1 2011-09-25T00:15:11 *** grad
2 2011-09-25T00:44:26 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 859:95d1033b7808 2.0/MoinMoin/ (5 files in 5 dirs): fix show/meta view
3 2011-09-25T01:05:52 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 860:de6ece090b7d 2.0/MoinMoin/items/__init__.py: fix download view
4 2011-09-25T01:53:16 *** MattMaker
5 2011-09-25T02:11:13 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 861:9a7889c04a25 2.0/MoinMoin/ (6 files in 4 dirs): fixed history view
6 2011-09-25T03:29:51 *** MattMaker
7 2011-09-25T09:00:28 <dreimark> moin
8 2011-09-25T09:00:49 <dreimark> i try to get familiar with the new api, some questions
9 2011-09-25T09:00:58 <dreimark> http://paste.pocoo.org/show/482040/
10 2011-09-25T09:01:14 <dreimark> replacement for format_date?
11 2011-09-25T09:02:01 <dreimark> and what is the previous revno, the global history code was also not in all cases correct by just doing -1
12 2011-09-25T09:02:58 <dreimark> i think we should not make linear rev numbers again.
13 2011-09-25T09:03:12 <dreimark> (for showing in history)
14 2011-09-25T09:03:42 *** MattMaker
15 2011-09-25T09:04:22 <dreimark> bbl
16 2011-09-25T09:12:47 *** marchael
17 2011-09-25T09:13:21 *** marchael
18 2011-09-25T09:14:31 *** marchael
19 2011-09-25T10:33:59 *** raignarok
20 2011-09-25T10:43:25 *** raignarok
21 2011-09-25T10:45:22 *** raignarok
22 2011-09-25T11:19:46 *** raignarok
23 2011-09-25T11:42:20 <ThomasWaldmann> dreimark: why do you want to replace format_date?
24 2011-09-25T11:43:02 <ThomasWaldmann> revno is gone, it is revid now and it does not sort
25 2011-09-25T11:43:19 <ThomasWaldmann> (or order)
26 2011-09-25T11:44:23 <ThomasWaldmann> what we'll need in the end is a parent revid in the revision, but for a quick hack, we can just order by MTIME now
27 2011-09-25T12:38:53 <dreimark> format_date crashes with that time info
28 2011-09-25T12:45:49 *** raignarok
29 2011-09-25T12:51:07 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 862:e1a68b3721cc 2.0/MoinMoin/ (app.py search/analyzers.py storage/middleware/indexing.py): use the right tokenizers/analyzers in the schema, use name_exact for 1:1 matching
30 2011-09-25T12:51:37 <ThomasWaldmann> dreimark: then it needs a fix, but not necessarily replacement
31 2011-09-25T12:59:46 <waldi> ThomasWaldmann: who did you say is the flatland developer, jek?
32 2011-09-25T13:06:55 <ThomasWaldmann> just join #discorporate
33 2011-09-25T13:07:11 <ThomasWaldmann> mainly jek, but also others there
34 2011-09-25T13:16:25 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 863:d421f075eb30 2.0/MoinMoin/templates/search.html: fix search to use revid
35 2011-09-25T13:26:23 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 864:653f12dbaf9f 2.0/MoinMoin/storage/middleware/routing.py: use colon to separate mountpoint from revid, not slash which makes it look like a path in the url
36 2011-09-25T13:42:37 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 865:8e63434d6c1a 2.0/MoinMoin/apps/frontend/views.py: show_item: switch off broken revision navigation
37 2011-09-25T13:42:38 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 866:03efad0ebc90 2.0/MoinMoin/templates/history.html: history: use revid
38 2011-09-25T14:12:25 <ThomasWaldmann> hmm, revid troubles
39 2011-09-25T14:12:48 *** raignarok
40 2011-09-25T14:13:09 <ThomasWaldmann> router wants mountpoint:uuid as revid, but sometimes just gets the uuid
41 2011-09-25T14:14:48 <ronny> ThomasWaldmann: i suppose thats cause the router only resuffles the name metadata, but leaves revid intact
42 2011-09-25T14:15:01 <ronny> (i think thats very important to kee for serialization)
43 2011-09-25T14:19:58 <ThomasWaldmann> btw, we have a problem with search also. the search view (MM.apps.frontend.view.search) uses whoosh directly, thus now acl protection by the acl middleware
44 2011-09-25T14:20:28 <ThomasWaldmann> but if one tries to make it use the filter search of the middleware, all sorts of stuff is missing because we only get revs
45 2011-09-25T14:21:01 <ThomasWaldmann> e.g. search stats, also I don't think it can deal with search results for other wikis than the local one then
46 2011-09-25T14:22:44 *** greg_f
47 2011-09-25T14:23:31 <ronny> ThomasWaldmann: we can ask the router for the backend name for a found item/revid pair
48 2011-09-25T14:24:22 <ThomasWaldmann> yeah, something like that
49 2011-09-25T14:25:04 <ronny> ThomasWaldmann: can we asume that whoosh returns name, backend revid for search and ask the root backend for its revid for the given name
50 2011-09-25T14:26:05 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 867:5db17e17624b 2.0/MoinMoin/storage/middleware/ (_tests/test_routing.py routing.py): router: fix some more places to use colon, not slash
51 2011-09-25T14:48:13 * ThomasWaldmann tries to make revid consistent mountpoint:be_revid above router, be_revid below router
52 2011-09-25T15:03:18 <ThomasWaldmann> ronny: do you know how to stop pytest from using "..." for values?
53 2011-09-25T15:11:22 <ThomasWaldmann> ronny: http://paste.pocoo.org/show/482139/
54 2011-09-25T15:11:51 <ThomasWaldmann> problem is that there are revids that do not have x:y form and I am not yet sure where they come from
55 2011-09-25T15:29:31 <ronny> ThomasWaldmann: paste me a examples please
56 2011-09-25T15:32:12 <ThomasWaldmann> http://paste.pocoo.org/show/482146/ first failure
57 2011-09-25T15:32:50 <ThomasWaldmann> line 256 you see revid has no x: prefix
58 2011-09-25T15:34:27 <ronny> ThomasWaldmann: so you want meta printed out in fulll?
59 2011-09-25T15:34:40 <ronny> ThomasWaldmann: just use py.test --pdb for debugging
60 2011-09-25T15:35:22 <ThomasWaldmann> i just wanted to see more, not use pdb
61 2011-09-25T15:35:50 <ThomasWaldmann> it often leaves out the interesting parts of data, so printing them that way is a bit pointless
62 2011-09-25T15:36:45 <ThomasWaldmann> btw, that is after applying the changes from the patch from previous pastebin
63 2011-09-25T15:37:20 <ThomasWaldmann> i am not sure how upper layers (above router) can even know revids without mountpoint: prefix
64 2011-09-25T15:53:44 <ronny> ThomasWaldmann: it will fuck with the serializing
65 2011-09-25T15:54:59 <ThomasWaldmann> hm?
66 2011-09-25T15:57:05 <ThomasWaldmann> you mean that mountpoint: prefix should not be in serialized output?
67 2011-09-25T16:09:27 <ronny> imagine desrializing to a different topology
68 2011-09-25T16:25:28 <ThomasWaldmann> yeah
69 2011-09-25T16:28:06 <ThomasWaldmann> but if we do not yield mountpoint:backend_revid from the router's __iter__ we can't find the right backend when using what it yielded
70 2011-09-25T16:34:48 <ThomasWaldmann> ronny: ^^
71 2011-09-25T16:35:41 <ronny> ThomasWaldmann: iter/getitem work with the m:b
72 2011-09-25T16:35:53 <ronny> ThomasWaldmann: however metadata items dont contain the m:b, but just the b
73 2011-09-25T16:44:45 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 868:8cf8ea7431ca 2.0/MoinMoin/search/_tests/test_analyzers.py: fix test_analyzers
74 2011-09-25T16:45:20 <ThomasWaldmann> ronny: ok, so we use m:b mostly for index building, but afterwards always name+revid?
75 2011-09-25T16:45:48 <ronny> ThomasWaldmann: yeah
76 2011-09-25T16:46:41 <ThomasWaldmann> waldi: E ParseError: unbound prefix: line 1, column 160
77 2011-09-25T16:46:55 <ThomasWaldmann> if you have a clue why that happens, please help fixing it
78 2011-09-25T16:47:29 <ThomasWaldmann> we are currently down to ~13 test fails, we should get to 0
79 2011-09-25T16:48:01 <ThomasWaldmann> ronny: you want to fix that while i work on other stuff?
80 2011-09-25T16:48:58 <ronny> ThomasWaldmann: lets see if i can find it quickyl
81 2011-09-25T16:54:13 <ThomasWaldmann> 12 fails currently
82 2011-09-25T16:58:14 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 869:1a7aa271851d 2.0/MoinMoin/storage/middleware/indexing.py: refactor indexing revision lazy retrieval
83 2011-09-25T17:25:04 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 870:04cdc3823cac 2.0/MoinMoin/ (3 files in 3 dirs): add index-dump script command
84 2011-09-25T17:25:34 <ThomasWaldmann> dreimark: reading?
85 2011-09-25T17:39:21 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 871:75de15de041b 2.0/MoinMoin/script/maint/serialization.py: add --file param to moin load/save script command
86 2011-09-25T17:41:57 *** raignarok
87 2011-09-25T17:54:19 <ronny> ThomasWaldmann: i have no idea where that parsere error comes from
88 2011-09-25T17:59:28 *** raignarok
89 2011-09-25T18:03:45 *** raignarok
90 2011-09-25T18:07:19 <ThomasWaldmann> ronny: parser error?
91 2011-09-25T18:10:09 <ThomasWaldmann> ronny: that was for waldi
92 2011-09-25T18:10:49 <ronny> oh, ok
93 2011-09-25T18:28:25 <ronny> ThomasWaldmann: ok, fixed the first 2 issues
94 2011-09-25T18:35:37 <ronny> rest needs some more digging
95 2011-09-25T18:56:42 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 872:135198715129 2.0/MoinMoin/script/ (__init__.py maint/modify_item.py): implement item-get / item-put script commands
96 2011-09-25T18:57:26 <ThomasWaldmann> ronny: that ^^ might be useful for the commandline hackers who want to edit wiki items
97 2011-09-25T18:58:41 <ThomasWaldmann> quite low-level currently, though
98 2011-09-25T19:02:49 <ronny> ThomasWaldmann: hm, item.rename is probably strange
99 2011-09-25T19:04:13 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 873:09534d362ed6 2.0/MoinMoin/script/ (__init__.py maint/create_item.py): remove create_item script, outdated/superceded
100 2011-09-25T19:16:48 <ThomasWaldmann> ronny: i did not fix rename and delete yet
101 2011-09-25T19:22:51 *** greg_f
102 2011-09-25T19:27:57 <ThomasWaldmann> https://bitbucket.org/RonnyPfannschmidt/moin-2.0/changeset/ba8c41979a44 if you change the api, please also fix the callers
103 2011-09-25T19:28:57 <ronny> ThomasWaldmann: the callers already used that other api
104 2011-09-25T19:36:13 <ThomasWaldmann> https://bitbucket.org/RonnyPfannschmidt/moin-2.0/src/ba8c41979a44/MoinMoin/script/maint/index.py#cl-104
105 2011-09-25T19:37:50 <ronny> ThomasWaldmann: then there is a test missing
106 2011-09-25T19:38:50 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 874:8fa57ebd113e 2.0/contrib/xml/preloaded_items.xml: remove xml dump with example items
107 2011-09-25T19:39:05 <ThomasWaldmann> :P
108 2011-09-25T19:41:59 <ronny> ThomasWaldmann: what do we do wrt the threadsavety of some stores?
109 2011-09-25T19:49:44 <ThomasWaldmann> what's the problem exactly?
110 2011-09-25T19:52:42 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 875:f15c538959aa 2.0/docs/devel/development.rst: storage docs updated
111 2011-09-25T19:56:15 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 876:1ceb6f3f2a2a 2.0/docs/examples/config/wikiconfig.py: maintaining one wikiconfig is enough for now
112 2011-09-25T19:57:54 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 877:f3b7fe1f3579 2.0/docs/todo/todo.rst: remove mercurial backend todo, does not exist any more
113 2011-09-25T20:05:31 <ThomasWaldmann> File "/home/tw/w/moin-2.0/MoinMoin/storage/middleware/serialization.py", line 61, in deserialize
114 2011-09-25T20:05:34 <ThomasWaldmann> assert limited.is_exhausted
115 2011-09-25T20:05:37 <ThomasWaldmann> AssertionError
116 2011-09-25T20:07:20 <ronny> ThomasWaldmann: that looks like something did fail to complete exthausing the stream from the deserializer
117 2011-09-25T20:07:47 <ronny> ThomasWaldmann: wrt threadsavety - httplib.HTTPConnecton is NOT
118 2011-09-25T20:14:37 <ThomasWaldmann> you mean the stdlib code is not threadsafe?
119 2011-09-25T20:31:55 <ronny> ThomasWaldmann: yes
120 2011-09-25T20:34:04 <ronny> ThomasWaldmann: httpconnection is not threadsafe
121 2011-09-25T20:35:21 *** raignarok_
122 2011-09-25T20:35:41 <ThomasWaldmann> hm, great. anything we could use instead?
123 2011-09-25T20:36:47 <ronny> ThomasWaldmann: i think it makes sense to investigate httplib2
124 2011-09-25T20:37:33 *** raignarok
125 2011-09-25T20:49:57 *** humbu
126 2011-09-25T20:52:02 <humbu> Hello, one question, i wanted to commit something in my moinmoinrepo with mercurial, but if i do this with "hg commit", a window with the default editor vim opens, but nothing is in there, really nothing - what could be wrong?
127 2011-09-25T20:52:50 <ronny> humbu: i should contain some information on the changes you did
128 2011-09-25T20:53:12 <ronny> humbu: whats the ourput of hg diff and how/where do you invoke commit?
129 2011-09-25T20:55:16 <humbu> i changed the variable EDITOR to gedit an i now ther is information about my changes, but not if i use vim !?
130 2011-09-25T20:55:37 <humbu> sure, hg diff gives me changes
131 2011-09-25T20:58:35 <humbu> Why does this not work with vim ?
132 2011-09-25T20:59:09 <ronny> humbu: it works fine here with vim
133 2011-09-25T20:59:16 <ronny> humbu: what distro are you on?
134 2011-09-25T21:00:09 <humbu> ubuntu 11.04
135 2011-09-25T21:03:04 <humbu> hm, how can i change the standard editor
136 2011-09-25T21:03:27 <ronny> with the editor var think
137 2011-09-25T21:03:38 <ronny> however there should be a list of changes in the editor
138 2011-09-25T21:03:41 <ronny> hmm
139 2011-09-25T21:05:29 <humbu> ok, the problem was that i defined EDITOR="/usr/bin/gnome-terminal -x vim" that i can edit textfiles with vim via double-klick --> deleted
140 2011-09-25T21:07:39 <ronny> humbu: there s something called gvim, use it ...
141 2011-09-25T21:07:54 <ronny> with an extra arg for blocking of course
142 2011-09-25T21:07:58 <ThomasWaldmann> I usually either use hg commit -m "..." for short commit comments, or
143 2011-09-25T21:08:06 <ronny> also bbl
144 2011-09-25T21:08:24 <ThomasWaldmann> hg commit -l filename # for longer comments i prepared in a file
145 2011-09-25T21:09:01 <ThomasWaldmann> and i often to hg diff > filename to start with such a file and then rewrite its contents while reading/deleting the diff inside it.
146 2011-09-25T21:17:45 <dreimark> re
147 2011-09-25T21:22:46 <dreimark> ThomasWaldmann: ping
148 2011-09-25T21:32:27 *** humbu
149 2011-09-25T21:32:53 <dreimark> minimum for format_date is format_date(datetime.utcfromtimestamp(tm))
150 2011-09-25T21:33:09 <dreimark> so i guess it was in the past a datetime object
151 2011-09-25T21:35:18 <ThomasWaldmann> dreimark: can you start with describing the problem?
152 2011-09-25T21:38:30 *** humbu
153 2011-09-25T21:39:35 <humbu> Hello, i have one last question - i made a config in .hg/hgrc an now he says i cannot trust this file ?
154 2011-09-25T21:46:07 *** humbu
155 2011-09-25T21:46:36 <dreimark> mom
156 2011-09-25T21:48:34 <dreimark> ThomasWaldmann: http://paste.pocoo.org/show/482333/
157 2011-09-25T21:52:44 <ThomasWaldmann> 15 data type of tm is ...?
158 2011-09-25T21:55:33 <ronny> we might have a missmatch between datetme and timestamps
159 2011-09-25T21:57:18 <dreimark> ThomasWaldmann: it is a int
160 2011-09-25T21:58:25 <dreimark> print tm
161 2011-09-25T21:58:25 <dreimark> 1316978801
162 2011-09-25T21:59:54 <ThomasWaldmann> looks like babel wants date or datetime
163 2011-09-25T22:37:01 *** raignarok_
164 2011-09-25T22:48:25 <CIA-65> Thomas Waldmann <tw AT waldmann-edv DOT de> storage-ng * 878:490222d8f98b 2.0/docs/admin/ (backup.rst configure.rst index.rst install.rst upgrade.rst): update some docs
165 2011-09-25T23:30:28 <dreimark> ThomasWaldmann: i have a problem with id in rev.editors
166 2011-09-25T23:30:32 <dreimark> http://paste.pocoo.org/show/482383/
167 2011-09-25T23:31:05 <dreimark> my json.loads does only accept the last input line 70
168 2011-09-25T23:37:44 <ronny> dreimark: thats python
169 2011-09-25T23:37:49 <ronny> dreimark: see the u
170 2011-09-25T23:37:52 <ronny> thats invalid in json
171 2011-09-25T23:38:46 <ThomasWaldmann> ronny: did you ever see the exhausted issue while testing deserialize?
172 2011-09-25T23:39:14 <ThomasWaldmann> i tried to debug it, but everything looks correct, except the asserted value
173 2011-09-25T23:39:33 <ronny> ThomasWaldmann: never before
174 2011-09-25T23:40:49 <ronny> dreimark: sae goes for the '', json only accepts "" for strings
175 2011-09-25T23:41:24 <ThomasWaldmann> why do we need that LimitedStream?
176 2011-09-25T23:41:31 <ThomasWaldmann> can't we just read the size we know?
177 2011-09-25T23:41:54 <ronny> gmm
178 2011-09-25T23:42:14 <ronny> ThomasWaldmann: then we cant serialize large streams
179 2011-09-25T23:42:27 <ronny> at least not sanely with small chunking
180 2011-09-25T23:42:57 <ThomasWaldmann> why?
181 2011-09-25T23:43:41 <ronny> ThomasWaldmann: limitedstream is basically a wrapper around a part of a file that you will iteratively read
182 2011-09-25T23:43:49 <ronny> read, will just give a big string
183 2011-09-25T23:44:38 <ThomasWaldmann> well, it looks broken
184 2011-09-25T23:44:50 <ThomasWaldmann> i created a rather trivial file
185 2011-09-25T23:45:05 <ThomasWaldmann> looks correct in hexdump, but that limitedstream fails
186 2011-09-25T23:45:56 <ronny> got a script for me?
187 2011-09-25T23:49:25 <ThomasWaldmann> wait, i think i found something
188 2011-09-25T23:49:51 <dreimark> if something is expected to become used with json, we should use it in the correct syntax
189 2011-09-25T23:51:33 <dreimark> something more to fix in global_history, if it is not somewhere else
190 2011-09-25T23:55:05 <dreimark> ah, my fault, hack in the morning