www.digitalmars.com         C & C++   DMDScript  

D - Standard Libraries

reply "Taylor Gautier" <tgautier yahoo.com> writes:
So the D specification seems to be an interesting forray in language design,
as seen by Walter and commented on by hundreds (thousands??).

I'm curious about the existence of a standard library of sorts for D?  These
days a language is defined as much by it's libraries -- libc; Java standard
classes - collections, string, awt, swing etc; Perl's CPAN -- as it is by
it's syntactical and semantic definitions.

So far D has the ability to snarf libc, but aside from that there doesn't
seem to be any information regarding a standard library (unless I missed
it).

I believe that such a library is every bit as crucial to the language's
success (or failure) as the language itself.

At the very least, I think it would be critically important to lay the
groundwork for how a standard library could grow and evolve and be
incorporated into production environments.  Java has done this relatively
well with it's libraries and deprecation mechanism (not to say there isn't
room for improvement)

I do see the deprecated keyword in the list of D keywords, so obviously
there is a start here.

Not that I am necessarily advocating the STL, but I do think D would need
something that provides similar functionality (same argument for iostreams
etc).

One thing to consider is how well the CPAN works for Perl.  Having the
ability to (almost) natively update from a common source is incredibly
convenient, and adds to the open-source, user contributed nature of perl.
It is basically as much a part of Perl philosophically as are the writings
of Larry Wall, the Camel book, and obfuscation contests (j/k ;-) ).

Comments?

-taylor
Aug 20 2001
parent "Kent Sandvik" <sandvik excitehome.net> writes:
"Taylor Gautier" <tgautier yahoo.com> wrote in message
news:9lrkbf$vi8$1 digitaldaemon.com...
 One thing to consider is how well the CPAN works for Perl.  Having the
 ability to (almost) natively update from a common source is incredibly
 convenient, and adds to the open-source, user contributed nature of perl.
 It is basically as much a part of Perl philosophically as are the writings
 of Larry Wall, the Camel book, and obfuscation contests (j/k ;-) ).

Yep, CPAN indeed makes a big difference, and one reason Python is catching up with this is that the language and the framework support system didn't have a standard and elegant way of updating libraries and announce the availability. I don't know if this falls into the language specification, but if the language had enough support for dynamic lookup of existing libraries and modules, and so forth, it would indeed help to put in place something that's even better than CPAN. As mentioned in other postings, the other 50% that makes a language interesting is the availability of standard modules. --Kent
Aug 20 2001