@bd808 So sorry for the hold up here. So we can delete the whole project. I just checked back with the rest of the team and backuped some potential useful data. I also deleted the instances already. Thanks for taking care.

I'll backup what might be useful and then we can sunset it.

I tested the examples form the description on beta and these cases seem to be fixed. So I'll close the ticket. If there are any remaining issues. Please create a new ticket.

Just validated on beta.

Validated locally.

FYI: This task was of the type "Error report"[1]. The form for that task did not include story points. @Aklapper was so kind of changing form for that type to include story points.

I wonder, what to do with extensions, that currently do not use eslint and do not have any JS code. Should we still add eslint then "just" for the json checks?

Should we consider the task done now with the patch merged? Or what about the comment on the last patch? Should this go into an extra ticket?

We're still not catching this edge case, but the patch makes our feature more complete nonetheless.

I roughly looked at the proof of concept patch and it works fine for me and demos how the whole thing could be done. So the question _if_ this can be done is answered. I'm not entirely sure if there might be an easier way without changes to the EditPage.php class. Let's maybe talk about this on Friday.

For now we decided on supporting this feature by using the technique described in T242141: Bad numbering when reusing a sub-reference before it's defined. We might come back to this but would create a new ticket for a different approach.

Deployed and back-ported, should be fine now.

I scheduled a little retro session for next Monday.

Solution above approved in person at desk.

During a brief review on the looks of the book references layout we agreed on reducing the indentation margin a bit. The patch for that got merged and the result can be demoed on beta now.

Everything left here is tracked in T240424: Decide whether to support "extends" in the references section

This ticket was meant as an umbrella for a first version of the feature. I consider it done. Any other things we want to work on in this context should go into their own tickets.

Before closing this ticket. Is there a condensed version of the findings where we have an answer for each question from above? Or is this just about the groundwork to answer the questions above and there's still something missing to put it all together?* @awight

Fixed now and deployed just confirmed on de.wikipedia

@JStrodt_WMDE Can be tested and confirmed e.g. using https://en.wikipedia.org/wiki/Special:SimulateTwoColEditConflict

@JStrodt_WMDE Can be tested and confirmed e.g. using https://de.wikipedia.org/wiki/Spezial:SimulateTwoColEditConflict

This seems to be fixed and deployed. I just confirmed using FF and the example link above.

Looks good, thanks for the work. I consider this done.

Just for the sanity of not just demoing to ourselves @JStrodt_WMDE :

In T240431#5751451, @awight wrote:

There might not be a way to measure the impact of this change. We aren't measuring the total memory extent of each thread, and garbage collection is (allegedly) disabled, with no recorded metrics that might show us decreased memory pressure.

This is fixed and deployed, I could just confirm this on de.wikipedia

In T239792#5751305, @awight wrote:

As can be seen in patch https://gerrit.wikimedia.org/r/#/c/mediawiki/core/+/530148/ , if we hook in after the external 3-way merge tool, we can detect whether automatic resolution failed.

@thiemowmde @awight To the remaining open steps above:

Solution is deployed and seems to work, see confirmation in T238991#5776382

In T238991#5776382, @Jack_who_built_the_house wrote:

Can anyone from the ruwiki community confirm?

I confirm: a newly registered account didn't get Reference Previews.