[06:30:10] !admin revdel needed [06:30:10] Please visit https://www.wikidata.org/wiki/WD:AN [06:31:52] What precisely do you need to revdel? [06:32:48] there was a post to WD:AN - https://www.wikidata.org/w/index.php?title=Q26752&diff=1099848962&oldid=1088878664 - I figured this would be faster [09:08:48] Change on 12meta.wikimedia.org a page Wikidata was modified, changed by 95.184.76.192 link https://meta.wikimedia.org/w/index.php?diff=19746003 edit summary: [-733] [09:10:06] Change on 12meta.wikimedia.org a page Wikidata was modified, changed by DannyS712 link https://meta.wikimedia.org/w/index.php?diff=19746008 edit summary: [+733] Reverted changes by [[Special:Contributions/95.184.76.192|95.184.76.192]] ([[User talk:95.184.76.192|talk]]) to last version by DannyS712 [09:10:08] 10[1] 04https://www.wikidata.org/wiki/Special:Contributions/95.184.76.19213 => [09:10:11] 10[2] 04https://www.wikidata.org/wiki/User_talk:95.184.76.192 [11:55:33] hi [11:56:03] i'm using wikidata for request into wikidata from a python script, it's extremely slow [11:56:21] so i'm considering to make my program multithread [11:56:29] so i could parallel the request [11:56:36] but i guess there is a limit [11:57:00] Is there a maximum number of requests i can launch in parallel [11:57:09] before it's considered an attack [13:43:24] Raminagrobis: there is guideline on 'edit' but no known restriction on mere 'read' actions afaik [13:43:46] ok thanks [13:43:55] i'll try will 10 threads or so i guess [13:44:00] *with [13:44:17] ho, those are not reads but sparql requests [13:44:37] Hmm, edits are supposed to have at least 5 seconds delays [13:44:48] (Just so you know) [13:45:35] basically i'm trying to shed some light on the everlasting debate on whether wikipedia is a sexist boy's club because only 16% of biographic articles areabout women [13:45:44] so i'm plotting the fraction of women [13:45:55] depending on birth year [13:46:03] and for some countries and some occupations [13:46:57] About 13% people born before 1900 that have an article on wpfr are women [13:47:07] for people born in the 1970's it's about 30% [13:47:40] I think asking at #wikimedia-operations (Ops guy lives there) might have better chance [13:47:51] (Though, it's weekend so people might not be around) [13:48:00] thx [15:34:32] 250 people on wikidata have "italy" as country of citizenship while they died before 1861 [15:34:38] https://query.wikidata.org/#SELECT%20%3Fperson%0A%20%20%20%20WHERE%20%7B%0A%20%20%20%20%3Fperson%20wdt%3AP31%20wd%3AQ5.%20%23%20instance%20of%20human%0A%20%20%20%20%3Fperson%20wdt%3AP570%20%3Fdeath%20.%20%23%20death%20date%0A%20%20%20%20%3Fperson%20wdt%3AP27%20wd%3AQ38%20.%0A%20%20%20%20filter%20%28%3Fdeath%20%3C%3D%20%221861-12-31%22%5E%5Exsd%3Ada [15:34:39] teTime%29%0A%20%20%20%20%7D [17:49:02] Hi, [17:49:07] Hi [17:49:35] The Galaxy SIII has many variants: https://en.wikipedia.org/wiki/Galaxy_SIII#Model_variants [17:49:55] Is it ok if I create entries for variants ? [17:50:15] It's a wiki .... [17:50:32] Just make sure that your variants fit in the ontology [17:50:51] ok, would the variant be instance of Galaxy SIII ? [17:51:32] or would it be a subclass of it ? [17:51:42] * GNUtoo is still confused with that for stuff like smartphones [17:51:49] For humains it's clear though [17:53:11] I have no idea if this is a car or a phone or a game [17:53:33] Why would it be necessary to have entries for the variants? What questions could one answer with this information? [17:53:56] I want to map precisely the hardware, and some variants are very different [17:54:07] And ideally I'd like to map the software support as well [17:54:27] For instance the I9300 variant has an exynos 4412 and is partially supported by upstream Linux, [17:54:32] (through arch/arm/boot/dts/exynos4412-galaxy-s3.dtsi) [17:54:58] Do you have sources for this information? [17:55:05] While the variant with some Qualcomm SOCs are not supported at all by upstream Linux and are like a totally different device inside [17:55:27] yes, [17:55:54] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/exynos4412-i9300.dts [17:56:04] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/exynos4412-i9305.dts [17:56:30] + the "Model variants" table on wikipedia [17:57:35] + sofware support for various Android distribution that only support certain model variant as the hardware support software is totally different [17:57:38] and I've proof of that too [17:58:08] https://wiki.lineageos.org/devices/i9300 [17:58:12] https://wiki.lineageos.org/devices/i9305 [17:58:16] The variants would probably be a subclass. [17:58:23] ok, thanks a lot [17:58:29] You would define Galaxy-Variant as a subclass [17:58:41] And then all the variants as instances of this sub-class. [17:58:56] In my way of looking at semantic modeling. There are differences of opinion ;) [18:03:09] Thanks a lot [18:03:14] I'll do that [18:37:42] As for software support there is platform, like in https://www.wikidata.org/wiki/Q20085696 [18:37:54] but it probaly wound't scale for Linux [18:38:45] So I wonder how to deal with it, an option would be for drivers for instance to represent the individual driver and state what it supports [18:38:52] But for devicetrees that's a bit harder [18:39:14] There is https://www.wikidata.org/wiki/Q16960371 [18:39:30] but It probably makes sense not to describe individual files? [18:40:20] like it would work with "exynos4412-i9300.dts" which is part of Linux, and supports the I9300, but it would create many entries [18:41:06] cd arch/arm/boot/dts && ls *.dts | wc -l => 1193 [18:41:15] So for 32bit ARM, that's about 1193 files [18:42:05] It probably makes sense to do it this way, because an entry with 1193 platforms would probably look weirder [18:42:24] *with up to something like 1193 platforms