Re: Intervention needed?

[Jere <jhb.chat@gmail.com>]

On Wednesday, March 13, 2019 at 5:10:31 AM UTC-4, Maciej Sobczak wrote:
> OK, playing the devils's advocate again.
> 
> > As far as I can see, Rust's only innovation is safer pointers,
> 
> Which is, arguably, the innovation many people have been waiting for.
> 
> > but it still has 
> > a bad design that requires the use of pointers.
> 
> Why it is bad?
> Before, the criticism of pointers was based on the fact that pointers are dangerous. So, Ada is better, because it avoids these dangers.
> Now we have guys with *safe* pointers and what? They are bad, because they use pointers? And pointers are bad, because they are dange... wait, something's went wrong.

The harder sell of Jeff's argument is the same issue we constantly 
have heated discussions about in comp.lang.ada: proper terminology. 
Rust technically does have pointers, but they are not what Jeff is
talking about.  He's talking about references, which, technically,
not really pointers (they don't dangle, can't be freed, cannot be
shared, and aren't required to be implemented via pointers under 
the hood...though they probably are).  Trying to use a pointer in
Rust will net you a compiler error out of the box.  Rust's whole
mission is to discourage the use of pointers.

> And, surprise, nobody cares. Because now their pointers are safe.
> 
> > Other than that, Rust seems to be a typical C-family language with all the 
> > dangers that implies.
> 
> What dangers? Like for example pointers? Oh, wait...
> 
> And you still wonder why you are not convincing and cannot attract newcomers to the language.
> 

I just think in discussions like this, especially when they cross over into
other communities, shouldn't devolve into "you're wrong!", which can easily 
happen since many of us get pretty passionate about Ada.  Ada is the better
language overall, but if we go in too strongly and can't even use the 
terminology that Rust uses properly, then our arguments will just be ignored
and Ada doesn't get the advocacy it deserves.  We just end up looking like
zealots who cannot even research what we are talking about.  Note, that I
am not saying this has happened yet, but based on how the discussion is 
going in this email chain, if it were to spill over into the Rust thread, 
I would be concerned that it would happen.  I apologize if this sounds
accusatory.  I didn't intend it, but not sure how to write this out 
without it sounding that way.

That said, I think it would be bad for the Ada community to not appreciate
some of the good things that Rust does better than Ada.  They aren't many
and certainly aren't enough for me to think it is a better language, but
to blindly ignore them is a long term mistake for the community as a
whole as it only hurts Ada.  I think there is a lot to be gained if both
the Rust community and Ada community collaborated instead of competed, 
for both languages.