[04:59:19] PROBLEM - puppet last run on wdqs1003 is CRITICAL: CRITICAL: Catalog fetch fail. Either compilation failed or puppetmaster has issues [05:30:29] RECOVERY - puppet last run on wdqs1003 is OK: OK: Puppet is currently enabled, last run 3 minutes ago with 0 failures [11:23:07] I wish there were some smaller partial data dumps available... [11:25:33] it's got so big that I don't have enough space to download it [11:27:08] and 20 million of them are instance of scholarly article which aren't relevant for what I want to do... that's 38% of the items and since they tend to have more statements than normal items, I wouldn't be surprised if the dump were half the size without them [12:12:24] nikki: If you have any good ideas on how exactly to split, feel free to open a ticket… we have been thinking about this for quite some time [12:12:28] input very much welcome [12:31:48] I would probably separate out things with a commonly used p31 value like scholarly article (~20 million), human (~4.8 million), taxon (~2.5 million) and wikimedia stuff like disambiguation pages, categories and templates (~6.7 million for the top five items) [12:32:10] and subclasses (p279) [12:32:53] (subclasses are ~2 million) [12:33:42] Did you try the linked data fragments nikki? [12:34:01] no [12:34:21] https://query.wikidata.org/bigdata/ldf [12:34:40] Info at http://linkeddatafragments.org/ [12:36:09] Not sure if it suits your use case, but it seems to try to fill the gap between SPARQL and dump [12:36:59] We seem to have a client at http://ldfclient.wmflabs.org/ (never used it) [12:42:37] I'm trying to work with items with geonames ids and find the ones which should have another id and don't [12:43:17] like something in america with a geonames id should normally have p590 too [12:48:13] I can query for that, at least, but other queries, e.g. things with p1566 which aren't in america or canada and don't have p2326 times out [13:36:27] abian: I see you're reporting things for me :D [13:36:53] Yeah :) [13:37:24] Although I was already writing the task about the dump of classes when you wrote it by here [14:41:01] * Praxidicae waves [14:54:21] hi Praxidicae [18:16:52] Wikidata is, since November, one of the top 10,000 websites according to Alexa's global traffic ranking :) [18:18:34] (Probably not enough, but let's continue improving) :P [18:19:27] abian: i'll help by deleting spam! :D [18:20:00] :D [18:20:10] hello all [18:20:19] or rather spamming dr with requests :P [18:20:24] Hi, elya [18:20:30] i wish tehre was a way to implement a sort of speedy for wd [18:21:15] I'm sure there's something on Wikipedia that we should be able to bring [18:21:52] could someone please help me with this query? http://tinyurl.com/yd4u5scv I don't understand why I get results with normal ranks, as I have preferred ranks in every item. [18:23:10] elya: You would get only the preferred value if you used wdt: everywhere [18:23:24] ... I think :) [18:23:41] But you're also using p:, ps:, etc. [18:23:46] well, but with wdt: I cannot get the qulifiers [18:24:02] Right [18:24:10] hm [18:24:21] it just seems silly to have to send it to DR abian [18:24:22] You could filter the rank if you prefer to keep your query as is [18:24:24] i understand why some need to be though [18:24:41] or just make it a gs wiki and say i can delete spam :P [18:24:53] * Praxidicae is ultimate lazy [18:25:04] elya: You could add FILTER(?rank = wikibase:PreferredRank) [18:25:24] @abian, ah, that's what I was missing. I'll try it [18:25:33] Or just replace ?rank with wikibase:PreferredRank and delete the variable from the SELECT [18:26:54] abian, beautiful, both work! Thank you! [18:27:13] Yay! No problem :) [18:27:59] bye.. [18:40:48] Praxidicae: there's a "requestdeletion" gadget in the preferences, maybe that would help? [18:40:53] I've never used it so I don't know how it works [18:41:27] i think i use that [19:00:45] I'm not sure how we would do a speedy deletion thing that doesn't use wd:rfd since items don't use normal wiki text [19:11:20] yah i dunno the technical side [19:11:23] i just wish it was a possibility