1 2011-06-12T00:11:00 *** sinha
2 2011-06-12T00:26:58 <ronny> re
3 2011-06-12T00:34:42 <CIA-69> Michael Mayorov <marchael@kb.csu.ru> * e6fb11f35a84 r265 moin-2.0/MoinMoin/search/indexing.py: Added wikiname/userid/address/hostname/ to whoosh schema
4 2011-06-12T00:45:04 *** Marchael
5 2011-06-12T01:43:04 *** m4k3r_
6 2011-06-12T07:47:53 *** MattMaker
7 2011-06-12T08:54:43 *** m4k3r
8 2011-06-12T09:32:03 *** moinBot`
9 2011-06-12T09:39:32 *** m4k3r
10 2011-06-12T09:39:38 *** dreimark
11 2011-06-12T09:39:38 *** jek
12 2011-06-12T09:39:38 *** eSyr
13 2011-06-12T09:39:40 *** aa_
14 2011-06-12T09:39:40 *** CIA-69
15 2011-06-12T09:39:41 *** franklin
16 2011-06-12T09:39:41 *** andialbrecht_
17 2011-06-12T09:39:42 *** ThomasWaldmann
18 2011-06-12T09:39:44 *** vpv
19 2011-06-12T09:39:44 *** ronny
20 2011-06-12T09:39:44 *** xjjk
21 2011-06-12T09:39:46 *** xorAxAx
22 2011-06-12T09:39:47 *** TheSheep
23 2011-06-12T09:39:47 *** ChanServ
24 2011-06-12T09:39:48 *** waldi
25 2011-06-12T09:39:49 *** moinBot
26 2011-06-12T09:46:30 *** ChanServ
27 2011-06-12T09:46:31 *** TheSheep
28 2011-06-12T09:46:31 *** vpv
29 2011-06-12T09:46:31 *** ThomasWaldmann
30 2011-06-12T09:46:31 *** andialbrecht_
31 2011-06-12T09:46:31 *** aa_
32 2011-06-12T09:46:31 *** CIA-69
33 2011-06-12T09:46:31 *** xorAxAx
34 2011-06-12T09:46:31 *** franklin
35 2011-06-12T09:46:31 *** xjjk
36 2011-06-12T09:46:31 *** ronny
37 2011-06-12T09:46:31 *** Marchael
38 2011-06-12T09:46:31 *** eSyr
39 2011-06-12T09:46:31 *** jek
40 2011-06-12T09:46:31 *** dreimark
41 2011-06-12T09:46:31 *** waldi
42 2011-06-12T09:46:31 *** m4k3r
43 2011-06-12T09:46:31 *** adams.freenode.net
44 2011-06-12T09:47:07 *** waldi
45 2011-06-12T09:47:07 *** waldi
46 2011-06-12T09:47:10 <Marchael> moin
47 2011-06-12T09:47:10 *** waldi
48 2011-06-12T09:47:11 *** dreimark
49 2011-06-12T09:47:11 *** jek
50 2011-06-12T09:47:11 *** eSyr
51 2011-06-12T09:51:45 *** waldi
52 2011-06-12T09:51:45 *** dreimark
53 2011-06-12T09:51:45 *** jek
54 2011-06-12T09:51:45 *** eSyr
55 2011-06-12T10:13:51 <Marchael> ThomasWaldmann: fixes in test_indexing.py http://paste.pocoo.org/show/404848/
56 2011-06-12T12:14:32 <dreimark> moin
57 2011-06-12T12:21:19 *** ronny
58 2011-06-12T12:22:22 *** ronny
59 2011-06-12T12:36:34 <ThomasWaldmann> Marchael: why hostname u"None"?
60 2011-06-12T12:36:51 <ThomasWaldmann> and userids are NOT usernames
61 2011-06-12T12:37:35 <dreimark> http://speed.pypy.org/
62 2011-06-12T12:37:57 <Marchael> "None" because we can't know which hostname complies this ip
63 2011-06-12T12:38:27 <Marchael> what correct format for userid?
64 2011-06-12T12:39:12 <ThomasWaldmann> find out
65 2011-06-12T12:40:17 <ThomasWaldmann> and also find out what it stores when it can't do reverse lookup. i doubt it stores u"None" in that case.
66 2011-06-12T12:43:16 <ThomasWaldmann> dreimark: last time i tried (not too long ago), pypy was 3..7 times slower than cpython with moin2
67 2011-06-12T12:46:14 <Marchael> ThomasWaldmann: but where? I looking in test.moinmo.in and metadata of all documents haven't userid field and hostname field always resolve
68 2011-06-12T12:46:19 <Marchael> *resolved
69 2011-06-12T12:47:41 <ThomasWaldmann> usually you can look for the metadata key constant (all defined in MoinMoin.config) and see how it is used
70 2011-06-12T12:48:40 <ThomasWaldmann> and for finding a userid, you just need to practically use moin2 with a logged in user and save something
71 2011-06-12T12:51:58 <ThomasWaldmann> and also look how userid behaves if user is not logged in
72 2011-06-12T12:56:28 <Marchael> I see only log_reverse_dns_lookups constant, but how it can helps me?
73 2011-06-12T12:59:34 <ThomasWaldmann> look where it is used
74 2011-06-12T13:03:06 <Marchael> :returns: host dns name (unicode) or
75 2011-06-12T13:03:06 <Marchael> None (if lookup is disallowed or failed)
76 2011-06-12T13:03:41 <ThomasWaldmann> ok, now look what is does with the result and what ends up in metadata
77 2011-06-12T13:07:13 <Marchael> seems like just nothing
78 2011-06-12T13:07:32 <Marchael> hostname field shouldn't appeared then
79 2011-06-12T13:07:35 <Marchael> in metadata
80 2011-06-12T13:07:49 <dreimark> why don't we use long type ? result[MTIME] = int(long(result[MTIME] or 0) / 1000000) # convert usecs to secs
81 2011-06-12T13:08:14 <dreimark> metadata['bookmarks'] = [(interwiki, int(long(bookmark)/1000000))
82 2011-06-12T13:08:38 <dreimark> e.g. why not just haveing it as long
83 2011-06-12T13:10:20 <ThomasWaldmann> dreimark: int and long has been unified in python, so I thought int is less special
84 2011-06-12T13:13:53 <ThomasWaldmann> Marchael: yes. but as you can't put just nothing into your whoosh field, you need to think about how to handle that sanely.
85 2011-06-12T13:14:32 <dreimark> ThomasWaldmann: a small difference is left over http://paste.pocoo.org/show/404909/
86 2011-06-12T13:15:20 <Marchael> ThomasWaldmann: it's not necessary what all fields of schema were filled.
87 2011-06-12T13:15:54 <Marchael> I mean documents
88 2011-06-12T13:16:10 <dreimark> size of the value seems to be the save in int and long
89 2011-06-12T13:16:11 <ThomasWaldmann> Marchael: that likely won't help you
90 2011-06-12T13:16:23 <dreimark> s/save/same/
91 2011-06-12T13:17:54 <dreimark> http://www.python.org/dev/peps/pep-0237/
92 2011-06-12T13:18:25 <Marchael> ThomasWaldmann: why? If I don't specify some fields in document, whoosh will be think what it just empty
93 2011-06-12T13:18:51 <dreimark> Long and short ints are handled different by the marshal module, and by the pickle and cPickle modules. This difference will remain (at least until Python 3.0).
94 2011-06-12T13:20:45 <ThomasWaldmann> Marchael: you'll see when you write that code
95 2011-06-12T13:23:29 <ThomasWaldmann> for your tests and now, maybe just use u'' in your to-be-indexed document when a String is not there in revision metadata
96 2011-06-12T13:24:44 <Marchael> hm, this code works without a failure http://paste.pocoo.org/show/404913/
97 2011-06-12T13:44:10 <ThomasWaldmann> ok.
98 2011-06-12T13:45:08 <ThomasWaldmann> (you removed some keys, you maybe can't do that in the real code later, but for the purpose of this test it maybe does not matter)
99 2011-06-12T13:45:48 <Marchael> ok, so can I commit?
100 2011-06-12T13:46:05 <ThomasWaldmann> yup
101 2011-06-12T13:47:39 <CIA-69> Michael Mayorov <marchael@kb.csu.ru> * 21860894d4dd r266 moin-2.0/MoinMoin/search/_tests/test_indexing.py: Rewrited some parts of test_indexing.py using new schemas
102 2011-06-12T13:49:30 <Marchael> ThomasWaldmann: so what's now?
103 2011-06-12T13:51:40 <ThomasWaldmann> index creation?
104 2011-06-12T13:52:59 <Marchael> from what should I start?
105 2011-06-12T13:54:11 <ThomasWaldmann> maybe the script that does a full index rebuild makes sense first
106 2011-06-12T13:55:40 <Marchael> ok, for rebuild we'll give data from backend, process it and them put it to index>
107 2011-06-12T13:55:41 <Marchael> ?
108 2011-06-12T13:56:03 <ThomasWaldmann> sure, what else? :)
109 2011-06-12T13:57:19 <Marchael> how we get documents from backend? point me to module?
110 2011-06-12T13:57:46 <ThomasWaldmann> storage packages
111 2011-06-12T13:57:50 <ThomasWaldmann> -s
112 2011-06-12T13:59:08 <ThomasWaldmann> you need iter_items_noindex() method from backend
113 2011-06-12T14:00:14 <ThomasWaldmann> and read the hints on EP
114 2011-06-12T14:01:50 <ThomasWaldmann> ehrm, first solve the trivial todo from EP 254
115 2011-06-12T14:02:05 <ThomasWaldmann> and remove todo afterwards
116 2011-06-12T14:03:07 <Marchael> ok
117 2011-06-12T14:03:58 <Marchael> but, MoinMoin/apps/__init__.py also is empty
118 2011-06-12T14:04:20 <Marchael> I don't know what should I write there :)
119 2011-06-12T14:04:45 <ThomasWaldmann> use _template.py
120 2011-06-12T14:05:03 <ThomasWaldmann> and just say what the search package contains
121 2011-06-12T14:05:37 <Marchael> done
122 2011-06-12T14:06:05 <Marchael> I'll push that on next commit
123 2011-06-12T14:14:26 *** aa_
124 2011-06-12T14:14:26 *** aa_
125 2011-06-12T14:29:15 <dreimark> xorAxAx: is pkumar working on memory leakage. I am asking because of http://moinmo.in/Himanshu%20Awasthi/Diary%20Page/2011-06-11
126 2011-06-12T14:29:43 *** grzywacz
127 2011-06-12T14:29:43 <dreimark> waldi: do both student work on the same ?
128 2011-06-12T14:29:43 *** grzywacz
129 2011-06-12T14:29:43 *** grzywacz
130 2011-06-12T14:32:27 <xorAxAx> dreimark: i advised him to work on memory leakage
131 2011-06-12T14:34:31 * ThomasWaldmann notes the lack of communication and presence of testing students
132 2011-06-12T14:35:42 <dreimark> xorAxAx: ok, what is left over for hardy89?
133 2011-06-12T14:35:49 <dreimark> waldi: ^
134 2011-06-12T14:36:15 <xorAxAx> dreimark: well, he is supposed to work on the stuff that we agreed on
135 2011-06-12T14:39:05 <dreimark> he writes on the diary page "Now only working on memory leak problem cos pkumar is working on logging."
136 2011-06-12T14:39:23 <dreimark> and none claimed http://etherpad.osuosl.org/moin2-unittest
137 2011-06-12T14:39:39 <dreimark> fix memory leaking
138 2011-06-12T14:40:55 <dreimark> xorAxAx: where can I see what he should work on. Is that all what is in the ep ?
139 2011-06-12T14:41:07 <dreimark> or do you have another table somewhere?
140 2011-06-12T14:41:10 <xorAxAx> dreimark: and his application
141 2011-06-12T14:42:06 <xorAxAx> but the EP is more up to date
142 2011-06-12T14:42:17 <xorAxAx> and it indeed says that HA is supposed to work on leaks
143 2011-06-12T14:43:26 <dreimark> xorAxAx: which line
144 2011-06-12T14:43:55 <ThomasWaldmann> Marchael: http://packages.python.org/Whoosh/indexing.html you read that?
145 2011-06-12T14:44:08 <xorAxAx> dreimark: 164-179
146 2011-06-12T14:44:43 <dreimark> oh, yes then it is now a conflict
147 2011-06-12T14:44:55 <Marchael> ThomasWaldmann: yep
148 2011-06-12T14:44:56 <dreimark> which likly happens because of absene
149 2011-06-12T14:45:05 <dreimark> absence of student and mentor
150 2011-06-12T14:58:18 <Marchael> ThomasWaldmann: In what time we should start building indexes? Only when wiki has starts work? But how we know what wiki admin wants to rebuild indexes?
151 2011-06-12T15:06:37 <ThomasWaldmann> Marchael: i already told you that you should use flask-script
152 2011-06-12T15:07:03 *** m4k3r_
153 2011-06-12T15:07:05 <ThomasWaldmann> that means it is a separate script for full index build / full index update
154 2011-06-12T15:08:17 <ThomasWaldmann> see MoinMoin/script/... for other flask-script based scripts
155 2011-06-12T15:10:23 <ThomasWaldmann> you need such a script at the very start, in case some index gets damaged, in case of incompatible whoosh or moin upgrades - in all these cases you just want to fully rebuild a fresh index
156 2011-06-12T15:11:31 *** m4k3r
157 2011-06-12T15:11:59 <Marchael> so, update script will be called each time when moin starts?
158 2011-06-12T15:18:21 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 253:37c117ad8f7e 2.0/MoinMoin/apps/__init__.py: add docstring to MoinMoin.apps
159 2011-06-12T15:18:38 <ThomasWaldmann> no, the script will be called when the wiki admin decides to call it
160 2011-06-12T15:20:06 *** sinha
161 2011-06-12T15:20:09 <ThomasWaldmann> (and that might be required by us to be done before first start of wiki, if the wiki can't work without index)
162 2011-06-12T15:20:45 * ThomasWaldmann brb
163 2011-06-12T15:26:33 *** franklin
164 2011-06-12T15:38:13 <Marchael> grep -r ' def iter_items_noindex' MoinMoin/storage/backends/ | wc -l
165 2011-06-12T15:38:13 <Marchael> 11
166 2011-06-12T15:38:33 <Marchael> which iter_items_noindex() I should choose?
167 2011-06-12T15:43:45 <ThomasWaldmann> flaskg.storage ones
168 2011-06-12T15:43:53 <Marchael> seems like for all...
169 2011-06-12T15:47:50 <Marchael> hm
170 2011-06-12T15:55:54 <Marchael> May be I do something wrong, but
171 2011-06-12T15:55:54 <Marchael> grep -r 'flaskg.storage' MoinMoin/storage/backends/ | wc -l
172 2011-06-12T15:55:54 <Marchael> 0
173 2011-06-12T16:01:24 <ThomasWaldmann> not everything can be solved by grep
174 2011-06-12T16:10:32 <Marchael> ThomasWaldmann: I guess what you mean MoinMoin/storage/backends/router.py
175 2011-06-12T16:11:54 <ThomasWaldmann> what i mean is using flaskg.storage.iter_items_noindex(), no more, no less
176 2011-06-12T16:27:41 *** franklin
177 2011-06-12T16:35:44 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 5784:5752497d5b03 1.9/MoinMoin/i18n/ (bg.MoinMoin.po da.MoinMoin.po ru.MoinMoin.po): updated i18n
178 2011-06-12T16:51:52 <ThomasWaldmann> Marchael: you're having progress?
179 2011-06-12T16:52:25 <Marchael> trying to understand how to access to flaskg.storage
180 2011-06-12T16:52:39 <Marchael> from ipython
181 2011-06-12T16:53:10 <Marchael> I activated virtual env, but I still got error
182 2011-06-12T16:55:29 <ThomasWaldmann> you need the app etc. initialized
183 2011-06-12T16:55:39 <ThomasWaldmann> just use a flask-script script
184 2011-06-12T16:58:11 <Marchael> but how can I test it? just write own flask-script and run it from terminal?
185 2011-06-12T17:00:41 <ThomasWaldmann> hmm, good question. currently most scripts don't have tests.
186 2011-06-12T17:06:35 <Marchael> btw, I mean how to run flask scripts
187 2011-06-12T17:10:15 <Marchael> Is there any way to run sripts instead of "from ... import ... & manager.add_command()"?
188 2011-06-12T17:10:29 <Marchael> in MoinMoin/script/__init__.py
189 2011-06-12T17:12:39 <ThomasWaldmann> you have a "moin" command
190 2011-06-12T17:12:52 <Marchael> ah, great
191 2011-06-12T17:13:12 <ThomasWaldmann> you're using flask-script all them time :)
192 2011-06-12T17:14:26 <dreimark> Marchael: look at some of the other scripts there
193 2011-06-12T17:34:20 <Marchael> ThomasWaldmann: may be we should put WhooshIndex object to flask.g? so it can be accessed from different modules of moin.
194 2011-06-12T17:42:21 <ThomasWaldmann> yes, maybe.
195 2011-06-12T17:43:16 * ThomasWaldmann is still a bit unsure whether that object is needed at all or whether we better live with just using whoosh directly.
196 2011-06-12T17:43:21 <ThomasWaldmann> we'll see.
197 2011-06-12T17:50:22 *** Marchael
198 2011-06-12T17:54:45 *** Marchael
199 2011-06-12T18:00:24 <Marchael> ThomasWaldmann: I'm not quite understand how some scripts running without options, for exmple MoinMoin.scripts.account.Set_Password requires some parametrs for start, but in MoinMoin/script/__init__.py I see only manager.add_command("account_password", Set_Password()) without options
200 2011-06-12T18:04:10 <Marchael> or it just take params from config?
201 2011-06-12T18:04:49 *** greg_f
202 2011-06-12T18:04:50 <ThomasWaldmann> Set_Password is the class, if it is run, framework call the run(...) method with arguments
203 2011-06-12T18:05:18 <ThomasWaldmann> flask-script has docs, btw
204 2011-06-12T18:15:55 *** greg
205 2011-06-12T18:20:51 <Marchael> seems like it's a big secret
206 2011-06-12T18:23:24 <ThomasWaldmann> what?
207 2011-06-12T18:23:52 <Marchael> http://flask.pocoo.org/docs/search/?q=scripts&check_keywords=yes&area=default
208 2011-06-12T18:24:33 <ThomasWaldmann> is it that hard to google for flask-script?
209 2011-06-12T18:25:17 <Marchael> heh
210 2011-06-12T18:25:48 <Marchael> google automatically rename "flask" to "flash"
211 2011-06-12T18:27:19 <Marchael> hm, "Feeling lucky" rocks
212 2011-06-12T18:27:39 <ThomasWaldmann> try googling for moin2 :P
213 2011-06-12T18:28:18 <Marchael> there is strange mmorpg in first results
214 2011-06-12T18:28:58 <Marchael> ah, russian google
215 2011-06-12T18:29:41 * ThomasWaldmann hates that "did you mean foo" if i searched for bar.
216 2011-06-12T18:41:04 <ronny> hmm, where are the testing students
217 2011-06-12T18:41:42 <ThomasWaldmann> in india :P
218 2011-06-12T18:48:47 * Marchael brb in 1 hour
219 2011-06-12T18:50:33 *** hardy89
220 2011-06-12T19:05:52 *** hardy89
221 2011-06-12T19:11:33 *** MattMaker
222 2011-06-12T19:22:44 *** hardy89
223 2011-06-12T19:22:56 *** pkumar
224 2011-06-12T19:45:35 <pkumar> xorAxAx: hi
225 2011-06-12T19:46:05 <xorAxAx> hi pkumar
226 2011-06-12T19:46:34 <hardy89> hi pkumar
227 2011-06-12T19:46:57 <pkumar> xorAxAx: how do we set fs2 backend? I tried making changes in conftest but didn't work out
228 2011-06-12T19:47:08 <pkumar> hardy89: hi
229 2011-06-12T19:47:12 <xorAxAx> no idea where its actually set
230 2011-06-12T19:47:21 <xorAxAx> did you read the logs of this channel?
231 2011-06-12T19:47:24 <xorAxAx> like you should always do
232 2011-06-12T19:48:06 <pkumar> xorAxAx : hardy89 , here will be working on memory leaking from now, so better I focus on fs2
233 2011-06-12T19:48:25 <hardy89> xorAxAx: we did
234 2011-06-12T19:48:29 <pkumar> xorAxAx : also, the one suggested by Thomas Waldmann
235 2011-06-12T19:48:31 <xorAxAx> didnt you write that you will focus on logging?
236 2011-06-12T19:48:46 <pkumar> yeah, it's almost done.
237 2011-06-12T19:48:48 <xorAxAx> ok
238 2011-06-12T19:49:30 <xorAxAx> well, if youdont know where the fs2 instance is created, add traceback.print_stack() in its constructor
239 2011-06-12T19:50:58 <pkumar> xorAxAx : as mentioned in Thomas repo, we need to change in init_test_app
240 2011-06-12T19:51:10 <pkumar> xorAxAx : but it's showing some error for me.
241 2011-06-12T19:51:21 <xorAxAx> i.e.?
242 2011-06-12T19:51:36 <pkumar> https://bitbucket.org/thomaswaldmann/moin-2.0/issue/10/pytest-run-with-fs2-shows-file-leak
243 2011-06-12T19:52:15 <pkumar> NameError: global name 'BackendError' is not defined
244 2011-06-12T19:52:21 <pkumar> xorAxAx: ^
245 2011-06-12T19:52:39 <xorAxAx> well, then fix the nameerror? :)
246 2011-06-12T19:54:10 <pkumar> xorAxAx : yeah, will do that
247 2011-06-12T19:55:41 <pkumar> xorAxAx : also, we have updated the EP
248 2011-06-12T19:55:52 <xorAxAx> ok
249 2011-06-12T19:56:26 <hardy89> waldi: the EP, diary entry etc has been updated.
250 2011-06-12T19:56:33 <xorAxAx> please pastebin a diff, pkumar
251 2011-06-12T19:57:22 <pkumar> xorAxAx : which diff?
252 2011-06-12T19:57:37 <xorAxAx> of the changes in the EP
253 2011-06-12T19:57:49 <pkumar> ok, sure
254 2011-06-12T19:59:23 <hardy89> xorAxAx: are we not supposed to fill in diary entry for saturday and sunday?
255 2011-06-12T19:59:35 <xorAxAx> hardy89: you may do so, why not?
256 2011-06-12T19:59:44 <hardy89> xorAxAx: right.
257 2011-06-12T19:59:45 <pkumar> xorAxAx : http://paste.pocoo.org/show/405137/
258 2011-06-12T19:59:46 <xorAxAx> no idea what thomas expects
259 2011-06-12T19:59:58 <xorAxAx> of course the working week consists of 5 days
260 2011-06-12T20:00:12 <xorAxAx> but we are in IT where unions are quite seldom :)
261 2011-06-12T20:00:17 <hardy89> xorAxAx: will ask him later. don't want to ping him just for this.
262 2011-06-12T20:00:19 <xorAxAx> so nobody will complain if you work on sunday
263 2011-06-12T20:00:24 <hardy89> xorAxAx: o :)
264 2011-06-12T20:00:58 <pkumar> hardy89 : If I remember right, he mentioned to have diary entries only for working days
265 2011-06-12T20:01:41 <hardy89> pkumar: what if we *do* work on weekends. cos i mostly do as i miss out on weekdays.
266 2011-06-12T20:01:50 <hardy89> cos of my exams
267 2011-06-12T20:02:01 <xorAxAx> hardy89: thats fine, i guess nobody will mind
268 2011-06-12T20:02:13 <hardy89> also now i'm full time working on this. done with my college :)
269 2011-06-12T20:02:21 <hardy89> xorAxAx: sure.
270 2011-06-12T20:02:54 <pkumar> hardy89 : actually I feel more peace on weekends to work on the project
271 2011-06-12T20:03:15 <pkumar> free from other stuffs :)
272 2011-06-12T20:03:22 <hardy89> pkumar: eye on the deadline ;)
273 2011-06-12T20:04:03 <pkumar> hardy89 : we really need to make up for earlier lags, one of the reason being your exams
274 2011-06-12T20:04:24 <hardy89> pkumar: i agree to that.
275 2011-06-12T20:04:45 <pkumar> hardy89 : np, we'll make it up :)
276 2011-06-12T20:04:55 <hardy89> pkumar: last was on 11th June. now i'm good to go :)
277 2011-06-12T20:06:31 <pkumar> hardy89 : sure
278 2011-06-12T20:13:18 <ThomasWaldmann> for the diary i expect an entry for each day when work has been done (and the diary entry has to happen on that day as last action)
279 2011-06-12T20:13:52 <ThomasWaldmann> and I don't want entries for days where no work has been done, because this give a false impression when looking at the calendar
280 2011-06-12T20:14:02 <hardy89> ThomasWaldmann: thats is being followed now :)
281 2011-06-12T20:14:37 <ThomasWaldmann> also, you don't need to fill it with blabla, just write what you did on that day :)
282 2011-06-12T20:15:18 <hardy89> ThomasWaldmann: was that exclusively for me? :)
283 2011-06-12T20:15:45 <xorAxAx> no, makes sense
284 2011-06-12T20:16:17 <hardy89> right.
285 2011-06-12T20:19:57 *** greg
286 2011-06-12T20:19:57 *** greg_f
287 2011-06-12T20:24:11 <ThomasWaldmann> hardy89: that's common for everybody
288 2011-06-12T20:24:37 <hardy89> ThomasWaldmann: all right. still will keep in mind.
289 2011-06-12T20:25:06 <ThomasWaldmann> and in general, we expect better communication. that's especially for testing people.
290 2011-06-12T20:25:35 <pkumar> ok
291 2011-06-12T20:26:27 <hardy89> ok.
292 2011-06-12T20:26:59 <ThomasWaldmann> did you already find out something about the mem leak?
293 2011-06-12T20:28:22 <pkumar> ThomasWaldmann : I think wrapper backend has some problem.
294 2011-06-12T20:28:31 <pkumar> in app
295 2011-06-12T20:29:39 <ThomasWaldmann> like?
296 2011-06-12T20:31:32 <pkumar> well, I was just using objgraph and even after deinit_test_app wrapper instance is not removed.
297 2011-06-12T20:31:59 <pkumar> gimme me mom, I'll paste it
298 2011-06-12T20:40:35 <pkumar> http://paste.pocoo.org/show/405169/
299 2011-06-12T20:41:35 <pkumar> though it gives a rough idea I guess
300 2011-06-12T20:43:58 <ThomasWaldmann> so, where exactly is the problem?
301 2011-06-12T20:45:09 <ThomasWaldmann> and what do the numbers show, precisely?
302 2011-06-12T20:45:31 <pkumar> after every destroy_app(app) we have increasing number of wrapper_descriptor
303 2011-06-12T20:46:32 <pkumar> these are not removed it seems
304 2011-06-12T20:47:18 <ThomasWaldmann> and what is that wrapper_descriptor?
305 2011-06-12T20:47:20 *** hardy89
306 2011-06-12T20:47:42 *** hardy89
307 2011-06-12T20:48:43 <pkumar> wrapper_descriptor is a type I guess
308 2011-06-12T20:49:20 <ThomasWaldmann> well, that's not very precise yet
309 2011-06-12T20:49:57 <ThomasWaldmann> also, how does it deal with garbage collection?
310 2011-06-12T20:50:24 <pkumar> yeah, need to work more on that to get more precise results.
311 2011-06-12T20:50:48 <pkumar> by the way, hardy89 is working on leaking problem and he is using dowser I guess
312 2011-06-12T20:50:56 <pkumar> right hardy89 ?
313 2011-06-12T20:51:28 <hardy89> pkumar: yes.
314 2011-06-12T21:00:21 *** hardy89
315 2011-06-12T21:05:17 <ronny> oO
316 2011-06-12T21:09:20 <pkumar> how do we add a path in sys.path. sys.path.append('pathname') is not permanent.
317 2011-06-12T21:10:51 <TheSheep> pkumar: it is if you put it in your code
318 2011-06-12T21:11:01 <TheSheep> pkumar: in the wsgi file for example
319 2011-06-12T21:13:38 <pkumar> Actually, I need to objgraph in virtual env, but it need some module from /usr/lib/pymodules
320 2011-06-12T21:14:02 <pkumar> that is not included in sys.path of virtual env
321 2011-06-12T21:14:49 <pkumar> so I can just add the file in the file that has to be run right?
322 2011-06-12T21:15:06 <pkumar> file => sys.path
323 2011-06-12T21:15:38 <pkumar> *
324 2011-06-12T21:20:40 <dreimark> pkumar: pip install into your env?
325 2011-06-12T21:21:25 <pkumar> actually there are more than one modules needed
326 2011-06-12T21:22:26 <pkumar> so, installing all of them for something not related to MoinMoin dependencies would not be a very good idea I guess
327 2011-06-12T21:22:58 <pkumar> so I was thinking if I could append the path to those modules
328 2011-06-12T21:23:43 <pkumar> in sys.path
329 2011-06-12T21:25:25 <ronny> pkumar: you can just pip install them into the venv, no need to change the deps of moin
330 2011-06-12T21:27:28 <pkumar> ronny: though that is another option, appending the path could be a simpler one.(if we can append the path easily)
331 2011-06-12T21:30:29 <ronny> pkumar: if you cant figure it in 5 minutes, rethink
332 2011-06-12T21:31:51 <ronny> your main goal is not to research the magic of the python import system
333 2011-06-12T21:33:08 <pkumar> ronny: I was just trying to run objgraph. http://mg.pov.lt/objgraph/
334 2011-06-12T21:34:38 <pkumar> I found it really useful and interesting.
335 2011-06-12T21:36:55 <ronny> pkumar: yes, and you can pip install that in the venv
336 2011-06-12T21:37:02 <ronny> so whats the problem?
337 2011-06-12T21:38:21 <pkumar> yeah, I did. But in case of objgraph.show_refs([y]) doesn't display the image.
338 2011-06-12T21:38:31 <pkumar> in virtual env
339 2011-06-12T21:38:57 <pkumar> it need some modules to run it.
340 2011-06-12T21:39:12 <pkumar> though it runs well outside.
341 2011-06-12T21:42:49 <pkumar> I think it may be due to graphviz
342 2011-06-12T21:45:06 <sinha> dreimark: reg that re-naming of bookmark as reminder, i left a comment at review
343 2011-06-12T21:46:05 * ThomasWaldmann too
344 2011-06-12T21:50:51 <Marchael> ThomasWaldmann: I trying call python2 MoinMoin/script/__init__.py rebuild_index, but nothing happens. code: http://paste.pocoo.org/show/405211/
345 2011-06-12T21:51:30 *** sinha
346 2011-06-12T21:51:46 *** sinha
347 2011-06-12T21:53:38 <Marchael> I expected what it prints message or traceback
348 2011-06-12T21:54:00 <pkumar> bbl
349 2011-06-12T21:54:19 *** pkumar
350 2011-06-12T22:06:45 <ThomasWaldmann> Marchael: you use the "moin" command
351 2011-06-12T22:06:54 <ThomasWaldmann> moin --help
352 2011-06-12T22:11:10 <Marchael> finally, hello world
353 2011-06-12T22:12:05 <Marchael> one moin to rule them all
354 2011-06-12T22:12:28 <ThomasWaldmann> yup :)
355 2011-06-12T22:12:51 <ThomasWaldmann> and you can even write "moin moin" :)
356 2011-06-12T22:17:27 *** grzywacz
357 2011-06-12T22:24:06 <Marchael> ThomasWaldmann: I try to use iter_items_noindex() but it fails http://paste.pocoo.org/show/405235/
358 2011-06-12T22:25:31 <ThomasWaldmann> Marchael: forget what i said about "with x as a, y as b:", that is py2.7 only
359 2011-06-12T22:27:19 <Marchael> ehrm
360 2011-06-12T22:27:29 <ThomasWaldmann> Marchael: hm, looks like you need the unprotected backend
361 2011-06-12T22:28:40 <Marchael> yep, I get one
362 2011-06-12T22:29:28 <Marchael> now I'm need to get values from fields and put them to index
363 2011-06-12T22:31:53 <ThomasWaldmann> codereview of what you have?
364 2011-06-12T22:32:40 <Marchael> I think what there is nothing to review yet
365 2011-06-12T22:32:55 <Marchael> I just print list of backend items
366 2011-06-12T22:33:09 <Marchael> now I'm need extract fields from them
367 2011-06-12T22:33:48 <Marchael> How can I get a list of fields for each item in backend?
368 2011-06-12T22:35:13 <ThomasWaldmann> a revision is dict-like
369 2011-06-12T22:35:25 <ThomasWaldmann> (for all the metadata stuff)
370 2011-06-12T22:36:07 <Marchael> s/list of/dict with/
371 2011-06-12T22:36:18 <ThomasWaldmann> and we are indexing revisions, not items (currently)
372 2011-06-12T22:37:29 <Marchael> I see RouterItem objects in that list, but RouterItem Inherited from SerializableItemMixin, IndexingItemMixin, BareRouterItem
373 2011-06-12T22:38:05 <ThomasWaldmann> you need to get the list of revisions in each item you get and iterate over the revisions
374 2011-06-12T22:39:00 <Marchael> list_revisions()?
375 2011-06-12T22:39:19 <ThomasWaldmann> it sounds good and has a docstring :)
376 2011-06-12T22:41:32 <Marchael> [[0], [0], [0], [0], [0]] seems like I need to populate my wiki :)
377 2011-06-12T22:44:17 <Marchael> I have got list of revisions but seems like it looking like ^
378 2011-06-12T22:44:32 <Marchael> s/seems like//
379 2011-06-12T22:45:15 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * 1a1dab306214 r259 default/MoinMoin/ (themes/__init__.py apps/frontend/views.py items/__init__.py): move the flatland form html generator to the global environment
380 2011-06-12T22:45:16 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * 20c18292bd02 r260 default/MoinMoin/ (4 files in 3 dirs): (log message trimmed)
381 2011-06-12T22:45:16 <CIA-69> use flatland and some html5 features for search form, details see below
382 2011-06-12T22:45:16 <CIA-69> generate the top-right search form using flatland.
383 2011-06-12T22:45:16 <CIA-69> currently, search form is only rendered for "show" view.
384 2011-06-12T22:45:16 <CIA-69> use a Validator for the search form, currently it rejects if search query
385 2011-06-12T22:45:17 <CIA-69> length is below 2 (such stuff usually causes high load and gives too many
386 2011-06-12T22:45:17 <CIA-69> results).
387 2011-06-12T22:45:17 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * bc16cd1f04e0 r261 default/MoinMoin/ (templates/layout.html apps/frontend/views.py): integrate the simple search into show view (makes it easier to show search query validation errors)
388 2011-06-12T22:45:19 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * 4654343b5abd r262 default/MoinMoin/themes/modernized/static/css/common.css: moin-searchform css: prettify validation error display, remove unused stuff
389 2011-06-12T22:45:19 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * 136a5245d05b r263 default/MoinMoin/storage/ (_tests/test_middleware_acl.py backends/acl.py): acl code/tests: meta ACL value is a unicode object, do not use bytestrings in tests, remove code that supports it being a list or tuple
390 2011-06-12T22:45:59 <CIA-69> Akash Sinha <akash2607@gmail.com> * 3076415fdaba r264 default/MoinMoin/ (3 files in 3 dirs): Bookmark functionality added for the global history results
391 2011-06-12T22:45:59 <CIA-69> Thomas Waldmann <tw AT waldmann-edv DOT de> * 37c117ad8f7e r265 default/MoinMoin/apps/__init__.py: add docstring to MoinMoin.apps
392 2011-06-12T22:45:59 <CIA-69> Akash Sinha <akash2607@gmail.com> * 7754844a1dbe r266 default/MoinMoin/ (10 files in 9 dirs): branch merged with main repo
393 2011-06-12T22:48:15 <ThomasWaldmann> Marchael: don't build a full list of items, nor of revision object, just iterate over them
394 2011-06-12T22:48:54 <Marchael> of course, I'm just playing with API
395 2011-06-12T22:53:52 <dreimark> sinha: can you have a look at http://moinmo.in/AkashSinha/Gsoc2011Diary/2011-06-11
396 2011-06-12T22:54:05 <dreimark> the cs of 0
397 2011-06-12T22:54:15 <dreimark> for item7
398 2011-06-12T22:54:35 <dreimark> when was it? I have a similiar example wehere it was on an other day
399 2011-06-12T22:54:41 *** RogerHaase
400 2011-06-12T22:55:22 <dreimark> we should only see the most recent changesets of a day
401 2011-06-12T22:55:32 <dreimark> and not all
402 2011-06-12T22:56:08 <sinha> yes i agree
403 2011-06-12T22:56:25 <sinha> this must be on 7th or after that i guess
404 2011-06-12T22:56:46 <sinha> i discard the changes which occured after the bookmark
405 2011-06-12T22:56:54 <dreimark> ok
406 2011-06-12T22:56:58 <dreimark> ehme
407 2011-06-12T22:57:05 <dreimark> not only after the bookmark
408 2011-06-12T22:58:04 <sinha> yes and on the bookmark also
409 2011-06-12T22:58:06 <dreimark> also if there is not bookmark, it should not show collected accumulated changes
410 2011-06-12T22:58:28 <sinha> that was item name grouping, right ?
411 2011-06-12T22:58:45 <dreimark> yes, i have overseen that
412 2011-06-12T22:58:56 <ThomasWaldmann> dreimark: hmm, why not?
413 2011-06-12T22:58:59 <dreimark> because i have not had many days changes in a copy of your repo
414 2011-06-12T22:59:10 <dreimark> ThomasWaldmann: because then we have all changes ever done
415 2011-06-12T22:59:16 <dreimark> we don't that in 1.x
416 2011-06-12T23:00:36 <ThomasWaldmann> dreimark: it would stop at now - N days
417 2011-06-12T23:00:55 <dreimark> it shows currently rev 0 which is from a day earlier
418 2011-06-12T23:01:35 <Marchael> gn/bbl
419 2011-06-12T23:01:58 <sinha> dreimark: have you tried the code ?
420 2011-06-12T23:02:11 <ThomasWaldmann> hmm, needs more thinking
421 2011-06-12T23:02:28 <dreimark> yes, but unfortunately i have now a cleaned wiki again
422 2011-06-12T23:02:44 <dreimark> but in an hour i have an other day
423 2011-06-12T23:04:01 <sinha> i mean if you are taking the example of that june7, then that rev0 has taken on that day itself
424 2011-06-12T23:04:49 <dreimark> sinha: i have seen a cs which was not from the day itselfs
425 2011-06-12T23:05:08 <dreimark> 1.x surpresses changes from other days
426 2011-06-12T23:05:35 <sinha> otherdays that is recent to bookmark time ?
427 2011-06-12T23:05:42 <dreimark> not logged in
428 2011-06-12T23:05:53 <dreimark> sinha: no bookmark, because not logged in
429 2011-06-12T23:06:17 <dreimark> e.g. if i change in a fresh wiki creole
430 2011-06-12T23:06:36 <dreimark> then i get [0] and [1]
431 2011-06-12T23:06:47 <dreimark> but 0 is from 2011-03-12
432 2011-06-12T23:07:42 <sinha> so if the bookmark is 2011-03-08
433 2011-06-12T23:07:53 <sinha> then this 0 will be seen ? or not to be seen ?
434 2011-06-12T23:07:59 <dreimark> sinha: i don't have now a bookmark
435 2011-06-12T23:08:07 <dreimark> for this example
436 2011-06-12T23:08:21 <ThomasWaldmann> hmm, without bookmark it should be grouped by day. because we have no bookmark as a reference point for diffs.
437 2011-06-12T23:08:58 <ThomasWaldmann> and because some readers maybe remember manually by day what they have read and what not
438 2011-06-12T23:09:37 <dreimark> if i set a bookmark then it works as expected
439 2011-06-12T23:19:55 <dreimark> sinha: not seen
440 2011-06-12T23:20:06 <sinha> why
441 2011-06-12T23:20:17 <sinha> it occured after the bookmark
442 2011-06-12T23:20:31 <dreimark> sinha: you should try without a bookmark
443 2011-06-12T23:20:50 <dreimark> and 2011-03-08 is before 2011-06-12
444 2011-06-12T23:21:36 <dreimark> with bookmark everything is fine
445 2011-06-12T23:21:47 <sinha> bookmark is set for 8th and rev took place on 12th
446 2011-06-12T23:22:02 <sinha> so it must be grouped with the the item there
447 2011-06-12T23:22:09 <dreimark> sinha: i am not talking about bookmark
448 2011-06-12T23:22:22 <dreimark> logout change creole and tell me what you get
449 2011-06-12T23:24:37 <sinha> dreimark: http://moinmo.in/AkashSinha/Gsoc2011Diary/2011-06-12?action=AttachFile&do=view&target=creole.png
450 2011-06-12T23:25:16 <dreimark> sinha: the [0] is not from today
451 2011-06-12T23:25:41 <dreimark> so it should not show up at today
452 2011-06-12T23:26:05 <dreimark> it should be surpressed
453 2011-06-12T23:26:10 <sinha> but item name grouping was done in that way
454 2011-06-12T23:26:25 <sinha> i discussed this thing earlier
455 2011-06-12T23:26:32 <dreimark> I know, and as I explained already I missed to tell this
456 2011-06-12T23:26:40 <dreimark> and found it in the morning
457 2011-06-12T23:26:40 <sinha> then where will i show this ?
458 2011-06-12T23:26:53 <dreimark> the [0] will be surpressed
459 2011-06-12T23:27:03 <dreimark> because you show only new changes
460 2011-06-12T23:27:10 <sinha> means all the changes of that item
461 2011-06-12T23:27:24 <dreimark> all new changes of that day
462 2011-06-12T23:27:27 <dreimark> not the past
463 2011-06-12T23:27:46 <dreimark> just look waht happens if there would be a few hundred changes per day
464 2011-06-12T23:28:08 <dreimark> if you edit your homepage on moinmo.in now
465 2011-06-12T23:28:18 <sinha> yes that would be problematic
466 2011-06-12T23:28:49 <dreimark> you see in RC after deleting the bookmark that it shows up only with today changes
467 2011-06-12T23:30:22 <sinha> yes it do
468 2011-06-12T23:30:42 <sinha> and also when bookmark is set it only shows that day change only
469 2011-06-12T23:30:48 <sinha> not the previous change
470 2011-06-12T23:31:42 <dreimark> yes, ok we should keep this logic
471 2011-06-12T23:34:21 * dreimark feels guiltily for not founding this earlier - sorry
472 2011-06-12T23:35:46 <sinha> okay not a big deal, just add a query condition during the item grouping, will checkif the revision's date is same as the latest revision for that item, then include it otherwise reject it. right ?
473 2011-06-12T23:36:15 <dreimark> yes
474 2011-06-12T23:37:43 <sinha> okay doing that before the paging stuff.
475 2011-06-12T23:37:48 <dreimark> good
476 2011-06-12T23:47:55 <ThomasWaldmann> why do we need paging?
477 2011-06-12T23:54:07 <dreimark> ThomasWaldmann: in 1.x we have 1 | 2 | 3 | 7 | 14 | 30 | 60 | 90 days.
478 2011-06-12T23:54:19 <dreimark> this is one kind of pageing
479 2011-06-12T23:54:22 <dreimark> sinha: ^^
480 2011-06-12T23:54:32 <ThomasWaldmann> no, it showed all on one page
481 2011-06-12T23:54:55 <sinha> paging like showing 20 results on one page like this ?
482 2011-06-12T23:55:02 <ThomasWaldmann> that's of course not very comfortable for 90d, but otoh, this is rarely used
483 2011-06-12T23:55:25 <ThomasWaldmann> usually one has a few days or just since bookmark
484 2011-06-12T23:55:36 <dreimark> the other is to show only on day 3 the 5th recent changes
485 2011-06-12T23:55:39 <ThomasWaldmann> which is mostly fine without paging
486 2011-06-12T23:55:48 <dreimark> and paginate others
487 2011-06-12T23:56:12 <dreimark> currently if we have users which did many edit per day
488 2011-06-12T23:56:42 <dreimark> then that item becomes only short in RC after editin it at the other day again
489 2011-06-12T23:57:20 <dreimark> ThomasWaldmann: i tried already to desctribe that on http://etherpad.osuosl.org/moin-2-ui line 92
490 2011-06-12T23:57:33 <dreimark> 89