2007-04-20T00:01:01  <dreimark> bbl
2007-04-20T00:09:27  <xorAxAx> there is a bug in the underlay tgz
2007-04-20T00:09:48  <xorAxAx> it still ships two pages that only differ in case and it has a wrong directory permission for (e383aae383b3e382afe38195e3828ce381a6e38184e381aae38184e3839ae383bce382b8)/revisions
2007-04-20T00:09:51  <xorAxAx> ThomasWaldmann:
2007-04-20T00:13:19  <CIA-28> moin: Alexander Schremmer <alex AT alexanderweb DOT de> * 2006:1339d6fdc4ff 1.7/ (35 files in 6 dirs): Changed tests to use py.test. Also changed the semantics a bit - the test wiki directory is only created freshly if it does not exist.
2007-04-20T00:15:00  <xorAxAx> grzywacz: if you feel like writing tests in SOC, py.test will severely ease it
2007-04-20T00:15:57  <grzywacz> xorAxAx, affirmative.
2007-04-20T00:16:24  <xorAxAx> flowhase: same for you :)
2007-04-20T00:16:29  <xorAxAx> fpletz: ping
2007-04-20T00:22:07  <dreimark> bb
2007-04-20T00:22:45  <flowhase> xorAxAx: i didn't use any unit testing frameworks until now
2007-04-20T00:22:55  <xorAxAx> flowhase: they are dead easy :)
2007-04-20T00:22:58  <flowhase> k
2007-04-20T00:23:13  <flowhase> i always hat problems imagining how to use them in webapp contexts
2007-04-20T00:23:18  <flowhase> s/hat/had
2007-04-20T00:23:55  <xorAxAx> well, if you write module tests, you usually dont need to replay a whole request cycle (but that is possible as well using some third party lib that is not yet integrated)
2007-04-20T00:24:15  <xorAxAx> but if you have some nice classes with some interface that you can test, it is easy
2007-04-20T00:24:30  <xorAxAx> the py.test docs are here - http://codespeak.net/py/dist/test.html
2007-04-20T00:29:24  <grzywacz> Hm, how about using JMeter?
2007-04-20T00:30:11  <fpletz> xorAxAx: pong, but on my way to bed now ;)
2007-04-20T00:30:24  <xorAxAx> grzywacz: whats that?
2007-04-20T00:30:25  <fpletz> something important?
2007-04-20T00:30:41  <xorAxAx> fpletz: there is still the -bug in the search module, for nearly a year now :)
2007-04-20T00:31:05  <fpletz> yep, dreimark already sent me a mail.. will fix it this weekend ;)
2007-04-20T00:31:11  <fpletz> just keep nagging *g*
2007-04-20T00:31:19  <grzywacz> xorAxAx, a piece of java software for stress testing, but can also be used to replay request sequences
2007-04-20T00:31:44  <grzywacz> xorAxAx, http://jakarta.apache.org/jmeter/
2007-04-20T00:32:32  <xorAxAx> grzywacz: hmm, i am not convinced :)
2007-04-20T00:32:39  <xorAxAx> grzywacz: have a look at paste.fixture
2007-04-20T00:32:39  <grzywacz> xorAxAx, sure, just an idea. :)
2007-04-20T00:34:27  <grzywacz> xorAxAx, what's been wrong with tests using python's unittest?
2007-04-20T00:34:34  <flowhase> does hg integrate unit-tests?
2007-04-20T00:35:04  <grzywacz> Um...
2007-04-20T00:35:20  <grzywacz> Why would hg have anything to do with testing? :\
2007-04-20T00:35:38  <flowhase> grzywacz: check-on-commit
2007-04-20T00:35:47  <xorAxAx> grzywacz: they are not so easy to use like with py.test
2007-04-20T00:36:30  <xorAxAx> py.test gives better tracebacks, better overviews, supports skipping, doesnt require calling weird methods (just write a plain assert),  doesnt require subclassing
2007-04-20T00:36:31  <flowhase> omg
2007-04-20T00:36:48  <flowhase> friend of mine just showed me the peter norvig python spellchecker
2007-04-20T00:36:50  <grzywacz> xorAxAx, ok, sounds nice :)
2007-04-20T00:36:52  <xorAxAx> py.test can easily be customized without writing custom test runners
2007-04-20T00:37:02  <xorAxAx> and it can automatically distribute tests to many machines :)
2007-04-20T00:37:08  <grzywacz> woot!
2007-04-20T00:37:31  <xorAxAx> flowhase: hmm?
2007-04-20T00:39:57  <flowhase> http://norvig.com/spell-correct.html
2007-04-20T00:40:02  <grzywacz> I can't connect to jabber.org ;<
2007-04-20T01:03:18  <dreimark> good night
2007-04-20T01:24:40  <grzywacz> xorAxAx, test.py indeed looks nice
2007-04-20T01:24:56  <grzywacz> xorAxAx, anyway, when my branch becomes stable, am I supposed to merge it into 1.7 as well?
2007-04-20T01:26:01  <grzywacz> Hm, maybe I can write test for test.py without bothering about it...
2007-04-20T01:26:02  <xorAxAx> grzywacz: your repo contains a snapshot that was pulled from 1.7
2007-04-20T01:26:38  <xorAxAx> grzywacz: you are encouranged to pull from 1.7 often
2007-04-20T01:26:38  <xorAxAx> merge, and push into your repo
2007-04-20T01:26:38  * grzywacz checks
2007-04-20T01:26:49  <xorAxAx> its called py.test btw :)
2007-04-20T01:26:58  <xorAxAx> because it needs the pylib, called "py"
2007-04-20T01:27:06  <xorAxAx> one of the worst names libraries that exist
2007-04-20T01:27:17  <grzywacz> Wait, I have an old clone locally, as it says 1.6
2007-04-20T01:27:18  * grzywacz fixes
2007-04-20T01:27:51  <xorAxAx> you need to clone again
2007-04-20T01:27:55  <grzywacz> In progress. :)
2007-04-20T01:28:11  <xorAxAx> (or with some branch magic restore to the 1.7 branch)
2007-04-20T01:28:35  <grzywacz> clone is simpler
2007-04-20T01:28:41  <grzywacz> (and safer, because I'm dead tired atm)
2007-04-20T01:29:02  <xorAxAx> yep
2007-04-20T01:29:32  <grzywacz> xorAxAx, is it enough if I test my changes with an instance launched from eclipse, or should I setup a separate instance with apache or something?
2007-04-20T01:30:10  <xorAxAx> grzywacz: you dont need apache, no
2007-04-20T01:30:22  <grzywacz> Good.
2007-04-20T01:30:54  <xorAxAx> good night
2007-04-20T01:31:27  <grzywacz> good night
2007-04-20T01:34:24  <grzywacz> xorAxAx, anyway, I think I'm going to use pyxmpp. It has quite a few feature Twisted Words does not, and probably won't have for some time.
2007-04-20T01:35:01  <grzywacz> Like data forms, that I wanted to experiment with. MUC, which could be handy for a'la CIA reporting.
2007-04-20T01:57:01  <grzywacz> ThomasWaldmann, do you happen to be awake? ;)
2007-04-20T01:59:41  <grzywacz> ThomasWaldmann, hg push prompts me for password
2007-04-20T02:08:45  <_keturn> xorAxAx: have you compared py.test to nose?
2007-04-20T02:17:47  <grzywacz> Night. I'll be back tomorrow (today) evening.
2007-04-20T08:32:20  <ThomasWaldmann> grzywacz: when doing what exactly?
2007-04-20T09:15:15  <grzywacz> I wanted to push changes I merged from 1.7
2007-04-20T09:16:48  <grzywacz> So I did "pull -u" first and then wanted to simply "push" them to my branch.
2007-04-20T09:16:57  <grzywacz> To test if it actually works. ;)
2007-04-20T09:35:54  <grzywacz> bbl
2007-04-20T09:41:21  <ThomasWaldmann> grzywacz: did you set the default-push in hgrc correctly?
2007-04-20T09:43:48  <ThomasWaldmann> grzywacz: especially check whether you used hg@hg.thinkmo.de
2007-04-20T09:44:10  <ThomasWaldmann> i didnt change your access entry except s/1.6/1.7/
2007-04-20T10:46:41  <xorAxAx> _keturn: not specifically,but something that only copies something that has been there for 3 years cant be better :)
2007-04-20T10:47:10  <xorAxAx> _keturn: can you think of anything that nose might do better than py.test?
2007-04-20T11:48:20  <neagulm> moin
2007-04-20T11:51:23  <xorAxAx> hi neagulm
2007-04-20T12:27:33  <dreimark> moin  neagulm
2007-04-20T12:27:58  <neagulm> moin dreimark
2007-04-20T12:28:20  <dreimark>  you have nice weather too ?
2007-04-20T12:28:23  <dreimark> did
2007-04-20T12:30:44  <dreimark> xorAxAx: do you have pushed that feature  'instantly moin-dev improrted to MM'
2007-04-20T12:31:03  <xorAxAx> dreimark: sure
2007-04-20T12:31:47  <dreimark> so its in ThomasWaldmann hand to establish it?
2007-04-20T12:31:55  <xorAxAx> yes, as always
2007-04-20T12:31:59  <dreimark> ;)
2007-04-20T12:33:27  <neagulm> dreimark, related to your email...
2007-04-20T12:34:03  <dreimark> ok
2007-04-20T12:34:36  <neagulm> we don't have to maintain separated db's for every language because mostly a wiki consists of pages in a single language...
2007-04-20T12:35:39  <neagulm> the db will be build via training and the text's in another language will be eliminated with the help of "noise reduction" (Eg. Bayesian Noice Reduction) or detected via SB
2007-04-20T12:38:52  <dreimark> ok, mostly it is used that way, and the language pragma is not used regulary
2007-04-20T12:39:52  <dreimark> I don't know a wiki with an hierachical structure and differnt languages yet
2007-04-20T12:40:41  <dreimark> the only common one with different languages always is MoinMaster
2007-04-20T12:41:26  <dreimark> that is because we use that for all help pages for all languages
2007-04-20T12:42:23  <xorAxAx> other wikis often mix english and a particular language
2007-04-20T12:43:13  <neagulm> mixing different languages is not a problem when SB is trained correctly...
2007-04-20T12:46:33  <dreimark> ok
2007-04-20T12:46:47  <xorAxAx> but SB only supports two categories, no?
2007-04-20T12:48:04  <dreimark> only spam and ham
2007-04-20T12:48:30  <xorAxAx> so its not useful for the goal, right?
2007-04-20T12:48:44  <xorAxAx> neagulm: you want to do n-category classification, no?
2007-04-20T12:48:58  <dreimark> xorAxAx: we discussed to first implement spam detection and to modularize this
2007-04-20T12:49:12  <xorAxAx> modularize what?
2007-04-20T12:49:26  <dreimark> so that an other classifier could be used too
2007-04-20T12:50:31  <dreimark> see moin-dev/2007-04-17
2007-04-20T12:51:32  <dreimark>  neagulm: btw.  that promised wiki page would be fine :)
2007-04-20T12:52:06  <xorAxAx> dreimark: hmm, ok
2007-04-20T12:53:45  <neagulm> dreimark, :) I will write today or tomorrow because I still wait for a email from my Neural Networks profesor about n-way classification, he sugested that we use a forward propagation neural network for text classification...
2007-04-20T12:54:18  <dreimark> ok
2007-04-20T12:54:27  <neagulm> I hope that it is not a problem that some of my school teacher are helping me...
2007-04-20T12:54:38  <neagulm> ..with ideas and their experience
2007-04-20T12:55:08  <dreimark> neagulm:that's welcome, may be they will become MM users too
2007-04-20T12:55:12  <dreimark> :)
2007-04-20T12:55:20  <neagulm> they are :D
2007-04-20T12:57:05  <neagulm> the ones that work at ieat.ro are using zwiki and MoinMoin for their projects (moist of them are formal verification, p-systems, distributed computing) and some of the remaining use the Faculty's Wiki and that one is MoinMoin...
2007-04-20T12:57:08  <neagulm> :)
2007-04-20T12:58:06  <dreimark> xorAxAx: currently we can look up how the whole thing could be implemented for SB and while this is discussed we can search for an available  unicode n-category classification tool too.
2007-04-20T12:58:30  <xorAxAx> dreimark: well, the interface would look different, never assume that you only have 2 categories :)
2007-04-20T12:58:47  <dreimark> neagulm:its always fine to have users :)
2007-04-20T13:00:25  <neagulm> I think that this is one of the greatest rewards of working in a open-source app: having thousands of users all around the world that use your work
2007-04-20T13:10:07  <dreimark> did you have thought about the possibility to share a db, so that we can train on e.
2007-04-20T13:10:13  <dreimark> g. MM
2007-04-20T13:10:34  <dreimark> and not every user has to start from zero
2007-04-20T13:15:02  <neagulm> the problem is that you don't know what content will the user have on his wiki... We just know about MoinMaster
2007-04-20T13:16:39  <neagulm> this could be out initial ham corpus... but again the spammer know about this and could build his spam pages so that SpamBayes thinks that his messages are ham
2007-04-20T13:17:41  <neagulm> I prefer to start with an empty db and Train-On-Error (TOE)
2007-04-20T13:18:07  <dreimark> ok
2007-04-20T13:19:33  <xorAxAx> which is rather troublesome compared to the current solution and ignores the fact that spammers work globally :)
2007-04-20T13:19:40  <xorAxAx> but you can use both methods ...
2007-04-20T13:20:41  <dreimark> a weighted mix of both
2007-04-20T13:23:19  <neagulm> xorAxAx, ... the spamassassin way
2007-04-20T13:58:06  <flowhase> moin
2007-04-20T14:04:16  <neagulm> moin flowhase
2007-04-20T14:33:29  <dreimark>  neagulm: did you fetched your new branch for 1.7 ?
2007-04-20T14:33:45  <neagulm> not yet
2007-04-20T14:34:11  <neagulm> am i still with the old one, currently I'm writing the Ideas page...
2007-04-20T14:39:38  * johill decided to do auth completely different again
2007-04-20T14:39:38  <johill> :P
2007-04-20T15:05:18  <dreimark>  neagulm: fine
2007-04-20T15:05:44  <dreimark> before coding you should have the new one
2007-04-20T15:07:12  <neagulm> moinmoin.wikiwikiweb.de is a bit slow in commiting a page modification, is there a problem?
2007-04-20T15:07:40  <neagulm> I have updated http://moinmoin.wikiwikiweb.de/MarianNeagul/ProjectIdeas
2007-04-20T15:08:20  <neagulm> dreimark, I will make major changes in the following days.
2007-04-20T15:09:58  <dreimark> I've seen it seconds before and Ive subscribed it, Skip I believe has subscribed all pages by a regex
2007-04-20T15:10:34  <dreimark> I will later add my comments
2007-04-20T15:11:20  <dreimark> or questions
2007-04-20T15:16:53  <neagulm> dreimark, It would be nice to have a collaborative antispam system between several MoinMoin instances, the paper “May I borrow Your Filter?” Exchanging Filters to Combat Spam in a Community is useful reading
2007-04-20T15:24:10  <dreimark> neagulm: yeah that is a good way to get the work in trainings splitted and you can start without answering lots of questions. Of course if you want you can start from the beginning but I would prefer that this is not the default. By exchanging filters we will see if we can avoid globally used spamtechnics.
2007-04-20T15:36:23  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2007:6cc2e0544657 1.7/MoinMoin/request/ (WSGI.py __init__.py):
2007-04-20T15:36:23  <CIA-28> moin: add finish handlers
2007-04-20T15:36:23  <CIA-28> moin: finish handlers can be added to the request by any part of the code
2007-04-20T15:36:23  <CIA-28> moin: to do some work when the request has been handled.
2007-04-20T15:36:25  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2008:52c474b8c02f 1.7/MoinMoin/ (i18n/__init__.py request/__init__.py):
2007-04-20T15:36:28  <CIA-28> moin: allow translating things early
2007-04-20T15:36:30  <CIA-28> moin: Authentication methods sometimes need to show strings to the user.
2007-04-20T15:36:32  <CIA-28> moin: These can be translated using the browser language but obviously
2007-04-20T15:36:34  <CIA-28> moin: not using user preferences (since no user object is present yet.)
2007-04-20T15:36:36  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2009:1b14cc05a54a 1.7/ (18 files in 8 dirs): refactor authentication and split out session handling
2007-04-20T15:36:39  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2010:d234621dbf2f 1.7/MoinMoin/user.py:
2007-04-20T15:36:42  <CIA-28> moin: don't store trail for first request
2007-04-20T15:36:44  <CIA-28> moin: avoids storing sessions to disk for requests from users
2007-04-20T15:36:46  <CIA-28> moin: that won't send back the session cookie anyway
2007-04-20T15:36:48  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2011:6cd0e4cc8e18 1.7/MoinMoin/user.py:
2007-04-20T15:36:56  <CIA-28> moin: refactor user.getUserId
2007-04-20T15:36:56  <CIA-28> moin: makes looking up by other keys easier (in the future)
2007-04-20T15:36:57  <CIA-28> moin: email lookup could use this as well
2007-04-20T15:36:59  <CIA-28> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2012:b2c71d9f95e7 1.7/MoinMoin/userform.py: split out userform to avoid using it with the enter key
2007-04-20T16:00:22  <grzywacz> re
2007-04-20T16:00:48  <grzywacz> ThomasWaldmann, yes, I have url of my repo there
2007-04-20T16:00:58  <grzywacz> default-push = ssh://hg.thinkmo.de/moin/1.7-jabber-knowak
2007-04-20T16:01:04  <grzywacz> ah
2007-04-20T16:01:11  <grzywacz> @
2007-04-20T16:01:33  <grzywacz> 8)
2007-04-20T16:01:35  <grzywacz> Thanks.
2007-04-20T16:11:16  <johill> let me know of any auth breakage, it's probably mine :)
2007-04-20T16:11:33  <johill> I've tested http, ssl, mysql and the regular auth, plus openid
2007-04-20T16:13:18  <johill> I have a few more things to polish in openid support but I think I can probably push it this weekend so you can test that too
2007-04-20T16:16:28  <dreimark> MM gots attacked currently
2007-04-20T16:17:09  <johill> ?
2007-04-20T16:17:25  <dreimark> a lot of spanish new pages
2007-04-20T16:28:27  <neagulm> bbl
2007-04-20T16:33:25  <johill> eek
2007-04-20T16:36:41  <dreimark> sometimes I wish I could trigger the surgeprotection for an user
2007-04-20T16:36:58  <dreimark> currently its only for the whole wiki.
2007-04-20T16:37:07  <grzywacz> Or revert changes via jabber with one click as they happen. ;o
2007-04-20T16:37:53  <dreimark> the spammer can do use this way too
2007-04-20T16:38:28  <grzywacz> Maybe...
2007-04-20T16:39:05  <johill> can you have buttons in jabber so it's really just one click?
2007-04-20T16:39:53  <dreimark> one of them entered an email adress JuanNacho
2007-04-20T16:40:29  <dreimark> does one speak spanish? and can send him an mail
2007-04-20T16:47:34  * johill slaps xorAxAx 
2007-04-20T16:48:40  <johill> is it possible to disable xmlrpc?
2007-04-20T16:48:50  <johill> I don't trust login methods that use just the user.id
2007-04-20T16:51:34  <grzywacz> johill, yes
2007-04-20T16:51:49  <grzywacz> johill, thanks to data forms extension
2007-04-20T16:53:01  <johill> cool
2007-04-20T16:53:06  <johill> I wonder if gossip supports it ;)
2007-04-20T16:54:31  <grzywacz> I don't think so.
2007-04-20T16:54:45  <dreimark> talk bbl
2007-04-20T16:54:52  <johill> I guess that can be changed
2007-04-20T16:55:17  <grzywacz> Only dev-psi and tkabber, iirc.
2007-04-20T16:58:09  <johill> k
2007-04-20T17:00:24  <johill> back in a bit, gotta test some linux changes that usually blow up spectacularly ;)
2007-04-20T17:02:00  <grzywacz> :D
2007-04-20T17:02:02  <grzywacz> good luck, then
2007-04-20T17:10:52  <johill> hey, their patch does fix it :)
2007-04-20T17:22:47  <johill> that's a relief, means it's not my fault :)
2007-04-20T17:33:27  <xorAxAx> johill: hmm?
2007-04-20T17:33:39  <xorAxAx> johill: just change it to use the new system
2007-04-20T17:33:57  <xorAxAx> johill: it was as secure as the remaining parts of moin at the time it was implmented
2007-04-20T17:43:40  <johill> yeah, I was looking at it but haven't figured out a good way yet
2007-04-20T17:43:42  <johill> I'll fix it
2007-04-20T17:43:56  <xorAxAx> "a good way yet"?
2007-04-20T17:44:11  <xorAxAx> like a normal moin needs to check a cookie, that one checks a token
2007-04-20T17:45:22  <johill> yeah, but I don't know how to clean up such tokens
2007-04-20T17:45:34  <johill> cookies always have a lifetime, for the tokens we'd also need one and I don't know how long
2007-04-20T17:45:42  <johill> short is probably good enough for wikisync ;)
2007-04-20T17:45:51  <johill> I'll look at it tonight, gotta go now
2007-04-20T17:46:09  <xorAxAx> johill: 15 minutes are ok
2007-04-20T17:46:20  <xorAxAx> it fetches a new token for every transaction
2007-04-20T17:58:36  <Fabi> re
2007-04-20T17:59:05  <dreimark>  xorAxAx: maketestwiki patch ready ?
2007-04-20T17:59:32  <xorAxAx> dreimark: its been in for more than 12 hours
2007-04-20T17:59:43  <xorAxAx> and its working fine for me
2007-04-20T17:59:50  <xorAxAx> (in 1.7)
2007-04-20T18:00:10  <dreimark> aha, 1.7 will you add it to 1.6 ?
2007-04-20T18:01:34  <flowhase> hi Fabi
2007-04-20T18:01:40  <flowhase> back in germany?
2007-04-20T18:01:44  <Fabi> yup
2007-04-20T18:01:57  <dreimark> bbl
2007-04-20T18:02:04  <Fabi> 30 hours of travel from Kona to STR
2007-04-20T18:02:08  <xorAxAx> dreimark: nope
2007-04-20T18:02:20  <flowhase> Fabi: i thought about meeting you in person some time
2007-04-20T18:02:24  <xorAxAx> Fabi: stressing?
2007-04-20T18:02:35  <flowhase> since you live nearby
2007-04-20T18:02:36  <Fabi> not as bad a I thought
2007-04-20T18:03:23  <Fabi> flowhase, sorry, where exactly are you living?
2007-04-20T18:03:29  <flowhase> Fabi: karlsruhe
2007-04-20T18:03:30  <dreimark> xorAxAx: stable and not ready code .... >
2007-04-20T18:03:41  <xorAxAx> dreimark: it should be stable
2007-04-20T18:03:46  <Fabi> xorAxAx, I just have to stay awake until the evening today
2007-04-20T18:03:59  <dreimark> bbl now
2007-04-20T18:04:01  <xorAxAx> Fabi: why?
2007-04-20T18:04:02  <Fabi> flowhase, oh, that's near
2007-04-20T18:04:13  <flowhase> :>
2007-04-20T18:04:23  <Fabi> xorAxAx, to get to the time zone
2007-04-20T18:04:29  <xorAxAx> ah
2007-04-20T18:05:07  <flowhase> Fabi: are you going to linuxtag in berlin? or would you prefer to join us at GPN6 in karlsruhe, june the 1st to june the 3rd
2007-04-20T18:05:11  <flowhase> ?
2007-04-20T18:05:46  <Fabi> flowhase, haven't arrived enough yet to think beyond this week end *sri*
2007-04-20T18:05:58  <flowhase> hehe no sweat
2007-04-20T18:06:28  <flowhase> i don't want to rush onto you
2007-04-20T18:07:49  <Fabi> yes, but we should meet soon and may/june is a good time frame
2007-04-20T18:08:40  <Fabi> I have to leave again, we can talk that WE
2007-04-20T18:09:39  <flowhase> hmm, a busy bunny
2007-04-20T20:15:32  <grzywacz> xorAxAx, is it possible to maintain a tcp connection unrelated to http traffic with mod_python?
2007-04-20T20:16:56  <xorAxAx> grzywacz: i guess that you can spawn a thread, yes
2007-04-20T20:17:12  <grzywacz> Good, memory servers me right, then.
2007-04-20T20:17:50  <grzywacz> Maintaining one connection may be more desirable than starting a new one for every notification. ;)
2007-04-20T20:18:13  <grzywacz> This won't work with cgi, though.
2007-04-20T20:18:37  <xorAxAx> grzywacz: well, you can pool it
2007-04-20T20:18:52  <xorAxAx> so you have a process wide singleton that handles all stuff
2007-04-20T20:19:38  <grzywacz> I'm writing more detailed architecture assumptions right now. :)
2007-04-20T21:00:50  <dreimark> xorAxAx: the package bug is no package bug
2007-04-20T21:03:09  <xorAxAx> dreimark: it is a failing test
2007-04-20T21:03:16  <xorAxAx> caused by a checkin of yours
2007-04-20T21:03:19  <dreimark> xorAxAx:because of MoinMoinBugs/DontOfferToUnpackOdtOpenOpenDocument a package has to end with .zip
2007-04-20T21:03:32  <xorAxAx> dreimark: why?
2007-04-20T21:03:35  <dreimark> http://hg.thinkmo.de/moin/1.7/diff/eb5328be394e/MoinMoin/action/AttachFile.py
2007-04-20T21:03:43  <xorAxAx> dreimark: unpacking zips != installing packages
2007-04-20T21:03:53  <xorAxAx> it has to accept every file extension
2007-04-20T21:04:46  <xorAxAx> (for packages)
2007-04-20T21:05:05  <xorAxAx> and a odt file is not a package
2007-04-20T21:05:08  <dreimark> I change that back,  if package.isPackage() and mt.minor == 'zip':
2007-04-20T21:05:26  <xorAxAx> yes, thats wrong
2007-04-20T21:05:38  <xorAxAx> and unrelated to his issue because OOo doesnt generate package files :)
2007-04-20T21:05:57  <xorAxAx> and mt.minor == 'zip' should go into the zip check, though
2007-04-20T21:06:04  <xorAxAx> (that makes some sense from the user POV)
2007-04-20T21:06:28  <dreimark> yeah, my fault and I do change it on both 1.6/1.7
2007-04-20T21:06:47  <xorAxAx> thanks
2007-04-20T21:06:56  <xorAxAx> let me write a test :-)
2007-04-20T21:07:05  <dreimark> I was believing its always a zip file extension, sorry
2007-04-20T21:07:40  <johill> xorAxAx: ok, I'll make it 15 minutes then, will do it later
2007-04-20T21:07:50  <johill> right now I want to look at the testing stuff and maybe add some auth tests
2007-04-20T21:08:02  <CIA-28> moin: Alexander Schremmer <alex AT alexanderweb DOT de> * 2013:c4cbabe9a02d 1.7/MoinMoin/conftest.py: Added a small comment
2007-04-20T21:08:04  <CIA-28> moin: Alexander Schremmer <alex AT alexanderweb DOT de> * 2014:b8de0a47ad95 1.7/ (21 files in 9 dirs): Merge with main.
2007-04-20T21:17:49  <CIA-28> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2015:41995fbf3d92 1.7/MoinMoin/action/AttachFile.py: AttachFile: bug fixed packages are not restricted to zip extension
2007-04-20T21:19:14  <CIA-28> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2004:6c5d5de31878 /MoinMoin/action/AttachFile.py: AttachFile.py: bug fixed packages are not restricted to zip extension
2007-04-20T21:19:49  <xorAxAx> dreimark: you added package installer methods but didnt update http://moinmaster.wikiwikiweb.de/HelpOnPackageInstaller
2007-04-20T21:21:35  <CIA-28> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2016:e55ae6646935 1.7/MoinMoin/parser/text_moin_wiki.py: text_moin_wiki: fixed nested {{{ bug
2007-04-20T21:22:51  <dreimark> xorAxAx: doing now
2007-04-20T22:02:05  <_keturn> xorAxAx: I haven't really looked at nose myself, I only wonder because in Ian Bicking's comparison of stuff vs turbogears, he said the py.test and nose were very similar, but turbogears was planning on moving from py.test to nose
2007-04-20T22:02:55  <xorAxAx> _keturn: i havent seen a single reason to use nose
2007-04-20T22:02:56  <CIA-28> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2017:dd3a1da58c28 1.7/MoinMoin/action/AttachFile.py: AttachFile: now really fixed extensions for packages
2007-04-20T22:03:10  <xorAxAx> not even by ian, besides "our users are to stupid to install py.test"
2007-04-20T22:03:20  <xorAxAx> but it is installable like any other python program via setup.py
2007-04-20T22:03:31  <xorAxAx> (and should be even easy_installable because of that)
2007-04-20T22:03:54  <johill> ohh, _keturn is here ;)
2007-04-20T22:04:18  <_keturn> hi johill
2007-04-20T22:23:40  <ThomasWaldmann> moin
2007-04-20T22:23:48  <johill> moin
2007-04-20T22:26:15  <dreimark> moin
2007-04-20T22:28:06  <dreimark> ThomasWaldmann:can you start this changed moin-dev logging tool
2007-04-20T22:28:23  <ThomasWaldmann> it is started by cron
2007-04-20T22:29:07  <dreimark> it should show instantly entries
2007-04-20T22:29:50  <ThomasWaldmann> i can start it more often
2007-04-20T22:30:03  <ThomasWaldmann> how often does it update currently?
2007-04-20T22:30:17  <dreimark> once per day
2007-04-20T22:30:42  <xorAxAx> ThomasWaldmann: i am running it every 3 minutes on my system
2007-04-20T22:39:28  <dreimark> bbl ~30min
2007-04-20T22:43:51  <ThomasWaldmann> xorAxAx: on what system?
2007-04-20T22:51:07  <xorAxAx> on a vserver ...
2007-04-20T22:51:39  <xorAxAx> hmm
2007-04-20T23:11:45  <johill> _keturn: so anyway, my openid stuff is working now, even rp
2007-04-20T23:12:58  <_keturn> cool.  I wasn't quite sure how to try it out.  is there a branch, or is that a set of patches against the 1.7 tree, or what?
2007-04-20T23:13:09  <johill> those three patches against 1.7
2007-04-20T23:13:20  <johill> http://johannes.sipsolutions.net/patches/moin/all/2007-04-20-13:39/
2007-04-20T23:13:36  <johill> or just moin-oid-common and openid-rp
2007-04-20T23:17:46  <johill> just download them now, the server is supposed to be moved to another datacenter tonight ;)
2007-04-20T23:18:14  <_keturn> do they depend on your auth refactor? (did that already go in to 1.7?)
2007-04-20T23:18:38  <johill> I put that into 1.7 today
2007-04-20T23:18:47  <johill> refactored again to split auth and session handling
2007-04-20T23:23:19  <dreimark> is there an easy way to find the one not translated e.g. MoinI18n/de
2007-04-20T23:24:13  <dreimark> ah founf it !!!!!
2007-04-20T23:24:53  <johill> what is it?
2007-04-20T23:24:58  <dreimark> Check your argument
2007-04-20T23:25:13  <johill> ?
2007-04-20T23:25:56  <dreimark> 1.6devMenuItemsNotTranslated I do check open bugs if they are now closeable
2007-04-20T23:27:31  <johill> btw talking about translation
2007-04-20T23:27:53  <johill> I had a weird problem earlier in that no strings were translated when I set my browser to de but all the pages were correctly loaded
2007-04-20T23:28:32  <johill> but it seems to be fine on test.
2007-04-20T23:28:38  <johill> so maybe it was a local problem
2007-04-20T23:30:40  <dreimark> could that be a browser cache problem?
2007-04-20T23:31:09  <johill> yeah
2007-04-20T23:31:52  <johill> dunno, I'll have to retest, don't worry about it
2007-04-20T23:38:14  <dreimark> ThomasWaldmann:can you update t.www (its not urgent)

MoinMoin: MoinMoinChat/Logs/moin-dev/2007-04-20 (last edited 2007-10-29 19:09:02 by localhost)