----- Original Message -----
From: "Geir Magnusson Jr." <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Saturday, June 09, 2001 21:58
Subject: Re: JavaOne Struts BOF
> "Diethelm Guallar, Gonzalo" wrote:
>
> > page. And, finally, he presented something that could be taken as a
> > disadvantage of the templating model: he said JSP is a specification,
> > not an implementation, and mentioned that with JSP you have the chance
> > to change from one implementation to another if you find the
> > particular implementation you are using does not meet your performance
> > needs (he even gave Jasper as an example of a less than optimal
> > implementation, in terms of performance). With templating languages
> > (and he specifically mentioned Velocity as an example), if you find
> > your performance needs are not satisfied, you are pretty much screwed,
> > since there is no other alternative. After this question, I had to
> > leave the BOF, since I had other things to do.
>
> I was there too (standing up against the left-side wall...) I was going
> to counter with the following, but figured it was his show :)
>
> Jon's sensible arguments regarding Velocity as having a defined spec
> notwithstanding, I will grant the spec-vs-impl issue to Craig for
> argument sake. The problem I had with his spec vs implementation
> argument is that it doesn't actually define *what* is specified.
>
> By this I mean that there are no provisions for basic looping and logic
> in the JSP spec, and once you toss out scriptlets due to danger ( <%
> while(true) { Integer arr[] = new Integer[1000000]; } %> ), the
> spec-compliant JSP is pretty useless. You can access beans, do
> compile-time includes, and generate static content.
>
> For logic and control, then have to resort to tag libraries, and there
> are no official specs for that yet - so to make a useful page, you have
> to use non-spec, implementation-only tag libraries, putting JSP squarely
> in the box that he put Velocity into...
>
> There is ongoing JSR work for a standard tag library that will provide
> useful stuff like 'if', and the ever-popular, indispensable 'otherwise'
> tag, but the current status of that work seems to require the use of
> scriptlets again
>
> <jsptl:if condition="<% [condition goes here] %>">
>
> (or something like that)
>
> as well as add tags for things like JDBC, XML-RPC and other I/O, etc, so
> it will certainly be interesting to see what comes out.
One interesting thing in the JSP spec is, as in many other Java related
specs, that it's owned by Sun. Although it has published it for free, there
may be some legal issues related to implementations and especially public
distribution of such implementations. If some implementation didn't
correspond to Sun's interpretation of the spec, the distributor might get
juridical troubles, which can cause vicarious nuisance to clients using that
implementation too. The best known case is of cource MS Java++, which was
also some kind of violation of an agreement, but milder cases can arise from
"non-Sun" implementations of such specs.
I'm not against specs and strongly appreciate Sun's work on keeping Java
clean, but just want to bring this up as one additional viewpoint that we've
faced when selecting open technologies to apply in commercial products. With
new and rapidly evolving technologies, it's an advantage in some conditions
to have the spec and implementation in the same package by the same
developers.
-- Ilkka
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
