MoinMoin: MoinMoinChat/Logs/moin-dev/2009-06-09

2009-06-09T00:00:02  <ThomasWaldmann> add a # (but don't commit THAT :)
2009-06-09T00:00:55  <dennda> shouldn't the test be added to < 2.0-st?
2009-06-09T00:01:37  <ThomasWaldmann> all will be add to 1.7
2009-06-09T00:01:38  <ThomasWaldmann> ed
2009-06-09T00:01:57  <ThomasWaldmann> and then pulled to 1.8/9/...
2009-06-09T00:02:28  <dennda> yes, but the test I should add should check if the example you outlined above really happens, right?
2009-06-09T00:03:09  <dennda> In that case, it'd need to be added to < 2.0 storage since it can't climb in 2.0-st anymore. (and if i shall not commit the testcase you can't backport it to 1.7) ;)
2009-06-09T00:03:21  <dennda> But I guess our discussion went out of sync
2009-06-09T00:03:40  <ThomasWaldmann> ok, 1.7.3 was declared as "final release", so we'll just add the changeset to the repo, but no new release 1.7.x
2009-06-09T00:04:19  <ThomasWaldmann> dennda: you comment out your fix temporarily to test the test failure
2009-06-09T00:04:38  <dennda> ahhh
2009-06-09T00:04:43  <ThomasWaldmann> if it fails, you remove the comment signs again and it should succeed then
2009-06-09T00:05:01  <dennda> and the test stays
2009-06-09T00:05:02  <dennda> sure
2009-06-09T00:07:30  <ThomasWaldmann> and if you want to get famous, send me a 1.7 changeset after testing the stuff in 2.0-st :D
2009-06-09T00:09:34  <dennda> the hardest part is
2009-06-09T00:09:53  <dennda> finding a variable name that fits in with the already existing fixed width naming pattern
2009-06-09T00:09:57  <dennda> !"$%
2009-06-09T00:10:37  <ThomasWaldmann> hm?
2009-06-09T00:11:02  <dennda> just a cosmetical thing
2009-06-09T00:11:26  <dennda> but the acl tests there are lined up and alined so that it looks nicely
2009-06-09T00:11:35  <dennda> the new item names have to be of the same lengths ;)
2009-06-09T00:11:50  <ThomasWaldmann> ah
2009-06-09T00:13:10  <dennda> are you ok with item_rwforall and subitem_4boss? ;)
2009-06-09T00:15:28  <ThomasWaldmann> ok
2009-06-09T00:16:28  <dennda> ok seems it's indeed the case
2009-06-09T00:18:21  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5917:6c4e788d8f99 2.0-storage/MoinMoin/security/_tests/test_security.py: storage: add tests for recently discovered security bug
2009-06-09T00:18:28  <dennda> ThomasWaldmann: review please
2009-06-09T00:19:16  <dennda> test fails due to >       AssertionError: u'AnyUser' may not write u'EveryoneMayReadWriteMe/OnlyTheBossMayWMe' (hierarchic)
2009-06-09T00:19:25  <dennda> if fix is commented out
2009-06-09T00:19:32  <dennda> if fix is active, no test fails
2009-06-09T00:21:40  * dennda clones 1.7
2009-06-09T00:23:29  <ThomasWaldmann> maywme?
2009-06-09T00:23:36  <dennda> may write me
2009-06-09T00:23:43  <ThomasWaldmann> and read
2009-06-09T00:23:54  <dennda> that's allowed by the after acl
2009-06-09T00:24:56  <dennda> There's no match for AnyUser on the item acl itself (since there's only 'JoeDoe:read,write') hence it does not check the parent but check the after acl, so anyuser may read
2009-06-09T00:25:41  <ThomasWaldmann> the other acls rather look like JaneDoe is teh boss :)
2009-06-09T00:26:22  <dennda> do you want me to change it for consistency or is it wurscht?
2009-06-09T00:26:56  <ThomasWaldmann> well, both or nothing :)
2009-06-09T00:27:13  <dennda> both what?
2009-06-09T00:33:17  <ThomasWaldmann> wme and jane
2009-06-09T00:33:50  <dennda> ??
2009-06-09T00:34:40  <dennda> don't you agree that the 'read' privilege is granted by the after acl?
2009-06-09T00:34:43  <ThomasWaldmann> either you change both or you change nothing, but not 2 changeset for cosmetic fixes :)
2009-06-09T00:34:52  <dennda> ah
2009-06-09T00:35:13  <dennda> well ok then it stays as it is ;)
2009-06-09T00:35:28  <dennda> how do you prefer to receive the 1.7 changesets? (one for the tests, one for the fix)
2009-06-09T00:36:44  <ThomasWaldmann> maybe 2 are better than one. first the tests, then the fix.
2009-06-09T00:37:10  <dennda> .oO(I just said that, didn't I?)
2009-06-09T00:37:28  <ThomasWaldmann> yes, i agree :)
2009-06-09T00:38:40  <dennda> so there will be a maint update for 1.8 then I assume?
2009-06-09T00:39:36  <ernestas> hmm, interesting sentence: "If you have a large image that you don't want to necessarily want/need to display at once, create a thumbnail preview."
2009-06-09T00:41:56  <ThomasWaldmann> patch goes into 1.7 for whoever wants to pick it from the repo, 1.8 maybe will see a maint release soon
2009-06-09T00:43:07  <ThomasWaldmann> ernestas: that sounds broken
2009-06-09T00:43:16  <ernestas> I see
2009-06-09T00:43:25  <ernestas> it's from HelpOnImage
2009-06-09T00:43:29  <ernestas> +s
2009-06-09T00:43:53  <ernestas> this page needs some proof-reading
2009-06-09T00:43:55  <ThomasWaldmann> s/want to//
2009-06-09T00:46:29  <dennda> ThomasWaldmann: How would I best send you the changesets now?
2009-06-09T00:46:40  <dennda> ah export
2009-06-09T00:47:27  <dennda> http://paste.pocoo.org/show/121876/
2009-06-09T00:47:48  <dennda> http://paste.pocoo.org/show/121877/
2009-06-09T00:48:04  <dennda> You need to select "download this paste" in the advanced options
2009-06-09T00:48:12  <dennda> and then save and hg import
2009-06-09T00:48:42  <ernestas> I'm not touching that page, should I?
2009-06-09T00:50:16  <ThomasWaldmann> ernestas: it's a wiki, feel free to improve it
2009-06-09T00:54:28  <ThomasWaldmann> dennda: patch unexpectedly ends in middle of line
2009-06-09T00:54:48  <ThomasWaldmann> "always end files with \n"
2009-06-09T00:57:08  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 3868:47c0ada5c1a2 1.7/MoinMoin/security/_tests/test_security.py: Add security tests for hierarchic acl processing. The tests fail.
2009-06-09T00:57:11  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 3869:897cdbe9e8f2 1.7/MoinMoin/security/__init__.py: Provide fix for recently added failing security tests.
2009-06-09T00:59:00  <dennda> ThomasWaldmann: yeah sorry, I'm too pep8ified
2009-06-09T00:59:15  <ThomasWaldmann> hm?
2009-06-09T01:02:35  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4340:36d4b0af2213 1.8/MoinMoin/security/ (__init__.py _tests/test_security.py): merged moin/1.7
2009-06-09T01:04:10  <dennda> ThomasWaldmann: a \n at the EOF is trailing whitespace :-)
2009-06-09T01:04:20  <dennda> http://hg.moinmo.in/moin/1.6/rev/61e81400d164
2009-06-09T01:05:04  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4745:8f913a5ac5f1 1.9/MoinMoin/security/ (__init__.py _tests/test_security.py): merged moin/1.8
2009-06-09T01:05:18  <ThomasWaldmann> dennda: no, it is not
2009-06-09T01:05:35  <dennda> hm ok, always thought it was
2009-06-09T01:05:42  <dennda> anyway, gonna get some rest
2009-06-09T01:05:59  <dennda> this little thing took quite a while now
2009-06-09T01:06:24  <ThomasWaldmann> yes, i know hacls were added in 1.6, but 1.6 doesn't get changes any more
2009-06-09T01:06:51  <ThomasWaldmann> (and that's not the first bug in it)
2009-06-09T01:07:25  <ThomasWaldmann> if it was, every line would have to be considered as "having trailing whitespace"
2009-06-09T01:10:55  <ernestas> when are you planning to release 1.9?
2009-06-09T01:12:13  <dennda> yeah sure, I know 1.6 is not maintained anymore
2009-06-09T01:12:21  <dennda> I was just being a nostalgic person ;)
2009-06-09T01:13:23  <dennda> ok, good night
2009-06-09T01:13:42  <ThomasWaldmann> ernestas: progress has been rather slow recently as we are all hacking on greater and newer stuff
2009-06-09T01:13:56  <ernestas> ok then
2009-06-09T01:13:59  <ThomasWaldmann> but i guess "this summer" :)
2009-06-09T01:16:08  *** ernestas has quit IRC
2009-06-09T03:32:36  *** grzywacz has quit IRC
2009-06-09T03:32:36  *** TheSheep has quit IRC
2009-06-09T03:33:29  *** TheSheep has joined #moin-dev
2009-06-09T03:48:44  *** PawelPac1na has quit IRC
2009-06-09T03:59:02  *** PawelPacana has joined #moin-dev
2009-06-09T05:58:20  *** amartani has quit IRC
2009-06-09T06:01:09  *** amartani has joined #moin-dev
2009-06-09T08:48:49  <dreimark> moin
2009-06-09T08:48:58  <ThomasWaldmann> moin
2009-06-09T09:14:34  *** PawelPacana has quit IRC
2009-06-09T09:19:47  *** PawelPacana has joined #moin-dev
2009-06-09T09:54:08  *** grzywacz has joined #moin-dev
2009-06-09T10:21:42  * xorAxAx sighs
2009-06-09T10:22:04  <xorAxAx> ... a melody in an indian raga
2009-06-09T11:11:39  <dreimark> he was too active before soc start and seems to be vanished.
2009-06-09T11:18:23  <xorAxAx> my student? no, he wasnt too active
2009-06-09T11:18:35  <xorAxAx> and he was here in the chat at the weekend
2009-06-09T11:25:00  *** johill has joined #moin-dev
2009-06-09T11:25:14  <johill> was there somebody working on openid for SoC?
2009-06-09T11:25:20  <johill> I thought I saw something
2009-06-09T11:25:25  <johill> http://wiki.openid.net/OpenID-Security-Best-Practices might be interesting
2009-06-09T11:25:26  * dennda waves
2009-06-09T11:25:45  <dennda> There were some commits, yeah, but that was no SoC task afaik
2009-06-09T11:26:36  <johill> ok, I was just going from the repo list :)
2009-06-09T11:29:40  <johill> anyway, I won't have time to work on openid security, just thought I'd throw that URL out there
2009-06-09T11:50:26  <xorAxAx> hi johill
2009-06-09T12:09:14  <ThomasWaldmann> moin johill. rowan kerr has been working on openid stuff since quite a while.
2009-06-09T12:12:21  <johill> yeah, I saw some of that
2009-06-09T12:13:34  <ThomasWaldmann> some peer review wouldn't be bad so we can merge that stuff some day
2009-06-09T12:24:43  <xorAxAx> currently doodle suggests sunday 18:00 CEST
2009-06-09T12:24:55  <johill> ThomasWaldmann: hmm, I'll try to find some cycles for that
2009-06-09T12:25:02  <xorAxAx> only reimar is away and my student hasnt voted yet
2009-06-09T12:28:27  <ThomasWaldmann> i was wondering a bit about how much time some students have not available for a 1h meeting
2009-06-09T12:28:49  <ThomasWaldmann> (and how they do their SOC project if they have not much time)
2009-06-09T12:30:40  <dennda> Should I add more yellow boxes? The green ones indicate my preferred day and time (which I thought the poll was about)
2009-06-09T12:30:50  <dennda> (Assuming you're talking to me) :-)
2009-06-09T12:32:05  <ThomasWaldmann> the colours are explained below the table
2009-06-09T12:33:01  * dennda edits his entry...
2009-06-09T12:33:21  <ThomasWaldmann> xorAxAx: what's your student's localtime for 18cest?
2009-06-09T12:35:10  <xorAxAx> he is 3.5 h in the future :)
2009-06-09T12:35:37  <ThomasWaldmann> half a timezone?
2009-06-09T12:35:40  <xorAxAx> yes
2009-06-09T12:36:40  * dennda added more yellow boxes
2009-06-09T12:37:23  <dennda> I could add even more, but then it'd be rather unpredictable for me
2009-06-09T12:38:59  <ThomasWaldmann> xorAxAx: ok, that should be doable for him (assuming that he has internet at his home/sleeping place)
2009-06-09T12:41:37  <ThomasWaldmann> dennda: please develop the habit to update the diary as last "soc thing" daily
2009-06-09T12:42:54  <dennda> hm ok, no problem. what's the reason you ask me to do that? sth wrong?
2009-06-09T12:44:09  <ThomasWaldmann> no, it's just about doing the docs near-time
2009-06-09T12:44:34  <ThomasWaldmann> what's your plan for today?
2009-06-09T12:45:14  <dennda> docs
2009-06-09T12:45:25  <dennda> perhaps the switch to permanent storage
2009-06-09T12:45:35  <dennda> and further exam preparations
2009-06-09T12:46:04  <dennda> oh and I was thinking about improving the access denied error page
2009-06-09T12:48:08  <ThomasWaldmann> sounds good :)
2009-06-09T12:49:22  <dennda> johill: while you're with us; do you remember how well your FS backend worked on posix platforms?
2009-06-09T12:49:36  <johill> 'how well'?
2009-06-09T12:49:59  <johill> I never had problems with it
2009-06-09T12:50:10  <dennda> That counts as 'well'
2009-06-09T13:01:19  <ThomasWaldmann> johill: btw, i did a bit cosmetics/cleanups/fixes to the non-revisioned flatfile backend, for basic needs that works now, too
2009-06-09T13:03:16  <ThomasWaldmann> (metadata handling is still lacking, though - and it of course can't handle binary items that way)
2009-06-09T13:05:23  <johill> why can it not handle binary?
2009-06-09T13:05:45  <johill> wI thought that was fine, since the arbitrary contents starts after \n\n
2009-06-09T13:08:02  <ThomasWaldmann> ehrm, i didn't try. it just fellt strange to mix text metadata header and binary data into 1 file.
2009-06-09T13:08:14  <ThomasWaldmann> -l
2009-06-09T15:02:33  <dimazest> ThomasWaldmann: you were right about processed_groups in _iter, recursive iterator it's something new for me
2009-06-09T15:03:15  <ThomasWaldmann> you mean to check it before recursing?
2009-06-09T15:03:28  <dimazest> yes
2009-06-09T15:03:58  <ThomasWaldmann> sure, one always tries the cheap stuff first (at least when coding :)
2009-06-09T15:19:54  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5918:9b548511594b 2.0-storage/create_persistent_dev_wiki.py: storage: Added simple script to create a persistent dev wiki for developers and beta testers
2009-06-09T15:32:09  *** ernestas has joined #moin-dev
2009-06-09T15:37:55  <dennda> damn pyc files
2009-06-09T15:39:17  <dennda> I think later we can have something similar to that script for conversion from 1.9 to 2.0
2009-06-09T15:53:13  <johill> ThomasWaldmann: yeah, but email or http does the same thing, so I'm pretty sure, modulo bugs, it should be ok -- I think I probably even tested it
2009-06-09T16:05:44  <CIA-19> Dmitrijs Milajevs <dimazest@gmail.com> default * 4773:d9151c006aab 1.9-groups-dmilajevs/MoinMoin/groups/backends/wiki_group.py:
2009-06-09T16:05:44  <CIA-19> Groups2009: wiki_page backend Group refinements. A _load_group() handles group
2009-06-09T16:05:44  <CIA-19> loading from cache or a wiki page. KeyError in the _load_group() returns a name
2009-06-09T16:05:44  <CIA-19> of not existing group. An _iter() does not add already stored group name to the
2009-06-09T16:05:44  <CIA-19> processed_groups.
2009-06-09T16:36:52  <ThomasWaldmann> dimazest: line109 is likely the same thing (not needed).
2009-06-09T16:40:11  <ThomasWaldmann> and this is because the processed_x param is just a reference to a mutable object (a set) and not a copy of it
2009-06-09T16:47:22  <ThomasWaldmann> dimazest: backend_type and backend_name needs to be given to Group(), not via usage of globals
2009-06-09T16:48:08  <ThomasWaldmann> maybe even better, use Group(...., self) with self being the Backend instance
2009-06-09T16:48:36  <ThomasWaldmann> and within Group, store it as self.backend
2009-06-09T16:49:36  <ThomasWaldmann> remove the two globals and do it directly in the Backend class
2009-06-09T16:51:16  <ThomasWaldmann> within Group, do not use "group_name" (we have a Group object g, thus g.name is of course the group's name)
2009-06-09T16:53:40  <ThomasWaldmann> same is true for Backend, with the slight problem there for backend_type that "type" alone is a builtin, so it should not be used
2009-06-09T16:57:04  <ThomasWaldmann> the Backend type is given by Backend.__class__ somehow, though
2009-06-09T16:57:18  <ThomasWaldmann> dimazest: still there?
2009-06-09T17:05:02  * ThomasWaldmann merges 1.9->1.9-st
2009-06-09T17:19:05  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5727:a78ed4cc1a22 1.9-storage/ (52 files in 9 dirs): merged moin/1.9
2009-06-09T17:23:41  <dimazest> ok
2009-06-09T17:26:26  <ThomasWaldmann> i guess we are soon finished with the wikigroup stuff, next would be some restructuring into different classes
2009-06-09T17:27:17  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5929:83001c1cb544 2.0-storage/ (52 files in 9 dirs): merged moin/1.9-storage
2009-06-09T17:27:24  <ThomasWaldmann> dimazest: do you like to do the config groups first or do you think we can already decide what's going into the misc classes?
2009-06-09T17:28:02  <dimazest> ThomasWaldmann: i was thinking about it
2009-06-09T17:28:41  <dimazest> i want to do config groups
2009-06-09T17:29:12  <ThomasWaldmann> hmm, lets look again at the manager stuff first...
2009-06-09T17:31:04  <dimazest> BackendManager can be used for config groups, it works with sets
2009-06-09T17:31:31  <ThomasWaldmann> i think you didn't implement the changes for groups/__init__.py yet
2009-06-09T17:31:43  <ThomasWaldmann> (we talked about it recently)
2009-06-09T17:32:06  <ThomasWaldmann> is there something unclear?
2009-06-09T17:32:17  <dimazest> ThomasWaldmann: to get rid of BackendManager and move mapping to backends?
2009-06-09T17:32:28  <ThomasWaldmann> yes
2009-06-09T17:32:56  <ThomasWaldmann> the less crap code we have, the easier it'll get :D
2009-06-09T17:33:38  <dimazest> as i said, BackendManager becomes config backend
2009-06-09T17:34:36  <ThomasWaldmann> hm?
2009-06-09T17:35:00  <dimazest> wait....
2009-06-09T17:35:38  <dimazest> it's easier to get rid of BackendManager, create Group and Backend for config group
2009-06-09T17:35:55  <ThomasWaldmann> btw, mapper_to_backend/backend_to_mapper - we have no such thing called a mapper, so maybe we need a better name somehow
2009-06-09T17:36:22  <dimazest> which, is very similar to wiki_group
2009-06-09T17:36:55  <dimazest> except of _load_group
2009-06-09T17:37:20  <ThomasWaldmann> how about backend_name(group_name) and group_name(backend_name)?
2009-06-09T17:38:14  <dimazest> instead of mappers? to_backend_name(..)
2009-06-09T17:38:57  <ThomasWaldmann> yes, something like that
2009-06-09T17:39:14  <ThomasWaldmann> in which class would you put it?
2009-06-09T17:39:15  <dimazest> because group_name is used in the code, and has slightly different meaning, even if we change self.group_name to self.name
2009-06-09T17:39:49  <dimazest> to Backend.
2009-06-09T17:40:33  <ThomasWaldmann> ok
2009-06-09T17:42:41  <dimazest> i'm not sure, but can decorators be used, like this http://paste.pocoo.org/show/122031/
2009-06-09T17:43:13  <dimazest> instead of writing return self._backend[self.mapper_to_backend(group_name)]
2009-06-09T17:43:51  <dimazest> are there decorators in 2.4? and should they be used for such a thing?
2009-06-09T17:44:54  <ThomasWaldmann> membergroups will also need to move
2009-06-09T17:45:32  <dimazest> from the GroupManager?
2009-06-09T17:45:53  <dimazest> btw, wiki_page does not use BackendManager
2009-06-09T17:46:29  <ThomasWaldmann> maybe a decorator adds more complexity than needed, and you can't use it for all needed usecases
2009-06-09T17:46:53  <dimazest> ok
2009-06-09T17:48:38  <ThomasWaldmann> (17:45) <  dimazest> btw, wiki_page does not use BackendManager
2009-06-09T17:48:46  <ThomasWaldmann> what do you mean exactly?
2009-06-09T17:50:29  <dimazest> that GroupManager takes wiki_page.Backend
2009-06-09T17:50:48  <dimazest> GroupManager(backends=[wiki_group.Backend(request)])
2009-06-09T17:51:53  <dimazest> what is different from GroupManager(backends=[BackendManager(backend=self.first_backend_groups)])
2009-06-09T18:02:40  <xorAxAx> ok, so the date is set
2009-06-09T18:04:38  <xorAxAx> amartani, devilsadvocate_, dennda, dimazest, dreimark, ThomasWaldmann, waldi: the date was fixed to sunday 16:00 UTC. see you then! and please, students, prepare your lines to be able to paste them
2009-06-09T18:05:20  <xorAxAx> devilsadvocate: ping
2009-06-09T18:06:41  <ThomasWaldmann> http://master18.moinmo.in/HelpOnInstalling/TwistedWeb?action=AttachFile&do=view&target=servicerunner.py
2009-06-09T18:07:01  <ThomasWaldmann> can someone comment on that? should we distribute that with master pages?
2009-06-09T18:10:13  <xorAxAx> looks like a helpful extension
2009-06-09T18:10:37  <ThomasWaldmann> dimazest: i am not sure I understand what you mean or what you want to know or what the problem is
2009-06-09T18:11:28  <dimazest> i'm saying that BackendManager is not used -> not needed
2009-06-09T18:11:58  *** ernestas is now known as ext
2009-06-09T18:12:01  <dimazest> ThomasWaldmann: why is backend_type is needed?
2009-06-09T18:12:28  <ThomasWaldmann> mostly for repr and debugging, but one could also use __class__ for that somehow
2009-06-09T18:12:51  <ThomasWaldmann> xorAxAx: thanks for arranging
2009-06-09T18:14:05  <dimazest> __class__ should be used, i think
2009-06-09T18:15:15  <ThomasWaldmann> so only name is left and that could be an optional param with a sane default
2009-06-09T18:15:43  <xorAxAx> ThomasWaldmann: no problem
2009-06-09T18:16:06  <xorAxAx> i am constantly recovering more from my illness and becoming better
2009-06-09T18:16:09  <dimazest> ThomasWaldmann: for Group now i have def __init__(self, request, name, backend):
2009-06-09T18:16:24  <xorAxAx> so i might find more time for moin
2009-06-09T18:16:27  <ThomasWaldmann> xorAxAx: good to hear
2009-06-09T18:17:07  <dimazest> and Group.backend moustly for debugging too?
2009-06-09T18:17:58  <ThomasWaldmann> dimazest: for anything needing access to the ("parent") backend
2009-06-09T18:19:09  <ThomasWaldmann> most stuff will want to access via request.groups, but who knows
2009-06-09T18:19:41  <ThomasWaldmann> at least repr can get the backend name and type via .backend
2009-06-09T18:20:27  <ThomasWaldmann> or just use repr(self.backend) internally
2009-06-09T18:22:06  <dimazest> ThomasWaldmann: http://paste.pocoo.org/show/122043/
2009-06-09T18:24:50  <ThomasWaldmann> add a blank after %
2009-06-09T18:26:12  <dimazest> ok
2009-06-09T18:28:57  <ThomasWaldmann> dimazest: the backend's name needs to be an option
2009-06-09T18:29:07  <ThomasWaldmann> ..al argument
2009-06-09T18:30:01  <ThomasWaldmann> like ..., name='moin_wiki_groups')
2009-06-09T18:30:59  <dimazest> ThomasWaldmann: name there is the name of the group, e.g AdminGroup
2009-06-09T18:31:26  <dimazest> should it be nameless?
2009-06-09T18:31:30  <ThomasWaldmann> that maybe should be group_name there (in the context of Backend)
2009-06-09T18:31:41  <ThomasWaldmann> it always depends on context
2009-06-09T18:32:04  <dimazest>     def __getitem__(self, group_name):
2009-06-09T18:32:04  <dimazest>         """
2009-06-09T18:32:04  <dimazest>         Return wiki group backend object.
2009-06-09T18:32:04  <dimazest>         """
2009-06-09T18:32:07  <dimazest>         return Group(request=self.request, name=group_name, backend=self)
2009-06-09T18:32:13  <ThomasWaldmann> and for the Group.__repr__ maybe rather use name= in the string, not group_name=
2009-06-09T18:32:41  <ThomasWaldmann> dimazest:  that code looks ok
2009-06-09T18:33:30  <ThomasWaldmann> (the named arguments are not really required, but make it easier readable)
2009-06-09T18:35:25  <dimazest> ThomasWaldmann: this is the reason, i've put it. self alone doesn't loos clear
2009-06-09T18:40:31  <ThomasWaldmann> btw, in the end, GroupManager will be rather a backend-like middleware
2009-06-09T18:41:44  <dimazest> ThomasWaldmann: i'm trying to add to_backend_name and to_group_name, it becomes quite messy
2009-06-09T18:44:40  <dimazest> http://paste.pocoo.org/show/122049/ ^^
2009-06-09T18:46:02  <dimazest> this is wrong
2009-06-09T18:46:37  <ThomasWaldmann> dimazest: if you are having a special backend, like the wiki page group backend, that can be just a method
2009-06-09T18:47:33  <dimazest> can property be used there?
2009-06-09T18:48:34  <dimazest> for the Graup.name?
2009-06-09T18:48:48  <dimazest> ThomasWaldmann: which method?
2009-06-09T18:49:07  <ThomasWaldmann> the mapping functions can be just methods
2009-06-09T18:49:22  <ThomasWaldmann> no need to give them to __init__
2009-06-09T18:49:35  <dimazest> aaa, clear
2009-06-09T18:50:39  <ThomasWaldmann> and someone (you) will have to code them :D
2009-06-09T18:50:40  <dennda> 16:00 UTC is 18:00 CEST, correct?
2009-06-09T18:50:49  <ThomasWaldmann> yes
2009-06-09T18:52:52  <ThomasWaldmann> btw, other than that, it looks ok and not messy
2009-06-09T18:53:47  <dimazest> :)
2009-06-09T18:54:00  <ThomasWaldmann> 25/26 should use backend_names not group_names
2009-06-09T18:54:27  <ThomasWaldmann> same at other places
2009-06-09T18:54:49  <ThomasWaldmann> (or backend_group_names ...)
2009-06-09T18:55:35  <ThomasWaldmann> then it reads less strange than to_group_name(group_name) :)
2009-06-09T19:02:06  <dimazest> ThomasWaldmann: but for wiki group it is not needed
2009-06-09T19:02:49  <dimazest> right?
2009-06-09T19:04:45  <ThomasWaldmann> for now, until we have dissected the stuff into misc. classes, maybe just use some identify function/method
2009-06-09T19:04:54  <ThomasWaldmann> ehrm, identity
2009-06-09T19:34:54  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4341:4de22494e7c7 1.8/docs/ (INSTALL.html UPDATE.html): make install-docs interwiki
2009-06-09T19:34:56  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4342:d1ef741f87ff 1.8/docs/CHANGES: updated CHANGES
2009-06-09T19:54:10  <dimazest> https://fedoraproject.org/wiki/FUDCon:Berlin_2009 looks interesting
2009-06-09T19:55:07  <dimazest> unfortunately it's not in Munich, too far for me, otherwise i would come
2009-06-09T20:00:53  <johill> I'll be there
2009-06-09T20:01:10  <johill> you can fly for ~130 eur ;)
2009-06-09T20:02:03  <dimazest> johill: i have exams
2009-06-09T20:02:26  <dimazest> why 130?
2009-06-09T20:02:26  <dennda> johill: What makes you visit a fedora event?
2009-06-09T20:04:15  <johill> nothing
2009-06-09T20:04:26  <johill> we're having a small wireless summit along with it
2009-06-09T20:04:40  <johill> dimazest: because that was the first quote I got easily?
2009-06-09T20:06:36  <dennda> johill: does it make sense to buy a draft-n access point?
2009-06-09T20:07:05  <johill> in what way?
2009-06-09T20:07:12  <johill> if you have regular DSL, no ;)
2009-06-09T20:08:09  <dennda> in the "linux supports most wirless n cards" way
2009-06-09T20:08:10  * ThomasWaldmann will use a fb7270 soon
2009-06-09T20:08:26  <dennda> ThomasWaldmann: and put debian on it? :-)
2009-06-09T20:08:33  <ThomasWaldmann> no
2009-06-09T20:08:53  <dennda> not too exciting then :-)
2009-06-09T20:09:13  <johill> dennda: other than broadcom
2009-06-09T20:09:24  <dennda> intel
2009-06-09T20:09:38  <dennda> ok
2009-06-09T20:09:54  <johill> intel's fine
2009-06-09T20:10:00  <johill> I have two 11n APs
2009-06-09T20:10:07  <johill> the dlink dir-855 is a nice device
2009-06-09T20:10:10  <johill> I like their web interface
2009-06-09T20:10:12  <ThomasWaldmann> dennda: i need it working stable with many of its features
2009-06-09T20:10:27  <dennda> stable != exciting
2009-06-09T20:14:51  * ThomasWaldmann just waits for telekom and kabelbw to port the phone numbers, then I can upgrade from g to n and have the n router near enough to get usable rates
2009-06-09T20:23:04  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4343:e630e9d340d1 1.8/MoinMoin/i18n/ (41 files): updated i18n
2009-06-09T20:49:04  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4344:bafe33aebad5 1.8/wiki/underlay.tar: updated underlay
2009-06-09T20:49:11  <dimazest> ThomasWaldmann: http://paste.pocoo.org/show/5Ill8v2lSWJgLGK8OvA9/ BaseGroup and BaseBackend
2009-06-09T20:53:05  <ThomasWaldmann> maybe it is easier to generally have the group names stored and only to convert to the backend name when needed
2009-06-09T20:53:10  *** johill has left #moin-dev
2009-06-09T20:55:50  <ThomasWaldmann> dimazest: or rather: that's the only way it'll work correctly
2009-06-09T20:56:26  <ThomasWaldmann> btw, do you commit once a while?
2009-06-09T20:56:37  <dimazest> so, group stores group names in moin moin format
2009-06-09T20:56:50  <dimazest> no, should i?
2009-06-09T20:57:40  <dimazest> changes are in a queue, so i can change them
2009-06-09T20:58:20  <ThomasWaldmann> well, do anything that yields clean changesets
2009-06-09T20:58:28  <ThomasWaldmann> and not too big ones
2009-06-09T20:59:24  <ThomasWaldmann>         for backend_name in self.member_groups:
2009-06-09T20:59:40  <ThomasWaldmann> you can't have backend_names in member_groups.
2009-06-09T21:00:34  <ThomasWaldmann> because when you parse e.g. a wiki group from a page, it'll get the moin names. and it won't know how they translated to the backend name at that place (and what backend handles them).
2009-06-09T21:02:08  <dimazest> ThomasWaldmann: life will be easier, if in member_groups are moin names
2009-06-09T21:04:11  <ThomasWaldmann> yes
2009-06-09T21:06:59  <dimazest> shoul i use moin group name, moin notation, moin format?
2009-06-09T21:32:13  <ThomasWaldmann> for?
2009-06-09T21:34:15  <dimazest> for comments, like @param name: name of the group in moin notation.
2009-06-09T21:34:47  <amartani> hi..
2009-06-09T21:34:59  <dimazest> hi
2009-06-09T21:35:29  <ThomasWaldmann> use moin group name
2009-06-09T21:47:53  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4345:3010c1a94185 1.8/ (12 files in 9 dirs): bump version number to 1.8.4
2009-06-09T22:06:40  <ThomasWaldmann> if someone likes to help testing: http://static.moinmo.in/files/moin-1.8.4.tar.gz
2009-06-09T22:15:30  <dennda> well, wget, tar -xvzf, cd, ./wikiserver.py, load localhost:8080 works at least ;)
2009-06-09T22:22:41  <ThomasWaldmann> yeah, even a bit more
2009-06-09T22:22:49  <ThomasWaldmann> ok, let's release
2009-06-09T22:52:25  <CIA-19> Alexandre Martani <amartani AT gmail DOT com> default * 5927:25b593d25d18 2.0-storage-editor-amartani/MoinMoin/ (4 files in 3 dirs): Real-time editor: \"Save and continue\" button on editor
2009-06-09T22:55:02  <ThomasWaldmann> amartani: btw, does it already indicate whether contents was modified (and needs saving)?
2009-06-09T22:56:47  <amartani> ThomasWaldmann: not yet.. I'll do it.
2009-06-09T22:57:51  <ThomasWaldmann> amartani: is it possible to count the changes done by all editors somehow?
2009-06-09T23:00:43  <ThomasWaldmann> (to have some sort of metric for amount of change)
2009-06-09T23:04:40  <amartani> ThomasWaldmann: I think it is possible, but a bit difficult, and I don't know if it would be reliable..
2009-06-09T23:06:52  <amartani> ThomasWaldmann: I'll think more about it..
2009-06-09T23:08:17  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4346:8b78ac8e8007 1.8/.hgtags: tagged release 1.8.4
2009-06-09T23:08:39  <amartani> ThomasWaldmann: I had an error when sending the post with no data file, I made a small change on items/__init__
2009-06-09T23:09:04  <ThomasWaldmann> amartani: btw, maybe you can steal some "be careful when moving away" js stuff from moin 1.8
2009-06-09T23:40:11  <CIA-19> Dmitrijs Milajevs <dimazest@gmail.com> default * 4774:64ff4ea0b597 1.9-groups-dmilajevs/MoinMoin/groups/backends/wiki_group.py:
2009-06-09T23:40:11  <CIA-19> Groups2009: wiki_group backend changes. backend_type and backend_name were
2009-06-09T23:40:11  <CIA-19> removed both for Group and Backend. Group.__init__ takes backed as a parameter.
2009-06-09T23:40:11  <CIA-19> Group.name is used instead of Group.group_name. Group.__repr__ prints its class.
2009-06-09T23:40:13  <CIA-19> Dmitrijs Milajevs <dimazest@gmail.com> default * 4775:7dd5d15de911 1.9-groups-dmilajevs/MoinMoin/groups/backends/ (__init__.py _tests/test_wiki_group.py wiki_group.py): Groups2009: BaseGroup and BaseBackend initial import. wiki_group.Group and wiki_group.Backend inherit from those classes. TestWikiGroupNameMapping is added to check group name mapping routines.
2009-06-09T23:47:55  * xorAxAx sighs in Asaveri
2009-06-09T23:49:48  <ThomasWaldmann> xorAxAx: you learned the musical language of your student? :)
2009-06-09T23:50:22  <xorAxAx> yep :)