Re: Missing snprintf in ucrt mingw + vc-refresh in find-file hook?

[Dmitry Gutov <dmitry@gutov.dev>]

On 14/02/2024 23:04, Arthur Miller wrote:

>> The idea behind it was something else: to be able to unify mode-line elements
>> (from vc and project).
> 
> Ok, I see. A question: is it importnat to keep display on modeline or
> elsewhere decoupled from the actual repo query?

Probably, yes. If I understood your question right.

>>> In case when list of files are checked, one can let-binding vc-display-status
>>> to nil in vc-refresh (or the responsible one) so to not trigger git?
>>
>> vc-display-status only affects vc's mode-line element.
> 
> Can I use to in the meaning "no status display" = "no git query"? That is why I
> ask if it is important to keep display decoupled from the query.

If we do it this way, it will just be an incremental increase in the 
complexity of the code, and it will support only a part of the 
applicable use cases. I think I've explained that in one of the previous 
emails.

>>>> So I would welcome such an experiment, especially if one is careful to retain
>>>> support for vc-follow-symlinks.
>>>>
>>>> vc-after-save could similarly avoid doing the full refresh until the file's
>>>> backend/state are requested again.
>>> When they request for the backend state; they could do so asyncrhonously to,
>>> by
>>> starting a timer? The update will not be immideate; run first when Emas is idle,
>>> if it is just the modeline; but third party apps, if there are such, can be more
>>> picky.
>>
>> That's not simpler.
> 
> After looking more at it; what I find problematic is that it is automatic. Also,
> use of hooks find-file/after-save (eventually) means either all or nothing.
> 
> I looked a bit in vc-hooks and vc-dir, and come up with this little ugly hack:
> 
> (define-minor-mode vc-mode
>    "vc-mode test"
>    :global nil :lighter " vcm "
>    (setq-local vc-handled-backends
>          (if vc-mode
>              '(RCS CVS SVN SCCS SRC Bzr Git Hg)
>            nil)))
> 
> Since vc-mode is just a dummy function; I per-used it. Now I can keep in my init
> file (setq vc-handled-backends nil) and nothing screems after Git when it is not
> in the path; and in projects where I wish to work with vc-mode and run vc-dir, I
> can start vc-mode and than vc-dir works as expected.

If you're just looking for a hack for personal use, this one's as good 
as any.

> I also don't see how Emacs could know when to start vc-mode and when not,
> without user taking an action. If complete automatic handling of version control
> is desired than something like that is not acceptable.

It would be whenever some actual feature requested the vc state to be 
computed because it's going to use it now.

> Anyway, I think I can personally live with my hack, unless there is something
> really nasty I am forseeing there. That achieves what I asked for, without you
> needed to change anything.

I don't think you're going to have any big problems with the above 
setup, as long as you remember to turn on vc-mode to use any features 
that require VC. Such as diff-hl-mode, for example.