1 2012-05-18T00:36:16 *** dwcramer
2 2012-05-18T00:43:13 *** dwcramer
3 2012-05-18T01:00:07 *** dwcramer
4 2012-05-18T01:40:12 *** dwcramer
5 2012-05-18T05:41:22 *** dwcramer
6 2012-05-18T09:12:52 *** MattMaker
7 2012-05-18T10:44:10 *** MattMaker
8 2012-05-18T10:46:49 *** MattMaker
9 2012-05-18T10:58:20 *** greg_f
10 2012-05-18T11:02:28 <ThomasWaldmann> moin
11 2012-05-18T11:05:26 <xiaq> moin
12 2012-05-18T11:05:46 * xiaq dumping thoughts on http://etherpad.osuosl.org/moin-render-method
13 2012-05-18T11:06:37 * xiaq almost dumped all of them...
14 2012-05-18T11:08:27 <xiaq> ThomasWaldmann: plz comment on http://etherpad.osuosl.org/moin-render-method
15 2012-05-18T11:08:31 <xiaq> DasIch: ^^
16 2012-05-18T11:08:43 * xiaq found spy is not here...
17 2012-05-18T11:09:26 <xiaq> MaikuMori: have a look there too, i wrote down some thoughts about plugins at the end of the EP
18 2012-05-18T11:14:29 <MaikuMori> ok
19 2012-05-18T11:31:40 *** jaiditya
20 2012-05-18T11:48:05 <MaikuMori> Got my google package
21 2012-05-18T11:50:51 <waldi> nice. what is in it this year?
22 2012-05-18T11:54:53 <breton> no
23 2012-05-18T11:55:12 <breton> don't spoil
24 2012-05-18T11:55:21 <MaikuMori> :D
25 2012-05-18T11:56:55 <breton> I want to see that awesome box and its contents by myself when it comes
26 2012-05-18T11:57:12 <MaikuMori> I was tad bit disapointed :(
27 2012-05-18T11:57:40 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5837:fb2a2ecd35c2 1.9/MoinMoin/action/rss_rc.py: [bugfix] action/rss_rc.py: proper rev comparison and diff generation when diffs=1
28 2012-05-18T11:57:40 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5838:7dce09f5edd4 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: preventing double escape - looks like XML generator performs escaping already when needed.
29 2012-05-18T11:57:40 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5839:d415cb95645c 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: proper links to appropriate diffs when ddiffs=1 and proper link/content for log entry related to new page.
30 2012-05-18T11:57:41 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5840:4a982b04461d 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: new option - lines - set max lines for diff/content shown in RSS.
31 2012-05-18T11:57:42 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5841:8b3c4b85921b 1.9/MoinMoin/ (action/rss_rc.py config/multiconfig.py): action/rss_rc.py: param parsing rewritten to support retrieval of default values from wiki configuration.
32 2012-05-18T11:57:43 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5842:97e36f417f6e 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: preserving show_names configuration option.
33 2012-05-18T11:57:44 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5843:89fbca5396a1 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: proper support of rename, revert, page creation.
34 2012-05-18T11:57:44 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5844:3d997b1235fe 1.9/MoinMoin/ (action/rss_rc.py config/multiconfig.py): action/rss_rc.py: added ability to show attachment-related actions in RSS.
35 2012-05-18T11:57:47 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5845:0e1b4214318f 1.9/MoinMoin/ (action/rss_rc.py config/multiconfig.py): action/rss_rc.py: added ability to set pagename-based filtering for RSS.
36 2012-05-18T11:57:47 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5846:a1054c9e7aae 1.9/MoinMoin/action/rss_rc.py: [bugfix] action/rss_rc.py: dealing with mixing unicode and non-unicode strings.
37 2012-05-18T11:57:47 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5847:2191676ba6fe 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: comment insertion code changed in flavor of using named placeholders.
38 2012-05-18T11:57:48 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5848:468e63254a4a 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: forcing unicode for comment strings.
39 2012-05-18T11:57:48 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5849:f7a570f3c1cb 1.9/MoinMoin/theme/__init__.py: theme: added RC RSS link for the front page, added page history RSS link for all other pages.
40 2012-05-18T11:58:00 <breton> oh, esyr
41 2012-05-18T11:58:10 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5850:8b4f02226670 1.9/MoinMoin/ (config/multiconfig.py theme/__init__.py): Option which gives ability to turn off addition of page history RSS feed link added.
42 2012-05-18T11:58:10 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5851:ab0bc2441518 1.9/MoinMoin/action/rss_rc.py: action/rss_rc.py: Retrieval of configuration options changed from getattr() since they are in config now already. Last rev which does not depend on configuration options presence: 468e63254a4a.
43 2012-05-18T11:58:10 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5852:3ae216a66dd8 1.9/docs/CHANGES: docs/CHANGES: information related rss_rc action update added.
44 2012-05-18T11:58:10 <CIA-146> Eugene Syromyatnikov <evgsyr@gmail.com> default * 5853:e1a4caec2513 1.9/MoinMoin/action/rss_rc.py: [minor] rss_rc: subaction in link for ATTNEW feed entries changed from get to view as less surprising.
45 2012-05-18T11:58:22 <eSyr-ng> hmmm.
46 2012-05-18T12:02:56 <eSyr-ng> (google's fedex delivery is fastest post delivery i ever observed)
47 2012-05-18T12:17:29 <breton> eSyr-ng: 1 month is fast?
48 2012-05-18T12:17:42 * breton have never used post for delivery
49 2012-05-18T12:18:31 <eSyr> breton: two days.
50 2012-05-18T12:20:45 <breton> eSyr: didn't they send it right after the acceptance announcement?
51 2012-05-18T12:21:25 * eSyr comparing with model m delivery where it took 20 days to transport it from aeroport to post office, about 20 kilometers
52 2012-05-18T12:21:58 <eSyr> breton: i'm mentor, so i recieved only t-shirt after end of gsoc :)
53 2012-05-18T12:22:04 <eSyr> *received
54 2012-05-18T12:52:22 <ThomasWaldmann> MaikuMori: eSyr: updated http://etherpad.osuosl.org/moin-render-method 46
55 2012-05-18T12:57:47 <eSyr-ng> ThomasWaldmann: well, currently i can't see any usecases where additional meta for render/behaviour changes is needed.
56 2012-05-18T12:58:04 <eSyr-ng> (except pastebin, maybe)
57 2012-05-18T12:58:28 <eSyr-ng> ThomasWaldmann: everything else can be successfully handled by mime.
58 2012-05-18T12:59:10 <ThomasWaldmann> we already had that problem for wiki items. highlight vs show.
59 2012-05-18T12:59:58 <ThomasWaldmann> e.g. if you have some docbook xml content, you can either render it or show the highlighted xml
60 2012-05-18T13:00:05 <eSyr-ng> ThomasWaldmann: these two just two separate Item methods, no?
61 2012-05-18T13:00:24 <eSyr-ng> i mean, in moin1.x we alse have multiple methods.
62 2012-05-18T13:00:25 <ThomasWaldmann> currently one can decide by choosing different views, but there is no way to tell the default view in metadata
63 2012-05-18T13:00:30 <eSyr-ng> (not precisely, but)
64 2012-05-18T13:00:40 <eSyr-ng> ah, see.
65 2012-05-18T13:01:22 <eSyr-ng> well, i think this should be configured the same way as other defaults: some setting in wiki config and some setting in meta.
66 2012-05-18T13:01:23 <ThomasWaldmann> so, that is a problem that even exists without special views like pastebin, tickets, etc.
67 2012-05-18T13:01:40 <eSyr-ng> ticket is not special view :)
68 2012-05-18T13:01:49 <eSyr-ng> text/x-moin=-ticket or so
69 2012-05-18T13:01:56 <eSyr-ng> *x-moin-ticket
70 2012-05-18T13:02:03 <ThomasWaldmann> ticket has markup, so THAT is the content type
71 2012-05-18T13:02:32 <ThomasWaldmann> it is only a ticket because of the special set of meta items
72 2012-05-18T13:02:36 <xiaq> eSyr-ng: that is not correct use of contenttype i guess
73 2012-05-18T13:02:43 <eSyr-ng> mmm.
74 2012-05-18T13:02:55 <eSyr-ng> xiaq: well, ok, let's start from the begginning.
75 2012-05-18T13:03:53 <xiaq> eSyr-ng: what we want is something that describe how an item's meta will look like
76 2012-05-18T13:04:07 <xiaq> plus how it will be rendered - no much to do with the content
77 2012-05-18T13:04:35 <eSyr-ng> xiaq: what's wrong if i define new syntax calling "ticket" with mime "text/x-moin-ticket", define some syntax for it (e.g, how i'm setting fields and values for them, transclusion/linking and such, may be moin- or xml- or yaml- base) and provide two methods: show (for showing) and edit (with custom editing gui to simplify the task)?
78 2012-05-18T13:04:55 <eSyr-ng> (values can have its own mime othervise text)
79 2012-05-18T13:05:27 <eSyr-ng> *yaml-based
80 2012-05-18T13:06:38 <eSyr-ng> (well, this is how i see ticket system for moin)
81 2012-05-18T13:08:03 * xiaq had bad wiki connection
82 2012-05-18T13:08:12 <eSyr-ng> (fields mostly can be meta but description and comments definitely should be part of body)
83 2012-05-18T13:09:12 <xiaq> hm
84 2012-05-18T13:12:10 <eSyr-ng> i mean, proposal reagrding meta_schemes and render_method are needed already, they trying to solve already existing important questions, but some significant part of the mentioned usecases can be solved without involving it.
85 2012-05-18T13:12:32 <eSyr-ng> *regarding
86 2012-05-18T13:13:15 <eSyr-ng> so, currently, i se only those render views: show, edit, hilight, raw.
87 2012-05-18T13:13:51 <eSyr-ng> (last two rather generic and first two stongly mime-based)
88 2012-05-18T13:14:26 <eSyr-ng> meta_schemes mostly can be determined by mime.
89 2012-05-18T13:14:39 <xiaq> eSyr-ng: well, as for ticket systems you would want some hybrid of "show" and "edit"
90 2012-05-18T13:14:48 <eSyr-ng> (currently i can't see usecases where we need different schemes with the same mime)
91 2012-05-18T13:15:59 <eSyr-ng> xiaq: well, that hybrid means "we incorporate some editing features in show method"
92 2012-05-18T13:16:12 <eSyr-ng> why not.
93 2012-05-18T13:16:41 <xiaq> eSyr-ng: hm
94 2012-05-18T13:16:56 <xiaq> eSyr-ng: i wonder why define a new *syntax* for tickets?
95 2012-05-18T13:17:13 <xiaq> why not make them just wiki markup in content?
96 2012-05-18T13:19:32 <ThomasWaldmann> well, i think we should not fix stuff to much when not needed
97 2012-05-18T13:19:58 <ThomasWaldmann> while most moin wiki users might be happy with moin markup, i know some people rather use rst or something else
98 2012-05-18T13:20:40 <ThomasWaldmann> so if we just keep the contenttype telling the (data) content's type :) we don't need to force people to use something specific
99 2012-05-18T13:21:06 <xiaq> ThomasWaldmann: i suppose eSyr-ng was suggesting x-rst-ticket etc.?
100 2012-05-18T13:21:33 <xiaq> eSyr-ng: yeah another point is that if we retain flexibility of the underlying markup for tickets
101 2012-05-18T13:21:36 <ThomasWaldmann> but, as I said: there are different ways to display the same data content. PLUS there are different ways to display the meta (and what meta we expect).
102 2012-05-18T13:21:44 <xiaq> eSyr-ng: we'll have to invent many more contenttypes...
103 2012-05-18T13:22:49 <ThomasWaldmann> i wouldn't make a lot of major/minor-type modifications. one option that could be considered is using parameters additionally to the type
104 2012-05-18T13:23:15 <ThomasWaldmann> but i am not sure about that, maybe that is abuse of the parameters. waldi, can you comment?
105 2012-05-18T13:23:48 * xiaq had really bad wifi connection
106 2012-05-18T13:23:51 <eSyr-ng> xiaq: no, by moin i meant MoinWikiEngine, not MoinWikiSyntax. I mean, in body there is somewhat like (in xml, for example): <ticket_body syntax="text/x-moin">Moin markup</ticket_body>
107 2012-05-18T13:24:08 <ThomasWaldmann> we currently use 2 parameters: charset for text content (obvious) and format=1.9 for slightly different interpretation of wiki markup (like 1.9 did)
108 2012-05-18T13:24:55 <ThomasWaldmann> eSyr-ng: and that is just backwards
109 2012-05-18T13:25:19 <ThomasWaldmann> in moin 1.9 we had #format xxx which is just the non-xml variant of what you just said
110 2012-05-18T13:25:36 <eSyr-ng> not precisely.
111 2012-05-18T13:25:39 <waldi> ThomasWaldmann: this would be some sort of parameter for the conversion, not for the start and end point
112 2012-05-18T13:25:46 <eSyr-ng> ThomasWaldmann: how comments are expected to be stored?
113 2012-05-18T13:25:57 <waldi> however, i'm not sure what different ways you mean
114 2012-05-18T13:25:59 <ThomasWaldmann> obviously you can only do that with text (xml), it is mixing meta into data
115 2012-05-18T13:26:43 <xiaq> eSyr-ng: flat or threaded, and ThomasWaldmann suggested implementing flat first to keep things simple
116 2012-05-18T13:26:48 <xiaq> eSyr-ng: http://etherpad.osuosl.org/moin-issue-tracker
117 2012-05-18T13:26:54 <ThomasWaldmann> eSyr-ng: we didn't discuss yet about comments at all and i guess this is just complicating things further
118 2012-05-18T13:26:54 <xiaq> line 7
119 2012-05-18T13:27:00 *** dwcramer
120 2012-05-18T13:27:13 <eSyr-ng> ThomasWaldmann: there's two ways: subitems and current item.
121 2012-05-18T13:27:41 <eSyr-ng> ThomasWaldmann: in the first case using mime for defining ticket content is fine, in the second i suppose it is not.
122 2012-05-18T13:28:10 <ThomasWaldmann> as the contenttype is only about the data, i think it must not also describe the meta at the same time
123 2012-05-18T13:28:49 <ThomasWaldmann> and a meta-schema would be somehow the corresponding thing for meta
124 2012-05-18T13:29:27 <xiaq> yeah, and proposed used_as (or render_method or whatever) should also affect the overall rendering of an item
125 2012-05-18T13:29:35 <xiaq> whilst contenttype only affects the content part...
126 2012-05-18T13:29:47 <ThomasWaldmann> this would keep the amount of types needed to n+m instead of n*m in the worst case (n*m when we mix both together)
127 2012-05-18T13:30:16 <xiaq> ThomasWaldmann: aha, i see moin-blog has the same pending decision of how to implement comments
128 2012-05-18T13:30:38 <ThomasWaldmann> xiaq: yes :)
129 2012-05-18T13:30:41 <xiaq> (is that one of the projects you officially mentor?)
130 2012-05-18T13:31:12 * ThomasWaldmann is primary mentor for branch-merge-sync and pastebin
131 2012-05-18T13:31:20 <xiaq> ok
132 2012-05-18T13:31:53 <ThomasWaldmann> (but obviously I have my fingers in elsewhere :D )
133 2012-05-18T13:33:26 <eSyr-ng> ThomasWaldmann: http://rn0.ru/show/ReczxS6huX1tztr5Joaa/
134 2012-05-18T13:34:28 <xiaq> http://etherpad.osuosl.org/moin-render-method line 48
135 2012-05-18T13:34:39 <xiaq> i think user page can be another use case of propsed "used_as"
136 2012-05-18T13:34:57 <xiaq> we can unify user meta and user homepage in this way
137 2012-05-18T13:35:23 <xiaq> in moin1 only homepage is possible, in (current) moin2 only user meta is possible
138 2012-05-18T13:36:07 <xiaq> besides, how does everyone think of the name i proposed, "used_as"?
139 2012-05-18T13:36:41 <xiaq> it is more generic than meta_schema or render_method, though may not be ideal and final
140 2012-05-18T13:36:46 <ThomasWaldmann> eSyr-ng: that is just a pseudo-xmlish way to do things you could also do in wiki markup (body stuff i mean)
141 2012-05-18T13:36:49 <eSyr-ng> xiaq: any page under homepage is user page
142 2012-05-18T13:37:01 <eSyr-ng> ThomasWaldmann: syntax is not significant.
143 2012-05-18T13:37:14 <eSyr-ng> ThomasWaldmann: as i said, it can be xml or moin or anything.
144 2012-05-18T13:37:34 <eSyr-ng> but it is separate syntax.
145 2012-05-18T13:37:41 <eSyr-ng> (well, as i see it)
146 2012-05-18T13:37:42 <ThomasWaldmann> we won't use numbers btw, but uuids
147 2012-05-18T13:37:44 <xiaq> hm, was test.moinmo.in purged recently?
148 2012-05-18T13:37:59 <ThomasWaldmann> xiaq: yes
149 2012-05-18T13:38:24 <eSyr-ng> currently i'm thinking that do custom thing based on mime is most consistent way of doing custom things :)
150 2012-05-18T13:38:24 <xiaq> ThomasWaldmann: what for? (just curious)
151 2012-05-18T13:38:59 <ThomasWaldmann> a user uploaded an exe, crashing moin. which led to me updating the site to latest code and also reinitializing index and storage.
152 2012-05-18T13:39:21 <eSyr-ng> hehe
153 2012-05-18T13:39:37 <eSyr-ng> ThomasWaldmann: moin2 has some parser for PE format already? Cool!
154 2012-05-18T13:39:59 <ThomasWaldmann> not yet, but moin knows application/octet-stream
155 2012-05-18T13:40:08 <eSyr-ng> ah, see.
156 2012-05-18T13:40:15 <ThomasWaldmann> but the user told it was text/x.moin.wiki, but it wasn't utf-8
157 2012-05-18T13:40:44 <xiaq> ThomasWaldmann: and that crashed the *entire* wiki??
158 2012-05-18T13:41:01 <eSyr-ng> xiaq: why not, it's beta.
159 2012-05-18T13:41:08 <ThomasWaldmann> ok, i think we need: contenttype, metaschema and render_as
160 2012-05-18T13:41:17 <eSyr-ng> xiaq: it can crash the whole vps and do rm -rf and stuff.
161 2012-05-18T13:41:25 <ThomasWaldmann> xiaq: no, only the front page
162 2012-05-18T13:41:48 <ThomasWaldmann> why are you wondering, did you put anything important there?
163 2012-05-18T13:42:01 <xiaq> ThomasWaldmann: nothing important, just curious :)
164 2012-05-18T13:42:39 <ThomasWaldmann> (while thinking about it, I noted that i did not update the virtualenv, so I think I'll redo that later)
165 2012-05-18T13:43:05 <xiaq> ThomasWaldmann: aha, first registered user of newborn test.moinmo.in XD
166 2012-05-18T13:45:54 <xiaq> ThomasWaldmann: http://etherpad.osuosl.org/moin-render-method line 43
167 2012-05-18T13:46:12 <xiaq> ThomasWaldmann: i propose one meta key (used_as) instead of two
168 2012-05-18T13:46:28 <xiaq> render_method and meta_schema are tightly coupled anyway
169 2012-05-18T13:47:49 * xiaq found eSyr-ng and ThomasWaldmann both using Fx on etherpad :)
170 2012-05-18T13:48:01 <xiaq> etherpad records the useragent. cool little feature
171 2012-05-18T13:50:01 *** MattMaker
172 2012-05-18T13:50:34 * xiaq read ThomasWaldmann talked about namespacing
173 2012-05-18T13:50:44 <xiaq> what is the progress of namespacing code merge?
174 2012-05-18T13:51:10 <ThomasWaldmann> there isn't much progress
175 2012-05-18T13:52:22 <xiaq> perhaps namespacing can achieve a similar goal like proposed used_as (ie. assign a default view and schema to each namespace)
176 2012-05-18T13:52:43 *** MattMaker
177 2012-05-18T13:52:47 <xiaq> the upside is that it is more explicit (present in the URL)
178 2012-05-18T13:52:53 <ThomasWaldmann> yes, but we don't have it yet, so we can't build on it
179 2012-05-18T13:52:53 <xiaq> the downside being lack of flexibility
180 2012-05-18T13:52:56 <eSyr-ng> why namespaces, we have tree structure already.
181 2012-05-18T13:53:31 <eSyr-ng> (if we want some settings to propagate on tree of entries, why not just provide some ability to do it)
182 2012-05-18T13:53:43 <breton> yay!
183 2012-05-18T13:53:50 <breton> fedex guy just called
184 2012-05-18T13:55:19 <xiaq> eSyr-ng: aha your comment got me thinking whether there are fundamental differences between namespacing and tree sructure
185 2012-05-18T13:55:27 <xiaq> ThomasWaldmann: thoughts?
186 2012-05-18T13:57:24 *** MattMaker
187 2012-05-18T13:57:30 <ThomasWaldmann> xiaq: we first had that fstab-mountpoint like thing, but we ran into all sorts of issues due to the use of NAMES there and decided it works better with separating stuff into namespaces instead of using mountpoints
188 2012-05-18T13:57:41 <eSyr-ng> xiaq: no difference, except in trees top-level item has some content.
189 2012-05-18T13:58:33 <eSyr-ng> ThomasWaldmann: but fses use names and mountpoints, why not just borrow (and use for some backends) decisions already implemented there?
190 2012-05-18T13:58:46 <eSyr-ng> (use == just rely)
191 2012-05-18T13:59:30 <ThomasWaldmann> well, i tried (had same idea as you :), but it didn't work
192 2012-05-18T13:59:31 <xiaq> ThomasWaldmann: what's that fstab-like thing? (i know fstab)
193 2012-05-18T13:59:33 <eSyr-ng> ThomasWaldmann: is there some document describing sort of issues?
194 2012-05-18T13:59:41 <eSyr-ng> (or may be link)
195 2012-05-18T13:59:54 <ThomasWaldmann> we had long discussions in brazil about that :)
196 2012-05-18T13:59:58 <eSyr-ng> xiaq: unix mounting concept
197 2012-05-18T14:00:19 <eSyr-ng> sadly i don't know brazil :(
198 2012-05-18T14:00:21 <ThomasWaldmann> iirc the main issue was that routing middleware
199 2012-05-18T14:00:36 <eSyr-ng> ehrrr, wait, i thought they're speaking portuguese.
200 2012-05-18T14:00:38 * xiaq afk for 5min
201 2012-05-18T14:01:07 <ThomasWaldmann> we discussed in english :)
202 2012-05-18T14:01:13 <xiaq> eSyr-ng: uh oh known that too. just wondering how is that used for wiki namespacing? usecase?
203 2012-05-18T14:01:16 <xiaq> ThomasWaldmann: ^^
204 2012-05-18T14:01:37 * xiaq afk for 5min, really :)
205 2012-05-18T14:02:06 <ThomasWaldmann> btw, we are getting off-topic a bit and we didn't solve the initial problem fully yet
206 2012-05-18T14:02:32 <eSyr-ng> xiaq: http://moinmo.in/MoinMoinTalks/DLR-2009-09-08#A2.0:_Storage_Middleware:_Router
207 2012-05-18T14:03:15 <eSyr-ng> ThomasWaldmann: what arguments are against move all specific stuff to mime-based as much as possible?
208 2012-05-18T14:03:20 <eSyr-ng> *moving
209 2012-05-18T14:04:48 <ThomasWaldmann> the main argument against is cleanliness (contenttype describes the data, not the meta, not the presentation)
210 2012-05-18T14:05:01 *** MattMaker
211 2012-05-18T14:05:39 <eSyr-ng> ThomasWaldmann: well, i though contenttype describes the Item.
212 2012-05-18T14:06:35 <ThomasWaldmann> e.g. if i have some song text as text/plain;charset=utf-8 that is just saying it is text, but not whether you should read it or sing it. nor it does say anything about metadata like author and date.
213 2012-05-18T14:06:41 *** MattMaker
214 2012-05-18T14:07:21 <ThomasWaldmann> eSyr-ng: we send that for the raw data if you download it
215 2012-05-18T14:10:53 *** jaiditya
216 2012-05-18T14:11:16 <xiaq> eSyr-ng: uhoh me is beginner in Deutschsprache :)
217 2012-05-18T14:12:15 <xiaq> s/Deutschsprache/Deutschesprache/
218 2012-05-18T14:14:12 <ThomasWaldmann> Deutsche Sprache :)
219 2012-05-18T14:14:23 <ThomasWaldmann> oder einfach "Deutsch"
220 2012-05-18T14:14:33 <xiaq> ThomasWaldmann: danke :)
221 2012-05-18T14:14:52 <ThomasWaldmann> i added some subitems in 51..60 to clarify what i mean
222 2012-05-18T14:15:16 <ThomasWaldmann> it pretty much looks like we are mixing two things into 1 value
223 2012-05-18T14:16:24 * ThomasWaldmann away for a while
224 2012-05-18T14:21:00 <xiaq> ThomasWaldmann: hm, i don't see why anyone will want a wiki item to be rendered as -highlight or -raw by default
225 2012-05-18T14:22:29 <xiaq> eSyr-ng: you occupied 1/4 of the precious 3-bit author address space of etherpad on http://etherpad.osuosl.org/moin-render-method :)
226 2012-05-18T14:22:48 *** MattMaker
227 2012-05-18T14:26:56 <ThomasWaldmann> xiaq: how about showing wiki source code examples?
228 2012-05-18T14:27:30 <xiaq> ThomasWaldmann: just access +raw/xxx directly
229 2012-05-18T14:27:32 <ThomasWaldmann> or rendering html vs showing html source?
230 2012-05-18T14:27:42 <xiaq> ThomasWaldmann: proposed_as only affects default rendering method
231 2012-05-18T14:27:57 <ThomasWaldmann> yes, you can do that, but the intention of the item author might be to do that by default
232 2012-05-18T14:29:17 <xiaq> ThomasWaldmann: hm
233 2012-05-18T14:29:49 <ThomasWaldmann> i think we really have 2 things here (and that problem is as old as moin :)
234 2012-05-18T14:30:01 <xiaq> ThomasWaldmann: but render_method and meta_schema are tightly coupled, which doesn't feel clean to me
235 2012-05-18T14:30:38 <ThomasWaldmann> if A has a B, it doesn't necessarily mean that there should be no B
236 2012-05-18T14:31:46 <ThomasWaldmann> btw, for stuff that should be rather downloaded than rendered, there is a similar issue
237 2012-05-18T14:31:58 <xiaq> ThomasWaldmann: hm, rephrase perhaps? (if A has a B, ...)
238 2012-05-18T14:32:35 <DasIch> wow you made a lot of progress on render_method
239 2012-05-18T14:32:36 <ThomasWaldmann> e.g. we can render zip content (will just show a table of contents of the zip), but maybe author wanted just that you download it.
240 2012-05-18T14:33:12 <ThomasWaldmann> (that "one click more" already caused quite some discussions)
241 2012-05-18T14:34:38 <ThomasWaldmann> DasIch: yes, but we are still struggling to get it clean and simple
242 2012-05-18T14:34:42 <xiaq> ThomasWaldmann: but {render_method="blog",meta_schema="ticket"} doesn't make much sense - in fact most combinations are "dumb" ones
243 2012-05-18T14:36:24 <ThomasWaldmann> but meta_schema="wiki" and render_method=raw/highlight/download/render makes sense
244 2012-05-18T14:37:10 <xiaq> ThomasWaldmann: yes that makes sense, but too many dumb possibilities doesn't feel clean to me :)
245 2012-05-18T14:37:58 <ThomasWaldmann> we don't need to use the invalid combinations :)
246 2012-05-18T14:38:39 <DasIch> we obviously need typed metadata which guarantees that invalid combinations don't occur ;)
247 2012-05-18T14:39:13 <ThomasWaldmann> you mean a meta schema?
248 2012-05-18T14:39:33 <ThomasWaldmann> or validation of values?
249 2012-05-18T14:40:00 <ThomasWaldmann> (well, in flatland, this somehow goes hand in hand)
250 2012-05-18T14:41:20 <DasIch> ThomasWaldmann: if the user can modify it we obviously need validation of some kind
251 2012-05-18T14:44:46 <ThomasWaldmann> there is storage/validation :)
252 2012-05-18T14:45:08 <ThomasWaldmann> being called, but not authoritative yet
253 2012-05-18T14:49:34 * ThomasWaldmann needs to do some other work...
254 2012-05-18T14:55:40 <xiaq> http://etherpad.osuosl.org/moin-render-method line 25
255 2012-05-18T14:55:41 <xiaq> comments?
256 2012-05-18T15:02:38 *** dwcramer
257 2012-05-18T15:19:17 *** kanha
258 2012-05-18T15:19:59 *** kanha_
259 2012-05-18T15:20:44 *** spy
260 2012-05-18T15:35:08 <eSyr-ng> ThomasWaldmann: ok, i understood.
261 2012-05-18T15:35:32 <eSyr-ng> looks like i thought that contenttype is type of item so i have no problems you're discussing :)
262 2012-05-18T15:35:45 *** spy
263 2012-05-18T15:35:48 <eSyr-ng> othervise we definitely need type of item.
264 2012-05-18T15:36:00 <eSyr-ng> you can name it meta_schema or something.
265 2012-05-18T15:36:08 <eSyr-ng> mmm, it can be named.
266 2012-05-18T15:36:46 <ThomasWaldmann> i guess you meant something like complete_item_type that says all
267 2012-05-18T15:37:12 <ThomasWaldmann> but if we do that, we would somehow have an explosion of combinations
268 2012-05-18T15:38:18 <eSyr-ng> well, ok. i want to define how item is showed only for item, not for content.
269 2012-05-18T15:38:37 <eSyr-ng> (so combinatorial explosion should take less damage)
270 2012-05-18T15:39:07 <xiaq> ThomasWaldmann: discussed much on render_method (or whatever), i'll work on metadata editor
271 2012-05-18T15:39:11 <eSyr-ng> because content is only content, it has nothing without metadata and such.
272 2012-05-18T15:39:40 <eSyr-ng> and we don't want to show content, we want to show item.
273 2012-05-18T15:39:56 <eSyr-ng> (or we can force this rule to make things simpler :) )
274 2012-05-18T15:41:17 <eSyr-ng> and, yes, we can define render(), hilight(), raw() methods, which can take content mime or not.
275 2012-05-18T15:41:53 <eSyr-ng> (well, in last several sentences i'm not very good with modalities, please apologise me)
276 2012-05-18T15:42:34 <eSyr-ng> (and my whole english is broken, currently thinking different, not in english)
277 2012-05-18T15:45:41 * xiaq gtg bbl
278 2012-05-18T15:45:54 <ThomasWaldmann> eSyr-ng: i think what you want is the stuff that is left now as "view_method"
279 2012-05-18T15:45:54 *** eSyr
280 2012-05-18T15:46:05 <ThomasWaldmann> oops
281 2012-05-18T15:48:26 *** eSyr
282 2012-05-18T15:50:48 <ThomasWaldmann> eSyr-ng: wb :) what was the last you got?
283 2012-05-18T15:51:07 <ThomasWaldmann> xiaq: cu :)
284 2012-05-18T15:52:41 <eSyr> ThomasWaldmann: you mean, me?
285 2012-05-18T15:52:54 <eSyr-ng> ThomasWaldmann: because i haven't been split.
286 2012-05-18T15:53:48 <ThomasWaldmann> ah, we have 2 of you. ok :D
287 2012-05-18T15:54:04 <ThomasWaldmann> redundant array of inexpensive IRCs
288 2012-05-18T15:56:12 <eSyr> (well, one of these ircs is quite expensive)
289 2012-05-18T15:56:36 <eSyr-ng> anyway.
290 2012-05-18T15:56:50 <eSyr-ng> not sure, whether view_method is what i meant.
291 2012-05-18T15:57:09 <eSyr-ng> i think the meta_schema is.
292 2012-05-18T15:57:37 <eSyr-ng> anyway, the issue looks settled.
293 2012-05-18T15:58:08 <eSyr-ng> so, why we can't do meta_schema mime-ish?
294 2012-05-18T15:58:33 <eSyr-ng> (and implicitly default it to content_type)
295 2012-05-18T15:59:04 <ThomasWaldmann> because metadata has not necessarily much to do with the data type?
296 2012-05-18T15:59:24 <eSyr-ng> of course.
297 2012-05-18T15:59:29 <eSyr-ng> it's type of item.
298 2012-05-18T15:59:41 <eSyr-ng> meta_scheme moin/ticket, ok.
299 2012-05-18T16:00:02 <eSyr-ng> or meta_scheme moin/blog-post.
300 2012-05-18T16:00:11 <ThomasWaldmann> isn't it all moin/* :)
301 2012-05-18T16:00:16 <eSyr-ng> no.
302 2012-05-18T16:00:24 <ThomasWaldmann> but?
303 2012-05-18T16:00:34 <eSyr-ng> also there are meta_scheme image/jpeg.
304 2012-05-18T16:00:38 <eSyr-ng> *there is
305 2012-05-18T16:01:15 <eSyr-ng> or meta_scheme text/plain
306 2012-05-18T16:01:16 <ThomasWaldmann> isn't that moin/wikiitem
307 2012-05-18T16:01:47 <eSyr-ng> mmm, everything is wiki item!
308 2012-05-18T16:01:58 <ThomasWaldmann> :)
309 2012-05-18T16:01:59 <eSyr-ng> ok, looks like another problem with termynology.
310 2012-05-18T16:02:08 <eSyr-ng> *therminology
311 2012-05-18T16:02:15 <ThomasWaldmann> almost
312 2012-05-18T16:02:24 <eSyr-ng> *terminology
313 2012-05-18T16:02:36 * ThomasWaldmann .oO(hot discussions ...)
314 2012-05-18T16:02:39 *** dwcramer
315 2012-05-18T16:03:07 <eSyr-ng> yeah, looks like i'm breaking everything with my own vision :)
316 2012-05-18T16:03:27 <ThomasWaldmann> i suggest we let that condense a bit in our minds and continue when some student is working on it
317 2012-05-18T16:04:12 *** Marchael1
318 2012-05-18T16:04:41 <ThomasWaldmann> heh. he accidentally joined here. :|
319 2012-05-18T16:08:50 <eSyr-ng> (i think i have written down something related to tickets, hmmm)
320 2012-05-18T16:22:06 *** spy
321 2012-05-18T16:37:31 *** spy
322 2012-05-18T17:26:54 *** RogerHaase
323 2012-05-18T17:41:17 *** DasIch_
324 2012-05-18T17:43:19 *** DasIch
325 2012-05-18T17:43:19 *** DasIch_
326 2012-05-18T18:30:34 <dreimark> ThomasWaldmann: eSyr-ng xiaq I think we should use the new render_method also for the user object
327 2012-05-18T18:30:56 <dreimark> there you want also different representations
328 2012-05-18T18:31:10 <dreimark> homepage, profile, etc-
329 2012-05-18T18:32:07 <eSyr-ng> "render user"
330 2012-05-18T18:32:23 <dreimark> and download ist also interesting
331 2012-05-18T18:32:34 <eSyr-ng> get?
332 2012-05-18T18:32:42 <dreimark> if you want the same setup on a different wiki, - yes
333 2012-05-18T18:33:50 <dreimark> upload for user isn't implemented yet
334 2012-05-18T18:34:11 <eSyr-ng> hm, interesting concept.
335 2012-05-18T18:34:19 <eSyr-ng> just upload instead of registering.
336 2012-05-18T18:35:00 <dreimark> I think we should look where else we can use that new concept
337 2012-05-18T18:35:35 <breton> "fork this wiki"?
338 2012-05-18T18:35:49 <breton> distributed wiki engine
339 2012-05-18T18:35:51 <breton> oh.
340 2012-05-18T18:37:30 <eSyr-ng> breton: with hg backend we already have one :)
341 2012-05-18T18:37:48 <dreimark> which is gone since the storage branch was merged
342 2012-05-18T18:37:54 <eSyr-ng> T_T
343 2012-05-18T18:38:32 <dreimark> for 1.9 i have a simple fork script (also for attachments)
344 2012-05-18T18:38:53 <eSyr-ng> well, 1.9 forking is just copying and cache flushing.
345 2012-05-18T18:39:26 <eSyr-ng> dreimark: btw, are there some notes regarding farm support in moin2?
346 2012-05-18T18:40:28 * dreimark looks
347 2012-05-18T18:41:42 <dreimark> eSyr-ng: at least index can handle search in many wikis of a farm
348 2012-05-18T18:42:00 <dreimark> http://readthedocs.org/docs/moin-20/en/latest/admin/index.html?highlight=farm
349 2012-05-18T18:43:21 <dreimark> I have not checked what all is missing for a real world useage of more than the builtin system
350 2012-05-18T18:44:07 <dreimark> the interwiki concept is also refactored for the namespace branch
351 2012-05-18T18:46:21 <dreimark> htg, bbl ~ 2h
352 2012-05-18T18:57:51 *** dwcramer
353 2012-05-18T19:02:00 *** dwcramer
354 2012-05-18T19:03:27 * breton thinking about distributed wiki with servers like github
355 2012-05-18T19:21:57 *** greg_f
356 2012-05-18T20:30:58 *** dwcramer
357 2012-05-18T20:33:14 *** dwcramer
358 2012-05-18T20:44:59 *** dwcramer
359 2012-05-18T21:02:06 *** dwcramer
360 2012-05-18T21:10:02 *** dwcramer
361 2012-05-18T21:11:12 <dreimark> breton: that needs a clean description, I don't know github too
362 2012-05-18T21:11:30 <dreimark> at least the parts different to bitbucket for that approach
363 2012-05-18T21:21:21 <dreimark> breton: the merging part does it consider extensions like, blog, tickets or paste too?
364 2012-05-18T21:27:12 *** dwcramer
365 2012-05-18T21:46:02 <dreimark> eSyr-ng: ThomasWaldmann xiaq http://etherpad.osuosl.org/moin-issue-tracker 170, 228, 251
366 2012-05-18T21:49:06 <eSyr-ng> dreimark: each github repo == each wikihub wiki.
367 2012-05-18T21:49:20 * eSyr-ng .oO( moinhub )
368 2012-05-18T21:52:48 *** spy
369 2012-05-18T21:55:08 <dreimark> the acl part is also relevant on blog, ticket and paste
370 2012-05-18T22:43:00 <RogerHaase> opened a bug report against whoosh 2.4.0 (but cannot reproduce problem): https://bitbucket.org/mchaput/whoosh/issue/262/240-traceback-on-line-117-of-codec
371 2012-05-18T22:43:41 *** kanha
372 2012-05-18T22:45:36 *** kanha
373 2012-05-18T22:55:12 <RogerHaase> maybe the above problem is related to the first item save after doing a quickinstall to get whoosh 2.4.0 installed. I did a quickinstall,but not a moin index-build -i
374 2012-05-18T23:32:20 *** spy
375 2012-05-18T23:36:40 <RogerHaase> ThomasWaldmann: fix for #191 validation errors on global index page ready for review: http://codereview.appspot.com/6218055/
376 2012-05-18T23:54:46 <ThomasWaldmann> hi roger
377 2012-05-18T23:54:57 * ThomasWaldmann currently at shackspace.de, but leaving soon
378 2012-05-18T23:56:09 *** bilal
379 2012-05-18T23:59:06 *** bilal
380