Actually no, I cannot agree with both arguments. First of all JSP is a very old and not even very good rendering tech, secondly, JSF delivered on promise and kept jsp backward compatibility even to the degree that this decision made the component model unnessesarily complicated, by introducing a compatible taglib layer.
Thirdly facelets is not a bugfix, it is simply another view layer, which fixes jsp (in my humble opinion) or replaces jsp with something way less broken. If you want to keep backward compatibility, to jsp, there is shale clay, there is jsf templates on dev.java.net both technologies are similar to facelets but keep backwards compatibility with all its disadavantages (limited caching possibilies, compilation against servlets etc..) So saying that facelets is a bugfix, then so is freemarker and velocity which cover similar domains with similar results which make them better than jsp. The success of facelets just basically shows that its approach was dead right and that it is easy to use (Which we also have to give kudos to Tapestry where a lot of things were derived from). I only agree to the point that JSF needs something like facelets out of the box. This is one of the major points open for discussion in JSF 2.0. I am rather sure facelets wont make it as default technology because they have to keep JSP backward compatibility but something along these lines will come. Just my 2c regarding this issue. The beauty of jsf is, in my opinion that they kept a rather small kernel in the first instances with open hook points which made it possible that those things exist, and now that things have been proven to work, they can be integrated into the core. I think the biggest mistake they could have probably made would be to push too many things into it in the first place, we would have ended up with another EJB1. Iordanov, Borislav (GIC) schrieb: > I think the existence of facelets, the motivation behind it, show JSF's > failure to deliver on its promise (after so many years!). I haven't > looked into Facelets, not that I'm afraid to learn some new "view > technology". I just don't want to impose this "bug fix" to people that > have already invested time in learning and using JSP pages, with custom > tag libraries developed etc (is Facelets backwards compatible with JSP? > probably not). > > -----Original Message----- > From: news [mailto:[EMAIL PROTECTED] On Behalf Of Werner Punz > Sent: Friday, April 06, 2007 4:36 AM > To: [email protected] > Subject: Re: New to MyFaces > > Iordanov, Borislav (GIC) schrieb: >> I'm not sure what "statistics" you are looking for. I haven't done an >> industry analysis. But in general, JSF is heavyweight machinery > without >> any substantial benefit. Simple things are complicated and complicated >> things impossible. It was obviously designed by (probably smart, Java >> knowledgeable) people that have no serious experience with web >> development. A well-known example is that it still doesn't work well >> with JSP (a technology for which JSF was designed from the start!) and >> it probably never will. >> >> > > JSF 1.2 does (myfaces soon will have jsf 1.2 level) > and facelets basically do what jsp does. You basically > speak about the mixin problems of html and jsf (verbatim tags) > > this problem is gone in the jsf 1.2 spec, and in facelets, facelets also > eliminates problems introduced by jsp... > >
