MoinMoin: MoinMoinChat/Logs/moin-dev/2009-04-26

2009-04-26T00:58:16  *** amartani has left #moin-dev
2009-04-26T01:23:43  *** dimazest_ has joined #moin-dev
2009-04-26T01:25:56  *** dimazest has quit IRC
2009-04-26T03:17:48  *** dimazest has joined #moin-dev
2009-04-26T03:33:55  *** dimazest_ has quit IRC
2009-04-26T05:37:53  *** dimazest_ has joined #moin-dev
2009-04-26T05:38:14  *** dimazest has quit IRC
2009-04-26T05:54:55  *** devilsadvocate_ has joined #moin-dev
2009-04-26T05:55:07  *** devilsadvocate has quit IRC
2009-04-26T06:23:04  *** amartani has joined #moin-dev
2009-04-26T07:02:53  *** amartani has left #moin-dev
2009-04-26T07:35:18  *** devilsadvocate_ has quit IRC
2009-04-26T07:37:41  *** devilsadvocate has joined #moin-dev
2009-04-26T09:56:19  *** devilsadvocate has quit IRC
2009-04-26T09:58:11  *** devilsadvocate has joined #moin-dev
2009-04-26T10:35:24  <xorAxAx> devilsadvocate: then it would disallow user modifications, right? you should probably aim for a solution which allows the user to edit the config and use the frontend
2009-04-26T10:36:02  <devilsadvocate> xorAxAx, user to edit the intermediate as well?
2009-04-26T10:36:18  <devilsadvocate> the user can still edit the .py files by hand
2009-04-26T10:37:25  <devilsadvocate> the files can even be exposed in the web front end itself as it is, if required. but that'd be a tad bit insecure (and there is always the possibility that tehy screw something up and cant reach it anymore)
2009-04-26T10:38:24  <xorAxAx> devilsadvocate: no, the user would want to edit the .py file
2009-04-26T10:38:36  <xorAxAx> devilsadvocate: but if he edits the py file, how do you get back the information?
2009-04-26T10:38:46  <xorAxAx> he might shuffle it completely and compute the values programmatically
2009-04-26T10:39:24  <devilsadvocate> so basically you're asking about how changes directly to the .py will come back into the intermediate
2009-04-26T10:39:30  <devilsadvocate> hm.
2009-04-26T10:39:36  <xorAxAx> thats why i suggest to have some kind of inheritance level in between, i.e. the gui-values are loaded after the default ones and then the .py file is loaded
2009-04-26T10:39:55  <xorAxAx> and no .py code generation is done
2009-04-26T10:41:27  <devilsadvocate> default ones as in moin defaults that are currently outside of wikiconfig.py and farmconfig.py?
2009-04-26T10:41:51  <xorAxAx> yes
2009-04-26T10:42:12  <xorAxAx> currently thats all done by python and the subclassing
2009-04-26T10:42:17  <xorAxAx> which is a bit weird
2009-04-26T10:43:00  <xorAxAx> but setting the attributes correctly isnt hard at all
2009-04-26T10:43:12  <xorAxAx> in python that is
2009-04-26T10:43:31  <xorAxAx> also yaml has the disadvantage that it needs a new dependency, right?
2009-04-26T10:43:58  <devilsadvocate> yes, it adds dependancies. the only reason i was considering is that google app engine uses yaml
2009-04-26T10:44:16  <devilsadvocate> i thought it might make things easier in the future
2009-04-26T10:44:42  <devilsadvocate> (if and when we go for GAE integration
2009-04-26T10:44:43  <devilsadvocate> )
2009-04-26T10:45:42  <devilsadvocate> to do this sort of inheritance without making major changes to the moin core, i'd have to do something similar to the current python subclassing, right?
2009-04-26T11:08:02  *** grzywacz has joined #moin-dev
2009-04-26T11:42:13  <ThomasWaldmann> moin
2009-04-26T11:46:30  <ThomasWaldmann> about yaml: i would not be too unhappy with that dependency, if we can bundle it (pure python)
2009-04-26T11:46:48  <ThomasWaldmann> i had the idea of doing more with yaml anyway
2009-04-26T11:50:50  <ThomasWaldmann> otoh, for the automatically created part of the config, i do not see that much advantage by using yaml
2009-04-26T11:51:39  <ThomasWaldmann> it could be just solved by another inheritance level
2009-04-26T11:54:20  <ThomasWaldmann> xorAxAx: devilsadvocate: ^^
2009-04-26T11:57:02  <xorAxAx> devilsadvocate: no, you would still subclass but have a small function that is called later on which does the setattrs
2009-04-26T11:57:04  <ThomasWaldmann> e.g. builtin -> autogen -> user
2009-04-26T11:58:59  <ThomasWaldmann> just patching in stuff with setattr reminds me a bit of the crappy test config code we had for some years :)
2009-04-26T12:15:16  <ronny> yaml is no fun
2009-04-26T12:25:18  *** cdyson37 has joined #moin-dev
2009-04-26T12:25:27  <ThomasWaldmann> ronny: hm, why?
2009-04-26T12:26:16  <ronny> ThomasWaldmann: many special cases that end up surprising
2009-04-26T12:27:30  <ronny> let me find the whole article
2009-04-26T12:59:36  *** devilsadvocate has quit IRC
2009-04-26T13:30:14  <dimazest_> hi
2009-04-26T13:31:12  <dimazest_> are Group classes in http://hg.moinmo.in/moin/0.0/1.8-ldapgroups-mmihaljevic/file/c9a888a7bdb8/MoinMoin/datastruct/__init__.py#l18 and http://hg.moinmo.in/moin/1.9/file/efde74175d2d/MoinMoin/wikidicts.py#l68
2009-04-26T13:31:25  <dimazest_> supposed to do similar thing?
2009-04-26T13:34:49  <ThomasWaldmann> yes
2009-04-26T13:36:39  <ThomasWaldmann> except that the old code intermixes group storage specific code with group code
2009-04-26T13:38:42  *** devilsadvocate has joined #moin-dev
2009-04-26T14:43:31  <ThomasWaldmann> dimazest_: ?
2009-04-26T14:44:57  <dimazest_> ThomasWaldmann: old code is in the 1.9?
2009-04-26T14:45:32  <dreimark> dimazest_: old code of wikidicts
2009-04-26T14:45:55  <dreimark> in current 1.9
2009-04-26T14:46:47  <dreimark> +is
2009-04-26T14:46:47  <dimazest_> should Group code from the wikidicts be moved to the datastruct?
2009-04-26T14:47:01  <dreimark> yes
2009-04-26T14:57:39  <ThomasWaldmann> dimazest_: not just moved, but refactored and cleaned up
2009-04-26T14:58:04  <dreimark> oops
2009-04-26T15:08:07  <dreimark> bbl
2009-04-26T16:32:23  *** amartani has joined #moin-dev
2009-04-26T17:42:35  <dreimark> re
2009-04-26T18:29:03  *** cdyson37 has quit IRC
2009-04-26T19:18:17  *** amartani has quit IRC
2009-04-26T19:37:40  * ThomasWaldmann works on 1.8.3
2009-04-26T20:06:30  *** amartani has joined #moin-dev
2009-04-26T20:32:00  *** amartani has quit IRC
2009-04-26T20:40:39  *** devilsadvocate has quit IRC
2009-04-26T20:40:56  *** devilsadvocate has joined #moin-dev
2009-04-26T20:50:24  *** amartani has joined #moin-dev
2009-04-26T21:24:40  <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4322:7d1c139a52b7 1.8/docs/CHANGES: updated CHANGES
2009-04-26T21:55:22  * dreimark is blocked by a missing realtime editor ;)
2009-04-26T21:57:18  <xorAxAx> ThomasWaldmann: "patching in"?
2009-04-26T21:57:24  <xorAxAx> ThomasWaldmann: thats how subclassing works
2009-04-26T21:57:36  <xorAxAx> which is the hacky part of the current/legacy design IMHO
2009-04-26T22:10:06  <ThomasWaldmann> xorAxAx: the old test config code did patch in the attrs, but did not call __init__ of the classes AFTERWARDS
2009-04-26T22:10:11  *** devilsadvocate has quit IRC
2009-04-26T22:12:45  *** devilsadvocate has joined #moin-dev
2009-04-26T22:13:39  <xorAxAx> ThomasWaldmann: ok. well, in this case the solution would be an integral part of the config system and not a hacky testcode thing
2009-04-26T22:56:00  *** dimazest has joined #moin-dev
2009-04-26T23:13:30  *** dimazest_ has quit IRC
2009-04-26T23:26:19  *** devilsadvocate has quit IRC
2009-04-26T23:59:48  <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4323:f6607543762e 1.8/MoinMoin/i18n/ (43 files): updated i18n