MoinMoin: MoinMoinChat/Logs/moin-dev/2007-10-20

2007-10-20T08:47:56  <ThomasWaldmann> moin
2007-10-20T09:12:50  <dreimark> moin
2007-10-20T09:27:42  <dreimark> xorAxAx: what killed the bot ?
2007-10-20T09:33:08  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2213:345c21aa7e25 1.6/MoinMoin/userform.py: Fixed small bug introduced by changeset 2204:4052b7c5dc1f
2007-10-20T09:33:12  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2214:2240f159783f 1.6/MoinMoin/ (action/edit.py userform.py): TextCha: also use it for user account creation
2007-10-20T09:36:53  <ThomasWaldmann> dreimark: the machine was restarted meanwhile
2007-10-20T09:43:30  <dreimark> ok, arrgh I recognized the missings logs earlier, but never thought they were lost
2007-10-20T10:45:04  * ThomasWaldmann enhances the user browser
2007-10-20T10:52:56  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2215:d21f3d32a843 1.6/ (MoinMoin/userform.py docs/CHANGES): SystemAdmin user browser: SuperUser now can enable/disable user accounts from there
2007-10-20T11:00:34  <dreimark> :)
2007-10-20T11:01:22  <dreimark> ThomasWaldmann: should I backport my acl group listings from 1.7 or did you?
2007-10-20T11:53:07  <ThomasWaldmann> dreimark: is all that code in it necessary?
2007-10-20T11:53:32  <ThomasWaldmann> dont we have the group stuff in "ready to use" form?
2007-10-20T11:54:44  <dreimark> We have had nothing to get which user is in which group, thats why I ve added it on that place
2007-10-20T11:55:12  <dreimark> ThomasWaldmann:
2007-10-20T11:56:33  <ThomasWaldmann> i suggest you rather add a get_groups(username)
2007-10-20T11:57:17  <dreimark> yep, sounds more useful
2007-10-20T11:58:05  <dreimark> I do that later today, currently I refactor the chapter I did for this python book
2007-10-20T11:58:09  <ThomasWaldmann> or even User.get_groups(self)
2007-10-20T11:58:19  <dreimark> Do you have time next week for reading?
2007-10-20T11:58:34  <ThomasWaldmann> ask me next week :D
2007-10-20T11:59:19  <dreimark> I think I add it to User
2007-10-20T12:00:09  <ThomasWaldmann> and check again if all that code is needed, i suspect it is not
2007-10-20T12:00:37  <dreimark> doing then
2007-10-20T12:01:10  <ThomasWaldmann> (try to avoid doing unnecessary things, that user admin page doesn't scale well already)
2007-10-20T12:02:08  * ThomasWaldmann cleaned up moinmaster accounts using it (disabled lots of spammer accounts)
2007-10-20T12:04:33  <dreimark> :)
2007-10-20T12:06:23  <ThomasWaldmann> if someone feels bored, we need a script that works on disabled accounts, checks whether the userid has edited pages
2007-10-20T12:06:37  <ThomasWaldmann> if no, the account can be purged
2007-10-20T12:06:50  <ThomasWaldmann> if yes, it should ask
2007-10-20T12:15:18  <dreimark> hmm, or something to put them in a honeypot, I geuss that makes them slower
2007-10-20T12:20:50  <ThomasWaldmann> [[SAVE]] click this button if you are a spammer and want to get kicked.
2007-10-20T12:23:04  <xorAxAx> dont forget i18n :)
2007-10-20T12:23:17  <xorAxAx> otherwise you will have many foreign users in the kick set :)
2007-10-20T13:11:19  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2216:59d8bd82ff5f 1.6/MoinMoin/security/textcha.py: TextCha: improve logging
2007-10-20T15:04:05  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2217:065bd554b8e9 1.6/MoinMoin/i18n/ (39 files): updated i18n
2007-10-20T15:08:11  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2218:15d56c2bfd97 1.6/wiki/htdocs/ (3 files in 3 dirs): TextCha: add CSS
2007-10-20T15:08:12  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2219:e3c46b7d0ff0 1.6/MoinMoin/i18n/ (39 files): merged main
2007-10-20T15:22:57  <dreimark> bbl
2007-10-20T15:38:18  <ThomasWaldmann> are there http headers that allow multiple usage of the same key?
2007-10-20T15:44:27  <ThomasWaldmann> (e.g. Status: makes only sense once, but is this the case in general, so we can safely use a dict?)
2007-10-20T15:48:58  <TheSheep> Multiple message-header fields with the same field-name MAY be present in a message if and only if the entire field-value for that header field is defined as a comma-separated list [i.e., #(values)]. It MUST be possible to combine the multiple header fields into one "field-name: field-value" pair, without changing the semantics of the message, by appending each subsequent field-value to the first, each separated by a comma.
2007-10-20T15:50:02  <TheSheep> from  http://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.2
2007-10-20T15:51:42  <ThomasWaldmann> ok, so i need to check against a list of field-names where a list is valid
2007-10-20T15:56:26  <TheSheep> ThomasWaldmann: I think that you can just do a ','.join whenver the header repeats...
2007-10-20T15:57:00  <ThomasWaldmann> that would be a bad idea when it is not allowed
2007-10-20T15:57:08  <ThomasWaldmann> e.g. for status
2007-10-20T15:57:24  <TheSheep>  It *MUST* be possible to combine...
2007-10-20T15:57:39  <ThomasWaldmann> i want to check that, there were lots of errors in moin's headers in the past
2007-10-20T15:57:40  <TheSheep> ThomasWaldmann: are you writing a http validator?
2007-10-20T15:57:48  <TheSheep> aah
2007-10-20T15:57:49  <TheSheep> ok
2007-10-20T16:01:47  <xorAxAx> would be stupid to have moin generate status: 200, 400 headers :)
2007-10-20T17:42:16  <ThomasWaldmann> xorAxAx: i am getting strange effects when trying to test emit_http_headers
2007-10-20T17:42:51  <xorAxAx> hmm
2007-10-20T17:42:57  <ThomasWaldmann> (I try to reset request.sent_headers to False in setup_method, but it thinks it has already sent the headers)
2007-10-20T17:42:57  <xorAxAx> what kind of effects?
2007-10-20T17:43:17  <xorAxAx> so setup_method is not executed multiple times?
2007-10-20T17:43:32  <ThomasWaldmann> it is, i put a print in it
2007-10-20T17:43:37  <xorAxAx> hmm
2007-10-20T17:43:54  <xorAxAx> and the if in moin only checks that attrib?
2007-10-20T17:44:01  <ThomasWaldmann> yes
2007-10-20T17:44:22  <xorAxAx> ah
2007-10-20T17:44:25  <xorAxAx> hmm
2007-10-20T17:44:40  <xorAxAx> check whether you see the same request object at all occurences
2007-10-20T17:44:51  <ThomasWaldmann> ok
2007-10-20T17:46:06  <ThomasWaldmann> same
2007-10-20T17:46:59  <xorAxAx> well, check in a failing test
2007-10-20T17:47:05  <xorAxAx> i can have a look now
2007-10-20T17:51:51  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2892:863060b2cfc5 1.7/MoinMoin/request/ (__init__.py _tests/test_request.py): cleanup http header emission, support list-type headers, add tests
2007-10-20T17:52:29  <ThomasWaldmann> strange, sent_headers is False at the beginning of the test methods
2007-10-20T18:03:24  <ThomasWaldmann> http://moinmo.in/DesktopEdition?action=diff&date=1192872671000000 ?
2007-10-20T18:04:09  <xorAxAx> well, it depends, now i added exe again
2007-10-20T18:06:56  <xorAxAx> ThomasWaldmann: ah! py.test executes every assert at least two times
2007-10-20T18:07:19  <xorAxAx> dont forget that you should never assert something with side effects
2007-10-20T18:07:35  <xorAxAx> and it checks for this invariant by doing it two times
2007-10-20T18:10:02  <ThomasWaldmann> is that a documented feature?
2007-10-20T18:10:13  <xorAxAx> yes
2007-10-20T18:10:30  <ThomasWaldmann> because it does not tell, just the test fails
2007-10-20T18:11:01  <xorAxAx> yes, it should say "incoherent behaviour" or something like that
2007-10-20T18:11:04  <xorAxAx> that might be a bug
2007-10-20T18:11:19  <xorAxAx> maybe it only does that if you do not get an exception for the second time
2007-10-20T18:11:21  <ThomasWaldmann> ok, i move the call out of the assertion
2007-10-20T18:19:50  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2893:408b301a5e44 1.7/MoinMoin/request/ (__init__.py _tests/test_request.py): emit_http_headers tests: fixed (never call functions with side-effects in assert, because they get executed twice)
2007-10-20T18:20:12  <xorAxAx> thats one side of the reason
2007-10-20T18:20:16  <xorAxAx> generally you shouldnt do so
2007-10-20T18:20:28  <xorAxAx> because asserts should be safe to be ommitted (cf. -O)
2007-10-20T18:21:00  <ThomasWaldmann> that would make the tests rather pointless :D
2007-10-20T18:22:18  <xorAxAx> and fast :)
2007-10-20T18:23:32  <ThomasWaldmann> ok, anything else todo for the headers? (I will backport that to 1.6)
2007-10-20T18:52:09  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2220:2096369b0601 1.6/MoinMoin/request/__init__.py: Fix: do not emit duplicate keys in http headers (but log warning), cleanup (backported from 1.7)
2007-10-20T19:49:44  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2221:3f5f01288c73 1.6/MoinMoin/request/__init__.py: Fix duplicate Expires http header emission
2007-10-20T20:40:00  <dreimark> ThomasWaldmann: wikidicts.membergroups looks strange to me. d = self.dictdict.get(dictname) dictname ???
2007-10-20T20:40:35  <dreimark> oh,m thats was entered wrong sorry
2007-10-20T20:41:29  <ThomasWaldmann> ?
2007-10-20T20:45:10  <dreimark> print group.name in that loop returns always None, but I have groups
2007-10-20T20:45:40  <dreimark> for group in self.dictdict.values():
2007-10-20T20:45:47  <dreimark> wikidicts.membergroups
2007-10-20T20:47:02  * ThomasWaldmann looks
2007-10-20T20:55:26  <ThomasWaldmann> well, the code looks ok, maybe insert some debug code and check the tests
2007-10-20T21:03:15  <dreimark> ThomasWaldmann: http://paste.pocoo.org/show/6861/
2007-10-20T21:04:12  <dreimark> the ugly admin tools shows both groups right
2007-10-20T21:05:56  <dreimark> some of the imports are not needed in my example, did some other tests before with that
2007-10-20T21:08:13  <dreimark> ThomasWaldmann: found the bug
2007-10-20T21:08:36  <dreimark> right is for group in self.groupdict.values():
2007-10-20T21:08:49  <dreimark> and not  for group in self.dictdict.values():
2007-10-20T21:08:56  <dreimark> do you agree?
2007-10-20T21:10:30  <ThomasWaldmann> no
2007-10-20T21:11:11  <ThomasWaldmann> well, it depends on what one wants
2007-10-20T21:11:38  <ThomasWaldmann> (whether one wants just direct group members or members of expanded groups)
2007-10-20T21:22:12  <dreimark> hmm, yeah but self.dictdict.values() are not explicity group pages
2007-10-20T21:40:16  <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2222:40944961d9c6 1.6/MoinMoin/ (3 files in 3 dirs): gedit formatter/converter: fix links/transclusion (suppressing default alt texts needs more work) (backport from 1.7)
2007-10-20T21:40:49  <dreimark> gn
2007-10-20T21:41:31  <ThomasWaldmann> gn