MoinMoin: MoinMoinChat/Logs/moin-dev/2007-05-03

2007-05-03T00:23:06  <lanius> n8
2007-05-03T08:06:41  <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2042:3fd01751ebf8 1.6/MoinMoin/PageEditor.py: PageEditor: bug fix for MyPages (wants to be tested)
2007-05-03T08:56:20  <dreimark> moin
2007-05-03T08:59:41  <ThomasWaldmann> moin dreimark
2007-05-03T09:00:08  <ThomasWaldmann> fixing that  stuff is a bit of a problem...
2007-05-03T09:01:41  <dreimark> I am not sure what is better fixing or removing
2007-05-03T09:02:19  <dreimark> it could be I guess totally replaced by the hierarchicaly patch
2007-05-03T09:03:12  <ThomasWaldmann> not easily
2007-05-03T09:05:58  <dreimark> if ME could be used I think I like to have a config var where it is on default disabled
2007-05-03T09:13:16  <ThomasWaldmann> or just using ##acl ... in the template and let the use "activate" it
2007-05-03T09:23:30  <dreimark> a switch gives more control over that feature, if its on or off some priviliged user can change the acl line too
2007-05-03T09:58:21  <ThomasWaldmann> i dont think i like that stuff. it does some magic and is not very precise about it.
2007-05-03T10:00:01  <ThomasWaldmann> and the code does pi processing although they are already processed. it also reads the page text twice.
2007-05-03T10:01:56  <dreimark> I do think I would prefer removing MyPages completly and if one likesto have something like this it could (not) easily but it could be done by hierarchical acls
2007-05-03T10:02:19  <ThomasWaldmann> it also does not really check the acls given to  (although the assumption will be mostly correct that they allow reading)
2007-05-03T10:04:09  <ThomasWaldmann> using hacls would need some intermediate page, like UserHomePage/ReadAble/SomeReadAblePage
2007-05-03T10:04:34  <ThomasWaldmann> and .../WriteAble/...
2007-05-03T10:06:14  <ThomasWaldmann> i would suggest we just add the # to the template page until we do templates differently
2007-05-03T10:07:53  <ThomasWaldmann> we can work on using some page cloning method as base for templating
2007-05-03T10:08:09  <ThomasWaldmann> (but not for 1.6)
2007-05-03T10:09:32  <ThomasWaldmann> (maybe even including sub-items on demand)
2007-05-03T10:15:12  <dreimark> ok, I do revert that patch and add that # change to the templates
2007-05-03T10:15:24  <dreimark> bbl
2007-05-03T10:29:15  <dreimark> bb
2007-05-03T12:20:29  <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2043:ea071f6f600a 1.6/MoinMoin/PageEditor.py: PageEditor: reverted 2042:3fd01751ebf8
2007-05-03T12:26:08  <dreimark> ThomasWaldmann: a comment on the template page like:  If you want this page saved as private page remove the comment sign from the acl line
2007-05-03T12:26:11  <dreimark> ?
2007-05-03T12:36:16  <ThomasWaldmann> sure
2007-05-03T12:45:28  <neagulm> hello
2007-05-03T12:45:50  <neagulm> dreimark, I working now on PoCv2
2007-05-03T12:46:16  <neagulm> dreimark, I have abandoned the idea of decorating the the Page() instance
2007-05-03T12:46:55  <neagulm> dreimark, my motivation is based on Knuth: "Premature optimization is the root of all evil" :)
2007-05-03T12:47:18  <dreimark> ThomasWaldmann: ohh, you have to change the acl stuff on the Template Pages.
2007-05-03T12:47:24  <dreimark> moin neagulm
2007-05-03T12:48:05  <neagulm> hello dreimark
2007-05-03T12:52:33  <xorAxAx> neagulm: note that you will have to merge with the storage refactoring code
2007-05-03T12:52:45  <xorAxAx> so track the current discussion a bit
2007-05-03T12:53:00  <neagulm> dreimark, k
2007-05-03T12:53:08  <xorAxAx> but indeed, there should not be too many issues with this as your code is one level above the storage layer
2007-05-03T12:55:56  <dreimark> neagulm: is it abandonded because of that's the design can't be as good as possible?
2007-05-03T12:56:51  <neagulm> neagulm, It is abandoned because I don't know enough MoinMoin internals :(
2007-05-03T12:56:59  <neagulm> dreimark,  It is abandoned because I don't know enough MoinMoin internals :(
2007-05-03T12:57:30  <xorAxAx> well, you have 4 weeks to get them to know :)
2007-05-03T12:57:42  <neagulm> dreimark, And I want my code as modular as possible
2007-05-03T12:58:16  <dreimark> neagulm: modularisation is a good reason
2007-05-03T12:58:55  <dreimark> can you please subscribe your wiki pages so if one leaves a message there you get informed
2007-05-03T13:09:59  <ThomasWaldmann> neagulm: and you can just ask for the stuff that is not obvious from the code
2007-05-03T13:44:52  <dreimark> neagulm: I guess someone on this channel can answer most of your questions :)
2007-05-03T13:53:08  <ThomasWaldmann> dreimark: acl stuff fixed on moinmaster
2007-05-03T13:54:50  <ThomasWaldmann> dreimark: yeah, even the one for sense of life, the universe and everything. But you have to find out yourself why the answer is 42.
2007-05-03T13:56:06  <dreimark> ThomasWaldmann: do you have already a mig script?
2007-05-03T13:57:17  <ThomasWaldmann> not really
2007-05-03T14:43:38  <neagulm> how can I get Config without using request ?
2007-05-03T14:57:18  <ThomasWaldmann> neagulm: if you mean the stuff from wikiconfig, the only way is using request.cfg
2007-05-03T17:38:59  <dreimark> ThomasWaldmann: what is the replacement for Page,clean_acl_cache ?
2007-05-03T17:42:37  <dreimark> if its only watching the edit-log then we may have to add to filesys.copytree something to create a log entry or ?
2007-05-03T17:43:15  <ThomasWaldmann> if you create a page, it should be visible in the edit-log, yes :)
2007-05-03T17:43:47  <ThomasWaldmann> (btw, if you work on the new templating stuff: in 1.7, please :)
2007-05-03T17:44:03  <flowhase> hmm?
2007-05-03T17:44:15  <flowhase> moin there btw.
2007-05-03T17:46:41  <dreimark> ThomasWaldmann: I know, currently I do check /MoinMoinBugs/1.6devStrangeACLbehaviour
2007-05-03T17:46:41  <flowhase> ah, you meant the other kind of templating :>
2007-05-03T17:46:41  <dreimark> flowhase: SomethingTemplate
2007-05-03T17:46:42  <dreimark> and I think xorAxAx is right,if I do first newpage = PageEditor(request, newpagename)
2007-05-03T17:46:42  <dreimark> I got not no acl protections if that page not exist
2007-05-03T17:47:21  <ThomasWaldmann> flowhase: moin templates, not related to your stuff :)
2007-05-03T17:47:30  <dreimark> and if I  then later use filesys.copytree(oldpath, newpath)
2007-05-03T17:47:53  <flowhase> ThomasWaldmann: yeah, i realized
2007-05-03T17:48:42  <dreimark> then I have not at this time changed the log entry
2007-05-03T17:49:30  <dreimark> so  request.user.may.write(newpagename): gets the old acls
2007-05-03T17:50:16  <dreimark> Page,clean_acl_cache is used in packages too and needs to be checked there too
2007-05-03T17:50:35  <xorAxAx> ?
2007-05-03T17:53:50  <dreimark> I was wondering in the past why it doesn't work to check request.user.may.write(newpagename): in PageEditor.copyPage while oldpagename works
2007-05-03T17:55:28  <dreimark> and both should have because of the copy the same acl settings
2007-05-03T17:55:30  <xorAxAx> and copyPage is using the clean call?
2007-05-03T17:56:49  <dreimark> the clean call is changed to only pass      # should not be necessary any more as the new cache watches edit-log for changes
2007-05-03T17:57:55  <xorAxAx> OMG
2007-05-03T17:58:19  <xorAxAx> write a test for the issue, its very critical
2007-05-03T18:05:41  <dreimark> I know and I am not knowing currently if we can abandon a separate clean cache routine
2007-05-03T18:08:50  <xorAxAx> well, if the assumption is wrong, find out why, if not, find out why the caching code is wrong
2007-05-03T18:22:17  <dreimark> bbl
2007-05-03T22:47:02  <dreimark> bb