Topic on User talk:Vojtěch Dostál

https://editgroups.toolforge.org/b/wikibase-cli/1d9061a46973c/ stopped 20:22 (last edit)

Bot misfunctioning, see "-1" edits: https://www.wikidata.org/w/index.php?title=Special:Contributions/Vojt%C4%9BchDost%C3%A1lBot&target=Vojt%C4%9BchDost%C3%A1lBot&offset=20240111202024 , e.g. https://www.wikidata.org/w/index.php?title=Q74895&diff=prev&oldid=2051448947

Also it seems the bot runs on QIDs sorted alphabetically, cf. https://www.wikidata.org/w/index.php?title=Special:Contributions/Vojt%C4%9BchDost%C3%A1lBot&target=Vojt%C4%9BchDost%C3%A1lBot&offset=20240111192150 :

1. 19:21, 11 January 2024 diff hist −3‎ Philippe Laval-Gilly (Q73980919) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
2. 19:21, 11 January 2024 diff hist −1‎ Sadia Dehlvi (Q7397899) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
3. 19:21, 11 January 2024 diff hist −3‎ Sami Coll (Q73978780) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
4. 19:21, 11 January 2024 diff hist −3‎ Robert Dulinski (Q73978087) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
5. 19:21, 11 January 2024 diff hist −3‎ Włodzimierz Biniaś (Q73975911) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
6. 19:20, 11 January 2024 diff hist −3‎ Stefan Heckel (Q73967953) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
7. 19:20, 11 January 2024 diff hist −3‎ S Ghoul-Mazgar (Q73967863) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
8. 19:20, 11 January 2024 diff hist −3‎ Claudine Blin-Wakkach (Q73967860) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
9. 19:20, 11 January 2024 diff hist −1‎ Sachchidananda Sinha (Q7396684) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli
10. 19:19, 11 January 2024 diff hist −3‎ Frédéric Causeret (Q73958891) ‎ ‎Updated Item: fix ISNI format (details) current Tag: WikibaseJS-cli

The first item here, starting with the oldest, having a -1 edit is Q7396684, the second is Q7397899, both having a 7-digit QID, while each item having a -3 edit has an 8-digit QID. Re Q7396684, KrBot changed the ISNI less than 6 hours before the -1 edit. .

KrBot works on items sorted by QID numerically. Maybe you do the same and work in a different range? If your bot would only operate on items having an 8- or 9-digit QID some issues would go away. Maybe start with 9-digit and then 8-digit sorted descendingly.

Yes, these edits unfortunately sometimes happened when the time zone in the preceding reference was different from 00:00. This is probably due to handling of references by the Wikibase-Cli tool. I just reported it here: https://github.com/maxlath/wikibase-cli/issues/182

I have no idea unfortunately how KrBot proceeds. My bot should skip 99% of the statements which had already been changed. But you have a point, this should be fixed.

KrBot just switched to do other things

1. 17:04, 12 January 2024 diff hist +435‎ Süßen station (Q224876) ‎ ‎Created claim: operator (P137): DB InfraGO AG (Q122870674), according to item-requires-statement constraint (Q21503247) constraint from ‎Deutsche Bahn station number (P12282) current thank
2. 16:51, 12 January 2024 diff hist −3‎ Warren Viessman (Q93431766) ‎ ‎Set a claim value: ISNI (P213): 000000011659071X, см. / see autofix на / on Property talk:P213 current thank

last item KrBot edited before the switch is Q93431766 (8-digit), and your bot is now working on Q93476256, which is higher, alphabetically and numerically.

1. 17:17, 12 January 2024 diff hist −3‎ Bernard Leveneur (Q93476256) ‎ ‎Updated Item: fix ISNI format (details) current thank Tag: WikibaseJS-cli

The most recent -1 , was Q9345599 (7-digit again) . So there will still be some of these -1, but soon the ones starting with 9 will be over, my report from above had QIDs starting with 7.

The changes of the -1 edits itself look good, it should always be "Timezone +00:00" if the timestamp has "Z". Whether Z is right and why there is something else from +00:00 could be talk for another thread. These edits may slow down your bot a bit, your first big run had ~64/min, the current () has currently 34/min, but I saw it with 24/min yesterday. Anyway, not many items left.

KrBot is missing the ISNI values starting 0000 0005, probably using an outdated regex . (I manually fixed the low Qid-items in that page that have P213 which are not ISNI at all, no need to look at these.)

The report is always a bit behind. Currently there are 94 5er ISNI, the highest: Q7282339: 0000 0005 0046 4697, but given the conversion is now at the Q9... items, there are probably more 5er ISNI.

If you could take care of the 5er ISNI, that would be great. I did some manually, but this is really something for bots.

Yeah, let's wait until the job ends. Then we take care of the remaining cases. Then we need to fix ISNIs in references...

KrBot already fixed refs when value stored in P213. Values of reference URLs could be checked too, they should be the canonical URLs.

How did you define the job for your bot? Did you provide a list of QID and ISNI? Or just the rules?

Yeah, a list of statement IDs and the new values for them.

Bot did change X to x 2024-01-08 : https://www.wikidata.org/w/index.php?title=Q38768212&diff=prev&oldid=2048395316 , it was "x" before 2023-12-28 when someone changed it to X.

Bot did change "Éditions de la nouvelle plume" to "Éditionsdelanouvelleplume" - https://www.wikidata.org/w/index.php?title=Q117309896&diff=prev&oldid=2044331161

Yeah, these are inevitable mustakes. I can't adapt to changes which happened after I queried for the values.

2023-12-28 to 2024-01-08 is 10 days diff. And it fails the regex: you "maybe" could have made sure to not write values that are not conforming. On the other hand: it would have taken you more time, and so you could start earlier. They were 19 format violations pre-conversion , so not a big problem. I reviewed each, hope each is now fixed, waiting for the first report after conversion.

If you could do the 94+ ISNI starting with 0000 0005 via your list2bot, it would be great.

0000 0005 should be fine now: https://www.wikidata.org/w/index.php?title=Property_talk:P213&diff=prev&oldid=2052281314 , sorry for not finding that earlier.