03:15:55] <cmlenz> well, a solid implementation isn't really possible with 0.10 I think
[03:16:11] <alect> hmmm, i could probably have the captcha page POST to the original URL, with all the args as hidden form variables
[03:16:17] <alect> yeah :(
[03:16:35] <cmlenz> with 0.11 you could use the manipulator APIs to add fields to the form
[03:16:39] * idnar has joined #trac
[03:16:47] <cmlenz> that would work as long as you don't have file upload fields
[03:17:00] <cmlenz> (the hidden form fields idea)
[03:17:09] <alect> hmm yeah
[03:17:17] <alect> are the manipulator apis in place yet?
[03:17:31] <alect> i mean the genshi ones
[03:17:47] <cmlenz> genshi manipulators?
[03:18:06] <cmlenz> nothing happened on the manipulator front... I thought that was part of workflow to some extent?
[03:18:35] <alect> ah, nm. i thought you were referring to the "template manipulation"
[03:18:58] <alect> but you are referring to ITicketManipulator et al
[03:19:09] <cboos> (about the POST issues: that's why I think that a simple workflow like: submit -> rejected as spam -> "please go identify yourself" -> settings + captcha, would perhaps be the cleanest solution for now)
[03:20:14] <alect> wouldn't you still have the same problem though? of preserving the form contents?
[03:24:35] <cboos> well, you'll lost it, unless you go 'Back', one time inconvenience (i.e. no better but no worse than what we currently do)
[03:24:44] <cboos> s/lost/loose/
[03:25:07] <alect> yeah
[03:25:21] * omry|work has joined #trac
[03:25:26] <alect> using the manipulator apis to add verification would be ideal, but it's not there yet
[03:25:42] <omry|work> check out this spam: http://firefang.net/projects/firestats/ticket/30#comment:6
[03:25:46] <omry|work> passed the spam filters
[03:26:05] <omry|work> the bastard changed all sorts of fields, I think this is why it passed.
[03:26:49] <omry|work> is it possible to delete a ticket comment?
[03:26:56] <alect> ticketdelete plugin on trac-hacks
[03:27:02] <alect> it will revert all those fields too
[03:27:04] <alect> invaluable
[03:27:14] <omry|work> my first ticket spam.
[03:27:17] <omry|work> thanks
[03:27:19] <alect> if you're brave, you could also try the captcha plugin ;)
[03:27:26] * alect pimps captcha
[03:27:35] <omry|work> I don't like captchas..
[03:28:43] <omry|work> its not citical enough yet, if it gets bad I`ll install the captcha
[03:28:51] <cboos> well, what about the "mathematical" captcha, simple, not intrusive, effective (for now...)
[03:28:55] * rsalveti has joined #trac
[03:29:42] <omry|work> well, I still don't like them.
[03:30:04] <cboos> may I ask why?
[03:30:39] * omry|wor1 has joined #trac
[03:30:54] * cboos looks around, fearing to see alec waving his trout...
[03:31:59] <alect> haha
[03:32:03] <alect> i'm curious as well
[03:32:17] <alect> the only decent reason i've heard is accessibility
[03:35:56] <omry|wor1> ticketdelete works great
[03:36:01] <omry|wor1> thanks.
[03:36:26] <omry|wor1> about the reason - I don't think we should bother humans to prove their identity if we can identify bots automatically.
[03:36:47] <cboos> obviously we can't ;)
[03:36:52] <alect> heh
[03:36:57] <alect> so true
[03:37:33] * omry|wor2 has joined #trac
[03:37:38] <cboos> I mean, even if you improve spam detection, it's an arm race, so attacking the problem from different front seems to always a winner
[03:37:47] <cboos> s/to /to be/
[03:38:14] <cboos> i.e. check that you're not a bot _and_ prove you're a human
[03:38:39] <cboos> even with that, we won't reach 100%
[03:38:52] <omry|wor2> cboos: so its just a matter of percentage.
[03:39:08] <omry|wor2> if I can get good enough results without bothering the user, I will.
[03:39:16] <omry|wor2> if its not good enough, I`ll bother the user.
[03:39:35] <cboos> not only that, I think the user is much more bothered by seeing his input rejected as spam than being shown a simple way through...
[03:40:10] <omry|wor2> cboos: this can be solved by challenging the user with a captcha IF he is detected as a bot.
[03:40:27] <cboos> ... that's what I'm arguing for, actually ;)
[03:40:28] <alect> that was suggested by cboos about 1 minute before you joined
[03:40:44] <omry|wor2> its a good suggestion in my opinion.
[03:40:53] * hpnadig_ has joined #trac
[03:41:34] <omry|wor2> 99% of the users will never be bothered with the captcha, and the 1% that would will have an easy way to say: hey, I`m a human!
[03:41:45] <cmlenz> captchas are still an arms race
[03:42:02] <omry|wor2> cmlenz: well, only if you use humans to solve them.
[03:42:16] <alect> people are solving them algorithmically
[03:42:17] <cmlenz> well, machines will get there
[03:42:20] <alect> neural nets etc.
[03:42:25] <cboos> he, the 2+2 = ? (hint: 4) is there since ages on CvsTrac ...
[03:42:28] <omry|wor2> alect: not any time soon.
[03:42:33] <alect> there are a couple of sites dedicated to it
[03:42:52] <omry|wor2> alect: besides, if spammers will push AI and improve it, maybe its a good thing.
[03:42:53] <alect> omry: sooner than you think: http://sam.zoy.org/pwntcha/
[03:43:26] <alect> just need to make sure the captchas are difficult enough, but even those will be cracked eventually :(
[03:43:30] * alect curses spammers
[03:43:54] <omry|wor2> alect: this site seem to be able to solve specific captchas.
[03:44:12] <omry|wor2> its not a real challenge to fight it.
[03:44:36] * cmlenz hates difficult captchas... a royal pita
[03:44:46] <omry|wor2> well, use images.
[03:44:54] <omry|wor2> I mean, pctures
[03:44:57] <cmlenz> erm, difficult image captchas
[03:45:03] <cmlenz> that's what I mean
[03:45:16] <cmlenz> where you have to guess what the text is, and get it wrong 2 times
[03:45:28] <alect> somebody has to classify the pictures though, and there can only be a limited set because they have to be done manually
[03:45:30] <omry|wor2> for example, show a picture, and have a question like : how many trees are in that picture.
[03:45:41] <omry|wor2> thats double captcha.
[03:45:48] <alect> ah
[03:46:00] <omry|wor2> you need visual recognision, and natural language decoding
[03:46:03] <cmlenz> so what about visually impaired users
[03:46:19] <alect> you'd need to make it say "ten trees"
[03:46:19] <omry|wor2> have a voice capcha for them
[03:46:27] <alect> effort
[03:46:29] <cmlenz> oh come on
[03:46:39] <omry|wor2> thats a general problem with capchtas
[03:46:40] <alect> tough luck if you don't speak english :)
[03:46:42] <omry|wor2> not just this one.
[03:47:02] <alect> i'll be interested to see how well a bayesian filter works with trac spam
[03:47:08] <omry|wor2> to remind you, I don't like captchas at all.
[03:47:18] * omry|work has quit IRC
[03:47:35] <cmlenz> alect: well, then help out on SpamFilter... it has an experimental bayesian filter and desperately needs testing
[03:47:50] <alect> cmlenz: how do you train it?
[03:48:00] <cmlenz> from webadmin
[03:48:05] <alect> aah
[03:48:12] <alect> does it still work with 0.10?
[03:48:13] <cmlenz> but you currently only can use existing submissions
[03:48:16] <cmlenz> yes
[03:48:30] <cmlenz> still need to add better ways for training, it's all pretty basic right now
[03:48:31] <alect> cool, happy to test
[03:48:59] <alect> does it train+delete or just train?
[03:49:17] <cmlenz> just train
[03:49:28] <cmlenz> have a look... it has a separate logging table
[03:49:57] <alect> ok
[03:53:56] * omry|wor1 has quit IRC
[03:55:31] * Hydra^ has joined #trac
[03:56:47] * omry|work has joined #trac
[03:59:46] <alect> cmlenz: did you consider divmod's reverend?
[04:00:07] <cmlenz> nope
[04:00:16] <cmlenz> didn't even know about that one
[04:00:46] <alect> it has a nice api
[04:01:04] <alect> but spambayes works, so ... :)
[04:04:16] <cmlenz> well, I'll look into reverend
[04:04:24] <cmlenz> looks pretty simple
[04:07:11] <alect> very much so
[04:09:07] * agile has joined #trac
[04:12:59] <alect> what version of spambayes are you using?
[04:14:01] * omry|wor2 has quit IRC
[04:16:26] <cmlenz> 1.1a3
[04:19:31] <alect> ok. 1.0.4 does not work
[04:19:39] <cmlenz> oh
[04:19:54] <cmlenz> didn't even try that
[04:20:00] <alect> ohmm
[04:20:05] <alect> maybe it's not that...
[04:20:11] * kelmo has left #trac
[04:21:02] <alect> http://paste.lisp.org/display/29086
[04:21:38] <cmlenz> hmm
[04:23:26] * Hydra_ has quit IRC
[04:25:31] * imajes has quit IRC
[04:26:30] * imajes has joined #trac
[04:34:35] <alect> cmlenz: i'm going to whip up a patch integrating those captcha methods as a fallback when spam detection matches
[04:40:02] <cmlenz> yeah, that may be nice
[04:40:19] <cmlenz> you mean for the spamfilter plug, right?
[04:40:28] <alect> yeah
[04:59:05] * shawn has quit IRC
[05:00:01] * shawn has joined #trac
[05:04:39] * jesterKinky is now known as jesterKing
[05:06:23] * Arnar has joined #trac
[05:06:29] <Arnar> hi folks..
[05:06:37] <Arnar> cmlenz: busy?
[05:06:57] <cmlenz> sort of... what's up?
[05:07:30] <Arnar> I upgraded my trac a few days ago.. and now I can't see the source of python files in my repository..
[05:07:39] <Arnar> they look like binary files would
[05:07:42] <Arnar> just the last commit message
[05:07:52] <Arnar> I have to click "plain text" or "original format" to get them
[05:08:12] <Arnar> is there some configuration I need to change.. perhaps wrt to mime-types or similar
[05:08:32] <Arnar> just ignore me if you're too busy - it's not a showstopper for me :)
[05:08:47] <cmlenz> definitely check the log
[05:08:50] <cmlenz> @logging
[05:08:50] <evil_twin> cmlenz: "logging" is http://trac.edgewall.org/wiki/TracLogging <-- Enable debug logging to file, ensure your environments log/ directory is writeable by your web server user, check for errors.
[05:08:59] <Arnar> nice.. thanks
[05:15:38] <sam`> hi
[05:15:52] <sam`> building trac svn fails with : error: package directory 'trac/scripts' does not exist
[05:16:23] <cmlenz> oh
[05:16:52] <cmlenz> thanks sam`
[05:16:55] <cmlenz> I'll ix that
[05:16:57] <cmlenz> fix, even
[05:17:32] <sam`> webadmin is now in the core, right ? i can safely remove the TracWebAdmin egg ?
[05:18:07] <sam`> btw, i was wondering : is edgewall a company or just a developers group name ?
[05:18:43] <sam`> the about page is kinda ... 404 :)
[05:19:45] <cmlenz> edgewall.org is "edgewall.org is a place where a community of software developers collaborate on creating exciting open source software based on the Python programming language. It is best known for the popular web-based project management tool Trac."
[05:20:11] <cmlenz> "Edgewall" itself is a one-man company in Sweden run by jonas
[05:20:29] <cmlenz> which basically only does edgewall.org hosting and adsense
[05:20:39] <alect> what happened to the other edgewall people?
[05:20:45] <alect> (.com)
[05:21:04] <cmlenz> there was only daniel lundin, and he moved on to other things (music, mostly)
[05:21:13] <alect> aha
[05:21:27] <sam`> ok
[05:21:51] <cmlenz> sam`: fixed now in svn
[05:21:58] <sam`> i'm looking for a 6 months internship, edgewall sounded interesting, but i don't think it employs interns :)
[05:21:59] <cmlenz> yeah, webadmin is now integrated
[05:22:02] <sam`> cmlenz: thx
[05:22:16] <cmlenz> heh, you'd probably be sitting alone in jonas apartment all day long ;-)
[05:22:22] <sam`> :P
[05:22:30] <sam`> cmlenz: sweet, works much better now :)
[05:26:19] <cmlenz> hmm, I'm gonna try and disable akismet on trac.edgewall.org for a couple hours
[05:26:39] <cmlenz> with ip blacklisting and limiting of external links we're catching a lot of spam without akismet' help
[05:27:05] <sam`> cmlenz: timeline's giving me a 500 ISE here, but only on a particular env
[05:27:06] <alect> it's been disabled on th for 2 or 3 days
[05:27:12] <alect> haven't had much spam at all
[05:27:44] <cmlenz> I suspect th may be getting a different level of spam than teo ;-)
[05:27:58] <alect> assuredly
[05:28:13] <cmlenz> teo has a page rank of 8, making it very attractive to link spammers
[05:28:14] <alect> but it was getting at least a couple a day
[05:28:24] <alect> yeah. th is 5
[05:29:35] <alect> one annoying thing about setuptools is that for its dependencies to work, all packages have to be installed with it
[05:29:49] <alect> which fails with typical system packages
[05:29:56] <sam`> damn, don't understand, there's no error in the log
[05:29:59] <sam`> even in DEBUG mode
[05:30:43] <cmlenz> what about the apache error log?
[05:31:05] <sam`> nothing much, not even a tracback
[05:31:33] <sam`> [Wed Nov 01 14:23:32 2006] [notice] mod_fcgid: process /usr/share/trac/cgi-bin/trac.fcgi(13419) exit(communication error), terminated by calling exit(), return code: 0
[05:33:12] <sam`> ticketdelete refuses to work too, now that webadmin is in the core
[05:34:04] <cmlenz> ah yeah... ping coderanger about that one, or asmodai
[05:34:18] <sam`> 2006-11-01 14:22:19,422 Trac[__init__] ERROR: Skipping "TracTicketDelete 1.1" ("TracWebAdmin" not found)
[05:34:26] <sam`> coderanger: ping ?
[05:35:58] <sam`> TicketDelete has TracWebAdmin as a dependency
[05:36:08] <sam`> but it won't find it as an egg since it's in trac
[05:37:27] <sam`> here
[05:37:28] <Arnar> cmlenz: I enabled logging.. it says the mime type is text/x-python and that it's trying to render a html preview with SilverCityRenderer.. then there is no error -- just to make sure I downloaded SilverCity 0.9.5 again and installed it but that didn't help..
[05:37:37] <Arnar> there's no error in the log however..
[05:37:48] <sam`> asmodai: you should consider removing install_requires = ['TracWebAdmin'] from ticketdelete's setup.py
[05:38:01] <cmlenz> sam`: TicketDelete will need to be changed to work with the webadmin in core
[05:38:11] <cmlenz> simply removing the dependency won't be enough
[05:38:27] <alect> sam: log a ticket on trac-hacks.org
[05:38:28] <sam`> damn :D
[05:38:31] <asmodai> sam`: k
[05:38:37] <sam`> asmodai: read up
[05:38:44] <sam`> might not be enough apparently
[05:38:56] <asmodai> yeah, but regardless it can be removed anyway
[05:39:08] <cmlenz> obviously the package names changed, from webadmin to trac.admin
[05:39:21] <cmlenz> and then some names in the interface changed, too
[05:39:39] <cboos> cmlenz:
[05:39:42] <cmlenz> asmodai: have a look at the tracspamfilter/admin.py which supports both 0.10 and 0.11
[05:39:48] <asmodai> cmlenz: I got a small meeting in a few minutes, but should be able to hack a bit after that
[05:39:54] <asmodai> cmlenz: That's the sandbox is it?
[05:39:59] <cmlenz> yeah
[05:40:16] <cboos> @wiki TracDev/ApiChanges/0.11
[05:40:16] <evil_twin> cboos: http://trac.edgewall.org/wiki/TracDev/ApiChanges/0.11
[05:40:25] <sam`> i'll be looking for updates tonight, thx
[05:40:30] <cboos>