1 2010-02-17T00:21:50 *** eSyr
2 2010-02-17T00:26:27 *** eSyr
3 2010-02-17T00:27:25 *** grzywacz
4 2010-02-17T00:35:41 <ThomasWaldmann> dreimark: ok, I found the editor_default problem
5 2010-02-17T00:36:17 <ThomasWaldmann> can someone test 1.8 please?
6 2010-02-17T00:37:35 <eSyr> what the problem?
7 2010-02-17T00:38:56 <ThomasWaldmann> crash in userprefs
8 2010-02-17T00:40:09 <eSyr> hm.
9 2010-02-17T00:48:44 <eSyr> ThomasWaldmann, btw, http://paste.pocoo.org/show/179086/ http://esyr.org:8080/?action=CheckTranslation&pageset=all_pages&language=fi (while i'm updating 1.8 hg)
10 2010-02-17T01:06:16 <ThomasWaldmann> strange, i can't reproduce locally yet
11 2010-02-17T01:08:28 <eSyr> me too
12 2010-02-17T01:09:43 <ThomasWaldmann> i think it is clean_input getting str, it wants unicode only
13 2010-02-17T01:10:19 * ThomasWaldmann hopes that is not in 1.8.7
14 2010-02-17T01:11:10 <ThomasWaldmann> eSyr: your test wiki is a bit slow from here :) it still loads...
15 2010-02-17T01:12:10 <ThomasWaldmann> gna, 1.8.7 has that bug also, fsck
16 2010-02-17T01:13:01 <eSyr> hm, looks like shh has hang. -_-
17 2010-02-17T01:14:23 <ThomasWaldmann> ok, time for monkeypatching
18 2010-02-17T01:15:56 <eSyr> ThomasWaldmann, plesae, try again
19 2010-02-17T01:16:00 <eSyr> *please
20 2010-02-17T01:21:31 <ThomasWaldmann> ah, better
21 2010-02-17T01:22:58 <ThomasWaldmann> ah, so it searches the right pages for the translator for updating i18n mapping?
22 2010-02-17T01:23:17 <eSyr> yes.
23 2010-02-17T01:40:32 <ThomasWaldmann> eSyr: is it possible to use search to also search for the master-page stuff you are looking for?
24 2010-02-17T01:40:57 *** evgSyr
25 2010-02-17T01:41:23 <ThomasWaldmann> eSyr: is it possible to use search to also search for the master-page stuff you are looking for?
26 2010-02-17T01:43:19 <evgSyr> hm...
27 2010-02-17T01:43:33 <evgSyr> what do you mean?
28 2010-02-17T01:44:25 *** eSyr
29 2010-02-17T01:44:55 <evgSyr> like, single query for pages with specific language and regez1:##master-page:(orig_page1|orig_page2|...)?
30 2010-02-17T01:45:08 <evgSyr> or query per orig_page?
31 2010-02-17T01:48:46 <ThomasWaldmann> reducing amount of Page objects you have to search "manually"
32 2010-02-17T01:51:23 <evgSyr> well, anyway one page object is created for every orig_page and trans_page.
33 2010-02-17T01:51:34 <evgSyr> i can cache those objects if you want.
34 2010-02-17T02:01:51 <ThomasWaldmann> so about how much is it total?
35 2010-02-17T02:02:19 <ThomasWaldmann> i just want to avoid that stuff getting to heavyweight. you never know how many bots will find it.
36 2010-02-17T02:02:38 <evgSyr> surge protection?
37 2010-02-17T02:04:19 <evgSyr> hm, i can reduce number of Page objects to one per page.
38 2010-02-17T02:04:33 <ThomasWaldmann> i only had a quick look at the diff, how much is it? N^2?
39 2010-02-17T02:05:18 <evgSyr> now? no, N*2 or N*3.
40 2010-02-17T02:05:28 <evgSyr> N*3 more likely.
41 2010-02-17T02:06:26 <ThomasWaldmann> * or **?
42 2010-02-17T02:06:32 <evgSyr> *
43 2010-02-17T02:07:01 <evgSyr> (i don't like complexity above linear)
44 2010-02-17T02:09:24 <ThomasWaldmann> ok, maybe look if you can so some simple optimization without needing too much code and effort
45 2010-02-17T02:09:33 <evgSyr> ok, i'll try.
46 2010-02-17T02:09:57 <ThomasWaldmann> and push it afterwards, then we'll just give it a try
47 2010-02-17T02:10:05 <evgSyr> ok.
48 2010-02-17T02:10:32 <ThomasWaldmann> strange, userprefs doesn't crash on some 1.8.7 wikis
49 2010-02-17T02:11:20 <evgSyr> hm... may be distro-specific patches?
50 2010-02-17T02:11:32 <evgSyr> or it depends on using farmconfig or something?
51 2010-02-17T02:11:40 <ThomasWaldmann> both lenny
52 2010-02-17T02:12:15 <ThomasWaldmann> not quite clear yet
53 2010-02-17T02:22:30 <CIA-50> Eugene Syromyatnikov <evgsyr@gmail.com> default * 499:25702a913bdd extensions/data/plugin/action/CheckTranslation.py: CheckTranslation: added supplying list of pages with the same master-page special comment for every page in page set.
54 2010-02-17T02:30:01 <ThomasWaldmann> evgSyr: works :)
55 2010-02-17T02:30:19 <evgSyr> yes.
56 2010-02-17T02:32:27 * ThomasWaldmann watches the cpu spikes with top
57 2010-02-17T02:33:00 <ThomasWaldmann> s/spikes/peaks/
58 2010-02-17T02:33:06 <evgSyr> yes, this action is rather heavy.
59 2010-02-17T02:33:45 <evgSyr> but, because it is intended only for translators i hope this is not big problem.
60 2010-02-17T02:33:58 <ThomasWaldmann> 39 root 15 -5 0 0 0 S 286 0.0 1000:13 ksoftirqd/4
61 2010-02-17T02:34:09 <ThomasWaldmann> 286% cpu!?
62 2010-02-17T02:34:18 <evgSyr> hm.
63 2010-02-17T02:34:54 <evgSyr> may be, iowait?
64 2010-02-17T02:35:23 <evgSyr> (but, anyway, it's kernel proccess with single-core affinity set)
65 2010-02-17T02:35:36 <evgSyr> (so it shouldn't be more than 100)
66 2010-02-17T02:38:00 <ThomasWaldmann> strange stuff
67 2010-02-17T02:38:56 <evgSyr> ok, i need some sleep.
68 2010-02-17T02:38:57 <evgSyr> gn
69 2010-02-17T02:40:10 <ThomasWaldmann> gn
70 2010-02-17T02:42:11 *** evgSyr
71 2010-02-17T07:25:53 *** kursor
72 2010-02-17T08:36:47 *** kursor
73 2010-02-17T10:20:19 <dreimark> moin
74 2010-02-17T10:20:21 <dreimark> gna
75 2010-02-17T11:27:13 *** moinBot`
76 2010-02-17T11:28:35 *** johill_
77 2010-02-17T11:33:53 *** johill
78 2010-02-17T11:33:53 *** moinBot
79 2010-02-17T11:36:16 *** johill_
80 2010-02-17T11:50:13 <ThomasWaldmann> moin
81 2010-02-17T12:21:31 *** kursor
82 2010-02-17T13:12:06 <CIA-50> Reimar Bauer <rb.proj AT googlemail DOT com> default * 500:de15d700d553 extensions/data/plugin/macro/Image.py: macro.Image._transform: exception handler for transpose_func added
83 2010-02-17T13:57:22 <ThomasWaldmann> hmm, i can't reproduce the TypeError on moinmo.in any more. wtf?
84 2010-02-17T14:03:37 <ThomasWaldmann> CIA-50: wb
85 2010-02-17T14:25:02 <dreimark> wow and cs 500 for extensions
86 2010-02-17T14:28:57 <dreimark> there is may be no need to handle this as special bug if it is one?
87 2010-02-17T15:24:09 *** valeuf
88 2010-02-17T15:52:28 <CIA-50> Reimar Bauer <rb.proj AT googlemail DOT com> default * 501:16aa5b518631 extensions/data/plugin/macro/Image.py: macro.Image: removed wrong used /div
89 2010-02-17T15:52:32 <CIA-50> Reimar Bauer <rb.proj AT googlemail DOT com> default * 502:f9845325cfcf extensions/data/plugin/macro/Image.py: macro.Image: optional parameter clear added. This is for clearing the floating style.
90 2010-02-17T16:43:28 *** eSyr
91 2010-02-17T16:54:38 <valeuf> hello ^^
92 2010-02-17T16:54:57 <valeuf> I changed quickly this afternoon the patch for Docbook export
93 2010-02-17T16:55:15 <valeuf> http://moinmo.in/MoinMoinPatch/IncludeMacroWithDocBookFormatter
94 2010-02-17T16:55:52 <valeuf> to use a regexp to remove the editlinx argument
95 2010-02-17T16:57:48 <ThomasWaldmann> hi valeuf
96 2010-02-17T16:58:44 *** cdyson37
97 2010-02-17T16:59:32 <ThomasWaldmann> valeuf: if you like coding in python, you may want to read pep8 to write even prettier code
98 2010-02-17T17:00:27 <valeuf> ThomasWaldmann: do ou mean http://www.python.org/dev/peps/pep-0008/
99 2010-02-17T17:00:28 <valeuf> ?
100 2010-02-17T17:00:33 <ThomasWaldmann> yes
101 2010-02-17T17:00:33 <eSyr> pep8 make code not prettier, but peppier.
102 2010-02-17T17:00:46 <ThomasWaldmann> :D
103 2010-02-17T17:01:27 <ThomasWaldmann> many people like it that way. I don't want 100% of it (there is also some strange stuff), but 95%. :)
104 2010-02-17T17:01:55 <eSyr> well, code style guides are better than their abscence.
105 2010-02-17T17:02:05 <ThomasWaldmann> valeuf: if tested that with different inputs?
106 2010-02-17T17:02:24 <ThomasWaldmann> s/if/you/ %)
107 2010-02-17T17:02:26 <valeuf> ThomasWaldmann: I should say that I am a completely beginer with Python, but PEP8 seems to be interesting I will have a look
108 2010-02-17T17:02:40 <valeuf> ThomasWaldmann: I usually use C/C++ without any style ...
109 2010-02-17T17:02:57 <eSyr> valeuf, in moin, there is py.test which also checks pep8 source code compiance.
110 2010-02-17T17:03:01 <ThomasWaldmann> valeuf: we even have a pep8 automated test, that'll fail with badly formatted source :)
111 2010-02-17T17:03:07 <eSyr> hivemind.
112 2010-02-17T17:03:10 <ThomasWaldmann> hehe
113 2010-02-17T17:03:16 <valeuf> ThomasWaldmann: the reg exp directly come from the macro/Include.py to match editlink
114 2010-02-17T17:03:35 <ThomasWaldmann> yeah, but you tested it nevertheless, right?
115 2010-02-17T17:03:46 <valeuf> ThomasWaldmann: tested with such of entry like <<(FooBar editlink,,editlink)>>
116 2010-02-17T17:04:56 <valeuf> or <<(FooBareditlink,"editlink",editlink)::
117 2010-02-17T17:05:19 <ThomasWaldmann> ok, I'll apply that patch (or a similar fix), thanks.
118 2010-02-17T17:06:23 <valeuf> s/::/>>
119 2010-02-17T17:06:23 <valeuf> ThomasWaldmann: but I probably did not test all the situation
120 2010-02-17T17:06:29 <valeuf> eSyr: you are talking about _test/pep8.py ??
121 2010-02-17T17:06:38 <eSyr> valeuf, yes.
122 2010-02-17T17:07:04 <ThomasWaldmann> test_sourcecode.py is the main script, but it uses pep8.py
123 2010-02-17T17:07:11 <eSyr> you have to install py.test in order to make use of those scripts in _test directories.
124 2010-02-17T17:08:06 <valeuf> eSyr: that's pretty cool, I worked before for video game company, and everybody was writting different style of C++ it was really annoying
125 2010-02-17T17:08:46 <ThomasWaldmann> the good thing about python is that you can't write really crappy formatted code
126 2010-02-17T17:08:58 <ThomasWaldmann> well you can, but it won't work then :)
127 2010-02-17T17:09:47 <valeuf> ThomasWaldmann: I started python few months ago, and I should say that I really hate to have so strong rules about code formatting
128 2010-02-17T17:09:51 <eSyr> ThomasWaldmann, oh really? i think, on third nested list comprehension this thesis will not work.
129 2010-02-17T17:10:19 <eSyr> valeuf, they are not strong when using rather smart editor.
130 2010-02-17T17:10:25 <valeuf> ThomasWaldmann: but I should say that now I like it, because it is really easier to read code from other developper
131 2010-02-17T17:10:32 <eSyr> (which adds indents and checks syntax realtime)
132 2010-02-17T17:10:34 <ThomasWaldmann> eSyr: that's hard to understand code, but you can format it pretty :)
133 2010-02-17T17:11:00 <valeuf> eSyr: I need to tweak a little bit more vim to use python correctly I saw many script to use correctly python
134 2010-02-17T17:11:27 <ThomasWaldmann> valeuf: http://linuxwiki.de/Python see 7.
135 2010-02-17T17:12:13 <ThomasWaldmann> http://moinmo.in/CodingStyle < better there
136 2010-02-17T17:13:40 <valeuf> ThomasWaldmann: yep I already have these settings in my .vimrc but I am more looking for such of advanced features : http://blog.dispatched.ch/2009/05/24/vim-as-python-ide/
137 2010-02-17T17:14:04 <valeuf> ThomasWaldmann: especially the autocompletion
138 2010-02-17T17:14:26 <eSyr> well, for me Ctrl+P is enough.
139 2010-02-17T17:14:28 <ThomasWaldmann> if i need such stuff, i use eclipse/pydev
140 2010-02-17T17:14:51 <valeuf> ThomasWaldmann: eclipse is too slow on my old laptop unfortunately
141 2010-02-17T17:15:07 <ThomasWaldmann> you need 1GHz/1GB
142 2010-02-17T17:15:29 <ThomasWaldmann> and a good screen resolution
143 2010-02-17T17:17:05 <eSyr> hehe
144 2010-02-17T17:17:29 <valeuf> seems like I need to check more carefully my code : http://paste.pocoo.org/show/179335/
145 2010-02-17T17:17:45 <valeuf> this really cool te check the syntax like this
146 2010-02-17T17:18:22 <ThomasWaldmann> yes, it finds quite some stuff, but not everything
147 2010-02-17T17:54:32 *** cdyson37
148 2010-02-17T17:59:43 <ThomasWaldmann> http://dev.pocoo.org/projects/zine/browser/zine/plugins < do we want to have a similar plugin system?
149 2010-02-17T18:00:40 <eSyr> btw, some time ago i patched trac plugins and found its plugin system rather interesting.
150 2010-02-17T18:02:02 <johill> eh, quick question -- how do I convert to html entities in moin?
151 2010-02-17T18:02:16 <eSyr> johill, convert what?
152 2010-02-17T18:02:50 <johill> arbitrary pre-formatted text containing <> etc
153 2010-02-17T18:03:13 <TheSheep> werkzeug.html(u'foo'), for example
154 2010-02-17T18:03:29 <eSyr> formatter.text()?
155 2010-02-17T18:03:32 <TheSheep> but I think moin uses some function from MoinMoin.wikiutil
156 2010-02-17T18:03:47 <TheSheep> eSyr: if you have a formatter, yes
157 2010-02-17T18:03:53 <ThomasWaldmann> wikiutil.escape
158 2010-02-17T18:03:59 <eSyr> TheSheep, in moin, i usually have formatter.
159 2010-02-17T18:04:07 <ThomasWaldmann> formatter.text() also works of course
160 2010-02-17T18:04:10 <TheSheep> eSyr: depends what you are writing
161 2010-02-17T18:04:17 <eSyr> (and have to escape different characters depending on output format)
162 2010-02-17T18:04:29 <TheSheep> yup
163 2010-02-17T18:04:46 <eSyr> well, i've supposed that talk is about plugins.
164 2010-02-17T18:05:03 <TheSheep> auth plugins don't have formatter ;)
165 2010-02-17T18:05:18 <TheSheep> and I think neither actions
166 2010-02-17T18:05:46 <eSyr> actions do have.
167 2010-02-17T18:05:57 <TheSheep> only if you make one
168 2010-02-17T18:06:02 <eSyr> anything that may output something, has formatter.
169 2010-02-17T18:06:10 <eSyr> TheSheep, request.formatter.
170 2010-02-17T18:06:17 <eSyr> if i have request, i have formatter.
171 2010-02-17T18:06:50 <TheSheep> maybe something changed recently, but I'm used to the idea that action is what actually decides what formatter to use and creates it
172 2010-02-17T18:07:17 <eSyr> TheSheep, formatter decision usually comes from request.
173 2010-02-17T18:07:17 <TheSheep> that's the 'view' action
174 2010-02-17T18:07:25 <eSyr> but action of course can do anything.
175 2010-02-17T18:07:40 <eSyr> ok, let's check out code.
176 2010-02-17T18:08:13 <TheSheep> you said it all without actually checking? O.o
177 2010-02-17T18:08:39 <eSyr> well, i said it bewcause i wrote action recently and used request.formatter.
178 2010-02-17T18:08:54 <eSyr> (CheckTranslation action, more precisely)
179 2010-02-17T18:09:10 <eSyr> but i haven't checked where formatter decision is made.
180 2010-02-17T18:09:38 <TheSheep> eSyr: see action/__init__.py and the do_show()
181 2010-02-17T18:10:06 <eSyr> so?
182 2010-02-17T18:10:16 <eSyr> in show action you can overwrite formatter by request parameter.,
183 2010-02-17T18:10:19 <eSyr> *.
184 2010-02-17T18:10:32 <eSyr> mmm, more likely, do_show parameters never passed directly.
185 2010-02-17T18:10:33 <TheSheep> the formatter is created in the Page class, which is instantiated (sometimes) by the action
186 2010-02-17T18:10:56 <eSyr> i'm not sure about this.
187 2010-02-17T18:11:23 <TheSheep> http://hg.moinmo.in/moin/1.9/file/bdf396ab799d/MoinMoin/Page.py#l154
188 2010-02-17T18:11:33 <TheSheep> formatter is created in Page
189 2010-02-17T18:13:20 <TheSheep> it's instantiated even later: http://hg.moinmo.in/moin/1.9/file/bdf396ab799d/MoinMoin/Page.py#l1041
190 2010-02-17T18:13:54 <TheSheep> in Page.send_page
191 2010-02-17T18:14:01 <eSyr> you're checking wrong place, i think.
192 2010-02-17T18:14:11 <TheSheep> so only actions that actually call Page.send_page have a formatter
193 2010-02-17T18:14:19 <eSyr> web/contexts.py, lines 117-118
194 2010-02-17T18:16:04 <eSyr> and then wsgiapp.py
195 2010-02-17T18:16:22 <TheSheep> eSyr: that's hardwired to html formatter
196 2010-02-17T18:16:27 <eSyr> yes.
197 2010-02-17T18:16:31 <eSyr> (as default formatter)
198 2010-02-17T18:17:08 <TheSheep> and this is new code in 1.9
199 2010-02-17T18:17:27 <eSyr> well, i think, ther was something the same in twisted code.
200 2010-02-17T18:17:30 <TheSheep> I can see it's only used in few places
201 2010-02-17T18:17:40 <TheSheep> twisted code?
202 2010-02-17T18:17:56 *** valeuf
203 2010-02-17T18:18:21 <eSyr> hm, iirc, previous versions of moin used twisted for serving http.
204 2010-02-17T18:18:27 <TheSheep> not really
205 2010-02-17T18:18:35 <TheSheep> there was a variety of different servers
206 2010-02-17T18:18:40 <TheSheep> twisted was one of them
207 2010-02-17T18:18:43 <eSyr> oh, ok.
208 2010-02-17T18:19:06 <TheSheep> and I'm sure there was no formatter in the request
209 2010-02-17T18:19:27 <TheSheep> if you look at where it's used, it only appears in the actions that have been rewritten
210 2010-02-17T18:20:23 <TheSheep> that formatter is basically used as a fallback, if there is no Page to get a formatter from
211 2010-02-17T18:20:48 <TheSheep> the real one is still created in the Page.send_page
212 2010-02-17T18:20:58 <TheSheep> and only parsers and macros get it
213 2010-02-17T18:21:18 <eSyr> http://paste.pocoo.org/show/179353/ — apply this patch and check out.
214 2010-02-17T18:22:05 <TheSheep> that's for .19
215 2010-02-17T18:22:18 <eSyr> i can write similar patch for 1.8 :)
216 2010-02-17T18:22:22 <TheSheep> we already established that 1.9 has a fallback formatter in the request
217 2010-02-17T18:22:35 <eSyr> (somewhere in dispatcher code)
218 2010-02-17T18:22:40 <TheSheep> except it will give you a different result
219 2010-02-17T18:22:51 <eSyr> oh -_-
220 2010-02-17T18:23:09 <eSyr> we have formatter attribute of request.
221 2010-02-17T18:23:32 <eSyr> and if action want to redefine it, action can do this.
222 2010-02-17T18:23:45 <eSyr> but we do have formatter.
223 2010-02-17T18:24:30 <eSyr> (of course, if we want to output a page, we should set parser and formatter for it and run the parsing/formatting stuff)
224 2010-02-17T18:24:37 <eSyr> but anyway we have formatter.
225 2010-02-17T18:24:43 <eSyr> and can format things using it.
226 2010-02-17T18:24:47 <TheSheep> ok, I was wrong
227 2010-02-17T18:24:54 <TheSheep> the html_formatter is created in MoinMoin/request/__init__.py
228 2010-02-17T18:25:23 <TheSheep> at least in 1.8
229 2010-02-17T18:25:55 <eSyr> looks like it's true.
230 2010-02-17T18:26:18 <eSyr> yes, i see.
231 2010-02-17T18:26:42 <eSyr> in 1.9 it is in web.contexts
232 2010-02-17T18:27:10 <TheSheep> Sat May 26 17:39:21 2007 +0200 (2 years ago)
233 2010-02-17T18:27:16 <TheSheep> I was taken by surprise
234 2010-02-17T18:28:46 <eSyr> m?
235 2010-02-17T18:28:51 <TheSheep> I nee to learn to adapt to changes faster than 2 years per change :D
236 2010-02-17T18:28:58 <eSyr> :)
237 2010-02-17T18:29:28 <TheSheep> it was introduced in 1.6
238 2010-02-17T18:31:06 <eSyr> yes, with parser/formatter concept, i think.
239 2010-02-17T18:32:07 *** valeuf
240 2010-02-17T18:32:34 <TheSheep> no, that dates back to 1.3
241 2010-02-17T18:32:37 <TheSheep> or even 1.2
242 2010-02-17T18:32:59 <eSyr> hm, iirc correctly, there was major changes in it in 1.6
243 2010-02-17T18:33:13 <eSyr> s/iirc correctly/iirc/
244 2010-02-17T18:33:22 <TheSheep> back in 1.3 there was separate plugin for parsing whole page and separate for color-highlighting code
245 2010-02-17T18:33:51 <TheSheep> 1.6 had the moin_wiki parser rewritten
246 2010-02-17T18:34:04 <TheSheep> but not much change to the plugin infrastructure
247 2010-02-17T18:34:45 <TheSheep> well s/rewritten/modified
248 2010-02-17T18:43:30 *** valeuf
249 2010-02-17T18:46:24 *** kursor
250 2010-02-17T19:11:38 *** eSyr
251 2010-02-17T19:14:05 * dreimark looks at zine plugins
252 2010-02-17T19:15:58 <ThomasWaldmann> sometimes I find me thinking about taking zine, ripping all out we don't need and use the rest to build moin2 :)
253 2010-02-17T19:16:28 <ronny> ThomasWaldmann: ali already ripped zines core ideas apart for glashammer
254 2010-02-17T19:18:31 <ThomasWaldmann> he ripped the ideas apart, or he used the zine code? :)
255 2010-02-17T19:20:41 <ThomasWaldmann> dreimark: that migration script was moin 1.3 or so :)
256 2010-02-17T19:21:27 <dreimark> hmm, that is long ago. but needed manual actions. also i remember at that time unicode trouble on suse systems
257 2010-02-17T19:22:20 <ThomasWaldmann> if utf-8 didn't work since then, moin would blow up at all places since long
258 2010-02-17T19:22:26 <dreimark> and because of the amount of data we have had in the past on suse systems we still have files which weren't migrated here
259 2010-02-17T19:32:10 <TheSheep> ThomasWaldmann: lets move to django ;)
260 2010-02-17T19:32:49 <ThomasWaldmann> for what?
261 2010-02-17T19:32:55 <TheSheep> moin2
262 2010-02-17T19:33:01 <TheSheep> just kidding, of coyrse
263 2010-02-17T19:33:03 <TheSheep> course
264 2010-02-17T19:34:57 <ThomasWaldmann> i just had a quick look at trac plugins, but somehow I am not feeling much love for setup.py
265 2010-02-17T19:42:18 <ThomasWaldmann> bbl
266 2010-02-17T19:45:40 <TheSheep> setuptools are pretty horrible
267 2010-02-17T20:23:20 <ronny> ThomasWaldmann: ali took the ideas, and made it pretty nice
268 2010-02-17T20:30:42 <TheSheep> what I like in trac, from the user's pov, is that you can have all plugins instaleld, but only some enabled
269 2010-02-17T20:30:50 <TheSheep> mercurial is like that too
270 2010-02-17T20:31:08 <TheSheep> which reminds me to investigate mercurial's plugin system
271 2010-02-17T21:22:25 <ThomasWaldmann> TheSheep: same for zine?
272 2010-02-17T21:24:43 <dreimark> we would need enabling and may be controlling who has which rights
273 2010-02-17T21:25:28 <dreimark> if a plugin has read/write capabilities and acls say only read then write should be disabled
274 2010-02-17T21:25:29 <TheSheep> ThomasWaldmann: didn't look at zine too closely
275 2010-02-17T21:27:28 <ronny> TheSheep: hg's plugin system is basically import a bunch of listed modules, call functions in them
276 2010-02-17T21:27:36 <ThomasWaldmann> TheSheep: it's OSS :)
277 2010-02-17T21:28:54 <TheSheep> ronny: KISS :)
278 2010-02-17T21:29:37 <dreimark> and also we have plugins we can only use as superuser
279 2010-02-17T21:30:02 <ronny> TheSheep: they use different hacks to integrate different plugins, monkeypatching __class__ in instances
280 2010-02-17T21:30:12 <ronny> well, sometimes
281 2010-02-17T21:58:22 * ThomasWaldmann still thinks about suitable k/v-store
282 2010-02-17T22:58:45 *** moin4
283 2010-02-17T23:09:50 *** moin4
284 2010-02-17T23:40:51 * dreimark xmlrpc needs some useablity fixes
285
