1 2012-06-26T00:02:24 *** RogerHaase
2 2012-06-26T00:10:36 *** kanha_
3 2012-06-26T00:10:53 *** kanha
4 2012-06-26T00:19:19 <dreimark> bretonium: how is it going?
5 2012-06-26T00:32:40 *** dwcramer
6 2012-06-26T00:43:25 *** spy
7 2012-06-26T00:59:28 * dreimark quite silent today
8 2012-06-26T01:14:09 <dreimark> gn
9 2012-06-26T01:17:10 *** yufra
10 2012-06-26T01:17:30 *** MaikuMori
11 2012-06-26T01:17:32 *** yufra
12 2012-06-26T01:18:44 *** MaikuMori
13 2012-06-26T01:29:35 *** MaikuMori_
14 2012-06-26T01:29:55 *** MaikuMori
15 2012-06-26T01:32:00 *** vpv
16 2012-06-26T01:32:08 *** vpv
17 2012-06-26T01:42:51 *** esyr
18 2012-06-26T01:43:01 *** esyr
19 2012-06-26T02:28:41 *** dwcramer
20 2012-06-26T03:32:46 *** sudo_dirk
21 2012-06-26T03:34:31 *** sudo_dirk1
22 2012-06-26T04:23:38 *** dwcramer
23 2012-06-26T05:08:07 *** dwcramer
24 2012-06-26T05:47:37 *** yufra
25 2012-06-26T05:47:45 *** yufra
26 2012-06-26T06:58:31 *** dwcramer
27 2012-06-26T07:24:32 <dreimark> moin
28 2012-06-26T07:31:46 <dreimark> seems often get_bool becomes implemented
29 2012-06-26T07:32:06 <dreimark> we can choose between ours and the one in pygments
30 2012-06-26T11:01:34 <ThomasWaldmann> MaikuMori_: are you working on your project? you should try to catch up the time you have lost due to your exams/coursework.
31 2012-06-26T11:19:30 *** greg_f
32 2012-06-26T11:20:04 <xiaq> http://hg.moinmo.in/moin/2.0/file/aa79e8413d9c/MoinMoin/items/__init__.py#l451
33 2012-06-26T11:20:12 <xiaq> # we use querystring param to FORCE content type
34 2012-06-26T11:20:18 <xiaq> is this used anywhere?
35 2012-06-26T11:21:03 <xiaq> the only use i can think of the creating a new *Draw item, since POSTing for them doesn't work yet
36 2012-06-26T11:21:21 <xiaq> otherwise the contenttype always get included in the meta text
37 2012-06-26T11:23:01 <ThomasWaldmann> yes, maybe that was one reason
38 2012-06-26T11:23:16 <xiaq> ThomasWaldmann: and the sole use case is actually a hack :)
39 2012-06-26T11:23:41 <ThomasWaldmann> there is another problematic "no meta" case (but that is not solved by this): multi-file drag-and-drop upload
40 2012-06-26T11:24:09 <xiaq> ThomasWaldmann: uh, drag-and-drop upload? what is it? is it documented anywhere?
41 2012-06-26T11:24:34 <ThomasWaldmann> go to index page and drop files onto it
42 2012-06-26T11:24:51 <xiaq> ThomasWaldmann: hm, the *Draw POSTing problem can be fixed by having the draw applets post to somewhere else, or include special arguments in the query string
43 2012-06-26T11:25:48 <xiaq> ThomasWaldmann: on my Linux box Firefox tries to open the file...
44 2012-06-26T11:27:16 <xiaq> Chromium does this also...
45 2012-06-26T11:28:59 <xiaq> oh, index page...
46 2012-06-26T11:29:19 <xiaq> i took it for the Home page... :)
47 2012-06-26T11:30:37 <ThomasWaldmann> +index/Foo
48 2012-06-26T11:31:26 <ThomasWaldmann> why do we need to "post somewhere else"? "include special arguments in the query string
49 2012-06-26T11:31:39 <ThomasWaldmann> " - well, isn't that exactly what we do? :)
50 2012-06-26T11:33:08 <xiaq> ThomasWaldmann: yeah, special arguments in qs. but not only contenttype, it needs to be said explicitly, "this POST comes from drawing applet"
51 2012-06-26T12:33:18 <dreimark> that is difficult, because you always have to patch 3rd aprty apps
52 2012-06-26T13:23:58 *** spy
53 2012-06-26T14:30:34 *** dave_largo
54 2012-06-26T14:51:23 *** kanha
55 2012-06-26T14:51:46 *** kanha
56 2012-06-26T15:24:10 *** dwcramer
57 2012-06-26T15:37:02 <xiaq> dreimark: not really, since the POST url is always configurable. actually current solution of embedding contenttype=foo is based on this fact. i'm just proposing a more explicit query string parameter
58 2012-06-26T16:57:43 <dreimark> ThomasWaldmann: i know, builtin stupidity, i file an issue or remove it
59 2012-06-26T16:58:41 <dreimark> have to look for updates also, may be someone has worked on that too
60 2012-06-26T16:59:18 <dreimark> xiaq: agreed, meta is currently not very good solved for external plugins
61 2012-06-26T17:14:39 * bretonium will post today something for review
62 2012-06-26T17:20:47 *** RogerHaase
63 2012-06-26T17:30:33 <dreimark> bbl
64 2012-06-26T18:28:45 <xiaq> ThomasWaldmann: http://codereview.appspot.com/6330064/ plz review
65 2012-06-26T18:28:47 <xiaq> dreimark: ^^
66 2012-06-26T18:28:55 <xiaq> and... gn :)
67 2012-06-26T20:17:57 *** greg_f
68 2012-06-26T21:07:23 <spy> ThomasWaldmann: could you once again paste your whoosh test example that you have already posted to dpaste.com ?
69 2012-06-26T21:12:42 <ThomasWaldmann> yes, wait a sec...
70 2012-06-26T21:13:46 <bretonium> err, what are "dynamic fields" in indexing middleware?
71 2012-06-26T21:14:06 <bretonium> I mean, what are they for and where are they used?
72 2012-06-26T21:15:00 <ThomasWaldmann> spy: http://paste.ubuntu.com/1061336/
73 2012-06-26T21:15:24 <ThomasWaldmann> bretonium: they are dynamic, so you can use them for anything you like
74 2012-06-26T21:16:06 <bretonium> how for example?
75 2012-06-26T21:16:18 <ThomasWaldmann> birthday_date
76 2012-06-26T21:16:30 <ThomasWaldmann> so you can index birthdays from metadata
77 2012-06-26T21:16:37 <spy> ThomasWaldmann: thanks
78 2012-06-26T21:17:13 <bretonium> so, if they don't match the pattern, they will not be searchable?
79 2012-06-26T21:18:47 <bretonium> they == custom meta entries
80 2012-06-26T21:29:22 <ThomasWaldmann> yes, dynamic stuff triggers only if it matches one of those glob patterns
81 2012-06-26T22:00:28 <bretonium> ThomasWaldmann, moin2-bms 172+ please. Cannot give the link, internets in the hotel suck, have to use the phone :(
82 2012-06-26T22:24:54 <ThomasWaldmann> well, that a question of what semantics we want
83 2012-06-26T22:25:39 <ThomasWaldmann> 1.9 semantics is to use latest acl for all revisions
84 2012-06-26T22:26:30 <ThomasWaldmann> so, it kind of uses branch head acl for all revisions in that branch
85 2012-06-26T22:27:03 <ThomasWaldmann> which can get kind of interesting for a fork point and its parents
86 2012-06-26T22:27:53 <ThomasWaldmann> bretonium: ^
87 2012-06-26T22:29:27 <bretonium> if we store in branches, there is a situation, where a userhead points to a non-branchhead revid. If we store in userhead, a branch without corresponding userheads can still exist
88 2012-06-26T22:29:46 <ThomasWaldmann> maybe it is easier for now if you just use the acl from the revision in question
89 2012-06-26T22:30:00 <ThomasWaldmann> that is different behaviour than in 1.9, but easier for now
90 2012-06-26T22:30:17 <ThomasWaldmann> you have to document that it is different :)
91 2012-06-26T22:30:27 <bretonium> so, we store it in "all" index?
92 2012-06-26T22:31:23 *** RogerHaase
93 2012-06-26T22:32:10 <ThomasWaldmann> yes
94 2012-06-26T22:37:19 <ThomasWaldmann> also interesting are hierarchical acls of non-current revisions...
95 2012-06-26T22:37:44 <ThomasWaldmann> for them, that simple solution does not work
96 2012-06-26T22:39:45 <ThomasWaldmann> bretonium: ^
97 2012-06-26T22:40:33 <bretonium> what's hierarchical acls?
98 2012-06-26T22:40:49 <bretonium> I mean, how do they work now?
99 2012-06-26T22:41:07 <dreimark> bretonium: they are broken in m2
100 2012-06-26T22:41:27 <ThomasWaldmann> dreimark: are they?
101 2012-06-26T22:41:38 <dreimark> in m1 you inherit acls from the major page
102 2012-06-26T22:41:51 <dreimark> ThomasWaldmann: we have a ticket for that
103 2012-06-26T22:42:53 <dreimark> https://bitbucket.org/thomaswaldmann/moin-2.0/issue/107/hierachical-acls
104 2012-06-26T22:44:11 <ThomasWaldmann> that's for namespaces branch, which is not merged yet
105 2012-06-26T22:44:40 <dreimark> bretonium: http://moinmo.in/HelpOnAccessControlLists#Hierarchical_ACL_processing
106 2012-06-26T22:44:51 <dreimark> true, my fault
107 2012-06-26T22:47:06 <ThomasWaldmann> hmm, guess there is no easy solution
108 2012-06-26T22:55:13 <ThomasWaldmann> bretonium: forget what i said, i think doing it on a per-revision basis would make more troubles than needed
109 2012-06-26T22:56:00 <ThomasWaldmann> i think doing it 1.9-like is easier, so the branch heads have the latest acl
110 2012-06-26T22:56:58 <ThomasWaldmann> (and we'll need a function that termines the branch heads for any revision)
111 2012-06-26T22:57:50 <ThomasWaldmann> determines*
112 2012-06-26T23:16:41 *** dave_largo
113 2012-06-26T23:35:52 <dreimark> gn