Problems with the "purge" command

[Dr Rainer Woitok]

Craig,

On Friday, 2023-06-16 10:07:22 -0600, you wrote:

> On Fri, Jun 16, 2023 at 9:54 AM John W <jwdevel at gmail.com> wrote:
> 
> > I find that putting "hgext.purge =" in my [extensions] makes the
> > difference.
> > ...
> 
> This is currently the only way to get purge to not query. When purge was
> moved to core, the default behavior was changed to always
> query the user.

In my world any discrepancy between some piece of code  and its documen-
tation clearly is a bug.  Either in the code or in the documentation.

Besides,  my experience has taught me  that having code  ask back before
doing something irreversibly destructive  as in the classical suggestion
of having "alias rm=rm -i" is futile due to human psychology:  after us-
ing such a command for a while  and getting used to typing "y" and hitt-
ing "ENTER" after each invocation, you will eventually do that automati-
cally,  and will only  realize your glitch  after the fact,  just as you
would without being asked back :-/

So for the case at hand I'm clearly voting for having the code implement
what the documentation describes, and for the time being I humbly revert
to the good ol' solution of using "!$HG status -nu0  |  xargs -0 rm -f",
which -- by the way -- is still advertised in "man 5 hgrc".

Sincerely,
  Rainer