[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#68568: transient.el interns potentially enormous symbols as commands
From: |
João Távora |
Subject: |
bug#68568: transient.el interns potentially enormous symbols as commands |
Date: |
Sun, 21 Jan 2024 08:10:33 +0000 |
On Sun, Jan 21, 2024 at 7:29 AM Psionic K <psionik@positron.solutions> wrote:
>
> > long descriptions of infix commands
>
> That's a user choice btw. Anyone can make long symbols through a
Not sure what you mean by "user". I'm a "user" of GPTel, but
I didn't make those long description. GPTel uses those descriptions
and, I think, reasonably.
> variety of means that are out of control of any package.
Sure, but that package shouldn't gratuitously have that choice
(questionable as it may be -- I think it's not) wreak havoc everywhere
else in Emacs. This is what happens with transient's interning
of such symbols
> definitely take alternative routes to avoid creating long symbols in
> transient.
I'd rather not bother transient-using packages with code complications.
GPTel's interface is fine as it is.
> 200 and how many instances?
About 8 or 9, all transient's. Doesn't matter because the greediness of the
flex matching algorithm, which was designed for short symbols,
eventually picks up on a pattern which is highly likely to appear in
the extremely long tail of the enormous symbol
So my usual searches like M-x vc SPC always where I expect to see
"vc" commands turn up some silly transient command, useless and
unwanted in that context.
> > transient.el works just fine with make-symbol instead of intern.
>
> It sounds like you have a patch ready for the upstream. Bon appetit:
> https://github.com/magit/transient
I take it you still can't think of a good reason why those symbols
should be internet. OK. Created a PR. I hope this gets fixed
quickly over here at the downstream. This is not something easy
to monkey-patch around.
João
bug#68568: transient.el interns potentially enormous symbols as commands, Eli Zaretskii, 2024/01/27