From unknown Sun Jun 15 01:08:25 2025 X-Loop: help-debbugs@gnu.org Subject: bug#26986: 26.0.50; Unsafe recovery should be off by default in batch mode Resent-From: Philipp
> From: Philipp <p.stephani2@gmail.com>
> Date: Thu, 18 May 2017 21:44:22 +0200
>
> In batch mode there will usually be no long-lived buffers with
> non-recoverable contents, therefore the reasoning for
> `attempt-stack-overflow-recovery' and
> `attempt-orderly-shutdown-on-fatal-signal' being t doesn't app= ly.=C2=A0 This
> means that these variables should be nil by default in batch mode.
I see your point, but what are the arguments for changing their values
in batch mode?=C2=A0 IOW, what advantages will this give us?When debugging Emacs internals or running unit tests,= setting these variables to nil will cause obvious and clear failures inste= ad of silent recovery attempts.=C2=A0
> From: Philipp Stephani <p.stephani2@gmail.com>
> Date: Thu, 18 May 2017 20:26:13 +0000
> Cc: 26986@d= ebbugs.gnu.org
>
>=C2=A0 I see your point, but what are the arguments for changing their = values
>=C2=A0 in batch mode? IOW, what advantages will this give us?
>
> When debugging Emacs internals or running unit tests, setting these va= riables to nil will cause obvious and
> clear failures instead of silent recovery attempts.
OK, but the batch mode is not used only for debugging or testing.
Perhaps we should change the values in these two cases.
> From: Philipp Stephani <p.stephani2@gmail.com<= /a>>
> Date: Thu, 18 May 2017 20:26:13 +0000
> Cc: 26986@d= ebbugs.gnu.org
>
>=C2=A0 I see your point, but what are the arguments for changing their = values
>=C2=A0 in batch mode? IOW, what advantages will this give us?
>
> When debugging Emacs internals or running unit tests, setting these va= riables to nil will cause obvious and
> clear failures instead of silent recovery attempts.
OK, but the batch mode is not used only for debugging or testing.
Perhaps we should change the values in these two cases.