| TechnoSquirrel69 | |
| Dec 25 2024, 3:28 AM |
| F58055285: image.png | |
| Dec 25 2024, 3:28 AM |
Happy holidays! As of 03:21 UTC on December 25, 2024, when I took the screenshot below, the Thalun-class frigate article was ostensibly marked as reviewed by a reviewer, but there was no log entry (log). It's unclear whether it was actually reviewed and the review action did not create a log entry or if it was mistakenly marked as reviewed by PageTriage with no manual action. I have added it back to the queue out of an abundance of caution.
https://en.wikipedia.org/w/index.php?title=Special:Log&logid=166688104
Ryan shell talk contribs marked revision 1265018253 of page Draft:Thalun-class frigate patrolled
This is probably what made PageTriage think the page was reviewed,
I think it's working as designed. Both a patrol action (clicking "mark this page as reviewed" in lower right corner, patrol log) and a review action (clicking "mark as reviewed" in the page curation toolbar, page curation log) should result in the page being marked as reviewed.
In general, most people use the "page curation log", but some few outliers and situations still use the "patrol log".
https://en.wikipedia.org/wiki/Wikipedia:New_pages_patrol#Patrol_versus_review
That's working as designed. What's not working as designed is that you can use the core patrol feature to mark a page as reviewed while it's still in draft namespace, and the reviewed status wasn't reset when it was mainspaced.
Looking at the logs for the draft reveals a potential complication: the draft page was marked as patrolled about 15 minutes after the page creator moved it to mainspace, so the edit conflict (patrol conflict?) may be part of the issue.
This ticket confuses me. Should I revert the edits I made to the original post? Was my original first comment correct and we should close this?
I've reverted my edits to the ticket. I think this is probably a case of it being in the patrol log rather than the page curation log, as mentioned above. I don't think there's a bug here.
@TechnoSquirrel69, if I'm missing something, feel free to edit the ticket to make it clearer what the bug is. Or feel free to close the ticket if the patrol vs page curation log explanation satisfies you.
I disagree, but I'm unsure exactly where the issue lies. Again, the only visible action in the logs is the out-of-date draftspace page being marked as patrolled after it had already been moved to mainspace, which somehow translated into the mainspace page being marked as reviewed. I agree with Pppery that this behavior doesn't seem at all intentional, and I can't imagine any reviewer is actually using a workflow of marking draftspace pages as patrolled when NPP-ing in mainspace.
According to this quote, it does sound like someone marked it as patrolled after it was moved to mainspace. An entry either in the patrol log or the page creation log counts towards an article being marked as reviewed. So this seems like it's working, at least from a technical standpoint.
Maybe the change to make here is to make this less confusing? Maybe we should make changes to the code that listens for a patrol action, and always make sure to do a duplicate/corresponding/identical page creation log action. Thoughts?
That would be nice. But this bug wasn't caused by that. Unrelated to anything PageTriage does, this log entry shouldn't have been possible:
13:46, 24 December 2024 Ryan shell talk contribs marked revision 1265018253 of page Draft:Thalun-class frigate patrolled (thank)
Revision 1265018253 belongs to "Thalun-class frigate", and not "Draft:Thalun-class frigate".
Filed T383108: Patrolling generates self-contradictory log entries for that impossible log.
Closing this ticket as invalid as it's turned into a confusing mixture of multiple different things, and it's best to start from scratch. Feel free to file a different ticket for improving the interactions between patrol and review if warranted.
And I think that's everything here.