[Controls] Options List Auto Select Top Option on First Load

[mattkime]

This will reduce the number of clicks for a user to see correct data in some circumstances.

[elasticmachine]

Pinging @elastic/kibana-app (Team:KibanaApp)

[ThomThomson]

If we add a setting auto select top option on first load to the Options List Control, we can give the author a way to make the dashboard always load with a selection in place, without having to specify which option should be selected.

If we combine this with sorting options, we unlock a lot of potential features, like a Control always selecting the latest Date, or always selecting the first department alphabetically, even when the departments change.

[Heenawter]

I'm not convinced that this is a worthy option to expose to the user.

If the author of a dashboard wishes to have a "default selection" when the dashboard first loads, they can simply make the selection and save it to the Dashboard. I think that having a "Auto select top option on first load" option adds extra complexity for a very specific use-case that already has a very easy workaround.

Thoughts @ThomThomson? If you agree, I think we can close this issue :)

[ThomThomson]

Hey @Heenawter, good call out! I agree and will close this, but with the caveat that we understand and write out all use-cases will think will be left uncovered by closing this.

I've seen only one use-case where this is valuable:

Having a dashboard which is supposed to always be filtered by a value and only one value. This is usually some sort of version of the data. Looking at multiple versions of the data at once doesn't make sense, and setting the dashboard to select a specific version on load means that the dashboard has to be edited every time documents are ingested with a new version value.

In these cases, selecting the top version numerically every time the dashboard loads would be the right call.

Do I know how prevalent this use case is? Not at all.
Do I think that this is a normal way to use Kibana? Not really, but I could be proven wrong.

Knowing these two things, is it worth the additional DX / UX complexity to introduce this option?
I don't think so, so I'm closing this one. Thanks for bringing this up!

If anyone can think of another potential use case for this feature please bring it up here! If it's compelling enough we can potentially reopen this.