Re: recent change to yanking behavior of the clipboard

[Stefan Monnier <monnier@iro.umontreal.ca>]

Po Lu [2022-02-07 09:18:14] wrote:
> I see no point in sticking to behaviour obsoleted by the operating
> system, such as storing cut text in the primary selection.  On X, every
> client has to cooperate with every other client in order to have working
> cut and paste, and a single client deviating from that is sufficient to
> bring down the entire deck-of-cards.

Now I'm lost.  What you say here seems to claim that the change under
discussion brings Emacs in line with the official/standard protocol.
Yet that change is very specific to the `CLIPBOARD` whereas I'd expect
such an official protocol to behave the same for all types of selections.

Elsewhere you said that the motivation was one of optimization
of performance.  Yet Noah's original message points out a change in
actual behavior rather than mere performance.

I'm having difficulty reconciling those somewhat contradictory facts.

Could someone walk me through step-by-step what happens in one or two cases,
(e.g. Noah's case, or the "normal" case) explaining what Emacs does and
why, how the patch affects the outcome and what the ideal behavior would
be according to the standard protocol?


        Stefan