1 2010-01-16T00:01:12 *** CIA-48
2 2010-01-16T00:08:32 *** CIA-48
3 2010-01-16T00:08:32 *** ChanServ
4 2010-01-16T00:08:32 *** dennda
5 2010-01-16T00:08:32 *** grzywacz
6 2010-01-16T00:08:32 *** TheSheep
7 2010-01-16T00:08:32 *** johill
8 2010-01-16T00:08:32 *** waldi
9 2010-01-16T00:08:32 *** elmo
10 2010-01-16T00:08:32 *** franklin
11 2010-01-16T00:08:32 *** vpv
12 2010-01-16T00:08:32 *** ThomasWaldmann
13 2010-01-16T00:08:32 *** xorAxAx
14 2010-01-16T00:14:39 *** ChanServ
15 2010-01-16T00:14:39 *** CIA-48
16 2010-01-16T00:14:39 *** TheSheep
17 2010-01-16T00:14:39 *** franklin
18 2010-01-16T00:14:39 *** vpv
19 2010-01-16T00:14:39 *** dennda
20 2010-01-16T00:14:39 *** grzywacz
21 2010-01-16T00:14:39 *** johill
22 2010-01-16T00:14:39 *** elmo
23 2010-01-16T00:14:39 *** waldi
24 2010-01-16T00:14:39 *** ThomasWaldmann
25 2010-01-16T00:14:39 *** xorAxAx
26 2010-01-16T00:14:39 *** irc.freenode.net
27 2010-01-16T00:19:46 *** CIA-48
28 2010-01-16T00:21:37 *** CIA-1
29 2010-01-16T00:22:21 *** CIA-1
30 2010-01-16T00:22:26 *** CIA-30
31 2010-01-16T00:24:41 *** eSyr
32 2010-01-16T00:30:19 *** eSyr
33 2010-01-16T00:35:18 *** eSyr
34 2010-01-16T00:37:19 *** eSyr
35 2010-01-16T00:40:07 *** eSyr
36 2010-01-16T00:42:34 *** eSyr
37 2010-01-16T00:45:30 *** eSyr
38 2010-01-16T00:48:05 *** eSyr
39 2010-01-16T00:51:32 *** eSyr
40 2010-01-16T00:54:17 *** eSyr
41 2010-01-16T00:59:05 *** eSyr
42 2010-01-16T01:00:06 *** eSyr
43 2010-01-16T01:02:20 *** eSyr
44 2010-01-16T01:02:32 *** evgSyr
45 2010-01-16T01:03:55 *** eSyr
46 2010-01-16T01:35:14 *** evgSyr
47 2010-01-16T01:35:24 *** evgSyr
48 2010-01-16T01:35:37 *** evgSyr
49 2010-01-16T01:36:10 <eSyr> i'm back.
50 2010-01-16T01:46:43 <ThomasWaldmann> wb eSyr
51 2010-01-16T01:48:37 <ThomasWaldmann> eSyr: i like that diff action navigation
52 2010-01-16T01:49:17 <eSyr> any remarks on this patch?
53 2010-01-16T01:51:21 <ThomasWaldmann> i am still reviewing it.
54 2010-01-16T01:51:25 <eSyr> ok.
55 2010-01-16T01:51:43 <ThomasWaldmann> it has unintended changes that are not necessary.
56 2010-01-16T01:52:01 <eSyr> yes, there are some, i may separate them.
57 2010-01-16T01:52:37 <eSyr> (but, i think, handling of determining prev/next revision is more proper)
58 2010-01-16T01:55:09 <ThomasWaldmann> (i mean stuff like removing some empty line)
59 2010-01-16T01:55:17 <eSyr> hm...
60 2010-01-16T01:56:16 <eSyr> mmm, can't see removing empty line in diff.
61 2010-01-16T01:56:56 <ThomasWaldmann> that args stuff maybe should be done with a dict with named args
62 2010-01-16T01:58:44 <eSyr> hm, maybe.
63 2010-01-16T01:59:52 <ThomasWaldmann> hmm, the current way you do things would work then
64 2010-01-16T01:59:59 <ThomasWaldmann> not
65 2010-01-16T02:00:21 <eSyr> mmm, what do you mean?
66 2010-01-16T02:01:55 <ThomasWaldmann> btw, in german i currently need full 1440px browser window with to see that navigation
67 2010-01-16T02:02:28 <eSyr> oh, sorry.
68 2010-01-16T02:02:46 <eSyr> i have 1920x1200 on my notebook and someime forgot to check pages in smaller resolutions.
69 2010-01-16T02:03:10 <eSyr> i thought about changing "next/previous revision" captions to <</>>
70 2010-01-16T02:04:39 <ThomasWaldmann> it should work with 1024px width and it would be nice if it didn't stay fixed width, but reflow somehow for lesser resolutions
71 2010-01-16T02:05:46 <ThomasWaldmann> mabye the prev/next rev could be part of the td
72 2010-01-16T02:06:21 <ThomasWaldmann> would make sense because they influence what is display in that (half of the table) td
73 2010-01-16T02:06:36 <eSyr> ok.
74 2010-01-16T02:07:03 <eSyr> should i use <</>> instead of current captions?
75 2010-01-16T02:07:19 <ThomasWaldmann> replacing prev/next rev by just < / > would be nice because of no i18n change required
76 2010-01-16T02:08:14 <ThomasWaldmann> and i think it should be rather < than <<. we just change by 1, << somehow suggests a bigger change
77 2010-01-16T02:08:25 <eSyr> ok.
78 2010-01-16T02:09:44 <ThomasWaldmann> do you think that < > buttons could go into the line that says "revision 1 of <date>", just after it?
79 2010-01-16T02:10:05 <eSyr> hm.
80 2010-01-16T02:10:33 <eSyr> well, i makes appearance of diff geader rather inconsistent, i think.
81 2010-01-16T02:10:49 <ThomasWaldmann> that would make some sense location-wise though
82 2010-01-16T02:10:59 <eSyr> no it is "row of buttons|block of info|diff contan"
83 2010-01-16T02:11:32 <eSyr> *now it is
84 2010-01-16T02:11:38 <eSyr> hm.
85 2010-01-16T02:11:47 <ThomasWaldmann> yes, but if you put < > near "revision x", it is rather logical what < > means
86 2010-01-16T02:12:00 <eSyr> then better if they will be links then.
87 2010-01-16T02:13:00 <ThomasWaldmann> but shoud be POST still?
88 2010-01-16T02:13:04 <ThomasWaldmann> l
89 2010-01-16T02:13:33 <eSyr> errr, butons these produce (and produced) get queries.
90 2010-01-16T02:14:14 <eSyr> (even more, i can't understand why they where buttons)
91 2010-01-16T02:14:47 <ThomasWaldmann> oh, ok, then it is easier even
92 2010-01-16T02:15:24 <ThomasWaldmann> ok, so just a link "next" and "previous" in that "revision x of date" line?
93 2010-01-16T02:15:31 <eSyr> ok.
94 2010-01-16T02:15:49 <ThomasWaldmann> that solves width issue also :)
95 2010-01-16T02:17:08 <ThomasWaldmann> that .sort() call, is it needed?
96 2010-01-16T02:17:47 <eSyr> well, it guarantees hat revlist is sorted, so i can get proper prev/nex revision from it.
97 2010-01-16T02:18:32 <eSyr> (old code used min/max, so i'm not sure is it sorted)
98 2010-01-16T02:19:48 * ThomasWaldmann is not sure either, but I suspect it is
99 2010-01-16T02:20:11 <eSyr> well, i'll check and if it is, i'll remove his call.
100 2010-01-16T02:23:51 <ThomasWaldmann> disable_next related code should be replaced by your disable_button call
101 2010-01-16T02:25:18 <eSyr> yes, i've notived and done it already.
102 2010-01-16T02:25:25 <eSyr> (locally)
103 2010-01-16T02:26:48 <ThomasWaldmann> and maybe disable_button(do_disable) -> enabled(val) # i like positive logic
104 2010-01-16T02:26:58 <eSyr> ok.
105 2010-01-16T02:43:09 <ThomasWaldmann> btw, maybe bisect -> bisect_right, just for symmetry
106 2010-01-16T02:43:19 <eSyr> ok.
107 2010-01-16T02:43:41 <ThomasWaldmann> and some of the args computation is duplicated expressions
108 2010-01-16T02:44:36 <ThomasWaldmann> I'll wait after the next changeset with looking at those expressions.
109 2010-01-16T04:20:46 <eSyr> ThomasWaldmann, i've updated the patch
110 2010-01-16T04:28:40 <eSyr> Also, little css cleanup (mostly, formatting): http://paste.pocoo.org/show/165928/
111 2010-01-16T04:35:18 <ThomasWaldmann> hehe, the night shift |-)
112 2010-01-16T04:36:26 <ThomasWaldmann> eSyr: are there tabs or other indentation in css? patch looks suspicious.
113 2010-01-16T04:36:39 <eSyr> no.
114 2010-01-16T04:37:07 <eSyr> only 4 spaces inside style definitions :)
115 2010-01-16T04:37:32 <eSyr> this patch fixes some missed whitespace, mostly.
116 2010-01-16T04:38:30 <ThomasWaldmann> so why does 402 look different than 403?
117 2010-01-16T04:39:36 <ThomasWaldmann> (maybe either keep same indentation or fix all to same level)
118 2010-01-16T04:39:40 <eSyr> hehe, looks like old csses used abs instead of spaces :)
119 2010-01-16T04:39:44 <eSyr> ok.
120 2010-01-16T04:39:59 <ThomasWaldmann> and no tabs
121 2010-01-16T04:40:20 <ThomasWaldmann> maybe do ws changes in separate changeset
122 2010-01-16T04:40:21 <eSyr> ThomasWaldmann, they already are in old csses.
123 2010-01-16T04:40:28 <eSyr> ok.
124 2010-01-16T04:41:06 <eSyr> can i separate this patch in two commits and push it, then?
125 2010-01-16T04:42:33 <ThomasWaldmann> that 50% width change, what does it exactly do?
126 2010-01-16T04:42:50 <ThomasWaldmann> well, it will do 50/50 split for normal diff display
127 2010-01-16T04:43:04 <ThomasWaldmann> but what about diffs with unbalanced content,
128 2010-01-16T04:43:12 <eSyr> it forces diff columns to be same widths, so when you switch to next/prev rev/diff, links do not jump
129 2010-01-16T04:43:21 <ThomasWaldmann> e.g. if in left column, there is few content (width)
130 2010-01-16T04:43:35 <ThomasWaldmann> and in right column there is wide content?
131 2010-01-16T04:44:12 <eSyr> well, the old patch uses buttons, which was in fixed place, but now that problem become in place.
132 2010-01-16T04:44:18 <ThomasWaldmann> you mean that you do not have to move mouse?
133 2010-01-16T04:44:27 <eSyr> well, i can't see nothing ad in unbalanced diffs.
134 2010-01-16T04:44:32 <eSyr> yes.
135 2010-01-16T04:44:52 <eSyr> (also, i want to implement inline diff view, like in pocoo or trac)
136 2010-01-16T04:45:02 <eSyr> *on
137 2010-01-16T04:45:09 <ThomasWaldmann> what about unbreakable content?
138 2010-01-16T04:45:37 <eSyr> layouting of it has higher priority.
139 2010-01-16T04:45:45 <ThomasWaldmann> e.g. -aaa left, +bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb right?
140 2010-01-16T04:45:58 <eSyr> yes, it crashes tale layout.
141 2010-01-16T04:46:55 <ThomasWaldmann> so there is no regression for those cases by 50/50?
142 2010-01-16T04:47:03 <eSyr> (we can force not to crash with divs and word-break and some oher hacks, but i don't see any sense in this, i think, that in this rare case unbalancing is better)
143 2010-01-16T04:47:08 <eSyr> no.
144 2010-01-16T04:47:38 <eSyr> *i mean, yes, no regressions
145 2010-01-16T04:47:43 <ThomasWaldmann> ok, commit
146 2010-01-16T04:47:51 <eSyr> thanks.
147 2010-01-16T05:02:23 <CIA-30> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5434:382cc546797e 1.9/MoinMoin/web/static/htdocs/ (4 files in 4 dirs): CSS: Whitespace cleanup of diff-info style definitions
148 2010-01-16T05:02:26 <CIA-30> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5435:b97aa05aee28 1.9/MoinMoin/web/static/htdocs/ (4 files in 4 dirs): CSS: Added width:50% to diff columns style
149 2010-01-16T05:02:31 <CIA-30> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5436:2c025e231b5f 1.9/MoinMoin/web/static/htdocs/ (4 files in 4 dirs): CSS: diff info header centered
150 2010-01-16T05:06:05 <eSyr> hm, and what about latest diff action patch?
151 2010-01-16T05:12:30 <ThomasWaldmann> did you pastebin a new one?
152 2010-01-16T05:13:53 <eSyr> well, i've added it on FR page.
153 2010-01-16T05:15:54 <eSyr> http://paste.pocoo.org/show/165930/
154 2010-01-16T05:17:34 <ThomasWaldmann> + return not val and u' disabled="disabled"' or u''
155 2010-01-16T05:17:48 <eSyr> yes.
156 2010-01-16T05:17:53 <eSyr> http://paste.pocoo.org/show/165931/
157 2010-01-16T05:17:57 <ThomasWaldmann> return val and u'' or u' disabled...'
158 2010-01-16T05:18:04 <eSyr> no.
159 2010-01-16T05:18:09 <eSyr> u'' is false
160 2010-01-16T05:18:43 <eSyr> (that's the difference between c's ?: and python's and or)
161 2010-01-16T05:18:45 <ThomasWaldmann> ah, right
162 2010-01-16T05:19:26 <eSyr> just because Guido thinks ?: is too complicated and no tpython-way <_<
163 2010-01-16T05:22:24 <ThomasWaldmann> 12..19 needed?
164 2010-01-16T05:24:03 <eSyr> errr, what?
165 2010-01-16T05:25:28 <ThomasWaldmann> referring to the highlighted patch line numbers on the wikipage
166 2010-01-16T05:25:53 <eSyr> ah, sorry.
167 2010-01-16T05:26:37 <eSyr> no, i've returned empy line back: http://paste.pocoo.org/show/165931/
168 2010-01-16T05:30:31 <ThomasWaldmann> btw, although querystr has ...str in the name, the recommended usage is as a dict
169 2010-01-16T05:30:53 <eSyr> ok.
170 2010-01-16T05:30:54 <ThomasWaldmann> (if not even required)
171 2010-01-16T05:36:42 <eSyr> any other remarks?
172 2010-01-16T05:42:20 <ThomasWaldmann> not yet, can you post a current diff, so i can apply and play with it?
173 2010-01-16T05:43:29 <eSyr> of course: http://paste.pocoo.org/show/165937/
174 2010-01-16T05:47:37 <ThomasWaldmann> if user may revert: revert_html = "..." else: revert_html = ""
175 2010-01-16T05:48:44 <ThomasWaldmann> page = Page(request, pagename)
176 2010-01-16T05:48:57 <ThomasWaldmann> see above: currentpage = same
177 2010-01-16T05:50:17 <ThomasWaldmann> querystr={ 'action': 'diff', 'rev1': old_rev, 'rev2': new_rev, }
178 2010-01-16T05:51:13 <ThomasWaldmann> just as a hint for the lazy coders: querystr=dict(action='diff', rev1=old_rev, rev2=new_rev)
179 2010-01-16T05:53:19 <eSyr> oh, yes, i've noticed it but forgot to fix.
180 2010-01-16T05:54:07 <eSyr> hehe, after js i'm highly addicted to first syntax :)
181 2010-01-16T05:55:01 <ThomasWaldmann> "No newer revision available fo diff"
182 2010-01-16T05:55:25 <ThomasWaldmann> btw, how many new i18n strings will that give us?
183 2010-01-16T05:55:46 <eSyr> four. Bu hey are in title hints only.
184 2010-01-16T05:55:50 <eSyr> *But
185 2010-01-16T05:55:55 <eSyr> *they
186 2010-01-16T05:56:10 <eSyr> so, i think, it is not a big issue.
187 2010-01-16T05:56:26 <ThomasWaldmann> ok
188 2010-01-16T05:56:38 * ThomasWaldmann tries it in practice now
189 2010-01-16T06:00:26 <ThomasWaldmann> deleted revisions cause minor navigation issues
190 2010-01-16T06:00:46 <ThomasWaldmann> likely because they are not in the revlist
191 2010-01-16T06:01:25 <eSyr> hm, can you provide example?
192 2010-01-16T06:02:30 <ThomasWaldmann> just create a page with rev 1, 2, deleted 3, 4 and navigate around
193 2010-01-16T06:02:53 <eSyr> well, looks correct for me.
194 2010-01-16T06:02:59 <eSyr> i. e., it skips rev 3.
195 2010-01-16T06:03:31 <eSyr> (i think, the same will be when acions with attachments happened)
196 2010-01-16T06:04:31 <ThomasWaldmann> when starting from info action, you can get diff 3-4 or 2-3, but once you lose 3 in the nav, it'll only show 2-4
197 2010-01-16T06:04:35 <eSyr> oh, no, new revisions are no create when attachments are changed.
198 2010-01-16T06:05:02 <eSyr> hm, well.
199 2010-01-16T06:05:18 <eSyr> i thought that using revlist is more correct approach.
200 2010-01-16T06:05:45 <eSyr> i can use simple inc/dec againg, anyway.
201 2010-01-16T06:06:29 <ThomasWaldmann> yeah, maybe less confusing
202 2010-01-16T06:07:05 <ThomasWaldmann> but at least /me learned about bisect :)
203 2010-01-16T06:07:21 <eSyr> :)
204 2010-01-16T06:08:31 <eSyr> hm, that what for revlist is used?
205 2010-01-16T06:08:41 <eSyr> (in old code)
206 2010-01-16T06:09:24 <eSyr> hm.
207 2010-01-16T06:09:25 <eSyr> no.
208 2010-01-16T06:09:46 <eSyr> i think, i should figure out, what is source of info rev list and why they differ.
209 2010-01-16T06:10:00 <eSyr> because some revisions can be deleted physically, for example.
210 2010-01-16T06:11:58 <ThomasWaldmann> hm?
211 2010-01-16T06:12:56 <ThomasWaldmann> i guess just using 1..n is maybe not far fetched. n being current rev (may be a deleted rev)
212 2010-01-16T06:13:20 <eSyr> when you physacally delete rev, then what happens?
213 2010-01-16T06:13:40 <eSyr> looks like getRevList handle that situation prperly.
214 2010-01-16T06:14:38 <ThomasWaldmann> not sure what you mean with physically
215 2010-01-16T06:14:53 <eSyr> rm data/pages/PageName/revisions/000005
216 2010-01-16T06:15:25 <ThomasWaldmann> deleted revs are that way in 1.x
217 2010-01-16T06:15:48 <ThomasWaldmann> there is just some code that fakes empty string for them
218 2010-01-16T06:15:50 <eSyr> (and there is no revision file for deleted rev, hm)
219 2010-01-16T06:15:59 <eSyr> oh -_-
220 2010-01-16T06:16:01 <ThomasWaldmann> (as content)
221 2010-01-16T06:16:14 <eSyr> ok, then you're right.
222 2010-01-16T06:16:26 <eSyr> in that case, i think, getRevList is useless.
223 2010-01-16T06:17:16 <eSyr> (in diff action)
224 2010-01-16T06:25:50 <ThomasWaldmann> yeah, looks like
225 2010-01-16T06:37:12 <TheSheep> do you guys never sleep?
226 2010-01-16T06:38:02 <eSyr> ThomasWaldmann, http://paste.pocoo.org/show/165946/
227 2010-01-16T06:38:17 <eSyr> TheSheep, we sleep when you sleep, sometimes.
228 2010-01-16T06:58:47 <ThomasWaldmann> diffing 2 deleted revs kills the nav
229 2010-01-16T06:59:23 <ThomasWaldmann> likely any diff with 2 same contents
230 2010-01-16T07:00:13 <ThomasWaldmann> maybe one should remove specialcasing this
231 2010-01-16T07:13:33 <eSyr> ThomasWaldmann, try this patch: http://paste.pocoo.org/show/165949/
232 2010-01-16T07:13:57 <eSyr> (i think, this is more proper way)
233 2010-01-16T11:58:35 * dreimark we need manual purging old sessions because cleansessions: checks.append(lambda session: session['expires'] < now) gives KeyError: 'expires'
234 2010-01-16T11:58:39 <dreimark> that is BAD
235 2010-01-16T11:59:38 <dreimark> so it think we better do this by a mig script
236 2010-01-16T11:59:43 <dreimark> s/it/i/
237 2010-01-16T12:11:51 <ThomasWaldmann> moin
238 2010-01-16T12:12:39 <ThomasWaldmann> dreimark: you need to run it once anyway otherwise it could not cleanup the old session files
239 2010-01-16T12:13:00 <dreimark> it can't it fails with the keyerror
240 2010-01-16T12:13:14 <ThomasWaldmann> --all
241 2010-01-16T12:14:58 <dreimark> ok, that helped, anyways why don't we want to add it to the mig cript?
242 2010-01-16T12:15:13 <dreimark> btw. good morning seems like you have a short night
243 2010-01-16T12:17:14 <dreimark> +had
244 2010-01-16T12:17:19 <ThomasWaldmann> was long enough, just unusually distributed :)
245 2010-01-16T12:21:15 <ThomasWaldmann> I'll add a HINT: to docs/CHANGES that people run it once manually and consider adding a cron script
246 2010-01-16T12:22:17 <dreimark> we should have a cron inside of moin
247 2010-01-16T12:22:30 <ThomasWaldmann> the 1.8 session code had autocleanup within moin
248 2010-01-16T12:23:05 <ThomasWaldmann> but it could lead to rather long response times now and then if you have many sessions
249 2010-01-16T12:23:17 <dreimark> i know, johil mentioned it. and it was by refactoring to werkzeug removed.
250 2010-01-16T12:23:52 <dreimark> how much time doese it costs to check at a limit e.g. 1k session files
251 2010-01-16T12:25:48 <dreimark> i mean i know it takes time but how much is it and is it worth to create more crontab dependencies instead living with it that once a day it checks for e.g. 5mins all
252 2010-01-16T12:32:37 <ThomasWaldmann> the questions is more about how much we need to do how often to avoid session file count runaway
253 2010-01-16T12:33:22 <ThomasWaldmann> processing all sessions every 1000 requests or so avoids that, but can make that request rather long if there are many sessions
254 2010-01-16T12:43:35 <dreimark> i have no idea how long "rather long" is. Or if we can trigger the cleanup process if the request takes more than x amount of time
255 2010-01-16T12:50:42 <ThomasWaldmann> that likely is O(n) with n being the amount of session files
256 2010-01-16T12:51:32 <ThomasWaldmann> it's more a mathematical problem :)
257 2010-01-16T12:51:37 <johill> it all depends on your IO speed and filesystem
258 2010-01-16T12:52:05 <johill> and I noticed that ext3 doesn't shrink directory inodes, apparently
259 2010-01-16T12:52:28 <johill> so after I had 200k session files and deleted them all, ls would still take forever on that dir
260 2010-01-16T12:53:08 <ThomasWaldmann> oh, yeah, that's bad :|
261 2010-01-16T12:53:34 <johill> after re-creating the dir it was much better
262 2010-01-16T12:53:53 * ThomasWaldmann had a similar issue with some shop software at a customer. it had never had session cleanup for some years and there was a gazillion of session files.
263 2010-01-16T12:55:20 <johill> now I have 18k session files again
264 2010-01-16T12:56:47 <johill> btw
265 2010-01-16T12:56:52 <johill> moin.py
266 2010-01-16T12:56:53 <johill> General options:
267 2010-01-16T12:56:56 <johill> --wiki-url=wiki.example.org/
268 2010-01-16T12:56:59 <johill> needs http now, no?
269 2010-01-16T12:58:19 <johill> KeyError: 'expires'
270 2010-01-16T12:59:04 <johill> ok if I remove all files that don't have 'expires' in them it works
271 2010-01-16T12:59:31 <johill> down to < 300 sessions now
272 2010-01-16T12:59:49 <johill> I need --all for the cronjob :)
273 2010-01-16T13:00:27 <ThomasWaldmann> yes, http:// needed
274 2010-01-16T13:00:51 <ThomasWaldmann> johill: no, don't put --all in the cronjob, just do that once
275 2010-01-16T13:01:56 <johill> why not?
276 2010-01-16T13:02:20 <johill> I mean instead of --wik-url
277 2010-01-16T13:04:04 * ThomasWaldmann fixes the http://
278 2010-01-16T13:04:28 <ThomasWaldmann> johill: what you mean doesn't exist yet
279 2010-01-16T13:04:54 <johill> I know :)
280 2010-01-16T13:10:57 <CIA-30> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5437:a20de9383481 1.9/MoinMoin/script/ (21 files in 8 dirs): fix --wiki-url examples, needs http:// now
281 2010-01-16T13:12:46 * ThomasWaldmann needs to go soon
282 2010-01-16T14:26:05 <dreimark> johill: do you have anonymous sessions?
283 2010-01-16T14:33:49 <johill> yes
284 2010-01-16T14:33:52 <johill> needed for openid
285 2010-01-16T14:35:28 <johill> there's another bug
286 2010-01-16T14:35:39 <johill> I used to not save sessions unless they were sent by the client
287 2010-01-16T14:35:54 <johill> it kinda looks like that might've gotten broken too
288 2010-01-16T14:37:07 <johill> correction, sent or required
289 2010-01-16T14:37:15 <johill> i.e. if there was only a single trail-item, don't save
290 2010-01-16T14:37:34 <dreimark> makes sense
291 2010-01-16T14:39:26 <eSyr> moin
292 2010-01-16T14:40:00 <johill> yup that is definitely broken
293 2010-01-16T14:40:07 <johill> creates an empty session file
294 2010-01-16T14:40:19 <johill> so it looks like the logic to not add the trail in that case is still there
295 2010-01-16T14:40:30 <johill> but not the logic to not save the session if it's empty
296 2010-01-16T14:41:24 <johill> actually, n/m that
297 2010-01-16T14:41:31 <johill> it seems the logic was restored now
298 2010-01-16T14:41:36 <johill> in a recent commit
299 2010-01-16T14:42:35 <eSyr> (and i still about patches)
300 2010-01-16T14:43:44 <johill> no
301 2010-01-16T14:43:47 <johill> it wasn't restored
302 2010-01-16T14:43:56 <johill> my config was just overwritten and I no longer had anonymous sessions
303 2010-01-16T14:47:33 <johill> http://moinmo.in/MoinMoinBugs/MoinSavesEmptySessions
304 2010-01-16T14:48:32 <eSyr> johill, i sill about my patches :)
305 2010-01-16T14:50:03 <johill> hm?
306 2010-01-16T14:50:33 <eSyr> never mind.
307 2010-01-16T15:01:10 <johill> uh wtf
308 2010-01-16T15:01:25 <johill> the groups code doesn't actually allow writing backend-independent code???!!
309 2010-01-16T15:07:40 <johill> fixed http://moinmo.in/MoinMoinBugs/GroupsNoLongerSorted
310 2010-01-16T15:08:07 <johill> ok that has a bug
311 2010-01-16T15:19:25 * dreimark has added a patch for the session bug - not sure if it is such easy
312 2010-01-16T15:20:57 <johill> you need to set the cookie actually
313 2010-01-16T15:21:00 <johill> but then not store the contents to disk
314 2010-01-16T15:21:30 <johill> bbl
315 2010-01-16T15:34:38 <dreimark> eSyr: sometimes i get http://paste.pocoo.org/show/166033/
316 2010-01-16T15:35:46 *** eSyr
317 2010-01-16T15:36:00 *** eSyr
318 2010-01-16T15:48:13 <dreimark> hmm, interesting why do we have 3 session file requests per page request?
319 2010-01-16T15:49:54 <dreimark> (of a logged in user)
320 2010-01-16T15:53:18 <dreimark> i mean get requsts of the session file
321 2010-01-16T15:53:50 <eSyr> ehrm.
322 2010-01-16T16:19:34 <CIA-30> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5438:bac3d5e9ffb9 1.9/MoinMoin/script/maint/globaledit.py: moin maint globaledit: Added support of master-revision special comment (patch by MoinMoin:RenatoSilva)
323 2010-01-16T16:19:37 <CIA-30> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5439:282276752499 1.9/MoinMoin/script/maint/globaledit.py: Merge
324 2010-01-16T17:16:01 <dreimark> johill: currently i try wget and ab while looking at memory. It looks like WantedPages needs a lot of memory
325 2010-01-16T19:45:56 <johill> dreimark: re file requests -- no idea
326 2010-01-16T19:46:31 <johill> dreimark: re wget, wantedpages makes sense, but memory should go down again, yet doesn't for me
327 2010-01-16T21:55:52 <dreimark> re
328 2010-01-16T22:08:31 <johill> wb I guess
329 2010-01-16T22:20:26 <dreimark> thx
330 2010-01-16T23:43:46 <CIA-30> Radomir Dopieralski <moindev@sheep.art.pl> default * 5440:6c0e03fcc067 1.9/ (MoinMoin/parser/_creole.py wikiconfig.py): update the wikicreole parser to version 1.1
331 2010-01-16T23:47:03 <dreimark> :(
332 2010-01-16T23:47:06 <dreimark> oops
333 2010-01-16T23:47:08 <dreimark> :)
334 2010-01-16T23:55:55 <TheSheep> ouch, I touched wikiconfig.py
335 2010-01-16T23:55:57 <TheSheep> sorry
336 2010-01-16T23:57:14 <TheSheep> just next to a warning to not change that file too! :)
337 2010-01-16T23:57:37 <dreimark> use a wikiconfig_local
338 2010-01-16T23:59:50 <TheSheep> yeah, I must have made that change months ago and forgotten about it