On Mon, 13 Aug 2001, Abigail <[EMAIL PROTECTED]> wrote,
> On Mon, Aug 13, 2001 at 09:37:28PM +0700, Hasanuddin Tamir wrote:
> > On Mon, 13 Aug 2001, Abigail <[EMAIL PROTECTED]> wrote,
> > >
> > > The point isn't that Java has Sun behind it (but not just Sun. Also
> > > IBM, Microsoft (even if they'd rather not) and more).
> > >
> > > The main difference between Java and Perl is what is being produced.
> > > The "great" Perl things that are being produced are modules. But modules
> > > aren't interesting for the majority of the enterprise world. Modules
> > > are just tools to make other tools. The enterprise world wants tools:
> > > that is applications.
> >
> > In that case, isn't Java just another tool? It's not application
> > either. I might get you wrong, but it seems to me that you compare
> > Java to Perl modules.
>
> Which part of "The main difference between Java and Perl is what is
> being produced." was unclear to you?
Ok, let me put it this way. It seems to me that you compare Java
applications to Perl modules. They look at the former since the
latter don't offer immediate solutions. I didn't judge your
statement, just tried to get your point.
> > I assume you you refer "enterprise" to people in the position making the
> > decision. But how do they get to the decision?
>
> They look at their needs, they look what's available, they look at the
> price. And yes, stupid decisions are being made. But many smart decisions
> are made as well.
Exactly. That's what our customers do here (and I think I can add
that support counts as part of the price). Only a few (if any) of
them, AFAIK, ever cared to ask what program language we use. And
there was only one potential customer that asked why we use Perl.
The fact is that most of them come back, still without question
about the language.
It's ok, as long as they happy with what we do for them, and we still
have fun doing it in Perl. Of course this doesn't imply whatsoever
that we've developed some killer applications.
> > > There are no killer Perl *applications*. Just killer modules. Not that
> > > there is something wrong with that.
> >
> > Once a friend asked me, "Why there are so many cool modules in CPAN but
> > there's no one real application written in Perl?" With a hard-to-hide
> > gasp I said, "Maybe because most of the great applications are developed
> > for internal usage." But I knew I wasn't so sure.
>
> Perhaps. And that's the problem. Buying a canned solution is often much
> cheaper than rolling your own.
So, can I say that, from the case above, they've bought canned solutions
from us?
> > Say there was a real killer Perl application, would they (the enterprises)
> > turn? And if the application really satistified them I don't really think
> > the would care what it's written in.
>
> "A" real killer application? No. A single killer application wouldn't turn
> the tables. A constant stream of good [1] applications, that's what is needed.
Let's make many of them, then. And make Perl as killer tool that produces
killer applications. But how we do this? I'm all ear to learn.
> [1] For various values of good. Sad as it is, "Point and drool but sometimes
> crashes your computer" is often better than a complicated interface
> that is always flawless.
Um, do you mean by "complicated interface" is command line interface?
For most computer users, yes, it's complicated.
> Now, I do *not* think Perl should go the Java way. I don't really care
> Java is "big" and Perl is "small". I don't think Perl has anything to
> gain by battling Java for the same market.
>
> Perl has its own role to play, and it plays it damned well. It isn't
> as big as Java, but it shines real well in its own theater.
>
> Let Perl be Perl.
*silently agrees*
san
--
Trabas - http://www.trabas.com
