Page MenuHomePhabricator

The limited list of suggested categories don't ease categories searching
Open, LowPublic


When you are looking for a category using VE, the list is limited and user have no idea about it.

For instance, typing "Série télévisée se déroulant en A" on French Wikipedia displays a limited list, ending with "Série télévisée se déroulant en Alabama", skipping places beyond like "Série télévisée se déroulant en Arkansas".

You don't know how many letters remain to be types to finally find what you are looking for, also because of T187341: Categories dropdown should be larger to ease categories' selection. So, at the end of the dropdown, it should be indicated something like 265 categories listed, please type more" or something similar. That's the case when you search for categories: the number of results is displayed at the top of the page.

Solve T59302: Suggest case insensitive results when searching for categories to add would help as well.

Event Timeline

Deskana added a project: Discovery-Search.
Deskana moved this task from To Triage to External and Administrivia on the VisualEditor board.
Deskana added a subscriber: Deskana.

I suspect that this is a more generic issue with how the search backend is calculating the relevance of the results to your query. Given that category searching is both niche functionality in the visual editor, and also niche functionality more generally, this is a low priority issue.

Agreed, Deskana, too niche to put a lot of time into, and if we get more use cases we'll take a deeper look.

I'm a little surprised by the decision to close this without any action, or even merging it to a related one. Category search is currently difficult for editors. Declaring categories to be "niche" and closing the request for improvements doesn't make the actual problem disappear.

The situation reminds me of when it was almost impossible to tag a task for the search team, because Phabricator's search only listed five options, and all of them were longer than the team's name. We fixed the Phab problem, and IMO we should (in due course, with appropriate consideration for competing priorities) fix this one.

@Whatamidoing-WMF I'm glad to leave this open, but the search team (which essentially is two engineers who could possibly do this) does not have the bandwidth to work on this anytime soon. The decision to decline it came after discussing it during our weekly backlog grooming, when we decide which of the items that have landed in our triage bucket (an average of 10 new requests per week or so) we may be able to take on between the planned work we have committed to for the year. Clearly, we can't commit to everything, and this ticket was declined with an invitation to provide more use cases, which would then raise it as a higher priority to look at in the future. Anyway, re-opening the ticket, feel free to elaborate the need.

@EBjune: To me that would translate to "lowest priority" if you are not actively against seeing this task fixed. Or removing the team project tag while keeping the (valid?) task open.

huh I thought the team project tag was removed...removing (again)