Ray Cromwell wrote:
> Justin Wells wrote:
> > Quoting Craig R. McClanahan ([EMAIL PROTECTED]):
> >
> > Find out what makes it slow and fix it? That's what's great about OSS
> > projects. Or maybe you won't want to do that, but someone else will do
> > it instead.
>
> > > I suppose
> > > I can go write my own implementation, but of course I can do that with
> > > the JSP syntax as well.
> >
> > However, that's a huge amount of work. It's better to take an existing
> > code base and improve it, then start all over from scratch.
>
> No it isn't, if the existing codebase sucks. Sometimes it's better to
> scrap the legacy and start over with a new design. Did you ever look
> at the codebase of Netscape 4? JMF1->JMF2? AWT -> Java2D -> Kestrel ->
> (Newest pluggable graphics primitive architecture, unannounced)
>
As it turns out, exactly this kind of discussion is going on in Jakarta as well --
see the "proposals/catalina" directory in the CVS source tree for a proposal
describing a completely different internal architecture for the Tomcat servlet
container.
<Disclaimer>
It's my proposal, so I obviously feel it's a good one :-).
</Disclaimer>
>
> That's actually one of the problems of OSS sometimes, is that people
> keep hacking away on the same codebase, because it's easy to patch,
> and hard to *design* GNUJSP could certainly be patched so much that
> it rivals Tomcat/Jakarta, but it would end up a mess.
>
Agreed. Sometimes it takes people willing to formet a "revolution" in order to
break out of that cycle.
>
> The first implementation of anything is usually a throw away design
> used to gain enough information and learn the lessons needed to write
> the final implementation.
>
> > > * If I need to hire a developer and I key "WebMacro" and "JSP" keywords
> > > into a resume web site, how many hits am I going to get on each?
> >
> > If you hire a developer who can't figure out how to use WebMacro in a day,
> > then either you hired the wrong person, or I'm doing something wrong. I
>
> Well, he might be able to learn it in a day, but would he neccessarily
> use it in the best way without experience? Would he use #parse
> directives effectively, or cut-and-paste? Would he rely on $Request
> too much, or delegate to the servlet?
>
> There's a big difference between "book learnin'" and actual real
> experience.
>
It's also not just the initial learning -- I need to train every new developer I
add for the lifetime of the application.
There is nothing wrong IMHO with WebMacro (I have stronger feelings about some of
the half-assed implementations of template systems) -- but I have to wear my
"project manager" hat as well, and say what's best for the overall product. In
many cases, that means mainstream technology choices, quite independent of
technical "elegance" or "beauty."
This seems to offend some open source folks. Sorry -- but that's the real world of
software development I've lived in for the last 25 years.
Craig McClanahan
===========================================================================
To unsubscribe: mailto [EMAIL PROTECTED] with body: "signoff JSP-INTEREST".
FAQs on JSP can be found at:
http://java.sun.com/products/jsp/faq.html
http://www.esperanto.org.nz/jsp/jspfaq.html
