On Fri, Sep 3, 2021, 13:30 Eli Zaretskii <
eliz@gnu.org> wrote:
> From: João Távora <joaotavora@gmail.com>
> Date: Fri, 3 Sep 2021 13:11:41 +0100
> Cc: Lars Ingebrigtsen <larsi@gnus.org>, "Philip K." <philipk@posteo.net>,
> Daniel Fleischer <danflscr@gmail.com>, Richard Stallman <rms@gnu.org>, emacs-devel <emacs-devel@gnu.org>,
> Stefan Monnier <monnier@iro.umontreal.ca>, Eli Zaretskii <eliz@gnu.org>,
> John Yates <john@yates-sheets.org>
>
> Additionally, I would like to point out that we need reproducible bug recipes from bug reporters, and the best
> way to get them is something like Emacs -Q and its defaults. It is unavoidable that offering profiles as
> suggested is going to have a negative impact in that department. But perhaps it can be controlled if we have
> a small number of profiles and said profiles are easily chosen from the command line and the bug reporting
> instructions should updated accordingly.
I don't think this is a significant problem. We could add the active
"profile(s)" to the report by report-emacs-bug, and that would take
care of reproducing the issues that don't happen in "emacs -Q".
Bear in mind that other packages are affected where bugs are not reported via report-emacs-bugs. But I agree that there is still a way -- there always is -- to determine the configuration that a user is using. My point is that it should be very easy to do that, i.e to report and control that big set of preferences, almost as easy as it is when one uses Emacs -Q, which is, for me, also a way to control to a very large number of preferences.
Even with the convenience of Emacs -Q, I already have a lot of trouble explaining and attaining reproducibility from some users. So, as a package developer, I'd appreciate Emacs -Qprofile-name (or equivalent). This is essential to interact with users who will, over time, become completely helpless when presented with a good-old-days Emacs -Q (with non-vi bindings, of all things...)
João