On 11.08.2015 14:12, Tom Browder wrote:
I have seen several lists of new Perl 6 features (versus Perl 5) but
they all seem to be lists that intermix features with varying degrees of
value to "ordinary" Perl 5 users. If one wants to sell long-time Perl 5
users (already using the latest Perl 5, Moose, etc.) on the value of
Perl 6, what should be on the important feature list?
on a more meta level: Perl 6 has the ability to evolve, and lots of
things that Perl 5 most likely will never have.
Just as an example: when I started to get involved with Perl around 2003
or '04 or so, people seemed to be aware that subroutine signatures were
a good thing to have. They probably were aware of that much longer. And
yet it took until perl 5.20 in 2014 to get even the most basic
subroutine signatures (and still marked as experimental).
Another one is the inability to reliably introspect strings for whether
they are text or binary data, which means enormous care must be taken to
not accidentally mix the two. Again, known for ages that it's a problem,
afaict no practical solution in sight.
These are just two examples of things that people take for granted in a
modern programming language, yet Perl 5 has real trouble with.
I could continue with other Perl 5 deficiencies (no strict by default,
lack of easy threading, too many globals, obscure regex syntax), but the
individual problems aren't the point. My main point is that large parts
of Perl 5 are still stuck in the past, with no good way forward.