[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Renaming eglot -- or at least add an alias?

From: Eli Zaretskii
Subject: Re: Renaming eglot -- or at least add an alias?
Date: Thu, 06 Oct 2022 09:17:13 +0300

> From: Tim Cross <theophilusx@gmail.com>
> Cc: Eli Zaretskii <eliz@gnu.org>, emacs-devel@gnu.org
> Date: Thu, 06 Oct 2022 11:13:43 +1100
> Richard Stallman <rms@gnu.org> writes:
> > How to deal with Eglot and Semantic is not clear to me.
> >
> > Is there any programming language which is supported both by Eglot and
> > by Semantic?
> I think this question still shows a fundamental lack of understanding
> regarding what eglot is and does.

I don't think there's a misunderstanding here, no.

> Eglot doens't know anything about programming languages. What languages
> it is able to work with is determined by the available language
> servers. The available language servers change as new ones are created
> and old ones die off. They are implemented in different languages and
> have nothing to do with Emacs or elisp.

That is immaterial.  To Emacs users, Eglot+language server is a single
entity that provides and/or enables some Emacs features.  Semantic
provides and enables some of the same features.

> When there is overlap, is it the same functionality? No, not
> necessarily.

Actually, yes, it's the same functionality.  Xref support is one area
of overlap, refactoring support is another, semantic analysis is yet

Semantic was developed with a similar goals in mind, so it isn't
surprising that there's overlap in functionality.  Those goals were
the reasons why we added parts of CEDET to Emacs years ago -- we hoped
they will be the basis of an Emacs-based modern IDE capabilities.

Experience taught us that these hopes didn't materialize, and
meanwhile the technology moved to other solutions.  So now we want to
adopt those solutions to Emacs, and naturally there will be clashes
with Semantic-based functionalities.  We'll need to resolve that as we
go, probably deprecating Semantic in the long run.

> This again brings us around to a core stumbling block - how do you have
> a name which describes what something does when there are multiple
> packages which do that thing? 

This dispute is from my POV a complete waste of time, energy, and
bandwidth.  I wish people have stopped arguing, because it won't
change anything in the short run.  As I already said at the beginning
of this useless thread.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]