On Fri, 05 Jan 2001, Dave Thomas wrote:
> ts <decoux / moulon.inra.fr> writes:
> 
> > >>>>> "D" == Dave Thomas <Dave / PragmaticProgrammer.com> writes:
> > 
> > D> Performance is important. But increasingly, time to develop, ease of
> > D> maintenance, and other more global factors trump performance.
> > 

> Mais oui! As I said, performance is important, and we should continue
> to ty to improve it. I'm just saying that there are other factors
> which are more important when it comes to a user choosing Ruby as a
> language.
> 

Actually, I believe a majority of the languages (and methodologies) considered
'de riguer" or 'mainstream' in the e-business world today were introduced
'subversively' by programmers who felt it was a better way to achieve the goal,
_all_ factors considered.

businessValue = results! An application in place generating income or solving a
problem is of far greater _current_ 'business value' than one in development,
regardless of the differences in processing "efficiency'.

This is not to say that processing efficiency should not be a goal towards
which to continually strive, only that it is not the _only_ goal, nor in the
overall context of providing concurrent 'business value' the most significant
one.

The "best" answer is both situational and time sensitive. If a client needs a
web based form to email system _today_, I would probably install FormMail with
Perl; if s/he granted a little more time and wanted something hihgly
customized, I would, at this point, probably go with Ruby. Process intensive
applications, especially with lots of looping and/or calculations involved are
most likely still the province of C, given there is time alotted for the
write/compile/run cycles involved. And if the goal is a fast track, totally
cross platform, GUI fra,ework, well, er, <sigh> :-)

All in all, my brief exposure to Ruby (_and_ to this list!) has been for me
very intriguing. Is Ruby perfect? No...but then again, what is? However, the
elegance of solutions offered here, together with the constant regactoring,
leads me to look at Ruby as one of, if not _the_ one, best current choice
towards proving "simple solutions to complex problems". 

Warm regards,

Kent Starr
elderburn / mindspring.com