Hello Eric,

I dug around for why I filed for the change in behavior.
Here is the bug report that prompted it:

    https://coq.inria.fr/bugs/show_bug.cgi?id=2726

ctags -e is not a perfect replacement for etags; in this
example, 'make tags' calls out to etags and expects the Emacs
version.

etags = emacs tags.  It seems only natural for it to have priority,
and tools which refer to it are likely to assume emacs tags. If
I am an emacs user, and I install 'ctags', I don't expect my 'etags'
executable to suddenly change; I should ask for it specifically.

Though, yes, I agree the behavior change is unexpected and confusing,
if you had been using ctags and then installed emacs.

Cheers,
Edward

Excerpts from Colin Watson's message of 2015-04-07 17:59:53 -0700:
> On Tue, Apr 07, 2015 at 06:58:23PM -0400, Eric Cooper wrote:
> > After installing exuberant-ctags:
> > 
> > $ sudo update-alternatives --config ctags
> > There are 2 choices for the alternative ctags (providing /usr/bin/ctags).
> > 
> >   Selection    Path                      Priority   Status
> >   ------------------------------------------------------------
> >   * 0            /usr/bin/ctags-exuberant   30        auto mode
> >     1            /usr/bin/ctags-exuberant   30        manual mode
> >     2            /usr/bin/ctags.emacs24     27        manual mode
> >       
> > 
> > $ sudo update-alternatives --config etags
> > There are 2 choices for the alternative etags (providing /usr/bin/etags).
> > 
> >   Selection    Path                      Priority   Status
> >   ------------------------------------------------------------
> >   * 0            /usr/bin/etags.emacs24     27        auto mode
> >     1            /usr/bin/ctags-exuberant   10        manual mode
> >     2            /usr/bin/etags.emacs24     27        manual mode
> > 
> > This was unexpected and confusing.
> 
> This was explicitly requested in https://bugs.debian.org/668560.  I
> don't use Emacs and have no intuition for what its users might want, so
> I've CCed the reporter of that bug and perhaps the two of you can argue
> it out and arrive at a consensus?  It doesn't seem like a good idea for
> me to flip-flop on this.
> 


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to