disable save button if visualization is dirty

[scampi]

close #11575

[elasticmachine]

Can one of the admins verify this patch?

[scampi]

I'm going to try to write a small test for this.

[scampi]

PR is now ready for review.

[thomasneirynck]

hi @scampi, my apologies for the little delay on this one. I'll take a look later today. In the meantime, do you mind rebasing and resolve merge conflict when you get a chance? thx!

[thomasneirynck]

This is a very nice UX improvement. thx!

The merge conflict is pretty minor, some other methods were inserted at the same location in the UI-test helpers, so rebase should be pretty straightforward.

LGTM pending rebase.

[thomasneirynck]

this is why you're awesome @scampi. Community PRs rarely include UI-tests!!!

[thomasneirynck]

@ppisljar you want to take 2nd look? thx

[scampi]

@thomasneirynck thanks for the kind words! I have rebased against master. I added a little 3rd commit so that the tests make more sense.

[ppisljar]

LGTM

[thomasneirynck]

5.x backport: #11816

[thomasneirynck]

I'm going to revert this as it introduced two UX regressions (#12082). I missed those. I think this is a pretty good change still, but we'll need to look into two issues:

• changing the time-filter should not cause the save-button to be disabled
• disabling the save button prevents users from doing a save-to-rename operation