1 2010-03-06T05:58:14 *** mrht
2 2010-03-06T07:12:14 *** mrht
3 2010-03-06T11:57:39 <ThomasWaldmann> moin
4 2010-03-06T17:03:38 <CIA-47> Reimar Bauer <rb.proj AT googlemail DOT com> default * 5601:0c94d0486a5c 1.9/MoinMoin/search/_tests/test_search.py: test_search: ignore domain underlay and system for various tests
5 2010-03-06T17:03:45 <CIA-47> Reimar Bauer <rb.proj AT googlemail DOT com> default * 5602:8d3799e97a01 1.9/MoinMoin/search/_tests/test_search.py: test_search: for failing tests show the wrong value and not the expression
6 2010-03-06T17:03:48 <CIA-47> Reimar Bauer <rb.proj AT googlemail DOT com> default * 5603:5e01b23e0a37 1.9/MoinMoin/search/_tests/test_search.py: test_search: replaced testpage name HelpOnEditing by TestOnEditing, because we need a page with a similiar name but without attachments.
7 2010-03-06T17:05:06 <dreimark> so now i have the situation that on my system the xapian index is not updated
8 2010-03-06T17:05:23 <dreimark> and i have something in the queue
9 2010-03-06T17:06:10 <dreimark> by the follwoing patch http://paste.pocoo.org/show/186340/
10 2010-03-06T17:06:37 <dreimark> i get by 400 trials the queue cleared and as result onle 3-4 failing tests
11 2010-03-06T17:54:15 <CIA-47> Reimar Bauer <rb.proj AT googlemail DOT com> default * 5604:bd4419916312 1.9/MoinMoin/search/_tests/test_search.py: test_search: doc string "exclude underlay/system pages" added
12 2010-03-06T17:54:58 <ThomasWaldmann> http://socghop.appspot.com/
13 2010-03-06T18:20:08 *** moin3
14 2010-03-06T18:21:02 *** moin3
15 2010-03-06T18:21:27 <moin5> hi guys, again.
16 2010-03-06T18:21:49 <ThomasWaldmann> hi
17 2010-03-06T18:22:09 <moin5> ThomasWaldmann, i came back as soon as possible
18 2010-03-06T18:22:51 <moin5> i'm working for you :-) mh, in the truth for me...
19 2010-03-06T18:23:23 * ThomasWaldmann won't pay :D
20 2010-03-06T18:27:04 * dreimark looks at indexing.do_queued_updates
21 2010-03-06T18:29:27 <moin5> i'm so sorry for my bad proof of concept i sent you before
22 2010-03-06T18:30:10 <ThomasWaldmann> no problem :)
23 2010-03-06T18:30:30 <moin5> i had not realize it was a "official demo"
24 2010-03-06T18:30:42 <moin5> without running demo
25 2010-03-06T18:30:44 <moin5> :-)
26 2010-03-06T18:31:43 <ThomasWaldmann> btw, if someone is interesting in mentoring for Google Summer of Code 2010 and not on page MentorGroup yet, please let us fix that soon
27 2010-03-06T18:31:49 <dreimark> fsck mtime
28 2010-03-06T18:32:06 <ThomasWaldmann> s/ing/ed/
29 2010-03-06T18:33:20 <ThomasWaldmann> http://moinmo.in/GoogleSoc2010/InitialProjectIdeas < need more work, please help
30 2010-03-06T19:04:14 <dreimark> ha - if I slow down the queueing system I get 0 failures
31 2010-03-06T19:04:48 <dreimark> http://paste.pocoo.org/show/186383/
32 2010-03-06T19:06:07 <dreimark> ThomasWaldmann: ^
33 2010-03-06T19:06:17 <dreimark> that sound somehow familiar
34 2010-03-06T19:13:07 <dreimark> looks like the ssd i run the tests by py.test 1.2.1 is to fast
35 2010-03-06T19:15:45 <ThomasWaldmann> need more info, that is just a simple loop, there shouldn't be access conflicts or mtime problems or whatever
36 2010-03-06T19:17:08 *** tpfennig
37 2010-03-06T19:17:18 <dreimark> whithout that second sleep i get after each pagename, attachmentname, revno = self.update_queue.get()
38 2010-03-06T19:17:27 <dreimark> an IndexError
39 2010-03-06T19:20:32 <ThomasWaldmann> that happens if queue is empty
40 2010-03-06T19:23:14 <dreimark> but it isn't
41 2010-03-06T19:24:26 <dreimark> i try to trace it
42 2010-03-06T19:24:45 <ThomasWaldmann> add debug logging or use debugger
43 2010-03-06T19:46:33 *** tpfennig
44 2010-03-06T19:52:49 <moin5> guys, i remember wrong or in moin1.7 were there something like debugerr / error tracer?
45 2010-03-06T19:53:30 <dreimark> see http://hg.moinmo.in/moin/1.9/file/bd4419916312/wiki/config/logging
46 2010-03-06T19:53:56 <dreimark> should be similiar in 1.7
47 2010-03-06T19:55:07 <moin5> yes but i mean something grafical. it was a html page with annotations about error, call trace ... ecc
48 2010-03-06T19:55:37 <dreimark> in 1.9 it is the werkzeug debugger
49 2010-03-06T19:56:19 <dreimark> http://hg.moinmo.in/moin/1.9/file/bd4419916312/wiki/server/wikiserverconfig.py#l34
50 2010-03-06T19:56:25 <moin5> um ok.
51 2010-03-06T19:57:13 <moin5> cool..thank you dreimark
52 2010-03-06T20:03:07 <moin5> ThomasWaldmann, if you want to take a look to the patch i tell you this afternoon http://paste.pocoo.org/show/186403/
53 2010-03-06T20:03:25 <moin5> told*
54 2010-03-06T20:07:19 <dreimark> multiple """ """ you can use a block
55 2010-03-06T20:09:01 <ThomasWaldmann> moin5: why is sending of the page dependant on result of send_event?
56 2010-03-06T20:09:03 <moin5> sorry dreimark, are you talking to me?
57 2010-03-06T20:09:16 <dreimark> yes
58 2010-03-06T20:10:00 <moin5> i'm proposing an hook, an event, to change the standard behaviour of moinmoin wiki after save action
59 2010-03-06T20:10:16 <moin5> so
60 2010-03-06T20:12:10 <ThomasWaldmann> change how?
61 2010-03-06T20:12:35 <moin5> with this event, you can decide to perform, for example, controls and validation, on the user had edited and then, at save time, you can re-propose the editor in 'preview mode' to correct the 'error annotaion' the validator returns
62 2010-03-06T20:12:37 <ThomasWaldmann> (the save has already happened, so what's the point?)
63 2010-03-06T20:13:02 <moin5> i can avoid the save with your event PagePreSave
64 2010-03-06T20:13:15 <ThomasWaldmann> ah
65 2010-03-06T20:14:20 <ThomasWaldmann> but then, it is no ConditionalSaveEvent, but rather a PagePostSave event?
66 2010-03-06T20:14:23 <moin5> i perform my checks / validation on the page, the i send an Abort from my handler of PagePreSave if page is 'not valid'
67 2010-03-06T20:14:40 <moin5> yes
68 2010-03-06T20:14:57 <moin5> for consistency is better your name
69 2010-03-06T20:15:41 <moin5> PagePostSave works, for me it's ok!
70 2010-03-06T20:15:59 <moin5> but for you ThomasWaldmann? guys?
71 2010-03-06T20:16:10 <moin5> is it useful for you?
72 2010-03-06T20:17:12 <ThomasWaldmann> i am not sure about the "not result" stuff
73 2010-03-06T20:17:21 <ThomasWaldmann> did you look at what result is?
74 2010-03-06T20:19:55 <moin5> 'if not result' handles the lack of PagePostSave handler...or if you perform some action in PostSave and you have not the need to change workflow you can return a no result ( in python is [] )
75 2010-03-06T20:20:39 <moin5> we should define an 'OK' EventResult!?!
76 2010-03-06T20:25:19 <moin5> sorry for my lack of experience programming python...and sorry for my english!
77 2010-03-06T20:30:51 <ThomasWaldmann> yeah, maybe we need some class derived from EventResult
78 2010-03-06T20:32:12 <moin5> ok
79 2010-03-06T20:33:27 <ThomasWaldmann> so, your event handler sends the editor again?
80 2010-03-06T20:34:05 <moin5> yes, in my specific case. yes.
81 2010-03-06T20:34:47 <moin5> is there something wrong?
82 2010-03-06T20:35:40 <ThomasWaldmann> maybe you could send Abort EventResult subclass
83 2010-03-06T20:36:06 <moin5> yeah
84 2010-03-06T20:37:05 <moin5> not result is returned only if there are no handler returns something better, you know.
85 2010-03-06T20:37:40 <moin5> otherwise i need to patch something else, right?
86 2010-03-06T20:39:28 <ThomasWaldmann> no handler -> []
87 2010-03-06T20:39:50 <ThomasWaldmann> many handlers -> [?]
88 2010-03-06T20:42:23 <moin5> yes
89 2010-03-06T20:42:36 <ThomasWaldmann> well, you need to find a good definition how it shall behave, what behaviour makes sense
90 2010-03-06T20:42:48 <ThomasWaldmann> esp. for multiple handlers (one is easy)
91 2010-03-06T20:43:55 <moin5> yes is difficult
92 2010-03-06T20:43:57 <moin5> ok
93 2010-03-06T20:46:11 <moin5> ThomasWaldmann, guys, thank you for all the support you give me! Can i come back in this IRC channel to talk and discuss about this stuff (PagePostSaveEvent) in these days?
94 2010-03-06T20:46:51 <moin5> naturally, if it's not a waste of time for you
95 2010-03-06T20:47:37 <dreimark> sure
96 2010-03-06T20:47:52 <moin5> i want to take precious time to nobody ...
97 2010-03-06T20:47:59 <moin5> uh ok, thank you dreimark
98 2010-03-06T20:49:38 <ThomasWaldmann> sure, that's interesting. i guess other people can use this, too.
99 2010-03-06T21:28:47 *** moin5
100 2010-03-06T22:31:03 <dreimark> looks like the queueing becomes confused if you delete a page and recreate it
101 2010-03-06T22:36:41 <ThomasWaldmann> it shouldn't
102 2010-03-06T22:50:00 <dreimark> http://paste.pocoo.org/show/186451/
103 2010-03-06T22:50:35 <dreimark> without a waitstate after nuking pages new pages don't go in the index
104 2010-03-06T22:52:25 <ThomasWaldmann> we don't have an indexer thread any more
105 2010-03-06T22:52:42 <ThomasWaldmann> so there should be no need for such stuff as it is all synchronous
106 2010-03-06T22:54:33 <CIA-47> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4495:6f975bdd5adf 1.8/MoinMoin/action/Load.py: fix orthography
107 2010-03-06T23:04:20 <ThomasWaldmann> http://socghop.appspot.com/document/show/program/google/gsoc2009/orgcriteria
