[07:16:34] (PS2) Mwjames: \SMW\ContentParser use ContentHandler where available [extensions/SemanticMediaWiki] - https://gerrit.wikimedia.org/r/77504 [07:16:46] (PS4) Mwjames: \SMW\Maintenance\RebuildConceptCache (rebuildConceptCache.php to replace SMW_conceptCache.php) [extensions/SemanticMediaWiki] - https://gerrit.wikimedia.org/r/77493 [07:17:03] (PS2) Mwjames: (Bug 52396) use getSubject()->getDBkey() instead [extensions/SemanticMediaWiki] - https://gerrit.wikimedia.org/r/77126 [13:29:55] (PS3) Himeshi: Add handling of other parameters for page sections [extensions/SemanticForms] - https://gerrit.wikimedia.org/r/77461 [14:31:47] yaron: "late" follow up question: 3 months ago, you told me to use getKey() on some SMW objects (SMWDIBlob::getSring() and SMWDIWikiPage::getTitle() ) but SMWDIWikiPage doesn't have getKey(). Are there any other functions I could use to achieve this? [14:32:01] I am not sure if it's clear what I am trying to do. So please just ask ;) [14:32:35] Oh man. I vaguely remember that. :) [14:32:52] What version of SMW are you using, and yes, what are you trying to do? [14:34:06] had a work break to study ;) [14:34:12] uhm, it's SMW 1.8 [14:35:18] it's not my code and it's badly documented so I can only guess: It walks down properties recursively. So it takes the property requests a string representation and goes on from there [14:35:43] the problem is this: PHP Fatal error: Call to undefined method SMWDIBlob::getTitle() [14:35:53] if I use getString() at that line [14:36:26] Calling getString() results in a "no getTitle() method" error? [14:36:26] it works for the SMWDIBlob, but then I get this (for some other property/page/object/.. it encounters on the recursive way: PHP Fatal error: Call to undefined method SMWDIWikiPage::getString() [14:36:43] This sounds like two different issues. [14:37:07] the problem is, that getString works for one object, getTitle for the other [14:37:20] so the easiest way would be to use a function that works for both [14:37:26] Right - I would think the code could benefit from some "if method_exists()" calls. [14:37:46] the code could benefit from a lot of things ;) [14:37:58] No need to call one method for all classes, though that would be the nicer solution. [14:38:00] I am not sure if it's crappy or too clever for me to understand [14:38:20] What extension is this, out of some curiosity? An unreleased extension? [14:38:26] yeah [14:38:49] Well, calling "if method_exists()" should do the trick. [14:39:01] it's called SemanticAssist and was part of a diploma thesis my boss coached [14:39:08] Oh, okay. [14:39:14] I have to get it to work 3 years after it was written ;) [14:39:18] I just saw that name somewhere... [14:39:23] and like 20 SMW versions later [14:39:29] Right. [14:39:59] Someone added info about that extension on mediawiki.org, or semantic-mediawiki.org, or something like that. [14:40:04] I think it was never published [14:40:07] but I am not sure [14:40:22] Alright - maybe it was something else. [14:40:29] it finds differences in pages a user provides [14:40:47] Oh. Property differences? [14:41:00] the purpose is to find pages/data/objects in the wiki with similar properties [14:41:09] and display them as a graph [14:41:14] I see, okay. [14:41:28] and it was written to use it for IT Service Management. [14:41:42] Oh. [14:41:52] e.g. "was there ever a computer object in this wiki with a similar graphics card that had driver problems" [14:41:55] or something like that [14:42:07] That's an unusual mix of academic/theoretical and a very applied use case. [14:43:55] yeah. it's part of my bosse's PhD Thesis. he was/is head of the IT department here and wanted to improve the service we do ;) [14:44:05] Ah! Now it all makes sense. :) [14:44:34] since the company is growing and by now more than 1500 computers make it harder and harder to keep track of/solve problems [14:44:51] Well, software written to serve one's own needs tends to have a much higher chance of being useful. [14:45:04] The company? [14:45:21] hmm? [14:45:34] I thought this was for an academic IT department. [14:45:47] Oh, never mind, now I get it. [14:45:53] it's kind of complicated [14:46:02] I understand. Well, that's pretty neat. [14:46:09] I don't know if this kind of thing exists in USA/UK/wherever [14:46:50] it's a government founded research facility which works together closely with an university [14:46:56] Oh, okay. [14:47:46] There are definitely things like that, as far as I know - JPL is an example, I think. [14:48:45] Well, I hope it gets released, after you fix it. :) [14:49:02] our board members are all professors and most of the employees are writing their PhD thesis here.. [14:49:26] so, anyways. Is it smarter to do that if method exists thing or should I rather do some checks if it is a specific class? [14:49:39] Oh, okay. Maybe there isn't anything like that, I don't know. [14:50:02] I would just go with method_exists() - it's more foolproof. [14:50:16] well, it's a collection of small/medium sized extensions. Some of them are kind of released, but noone seemed interested yet [14:50:40] Oh - "SemanticAssist" is the whole collection? [14:50:44] nah [14:51:07] I just renamed it ITSMWiki. I lack creativity in naming [14:51:34] You renamed the extension to that, or the collection to that? [14:51:39] the collection [14:51:44] there is no extension page [14:51:54] Alright. [14:52:12] I am working on this for like 4 years now. Most of the code is written by me, some stuff comes from diploma thesises [14:52:27] (working = 35h a month, so not that much) [14:52:29] Man. [14:52:51] Oh, okay - I assumed that wasn't the only thing you had done over 4 years. :) [14:53:13] it's not very good, I think. I do too much page parsing instead of using SMW/MW functions [14:53:16] but it works [14:53:51] Well, that's the important thing... [14:53:52] Why is it called "SemanticAssist", by the way? [14:53:59] I dunno :P [14:54:05] there is a paper... sec [14:55:08] I think the reason is "it provides assistent to problem solving using semantic stuff" :P [14:56:25] Oh, wonderful. :) [14:56:39] "Incident and Problem Management Using a Semantic Wiki-enabled ITSM Platform" =>http://arnetminer.org/publication/incident-and-problem-management-using-a-semantic-wiki-enabled-itsm-platform-3440957.html;jsessionid=82F3EF60DC4179734FEA3C8459252C0E.tt [14:57:25] ah, my bad [14:58:00] the extension was called "SemPA - Semantic Incident and Problem Analyizer" in the first place [15:00:47] I think I will try to publish the extension as soon as it's done. It's kind of a work in progress and I never felt it "worked enough" to release it [15:00:56] and I am not quite sure how the legal thingy works [15:01:38] Oh - I assume you would need to get the permission of everyone who helped write it. [15:10:15] In any case, I hope you release the extension, and I hope you rename it before releasing it. :) [15:11:07] why rename it? :P [15:12:25] the permission is not the problem, 8500 lines of the code were written by me, 1000 are the stuff I am trying to get to work right now. [15:12:29] Well, because "Assist" is kind of meaningless. [15:12:40] ...or were you being sarcastic? [15:13:16] Ah - by legal thingy I guess you meant whether you can release work that someone else paid for as open source. [15:13:21] with "legal thingy" I rather meant that I am not sure if I/we can just publish diploma/PhD thesises or if it "belongs" to the university in some way [15:13:41] Right. [15:14:44] right now the name is "ITSMWiki". I will rename the "Semantic Assist" part back to "Semantic Incident and Problem Analyizer". Together with the ITSM-part it should be clear what it does [15:15:53] Alright. Well, that sounds better already... although, is it really a problem analyzer? [15:16:06] yeah [15:16:12] I mean, isn't analyzing problems just one possible usage of this extension? [15:16:13] I just got it to work. At least half of it [15:16:17] hmm [15:16:20] It sounded more generic than that. [15:18:01] well, it builds a graph comparing 2 pages with nodes containing 3 fields: the property value of page one, the value of page 2 and a boolean if those values match. You can do some other stuff too, like comparing the timeline of a single page and see which properties changed in which way [15:18:08] basically it's not very "magic" [15:19:20] That sounds good. [15:19:58] ...and as I thought, there's nothing problem-specific about it. [15:20:05] yes, you are right ;) [15:20:53] but anyways, I just realized it actually doesn't do very much. In most cases it's faster to compare the properties by looking at the pages than going to that special page and c&p'ing the page names [15:21:08] I think its application is 90% academic [15:21:58] What about the graphing stuff? [15:22:10] it's ugly ;) [15:22:34] it's only 1 root node and X leaves [15:22:48] using the GraphViz extension [15:22:52] nothing special about that [15:23:07] Oh, okay. [15:23:53] well, but it only works partly yet [15:23:57] maybe there is more to come [15:24:27] comparing 2 pages seems boring. The timeline comparison and graphing a single page don't produce anything yet. I have to dig into that [15:26:29] Well - there's always potential for improving the interface. [15:27:55] yep. I will continue tomorrow. My brain is kind of dead right now [16:00:14] (CR) Yaron Koren: [C: 2 V: 1] "Looks great! Removing the hardcoded English was quite useful." [extensions/SemanticForms] - https://gerrit.wikimedia.org/r/77461 (owner: Himeshi) [16:00:17] (Merged) jenkins-bot: Add handling of other parameters for page sections [extensions/SemanticForms] - https://gerrit.wikimedia.org/r/77461 (owner: Himeshi) [17:08:07] (PS6) Mwjames: \SMW\SimpleDependencyBuilder + \SMW\DependencyInjector [extensions/SemanticMediaWiki] - https://gerrit.wikimedia.org/r/76956 [17:13:14] (PS7) Mwjames: \SMW\SimpleDependencyBuilder + \SMW\DependencyInjector [extensions/SemanticMediaWiki] - https://gerrit.wikimedia.org/r/76956 [18:19:31] yaron: It seems a lot of extensions seem to have been bothered by my GSoC work :p [18:19:49] Oh - you did that? :) [18:20:07] Oh, I guess that was part of the whole SQLStore3 transition. [18:21:02] There's some EnhancedRetrieval extension which uses the table names directly [18:21:18] Wasn't that one part of SMW+? [18:21:26] yeah I think [18:21:29] i.e., a dead extension? [18:21:56] There's some work going on [18:21:59] Oh. [18:22:16] The Semantic Drilldown extension also uses SMW table names directly - I had to update it. [18:23:18] It was a useful change for SMW's performance, though I hope there isn't an SQLStore4 any time soon. :) [18:23:28] So the guy working on it wanted to know the names to replace with [18:23:33] I see. [18:23:45] And I had to answer its not that easy :) [18:24:11] Jeroen and I tried to convince Markus to not name it '3' [18:24:11] Yes, it takes some work. [18:24:24] What was your alternate name? [18:24:35] just to drop the number [18:24:53] But there was already an SQLStore. [18:25:03] thinking about it now, something fancy would have been good :p [18:25:16] SQLStoreDeluxe. :) [18:25:21] yeah that could become SQLStore2 and then back to SQLStore [18:25:59] And if there were another version after that, what would it be called? [18:26:17] we planned to not change names after that [18:26:28] SQLStoreRises [18:26:31] ;) [18:26:36] There you go. :) [18:26:51] I think Markus made the right choice. [18:27:19] seems so, otherwise I can see how much confusion had been [18:27:46] Its an enjoyment to keep others confused :D [18:28:18] You're not the only one who feels that way, it seems. :) [18:31:55] yaron: Apparently I am one of the top 10% freelancers in oDesk [18:32:11] Wow! Nice. [18:32:38] though I doubt if its a publicity stunt of theirs [18:33:10] Oh - they sent you an email about it? [18:33:31] yeah [18:33:52] signed by the Ceo :p [18:33:57] Well... there are probably a lot of user accounts created, that haven't done any work via the site. [18:33:59] *CEO [18:34:09] true [18:34:23] That's nice, though. [22:46:22] Hello. I have a question about $smwgNamespacesWithSemanticLinks [22:46:32] Does this need to be defined before including SMW? [23:12:50] From what I'm reading they look like they're pure definitions that must be english aliased, they are constants [23:15:23] I don't think they'll break anything if not there. [23:16:11] Demos: I'm adding it because of custom namespaces [23:16:33] But I can't get it to work, I placed it after the includes and the pages still didn't have properties [23:17:04] So I moved it before the includes, and it still didn't work [23:34:11] So you want to probably create something like Wiki:Mainthat inclues your main namespace and add your property to it, then anything within that namespace will be associated to it [23:34:19] ...and hes gone