MoinMoin: MoinMoinChat/Logs/moin-dev/2007-07-26

2007-07-26T00:03:23  <ThomasWaldmann> re
2007-07-26T04:43:46  <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2561:f42fb1ae0a38 1.7-jabber-knowak/MoinMoin/events/ (__init__.py emailnotify.py notification.py): Fixes to the Event.name commit.
2007-07-26T04:43:49  <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2562:53a79ea2ad93 1.7-jabber-knowak/jabberbot/oob.py: Make url description mandatory, pass args in a dictionary.
2007-07-26T04:43:52  <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2563:8ec6cc1a909f 1.7-jabber-knowak/ (5 files in 2 dirs): Change argument passing in notifications; some pylint and event.name fixes too
2007-07-26T04:44:17  <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2579:e76789b0e537 1.7-jabber-knowak/ (11 files in 7 dirs): Merge main.
2007-07-26T04:45:52  <grzywacz> Ok, what I really need now is a way to generate URLs that would allow to: revert a page, view a certain revision of a page, view a diff... and possibly more
2007-07-26T04:58:07  <starshine> grzywacz: wouldn't that risk bots setting them off?
2007-07-26T08:59:09  <ThomasWaldmann> grzywacz: see Page.url()
2007-07-26T11:59:52  <xorAxAx> ARGH
2007-07-26T12:00:02  <xorAxAx> again a place where reimar didnt fix the server module names
2007-07-26T12:09:44  <ThomasWaldmann> we should ask him if he used Eclipse search function
2007-07-26T12:10:14  <ThomasWaldmann> brb
2007-07-26T12:24:51  <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2692:3e795dbfb01e 1.7-storage-hwendel/MoinMoin/ (storage/external.py events/storagenotify.py): remove cache for now
2007-07-26T12:35:11  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2705:1fb8d43ee759 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: remove code that is unused
2007-07-26T12:37:44  <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2693:7004ddffb860 1.7-storage-hwendel/MoinMoin/user.py: remove unused code
2007-07-26T12:44:25  <xorAxAx> lanius_: so, do you have a plan for caching?
2007-07-26T13:03:49  <CIA-27> moin: Alexander Schremmer <alex AT alexanderweb DOT de> * 2120:b6c92549b7cb 1.6/wiki/server/moinmodpy.htaccess: Yet another place where request module renaming did not take place.
2007-07-26T13:16:26  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2706:6986d581880d 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: Refactoring, extract method parse_repomd()
2007-07-26T13:28:17  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2707:8de51ce63d8d 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: Refactoring, extract method handle_mirrorlists()
2007-07-26T14:32:55  <grzywacz> http://python.net/~goodger/projects/pycon/2007/idiomatic/handout.html
2007-07-26T14:38:58  <lanius_> xorAxAx: no
2007-07-26T14:39:03  <xorAxAx> does it contain "change the API every 3 weeks as well"? :)
2007-07-26T14:39:13  <xorAxAx> lanius_: a wikipage at least with brainstorming stuff? :)
2007-07-26T14:39:31  <lanius_> soon
2007-07-26T14:41:02  <vpv> grzywacz: nice article, bookmarked
2007-07-26T14:41:43  <grzywacz> :)
2007-07-26T14:51:05  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2708:b423a90b13e7 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: small comment etc. fixed
2007-07-26T14:51:07  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2709:acd4a7559d23 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: move ElementTree importing out of handle_repository() so that the interfaces are a bit cleaner
2007-07-26T14:52:25  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2710:0caf5274a62c 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: fix previous commit, should never commit before testing ;)
2007-07-26T16:06:54  <johill> xorAxAx: I fixed *all* of the ID bugs now
2007-07-26T16:10:45  <johill> uh
2007-07-26T16:10:56  <johill> somehow the toc causes the page to show up completely empty now. something's wrong
2007-07-26T16:13:26  <johill> wtf
2007-07-26T16:13:28  <johill> oh btw
2007-07-26T16:13:36  <johill> what are you people smoking :/
2007-07-26T16:13:42  <johill> result = <some boolean expression>
2007-07-26T16:13:45  <johill> expected = True
2007-07-26T16:13:49  <johill> assert expected == result
2007-07-26T16:13:51  <johill> wtf?
2007-07-26T16:13:54  <ThomasWaldmann> johill: you know the wtf public license? :)
2007-07-26T16:14:42  <johill> can't say I do
2007-07-26T16:18:40  <johill> ok that was a weird bug
2007-07-26T16:19:12  <ThomasWaldmann> http://sam.zoy.org/wtfpl/  - see also monsterz on same site for a productivity killer :)
2007-07-26T16:22:43  <johill> :)
2007-07-26T16:24:46  <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2569:5c9e46ab917c 1.7/MoinMoin/parser/text_moin_wiki.py: parser should use anchorlink() instead of url()
2007-07-26T16:25:39  <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2570:896b81e9764e 1.7/MoinMoin/formatter/text_html.py: get rid of the text_html formatter tag stacks, they break with cached stuff anyway
2007-07-26T16:26:18  <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2571:a7dc3cc36362 1.7/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): apparently underscores aren't valid in IDs, fix this
2007-07-26T16:28:45  <johill> this one's large
2007-07-26T16:28:46  <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2572:ce0787373150 1.7/MoinMoin/ (8 files in 4 dirs):
2007-07-26T16:28:46  <CIA-27> moin: introduce ID cache in request, make everybody generate valid IDs even inside includes
2007-07-26T16:28:46  <johill> sorry
2007-07-26T16:28:48  <CIA-27> moin: fixes the TOC macro, top/bottom etc. links inside includes, also [#anchorlinks] etc.
2007-07-26T16:29:48  <johill> but we now generate valid html again
2007-07-26T16:29:49  <johill> :))
2007-07-26T16:30:24  <johill> even with the buggered line anchors
2007-07-26T16:34:20  <ThomasWaldmann> johill: if you touch wiki parser or html formatter, be very careful and do lots of tests
2007-07-26T16:34:51  <johill> the whole thing is so broken anyway... :P
2007-07-26T16:35:39  <johill> we really need to clean away all the internal formatter state
2007-07-26T16:38:55  <johill> at least now you can have an [[Anchor(test)]] and [#test] in a page and still expect it to work inside includes
2007-07-26T16:40:20  <ThomasWaldmann> the important thing is that it gets better and doesnt break important stuff
2007-07-26T16:40:31  <johill> I haven't found anything that broke
2007-07-26T16:40:48  <johill> rendered many pages
2007-07-26T16:41:42  <johill> it did, however, get slower
2007-07-26T16:41:49  <johill> but correctness trumps speed :)
2007-07-26T16:44:15  <xorAxAx> johill: did you test pages with a TOC macro  on them and including those?
2007-07-26T16:44:21  <xorAxAx> so that you see two tocs?
2007-07-26T16:44:42  <johill> yeah
2007-07-26T16:44:43  <johill> works now
2007-07-26T16:44:51  <johill> even mulptilpe levels deep
2007-07-26T16:44:52  <xorAxAx> cool
2007-07-26T16:45:11  <johill> however, something else is wrong now, sometimes pages stay blank on sipsolutions.net
2007-07-26T16:45:16  <johill> not locally in my test though, which is weird
2007-07-26T16:46:54  <johill> it's like everything stops when the TOC starts, and no more output comes
2007-07-26T16:47:18  <johill> nothing in the error log though, strange
2007-07-26T16:48:00  <xorAxAx> i had a similar issue once - the only way to track it down is print debugging ...
2007-07-26T16:48:12  <johill> yeah hard to do though on a server that runs fcgi :)
2007-07-26T16:49:27  <xorAxAx> well, you can also use pdb
2007-07-26T16:49:35  <xorAxAx> on a socket e.g.
2007-07-26T16:50:47  <johill> I'll just print to a logfile
2007-07-26T16:51:06  <ThomasWaldmann> logging.*
2007-07-26T16:57:11  <johill> huh now it's doing nothing any more
2007-07-26T16:57:49  <xorAxAx> the anti-matter TOC?
2007-07-26T17:01:36  <johill> something seems to go wrong with multiple request.redirectedOutput
2007-07-26T17:02:04  <johill> ah no
2007-07-26T17:02:07  <johill> it recurses
2007-07-26T17:05:20  <lanius_> ThomasWaldmann: i have a question about deleted pages
2007-07-26T17:05:34  <lanius_> ThomasWaldmann: in the current code they appear as empty page with an own revision in the edit-log
2007-07-26T17:05:37  <johill> how do I print a stack trace to a log?
2007-07-26T17:05:40  <lanius_> ThomasWaldmann: is this a behaviour which should be kept?
2007-07-26T17:07:21  <johill> uh huh, I'm dumb
2007-07-26T17:07:37  <lanius_> xorAxAx: what do you think about it?
2007-07-26T17:09:26  <johill> fixed
2007-07-26T17:09:32  <johill> that's what you get for local formatter modifications
2007-07-26T17:11:55  <johill> that's an interesting other bug though
2007-07-26T17:13:54  <johill> (probably because I disabled tracebacks0
2007-07-26T17:14:53  <johill> moin should email you the bugs ;)
2007-07-26T17:16:19  <xorAxAx> well, its a bug if it doesnt give you a small message with "server broken"
2007-07-26T17:17:30  <johill> it doesn't
2007-07-26T17:17:32  <johill> locally it does
2007-07-26T17:17:36  <johill> so I'm not sure why
2007-07-26T17:17:40  <johill> fcgi?
2007-07-26T17:18:08  <xorAxAx> because it uses the formatter which is "your" formatter?
2007-07-26T17:18:14  <xorAxAx> ah, wait
2007-07-26T17:18:14  <johill> oh no no
2007-07-26T17:18:22  <xorAxAx> it uses request.write which is stringio
2007-07-26T17:18:23  <johill> the bug was caused by me adding something to the formatters
2007-07-26T17:18:26  <xorAxAx> thats the bug
2007-07-26T17:18:30  <johill> but it works fine locally
2007-07-26T17:18:39  <xorAxAx> there should be a request._write that is always something usable
2007-07-26T17:18:45  <johill> if I cause a bug in the toc code then I see the exception in the webbrowser
2007-07-26T17:18:46  <johill> that's not it
2007-07-26T17:18:55  <xorAxAx> ?
2007-07-26T17:19:02  <johill> s/toc code/toc formatter/
2007-07-26T17:19:10  <xorAxAx> my theory is invalid?
2007-07-26T17:19:17  <johill> yeah
2007-07-26T17:19:23  <johill> because the exception propagates out
2007-07-26T17:19:24  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2711:2ad562292bce 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: Implement parse_primary(), this will fix a serious bug in rpm/srpm handling, even though it means parsing one more XML file.
2007-07-26T17:19:27  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2716:be07231a94d7 1.7-maninfo-vpv/MoinMoin/ (11 files in 7 dirs): merge from main
2007-07-26T17:19:33  <xorAxAx> johill: ?
2007-07-26T17:19:33  <johill> and redirectedOutput has try/finally
2007-07-26T17:19:45  <xorAxAx> hmm
2007-07-26T17:19:55  <vpv> I'm glad Thomas fixed that bot :)
2007-07-26T17:20:04  <johill> so the exception goes out of redirectedOutput which resets to the real write
2007-07-26T17:20:05  <johill> that's fine
2007-07-26T17:20:16  <johill> I'm not sure why on my fcgi installation the page just stays blank at that point
2007-07-26T17:23:55  <johill> well
2007-07-26T17:23:59  <johill> it never drops into failure.handle
2007-07-26T17:24:40  <xorAxAx> hmm
2007-07-26T17:33:45  <johill> ah
2007-07-26T17:33:47  <johill> found the bug
2007-07-26T17:34:29  <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2573:fd896a933d58 1.7/MoinMoin/request/__init__.py: fcgi completely stops output at request.finish()
2007-07-26T17:42:00  <johill> oh btw
2007-07-26T17:42:08  <johill> I fixed a bug in the toc macro so it generates valid html
2007-07-26T17:42:17  <johill> but now if you don't have a level 1 heading you get an empty list item
2007-07-26T17:42:18  <xorAxAx> you mean that div(0)? :)
2007-07-26T17:42:25  <johill> that too :)
2007-07-26T17:42:38  <xorAxAx> who generates div(1)?
2007-07-26T17:42:40  <johill> actually not div(0) but ordered_list(0) I think
2007-07-26T17:42:54  <johill> I do
2007-07-26T17:42:58  <johill> line 160
2007-07-26T17:43:52  <johill> look at HelpOnMacros or such
2007-07-26T18:00:09  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2717:ffade0778123 1.7-maninfo-vpv/MoinMoin/ (action/SisterDiff.py script/xmlrpc/manimport.py): add a couple of FIXMEs
2007-07-26T18:13:44  <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2718:9f94fddfe892 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: add yet another FIXME
2007-07-26T21:15:39  <CIA-27> moin: Marian Neagul marian@info.uvt.ro work * 2519:879529c63f74 1.7-classify-mneagul/MoinMoin/classifier/__init__.py: Remove whitespace
2007-07-26T21:16:09  <CIA-27> moin: Marian Neagul marian@info.uvt.ro work * 2520:e98fa1965dc8 1.7-classify-mneagul/MoinMoin/classifier/__init__.py: Small tokenizer update
2007-07-26T21:16:11  <CIA-27> moin: Marian Neagul marian@info.uvt.ro work * 2521:2013b411fc72 1.7-classify-mneagul/MoinMoin/classifier/RecommendationSystem.py: Initial feature selection for the clustering system
2007-07-26T21:32:03  <ThomasWaldmann> moin
2007-07-26T21:33:22  <xorAxAx> sigh, neagulm not reachable
2007-07-26T21:36:41  <grzywacz> :S
2007-07-26T21:37:41  * ThomasWaldmann .oO(use the force\h\h\h\h\hflatrate)
2007-07-26T22:00:47  <grzywacz> I looks like I should have separate notification text generators for email and jabber anyways... -_-;
2007-07-26T22:00:50  <grzywacz> *It
2007-07-26T22:01:00  <xorAxAx> why?
2007-07-26T22:08:03  <grzywacz> Because it makes sense to, for example, add URLs directly to email body.
2007-07-26T22:08:17  <grzywacz> But it looks nicer if they're separated (using oob data) in jabber.
2007-07-26T22:08:36  <grzywacz> Then you have a notification text with brief explanation of what has happened, and a list of relevant addresses.
2007-07-26T22:08:52  <grzywacz> Well, at least IMHO.
2007-07-26T22:09:29  <xorAxAx> well
2007-07-26T22:09:44  <xorAxAx> how about providing some special handler for urls so that you onlyhave to duplicate the URL stuff?
2007-07-26T22:13:46  <grzywacz> Doable, but still requires changes. Too bad I didn't think about that at the beginning.
2007-07-26T22:22:47  <xorAxAx> neagulm: please rename your modules to be PEP8 compatible (This message has been postponed.)
2007-07-26T23:48:01  <zenhase> still working on fixing the tests, now i am stuck on how to skip tests in absence of genshi
2007-07-26T23:48:17  <zenhase> any suggestions? :)
2007-07-26T23:48:55  <ThomasWaldmann> try: import genshi except: genshi = None
2007-07-26T23:49:24  <ThomasWaldmann> if genshi: do genshi stuff else: py.test.skip("No genshi available")
2007-07-26T23:50:02  <zenhase> ok, will try that
2007-07-26T23:51:30  <grzywacz> hello zenhase :)
2007-07-26T23:51:36  <zenhase> hello grzywacz :)
2007-07-26T23:51:48  <xorAxAx> ThomasWaldmann: nah
2007-07-26T23:51:54  <xorAxAx> do it like this
2007-07-26T23:52:13  <xorAxAx>  try: import genshi\nexcept ImportError: py.test.skip("...")
2007-07-26T23:53:28  <zenhase> hmm
2007-07-26T23:53:56  <zenhase> i put it at the occurence of the first test-error ... which was in conftest.py
2007-07-26T23:54:04  <xorAxAx> ??
2007-07-26T23:54:16  <zenhase> now: tests finished ... 17 skipped in 3.11 seconds
2007-07-26T23:54:17  <grzywacz> Wrong, put it in your test. 8)
2007-07-26T23:54:28  <xorAxAx> yes, into your test file
2007-07-26T23:54:30  <xorAxAx> at the top
2007-07-26T23:54:34  <xorAxAx> below import py
2007-07-26T23:54:43  <xorAxAx> above your test classes
2007-07-26T23:55:10  <zenhase> erm
2007-07-26T23:55:16  <zenhase> there is no test file
2007-07-26T23:55:50  <xorAxAx> ?
2007-07-26T23:55:59  <xorAxAx> so who is importing genshi?
2007-07-26T23:56:16  <zenhase> several moin-modules
2007-07-26T23:56:54  <xorAxAx> then you dont need to skip any tests
2007-07-26T23:57:03  <xorAxAx> because moin has a hard dependency on genshi
2007-07-26T23:57:09  <zenhase> in this branch, yes
2007-07-26T23:57:15  <xorAxAx> yes
2007-07-26T23:58:10  <grzywacz> Hm, true. :)
2007-07-26T23:59:07  <zenhase> perhaps i got confused there :>
2007-07-26T23:59:13  <ThomasWaldmann> zenhase: we can't have a hard dependency on genshi