One keeps hearing that "Struts is dead." (At least, in French
http://www.zdnet.fr/builder/commentaires/0,39021559,2134284-3,00.htm
I'm late following up on Tom's post, but hopefully better late....
Whether or not one agrees with all of the French author's views, the article that is cited, which actually begins at http://www.zdnet.fr/builder/commentaires/0,39021559,2134284-1,00.htm , really reinforces for me the extraordinary importance of Java Server Faces, which will arrive in the mainstream in development tools in the next couple of months.
I can precis some of the article. It starts by going back in history to 1996-1997. Servlets were introduced around this time, but their acceptance was limited because they lacked model-view separation.
Microsoft introduced ASP to wide acceptance. Sun came back with JSP, which really did take off in the Java community, despite architectural problems and awkwardnesses mixing code snippets with HTML. Developers, by adding MVC Controller approaches on top of JSP's, rescued them to a large extent and they achieved wide use.
But Microsoft, intent on recovering its lead and beating back JSP's, then introduced .Net, with web design based on Web Forms and totally centered around graphic composition. To date, Java does not have that and is vulnerable to being replaced by .Net in many situations, for that reason.
Microsoft, however, may have made a strategic error by omitting MVC concepts. Microsoft thought they were too complex.
JSF is the new Java answer that brings both graphic composition and an MVC model to web design. So we now have the possibility for the superiority of Java web design, after all, over .Net and WebForms.
The author says there were 4 evolutionary stages:
Servlets ->JSP -> MVC -> JSF.
(That is the end of my very-partial precis. The following comments are my own).
So that is some of what is at stake here. It is not small.
Now, whether a JSF/Struts combo, or JSF alone, ends up as the centerpiece is not completely clear yet. That is the issue Tom was talking about.
IBM's tooling will, it appears, combine JSF and Struts. (They need combining because they are otherwise not 100% compatible out of the box!)
Well, there are more major issues too.
(1) It is being left to the vendors to define most of the graphic web components that will come with JSF. That means IBM's will be different from Sun's, which will be different from BEA's etc. Is this a major error?
(2) Open source developers were largely left out of the JSF JSR 127 (see http://java.sun.com/j2ee/javaserverfaces/ ). They will left to a very late start compared to the members of the JSR, and will not get source code,as far as I know (or at least have not got it yet). Is this a major error?
Well I hope I have provoked some of you into some posts by now.
By the way, IBM's JSF Team Lead will be presenting, including a demo of tooling, at the Websphere User Group on the 30th (see http://www.rtpwug.org/meetings.html ) and it is free. I hope we can also hear at a later date of other JSF (and alternative) implementations, in either the JUG or the WUG...
Regards,
Don
-
_______________________________________________ Juglist mailing list [EMAIL PROTECTED] http://trijug.org/mailman/listinfo/juglist_trijug.org
