Andrei's list of barriers to D adoption

[Jonathan M Davis via Digitalmars-d]

On Tuesday, June 07, 2016 15:47:10 Walter Bright via Digitalmars-d wrote:
> On 6/7/2016 3:23 PM, Timon Gehr wrote:
> > Obviously they proved the virtual machine itself memory safe,
>
> As I recall, the proof was broken, not the implementation. People do make
> mistakes and overlook cases with proofs. There's nothing magical about them.

Yeah. I recall an article by Joel Spoelsky where he talks about deciding
that proofs of correctness weren't worth much, because they were even harder
to get right than the software.

I do think that there are situations where proofs are valuable, but they do
tend to be very difficult to get right, and their application is ultimately
fairly limited.

- Jonathan M Davis