MoinMoin: MoinMoinChat/Logs/moin-dev/2009-05-30

2009-05-30T00:00:56  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5814:4dd9dca3af6c 2.0-storage/MoinMoin/items/__init__.py: items: fix undelete behaviour
2009-05-30T00:02:16  *** ThomasWaldmann has quit IRC
2009-05-30T00:02:16  *** xorAxAx has quit IRC
2009-05-30T00:03:22  *** xorAxAx has joined #moin-dev
2009-05-30T00:03:22  *** ThomasWaldmann has joined #moin-dev
2009-05-30T00:05:05  *** xorAxAx has quit IRC
2009-05-30T00:05:08  *** xorAxAx has joined #moin-dev
2009-05-30T00:09:25  *** grzywacz has quit IRC
2009-05-30T01:10:40  *** devilsadvocate has quit IRC
2009-05-30T01:11:06  *** devilsadvocate has joined #moin-dev
2009-05-30T01:20:12  *** rkerr has quit IRC
2009-05-30T03:34:24  *** coreyleong has joined #moin-dev
2009-05-30T03:38:51  *** coreyleong has quit IRC
2009-05-30T08:30:34  <CIA-19> Dmitrijs Milajevs <dimazest@gmail.com> default * 4735:8dcf18d0ae41 1.9-groups-dmilajevs/MoinMoin/groups/ (6 files in 2 dirs): Groups 2009: pep8. MoinMoin.groups _test renamed to _tests
2009-05-30T08:30:37  <CIA-19> Dmitrijs Milajevs <dimazest@gmail.com> default * 4736:c3e904a9c00b 1.9-groups-dmilajevs/MoinMoin/ (6 files in 4 dirs): Groups 2009: BackendManager gets request object in __init__
2009-05-30T11:09:43  *** grzywacz has joined #moin-dev
2009-05-30T11:37:09  *** grzywacz has quit IRC
2009-05-30T11:37:25  <ThomasWaldmann> moin
2009-05-30T11:42:35  *** grzywacz has joined #moin-dev
2009-05-30T11:47:14  <ThomasWaldmann> dimazest: please search for a better place for group_manager and stop duplicating code
2009-05-30T11:55:41  <dreimark> moin
2009-05-30T11:56:06  <dreimark> ThomasWaldmann: that second ThomasWaldmann is confusing my url
2009-05-30T11:58:06  <ThomasWaldmann> ?
2009-05-30T11:58:33  <dreimark> http://moinmo.in/ThomasWaIdmann
2009-05-30T11:58:48  <dreimark> the I looks like an l
2009-05-30T12:00:21  <dreimark> see RC
2009-05-30T12:04:49  <ThomasWaldmann> fixed :)
2009-05-30T12:12:11  <dreimark> name dubs is always a problem and we didn't hit them often.
2009-05-30T12:12:45  <ThomasWaldmann> he said he was just joking
2009-05-30T12:13:33  <dreimark> he was knowing that it makes the reading of the url confusing
2009-05-30T12:16:00  <dreimark> that can be softssecurity problem
2009-05-30T12:16:52  <ThomasWaldmann> yeah, but as you see, softsecurity worked :P
2009-05-30T12:17:56  <dreimark> yes. because you and your page are known
2009-05-30T12:18:44  <dreimark> if the page is not such often visible in RC it is difficult to get that is a new one and not the one you know.
2009-05-30T12:18:47  <ThomasWaldmann> dimazest: i don't think the group_manager should be created in security/__init__, but in request class init
2009-05-30T12:19:27  <ThomasWaldmann> (and IIRC I already told you that)
2009-05-30T12:20:39  <dreimark> of course you can look at info action. a solution could be to show the slightly different named one as an additional tag similiar to sisterpages
2009-05-30T12:21:30  <ThomasWaldmann> guess we have better stuff to do
2009-05-30T12:22:03  <dreimark> yes but I will add a FR EasyToDo task about this.
2009-05-30T12:22:36  <ThomasWaldmann> that is not easy
2009-05-30T12:23:19  <ThomasWaldmann> (you would have to cover char similarities considering all fonts and all unicode chars)
2009-05-30T12:23:36  <ThomasWaldmann> so just forget it ;)
2009-05-30T12:23:46  <dreimark> we have a lists already at page creation time
2009-05-30T12:25:10  <ThomasWaldmann> maybe we can concentrate on getting groups done better?
2009-05-30T12:25:17  <dreimark> he
2009-05-30T12:28:02  <ThomasWaldmann> I am not too happy with all that *Manager stuff, maybe it can be done better with some classmethod
2009-05-30T12:28:16  <ThomasWaldmann> see items.Item.create
2009-05-30T12:29:50  <ThomasWaldmann> i started reading the current source and after 5 seconds i find a typo...
2009-05-30T12:33:03  <dreimark> s/avaliable/available
2009-05-30T12:33:53  <dreimark> s/for an containment/for a containment
2009-05-30T12:36:29  <ThomasWaldmann> i am editing the stuff...
2009-05-30T12:37:49  * dreimark don't agree
2009-05-30T13:22:00  <dreimark> bbl
2009-05-30T14:50:03  <ThomasWaldmann> devilsadvocate: one week of SOC has passed and you have failed to visibly work on it (no diary entries, no commits AFAICS)
2009-05-30T14:53:00  <ThomasWaldmann> xorAxAx: ^^
2009-05-30T15:00:10  <ThomasWaldmann> waldi: as amartani didn't work on Mon/Fri he maybe wants to catch up a bit at the weekend
2009-05-30T15:00:51  <ThomasWaldmann> waldi: and I would also appreciate if he finally manages to join us here more often / permanently
2009-05-30T15:02:39  <ThomasWaldmann> also, some more planning on his page would be nice (and maybe general cleanup there, too)
2009-05-30T15:08:02  <ThomasWaldmann> devilsadvocate: ok, you did some minor work on your wiki page, but still: you need to get up to speed
2009-05-30T15:11:08  <ThomasWaldmann> dennda: please check what you consider done (on your project wiki page) and create a plan for 2nd week
2009-05-30T15:22:54  <waldi> class MobwriteWorkerProxy():
2009-05-30T15:22:57  <waldi> oh
2009-05-30T15:23:09  <waldi> i think it misses an "object"
2009-05-30T15:55:48  <dreimark> dimazest: we/you may want to have request.groups so it need to be initialized in contexts and not in security/__init__. I missed that question yesterday so it is half of my fault
2009-05-30T15:55:56  <ThomasWaldmann> waldi: about the "getting rid of the daemon" stuff, did you both decide whether it will be a wsgi app or action?
2009-05-30T15:56:44  <dreimark> dimazest: look how it is done for user and do it similiar
2009-05-30T17:28:28  * ThomasWaldmann refactors RootPage code
2009-05-30T18:12:13  <ThomasWaldmann> dennda: how about s/search_item/search_items/g?
2009-05-30T18:38:07  <dennda> yeah that likely better describes what it does
2009-05-30T18:40:11  <dennda> i'd stay with privilege though
2009-05-30T18:41:32  <dennda> 18:45 < ThomasWaldmann> dennda: handling of non-hierarchic default acl is missing, please check / add test
2009-05-30T18:41:36  <dennda> 18:53 < ThomasWaldmann> hmm, not sure about the default acl. it has to get checked and tested for both implementations.
2009-05-30T18:41:42  <dennda> not sure what you meant here, btw
2009-05-30T19:01:39  <dennda> ThomasWaldmann: we have a problem, btw
2009-05-30T19:02:17  <dennda> ThomasWaldmann: how can a backend implementor specify his own methods? they won't be available on the wrapped item / wrapped backend
2009-05-30T19:04:21  <dennda> i think that can only be done properly by *not* inheriting from backend / item (what does that give us, anyway?) and do some __getattr__ trick. the problem with that is, that every isinstance(foo, Backend) check would fail, but I don't even see a usecase for that
2009-05-30T19:04:35  <dennda> i'll provide a proof of concept
2009-05-30T19:19:26  <dennda> ok that was simple, two lines
2009-05-30T19:20:13  <dennda> the backend implementor should at most define some extra methods on the backend, not on the item anyway...
2009-05-30T19:42:16  <dennda> http://paste.pocoo.org/show/119927/ wtf?
2009-05-30T19:42:32  <dennda> why does that first work and then fail? (see beginning and end, skip the middle part)
2009-05-30T20:09:47  <ThomasWaldmann> dennda: what "own methods"?
2009-05-30T20:12:03  <ThomasWaldmann> from aclfrickeln import awm_maker
2009-05-30T20:12:09  <ThomasWaldmann> ^^ LOL :D
2009-05-30T20:38:43  <dennda> better than from foo ... ;-)
2009-05-30T20:40:11  <dennda> ThomasWaldmann: well the implementor may decide he needs some helper methods that his specialiced items call
2009-05-30T20:40:16  <dimazest> ThomasWaldmann: which code repetition are you talking about? in unittests?
2009-05-30T20:42:20  <ThomasWaldmann> dimazest: if the init is at the right place, you don't need to do it at misc. places
2009-05-30T20:42:21  <dimazest> ThomasWaldmann, dreimark: should i make GroupManager creation there http://hg.moinmo.in/moin/1.9-groups-dmilajevs/file/c3e904a9c00b/MoinMoin/wsgiapp.py#l242 ?
2009-05-30T20:42:51  <dimazest> ThomasWaldmann: i'm waiting for that right place
2009-05-30T20:43:46  <ThomasWaldmann> no
2009-05-30T20:44:50  <dimazest> but where request is created?
2009-05-30T20:45:25  <dimazest> and what is the difference between request and contexts, because dreimark mentioned contexts
2009-05-30T20:48:34  <ThomasWaldmann> dimazest: http://hg.moinmo.in/moin/2.0-storage/rev/eb55e2023c93
2009-05-30T20:48:43  <ThomasWaldmann> do it in a similar way
2009-05-30T20:49:08  <dimazest> thank you
2009-05-30T20:50:54  * dimazest tried this, but changes in MoinMoin/conftest.py were missing, hope now everything will be fine.
2009-05-30T21:13:31  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5815:a2bb7120f197 2.0-storage/MoinMoin/ (Page.py items/__init__.py): items: move most RootPage code to item (list_items, count_items methods)
2009-05-30T22:00:26  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5816:d1aba4f4f304 2.0-storage/MoinMoin/ (Page.py macro/PageCount.py search/results.py web/contexts.py): request.rootitem (use it instead of rootpage)
2009-05-30T22:08:22  <CIA-19> Alexandre Martani <amartani AT gmail DOT com> default * 5806:36acc78e8780 2.0-storage-editor-amartani/MoinMoin/action/mobwrite.py: pep8 fix
2009-05-30T22:10:59  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5807:3653778ad68c 2.0-storage-editor-amartani/MoinMoin/storage/_tests/test_middleware_acl.py: storage: tests: use non-confusing variable name
2009-05-30T22:11:06  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5808:ab9f0d7fdd93 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: Backed out changeset 8f3f50a16d31 (passing through method calls to the real backend)
2009-05-30T22:11:07  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5809:a33db3d28c7e 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: backout
2009-05-30T22:11:09  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5810:16a07731d015 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: Backed out changeset fbe8c3246354; (manual replacing of item._backend attribute)
2009-05-30T22:11:11  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5811:56c695215a1e 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: backout
2009-05-30T22:11:13  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5812:430a38406212 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: Backed out changeset 4f730ebdf78f; (manual replacing of item._backend)
2009-05-30T22:11:18  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5813:1d34d79b8cea 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: backout
2009-05-30T22:11:21  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5814:8cc97414fd7d 2.0-storage-editor-amartani/MoinMoin/ (5 files in 3 dirs): merged remote 2.0-storage
2009-05-30T22:11:24  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5815:26b09e362a5d 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: storage: Introducing AclWrapperItem. Not really functional yet.
2009-05-30T22:11:28  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5816:d1b02cc816bf 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: storage: AclWrapperItem refactoring. Introduce require_privilege decorator
2009-05-30T22:11:34  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5817:a6f0334b9965 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: storage: AclWrapperItem: Now completely covered the public Item interface
2009-05-30T22:11:40  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5818:7ca81a8e9ac0 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: storage: Adding myself to authors for MoinMoin.storage.backends.acl
2009-05-30T22:11:44  <CIA-19> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 5819:160e1c83245e 2.0-storage-editor-amartani/MoinMoin/storage/backends/acl.py: storage: Make AclWrapperBackend.history only return readable revisions. Fix 2 permission checks
2009-05-30T22:11:48  <CIA-19> Alexandre Martani <amartani AT gmail DOT com> default * 5826:0a72eed33934 2.0-storage-editor-amartani/ (27 files in 16 dirs): merge 2.0-storage
2009-05-30T22:16:15  <dimazest> now the problem is, that the group_manager_init the function which creates group_manager, is needed to be defined in the Config class like http://paste.pocoo.org/show/119954/
2009-05-30T22:17:18  <dimazest> if i define it like in the http://hg.moinmo.in/moin/1.9-groups-dmilajevs/file/c3e904a9c00b/MoinMoin/groups/_tests/test_group_manager.py#l46 then the value from the wikiconfig is taken
2009-05-30T22:18:25  <dimazest> i do not like that admin_group and so on became global variables (in the snippet on the pocoo), any suggestions?
2009-05-30T22:24:18  <ThomasWaldmann> dimazest: what's the problem with it being defined in the Config?
2009-05-30T22:26:06  <ThomasWaldmann> and of course the stuff shall not be globals, please test realistic scenarios (stuff in Config)
2009-05-30T22:26:58  <dimazest> if i define group_manager_init in setup_class, conftest.init_test_request doesn't see it, and use value from the MoinMoin._tests.wikiconfig
2009-05-30T22:27:21  <ThomasWaldmann> then don't do it
2009-05-30T22:28:44  <dimazest> not to define it in the wikiconfig?
2009-05-30T22:29:13  <ThomasWaldmann> and the trick to make the lambda work is lambda ..., cfg=self: ... cfg.first_backend_groups ...
2009-05-30T22:30:39  <ThomasWaldmann> the Config class is to define test specific configuration
2009-05-30T22:31:00  <ThomasWaldmann> (the one you define within the test class)
2009-05-30T22:41:13  <dimazest> ThomasWaldmann: used backends is test specific configuration, how is this variant? http://paste.pocoo.org/show/119963/
2009-05-30T22:42:54  <ThomasWaldmann> move stuff into Config class
2009-05-30T22:44:24  <dimazest> but it is used by the test
2009-05-30T22:45:04  <dimazest> or then, it can be accessed by self.Config.whatever?
2009-05-30T22:46:01  <ThomasWaldmann> self.request.cfg, as usual
2009-05-30T22:49:56  <ThomasWaldmann> dennda: we have to think about search_item ACL behaviour (the result list maybe shall not contain items the user can't read [and shall not even see])
2009-05-30T22:51:14  <ThomasWaldmann> ah, ignore that, the wrapper for iteritems already does it
2009-05-30T22:59:53  * dimazest can share local repo http://87.15.110.98:8000 
2009-05-30T23:02:48  <CIA-19> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5817:6940a95a3d3c 2.0-storage/MoinMoin/action/titleindex.py: make titleindex action use the rootitem
2009-05-30T23:28:03  <dimazest> dreimark: is it ok if in the diary i publish diffs. i do not want to commit, because, i use mercurial queues, so i can change history, in the case of typos or wrong implementation