C++ launched its community survey, too

[H. S. Teoh]

On Tue, Feb 27, 2018 at 05:46:58PM +0000, Adam D. Ruppe via Digitalmars-d wrote:
> On Tuesday, 27 February 2018 at 17:41:07 UTC, H. S. Teoh wrote:
> > And just about every new dmd release, people fume on this forum
> > about regressions and gratuitous code breakages.
> 
> Not all deprecations/code breakages are *regressions* and
> *gratuitous*.
> 
> You just need to do a cost/benefit look at it. For C++ though,
> supporting decades of very widespread use is doing to adjust the cost
> calculus of a change relative to D, of course.

Of course.  The amount of code breakage caused by a new release of dmd
is surely less than the amount of code breakage that would be caused by
an implementation of a new backward-incompatible C++ standard.  That's
why it's not going to happen.

Well, OK, there *have* been backward-incompatible changes in the C++
standard (I experienced some myself just these past 2 weeks while
updating an old C++98 project of mine... which was also motivation for
ditching C++ completely and migrating the whole codebase to D).  But
you're not going to see the sort of fundamental change that will fix
some of the longstanding inherent design problems with C++, because that
would mean alienating the majority of existing C++ projects out there.
Even if the C++ committee went ahead with such a revision, it will
surely not fly: nobody will want to implement it.

We saw a similar situation, though on a smaller scale, with the whole
D1/D2 Tango vs. Phobos fiasco. Even today that scar still persists in
outsiders' perception of D, despite all our efforts to bury that ugly
past.


T

-- 
Winners never quit, quitters never win. But those who never quit AND never win are idiots.