Re: Delete variables obsolete since Emacs 23

[Gregory Heytings]

> > In that case, would a two-step process not be better?  First declaring 
> > the function "obsolete", and when it is known that it will be removed 
> > declare it "pending-removal" with a target major version.
>
> I'm not sure it would make much difference to the speed at which people 
> move away from using the obsolete feature (which is really all that 
> matters, in the end).

FWIW, I think it would.  With the current situation there is no concrete 
reason to move away from such a feature.  As Eli wrote "we don't really 
know whether we will remove [an absolete feature], let alone when", so it 
could very wall stay there forever, and I can understand Drew's viewpoint 
who did not make the effort to update the 2500 lines of his codebase in 
which 'interactive-p' is used.  If he had been warned, when Emacs 25 was 
released, with:

(declare (pending-removal called-interactively-p "27"))

he would have known that it would be removed in a medium term, and would 
have had a reason to update his codebase.

That being said, about this particular feature, I still don't understand 
why it is obsolete.  It is just a way to express a common pattern in a 
shorter way, like say (1+ x) and (+ x 1).

Gregory