[09:08:17] <eidax00>	 hi
[09:08:21] <eidax00>	 why does this not work: https://w.wiki/8gL
[09:08:42] <eidax00>	 but this works: https://w.wiki/8gM
[09:15:32] <eidax00>	 Also tried: ?item wdt:P780/wdt:P31*/wdt:P279* wd:Q737460.
[09:35:24] <SothoTalKer>	 probably the query optimizer. maybe file a bug :)
[09:53:36] <eidax00>	 okay
[09:53:37] <eidax00>	 thanks
[09:54:49] <SothoTalKer>	 it works for a single query for either wd:, or it works for the second one if you leave out either wdt:P31* or wdt:P279*
[09:55:09] <SothoTalKer>	 it should not produce an error, though
[09:55:48] <SothoTalKer>	 maybe you can divide that into subqueries
[09:57:19] <SothoTalKer>	 i do hate the query service. very much.
[10:07:43] <eidax00>	 will look at examples of subqueries
[10:19:50] <eidax00>	 SothoTalKer, subqueries work thanks
[10:20:03] <SothoTalKer>	 yw
[10:21:57] <eidax00>	 wow now it started workihng, the other query
[10:22:12] <eidax00>	 oh wait no
[10:22:20] <eidax00>	 adding this fixes it: hint:Query hint:optimizer "None" .
[10:24:46] <SothoTalKer>	 i tried that before and it did not work for me. dunno why.
[10:26:08] <eidax00>	 hmm, adding LIMIT 10 fixes it ...
[10:27:07] <eidax00>	 hmm, subquery also dies
[10:27:11] <eidax00>	 Only works with limit
[10:30:47] <eidax00>	 Put it here: https://www.wikidata.org/wiki/Wikidata_talk:SPARQL_query_service#Possible_bug
[11:04:49] <eidax00>	 if I have an rdf file is there some easy way to load it into some explorer where I can query and view relations?
[11:09:35] <bennofs[m]>	 you can load it into a local blazegraph instance, that'll give you a query interface to explore it via SPARQL. It doesn't do any fna
[11:09:50] <bennofs[m]>	 *any fancy visualiations though
[11:25:17] <eidax00>	 okay will check
[11:25:22] <eidax00>	 How is this query dubious: <bennofs[m]> *any fancy visualiations though
[11:25:25] <eidax00>	 sorry wrong paste
[11:25:35] <eidax00>	 How is this query dubious: https://www.wikidata.org/wiki/Wikidata_talk:SPARQL_query_service#Possible_bug
[11:26:11] <eidax00>	 what is dubious about: ?item wdt:P780/wdt:P31*/wdt:P279* wd:Q737460, wd:Q86, wd:Q21120251.
[11:26:14] <eidax00>	 I don't get it
[11:26:25] <bennofs[m]>	 eidax00: i don't understand that answer either. looks like a bug to me, I don't think you should be getting internal blazegraph exceptions with such a query
[11:26:39] <eidax00>	 Yea
[11:26:55] <eidax00>	 Well I don't know - but I know it works fine with limit X
[11:27:29] <bennofs[m]>	 eidax00: it seems to be caused by the wikibase label service
[11:27:32] <bennofs[m]>	 if you remove the label it works
[11:28:03] <eidax00>	 Hmm
[11:28:06] <eidax00>	 nice, thanks bennofs[m] 
[11:28:41] <eidax00>	 I think there is alternate ways to get label? (can't recall, maybe not but I think wikibase label is recent)
[11:30:46] <eidax00>	 On a slightly unrelated note, say I want to import data from here with bot: https://www.malacards.org/card/pneumonia#clinical_features
[11:30:55] <eidax00>	 Do I have to worry about copyright?
[11:32:29] <bennofs[m]>	 eidax00: https://w.wiki/8gb is a query that works but not sure if that's exactly the same as what the labelling service would do (for example, I don't know if the labeling service uses some matching heuristic to match en-US, en-GB etc and not just en as lang tag)
[11:33:23] <eidax00>	 bennofs[m], thanks a bunch, it is fine if it just takes en label
[11:33:26] <bennofs[m]>	 eidax00: might be the same as https://phabricator.wikimedia.org/T216271
[11:36:01] <bennofs[m]>	 actually, https://phabricator.wikimedia.org/T147577 looks more similar
[11:36:46] <bennofs[m]>	 eidax00: that last phabricator task also has a better fix: add "hint:Query hint:analytic "false" ." to your original query and it works fine
[11:39:08] <eidax00>	 bennofs[m], thanks a bunch
[11:39:20] <eidax00>	 will use the hint for now then, works great
[19:44:46] <SothoTalKer>	 eidax00: there is "OPTIONAL { ?item rdfs:label ?itemLabel filter (lang(?itemLabel) = "en") }"
[19:45:15] <SothoTalKer>	 which shall be faster than "SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }"
[19:49:31] <eidax00>	 thanks SothoTalKer 
[19:49:59] <eidax00>	 SothoTalKer, that was in query that benno shared also
[19:50:18] <SothoTalKer>	 the difference between the two: if there is no label, it is left blank instead of the item number
[19:50:28] <eidax00>	 I see
[19:50:47] <SothoTalKer>	 still, my queries time out (:
[19:51:13] <eidax00>	 what query are you sending?
[19:51:31] <eidax00>	 I had some other that I was running, but it was timing out unless I used union
[19:51:54] <SothoTalKer>	 they did not time out half a year ago
[19:57:55] <SothoTalKer>	 looks like 77k entries are too much :x
[20:03:08] <amz3`>	 SothoTalKer: winter is coming.
[20:03:24] <SothoTalKer>	 that's great :)
[20:04:22] <SothoTalKer>	 any hints?
[20:07:23] <amz3`>	 SothoTalKer: pagination ?
[20:32:04] <SothoTalKer>	 hm, i turned this https://w.wiki/8hJ into this: https://w.wiki/8hJ
[20:32:49] <SothoTalKer>	 https://w.wiki/8hK <- this '-'
[20:34:32] <SothoTalKer>	 old query took 55s, new one took 13s.
[20:34:57] <SothoTalKer>	 let's see if it works for the 77k entries :p
[20:35:33] <SothoTalKer>	 nope, still times out :x
[20:40:14] <SothoTalKer>	 instead of MINUS using FILTER NOT EXISTS did work, albeit quite close.
[20:48:12] <SothoTalKer>	 and replacing the label service with the faster variant made it finish in under 40 seconds.