Re: emacs rendering comparisson between emacs23 and emacs26.3

[Dmitry Gutov]

On 07.04.2020 11:32, martin rudalics wrote:
> It depends on what we consider a failure: I do not think that inserting
> a "/*" in my window showing the top of my C buffer should cause the
> window showing the bottom of my buffer consider everything part of a
> comment starting at that "/*".  I think that the comment should stop at
> the first paren in column zero after that "/*" even if that paren is
> shown in the window showing the top of the buffer and even if that's
> syntactically wrong (and hence the compiler would see it differently).

To be accurate, the comment started by '/*' would end at the first 
occurrence of '*/' somewhere down the buffer.

If you insert or close a paren, however, that would definitely change 
syntax state at EOB.

The former case is possible to support better in syntax-ppss, FWIW. The 
latter we probably never will (in syntax-ppss).

> So what I would vote for in this case is that font-lock would specially
> highlight the open paren that "stops seeing the comment as comment" so
> the user is aware of the fact.  Would that be so hard to implement?
> Even with a slight change in 'parse-partial-sexp' and 'syntax-ppss' that
> would accompany it.  Optionally, by default turned off, so nobody who
> wants the correct syntactic interpretation would be bothered.

According to Alan's experiment, you can get this behavior now already 
simply by setting comment-use-syntax-ppss to nil and 
open-paren-in-column-0-is-defun-start to t.

Does it improve your experience significantly?