1 2010-03-14T07:56:41 *** y84
2 2010-03-14T09:23:11 *** kursor
3 2010-03-14T09:24:08 *** kursor
4 2010-03-14T12:46:45 *** kursor
5 2010-03-14T13:23:27 *** kursor
6 2010-03-14T13:30:05 *** kursor
7 2010-03-14T13:59:14 *** kursor
8 2010-03-14T13:59:40 *** kursor
9 2010-03-14T15:07:55 *** kursor_
10 2010-03-14T15:08:18 *** kursor
11 2010-03-14T15:08:18 *** kursor_
12 2010-03-14T15:57:38 *** kursor
13 2010-03-14T16:24:48 *** moin3
14 2010-03-14T16:26:18 *** moin3
15 2010-03-14T16:26:45 <moin5> hi guys
16 2010-03-14T16:27:03 <ThomasWaldmann> hi
17 2010-03-14T16:27:47 <moin5> ThomasWaldmann, i'm the PagePostEvent guy
18 2010-03-14T16:28:17 <ThomasWaldmann> /nick works
19 2010-03-14T16:28:27 <moin5> have you time to spent for a little discussion on this topic?
20 2010-03-14T16:28:37 <ThomasWaldmann> yes
21 2010-03-14T16:29:26 <moin5> oh yes...i have choosen this nick for consistency to the first time i join here
22 2010-03-14T16:29:58 <ThomasWaldmann> X)
23 2010-03-14T16:31:30 *** moin5
24 2010-03-14T16:32:21 *** fd
25 2010-03-14T16:38:30 <misterfd> ThomasWaldamm, if you remenber, one week before today, we were on this http://paste.pocoo.org/show/186403/. The chat log is http://moinmo.in/MoinMoinChat/Logs/moin-dev/2010-03-06
26 2010-03-14T16:40:10 <misterfd> you ask me to find a good definition how PagePostSaveEvent shall behave in case of are multiple handler on this event
27 2010-03-14T16:40:42 <ThomasWaldmann> yes?
28 2010-03-14T16:41:41 <misterfd> i have analyzed the whole code of MoinMoin and i find a good behaviour definition in PageEditor.py in PagePreSaveEvent
29 2010-03-14T16:43:23 <misterfd> http://docs.moinmo.in/moin/1.9/ between #1101 #1108
30 2010-03-14T16:43:29 <misterfd> uh
31 2010-03-14T16:43:56 <misterfd> http://docs.moinmo.in/moin/1.9/MoinMoin.PageEditor-pysrc.html is better ;-)
32 2010-03-14T16:45:18 <misterfd> now i try to paste this definition in pseudo code
33 2010-03-14T16:45:31 <misterfd> if handler1.isNotOk() V handler2.isNotOk() V handler3.isNotOk() V ... : sendEditor(with annotated errors or warn messages) else: sendPage(with msg)
34 2010-03-14T16:45:33 <misterfd> ok
35 2010-03-14T16:47:26 <ThomasWaldmann> that present code looks a bit crappy, so I am not sure it is too helpful
36 2010-03-14T16:47:55 <misterfd> http://paste.pocoo.org/show/189526/
37 2010-03-14T16:48:26 <misterfd> a little little little bit crappy!!! :-D
38 2010-03-14T16:50:10 <misterfd> it's something similar to the "for result in results" in Page.py
39 2010-03-14T16:50:27 <misterfd> for PagePreSaveEvent, of course
40 2010-03-14T16:50:33 <ThomasWaldmann> (i meant moin's present PagePreSaveEvent)
41 2010-03-14T16:51:19 <ThomasWaldmann> also, what is "OK" and what not?
42 2010-03-14T16:52:04 <misterfd> sorry, what do yo mean with "(i meant moin's present PagePreSaveEvent)"?
43 2010-03-14T16:53:30 <ThomasWaldmann> that it is a bit crappy at the event result processing place
44 2010-03-14T16:53:44 <misterfd> what is OK and what not, depends by handler
45 2010-03-14T16:54:21 <ThomasWaldmann> yes, but what does OK (or not OK) have as consequence?
46 2010-03-14T16:55:52 <ThomasWaldmann> and is it really OR?
47 2010-03-14T16:56:12 <ThomasWaldmann> iirc I suggested creating a wiki page to get more comments
48 2010-03-14T16:57:19 <misterfd> the consequence is to send PageEditor again. My proposal is something like MoinMoin handles conflict version of page, but general purpose
49 2010-03-14T16:58:43 <misterfd> this events "will be used" to perform check on page before save it
50 2010-03-14T16:58:58 <misterfd> and you can avoid save with the pagePreSave event
51 2010-03-14T16:59:31 <misterfd> yes is reallty or like in PagePreSave handler in Page.py
52 2010-03-14T17:00:04 <misterfd> MoinMoin performs a for cycle and then check if there are msgs or not
53 2010-03-14T17:00:30 <misterfd> it's enough at least one Abort message to abort save process, no?
54 2010-03-14T17:02:14 <ThomasWaldmann> the question is whether a check like that is generally useful
55 2010-03-14T17:03:06 <ThomasWaldmann> also, the processing in the existing code is a bit crappy, so don't just duplicate that
56 2010-03-14T17:04:22 <misterfd> do you mean "conflict version handler" code?
57 2010-03-14T17:05:32 <misterfd> BTW, the generally usefullness is due to the fact that the checker can be implemented in event plugin
58 2010-03-14T17:06:28 <misterfd> you can perform all the checks you want to perform/implement to avoid the Page View() action at the and of editing process
59 2010-03-14T17:06:41 <misterfd> let me show you an example
60 2010-03-14T17:07:06 <misterfd> SpellChecker
61 2010-03-14T17:08:01 <misterfd> at PagePreSaveEvent (before save effectively) my plugin perform SpellCheck control
62 2010-03-14T17:08:35 <misterfd> my SpellChecker reveal that i wrote ThmasWaldmann
63 2010-03-14T17:09:53 <misterfd> the PagePostSaveEvent will check thant SpellChecker has returned a Abort() event result
64 2010-03-14T17:11:07 <misterfd> finally the code in edit.py will re-send PageEditor with a message like "i have found some errors: ..."
65 2010-03-14T17:13:13 <ThomasWaldmann> how do you know the returncode of presave in the postsave handler?
66 2010-03-14T17:14:11 <ThomasWaldmann> and, if it is based on the presave returncode, do we need postsave event at all?
67 2010-03-14T17:15:07 <misterfd> cookie?
68 2010-03-14T17:16:27 <ThomasWaldmann> it is same request
69 2010-03-14T17:16:29 <misterfd> we need a postsave to give a chance to correct error checker have founded
70 2010-03-14T17:16:55 <misterfd> has founded*
71 2010-03-14T17:17:12 <ThomasWaldmann> found :)
72 2010-03-14T17:17:24 <misterfd> :-)
73 2010-03-14T17:19:45 <ThomasWaldmann> btw, emitting content in postsave might be a problem
74 2010-03-14T17:20:11 <misterfd> ok
75 2010-03-14T17:20:15 <ThomasWaldmann> e.g. if your postsave spellchecker eventhandler sends the editor again so you can fix some typos,
76 2010-03-14T17:20:29 <misterfd> ok i trust you
77 2010-03-14T17:20:39 <ThomasWaldmann> that could conflict with another postsave eventhandler also wanting to do such a thing
78 2010-03-14T17:21:21 <misterfd> ok
79 2010-03-14T17:22:04 <misterfd> my handler send nothing
80 2010-03-14T17:22:20 <misterfd> is the code in edit.py send editor again
81 2010-03-14T17:23:45 <y84> ThomasWaldmann: which seems to imply that plugins in general shouldn't send_page
82 2010-03-14T17:23:54 <misterfd> http://paste.pocoo.org/show/189526/ it's the definition of behaviour in edit.py, when edit.py collects all the eventResult
83 2010-03-14T17:27:00 <misterfd> it's the same of except branch at the end of edit.py, when MoinMoin catch EditConflict exception
84 2010-03-14T17:27:47 <misterfd> in this case MoinMoin re-sends editor according with a specific reason "edit conflict"
85 2010-03-14T17:29:28 <misterfd> my proposal is to add PostSave event to extend this behaviour for general purpose e.g. spellcheck, length check
86 2010-03-14T17:30:04 <misterfd> ThomasWaldmann, note that this checkers are not costraints
87 2010-03-14T17:30:37 <misterfd> you can save, even if page is not valid
88 2010-03-14T17:30:51 <misterfd> wiki philosophy is completely preserved
89 2010-03-14T17:35:56 *** y84
90 2010-03-14T17:37:11 <ThomasWaldmann> yes, but where is decision taken?
91 2010-03-14T17:37:36 <ThomasWaldmann> if you just have that postsave stuff to do some stuff that has been decided in presave already, it is pointless
92 2010-03-14T17:39:39 <misterfd> ok, this is a good point
93 2010-03-14T17:40:21 <ThomasWaldmann> (this is maybe(?) the reason why there is no PostSave, not sure)
94 2010-03-14T17:44:14 <misterfd> ok i my hed
95 2010-03-14T17:44:44 <misterfd> head* postsave is used only t avoid Page.send_page()
96 2010-03-14T17:45:57 <ThomasWaldmann> waldi: we need to tear the DOM SOC projects into multiple projects, otherwise they look like one
97 2010-03-14T17:46:51 <misterfd> ThomasWaldmann: for you is not enough to insert a postsave event in the main-line
98 2010-03-14T17:47:28 <misterfd> effectively postsave seems an appendix of presave
99 2010-03-14T17:47:30 <ThomasWaldmann> waldi: i still think that dividing them by format and keeping forward and reverse direction in one project might be better than otherwise, because of roundtripping
100 2010-03-14T17:47:48 <waldi> well, okay
101 2010-03-14T17:48:21 <ThomasWaldmann> waldi: can you estimate it, is 1x forward + 1x backwards == SOC scale?
102 2010-03-14T17:48:35 <ThomasWaldmann> (for a normally talented student)
103 2010-03-14T17:49:38 <ThomasWaldmann> and, can we do the backwards stuff or are there fundamental blockers?
104 2010-03-14T17:53:29 <waldi> i would like to pack the wiki-like formats. the backward conversion is rather similar
105 2010-03-14T17:54:35 <waldi> yes, we can do backwards stuff, it is just not used yet
106 2010-03-14T17:54:43 <ThomasWaldmann> ok, can you just update the wiki page with your bundling suggestion? maybe use multiple sub-headlines, one per project.
107 2010-03-14T17:54:48 <waldi> yep
108 2010-03-14T17:57:01 <ThomasWaldmann> btw, general security-related question about html/js/java/flash etc.:
109 2010-03-14T17:58:00 <ThomasWaldmann> how easy is it to create a full http (post) request, including some specific headers (cookie) and some specific form values and send the request to ANOTHER site?
110 2010-03-14T17:59:09 <ThomasWaldmann> (so that cookie value seen by target site == some form field value seen by target site)
111 2010-03-14T18:00:06 <waldi> html: not at all, it is no programming language
112 2010-03-14T18:00:57 <waldi> js: can do any request it wants, but only access the information of the current security domain
113 2010-03-14T18:01:28 <misterfd> ThomasWaldmann: i see you are busy now, can i join tomorrow to finish this discussion about PostSave? will you be here tomorrow?
114 2010-03-14T18:02:05 <waldi> java/flash: can at most read informations from current domain enforced by browser, can usually connect to the home server but nowhere else (AFAIK)
115 2010-03-14T18:03:45 <ThomasWaldmann> waldi: it is NOT needed to read an existing cookie, just to make it send a specific cookie value to another site
116 2010-03-14T18:04:16 <ThomasWaldmann> misterfd: make a wiki page
117 2010-03-14T18:04:49 <ThomasWaldmann> misterfd: find misc. scenarios
118 2010-03-14T18:05:01 <ThomasWaldmann> misterfd: find their needs
119 2010-03-14T18:05:11 <ThomasWaldmann> misterfd: find a system that can do it
120 2010-03-14T18:07:03 <misterfd> ok
121 2010-03-14T18:07:15 <misterfd> thank you ThomasWaldmann
122 2010-03-14T18:09:06 <ThomasWaldmann> (if you find ways to cleanup existing PageEditor stuff, that is welcome, too :)
123 2010-03-14T18:09:56 <misterfd> :-) ok i will try to
124 2010-03-14T18:10:39 <misterfd> last question: what are the criteria wich by discussion on this irc channel are logged?
125 2010-03-14T18:10:48 <misterfd> all the discussion?
126 2010-03-14T18:11:27 <misterfd> one of you select some that seems intesting
127 2010-03-14T18:11:30 <misterfd> ?
128 2010-03-14T18:12:50 <misterfd> this discussion will be logged?
129 2010-03-14T18:15:22 <ThomasWaldmann> all is logged here
130 2010-03-14T18:19:06 <ThomasWaldmann> waldi: that does undecided mean for pygments/dom stuff?
131 2010-03-14T18:19:49 <waldi> the one who wants it can have it
132 2010-03-14T18:20:40 <misterfd> ok salut
133 2010-03-14T18:21:36 <ThomasWaldmann> cu misterfd
134 2010-03-14T18:22:35 <ThomasWaldmann> waldi: so, it is 2 projects now + time filling stuff for someone?
135 2010-03-14T18:23:03 <waldi> yes
136 2010-03-14T18:23:06 <misterfd> ThomasWaldmann: sorry, cu stands for?
137 2010-03-14T18:23:31 <ThomasWaldmann> "see you"
138 2010-03-14T18:24:22 <misterfd> X-) ok
139 2010-03-14T18:24:25 <misterfd> cu!
140 2010-03-14T18:24:36 *** misterfd
141 2010-03-14T18:54:43 <ThomasWaldmann> waldi: it currently sounds like it is 1 SOC project that has multiple parts, is that intended?
142 2010-03-14T18:56:28 <waldi> no, it is not
143 2010-03-14T19:02:39 <ThomasWaldmann> ok, i clarified it
144 2010-03-14T19:15:45 <ThomasWaldmann> waldi: shall we add the "image converter" to the part-time projects?
145 2010-03-14T19:27:10 <dreimark> ThomasWaldmann: look at edit moin it could be used to send cookies to another site too
146 2010-03-14T19:34:56 <ThomasWaldmann> dreimark: we are talking about web content
147 2010-03-14T19:38:19 <dreimark> oh, ok
148 2010-03-14T20:52:50 *** moin6
149 2010-03-14T22:37:49 *** grzywacz
150
