1 2014-05-31T01:12:53 *** sl33k_
2 2014-05-31T01:52:59 *** derdon
3 2014-05-31T02:08:19 *** penguinRaider
4 2014-05-31T02:08:59 *** penguinRaider
5 2014-05-31T02:51:54 *** sl33k_
6 2014-05-31T05:34:11 *** sl33k_
7 2014-05-31T05:41:02 *** sl33k_
8 2014-05-31T06:59:29 *** sl33k_
9 2014-05-31T07:10:09 *** sl33k_
10 2014-05-31T07:36:23 *** sl33k_
11 2014-05-31T08:08:58 *** randomax
12 2014-05-31T09:44:16 *** penguinRaider
13 2014-05-31T09:44:27 *** penguinRaider_
14 2014-05-31T10:51:47 *** penguinRaider_
15 2014-05-31T11:02:19 *** skathpalia
16 2014-05-31T11:02:33 <skathpalia> moin
17 2014-05-31T11:03:39 <skathpalia> dimazest, I have updated the cr https://codereview.appspot.com/93570043
18 2014-05-31T11:03:55 <skathpalia> made a minor change in notifications.py
19 2014-05-31T11:28:28 <dimazest> skathpalia: see my comments
20 2014-05-31T11:37:26 *** derdon
21 2014-05-31T11:53:04 <skathpalia> dimazest, I think which method do you prefer using fqname[1] directly or having one else statement
22 2014-05-31T11:53:25 *** derdon
23 2014-05-31T11:54:15 *** derdon
24 2014-05-31T11:55:01 <dimazest> i'm not sure
25 2014-05-31T11:55:07 <dimazest> why not to show always the id
26 2014-05-31T11:56:00 <skathpalia> If the item has name then sending name would be more helpful
27 2014-05-31T11:56:18 <skathpalia> than always sending itemid
28 2014-05-31T11:57:33 <skathpalia> I personally prefer the else statement method as there are only 3 unique identifiers (name, itemid, revid)
29 2014-05-31T11:57:55 <dimazest> but this can change
30 2014-05-31T11:57:58 <dimazest> item_name=fqname[2]
31 2014-05-31T11:58:11 <dimazest> what is fqname[2]? is it a name?
32 2014-05-31T11:58:40 <skathpalia> it depends on field which is fqname[1]
33 2014-05-31T11:58:41 <dimazest> actually, you can avoid if else by using a dictionary
34 2014-05-31T11:58:49 <dimazest> ok
35 2014-05-31T11:59:14 <dimazest> so, is it important for user to know how his item was change by name or by item_id?
36 2014-05-31T11:59:25 <dimazest> maybe we can drop that part of notification
37 2014-05-31T12:00:42 <skathpalia> But he has to be notified that which item has been modified
38 2014-05-31T12:00:54 <skathpalia> We have to give either name or itemis
39 2014-05-31T12:00:58 <skathpalia> *itemid
40 2014-05-31T12:01:54 <dimazest> but you always show fqname[2]
41 2014-05-31T12:02:36 <dimazest> why not to use the original message form
42 2014-05-31T12:03:00 <dimazest> without "with ITEMID" or "with name"
43 2014-05-31T12:03:56 <skathpalia> But the user should be notified by which means he can access the item
44 2014-05-31T12:04:45 <skathpalia> I mean if ITEMID is mentioned then he can use this <domain>/@itemid/some_ITEMID and if it is name then <domain>/name
45 2014-05-31T12:05:45 <skathpalia> I think the dictionary method is better for this
46 2014-05-31T12:06:51 <skathpalia> I mean we can define the label[NAME_EXACT] = 'Name', label[ITEMID] = 'Itemid', label[REVID] = 'Revisionid'
47 2014-05-31T12:07:28 <skathpalia> and then use this label in subject
48 2014-05-31T12:07:38 <skathpalia> what do you think about this?
49 2014-05-31T12:15:33 <skathpalia> dimazest, ?
50 2014-05-31T12:17:05 <dimazest> ok
51 2014-05-31T12:17:12 <dimazest> but use dict.get() method
52 2014-05-31T12:17:26 <dimazest> and give some default in case some other method is used
53 2014-05-31T12:17:34 <dimazest> why not to give a link to the item
54 2014-05-31T12:18:39 <skathpalia> Yeah that can be given in the message text
55 2014-05-31T12:19:43 <skathpalia> That can be added later as the there will be lots of changes in one commit only which would make it difficult to understand
56 2014-05-31T12:22:12 <skathpalia> dimazest, can I do it in my later commit?
57 2014-05-31T12:22:21 <dimazest> ok
58 2014-05-31T12:23:00 <dimazest> but if the design will be changed later
59 2014-05-31T12:23:11 <dimazest> why to write complex code now?
60 2014-05-31T12:24:40 <skathpalia> Now change is done only in the subject
61 2014-05-31T12:25:08 <skathpalia> and the adding link part will be done in message text
62 2014-05-31T12:25:48 <skathpalia> so the design would remain same
63 2014-05-31T12:26:01 <skathpalia> just an extra information will be added in message text
64 2014-05-31T12:32:18 *** randomax
65 2014-05-31T12:34:16 <dimazest> but this extra infornation brings a lot of code
66 2014-05-31T12:35:01 <skathpalia> Yeah for that reason only I wanted to do it in a separate commit
67 2014-05-31T12:35:30 <skathpalia> For that changes would only be done in msgs()
68 2014-05-31T12:36:05 <skathpalia> so it would be kind of an independent change
69 2014-05-31T12:37:30 <dimazest> use item_name=fqname[2]
70 2014-05-31T12:37:51 <dimazest> and leave the nice message for later
71 2014-05-31T12:38:08 <dimazest> though, item_name is not a good identifier anymore
72 2014-05-31T12:39:25 <skathpalia> Okay I am working on it now :)
73 2014-05-31T12:42:11 <skathpalia> dimazest, What should be default value for the label if key is not found?
74 2014-05-31T12:44:31 <dimazest> the value itself
75 2014-05-31T12:45:42 <skathpalia> the fqname[2]?
76 2014-05-31T12:48:59 <skathpalia> dimazest, ^^
77 2014-05-31T12:49:28 <dimazest> fqname[1]
78 2014-05-31T12:52:27 <skathpalia> dimazest, I have made changes in the cr https://codereview.appspot.com/93570043/
79 2014-05-31T12:52:39 <skathpalia> Removed code duplication part
80 2014-05-31T13:02:50 * dimazest is having a lunch
81 2014-05-31T13:03:04 <skathpalia> dimazest, Once this is Ok then I will proceed to adding link in the message
82 2014-05-31T13:07:23 *** skathpalia
83 2014-05-31T13:40:47 *** RogerHaase
84 2014-05-31T13:57:01 *** skathpalia
85 2014-05-31T14:11:01 <dimazest> skathpalia: looks good
86 2014-05-31T14:11:59 <skathpalia> Now I am working on adding link in the message :)
87 2014-05-31T14:26:32 <skathpalia> dimazest, I have added item_url in the message
88 2014-05-31T14:26:42 <skathpalia> You can review it here https://codereview.appspot.com/93570043/
89 2014-05-31T14:27:37 <dimazest> is it the right link?
90 2014-05-31T14:29:13 <skathpalia> I have earlier tried using url_for_item and it produces the correct link
91 2014-05-31T14:30:02 <dimazest> i mean the link to the review
92 2014-05-31T14:30:48 <skathpalia> Yeah it is right link
93 2014-05-31T14:32:05 <dimazest> but i don't see the change that shows the link to the item
94 2014-05-31T14:32:17 <dimazest> (sorry, maybe i'm looking to a wrong place)
95 2014-05-31T14:32:44 *** sl33k_
96 2014-05-31T14:39:01 <skathpalia> dimazest, may be you are looking on patchset 9. The updated one is patchset 10
97 2014-05-31T14:39:28 <dimazest> what file it is? notifications.py?
98 2014-05-31T14:39:52 <skathpalia> yeah
99 2014-05-31T14:40:05 <dimazest> ahh, it's lines ~50
100 2014-05-31T14:40:40 <skathpalia> yeah around that only
101 2014-05-31T14:41:07 <dimazest> i would make this change as a separate commit with a separate review
102 2014-05-31T14:42:44 <skathpalia> So can I commit the old patchset 9 and then I can make a separate cr?
103 2014-05-31T14:43:27 <skathpalia> I mean committing it to my repo
104 2014-05-31T14:45:18 <skathpalia> dimazest, ^^
105 2014-05-31T14:47:09 <dimazest> actually, once we give a link, it doesn't matter, how to retrieve an item
106 2014-05-31T14:48:05 <dimazest> what if we just have subject = L_('[%(moin_name)s] Update of "%(item_name)s" by %(user_name)s', moin_name=app.cfg.interwikiname, item_name=fqname[2], user_name=u.name0)
107 2014-05-31T14:48:49 <dimazest> and replace item_name with item_id or something like this (so it reflects what is being passed, which is not always a name)
108 2014-05-31T14:49:07 <skathpalia> But if we give the information regarding the field then the subject would be user friendly
109 2014-05-31T14:50:12 <skathpalia> I mean the user can some information without even going to the link
110 2014-05-31T14:52:45 <skathpalia> dimazest, may we can have ThomasWaldmann views also on this
111 2014-05-31T14:55:59 <skathpalia> Then decide upon this
112 2014-05-31T14:56:18 <skathpalia> dimazest, what do you say about this?
113 2014-05-31T14:56:43 <dimazest> right, we can discuss it with thomas
114 2014-05-31T14:57:09 <dimazest> though, name and item id are totally different items and they are easy to distinguish
115 2014-05-31T14:57:43 <skathpalia> But if name is also having length 32 then it is difficult to distinguish
116 2014-05-31T15:00:12 *** randomax
117 2014-05-31T15:01:02 <randomax> RogerHaase, moin
118 2014-05-31T15:01:47 <randomax> the current whoosh schema does not have much info about the items
119 2014-05-31T15:02:09 <randomax> like the item size i guess..
120 2014-05-31T15:02:44 <randomax> so should i add fields into that ?
121 2014-05-31T15:14:11 <RogerHaase> randomax: moin, dk ask at meeting (size seems helpful to me)
122 2014-05-31T15:15:56 <RogerHaase> if students were to prepare Done/Blocked/TODO and do copy and paste, the meeting may go faster
123 2014-05-31T15:39:34 <ThomasWaldmann> skathpalia: https://bitbucket.org/saurabh_k/moin-2.0/commits/all it looks like you pulled in changesets, but you did not merge them with your stuff
124 2014-05-31T15:40:10 <ThomasWaldmann> (or update your repo workdir?)
125 2014-05-31T15:40:37 <skathpalia> Yeah I think I am around 3-4 commits behind your repo
126 2014-05-31T15:42:21 <skathpalia> ThomasWaldmann, would syncing with default work https://bitbucket.org/saurabh_k/moin-2.0/branches
127 2014-05-31T15:42:24 <skathpalia> ?
128 2014-05-31T15:42:50 <ThomasWaldmann> i specifically mean the number of heads in your repo
129 2014-05-31T15:43:19 <ThomasWaldmann> (see those graphical lines at the url i posted)
130 2014-05-31T15:43:33 <ThomasWaldmann> or use hg heads
131 2014-05-31T15:47:22 <skathpalia> Yeah I committed my changes without syncing with you repo
132 2014-05-31T15:47:52 <ThomasWaldmann> that's no problem, but you should merge the 2 "default" branch heads now
133 2014-05-31T15:48:40 <ThomasWaldmann> in general, it is a good idea to pull changesets from main repo before starting to work, you can avoid some merges that way
134 2014-05-31T15:49:02 <ThomasWaldmann> if changes happen in parallel, you have to pull them from main repo and then merge them
135 2014-05-31T15:50:41 <skathpalia> Can you tell me the command to do that ?
136 2014-05-31T15:50:48 <skathpalia> I am not able to find it
137 2014-05-31T15:51:00 <RogerHaase> https://moin-20.readthedocs.org/en/latest/devel/development.html#publish-your-change
138 2014-05-31T15:51:10 <RogerHaase> hg merge
139 2014-05-31T15:53:32 <ThomasWaldmann> skathpalia: after merging and meeting, please clean up etherpad, remove stuff that is solved / not needed any more.
140 2014-05-31T15:54:45 <skathpalia> Okay sure will do that
141 2014-05-31T15:57:37 <skathpalia> ThomasWaldmann, merging done :)
142 2014-05-31T15:59:29 <dimazest> i'll be back in 5 min (need to go to the toilet)
143 2014-05-31T15:59:29 <ThomasWaldmann> btw, merging needs to be done very carefully.
144 2014-05-31T16:00:17 <ThomasWaldmann> look out for / resolve merge conflicts, review all merges by hg diff
145 2014-05-31T16:00:59 <ThomasWaldmann> guess we can wait a few minutes until everybody is finished with important business ;D
146 2014-05-31T16:02:02 * dimazest is back
147 2014-05-31T16:03:37 <ThomasWaldmann> ok, welcome to our weekly meeting
148 2014-05-31T16:03:48 <ThomasWaldmann> skathpalia: please begin
149 2014-05-31T16:04:02 <skathpalia> Okay so done part is as follows:
150 2014-05-31T16:04:30 <skathpalia> According to this cr https://codereview.appspot.com/93570043/
151 2014-05-31T16:04:34 <skathpalia> Now ticket update is working properly
152 2014-05-31T16:04:41 <skathpalia> Even the items having no name can also be successfully updated
153 2014-05-31T16:04:59 <skathpalia> The notification module now doesn't depend on item_name. It can ne None also.
154 2014-05-31T16:05:09 <skathpalia> Also the unit test which was earlier failing has passed now. Now no test is failing
155 2014-05-31T16:06:05 <skathpalia> TODO Part
156 2014-05-31T16:06:17 <skathpalia> Improve the UI of ticket items
157 2014-05-31T16:06:31 <skathpalia> Make a +TicketList view having list of tickets with their links and important meta data information.
158 2014-05-31T16:07:16 * ThomasWaldmann added a comment to your EP about maybe rather using +tickets than +TicketList
159 2014-05-31T16:10:11 <skathpalia> Yeah I should use that only
160 2014-05-31T16:11:04 <ThomasWaldmann> do you have specific ideas about that ticket list?
161 2014-05-31T16:11:22 <skathpalia> yeah
162 2014-05-31T16:11:50 <skathpalia> we will show the summary and priority, difficulty and Assigned to
163 2014-05-31T16:12:39 <skathpalia> A table would be displaying tickets and will also have features such as sorting according to difficulty etc
164 2014-05-31T16:13:06 <ThomasWaldmann> why show assigned to?
165 2014-05-31T16:13:43 <dimazest> maybe columns should be configurable
166 2014-05-31T16:13:50 <skathpalia> I saw that in bitbucket also assigned to is shown thats why I was thinking of that
167 2014-05-31T16:14:04 <dimazest> so there is some default config, but it can be overwritten
168 2014-05-31T16:14:31 <ThomasWaldmann> maybe rather show all metrics (so it can be user for re-sorting)
169 2014-05-31T16:15:40 <ThomasWaldmann> yeah, configurable is ok also. not sure we really need it though.
170 2014-05-31T16:16:16 <dimazest> we can start easy
171 2014-05-31T16:16:22 <dimazest> and then add configuration
172 2014-05-31T16:16:24 <ThomasWaldmann> what else needs to be on +tickets?
173 2014-05-31T16:16:30 <skathpalia> You mean showing all the difficulty, priority etc all should be shown so that user can use all of them for sorting purpose?
174 2014-05-31T16:16:30 <RogerHaase> it would be helpful to know what tickets one is assigned to, how may metrics are there and why not show all?
175 2014-05-31T16:16:57 *** penguinRaider
176 2014-05-31T16:18:46 <skathpalia> RogerHaase, 4 are there Effort, Difficulty, Severity, Priority
177 2014-05-31T16:19:11 <skathpalia> Yeah may be we can show all
178 2014-05-31T16:19:50 <ThomasWaldmann> you must show and be able to sort by all
179 2014-05-31T16:20:00 <ThomasWaldmann> all these 4 i mean
180 2014-05-31T16:20:09 <skathpalia> Okay will do that :)
181 2014-05-31T16:20:55 <RogerHaase> that could be easily done with one of the js table sorters
182 2014-05-31T16:20:57 <ThomasWaldmann> because they are very fundamental and E D S are orthogonal
183 2014-05-31T16:21:35 <skathpalia> Will look for those sorters
184 2014-05-31T16:21:42 <ThomasWaldmann> either js table sorter or whoosh sort order
185 2014-05-31T16:21:54 <ThomasWaldmann> js is more interactive, of course
186 2014-05-31T16:22:17 <ThomasWaldmann> aside from misc columns, what else should be on +tickets?
187 2014-05-31T16:22:56 <RogerHaase> a js table sorter would be a nice feature to add. xiaq already packaged mottie's table sorter and DataTables for xstatic
188 2014-05-31T16:23:27 <skathpalia> ThomasWaldmann, there would be links to tickets also
189 2014-05-31T16:23:37 <skathpalia> I didn't think something else
190 2014-05-31T16:24:41 <ThomasWaldmann> how about button with "Create ticket" link?
191 2014-05-31T16:24:53 <ThomasWaldmann> iirc, we have that nowhere yet
192 2014-05-31T16:27:58 *** skathpalia
193 2014-05-31T16:28:16 *** skathpalia
194 2014-05-31T16:28:37 <skathpalia> sorry I got disconnected due to internet failure
195 2014-05-31T16:28:46 <dimazest> that's useful
196 2014-05-31T16:29:01 <dimazest> not the disconnects, but the button :)
197 2014-05-31T16:29:15 <ThomasWaldmann> (there also needs to be a lot more later, but lets start with the basics)
198 2014-05-31T16:30:06 <ThomasWaldmann> any specific js table sorter, that is known-good and proven? :)
199 2014-05-31T16:30:27 <ThomasWaldmann> (or do we already have any such in moin code?)
200 2014-05-31T16:30:51 <skathpalia> ThomasWaldmann, I have used stupid-table sorter earlier
201 2014-05-31T16:31:00 <RogerHaase> I rather like motties table sorter and have been using in in 1.9
202 2014-05-31T16:31:29 <dimazest> i don't have any experience with sorters in js ;(
203 2014-05-31T16:31:48 <RogerHaase> it sorts on multiple colums and guesses correct format for things like dates
204 2014-05-31T16:31:52 <skathpalia> ThomasWaldmann, But the ticket can be created by just adding <domain>/name
205 2014-05-31T16:31:58 <skathpalia> as other items are created
206 2014-05-31T16:32:08 <RogerHaase> and is less complicated than DataTables
207 2014-05-31T16:32:39 <ThomasWaldmann> skathpalia: no, that's no good way
208 2014-05-31T16:33:00 <ThomasWaldmann> it's unintuitive and also, tickets have no name, so why would one give a name for creating?
209 2014-05-31T16:33:33 <skathpalia> Okay will work on adding that button also
210 2014-05-31T16:34:26 <ThomasWaldmann> also, the generic path ask question about item type (which is already defined for that use case)
211 2014-05-31T16:35:45 <ThomasWaldmann> skathpalia: any blockers or any question left for now?
212 2014-05-31T16:36:02 <RogerHaase> I am still unclear as to advantage of nameless tickets vs. how tickets will work with Add Link and Subscribe
213 2014-05-31T16:36:08 <ThomasWaldmann> skathpalia: i did a review of https://codereview.appspot.com/93570043/ in parallel
214 2014-05-31T16:36:19 <skathpalia> Yeah I just saw that
215 2014-05-31T16:36:44 <ThomasWaldmann> RogerHaase: not sure i understand your question
216 2014-05-31T16:36:57 <skathpalia> ThomasWaldmann, I have one doubt in the subject part of the mail
217 2014-05-31T16:37:05 <skathpalia> *notification mail
218 2014-05-31T16:37:43 <skathpalia> Currently the subject just contains the item_name but items can be such that they have no name
219 2014-05-31T16:37:51 <RogerHaase> can a user do an Add Link for a ticket, and if he does, what shows up in place of the item name?
220 2014-05-31T16:38:07 <ThomasWaldmann> in general, likely all places where a name has been should work with fqname in the future
221 2014-05-31T16:38:43 <skathpalia> So I thought of specifying the field and the value in the subject which is there in my cr also
222 2014-05-31T16:39:10 <skathpalia> I mean name with "some_name", Item Id with "some_itemid"
223 2014-05-31T16:39:23 <skathpalia> something like this ^
224 2014-05-31T16:39:55 <ThomasWaldmann> maybe such stuff should be solved at 1 place, maybe in CompositeName
225 2014-05-31T16:42:49 *** sl33k_
226 2014-05-31T16:43:09 <ThomasWaldmann> until we know sth better, we can use same format as in url
227 2014-05-31T16:43:55 <ThomasWaldmann> ok, guess we should continue with randomax (we can continue discussion later about above topics)
228 2014-05-31T16:44:54 <randomax> TODO
229 2014-05-31T16:45:08 <randomax> Improving search GUI
230 2014-05-31T16:45:12 <randomax> - include stuff like item size, content type, wikiname
231 2014-05-31T16:45:21 <randomax> - make it a bit color coded
232 2014-05-31T16:45:44 <randomax> Also i made a small mistake
233 2014-05-31T16:45:57 <randomax> A minor bug in the previous commit - it doesnt affect anything
234 2014-05-31T16:46:14 <randomax> just a code overlap issue
235 2014-05-31T16:46:30 <randomax> Done
236 2014-05-31T16:46:36 <randomax> Shortening of the sidebar
237 2014-05-31T16:46:41 <randomax> Removing multiple occurences of "settings"
238 2014-05-31T16:46:41 *** sl33k_
239 2014-05-31T16:46:46 <randomax> Blocked
240 2014-05-31T16:46:51 <randomax> Quicklinks exceeding the sidebar - Big links cause the a new quicklink to be created each time the "X" is clicked upon and the new quicklink has a shortened url link
241 2014-05-31T16:47:36 <ThomasWaldmann> btw, some picpaste links in your blog post are broken
242 2014-05-31T16:48:00 <randomax> yeah i noticed that..
243 2014-05-31T16:48:08 <randomax> i'll reupload the pics
244 2014-05-31T16:48:43 <ThomasWaldmann> maybe check whether that upload site is meant to permanently store media for your blog posts
245 2014-05-31T16:48:55 <ThomasWaldmann> usually paste sites are rather temp storage
246 2014-05-31T16:50:02 <ThomasWaldmann> the stuff you noted as blocker above - is this really a blocker in the sense that you don't know how to solve or rather a todo?
247 2014-05-31T16:50:21 <randomax> yeah its more of a TODO
248 2014-05-31T16:50:55 <ThomasWaldmann> just for the definition: blocker means that you can't continue on your own.
249 2014-05-31T16:51:42 <randomax> Oh.. ok..
250 2014-05-31T16:52:00 <randomax> I'll revisit it coming week..
251 2014-05-31T16:52:18 <ThomasWaldmann> which can be for a partial problem (then you can still continue working on other stuff) or completely (you can't do anything until solved, this is rather rare)
252 2014-05-31T16:52:33 <randomax> i was currently working looking into the search gui stuff
253 2014-05-31T16:53:22 *** skathpalia
254 2014-05-31T16:53:32 <RogerHaase> the long name for a quicklink is rather easily solved if you would have asked for help (at least I think so)
255 2014-05-31T16:54:55 <randomax> RogerHaase, i solved the long name shortening problem.. but finally there was this problem that the quicklink on trying to close
256 2014-05-31T16:55:23 <randomax> started to make new quicklinks with urls set to the shortened name of the item
257 2014-05-31T16:55:30 <randomax> i was talking about that..
258 2014-05-31T16:56:08 <ThomasWaldmann> (the blog post talks about much more than what's visible in the repo, so before starting on search stuff, first finalize and commit the existing stuff)
259 2014-05-31T16:56:25 <RogerHaase> I understand, in basic layout.html text must be replaced with url <a class="moin-quicklink-delete" href="{{ url_for('frontend.quicklink_item', item_name=text) }}">
260 2014-05-31T16:58:04 <RogerHaase> that will more or less solve the creating a link to an item with a shortened name problem
261 2014-05-31T16:59:44 <randomax> oh.. ok.. i'll make that change
262 2014-05-31T17:01:08 <randomax> ThomasWaldmann, yeah the third code review i'll make the change according to what RogerHaase said and send a pull request..
263 2014-05-31T17:01:30 * ThomasWaldmann just added some comments to that CR
264 2014-05-31T17:01:58 <ThomasWaldmann> RogerHaase: also have a look, what do you think? ^
265 2014-05-31T17:03:19 <ThomasWaldmann> https://codereview.appspot.com/91680043/
266 2014-05-31T17:05:15 <RogerHaase> line 242 has the problem, I think, text should be ~ url[1:]
267 2014-05-31T17:05:48 <randomax> ThomasWaldmann, the problem was that different quicklinks would have different titles and the loop is running in the template itself so i have to make the code there itself i guess
268 2014-05-31T17:07:03 <randomax> regarding the "url[1:]" i can make another variable and send it in..
269 2014-05-31T17:07:57 <ThomasWaldmann> is there anywhere a description of the problem and the way you want to solve?
270 2014-05-31T17:09:23 <RogerHaase> maybe too much detail on that issue for meeting, we should note that mistake mentioned earlier is that if you do ./m css on current tip, then there is a change to basic.css and that needs to be fixed
271 2014-05-31T17:10:34 <RogerHaase> I am concerned that we are always testing with small bits of wiki data. It would be helpful if we had a bigger wiki for occasional testing - could we get a converted moinmo.in that we could download for testing?
272 2014-05-31T17:10:56 <randomax> ThomasWaldmann, there is no explicit issue reported regarding that.. but i had mentioned about it in the last meeting
273 2014-05-31T17:11:18 <RogerHaase> For example, the suggestions whoosh generates after a search for input, name term, and content term seem rather silly. Maybe more data would create helpful results so we can determine if showing those are worthwhile.
274 2014-05-31T17:11:33 <ThomasWaldmann> RogerHaase: such a dump has a lot of personal data in it...
275 2014-05-31T17:13:14 <RogerHaase> have any other ideas on getting a large wiki for testing?
276 2014-05-31T17:13:38 <ThomasWaldmann> yeah, more data would help for some tests.
277 2014-05-31T17:15:14 <ThomasWaldmann> we would either create new stuff (just like with the sample wiki content, just more). writing an anonymizer is maybe not doable.
278 2014-05-31T17:15:35 <ThomasWaldmann> -either
279 2014-05-31T17:17:18 <RogerHaase> for testing whoosh, having some realistic data would be better. Would you object to a bot sucking all pages from moinmo.in?
280 2014-05-31T17:17:36 <ThomasWaldmann> btw, just as a general note "
281 2014-05-31T17:17:36 <ThomasWaldmann> Fixing quicklinks and their tooltips
282 2014-05-31T17:17:38 <ThomasWaldmann> "
283 2014-05-31T17:18:19 <ThomasWaldmann> -> such text is not very useful as it is not saying what "fixing" means or what was "broken".
284 2014-05-31T17:18:42 <ThomasWaldmann> so i'ld appreciate if you just said that instead of "fixing".
285 2014-05-31T17:19:18 <ThomasWaldmann> applies for codereview comments as well as for commit comments btw
286 2014-05-31T17:19:19 <randomax> ok.. i'll do that
287 2014-05-31T17:20:37 <ThomasWaldmann> randomax: can you quickly give file and lineno feeding that template?
288 2014-05-31T17:23:51 <randomax> https://bitbucket.org/thomaswaldmann/moin-2.0/src/600a9b4de684a67b5a247b3ff55777f596c71aaa/MoinMoin/themes/__init__.py?at=default#cl-389
289 2014-05-31T17:26:07 <ThomasWaldmann> ok, so split_navilink is called from exactly 1 place (==there)
290 2014-05-31T17:29:25 <ThomasWaldmann> so, if it doesn't give you the right "title", shouldn't that get fixed either in split_navilink or right after calling it?
291 2014-05-31T17:32:10 <randomax> its not giving me a wrong title as such.. title is supposed to be the title of the wiki itself.. but the quicklink title which shows up in the tooltip should contain both the wikiname and itemname as we discussed in the last meeting
292 2014-05-31T17:32:51 <randomax> thats what i was doing in the template..
293 2014-05-31T17:34:02 <ThomasWaldmann> 388 url, link_text, title = self.split_navilink(text)
294 2014-05-31T17:34:22 <ThomasWaldmann> that is giving you an url, a link text for that url and a title attribute for the link
295 2014-05-31T17:35:02 <randomax> yes
296 2014-05-31T17:35:17 <ThomasWaldmann> and these "items" are just tuples css-class, url, linktext and title
297 2014-05-31T17:35:50 <ThomasWaldmann> ALL are made that way, so there is little reason having to specialcase stuff in the template
298 2014-05-31T17:36:19 <RogerHaase> randomax: change the code in split_navilink and simplify the lines in layout.html
299 2014-05-31T17:36:47 <randomax> ok.. i'll do that
300 2014-05-31T17:37:29 <ThomasWaldmann> the only reason the original code is having that "if title" is that we do not want to generate a title attribute for the link if the title is empty
301 2014-05-31T17:38:20 <randomax> ok..
302 2014-05-31T17:38:45 <ThomasWaldmann> randomax: and also put a comment "remove leading /" above the right place in the code, as that is not totally obvious by [1:]
303 2014-05-31T17:39:55 <randomax> ok.. i'll do that..
304 2014-05-31T17:41:35 <ThomasWaldmann> ok, guess we are through with the meeting. thanks for attending.
305 2014-05-31T17:41:53 <ThomasWaldmann> randomax: if you have questions, just ask :)
306 2014-05-31T17:42:50 <randomax> nope.. :)
307 2014-05-31T17:44:38 <RogerHaase> back in an hour
308 2014-05-31T17:48:14 *** randomax
309 2014-05-31T19:44:57 *** sl33k_
310 2014-05-31T19:48:33 *** sl33k_
311 2014-05-31T19:54:06 <RogerHaase> if you create an empty "user" directory within a moin 1.9 wiki/underlay directory, you can then use ./m import <path>/wiki/underlay to create a bigger test wiki. There is a bit of junk, but it is 20 times bigger and has 20 times more files than what you get from ./ sample. I put a zip file at http://www.digitalrockart.org/Temp/underlay.zip . Unzip, then do ./m restore <path>/underlay.moin
312 2014-05-31T22:04:48 <RogerHaase> ThomasWaldmann: #431 moin index-build https://codereview.appspot.com/98660043/
313 2014-05-31T22:11:33 *** derdon
314 2014-05-31T22:12:18 *** RogerHaase
315 2014-05-31T23:38:38 *** sl33k_
316 2014-05-31T23:45:41 *** sl33k_
317
