You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Nov 25, 2021. It is now read-only.
Repository navigation
This repository was archived by the owner on Nov 25, 2021. It is now read-only.
TOOLTIP_DISPLAY_DELAY is not respected for continuous mouse movement within a token #56
What do you mean by considering mouse movement by pixels to tokens?
It currently(and always has) gotten the events from the tokens dom element itself. This means that the event is triggered from the instant the mouse first goes over the token and that point in time is where the delay starts counting.
When we fully switch over to event-positions and are using pixels rather than the event target we can easily start the delay when it stops moving. It’s still possible now but would probably require some tricky logic.
I would agree it's not a big problem (especially if it's already changing as part of a bigger change), but the UX of just moving the cursor normally and having many tooltips momentarily fade in and out is pretty obnoxious and distracting. If this library is entirely about adding tooltips to code, I'd argue that producing these tooltips in an overly-noisy manner is very much a bug.
Activity
ijsnow commentedon Sep 8, 2018
I don’t think this is a regression.
What do you mean by considering mouse movement by pixels to tokens?
It currently(and always has) gotten the events from the tokens dom element itself. This means that the event is triggered from the instant the mouse first goes over the token and that point in time is where the delay starts counting.
When we fully switch over to event-positions and are using pixels rather than the event target we can easily start the delay when it stops moving. It’s still possible now but would probably require some tricky logic.
chrismwendt commentedon Sep 8, 2018
Oh, it might not be a regression in the browser extension. I was thinking of https://github.com/sourcegraph/sourcegraph/issues/11457 (in the Sourcegraph web app)
Could one of the solutions in https://github.com/sourcegraph/sourcegraph/pull/11467 or https://github.com/sourcegraph/sourcegraph/pull/11768 be applied here?
felixfbecker commentedon Sep 8, 2018
That code is in codeintellify: https://sourcegraph.sgdev.org/github.com/sourcegraph/codeintellify@e6a1f16c94058c12cc58a8085d623ee2171250f6/-/blob/src/hoverifier.ts#L308-329
chrismwendt commentedon Sep 8, 2018
@felixfbecker Any idea why this behavior is happening? I'm noticing this on the web app too
felixfbecker commentedon Sep 8, 2018
No, but I would say it's not really a big problem, is it? It is probably not worth investigating if this all changes with
event-positionsnnchang commentedon Oct 4, 2018
I would agree it's not a big problem (especially if it's already changing as part of a bigger change), but the UX of just moving the cursor normally and having many tooltips momentarily fade in and out is pretty obnoxious and distracting. If this library is entirely about adding tooltips to code, I'd argue that producing these tooltips in an overly-noisy manner is very much a bug.