[00:03:25] anyway thanks [00:13:27] Amqui: the intersting thing is how diffent sources are handled. e.g. if sb. dies he must not be dead for everybody [00:14:42] Change merged: Jeroen De Dauw; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/33045 [00:14:43] he could also have only left his body and may come back later. (i learned that from enwiki) [00:18:34] Sannita you give good answers [00:18:49] thanks jeblad_WMDE [00:19:49] There will be statements on the item page, and those statements will be connected to properties and have values [00:20:03] Sort of oversimplified [00:20:14] Merlissimo what happens on enwiki stays on enwiki :P [00:21:55] and it will be possible to extract values marked as a specific property inside an item [00:21:57] Amqui: but you cannot force enwiki to include the infomation that sb is death if they think he only disappeard from his body [00:22:21] it's bad though the interwikis phase had been implemented without a property to say if it's a FA on this wiki [00:22:50] FA or GA [00:23:04] New review: Aude; "not been able to test this yet because pollforchanges stuff is broken" [mediawiki/extensions/Wikibase] (master); V: 0 C: -1; - https://gerrit.wikimedia.org/r/33033 [00:23:05] Amqui: FA was included at the first version [00:23:22] this information was dropped later [00:23:42] Amqui: jeblad_WMDE removed it ;-) [00:23:45] why [00:24:12] it will have to be re-added at one point [00:24:23] Amqui: FA and GA will probably be one of the first property stuff we do [00:24:47] not exactly sure how yet, as it's somewhat different than all the other properties we'll have for items [00:24:57] they are properties of site links [00:25:18] aude not really different from "article X's title in XYZ language is "BLABLA"" [00:25:28] "article X in XYZ language is a FA" [00:26:36] true but then people will want more stuff attached to site links [00:26:51] Its a property of the sitelink [00:26:52] such what? [00:26:56] yes [00:27:03] aude and in what is that a bad thing? [00:27:05] maybe all the article ratings? [00:27:09] stub, start, etc. [00:27:13] importance, etc. [00:27:33] I see [00:27:34] if we use the property stuff and attach them to siteslinks, then it should be possible and more usable [00:27:46] rather than hack it in there just for GA and FA [00:28:11] yes it is property of the sitelinks, or in other world "properties of the article which is in fact a real world entity" [00:28:28] sure [00:28:43] they are not treated that way yet but it's possible [00:28:57] It is a property of the link itself, not the real world entity [00:29:26] jeblad_WMDE I just said, I we push that to its limits, the article itself could be an entity [00:29:31] if we push* [00:29:46] Amqui: perhaps but baby steps :) [00:30:19] The entity is out there in the real world, the item is the page describing the entity [00:30:35] The pages at Wikipedia is not the entity [00:30:41] it could be [00:30:53] there a pages on Wikipedia about other pages around the interweb [00:30:54] They are also descritions but in prose [00:31:33] someday there will be an article on Wikipedia about the history of another article on Wikipedia ;) [00:31:33] Sitelinks connects the pages about the entities with items which is also about the same entities [00:32:37] jeblad_WMDE yes I know that [00:32:59] but I just didn't see why it's "bad" for those sitelinks to have more properties [00:33:32] Amqui: some day there will be an article about the end of wikidata [00:33:54] Its not bad [00:33:55] Merlissimo xD [00:34:11] The code was removed because it was dead [00:35:03] ok, even the example mentioned, having a central repersitory to know if that article in another language is a stub or not, might be good [00:35:34] will just need some thinking about its maintenance [00:35:50] Amqui: perhaps in 500 years an historian reads a book about jeblad being member of the wikidata team. ;-) [00:36:50] perhaps [00:37:09] More like the moron that messed up [00:37:49] how disambiguation page are treated on wikidate? [00:38:10] wikidata* [00:38:59] since they don't correspond to any entity [00:39:28] some month ago i was asked to rewrite some code in fortran. the programm was developed in 1970 and is still in use. [00:40:15] Merlissimo I'm not surprised [00:40:20] It doesn't really matter if som items are used just to connect pages and have no real data [00:40:35] ok, so that entity rule is flexible [00:40:59] there could be items to connect wikiprojects for example [00:41:01] break the rules, but community decides [00:41:21] of course [00:43:03] Merlissimo at university one teacher gave us his program in fortran [00:43:09] if more interwikis would be available for sitelinks we could create an item about this channel [00:44:10] I showed him how to do the samething with MS Excel in 30sec lol [00:44:27] good night [00:44:44] anyway thanks jeblad_WMDE [00:44:59] his fortran program wold run 30 times faster [00:45:14] yes, that's why I rewrote it [00:45:30] in C++ after [00:45:39] but that's outside the story [00:46:08] fortran is usually faster than c++ on math problems [00:46:32] if you say so [00:46:45] I'm not a programmer [00:48:09] fortran is more or less optimized for numerical problems [00:49:04] ok [00:49:48] he must have liked what I've done, since now he's giving mine [00:52:10] Personally I'm impressed by the latest versions of javascript [00:52:31] but he also said I should loose less time programming that, and more time learning the subject of his classes lol [00:53:29] I don't know about JS [00:55:02] aude? [01:02:11] jeblad_WMDE: ? [01:04:32] Sorry, got lost in a movie [01:05:07] ok :) [01:05:08] aude, can you poke some at the ULS team [01:05:20] Seems like something is broken [01:05:47] Nikerabbit: poke ^ [01:05:50] And it takes our js-code with it [01:06:22] Nikerabbit or siebrand can look at it but doubt we'll see them online now [01:06:47] * aude just won't update my install of ULS now [01:06:56] I wrote an email to the list, but you know how hard it is to keep up with that list [01:07:02] sure [01:07:11] ;) [01:07:46] Seems like it is something after friday [01:08:15] i know they were doing stuff with the geolocation [01:08:19] not sure what else [01:08:42] someting with a position is missing [01:08:54] hmmm [01:09:03] is there a bugzilla ticket for this? [01:09:38] No, I thought it would be fixed asap [01:09:53] they also did Update jquery.uls from upstream [01:10:02] that could be a big pile of code [01:10:21] But there is a log of the failure in my email [01:10:23] but sounds geo related [01:12:12] Uncaught TypeError: Object #