1 2010-09-19T00:10:50 <ThomasWaldmann> ok, what do we hack next? :)
2 2010-09-19T04:42:17 <valeuf> dreimark: ah sad that the guys lef
3 2010-09-19T04:42:40 <valeuf> dreimark: btw, what about adding some documentation about the command for the converters in http://moinmo.in/MoinMoin2.0/Converters ?
4 2010-09-19T09:07:14 <ThomasWaldmann> moin
5 2010-09-19T09:12:25 <valeuf> ThomasWaldmann: moin
6 2010-09-19T09:22:17 <ThomasWaldmann> hi valeuf
7 2010-09-19T09:28:49 *** RaphaelBossek
8 2010-09-19T09:43:22 <ThomasWaldmann> btw, we have to review the dom conversion chain about order and the right place to put caching in it
9 2010-09-19T09:43:43 <RaphaelBossek> ThomasWaldmann, moinmoin
10 2010-09-19T09:44:31 * RaphaelBossek try to find out how to disable PON'T PANIC feature of werkzeug in order to be able to debug MoinMoin in WingIDE
11 2010-09-19T09:44:55 <ThomasWaldmann> moin -d ?
12 2010-09-19T09:50:05 <RaphaelBossek> "moin -d" does not work for me; nobody calls http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/support/flaskext/script.py#l300
13 2010-09-19T09:51:29 <ThomasWaldmann> ehrm, sorry, try "moin moin -d -r"
14 2010-09-19T09:51:37 <ThomasWaldmann> (no debugger, no reloader)
15 2010-09-19T09:52:46 <ThomasWaldmann> just "moin" is a special case that hacks some stuff so it just starts the standalone server with the usual params
16 2010-09-19T09:54:05 <RaphaelBossek> it's not enought to stop PON'T PANIC from werkzeug
17 2010-09-19T09:54:11 * RaphaelBossek is Semmeln holen
18 2010-09-19T09:55:39 <ThomasWaldmann> there is also a DEBUG config var, not sure whether it is something different
19 2010-09-19T09:56:35 <ThomasWaldmann> valeuf: TheSheep: waldi: about smileys
20 2010-09-19T09:56:49 <ThomasWaldmann> a) can we live with just one set?
21 2010-09-19T09:57:13 <ThomasWaldmann> currently they are served from a place independant of the theme the user has selected
22 2010-09-19T09:59:45 <ThomasWaldmann> b) the smiley converter is rather late in the conversion chain. together with how include converter works right now, smiley converter runs multiple times over same data.
23 2010-09-19T10:01:21 <ThomasWaldmann> if what it outputs is just a function of the input, we could do smileys early, and have the resulting tree cached
24 2010-09-19T10:01:28 <valeuf> ThomasWaldmann: I thought that the implementation I wrote was using a directory depending of the them
25 2010-09-19T10:01:45 <ThomasWaldmann> i killed the modernized hardcoded part
26 2010-09-19T10:02:00 <valeuf> otherwise, I need to check why the smiley converter run multiple times over same data
27 2010-09-19T10:02:32 <ThomasWaldmann> because of include using internal_representation()
28 2010-09-19T10:02:59 <valeuf> I see
29 2010-09-19T10:03:24 <ThomasWaldmann> i would like to change/reorder that
30 2010-09-19T10:03:53 <ThomasWaldmann> internal_repr should give the dom tree with all stuff converted that is just a function of input document
31 2010-09-19T10:03:54 <valeuf> ThomasWaldmann: cannot we try to run the smiley converter at the very end of conversion chain?
32 2010-09-19T10:04:07 <ThomasWaldmann> (not of user, not of other items)
33 2010-09-19T10:04:32 <ThomasWaldmann> and then we just make internal_repr use the cache
34 2010-09-19T10:04:54 <ThomasWaldmann> valeuf: running it once at the end is the 2nd best option
35 2010-09-19T10:05:21 <ThomasWaldmann> we do that if it depends on other stuff (like user's theme), if we want that
36 2010-09-19T10:06:20 <ThomasWaldmann> btw i just had an idea how we could make the browser do the smiley theming for us :)
37 2010-09-19T10:06:39 <ThomasWaldmann> without having to generate different domtree
38 2010-09-19T10:07:31 <ThomasWaldmann> couldn't we just emit an empty span with some smiley-xy class and then use css images like we have for links?
39 2010-09-19T10:08:10 <ThomasWaldmann> or maybe even put the smiley markup into the span
40 2010-09-19T10:08:28 <ThomasWaldmann> if css works, we hide content, but have an image
41 2010-09-19T10:08:38 <ThomasWaldmann> if css doesn't work, we have the markup
42 2010-09-19T10:09:16 <ThomasWaldmann> and if css refers to the image using a relative url (see link styles), we have themed smileys
43 2010-09-19T10:10:29 <ThomasWaldmann> and we can still do smileys early and cache result
44 2010-09-19T10:12:29 <ThomasWaldmann> hmm, is it possible to hide span content, but have :before image rendered?
45 2010-09-19T10:21:57 <ThomasWaldmann> hmm, just setting it display:none hides both content and :before stuff
46 2010-09-19T10:27:11 <valeuf> ThomasWaldmann: I thought about it too when I did the implementation, however there is some problem
47 2010-09-19T10:27:37 <valeuf> you cannot really manage well the image postion with such technic
48 2010-09-19T10:27:56 <valeuf> if I remember my test, depending of the browser, theimage is not always render at the same place
49 2010-09-19T10:28:07 <valeuf> (not aligned correctly with the text I mean)
50 2010-09-19T10:28:23 <valeuf> but maybe I did not use correc CSS code
51 2010-09-19T10:31:32 <ThomasWaldmann> major issue or just a pixel off?
52 2010-09-19T10:32:51 <valeuf> few pixels, it was not that big in my memory, but I can chek if there is not some fix for that
53 2010-09-19T10:36:17 * ThomasWaldmann tries http://www.kryogenix.org/code/browser/lir/
54 2010-09-19T10:40:34 <TheSheep> ThomasWaldmann: using a large negative text-indent is better
55 2010-09-19T10:44:28 <ThomasWaldmann> hm, funny, currently it displays /!\ with the image ON TOP of it X)
56 2010-09-19T10:45:24 <ThomasWaldmann> http://paste.pocoo.org/show/263997/
57 2010-09-19T10:46:28 <ThomasWaldmann> that !important, is that part of the ie5.5 support hack?
58 2010-09-19T10:46:47 <ThomasWaldmann> (see http://www.kryogenix.org/code/browser/lir/ for complete hack)
59 2010-09-19T10:47:02 <ThomasWaldmann> TheSheep: can you fix the pastebin?
60 2010-09-19T10:47:36 <ThomasWaldmann> (and does your method still have good accessibility, like screenreader)
61 2010-09-19T10:51:47 <ThomasWaldmann> somehow that height 0 doesn't work for me
62 2010-09-19T10:52:54 <TheSheep> ThomasWaldmann: that hack of theirs will only work on h1 and other block elements
63 2010-09-19T10:54:00 <ThomasWaldmann> ah :|
64 2010-09-19T10:54:27 <TheSheep> you can't set height on inline elements
65 2010-09-19T10:54:29 <TheSheep> http://paste.pocoo.org/show/263998/
66 2010-09-19T10:56:09 <TheSheep> note that inline-block is not supported in msie6
67 2010-09-19T10:56:34 <TheSheep> but!
68 2010-09-19T10:56:57 <TheSheep> there was some trick around that
69 2010-09-19T10:58:50 <TheSheep> a screenreader will read that as 'foo slash exclamation point backslash bar'
70 2010-09-19T10:59:36 <TheSheep> perhaps better accessibility would be achieved if it read it as 'foo alert icon bar', but whatever
71 2010-09-19T11:08:41 <ThomasWaldmann> ok, we do it that way for the icon conversion or is there some catch?
72 2010-09-19T11:11:48 <TheSheep> I didn't test it on msie
73 2010-09-19T11:13:19 <TheSheep> you probably want class="moin-icon-alert"
74 2010-09-19T11:18:05 <dreimark> valeuf: sure good place
75 2010-09-19T11:18:08 <dreimark> moin
76 2010-09-19T11:19:34 <valeuf> dreimark: alright
77 2010-09-19T11:20:20 <ThomasWaldmann> i need to go soon
78 2010-09-19T11:20:37 <ThomasWaldmann> valeuf: can you make the appropriate changes in smiley converter and css?
79 2010-09-19T11:21:28 <ThomasWaldmann> the smiley images will move back to themes/modernized/static/img/smileys/ then
80 2010-09-19T11:21:35 <ThomasWaldmann> (hg mv)
81 2010-09-19T11:23:57 <valeuf> ThomasWaldmann: yep, but I om not sure I can do it today
82 2010-09-19T11:26:42 <ThomasWaldmann> well, just have a look at main repo before you do it
83 2010-09-19T11:26:49 <ThomasWaldmann> first come, first commit :D
84 2010-09-19T11:27:04 <ThomasWaldmann> bbl
85 2010-09-19T11:27:19 * dreimark aha a race
86 2010-09-19T11:35:28 <dreimark> eSyr: please forward the recent request for rest highlight on the ML to Dmitry
87 2010-09-19T13:38:10 <ThomasWaldmann> re
88 2010-09-19T13:41:19 <RaphaelBossek> ThomasWaldmann, is MoinMoin/support/flaskext a wrapper over flask ? To get the DEBUG mode working I have to put some MoinMoin specific stuff to it...
89 2010-09-19T13:42:08 <ThomasWaldmann> flaskext are some flask extensions
90 2010-09-19T13:42:19 <RaphaelBossek> so it's a upstream project
91 2010-09-19T13:42:21 <ThomasWaldmann> nothing moin specific
92 2010-09-19T13:42:23 <ThomasWaldmann> yes
93 2010-09-19T13:42:30 <ThomasWaldmann> google flask extensions
94 2010-09-19T13:42:50 <ThomasWaldmann> what do you have to change there?
95 2010-09-19T13:43:50 <RaphaelBossek> http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/support/flaskext/script.py#l345 -- they do not deal in the right way with the DEBUG mode
96 2010-09-19T13:44:43 <RaphaelBossek> if you intent to catch exceptions by the IDE you have to set debug=True,use_debugger=False,use_reloader=False but this is not possible with debug=use_debugger
97 2010-09-19T13:47:09 <RaphaelBossek> related description for the Flask.run() method @ http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/support/flask/app.py#l373
98 2010-09-19T13:49:44 <RaphaelBossek> it's also mandatory to passthrough_errors parameter for werkzeug.run_simple() function
99 2010-09-19T13:50:48 * RaphaelBossek have to find out how to read the DEBUG configuration parameter from wikiconfig(_local).py
100 2010-09-19T13:55:55 <ThomasWaldmann> RaphaelBossek: ok, sounds like you have a bug for flask-script :)
101 2010-09-19T14:06:30 <RaphaelBossek> here is my proposal for a fix : http://paste.pocoo.org/show/264055/
102 2010-09-19T14:10:42 <ThomasWaldmann> line 14 is a bit unclear
103 2010-09-19T14:12:55 <RaphaelBossek> flask calls werkzeug.serving.py:run_simple() function @ http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/support/werkzeug/serving.py#l463
104 2010-09-19T14:13:12 <ThomasWaldmann> == could there be a case when one wants passthrough false and debug true?
105 2010-09-19T14:13:35 <RaphaelBossek> if I'm working within a IDE I do not like that DON'T PANIC werkzeug feature print out the backgrace within the browser
106 2010-09-19T14:13:44 <RaphaelBossek> I want to see the exception within my IDE
107 2010-09-19T14:14:53 <RaphaelBossek> I tweaked the patch a little bit to be 100% backward compatible, http://paste.pocoo.org/show/264058/
108 2010-09-19T14:15:38 <RaphaelBossek> with this patch nothing change with the parameter values because 'passthrough_errors' is False by default in werkzeug
109 2010-09-19T14:15:54 <RaphaelBossek> only if someone set the DEBUG configuration parameter the defaults change
110 2010-09-19T14:19:31 <ThomasWaldmann> but DEBUG doesn't necessarily mean you use a ide with own debugger
111 2010-09-19T14:20:07 <RaphaelBossek> ok, where can I find the definition of DEBUG ? ;-)
112 2010-09-19T14:20:40 <ThomasWaldmann> no idea :) maybe in the main flask docs?
113 2010-09-19T14:20:50 <RaphaelBossek> until nobody care about it and there is no definition about it, why not do it right now?
114 2010-09-19T14:22:43 <RaphaelBossek> grep for DEBUG and you will not find one line of code where DEBUG is used by flask or werkzeug; it's only an example for an uppercase configuration parameter
115 2010-09-19T14:22:49 <ThomasWaldmann> the only issue with patching it could be to introduce similar problems (just somehow the other way round :)
116 2010-09-19T14:22:49 * RaphaelBossek check the flesk docs
117 2010-09-19T14:23:28 <RaphaelBossek> I would like to contact the upstread author and see how his explanation will be
118 2010-09-19T14:23:38 <RaphaelBossek> he should decide what should happen
119 2010-09-19T14:24:37 <ThomasWaldmann> dan jacob
120 2010-09-19T14:24:55 <RaphaelBossek> do you have email contact to dan?
121 2010-09-19T14:25:06 <RaphaelBossek> I would put you on CC
122 2010-09-19T14:26:29 <ThomasWaldmann> http://bitbucket.org/danjac/flask-script/ there is a issue tracker
123 2010-09-19T14:28:36 * RaphaelBossek afk
124 2010-09-19T14:47:59 * RaphaelBossek re
125 2010-09-19T15:00:29 <dreimark> re
126 2010-09-19T15:08:13 <ThomasWaldmann> TheSheep: that -2000px how is that with full hd displays? :D
127 2010-09-19T15:08:40 <ThomasWaldmann> or the generation after them? :)
128 2010-09-19T15:10:29 * dreimark wants something linke atiny url generator in m-2.0
129 2010-09-19T15:11:05 <dreimark> based on the moin url
130 2010-09-19T15:28:05 <ThomasWaldmann> aren't there enough on the internet?
131 2010-09-19T15:47:46 <RaphaelBossek> ThomasWaldmann, could it be that flaskext (flask-scripts) does not depend on werkzeug? In that case I've to modify my patch a little bit more, because flaskext does not pass though further parameters...
132 2010-09-19T15:48:18 <valeuf> dreimark: how would use such of tiny url generator?
133 2010-09-19T15:51:17 <ThomasWaldmann> RaphaelBossek: flask depends on werkzeug
134 2010-09-19T15:51:43 <RaphaelBossek> ThomasWaldmann, I've found it just in this second :-)
135 2010-09-19T15:52:20 <RaphaelBossek> ThomasWaldmann, so I can summe that flask-skripts has to know about this
136 2010-09-19T15:56:11 <ThomasWaldmann> know what?
137 2010-09-19T15:57:42 <RaphaelBossek> here the official issue agains flask-script to get DEBUG support working @ http://bitbucket.org/danjac/flask-script/issue/15/do-not-catch-exceptions-support-for-debug
138 2010-09-19T16:00:29 <ThomasWaldmann> great :)
139 2010-09-19T16:02:29 <RaphaelBossek> let us see how long it will take until the patch is accepted
140 2010-09-19T16:06:14 <ThomasWaldmann> btw, pep8 wants more spaces :)
141 2010-09-19T16:11:48 * RaphaelBossek check why http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/apps/frontend/views.py#l189 throws an exception
142 2010-09-19T16:12:35 <ThomasWaldmann> traceback?
143 2010-09-19T16:13:40 <RaphaelBossek> traceback : http://paste.pocoo.org/show/264204/
144 2010-09-19T16:16:16 <RaphaelBossek> how to reproduce : http://localhost:8080/+convert/Home?mimetype=application/docbook+xml
145 2010-09-19T16:16:56 <ThomasWaldmann> RaphaelBossek: check mimetype
146 2010-09-19T16:17:40 <RaphaelBossek> ThomasWaldmann, how to solve this exception ?
147 2010-09-19T16:18:07 <ThomasWaldmann> _convert for the docbook type looks ok and matchined, the one in the bases not
148 2010-09-19T16:18:47 <ThomasWaldmann> s/ed/g/
149 2010-09-19T16:22:13 <RaphaelBossek> check http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/items/__init__.py#l628 -- is Binary not the right type ? The Binary._convert() method does not except any parameters...
150 2010-09-19T16:23:00 <RaphaelBossek> to be more specific, converted_item is of type MoinMoin.items.Binary, should it not be ?
151 2010-09-19T16:23:05 <ThomasWaldmann> shouldn't it end up in the docbook class?
152 2010-09-19T16:23:45 <RaphaelBossek> mimetype=u'application/docbook xml'
153 2010-09-19T16:26:34 <dreimark> valeuf: for references in comments or commit messages
154 2010-09-19T16:26:54 <ThomasWaldmann> RaphaelBossek: strange, your url looks ok
155 2010-09-19T16:28:53 <RaphaelBossek> http://hg.moinmo.in/moin/2.0-dev/file/c9cf24017419/MoinMoin/items/__init__.py#l146 : _find_item_class() method does not find any subclass for the mimetype=u'application/docbook xml'
156 2010-09-19T16:29:08 * RaphaelBossek is happy that WingIDE is working in DEBUG mode with MoinMoin 2.0-dev :-)
157 2010-09-19T16:30:10 <RaphaelBossek> where is the + (plus) in the mimetype?
158 2010-09-19T16:30:21 <RaphaelBossek> it should be "application/docbook+xml"
159 2010-09-19T16:31:45 <ThomasWaldmann> ehrm, is there confusion + and blank?
160 2010-09-19T16:32:03 <RaphaelBossek> yes
161 2010-09-19T16:33:24 <RaphaelBossek> I cachanged the mimetype=u'application/docbook xml' value with the debugger to mimetype=u'application/docbook+xml' and it's working
162 2010-09-19T16:33:26 <ThomasWaldmann> looks like url-unquote-plus or so
163 2010-09-19T16:33:35 <ThomasWaldmann> or quote the +
164 2010-09-19T16:42:20 <RaphaelBossek> the following URL works fine : http://localhost:8080/+convert/Home?mimetype=application/docbook%2Bxml
165 2010-09-19T16:43:39 <RaphaelBossek> what about this exception -- is this a low priority bug ? I would except to get a MoinMoin error message saying for example "not supported mime type"
166 2010-09-19T16:44:11 <ThomasWaldmann> the base classes methods need fixing
167 2010-09-19T16:44:27 <RaphaelBossek> ThomasWaldmann, which one ?
168 2010-09-19T16:44:42 <ThomasWaldmann> _convert
169 2010-09-19T16:56:39 <RaphaelBossek> bye
170 2010-09-19T16:56:43 *** RaphaelBossek
171 2010-09-19T17:43:23 * ThomasWaldmann tries changing converter order
172 2010-09-19T17:43:39 <ThomasWaldmann> input, smiley, link first
173 2010-09-19T17:43:46 <ThomasWaldmann> cache
174 2010-09-19T17:43:59 <ThomasWaldmann> include, macro afterwards
175 2010-09-19T18:10:06 <ThomasWaldmann> can anyone do a quick converter review?
176 2010-09-19T18:23:26 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9981:6c01506e1a93 2.0-dev/MoinMoin/items/__init__.py: (log message trimmed)
177 2010-09-19T18:23:26 <CIA-36> change converter order, see below
178 2010-09-19T18:23:26 <CIA-36> first we call internal_representation(): this is just calling input and smiley
179 2010-09-19T18:23:26 <CIA-36> converter - output of this conversion does not change except when the local
180 2010-09-19T18:23:26 <CIA-36> item data changes. We can easily add caching for this later.
181 2010-09-19T18:23:26 <CIA-36> then we call _expand_document(): this expands macros and includes (can't be
182 2010-09-19T18:23:27 <CIA-36> cached easily)
183 2010-09-19T18:23:32 <ThomasWaldmann> ^^ that
184 2010-09-19T19:03:53 <dreimark> diofeher: TheSheep how can i try the userbrowser ?
185 2010-09-19T19:04:23 <dreimark> ah, found, +admin
186 2010-09-19T19:05:57 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9982:11325084fa7a 2.0-dev/MoinMoin/items/__init__.py:
187 2010-09-19T19:05:58 <CIA-36> cache internal_representation()
188 2010-09-19T19:05:58 <CIA-36> you need to configure flask-cache to see it working:
189 2010-09-19T19:05:58 <CIA-36> DEBUG = False
190 2010-09-19T19:05:58 <CIA-36> CACHE_TYPE = 'filesystem'
191 2010-09-19T19:05:58 <CIA-36> CACHE_DIR = '/path/to/flask-cache-dir'
192 2010-09-19T19:06:58 <dreimark> +admin/ behaves different than +admin - the + becomes %2B
193 2010-09-19T19:07:32 * dreimark tries to find that
194 2010-09-19T19:07:38 <dreimark> also the handler thingie
195 2010-09-19T19:20:42 <ThomasWaldmann> ok, that domtree caching saves quite a bit of time
196 2010-09-19T19:27:07 *** grzywacz
197 2010-09-19T19:27:07 *** grzywacz
198 2010-09-19T19:27:09 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9983:691ecee81ba8 2.0-dev/MoinMoin/ (templates/layout.html theme/__init__.py): add a render_template timer
199 2010-09-19T19:27:34 * dreimark cries
200 2010-09-19T19:28:07 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9984:e0ca2a47a4db 2.0-dev/MoinMoin/datastruct/backends/_tests/test_wiki_groups.py: test_wiki_groups: fixed docstrings
201 2010-09-19T19:28:07 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9985:8863250c7b85 2.0-dev/MoinMoin/ (_tests/__init__.py storage/_tests/test_serialization.py): update_item moved to tests
202 2010-09-19T19:28:08 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9986:51d2b40821d9 2.0-dev/MoinMoin/_tests/__init__.py: _test.__init__: import list sorted
203 2010-09-19T19:28:09 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9987:84c4f9dea1c0 2.0-dev/MoinMoin/ (_tests/__init__.py storage/_tests/test_serialization.py): tests: moved become_trusted into test_serialization
204 2010-09-19T19:28:11 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9988:0227c2ba23df 2.0-dev/MoinMoin/ (3 files in 2 dirs): tests: simplified meta keys api for create_item append_item. meta has to be given as dict.
205 2010-09-19T19:28:12 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9989:252f66104e5e 2.0-dev/MoinMoin/_tests/__init__.py: unified itemname, item_name to name
206 2010-09-19T19:28:15 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9990:8af8c5daada8 2.0-dev/MoinMoin/ (2 files in 2 dirs): tests.append_item: appends content to data too. For testing of wiki groups this is not necessary.
207 2010-09-19T19:28:16 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9991:62333ef0e0af 2.0-dev/ (21 files in 13 dirs): meged main
208 2010-09-19T19:28:19 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9992:0a28af6360a3 2.0-dev/MoinMoin/ (templates/layout.html theme/__init__.py): merged main
209 2010-09-19T19:28:28 <dreimark> merged main is silly
210 2010-09-19T19:28:45 <ThomasWaldmann> ?
211 2010-09-19T19:29:17 <dreimark> well I merged before my push you commited faster than I have to merge again
212 2010-09-19T19:29:54 <ThomasWaldmann> that's life :)
213 2010-09-19T19:30:29 <dreimark> later may be some of those utils should be moved into a cli
214 2010-09-19T19:36:17 <ThomasWaldmann> E TypeError: create_item() got an unexpected keyword argument 'acl'
215 2010-09-19T19:36:23 <ThomasWaldmann> is that new?
216 2010-09-19T19:37:21 <dreimark> where ?
217 2010-09-19T19:38:00 <dreimark> yes, it should be a dict now
218 2010-09-19T19:38:15 * ThomasWaldmann just ran py.test from toplevel dir
219 2010-09-19T19:38:37 * dreimark looks
220 2010-09-19T19:40:19 * dreimark fixes
221 2010-09-19T19:40:35 <dreimark> sorry missed test_security
222 2010-09-19T19:42:27 <ThomasWaldmann> maybe create_item and append_item should be killed and just update_item be used
223 2010-09-19T19:43:14 <ThomasWaldmann> it's all more or less the same, but with different implementations
224 2010-09-19T19:43:50 <dreimark> i don't think so. at least not now.
225 2010-09-19T19:44:52 <dreimark> if you want to keep the content or the old meta you have to do this than in tests
226 2010-09-19T19:47:32 <ThomasWaldmann> and the meta stuff in append_item is rather strange anyway
227 2010-09-19T19:47:56 <dreimark> what is strange?
228 2010-09-19T19:48:39 <ThomasWaldmann> well, maybe explain what it shall do
229 2010-09-19T19:50:12 <ThomasWaldmann> but {} and .extend() don't look like mixing well
230 2010-09-19T19:50:21 <dreimark> it overwrite existing meta data with old + new
231 2010-09-19T19:51:11 <ThomasWaldmann> .extend only works for lists
232 2010-09-19T19:51:26 <ThomasWaldmann> and it expects a list as arg
233 2010-09-19T19:51:48 <ThomasWaldmann> nothing tells that the value is a list
234 2010-09-19T19:52:02 <ThomasWaldmann> and even if, it wouldn't work as expected
235 2010-09-19T19:52:26 <dreimark> it works with my group tests or do you day they are broken too
236 2010-09-19T19:52:33 <dreimark> s/d/s/
237 2010-09-19T19:53:05 <ThomasWaldmann> >>> {}.extend(1)
238 2010-09-19T19:53:05 <ThomasWaldmann> Traceback (most recent call last): File "<stdin>", line 1, in <module>
239 2010-09-19T19:53:05 <ThomasWaldmann> AttributeError: 'dict' object has no attribute 'extend'
240 2010-09-19T19:53:39 <ThomasWaldmann> >>> [].extend(1)
241 2010-09-19T19:53:40 <ThomasWaldmann> Traceback (most recent call last): File "<stdin>", line 1, in <module>
242 2010-09-19T19:53:40 <ThomasWaldmann> TypeError: 'int' object is not iterable
243 2010-09-19T19:54:35 <dreimark> true that is wrong
244 2010-09-19T19:54:44 <ThomasWaldmann> (if you give a string or some other sequence, it won't crash, but it won't do the right thing either)
245 2010-09-19T19:55:11 * ThomasWaldmann would say "kill it completely", see above :)
246 2010-09-19T19:56:13 <dreimark> hmm, at least i learned how to access all item data.
247 2010-09-19T19:57:04 <dreimark> ok, agreed otherwise these utils become much bigger than the tests
248 2010-09-19T19:57:43 <ThomasWaldmann> the only thing what's useful from the other methods is the unicode check and encoding, that can be added to update_item
249 2010-09-19T19:58:35 <ThomasWaldmann> and revno=None could determine revno automatically
250 2010-09-19T20:05:33 <ThomasWaldmann> the stuff calling append_item does not really make sense any more
251 2010-09-19T20:06:06 <ThomasWaldmann> it made sense back then when the group definition was a piece of wiki markup text and we appended to that text
252 2010-09-19T20:06:32 <dreimark> too but update_item sounds also not perfect
253 2010-09-19T20:07:09 <dreimark> if you don't care it removes groups while you add acls
254 2010-09-19T20:07:29 <ThomasWaldmann> you have to call it correctly :)
255 2010-09-19T20:08:42 *** RogerHaase
256 2010-09-19T20:11:26 <ThomasWaldmann> RogerHaase: we maybe have solved theming of smileys
257 2010-09-19T20:15:12 <RogerHaase> Moin
258 2010-09-19T20:16:09 <ThomasWaldmann> (just use spans with some class, do image by css)
259 2010-09-19T20:16:54 <dreimark> RevisionNumberMismatchError: You cannot create a revision with revno None. Your revno must be greater than the item's last revision, which is -1.
260 2010-09-19T20:17:45 <ThomasWaldmann> you can't directly use None, it is just the default value to trigger determination of the correct value
261 2010-09-19T20:18:34 <RogerHaase> That will be easy and flexible. You do as background image with padding on the text and avoid the annoying line breaks as well.
262 2010-09-19T20:18:57 <ThomasWaldmann> annoying linebreaks?
263 2010-09-19T20:19:13 <dreimark> linebreaks ?
264 2010-09-19T20:19:30 <dreimark> how and where
265 2010-09-19T20:21:35 <RogerHaase> Isn't there sometimes a problem when the icon preceding an anchor is on the far right of a line and the link wraps to the next line? (I should have said wrap, not break).
266 2010-09-19T20:22:56 <ThomasWaldmann> that's a different thing (not a smiley)
267 2010-09-19T20:23:08 <ThomasWaldmann> but it is also done by css
268 2010-09-19T20:25:04 <RogerHaase> OK, you are thinking of an empty span tag for a smiley?
269 2010-09-19T20:29:08 <ThomasWaldmann> maybe even :-) inside the span
270 2010-09-19T20:29:15 <ThomasWaldmann> or the smiley name
271 2010-09-19T20:29:39 * ThomasWaldmann hacks the smiley converter to do that
272 2010-09-19T20:34:05 <ThomasWaldmann> hmm
273 2010-09-19T20:35:12 <ThomasWaldmann> that hack just works for html output converter
274 2010-09-19T20:37:52 <ThomasWaldmann> waldi: ^^
275 2010-09-19T20:38:23 <ThomasWaldmann> waldi: which kind of domtree object do you think is appropriate as smiley converter output?
276 2010-09-19T20:39:18 <waldi> image
277 2010-09-19T20:39:55 <ThomasWaldmann> even if we don't know whether the output converter will really make one?
278 2010-09-19T20:40:35 <ThomasWaldmann> i basically want to pass through smiley_markup and smiley_name,
279 2010-09-19T20:40:47 <ThomasWaldmann> and the output converter then decides what to make from that
280 2010-09-19T20:41:18 <ThomasWaldmann> for html out, it will be <span class="smiley_name">smiley_markup</span> then
281 2010-09-19T20:41:58 <ThomasWaldmann> some text plain out could also just emit smiley_markup
282 2010-09-19T20:42:45 <ThomasWaldmann> or some funny unicode char :)
283 2010-09-19T20:45:02 <ThomasWaldmann> waldi: ^^
284 2010-09-19T20:45:49 <waldi> puh
285 2010-09-19T21:08:00 <ThomasWaldmann> >>> smiley_conv
286 2010-09-19T21:08:01 <ThomasWaldmann> <MoinMoin.converter2.everything.Converter object at 0xa2038cc>
287 2010-09-19T21:08:04 <ThomasWaldmann> huh?
288 2010-09-19T21:11:15 <ThomasWaldmann> ah found it :) syntax error in the smiley conv and it takes the everything conv
289 2010-09-19T21:13:21 <ThomasWaldmann> if noone has a better idea for some appropriate element, I'll create that span in the dom tree
290 2010-09-19T21:27:48 <dreimark> ThomasWaldmann: acl None makes my problem
291 2010-09-19T21:29:28 <ThomasWaldmann> yeah, that is not a valid acl
292 2010-09-19T21:30:12 <ThomasWaldmann> TheSheep: RogerHaase: the class for the span, currently it is "smile" for :)
293 2010-09-19T21:30:25 <ThomasWaldmann> maybe better class="icon smile"?
294 2010-09-19T21:30:47 <TheSheep> moin-smile
295 2010-09-19T21:30:51 <RogerHaase> or moin-smile
296 2010-09-19T21:31:00 <xorAxAx> or maybe even moin-smile
297 2010-09-19T21:31:12 <ThomasWaldmann> :)
298 2010-09-19T21:31:20 <ThomasWaldmann> not 2 classes?
299 2010-09-19T21:31:22 <dreimark> http://hg.moinmo.in/moin/2.0-dev/file/0a28af6360a3/MoinMoin/security/_tests/test_security.py#l275
300 2010-09-19T21:31:31 <ThomasWaldmann> would maybe make css simpler?
301 2010-09-19T21:32:31 <dreimark> looks like None introduces a special rule different to u''
302 2010-09-19T21:33:19 <ThomasWaldmann> it is a difference whether an item has no acl or an empty acl
303 2010-09-19T21:33:49 <dreimark> ok
304 2010-09-19T21:34:59 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9993:98d960404052 2.0-dev/MoinMoin/converter2/smiley.py:
305 2010-09-19T21:34:59 <CIA-36> smiley converter: emit <span class="moin-SMILEY_NAME">SMILEY_MARKUP</span>
306 2010-09-19T21:34:59 <CIA-36> fixed some comments referring to <object> appropriately
307 2010-09-19T21:34:59 <CIA-36> fixed wrong year bug in (c)
308 2010-09-19T21:35:17 <ThomasWaldmann> can i haz some css for that?
309 2010-09-19T21:37:56 <dreimark> ThomasWaldmann: http://paste.pocoo.org/show/264364/
310 2010-09-19T21:38:11 <dreimark> 5 failed, 143 passed
311 2010-09-19T21:38:24 <ThomasWaldmann> (see the diff of the changeset for icon w/h)
312 2010-09-19T21:39:13 <ThomasWaldmann> dreimark: 300 "is not None" ?
313 2010-09-19T21:40:50 <ThomasWaldmann> same for 383
314 2010-09-19T21:41:25 <dreimark> indeed
315 2010-09-19T21:41:58 <dreimark> :)
316 2010-09-19T21:43:57 <ThomasWaldmann> btw, moin is getting nearer to that picture that juergen has drawn 7y ago or so
317 2010-09-19T21:44:22 <dreimark> s/nearer/closer ?
318 2010-09-19T21:44:32 <ThomasWaldmann> ehrm, yeah
319 2010-09-19T21:47:23 <ThomasWaldmann> http://en.wikipedia.org/wiki/Moin_moin
320 2010-09-19T21:48:17 <dreimark> typo
321 2010-09-19T21:48:34 <dreimark> -n
322 2010-09-19T22:02:17 <CIA-36> Reimar Bauer <rb.proj AT googlemail DOT com> default * 9994:a0b8065c3ab1 2.0-dev/MoinMoin/ (4 files in 3 dirs): tests: killed create_item and append_item. All related tests refactored for the successor update_item.
323 2010-09-19T22:02:26 <dreimark> btw. now only 6 left for the next magic cs
324 2010-09-19T22:03:40 <dreimark> 10.x years nearly 10K cs
325 2010-09-19T22:06:04 <ThomasWaldmann> no, 5y 10k cs
326 2010-09-19T22:08:31 <dreimark> better rate :)
327 2010-09-19T22:08:55 <dreimark> I can't run the testwiki I get a KeyError: 'sha1'
328 2010-09-19T22:10:43 <dreimark> ThomasWaldmann: does it run for you?
329 2010-09-19T22:11:08 <dreimark> any new item gives an error
330 2010-09-19T22:11:31 <dreimark> http://test.moinmo.in:8080/Exea
331 2010-09-19T22:13:09 <ThomasWaldmann> oops
332 2010-09-19T22:14:07 <ThomasWaldmann> ah, clear
333 2010-09-19T22:18:39 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9995:96037177782e 2.0-dev/MoinMoin/items/__init__.py: caching of internal_representation: fix for non-existing items
334 2010-09-19T22:18:40 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9996:11106c185614 2.0-dev/MoinMoin/ (4 files in 3 dirs): merged main
335 2010-09-19T22:21:32 <dreimark> hmm Disallow can that use a regex. it is strange that one could create an item /+convert/foo
336 2010-09-19T22:22:44 <dreimark> no, not with that example that gives another traceback
337 2010-09-19T22:23:04 <dreimark> but +admin/foo one can create an item
338 2010-09-19T22:23:50 <ThomasWaldmann> ?
339 2010-09-19T22:24:41 <dreimark> http://localhost:8080/+convert/foo
340 2010-09-19T22:24:57 <dreimark> http://localhost:8080/+admin/foo
341 2010-09-19T22:25:15 <ThomasWaldmann> that is another bug
342 2010-09-19T22:25:42 <ThomasWaldmann> (and somehow invalid as you need to give a mimetype to convert to)
343 2010-09-19T22:30:14 <ThomasWaldmann> noone going for the css?
344 2010-09-19T22:30:56 <dreimark> i guess TheSheep diofeher or RogerHaase is offline
345 2010-09-19T22:31:03 <TheSheep> ThomasWaldmann: going for css?
346 2010-09-19T22:31:20 <ThomasWaldmann> (21:35) < ThomasWal> can i haz some css for that?
347 2010-09-19T22:31:34 <TheSheep> ThomasWaldmann: that does't explain anything
348 2010-09-19T22:31:51 <TheSheep> +n
349 2010-09-19T22:32:17 <ThomasWaldmann> well, for the smileys
350 2010-09-19T22:32:29 <ThomasWaldmann> but maybe i do it myself, faster than explaining
351 2010-09-19T22:34:00 <RogerHaase> still working on css, had trouble with pocoo example
352 2010-09-19T22:34:47 <ThomasWaldmann> RogerHaase: on the smiley stuff?
353 2010-09-19T22:35:12 <RogerHaase> positioned smiley over text; trying font-size: 0px; but safari and chrome leave 1 pixel artifact
354 2010-09-19T22:35:57 <RogerHaase> .moin-smile { background-image:url('/static/images/smileys/smile.png'); background-repeat: no-repeat;
355 2010-09-19T22:35:57 <RogerHaase> font-size: 0px; padding-top: 16px; padding-left: 16px; overflow: hidden; }
356 2010-09-19T22:38:05 <CIA-36> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 9997:3fd31812b9ee 2.0-dev/MoinMoin/ (50 files in 2 dirs): move smileys to theme
357 2010-09-19T22:38:42 <ThomasWaldmann> RogerHaase: http://paste.pocoo.org/show/264385/ < tried this?
358 2010-09-19T22:39:08 <ThomasWaldmann> (from TheSheep )
359 2010-09-19T22:40:11 <RogerHaase> will try
360 2010-09-19T22:45:58 <RogerHaase> seems to have about 20px of padding on right of icon on all browers
361 2010-09-19T22:47:41 <RogerHaase> works better if I delete: display: inline-block; -- still testing
362 2010-09-19T22:48:56 <dreimark> test wiki looks like very slow
363 2010-09-19T22:50:20 <ThomasWaldmann> retry
364 2010-09-19T22:50:49 <RogerHaase> IE8 and FF ok, Safari and Chrome (Iron) have same 1 pixel artifact on lower-right of smiley that I had with font-size: 0px
365 2010-09-19T22:52:46 <dreimark> ThomasWaldmann: http://test.moinmo.in:8080/ASA Internal Server Error
366 2010-09-19T22:54:51 <ThomasWaldmann> yeah, now at tip-1
367 2010-09-19T23:06:55 <RogerHaase> My error on TheSheep's css, it works great when I remember to comment out all the other variations I tried. :)
368 2010-09-19T23:10:33 <ThomasWaldmann> ok, make a changeset :)
369 2010-09-19T23:12:53 <ThomasWaldmann> btw, maybe rather use -3000px
370 2010-09-19T23:13:12 <ThomasWaldmann> 2000 is maybe to near at current displays
371 2010-09-19T23:15:02 <RogerHaase> Maybe even higher, is there a limit?
372 2010-09-19T23:15:31 <ThomasWaldmann> no idea
373 2010-09-19T23:16:00 <ThomasWaldmann> 4000? should be enough for 2 fullhd cinerama displays :)
374 2010-09-19T23:16:16 <ThomasWaldmann> not that anyone would run a that big browser window, but...
375 2010-09-19T23:18:56 <RogerHaase> http://reference.sitepoint.com/css/text-indent recomments 9999px or 999em as limit or Opera will ignore higher.
