Don:
I agree completely. I would say that the #1 problem in D is the
paranoid fear of breaking backwards compatibility. I said that
in my 2013 talk. It is still true today.
Sociomantic says, PLEASE BREAK OUR CODE! Get rid of the old
design bugs while we still can.
I keep a large amount of working D2 code, mostly for
scientific-like usages (plus other code outside work, like almost
one thousand of Rosettacode programs, a lot of lines of very
carefully written good code). Most of such code is divided in
many small programs. I have plenty of unittests and I also use
contract programming. A planned breaking change, if it gives
clear and nice error messages, allows me to fix the whole code
base in a limited amount of time (1 hour, or 2 hours or a little
more), and often it's an easy work that I can do late in the day
when I am too tired to do more intelligent work anyway. Compared
to the work to understand the problems, invent the problems,
invent the solutions, and write working code, the amount of brain
time & work to keep code updated is not significant for me.
So I suggest to deprecate the built-in sort, deprecate other
things that are waiting for it since some time, do the other
language updates that you think are good.
Bye,
bearophile
