bug#61637: 30.0.50; Fix Eglot tests that need HOME=~USER

[Basil L. Contovounesios]

João Távora [2023-02-19 19:38 +0000] wrote:

> Both patches look generally good, though I haven't tested them,
> so I'm trusting you did and they generally add robustness.

Thanks, indeed 'make check' now/still succeeds locally.

> If that's the case, go ahead and push (to emacs-29, presumably
> since this is fixing bugs in the test suite, but master isn't that
> bad either).

The small change in ert-x.el, and the proximity to an emacs-29 RC, are
the reasons I wanted to double check with Eli about where these changes
should go.

Perhaps I should split out the ert-x.el change for master, and the rest
can go to emacs-29?

> The only nit I would point out is that there seems to be some
> unrelated housekeeping already in the first patch that could
> be moved to the second patch, or maybe a separate commit.

You mean, the indentation and commentary fixes?  The impression I got is
that these kinds of changes are more welcome in emacs.git when the
surrounding code is already being touched, as opposed to making small
whitespace-only changes to functions that are not otherwise being
changed.

I have little personal preference either way.

> The yasnippet-related fix could also be its own commit.  But
> again, that's only a minor nit.

Sure, I can break out the autopep8 and YASnippet changes if that's
preferred.

Thanks,

-- 
Basil