RE: AW: delete-selection-mode

[Drew Adams]

> > This will be a good occasion to get rid of the butt-ugly and
> > bug-inducing code that implements the feature now.  And it 
> > will provide the DEL part of delete-selection-mode.
> >
> > I'd be happy to also provide the self-insert behavior of
> > delete-selection-mode, but at least this intermediate step 
> > sounds very good to me.
> >
> > The implementation should keep an eye towards generalizing it to
> > self-insertion keys (i.e. maybe it should just use the code from
> > delete-selection-mode).
> 
> That all sounds perfect.
> 
> The "DEL deletes" functionality seems the main thing; a mode 
> which only did that would probably be fine, and solve 95% of
> the muscle-memory-from-other-apps issues.

Why? How will that help users who expect the usual type-to-replace behavior?

What's so special about DEL?  IIUC, your proposal essentially binds DEL to
`delete-region' when the region is active.

I disagree that DEL amounts to 95%, or even 5%, of the use of an active region
in other apps. DEL is simply one minor, special case of type-to-replace. Most of
the time, users of other apps select text to replace it - THAT's the 95% or more
case.

Your proposal is a miniscule improvement, one baby step on the road to the
d-s-mode behavior that newbies expect (and many oldbies prefer). What's the
point of such a watered-down compromise?

An active region should be active, including in the sense that most people
expect: type-to-replace.

If someone doesn't want the region to be active, there are trivial remedies:
turn it off either temporarily (C-g) or permanently (turn off t-m-mode).