2009-07-25T00:05:21 <ThomasWaldmann> dennda: using a relative path for the dynamically created backends sucks
2009-07-25T00:05:52 <ThomasWaldmann> everywhere you run py.test <whatever>, you'll find instance/ and hgback/ later
2009-07-25T00:07:52 <dennda> hm
2009-07-25T00:08:58 <dennda> obviously I can't just use /home/dennda/dev/py/moin/2.0-storage/instance... :-)
2009-07-25T00:09:02 <dennda> thinking about a sane solution
2009-07-25T00:09:08 <ThomasWaldmann> __file__
2009-07-25T00:10:35 <dennda> that's just the filename
2009-07-25T00:11:15 <dennda> (i.e., relative as well)
2009-07-25T00:11:32 <CIA-43> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 6275:bf782347bea8 2.0-storage/MoinMoin/search/builtin.py: fix a search test failure: request.editlog is not needed any more
2009-07-25T00:14:07 <dennda> os.getcwd
2009-07-25T00:17:38 <dennda> ThomasWaldmann: I don't actually see where those folders should come from, though. The tests have a teardown_method that performs kill_backend
2009-07-25T00:21:57 <ThomasWaldmann> try it
2009-07-25T00:25:21 <ThomasWaldmann> dimazest: dreimark:
2009-07-25T00:25:22 <ThomasWaldmann> # Clear caches
2009-07-25T00:25:22 <ThomasWaldmann> # TODO Code from MoinMoin/script/maint/cleancache.py may be used
2009-07-25T00:25:22 <ThomasWaldmann> - page.clean_acl_cache() # It is not necessary should be removed.
2009-07-25T00:25:32 <ThomasWaldmann> why is the comment still there?
2009-07-25T00:25:51 <ThomasWaldmann> can we finally clear whether it's needed or not?
2009-07-25T00:27:07 <CIA-43> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 6282:84c824e4ae99 2.0-storage/MoinMoin/ (26 files in 8 dirs): merged moin/1.9
2009-07-25T00:32:05 <dennda> for the record: http://the-space-station.com/~dennda/tmp/underlay_version_1_from_19.xml -- 12.5 MB
2009-07-25T00:33:41 <ThomasWaldmann> lets see how FF swallows that
2009-07-25T00:36:25 <dennda> takes a few moments for me
2009-07-25T00:36:28 <ThomasWaldmann> ok
2009-07-25T00:36:35 <dennda> the bloody thing
2009-07-25T00:36:39 <dennda> chrome wouldn't die that easily
2009-07-25T00:38:34 <ThomasWaldmann> ff didn't die
2009-07-25T00:38:52 <dennda> well not "die" in a "crash" sense, but in a "become completely unresponsive" way
2009-07-25T00:39:04 <dennda> because it's one process for all tabs
2009-07-25T01:00:22 <dennda> ThomasWaldmann: Should I find out how to do the syspages UI or do you want to/can you do that?
2009-07-25T01:18:46 <ThomasWaldmann> you mean the tagging script?
2009-07-25T01:20:02 <dennda> no
2009-07-25T01:20:08 <dennda> I mean the interface for the admin
2009-07-25T01:20:53 *** grzywacz has quit IRC
2009-07-25T01:21:32 <dennda> without that I can't really implement upgrade_syspages because I need access to the backend defined in wikiconfig and I have no idea how I would access that from MoinMoin.storage
2009-07-25T01:24:04 <ThomasWaldmann> reimar has just committed something similar to 1.9 (and iirc i merged it into 2.0)
2009-07-25T01:24:45 <dennda> oh
2009-07-25T01:25:24 <ThomasWaldmann> of course it need major changes because it uses attachfile and packages
2009-07-25T01:25:37 <ThomasWaldmann> but you can keep the ui part maybe
2009-07-25T01:26:50 <dennda> looking for the changeset
2009-07-25T01:27:36 <dennda> anyway, gotta get some rest
2009-07-25T01:27:37 <dennda> n8
2009-07-25T01:30:33 <ThomasWaldmann> gn
2009-07-25T07:58:54 <CIA-43> Alexandre Martani <amartani AT gmail DOT com> default * 6170:22b0e0b55704 2.0-storage-editor-amartani/MoinMoin/items/_tests/test_ItemLock.py: Real-time editor: Add tests to ItemLock.
2009-07-25T07:58:56 <CIA-43> Alexandre Martani <amartani AT gmail DOT com> default * 6171:50f5f22502a6 2.0-storage-editor-amartani/MoinMoin/items/ (_tests/test_ItemLock.py lock.py): Real-time editor: rewrite ItemLock class to support multiple editors locking and suggest enable sync.
2009-07-25T07:58:58 <CIA-43> Alexandre Martani <amartani AT gmail DOT com> default * 6172:245bdb5a9bfb 2.0-storage-editor-amartani/MoinMoin/ (items/_tests/test_ItemLock.py items/lock.py wikiutil.py): Real-time editor: Use more constants and pep8 fix.
2009-07-25T07:59:00 <CIA-43> Alexandre Martani <amartani AT gmail DOT com> default * 6173:7461b4e731cb 2.0-storage-editor-amartani/MoinMoin/web/static/htdocs/common/js/text_editor.js: Real-time editor: todo about memory leak in xmlhttp wrapper
2009-07-25T09:22:06 <ThomasWaldmann> moin
2009-07-25T09:49:51 <dreimark> moin
2009-07-25T09:50:18 <dreimark> ThomasWaldmann: I don't see this ToDo in my version of cleancache
2009-07-25T09:53:02 <dreimark> found that comment in packages
2009-07-25T09:54:59 <dreimark> it is needed because underlay pages aren't saved by page editor
2009-07-25T10:03:11 *** devilsadvocate_ has joined #moin-dev
2009-07-25T10:04:41 *** devilsadvocate has quit IRC
2009-07-25T10:11:53 <dreimark> dimazest: ^^^^
2009-07-25T10:12:33 <dreimark> bbl
2009-07-25T10:14:06 *** devilsadvocate_ has quit IRC
2009-07-25T10:15:00 *** devilsadvocate has joined #moin-dev
2009-07-25T10:27:22 <ThomasWaldmann> dreimark: recently code below the comment was removed, because it called other code that was a NOP
2009-07-25T10:28:12 <ThomasWaldmann> and iirc, when you install a underlay page package, you have to restart moin (1.8) to make it see the pages
2009-07-25T10:45:24 *** Twaldmann has joined #moin-dev
2009-07-25T10:46:07 <ThomasWaldmann> test
2009-07-25T10:46:35 <ThomasWaldmann> üöä߀
2009-07-25T10:48:08 <Twaldmann> Works
2009-07-25T10:48:29 *** Twaldmann has quit IRC
2009-07-25T11:16:40 <dreimark> ThomasWaldmann: package does overwrite underlay pages so you can get stale cache files
2009-07-25T11:17:38 <ThomasWaldmann> i rather suspect the in-memory page cache
2009-07-25T11:21:55 <dreimark> hmm, the one who is allowed to install underlay pages must not be the same person who can restart the server process
2009-07-25T11:24:47 <dreimark> that would also fit more into a general adminstratioon tool
2009-07-25T11:25:18 * dreimark goes shopping
2009-07-25T11:39:36 <dimazest> ThomasWaldmann: dreimark i left that comment, because it is not clear for me should cache be cleaned or not
2009-07-25T11:40:15 <dimazest> suppose, that new underlay has page which is older than one in cache
2009-07-25T11:50:42 <ThomasWaldmann> how is "older" defined?
2009-07-25T11:54:06 <ThomasWaldmann> dimazest: what are you currently doing wrt xapian?
2009-07-25T12:38:30 <CIA-43> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4349:3c7e2e535d67 1.8/wiki/htdocs/rightsidebar/css/common.css: fix typo in rightsidebar css
2009-07-25T12:56:55 <dimazest> ThomasWaldmann: older: page form new underlay has modification time which is before cache creation time (assume this check http://localhost:8000/file/19e9f1edb2f1/MoinMoin/datastruct/backends/wiki_groups.py#l43 )
2009-07-25T12:57:38 <dimazest> ThomasWaldmann: currently i trying to query moin index using xappy
2009-07-25T14:57:52 <dreimark> dimazest: with which version do you play?
2009-07-25T15:01:39 <dimazest> xapian 1.0.14
2009-07-25T15:01:45 <dreimark> and xappy?
2009-07-25T15:01:52 <dimazest> 0.5
2009-07-25T15:02:08 <dreimark> may be it is woth to look at the svn version
2009-07-25T15:02:17 <dreimark> because of http://xapian.wordpress.com/2009/03/11/xappy-now-supports-image-similarity-searching/
2009-07-25T15:04:22 <dreimark> that can become a feature in the future
2009-07-25T15:05:32 <dennda> funny
2009-07-25T15:05:34 <dennda> "how does it work"
2009-07-25T15:05:56 <dennda> learning something like that (and friends) just right now ;)
2009-07-25T15:06:45 <ThomasWaldmann> btw, we need xapian/xappy working with some "usual" xapian lib version (e.g. the one in ubuntu/debian), so I would favour NOT using the latest and greatest
2009-07-25T15:07:26 <dimazest> :) which is the latest xapian version in ubuntu?
2009-07-25T15:08:33 <dennda> 1.0.7
2009-07-25T15:09:26 <dimazest> i can build 1.0.7 and work with it
2009-07-25T15:10:31 <dennda> that's jaunty
2009-07-25T15:10:37 <dreimark> there seems to be a problem with unicode
2009-07-25T15:10:39 <dennda> check which version is used in hardy and use that
2009-07-25T15:10:42 <dennda> packages.ubuntu.com
2009-07-25T15:10:44 <dreimark> http://bazaar.launchpad.net/~miracle2k/django-xappy/trunk/revision/38
2009-07-25T15:10:53 <dreimark> django has workaround this
2009-07-25T18:14:49 *** grzywacz has joined #moin-dev
2009-07-25T20:07:24 *** grzywacz has quit IRC
2009-07-25T20:08:08 *** grzywacz has joined #moin-dev
2009-07-25T20:19:42 <dreimark> dimazest: please add some content to Xapian2009 about what you are checking
2009-07-25T20:19:59 <dreimark> also you could add some examples
2009-07-25T20:22:40 <dimazest> dreimark: today i tried query min index with xappy, but no luck
2009-07-25T20:23:01 <dimazest> may be it worth to start not with searching, but with building index
2009-07-25T20:23:31 <dimazest> and then refactor searching, the problem is that it is a lot of changes
2009-07-25T20:23:35 <dreimark> yes
2009-07-25T20:24:06 <dimazest> i mean it is one massive change
2009-07-25T20:24:52 <dreimark> in the meaning of much amount of work or in the meaning of one change set?
2009-07-25T20:25:08 <dimazest> and moin search code is tricky
2009-07-25T20:25:27 <dreimark> yes, ThomasWaldmann and FranzPletz has done much of it
2009-07-25T20:26:04 <dimazest> dreimark: in amoiunt of work and it is atomic, i think it is not possible to change indexing without breaking searching
2009-07-25T20:27:15 <dreimark> breaking in which way - you mean while you refactor it can't be used?
2009-07-25T20:27:29 <dreimark> or does it behave afterwards different?
2009-07-25T20:27:31 <dimazest> btw, xapian 1.0.7 does not support facets
2009-07-25T20:27:49 <dimazest> dreimark: it cannot be used
2009-07-25T20:30:39 <dimazest> may be search.Xapian should be reviwed
2009-07-25T20:31:52 <dimazest> and i need to understand how Tokenizer and indexing works and how to use them with xappy
2009-07-25T20:33:06 <dreimark> without the hacks th tokenizer is only about words.
2009-07-25T20:33:35 <dreimark> while moin has camelcase words ThomasWaldmann has introduced to split these words
2009-07-25T20:33:40 <dreimark> also on numerical parts
2009-07-25T20:33:52 <dreimark> so that we get more words in the index
2009-07-25T20:34:46 <dreimark> also xapian can use word stemming
2009-07-25T20:34:53 <dimazest> yes, this i understand, but i have not found a way how to apply it to xappy
2009-07-25T20:36:30 <dimazest> i suppose, xappy support it internally
2009-07-25T20:37:51 <dreimark> can xappy read from any xapian database?
2009-07-25T20:38:37 <dimazest> i do not know, it is very strange, why i can not get content of documents
2009-07-25T20:38:50 <dimazest> and make queries
2009-07-25T20:39:43 <dimazest> e.g. conn.query_field('content', 'moin')
2009-07-25T20:39:55 <dreimark> do you have asked on the ml of xappy?
2009-07-25T20:39:56 <dimazest> returns 'Xapian::Query()'
2009-07-25T20:40:09 <dimazest> no, i should do it
2009-07-25T20:41:23 * dreimark lloks also if I can figure it out
2009-07-25T20:41:44 <dimazest> probably i query wrong field
2009-07-25T20:45:52 <dreimark> http://paste.pocoo.org/show/130745/
2009-07-25T20:46:03 <dreimark> python search.py SystemInfo
2009-07-25T20:46:16 <dreimark> founds it 22 times
2009-07-25T20:48:13 <dimazest> dreimark: http://paste.pocoo.org/show/130748/
2009-07-25T20:48:30 <dimazest> 73 results, but no document IDs
2009-07-25T20:48:50 <dreimark> for me too.
2009-07-25T20:49:08 <dreimark> besides that I have 18 results only
2009-07-25T20:49:41 <dreimark> that just means that the way i tried to access the data is wrong whild it can find the word
2009-07-25T20:56:23 <dimazest> dreimark: i'll write to the xappy ml
2009-07-25T20:56:35 <dreimark> good
2009-07-25T21:04:25 <dimazest> the ml is not very active
2009-07-25T21:08:00 <dreimark> I hope someone has subscribed and recogn. after ep that people are interested in using it
2009-07-25T21:08:53 <dimazest> we will see, and there is #xapian :)
2009-07-25T21:12:32 * dreimark reads doctests
2009-07-25T21:21:39 <dimazest> bbl
2009-07-25T23:18:41 <ThomasWaldmann> re MoinMoin: MoinMoinChat/Logs/moin-dev/2009-07-25 (last edited 2009-07-24 22:15:02 by IrcLogImporter)
