Re: Case Study: XMLC vs. Velocity

On Monday 15 September 2003 02:58, Matthew Hixson wrote:
> I got a laugh out of this "case study."
>
> http://jakarta.apache.org/velocity/casestudy2.html
>
> I've never used Velocity, but after working with JSP for the past 3
> years I've come to the conclusion that it is the _wrong_ way to do web
> development.  The biggest thing that bothers me is the lack of compile
> time sanity checking.  JSP promised a separation of code and HTML.
> You've gotta be kidding me if you think it accomplished its goal.  If
> anything JSP leads to nothing but a big inconsistent, poorly
> maintained, highly fragile, unmaintainable nightmare.  Ugh.  I'm trying
> to get away from JSP/custom tags/Struts and move towards a servlet+XMLC
> world.  I think life will be much nicer there.

I couldn't agree with you more.  And I suspect of course, you'll find the same 
from most list members.   We're using XMLC on one current project.  We used 
it from the beginning, because I hate JSPs and like the XMLC concepts (plus 
I'd used to before).  Both of the two other developers were willing to try 
it.   

We've built our own "widget" framework on top of XMLC, and most of the time I 
find it very easy to slap out a re-usable UI snippet, or to just make a new 
page.  However, I apparently am not the norm.  

13 months into the project, and now the team is 10 developers.  Nobody but me 
likes it.  They gripe about it all the damn time.  For many programmers, it's 
just too abstract.  With JSPs, they can "see" what's going to happen, but 
they seem to have trouble with visualizing the DOM manipulation that's 
occurring.   Of course, these same programmers that "see" what's happening in 
JSPs are (in my oppinion) way to prone to copying code, too (which JSP 
forces/encourages).
-- 
David Corbin <dcorbin@xxxxxxxxxxxxxx>