🚀 go-pugleaf

In article <7fgg6c$3it@cantine.wu-wien.ac.at>,
 <mottl@miss.wu-wien.ac.at> wrote:
>I am not sure whether we can compare these two developments with OCAML. I
>would consider "OCAML" and most (all?) other functional languages as
>quite "fragile": their "market share" is extremely small. If commercial
>firms were allowed to just rip out all good stuff to brush up their
>inferior products (in this case: Java compilers) this could well lead
>to a problem. It could draw away the still low attention of the public.

   Frankly, I don't consider forking to be a major danger unless the
maintainers are unfriendly to patches (I believe this was the problem
that caused the fork in GCC) or the maintainers have difficulty
interacting with people (witness OpenBSD).  Even if the maintainers were
somewhat averse to patches, I don't think you'd get a full-blown fork
until the language had a wide-enough base to support it.  (What you
might get is one or two people who maintain a port to a machine, or
incorporate some nonstandard features, and try to keep their release in
sync with the official distribution).

>
>If there is hardly any competition, this does not apply. E.g. I'd say
>there is/was no real competition to Emacs - the split didn't hurt,
>better: it helped the project. In the case of GNU C / EGCS: both are
>just direct substitutes for commercial products and there are a lot of
>better commercial C/C++-compilers. C/C++ (unfortunately?) wouldn't die
>out if EGCS / GNU C ceased to exist. But if people used Java instead of
>OCAML because of technology-transfer (OCAML->Java), I'd regard this as a
>loss not only to the FP-community but to software development in general.
>
   I think it's interesting to note that in the case of text editors,
Emacs (and vi for vi-lovers) has been around a very long time (I count
it as being around since it's original creation in the mid-70's) and has
been ported to FSF knows how many machines and used by most programmers,
though it's certainly not "hyped".  It's availabilty and superiority
(for a certain class of users) have given it dominance.  Likewise, vi
has been adopted by masses of programmers.  What you have are
essentially two juggernauts, each of which represents its users think a
text editor should be (the barebones vi or the full-bodied Emacs).  I
submit that this is very analogous to the situation with languages -
some people are going to prefer imperative, and no amount of arguing is
going to make them change their minds, and likewise with functional.  If
you make a superior free implementation of your preferred paradigm
(OCAML in this case) the users you can hope to get will come to you.
The people who use imperative languages are not very likely to jump over
to functional languages just for the technology, so I see no reason to
prevent their enjoying the benefits of the compiler technology.

Lynn