1 2011-04-05T00:03:40 <ThomasWaldmann> http://wiki.list.org/display/DEV/Google+Summer+of+Code+2011 < see last project idea before "Mentors" :)
2 2011-04-05T00:06:04 <dreimark> :)
3 2011-04-05T00:07:07 <sinha> Installed : libjpeg-turbo-devel-1.1.0-2.fc14.i686 , libjpeg-turbo-static-1.1.0-2.fc14.i686 but still getting same error ..seems like libjpeg62 package is different for redhat/fedora
4 2011-04-05T00:08:22 *** m4k3r
5 2011-04-05T00:09:06 <ThomasWaldmann> mine is without turbo :)
6 2011-04-05T00:15:33 <sinha> okay i will try to figure it out
7 2011-04-05T00:16:42 <sinha> but again the same question isnt any way to add the styles in rendering a image object rather than just width and height
8 2011-04-05T00:17:09 <sinha> in embeding into the content
9 2011-04-05T00:17:16 <ThomasWaldmann> {{png||class="foo"}}
10 2011-04-05T00:17:20 <sinha> not displayg it separately
11 2011-04-05T00:17:26 <sinha> it works !
12 2011-04-05T00:17:38 <ThomasWaldmann> but it doesn't currently work, likely due to same reason as &w and &h
13 2011-04-05T00:18:36 <sinha> yeah i am also seeing that
14 2011-04-05T00:19:58 <sinha> but one more thing here how i could one add a new class ,means any user through web interface wont be able to do it ,right
15 2011-04-05T00:21:20 <sinha> in that case adding inline styling would be convenient
16 2011-04-05T00:21:44 <sinha> like {{ png || style="width:100px;height:100px;" }}
17 2011-04-05T00:21:56 <dreimark> inline styink is a risk
18 2011-04-05T00:22:11 <dreimark> s/k/g
19 2011-04-05T00:23:22 <sinha> you mean something like this {{ png || style="s/k/g" }}
20 2011-04-05T00:24:24 <sinha> but will it a risk for server side ,while just adding this style part as it as to the output,like rendered img will be <img src="" style="s/k/g" />
21 2011-04-05T00:24:29 <dreimark> with style you can add js
22 2011-04-05T00:25:32 <dreimark> and one can use it for xss attacks
23 2011-04-05T00:26:14 <dreimark> sinha: s/k/g was meaning to fix a typo in styink - which were two in that word. "styling"
24 2011-04-05T00:26:42 <sinha> okay :P
25 2011-04-05T00:27:00 <sinha> yeah but to prevent js attack we can remove those "<" chars
26 2011-04-05T00:27:23 <sinha> we will be having only those chars which are used in styling
27 2011-04-05T00:27:34 <sinha> and i guess that would take care of many cases
28 2011-04-05T00:28:11 <sinha> only using chars, numbers and special symbol like "- ; : " thats it
29 2011-04-05T00:29:04 <sinha> chars means alphabets
30 2011-04-05T00:30:17 <ThomasWaldmann> sinha: i did a long investigation about making style safe. and found people who tried same and finally gave up.
31 2011-04-05T00:30:52 <sinha> okay i understand
32 2011-04-05T00:31:57 <sinha> but we should provide some minimal class which could satisfy user's needs
33 2011-04-05T00:32:08 <sinha> like if he want to make it float left or right
34 2011-04-05T00:32:53 <ThomasWaldmann> you mean for the show view of images?
35 2011-04-05T00:32:54 <sinha> but as i see there use to be a div tag around each object
36 2011-04-05T00:33:00 <sinha> yes
37 2011-04-05T00:33:05 <sinha> on the content page
38 2011-04-05T00:33:13 <dreimark> gn
39 2011-04-05T00:33:26 <ThomasWaldmann> can be done, not a big issue
40 2011-04-05T00:34:11 <sinha> okay i guess forst this argument passing things need to be handled
41 2011-04-05T00:34:14 <sinha> *first
42 2011-04-05T00:34:27 <sinha> correct parsing of width/height
43 2011-04-05T00:35:00 <ThomasWaldmann> that's unrelated to the show view
44 2011-04-05T00:35:51 <sinha> you said this thing {{png||w=100 }} not working
45 2011-04-05T00:35:53 <sinha> now
46 2011-04-05T00:39:53 <ThomasWaldmann> &w
47 2011-04-05T00:40:36 <ThomasWaldmann> (that makes w go into the url args, otherwise it'll go into the tag)
48 2011-04-05T00:42:09 <sinha> okay as not seeing any change now even in the resulting html tag,maybe libjpeg issue
49 2011-04-05T00:42:31 <ThomasWaldmann> you can try with a png
50 2011-04-05T00:48:19 <sinha> working fine with png for opening via url ,but not in embedding i tried {{ PNG ||&w=100 }} and {{ PNG || w=100 }}
51 2011-04-05T00:54:04 <ThomasWaldmann> we already discussed that arg parsing seems broken
52 2011-04-05T00:57:56 *** m4k3r
53 2011-04-05T01:05:03 *** not-xjjk
54 2011-04-05T01:09:55 *** sinha
55 2011-04-05T02:56:11 *** kasun
56 2011-04-05T03:12:51 *** yufra
57 2011-04-05T04:43:10 *** Marchael
58 2011-04-05T05:05:34 *** ronny
59 2011-04-05T05:06:03 *** ronny
60 2011-04-05T05:23:56 *** kursor
61 2011-04-05T06:04:30 *** Marchael
62 2011-04-05T06:29:33 <sl33k_> dreimark: yea. what do i do now?
63 2011-04-05T06:58:42 *** Marchael
64 2011-04-05T07:07:28 *** Marchael
65 2011-04-05T07:38:49 *** andialbrecht_
66 2011-04-05T07:58:14 *** sl33k_
67 2011-04-05T08:14:12 *** sl33k_
68 2011-04-05T08:18:39 *** sl33k_
69 2011-04-05T08:18:39 *** sl33k_
70 2011-04-05T09:04:28 *** greg_f
71 2011-04-05T09:33:22 <ThomasWaldmann> moin
72 2011-04-05T09:37:30 *** andialbrecht
73 2011-04-05T09:45:09 *** cemycc
74 2011-04-05T09:45:32 <cemycc> Hello
75 2011-04-05T09:45:56 <cemycc> I am a GSoC student, want to apply for a project. I still have time to do it ?
76 2011-04-05T09:46:30 <ThomasWaldmann> yes, read the timeline
77 2011-04-05T10:01:40 <cemycc> I will want to apply for the first User interface project
78 2011-04-05T10:01:54 <cemycc> Can you please give me some hints about what I can write on my wiki ?
79 2011-04-05T10:02:04 <cemycc> wiki page*
80 2011-04-05T10:03:42 <dreimark> sl33k_: you should write a msg, and avoid to give an empty results list
81 2011-04-05T10:04:54 <dreimark> e.g. you can check if a logged in user calls it and only then it creates a list
82 2011-04-05T10:04:58 <dreimark> w
83 2011-04-05T10:05:17 <ThomasWaldmann> cemycc: well, it is expected that you write your own proposal. after installing moin2 and playing with it, some ideas should be rather obvious if you are into UI stuff.
84 2011-04-05T10:09:13 <ThomasWaldmann> (you may, but you are not require to write your proposal on a wiki page. all what counts is the application in melange.)
85 2011-04-05T10:13:52 <cemycc> Ok then
86 2011-04-05T10:14:58 <ThomasWaldmann> cemycc: as the deadline is in a few days, you'll have lots to do now due to starting so late
87 2011-04-05T10:16:15 <ThomasWaldmann> is this your first SOC?
88 2011-04-05T10:16:38 <cemycc> ThomasWaldmann: Yes, I know - I will try to do my best in this days
89 2011-04-05T10:16:40 <cemycc> Yes
90 2011-04-05T10:17:16 <cemycc> A teacher from faculty inform me about GSoC.
91 2011-04-05T10:17:42 <ThomasWaldmann> you need to read slashdot.org and similar media :)
92 2011-04-05T10:18:19 <ThomasWaldmann> (not right now, but in general :)
93 2011-04-05T10:19:55 <ThomasWaldmann> cemycc: with "first ui project" you mean the flatland/jinja2 stuff?
94 2011-04-05T10:20:17 <cemycc> Yep
95 2011-04-05T10:21:04 <ThomasWaldmann> did you work already with flatland or jinja2?
96 2011-04-05T10:21:29 <cemycc> i haven't work with flatland or jinja2 but I have work with django
97 2011-04-05T10:21:45 <cemycc> I was reading some tutorials about flatland is easy to understand
98 2011-04-05T10:22:57 <cemycc> and about jinja2 he is just a template engine :) Not so hard to understand :)
99 2011-04-05T10:23:14 <ThomasWaldmann> ok, I suggest you install moin2 from a repo clone (see the moin2 docs) and have a look at MoinMoin.apps.frontend.views
100 2011-04-05T10:24:30 <cemycc> yep I was reading the wiki for devs right now . I will try to do my best this days
101 2011-04-05T10:24:59 <ThomasWaldmann> then look at user settings (esp. what's missing still), item revision metadata (how it is edited right now, how you can improve that by flatland and jinja2)
102 2011-04-05T10:27:29 <ThomasWaldmann> if you can provide some patch for moin to prove that you can work with our tools and have skills, that would be appreciated (as you are rather late, you can also do that after the application deadline)
103 2011-04-05T10:28:11 <ThomasWaldmann> if you can show some own (python) code, link to it from your application
104 2011-04-05T10:28:30 <cemycc> oh awesome, I will do it after application deadline
105 2011-04-05T10:29:35 <cemycc> I have only a chess application written in Python without AI . I will upload it on my shared hosting asap
106 2011-04-05T10:30:57 <ThomasWaldmann> do you use bitbucket.org?
107 2011-04-05T10:31:46 <cemycc> no, you want me to upload it there ?
108 2011-04-05T10:32:20 <ThomasWaldmann> if you want to contribute to moin, forking moin-2.0 there to your own clone of it is a good way
109 2011-04-05T10:33:07 <ThomasWaldmann> http://moinmo.in/MoinMoin2.0 there are some links and general infos
110 2011-04-05T10:33:19 <ThomasWaldmann> the moin 2 docs are also linked from there
111 2011-04-05T10:33:52 <cemycc> I haven't used because I don't have so much projects: I was learning python, after that working on this application, trying to implement AI but without success and after that I was starting to learn Django.
112 2011-04-05T10:35:15 <cemycc> I don't have any official application written in Django, but I can handle because in the past I was working allot with PHP and frameworks.
113 2011-04-05T10:35:24 <cemycc> So is not so hard for me to learn Flask for example
114 2011-04-05T10:36:01 <cemycc> I have seen that you use it on Moin 2
115 2011-04-05T10:36:34 <ThomasWaldmann> well, you are in good company :D many people tried to "implement AI" without success. ;)
116 2011-04-05T10:38:04 <cemycc> yea, i will stick to web development. :D
117 2011-04-05T10:39:15 <ThomasWaldmann> yes. but for the widgets project you maybe won't have to do much with flask stuff. rather with flatland, jinja2, html, css.
118 2011-04-05T10:46:28 <cemycc> Yes
119 2011-04-05T10:49:52 <cemycc> I have another question about GSoC. In their FAQ says: Google allocates a particular number of student slots to each organization.
120 2011-04-05T10:50:05 <cemycc> So every project have his priority ?
121 2011-04-05T10:54:37 <ThomasWaldmann> we don't know our slot count yet
122 2011-04-05T10:55:09 <ThomasWaldmann> but student projects will be ranked and we'll take the topmost N at the end
123 2011-04-05T11:04:38 <CIA-25> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 164:ed61511b9ec2 2.0/docs/admin/install.rst: install docs: installing pil with jpeg support
124 2011-04-05T11:09:47 <cemycc> Ok then
125 2011-04-05T11:24:30 <sl33k_> dreimark: okay i would need an else or elif clause there. what could go in there?
126 2011-04-05T11:31:32 *** kursor
127 2011-04-05T11:37:33 *** Marchael
128 2011-04-05T11:56:11 *** ronny
129 2011-04-05T12:23:37 *** Marchael
130 2011-04-05T12:56:11 <dreimark> sl33k_: what do you think?
131 2011-04-05T13:00:11 *** ronny
132 2011-04-05T13:41:45 *** kasun
133 2011-04-05T13:54:05 <cemycc> ThomasWaldmann: How are widgets handled in moin2 ? In moin-1.9.3 I have seen a widget system.
134 2011-04-05T13:55:45 <ThomasWaldmann> there are a few jinja2 macros for flatland, but nothing more yet
135 2011-04-05T14:00:11 <cemycc> then I will need to implement a system that will need to handle more types of widgets
136 2011-04-05T14:02:26 <ThomasWaldmann> or just call the macros (including your new ones) from the j2 templates
137 2011-04-05T14:03:17 <ThomasWaldmann> the old widget system was not too popular btw
138 2011-04-05T14:10:54 *** andialbrecht_
139 2011-04-05T14:12:22 <cemycc> ok then so my first improvements will be for metadata editing and for user settings validation . Then I will come with my ideas ?
140 2011-04-05T14:13:11 <cemycc> Also, I will like to work with Ajax for example at validation or at metadata editing
141 2011-04-05T14:14:00 <cemycc> first I will try to understand the macros
142 2011-04-05T14:17:57 <ThomasWaldmann> maybe ajax won't be needed, at least not initially
143 2011-04-05T14:18:08 <ThomasWaldmann> jek: ^^
144 2011-04-05T14:27:41 *** m4k3r
145 2011-04-05T14:35:48 *** m4k3r
146 2011-04-05T14:43:15 <Marchael> ThomasWaldmann: I designing one simple whoosh`s schema which can be using for indexing data http://paste.pocoo.org/show/365861/
147 2011-04-05T14:43:54 <Marchael> there is one simple example which describes the whoosh schema and add document to it
148 2011-04-05T14:44:42 <Marchael> I thought what it will be better if revisions will be storing together with item data
149 2011-04-05T14:45:48 <Marchael> Weak point of this schema what I don`t know current revision of wiki`s page
150 2011-04-05T14:46:01 <Marchael> but it is fixible
151 2011-04-05T14:48:46 <Marchael> TheSheep: hi
152 2011-04-05T14:49:00 <Marchael> can you looking for link above?
153 2011-04-05T15:12:49 *** cemycc
154 2011-04-05T15:14:44 <Marchael> there is example with current revision http://paste.pocoo.org/show/365881/
155 2011-04-05T15:15:47 <Marchael> $ ./test_index.py
156 2011-04-05T15:15:47 <Marchael> 1 <Top 1 Results for Term('current', u't') runtime=0.000104904174805>
157 2011-04-05T15:16:05 <Marchael> not so bad
158 2011-04-05T15:30:52 <ThomasWaldmann> Marchael: that might be a bit too simple / incomplete
159 2011-04-05T15:33:39 <ThomasWaldmann> maybe think about what users usually search for
160 2011-04-05T15:33:54 <ThomasWaldmann> also think about what moin internally may want to search for
161 2011-04-05T15:34:18 <ThomasWaldmann> think about how you update the index when a new revision gets saved
162 2011-04-05T15:34:59 <ThomasWaldmann> think about what meta items are most important
163 2011-04-05T15:35:12 <ThomasWaldmann> think about how to deal with arbitrary metadata
164 2011-04-05T15:36:24 <ThomasWaldmann> btw, if you see something like uuid=itemname that is just a temporary thing, in the end, the uuid is of course NOT the item name
165 2011-04-05T15:36:44 <ThomasWaldmann> but just a big random number or hex string
166 2011-04-05T15:38:51 <ThomasWaldmann> (of course you don't need to code that now, just develop some vision what could be done and what not :)
167 2011-04-05T15:40:58 <Marchael> > think about how you update the index when a new revision gets saved
168 2011-04-05T15:40:58 <Marchael> They are always updating automatically when new item commit
169 2011-04-05T15:40:58 <Marchael> > think about what meta items are most important
170 2011-04-05T15:41:17 <Marchael> oops
171 2011-04-05T15:41:46 <Marchael> flooding control :)
172 2011-04-05T15:41:54 <Marchael> > think about how you update the index when a new revision gets saved
173 2011-04-05T15:41:54 <Marchael> They are always updating automatically when new item commit
174 2011-04-05T15:42:00 <Marchael> > think about what meta items are most important
175 2011-04-05T15:42:00 <Marchael> Why? I can use `meta` field for all metas
176 2011-04-05T15:42:23 <Marchael> sorry, if I repeated my previous messages
177 2011-04-05T15:43:32 <Marchael> > think about how to deal with arbitrary metadata
178 2011-04-05T15:43:32 <Marchael> I`m not quite understand what you mean.
179 2011-04-05T15:44:52 <ThomasWaldmann> well, if you mix all metadata together and put that into one field, the search might be not that efficient
180 2011-04-05T15:46:34 <Marchael> searcher.search(query) returns all fields then I can parse them for metadata
181 2011-04-05T15:47:51 <Marchael> for example $ ./test_index.py
182 2011-04-05T15:47:51 <Marchael> 1 <Hit {'mimetype': u'text/x.moin.wiki', 'rev_no': 1, 'metas': u'foo:bar, key:val', 'acl': u'u:rwx,g:rwx,o:r-x', 'tags': u'test moin new'}>
183 2011-04-05T15:49:59 <Marchael> if I have more than one object matching search criterian then I can limit output
184 2011-04-05T15:50:15 <Marchael> s/criterian/criteria/
185 2011-04-05T15:51:00 <ThomasWaldmann> if you know what you search for, you don't want to get lots of results and postprocess them, you just want to get what you want as result of your query
186 2011-04-05T15:52:29 <Marchael> So then I will be using searching by document`s ID
187 2011-04-05T15:52:59 <Marchael> and there will be only one object what matching this criteria
188 2011-04-05T15:53:54 <ThomasWaldmann> show how you'ld search for a part of an item name
189 2011-04-05T15:54:58 <Marchael> query = QueryParser("uuid", ix.schema).parse(u"Help*")
190 2011-04-05T15:54:59 <Marchael> results = searcher.search(query)
191 2011-04-05T15:55:28 <ThomasWaldmann> do you read what i write? :)
192 2011-04-05T15:55:46 <ThomasWaldmann> (15:36) < ThomasWal> btw, if you see something like uuid=itemname that is just a temporary thing, in the end, the uuid is of course NOT the item name
193 2011-04-05T15:56:27 <Marchael> oh, sorry :)
194 2011-04-05T15:57:04 <Marchael> So I`m will add item_name field to shema
195 2011-04-05T15:58:33 <ThomasWaldmann> how do you search for a word in the revision data?
196 2011-04-05T15:59:19 <ThomasWaldmann> why is mimetype TEXT?
197 2011-04-05T15:59:22 <ThomasWaldmann> ...
198 2011-04-05T16:03:16 <Marchael> hm, I don`t have `content` field so I will add this
199 2011-04-05T16:03:32 <Marchael> > why is mimetype TEXT?
200 2011-04-05T16:03:32 <Marchael> is it bad?
201 2011-04-05T16:03:59 <dreimark> moin has different mimetypes
202 2011-04-05T16:04:20 <dreimark> but text sounds like a major one and not like any defined
203 2011-04-05T16:06:33 <Marchael> no, type TEXT using only for specifing searching, indexing and storing attributes
204 2011-04-05T16:07:13 <ThomasWaldmann> TEXT is maybe a bit too much for mimetype
205 2011-04-05T16:07:15 <Marchael> but, I don`t think what users often wil searching for metadata
206 2011-04-05T16:07:38 <ThomasWaldmann> Marchael: sure they will :)
207 2011-04-05T16:07:47 <Marchael> so may be better change TEXT to STORED
208 2011-04-05T16:08:08 <Marchael> mm
209 2011-04-05T16:08:19 <ThomasWaldmann> e.g. if you are interested in text items only (not in pictures, binaries, whatever), you'll use something like mimetype:text
210 2011-04-05T16:09:37 <ThomasWaldmann> otherwise, if you search for FooBar, it'll return FooBar, FooBar/foo.pdf, FooBar/bar.png, FooBar/baz.mp3
211 2011-04-05T16:10:22 <ThomasWaldmann> i think you'll need some special Analyzers at least
212 2011-04-05T16:10:29 <Marchael> mimetype can also be searching criteria
213 2011-04-05T16:11:09 <Marchael> I`m not quite understand why TEXT is really bad for `mimetype`
214 2011-04-05T16:11:22 <ThomasWaldmann> but, you're usually not interested in word frequency or position for the mimetype string
215 2011-04-05T16:11:50 <ThomasWaldmann> it does too much. and maybe the standard analyzer is no good, not sure.
216 2011-04-05T16:15:33 <Marchael> hm, I`m need some time to thinking ^)
217 2011-04-05T16:15:35 <Marchael> :)
218 2011-04-05T16:20:56 <ThomasWaldmann> also thing about that "current" field. how do you want to deal with it?
219 2011-04-05T16:21:00 <ThomasWaldmann> think*
220 2011-04-05T16:23:02 <Marchael> What you mean? If I`m need current revision of wiki`s page i just using query like query = QueryParser("current", ix.schema).parse(u"1") and get current revision
221 2011-04-05T16:23:28 <ThomasWaldmann> what do you do if you save a new revision?
222 2011-04-05T16:23:53 <Marchael> add it as new document to existing schema
223 2011-04-05T16:25:03 <ThomasWaldmann> and then you have 2x current? :)
224 2011-04-05T16:25:09 <Marchael> but I`m not sure about how it will affect to schema size
225 2011-04-05T16:25:18 <Marchael> hm
226 2011-04-05T16:25:23 <Marchael> no ofcorse
227 2011-04-05T16:25:37 <Marchael> I can set Stored=False
228 2011-04-05T16:25:53 <Marchael> but I`m not sure
229 2011-04-05T16:26:09 <ThomasWaldmann> that doesn't fix your problem
230 2011-04-05T16:27:15 <Marchael> or storing not full copy with new changes, but only diff between new revision and previous
231 2011-04-05T16:28:48 <Marchael> And also why I need storing page`s text to whoosh schema, it is storing into moin backends...
232 2011-04-05T16:29:25 *** ronny
233 2011-04-05T16:29:51 *** andialbrecht
234 2011-04-05T16:29:59 *** ronny
235 2011-04-05T16:32:35 <ThomasWaldmann> better forget about diffs
236 2011-04-05T16:34:03 *** RogerHaase
237 2011-04-05T16:35:38 <RogerHaase> I have a problem running tests on Ubuntu 10.10, AMD X2, 2 GB memory.
238 2011-04-05T16:36:37 <RogerHaase> After starting, system monitor shows memory steadily increasing to 98%, then swap memory starts steady increase.
239 2011-04-05T16:37:22 *** Marchael
240 2011-04-05T16:37:30 *** Marchael
241 2011-04-05T16:38:00 <RogerHaase> System becomes unusable (sluggish). I cancelled py.test after 40 minutes, memory does not free up until I close terminal window.
242 2011-04-05T16:39:10 *** sinha
243 2011-04-05T16:39:52 <RogerHaase> Anyone else see similar issue running tests on Ubuntu?
244 2011-04-05T16:41:02 <ThomasWaldmann> i have same ubuntu, but 4gigs, works for me
245 2011-04-05T16:41:24 <ThomasWaldmann> RogerHaase: does this happen with moin-2.0 tip?
246 2011-04-05T16:42:00 <RogerHaase> I have 4 GB on windows, have not noticed it there either. Yes I am at moin 2.0 tip, Ubuntu is current as well.
247 2011-04-05T16:42:19 <ThomasWaldmann> you just run py.test from the toplevel dir?
248 2011-04-05T16:43:36 <RogerHaase> From top level: source env/bin/activate
249 2011-04-05T16:43:36 <RogerHaase> py.test > pytest2.txt
250 2011-04-05T16:44:18 <ThomasWaldmann> yes, it eats memory :|
251 2011-04-05T16:45:28 *** m4k3r
252 2011-04-05T16:45:53 *** sinha
253 2011-04-05T16:46:00 <RogerHaase> I tried googling for pytest memory leak, but did not find anything.
254 2011-04-05T16:46:10 *** m4k3r
255 2011-04-05T16:46:23 *** m4k3r
256 2011-04-05T16:46:27 <ThomasWaldmann> google for moin2 py.test memory leak :)
257 2011-04-05T16:48:24 <RogerHaase> I will try again tomorrow :-) maybe I will have better luck.
258 2011-04-05T16:48:47 * ThomasWaldmann was just joking
259 2011-04-05T16:49:21 <RogerHaase> me too :-)
260 2011-04-05T16:49:53 <ThomasWaldmann> i think the problem is in conftest.py
261 2011-04-05T16:53:29 <ThomasWaldmann> it needs ~1.5GB memory to run all the tests
262 2011-04-05T16:53:45 <Marchael> ThomasWaldmann: can you point me to documentation about storaging documents in moin2?
263 2011-04-05T16:53:51 <ThomasWaldmann> (on 32bit system)
264 2011-04-05T16:54:33 <ThomasWaldmann> Marchael: like api docs for MoinMoin.storage?
265 2011-04-05T16:56:38 <RogerHaase> I have 64 bit, I will try closing off everything else and see how far I can get. back later.
266 2011-04-05T16:56:52 <Marchael> mmm, no quite. Just describing how documents reach the storage
267 2011-04-05T16:57:18 <ThomasWaldmann> RogerHaase: as 64bit consumes more, that'll likely eat 2GB then
268 2011-04-05T16:57:39 <Marchael> all stages from creating in editor to storing in backend
269 2011-04-05T16:58:52 <RogerHaase> Yes, I am expecting that it will require more than 2 GB I will probably just buy more memory if that is what it needs.
270 2011-04-05T17:03:03 <ThomasWaldmann> of course fixing the code would be better, but I currently only have some suspicions, but no solution
271 2011-04-05T17:04:06 *** RogerHaase
272 2011-04-05T17:04:09 <ThomasWaldmann> it is likely conftest.py and what it does (does not) in init_test_app and deinit_test_app, and also the nesting of that for module, class, method
273 2011-04-05T17:04:53 <Marchael> ThomasWaldmann: ^
274 2011-04-05T17:05:22 <Marchael> I searching for docs, but all what I see it`s api docs...
275 2011-04-05T17:05:28 <ThomasWaldmann> Marchael: you likely will have to read some code or api docs for that. it starts in MoinMoin.apps.frontend.views
276 2011-04-05T17:05:37 <Marchael> ok
277 2011-04-05T17:09:44 *** kursor
278 2011-04-05T17:46:13 *** sinha
279 2011-04-05T18:15:28 *** andialbrecht_
280 2011-04-05T18:15:32 *** andialbrecht_
281 2011-04-05T18:16:20 *** xorAxAx
282 2011-04-05T18:17:00 *** xorAxAx
283 2011-04-05T18:18:38 *** Marchael
284 2011-04-05T18:21:39 <CIA-25> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 165:76430561b75b 2.0/docs/ (admin/backup.rst devel/support.rst): docs: add backup/restore docs
285 2011-04-05T18:35:24 *** waldi
286 2011-04-05T18:37:59 *** dreimark
287 2011-04-05T18:38:11 *** waldi
288 2011-04-05T18:38:32 *** dreimark
289 2011-04-05T18:50:30 *** chickenzilla
290 2011-04-05T18:50:54 *** z^ck
291 2011-04-05T18:51:22 *** dreimark
292 2011-04-05T18:51:22 *** dreimark
293 2011-04-05T18:51:28 <dreimark> bbl
294 2011-04-05T18:59:55 *** cemycc
295 2011-04-05T19:06:50 *** Marchael
296 2011-04-05T19:17:52 *** grzywacz
297 2011-04-05T19:30:12 *** greg_f
298 2011-04-05T19:38:40 *** RogerHaase
299 2011-04-05T19:43:52 <RogerHaase> ThomasWaldmann: FYI: == 51 failed, 2106 passed, 247 skipped, 6 error, 7 xpassed in 2648.54 seconds == Although this message came out after 44 minutes, it took another 20 minutes of heavy disk activity for pytest to end and release memory. Swap memory went up to 1.3 GB from a few hundred MB before the test was started.
300 2011-04-05T19:48:22 <jek> whoa
301 2011-04-05T20:08:26 <dreimark> i also have a similiar machine
302 2011-04-05T20:08:31 <dreimark> 64bit
303 2011-04-05T20:08:35 <dreimark> RogerHaase: ^
304 2011-04-05T20:10:01 <dreimark> top should show the amount it costs
305 2011-04-05T20:12:43 *** kursor
306 2011-04-05T20:13:06 <dreimark> 52 failed, 2241 passed, 246 skipped, 36 error, 8 xpassed in 279.28 seconds
307 2011-04-05T20:13:18 <dreimark> 2.3GB allocated
308 2011-04-05T20:14:56 <dreimark> we should easily find out which kind of test needs that
309 2011-04-05T20:17:33 <dreimark> e.g. py.test -k storage seems to need a constant amount of 182m without increasing
310 2011-04-05T20:18:10 <dreimark> haha, stopped top
311 2011-04-05T20:18:21 <dreimark> it increases
312 2011-04-05T20:29:13 <RogerHaase> dreimark: memory increases slowly throughout run for me.
313 2011-04-05T20:40:05 *** franklin
314 2011-04-05T20:59:13 *** Marchael
315 2011-04-05T21:09:27 <ThomasWaldmann> RogerHaase: there is a bug for that
316 2011-04-05T21:14:00 <RogerHaase> gtg
317 2011-04-05T21:14:06 *** RogerHaase
318 2011-04-05T21:19:41 *** m4k3r
319 2011-04-05T21:27:06 *** m4k3r
320 2011-04-05T21:28:47 * dreimark runs tests with fs2 backend
321 2011-04-05T21:31:05 <ThomasWaldmann> btw, good news about memory usage: moin2 seems to constantly use ~24MB (tested with ab and melae)
322 2011-04-05T21:31:51 <dreimark> i guess we have a problem with the memory backend
323 2011-04-05T21:45:34 <dreimark> fs2 backend results are very strange http://paste.pocoo.org/show/366117/
324 2011-04-05T22:04:02 *** hardy89
325 2011-04-05T22:13:07 * dreimark fixes test_user
326 2011-04-05T22:13:20 *** hardy89
327 2011-04-05T22:13:54 <dreimark> for whatever reason that test did not fail with mem backend
328 2011-04-05T22:14:04 *** grzywacz
329 2011-04-05T22:32:18 *** ronny
330 2011-04-05T22:41:18 <ThomasWaldmann> some tests might depend on a clean backend
331 2011-04-05T22:41:48 <dreimark> after fixing the unicode bug the tests are running
332 2011-04-05T22:41:57 <dreimark> but slooooow
333 2011-04-05T22:42:40 <dreimark> currently at test_backends and 800MB
334 2011-04-05T22:43:35 <dreimark> all the time ROUTER_index.sqlite-journal and ROUTER_index.sqlite
335 2011-04-05T22:43:41 <dreimark> become rebuilded
336 2011-04-05T22:45:59 <ThomasWaldmann> it is likely unrelated to the backend. i suspect either router_index or general app cleanup.
337 2011-04-05T22:46:07 <ThomasWaldmann> (as i wrote in the bug)
338 2011-04-05T22:46:15 *** ronny
339 2011-04-05T22:47:05 <dreimark> found a new problem
340 2011-04-05T22:47:21 <dreimark> OSError: [Errno 24] Too many open files: '/dev/null'
341 2011-04-05T22:48:11 <dreimark> http://paste.pocoo.org/show/366149/
342 2011-04-05T22:51:02 <dreimark> ValueError: cannot resume, already capturing with 'fd
343 2011-04-05T22:53:01 <ThomasWaldmann> out of files is likely related to out of memory, same root cause
344 2011-04-05T22:53:43 <dreimark> yes but I was not out of memory
345 2011-04-05T22:54:27 <ThomasWaldmann> not yet :)
346 2011-04-05T22:54:44 <dreimark> it crashed already at 800MB
347 2011-04-05T22:54:52 <ThomasWaldmann> btw, meliae has internal corruption problems, the json dumps it generates are invalid
348 2011-04-05T22:55:52 <dreimark> i wonder a bit that this did not happen at all in the memory backend.
349 2011-04-05T22:56:04 <sinha> dreimark: hi,that tree view for index is working for almost all chars but failing when someone enter quotes in the item name ?
350 2011-04-05T22:56:22 <sinha> what should be done ,is this needed to be corrected
351 2011-04-05T22:56:24 <dreimark> sinha: quotes must be escaped
352 2011-04-05T22:56:38 <dreimark> in treeview
353 2011-04-05T22:57:15 <sinha> okay,this kind of item name is completely valid-> Tre;;;'" ?
354 2011-04-05T22:57:40 *** m4k3r
355 2011-04-05T22:58:33 <dreimark> it is
356 2011-04-05T22:58:50 <dreimark> and look at the html source of the item
357 2011-04-05T22:59:03 <dreimark> >Tre;;;'"
358 2011-04-05T23:01:10 <sinha> yes
359 2011-04-05T23:01:25 <sinha> now i got it working
360 2011-04-05T23:02:32 <sinha> only character is " ,giving problem
361 2011-04-05T23:02:55 <sinha> i guess there would be any filer in jinja which encodes all these characters ?
362 2011-04-05T23:03:05 <sinha> as url encoding is done
363 2011-04-05T23:03:29 <dreimark> there is a raw feature
364 2011-04-05T23:04:55 <sinha> okay i will try that one and make it work for every character
365 2011-04-05T23:05:19 <sinha> dreimark: and also i have submitted my proposal at google-melange site
366 2011-04-05T23:05:33 <ronny> sup
367 2011-04-05T23:05:37 <dreimark> :)
368 2011-04-05T23:06:04 <dreimark> thx for notifying I haven't got an email
369 2011-04-05T23:06:05 <ronny> a guy with the nick pkumar just reminded me of the lack of my pytest2 fixes
370 2011-04-05T23:06:16 <ronny> he wound up in #pylib
371 2011-04-05T23:06:42 <dreimark> ronny: http://paste.pocoo.org/show/366149/
372 2011-04-05T23:06:43 <ronny> is he in #moin? her missing here
373 2011-04-05T23:07:03 *** pkumar
374 2011-04-05T23:07:08 <ronny> ah :)
375 2011-04-05T23:07:10 <dreimark> ronny: ^
376 2011-04-05T23:07:42 <ronny> i'll need some time to sort out the patches and apply them on the new history, i never got that finished
377 2011-04-05T23:08:02 <ronny> i'll make a named branch
378 2011-04-05T23:08:20 <ronny> since just fixing running on pytest2 will not fix the tests
379 2011-04-05T23:08:29 <ronny> and the generative tests will actually make it hard
380 2011-04-05T23:08:50 <dreimark> true and anything else than a memory backend is slooooow
381 2011-04-05T23:09:59 <dreimark> (for doing all the tests)
382 2011-04-05T23:10:36 <pkumar> Since a lot of changes have been made in pytest, after py 0.9.1
383 2011-04-05T23:10:48 <ronny> pkumar: pytest up to 1.2 works fine
384 2011-04-05T23:11:16 <ronny> pkumar: the pytest.py file in moinmoin is from the release before pytest 2.0
385 2011-04-05T23:11:23 <ronny> (and it works)
386 2011-04-05T23:12:05 <ThomasWaldmann> are we talking about py.test in moin 1.9 or moin2 (SOC)?
387 2011-04-05T23:13:17 <ThomasWaldmann> 'py==1.3.4', # py.test 1.3.4 is needed by unit tests
388 2011-04-05T23:13:19 *** sinha
389 2011-04-05T23:13:26 <ThomasWaldmann> that's what we have in moin2 setup.py
390 2011-04-05T23:14:05 <ronny> ThomasWaldmann: not sure
391 2011-04-05T23:14:16 <ronny> pkumar: what exacly are you working on?
392 2011-04-05T23:14:24 <pkumar> moin 1.9
393 2011-04-05T23:14:41 <ronny> pkumar: then porting to a new pytest version is pointless
394 2011-04-05T23:14:59 <ronny> 1.9 is in maintaince mode
395 2011-04-05T23:15:15 <ThomasWaldmann> pkumar: are you a potential SOC student?
396 2011-04-05T23:15:39 <pkumar> yes.
397 2011-04-05T23:15:52 <ThomasWaldmann> all projects are for moin2
398 2011-04-05T23:16:08 <pkumar> ok.
399 2011-04-05T23:16:25 <ronny> ok, i'll finish getting my patches over, i'll put them into a named branch
400 2011-04-05T23:16:40 <dreimark> pkumar: http://static.moinmo.in/m2/html/
401 2011-04-05T23:17:38 <dreimark> http://moinmo.in/MoinMoin2.0
402 2011-04-05T23:19:00 <pkumar> dreimark:thanks.
403 2011-04-05T23:20:32 <pkumar> ronny: It would be great if you could send me link of patches.
404 2011-04-05T23:20:56 <pkumar> my id: contactprashantat@gmail.com
405 2011-04-05T23:24:35 <ronny> pkumar: im applying them to my clone of moin2
406 2011-04-05T23:24:51 <ronny> pkumar: its 18 patches, will take some time to sort them out
407 2011-04-05T23:24:58 <ronny> in particular wrt the branches
408 2011-04-05T23:26:42 <pkumar> I would start working on moin2 and try to fix it to run on pytest2
409 2011-04-05T23:26:52 <pkumar> to begin with.
410 2011-04-05T23:33:12 <ThomasWaldmann> pkumar: if you want to do a SOC project on the tests topic, primarily work on your application now
411 2011-04-05T23:33:28 <ThomasWaldmann> it is not much time left and you should know what you want to do :)
412 2011-04-05T23:34:44 <pkumar> Though I was working on moin 1.9 it gave me familiarity with MoinMoin
413 2011-04-05T23:34:55 <pkumar> and test codes
414 2011-04-05T23:35:25 <dreimark> pkumar: a clone of moin-2 is only 1.6 MB
415 2011-04-05T23:35:31 <dreimark> you should look in that
416 2011-04-05T23:35:39 <dreimark> because the code is much better
417 2011-04-05T23:35:56 <dreimark> and smaller
418 2011-04-05T23:37:57 <CIA-25> Reimar Bauer <rb.proj AT googlemail DOT com> default * 166:9c4fdb103498 2.0/MoinMoin/_tests/test_user.py: test_user: fix failures related to kvstore type checking, unicode vs. str
419 2011-04-05T23:39:30 <pkumar> ThomasWaldmann: Thanks, I'm also working on the application currently.
420 2011-04-05T23:45:53 <ronny> pkumar: ok, i just added my stuff on https://bitbucket.org/RonnyPfannschmidt/moin-2.0/ its a bit unordered, but something one can start to work with
421 2011-04-05T23:46:31 <ronny> it uses a extra named branch, so no confusion comes up, and one cant screw up default by accident even if its pushed
422 2011-04-05T23:47:00 <pkumar> ronny:thanks, It would be of great help.
423 2011-04-05T23:47:26 <ronny> pkumar: bascially, run the quickinstall script, then see whats broken, there is a lot
424 2011-04-05T23:50:37 <ronny> (use env/bin/py.test)
425 2011-04-05T23:51:29 <pkumar> ok.
426 2011-04-05T23:54:16 <ThomasWaldmann> env/bin is in path if you have activated the env
427 2011-04-05T23:54:53 <ronny> ok, i'll grab myself a beer and read a manga, since im tired from the martial arts training, expect higher lags from me since i'll only look once every chapeter
428 2011-04-05T23:57:21 <pkumar> ThomasWaldmann: ok, got it.
