[Ocaml-biz] IDEs

[Brandon J. Van Every]

Brian Hurt wrote:
>
> Emacs has more *developers* writting emacs lisp scripts for
> it.  But this
> is because vi doesn't have a built-in equivelent scripting
> language.

Yeah, so leave it to the sysadmins.  Show me fantastic OCaml support
under Vim if you want me to think otherwise.

> And
> before you start talking about emacs using it's "superior"
> user base to
> supplant vi, remember that it hasn't succeeded to any
> signifigant degree in 25 years.

This is a sideshow.  We haven't had the Emacs vs. Eclipse discussion
yet.

> But you didn't answer my question: right this second, what's
> wrong with
> either emacs or vi (any versions thereof) that needs to be
> fixed?

I will get back to you at a later time why I don't think Vim's support
of OCaml is as good as the Emacs support.  If someone more knowledgeable
can beat me to commenting upon it, that would be great.  A true OCaml
Vim advocate hasn't spoken up yet, so I feel this is a theoretical
discussion at present.  The sentiment of "gee we should support
everything" vs. "gee we should focus on best of breed."  If I can't
provoke an OCaml Vim or XEmacs advocate into the open with my unkind
words, that says buckets right there.

I've spoken pretty clearly about what's wrong with XEmacs.  It's
permanently in GNU's shadow.  *.el's will always work well on GNU Emacs
and be fragile on XEmacs.  You yourself agreed that the XEmacs user base
is, like, 10%.  This is not fixable.  I see nothing about XEmacs to
indicate a superior OCaml IDE compared to GNU Emacs.  Ergo, XEmacs
loses.  We want volume for OCaml, not maintenance headaches.

> What editor you use shouldn't even come up when discussing
> programming languages.

As far as businesses are concerned, it should and does.

> Nothing turns me off to a language faster than
> learning you
> need to use the special language environment for that
> language.  They're
> text files.  If you want to use notepad, that should work.

We are discussing IDEs, and your view regarding notepad is unrealistic.
It is not an IDE.  It is a text editor, and a poor one at that.

The question amounts to which IDEs 'have legs' for mission critical
commercial development in OCaml.  What I'm definitely sure of at this
point, is Visual Studio doesn't, and Cameleon doesn't.

> My point in this isn't that the choice of emacs over vi is
> wrong, it's that *choosing* between emacs and vi is wrong.

We are going to continue to differ on what businesses need and want for
mission critical commercial development, until Oct. 1st.  At that time,
we'll have either reached a consensus about what the best tools and
practices are, or we'll have gone our separate ways, relying on "open
source diffusion" to advance the cause of OCaml.  We do not have the
resources on ocaml-biz to promote and support everything and anything,
the development culture you seem to prefer.  We're going to have to pick
winners and losers.  It might be a circle of 2 winners, or at a stretch
3, but we must choose.

As I said, 1 month to establish that political will exists for the plan:
> > 1) Determine market
> > 2) Determine useful, desired OCaml toolbox for this market
> > 3) Find existing best fit tools

If the will doesn't exist, oh well.  Life goes on, if not on OCaml's
terms.


Cheers,                     www.indiegamedesign.com
Brandon Van Every           Seattle, WA

When no one else sells courage, supply and demand take hold.