MoinMoin: MoinMoinChat/Logs/moin-dev/2009-07-11

2009-07-11T01:11:14  <amartani> is there any locking when editing items already implemented on 2.0?
2009-07-11T01:53:20  *** grzywacz has quit IRC
2009-07-11T05:56:17  *** amartani_ has joined #moin-dev
2009-07-11T05:56:17  *** amartani has quit IRC
2009-07-11T06:46:01  <CIA-38> Alexandre Martani <amartani AT gmail DOT com> default * 6104:f2f57e2d034a 2.0-storage-editor-amartani/MoinMoin/ (2 files in 2 dirs): Real-time editor: Enable mobwrite on enable sharing button
2009-07-11T06:46:03  <CIA-38> Alexandre Martani <amartani AT gmail DOT com> default * 6105:dcd84832416f 2.0-storage-editor-amartani/MoinMoin/ (3 files in 3 dirs): Real-time editor: Check mobwrite availability before enable sharing
2009-07-11T06:48:09  *** amartani_ has quit IRC
2009-07-11T07:34:34  *** devilsadvocate has quit IRC
2009-07-11T07:34:55  *** devilsadvocate has joined #moin-dev
2009-07-11T08:17:17  *** TheSheep has quit IRC
2009-07-11T08:17:34  *** TheSheep has joined #moin-dev
2009-07-11T08:42:44  <ThomasWaldmann> moin
2009-07-11T08:43:23  <TheSheep> moin
2009-07-11T09:06:55  *** devilsadvocate has quit IRC
2009-07-11T09:06:55  *** ser has quit IRC
2009-07-11T09:06:55  *** CIA-38 has quit IRC
2009-07-11T09:06:55  *** dreimark has quit IRC
2009-07-11T09:07:44  *** devilsadvocate has joined #moin-dev
2009-07-11T09:09:19  *** dreimark has joined #moin-dev
2009-07-11T09:09:25  *** ser has joined #moin-dev
2009-07-11T09:09:25  *** CIA-38 has joined #moin-dev
2009-07-11T09:10:18  *** ser has quit IRC
2009-07-11T09:10:25  *** CIA-78 has joined #moin-dev
2009-07-11T09:10:25  *** ser has joined #moin-dev
2009-07-11T09:12:23  *** CIA-38 has quit IRC
2009-07-11T12:35:24  <CIA-78> Dmitrijs Milajevs <dimazest@gmail.com> default * 4835:193bbbdd0ace 1.9-groups-dmilajevs/MoinMoin/formatter/_tests/test_groups.py: Groups2009: Cosmetic changes in MoinMoin.formatter.groups tests: variable and method names.
2009-07-11T12:35:26  <CIA-78> Dmitrijs Milajevs <dimazest@gmail.com> default * 4836:916c28ef9026 1.9-groups-dmilajevs/MoinMoin/datastruct/backends/wiki_groups.py: Groups2009: Bug fix. Access to the group page do not cause infinite recursion.
2009-07-11T12:43:49  *** grzywacz has joined #moin-dev
2009-07-11T13:12:33  <dimazest> ThomasWaldmann: dreimark may i use this code? http://moinmo.in/DictsFormatter
2009-07-11T13:40:00  <dennda> Guys, don't forget to fill out the evaluation form...
2009-07-11T13:40:27  * dennda needs to do that as well
2009-07-11T13:50:25  <xorAxAx> or dont do it if you want to see googles revenue increase :)
2009-07-11T13:57:29  <dennda> ThomasWaldmann: How will we be feeding new syspages to wikis? By a script?
2009-07-11T14:53:22  <ThomasWaldmann> dennda: i think we'll have a bunch of xml files (different sized sets, per language) and some action to feed them into the wiki
2009-07-11T14:55:33  <ThomasWaldmann> and a script (so basically similar to now, just not zip packages and not a full set of english pages preinstalled)
2009-07-11T14:57:28  <dimazest> ThomasWaldmann: dreimark there is a problem with dicts
2009-07-11T14:57:42  <dimazest> if dict page is written in wiki format
2009-07-11T14:57:58  <dimazest> and dict is used on a page which uses creaole
2009-07-11T14:58:16  <dimazest> what should dict return?
2009-07-11T14:59:06  <ThomasWaldmann> you mean if you use markup in the values?
2009-07-11T14:59:11  <dimazest> i think, it should be kind of tree of element, which then are formatted to wiki/creole/whatever
2009-07-11T14:59:16  <dimazest> yes
2009-07-11T14:59:29  <ThomasWaldmann> you shouldn't care for that now
2009-07-11T15:19:21  <CIA-78> Dmitrijs Milajevs <dimazest@gmail.com> default * 4837:982f706482e7 1.9-groups-dmilajevs/MoinMoin/ (4 files in 3 dirs):
2009-07-11T15:19:24  <CIA-78> Groups2009: MoinMoin.formatter.dicts formatter was introduced. It collects dict
2009-07-11T15:19:27  <CIA-78> items defined on a page. MoinMoin.datastruct.backends.wiki_dicts uses this
2009-07-11T15:19:31  <CIA-78> formatter to get items from a dict page. _dict_page_parse_regex was killed.
2009-07-11T15:35:49  <dennda> ThomasWaldmann: ok so we can give them the proper metadata ourselves already
2009-07-11T16:51:21  <ThomasWaldmann> hm?
2009-07-11T16:52:33  * ThomasWaldmann completed the surveys and gave google some non-secrets from his past.
2009-07-11T16:55:44  <ThomasWaldmann> (and became aware about that his first "coding" must have been about 30y ago O_O )
2009-07-11T16:59:00  <ThomasWaldmann> http://www.ti59.com/ < my brother had one of those
2009-07-11T17:53:12  <dennda> You're old...
2009-07-11T18:38:50  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6034:543e8ee9a968 2.0-storage/DEVELOPER_INFORMATION: storage: Remove obsolete comments from DEVELOPER_INFORMATION
2009-07-11T18:38:53  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6035:4872a98e6dd7 2.0-storage/MoinMoin/storage/__init__.py: storage: API CHANGE: Add item.erase(). This method attempts to erase an Item from disk.
2009-07-11T18:38:55  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6036:449b5f58cc32 2.0-storage/MoinMoin/storage/backends/memory.py: storage: Implement MemoryBackend._erase_item
2009-07-11T18:38:56  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6037:fc4df9879818 2.0-storage/MoinMoin/storage/ (__init__.py backends/router.py): storage: Make RouterItem.rename work for cross backend renames. Fixes last failing routerbackend test.
2009-07-11T18:38:58  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6038:7e2c832d98c9 2.0-storage/MoinMoin/storage/backends/acl.py: storage: Handle Item.erase in AMW
2009-07-11T18:39:03  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6039:65d4d6106e51 2.0-storage/MoinMoin/storage/backends/router.py: storage: Disable RouterItem.erase until someone has a use-case for that
2009-07-11T18:39:08  <dennda> RFC ^^^
2009-07-11T18:42:51  <dreimark> moin
2009-07-11T18:52:52  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6040:7d258d03447c 2.0-storage/MoinMoin/storage/_tests/test_backends.py: storage: Add erasure testcase
2009-07-11T18:57:17  <dreimark> dennda: is it possible with the router backend to have some pages using hierarchical acls and some others standard definition?
2009-07-11T19:50:11  <ThomasWaldmann> omg X)
2009-07-11T20:02:25  <ThomasWaldmann> (that was referring to even more complex acl setups)
2009-07-11T20:02:37  * ThomasWaldmann looks at dennda's changesets now
2009-07-11T20:04:47  <ThomasWaldmann> dennda: decide whether it is erase or nuke
2009-07-11T20:06:44  * ThomasWaldmann likes both with a slight tendency towards nuke
2009-07-11T20:09:29  <ThomasWaldmann> some comments:
2009-07-11T20:10:20  <ThomasWaldmann> you didn't really document history behaviour in the base class, but in the memory backend class one sees that you also nuke the history related to the item
2009-07-11T20:13:14  * ThomasWaldmann would like to have a 'nuke'/'kill'/'destroy' permission related to that
2009-07-11T20:13:33  <ThomasWaldmann> destroy also sounds nice as opposite of create
2009-07-11T20:14:45  <ThomasWaldmann> http://hg.moinmo.in/moin/2.0-storage/rev/7e2c832d98c9 doing inter-backend moves is not the only application for nuke
2009-07-11T20:15:10  <dennda> what privs would rename need, then?
2009-07-11T20:15:20  <ThomasWaldmann> e.g. if some idiot put something unlawful into your wiki, you also rather might want to erase that rather than getting sued
2009-07-11T20:16:14  <dennda> sure
2009-07-11T20:16:17  <ThomasWaldmann> and for practical and symmetry reasons, we need nuke_item (see create_item) and nuke_revision (see create_revision)
2009-07-11T20:17:37  <ThomasWaldmann> rename is a special case. while in the end it does copy/nuke, it also has knowledge that it first successfully copied the item, thus we can have a more relaxed check than just for nuke.
2009-07-11T20:19:18  <ThomasWaldmann> (another reason why we need nuke_revision is thinning out history from time to time)
2009-07-11T20:19:52  <ThomasWaldmann> (esp. for large binary items with many revisions this is rather important)
2009-07-11T20:21:19  <ThomasWaldmann> btw, for interbackend rename, it must not kill all the history, but rather rename that, to
2009-07-11T20:21:23  <ThomasWaldmann> o
2009-07-11T20:23:05  <dreimark> s/destroy/destruct/ ?
2009-07-11T20:23:19  <ThomasWaldmann> ok, so maybe we first agree on the name
2009-07-11T20:23:34  <ThomasWaldmann> i think the verb is destroy
2009-07-11T20:24:45  <ThomasWaldmann> hmm, both is valid english, but I think destroy is more common, right?
2009-07-11T20:25:35  <dreimark> hmm it sounds a bit too military
2009-07-11T20:26:22  <dreimark> but that is not so important destroy is ok
2009-07-11T20:27:10  <dennda> ldestroy is good
2009-07-11T20:29:04  <dennda> ThomasWaldmann: 20:21, don't get that
2009-07-11T20:30:27  <ThomasWaldmann> well, the item still exists afterwards (in the other backend)
2009-07-11T20:32:17  <dennda> other == new or old?
2009-07-11T20:33:57  <ThomasWaldmann> new
2009-07-11T20:34:43  <dennda> well, that is the intention of rename, isn't it?
2009-07-11T20:35:20  <dennda> the item is copied with all its history
2009-07-11T20:42:04  <dennda> dreimark: afaict hierarchic is not a per backend setting
2009-07-11T20:42:16  <dennda> hence: not possible
2009-07-11T20:42:29  <ThomasWaldmann> dennda: it is ok if all history is preserved, maybe i misunderstood your code
2009-07-11T20:42:58  <dennda> i think so. copy_item copies revs as well
2009-07-11T20:43:27  <ThomasWaldmann> dennda: dreimark: i think that would be too complex for most users anyway. quite some don't get acls anyway, thus no need to make them even more complex.
2009-07-11T20:43:46  <dennda> i agree
2009-07-11T20:55:04  <dennda> 20:10 < ThomasWaldmann> you didn't really document history behaviour in the base class, but in the memory backend class one sees that you also nuke the history related to the item
2009-07-11T20:55:09  <dennda> I thought that was obvious
2009-07-11T20:55:14  <dennda> but I'll make it more explicit
2009-07-11T21:06:06  <dreimark> ThomasWaldmann: ok
2009-07-11T21:34:09  * dennda wants the werkzeug debugger
2009-07-11T21:45:05  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6041:4c7ac48b970b 2.0-storage/MoinMoin/storage/ (5 files in 3 dirs): storage: erase -> destroy
2009-07-11T21:45:07  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6042:7bf923f18603 2.0-storage/MoinMoin/storage/backends/router.py: storage: OK, let's expose RouterItem.destroy() and hope the new name raises awareness as to the severeness of the method
2009-07-11T21:45:09  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6043:3f51b8ef51d2 2.0-storage/MoinMoin/storage/ (__init__.py backends/memory.py): storage: Add clarifying comment
2009-07-11T21:45:10  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6044:c69b42e0d8ac 2.0-storage/MoinMoin/items/__init__.py: storage: Fix for traceback that was thrown for implicit revs
2009-07-11T22:01:43  <dennda> hm
2009-07-11T22:01:56  <dennda> I suggest destroy -> harakiri
2009-07-11T22:02:04  <dennda> or suicide
2009-07-11T22:02:07  <dennda> :-)
2009-07-11T22:06:18  <ThomasWaldmann> or slay
2009-07-11T22:19:29  <dennda> gah
2009-07-11T22:20:06  <dennda> now I need to create an aclwrapper just for StoredRevision.destroy()...
2009-07-11T22:21:20  <dennda> hmm
2009-07-11T22:37:32  *** amartani has joined #moin-dev
2009-07-11T22:44:24  * dennda loves ducktyping
2009-07-11T22:46:10  <ThomasWaldmann> at the europython hotel we had real ducks :D
2009-07-11T22:51:03  <dennda> How do I formally introduce a new privilege?
2009-07-11T22:55:38  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6045:bed6641ca8ae 2.0-storage/MoinMoin/storage/__init__.py: storage: API CHANGE: Introduce StoredRevision.destroy()
2009-07-11T22:55:39  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6046:0b10a0b56f14 2.0-storage/MoinMoin/storage/ (__init__.py _tests/test_backends.py): storage: Testcase for revision.destroy
2009-07-11T22:55:41  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6047:98c1135a5fcd 2.0-storage/MoinMoin/storage/backends/memory.py: storage: Implement MemoryBackend._destroy_revision
2009-07-11T22:55:44  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6048:6fed017b7293 2.0-storage/MoinMoin/storage/ (_tests/test_middleware_acl.py backends/acl.py): storage: Informally introduce DESTROY privilege and make acl code aware of Revision.destroy
2009-07-11T22:55:48  <CIA-78> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com> default * 6049:76e97217f509 2.0-storage/MoinMoin/storage/backends/router.py: storage: Implement RouterRevision.destroy()
2009-07-11T22:56:27  <dennda> ok, now that makes quite a few tests fail because a) 'destroy' priv is not yet formally introduced to moin and b) several backends don't yet support Revision.destroy() and Item.destroy()
2009-07-11T23:00:42  <dennda> The former I have to fix, but I don't know hg/fs backends internally
2009-07-11T23:30:20  *** amartani has quit IRC
2009-07-11T23:30:29  *** amartani has joined #moin-dev