[jsr-314-open] Pre-JCP filed draft of JSF 2.2 JSR

[Pete Muir]

Hi Ed,

Red Hat would like to support this JSR. In general we are happy with the scope, content and process associated with the JSR, however we would like to see these issues addressed as well as those discussed in the JSR proposal:

• Specification missing list of retargetable handlers
	• JAVASERVERFACES_SPEC_PUBLIC-922
	• Easy to do with large impact
	• Mojarra, and MyFaces currently do this differently
• Deprecate or remove legacy JSF 1.x APIs that are now made obsolete with JSF 2.0+
	• Example : ConfigurableNavHandler replaced NavHandler, and etc.
• De-couple the Facelet template engine from the Web Request/Response lifecycle.
	• Clearly explain diff

Additionally we see these issues planned in JIRA but not in the JSR proposal, and would like confirmation that they are scheduled for 2.2 (otherwise they move to the above list ;-).

• PartialViewContext interoperability
	• JAVASERVERFACES_SPEC_PUBLIC-658
• Add new client event type - between complete and success
	• JAVASERVERFACES_SPEC_PUBLIC-678 
• Cancelable Ajax requests
	• JAVASERVERFACES_SPEC_PUBLIC-621
• Response containing the JSF spec version
	• Likely good to have impl info as well
	• JAVASERVERFACES_SPEC_PUBLIC-149
• Request timeout support
	• JAVASERVERFACES_SPEC_PUBLIC-682


Additionally, from a CDI integration perspective (with our CDI spec lead hat on):

• Deprecate, remove, or make optional @javax.faces.managedbean
	• This is confusing to new developers as it overlaps with CDI, especially when in the Java EE platform.
	• Consider making this an "optional" part of the spec which isn't available when running in the Java EE platform
• All developer-facing components must be CDI-aware and available via @Inject 
	• Phase Listeners, Components, Nav Handlers, Applications, View Handlers, Converters, Validators, and so on...
• All JSF lifecycle and component events must be propagated to the CDI event bus, without the need to register a JSF-coupled listener.
	• Less critical from an end-developer point-of-view; however
	• Still critical from the extension-developers perspective

For this JSR Red Hat's primary representation will come from Jay Balunas and Lincoln Baxter, however as this is an open list, I'm sure others will chip in :-)

I'll continue to help with any CDI integration issues.

Best,

Pete

On 16 Feb 2011, at 15:32, Ed Burns wrote:

>>>>>> On Mon, 14 Feb 2011 11:27:30 +0000, Pete Muir <pmuir at redhat.com> said:
> 
> PM> Ed, could you discuss how this, and plans for future revisions of
> PM> JSF, fit into the Java EE 7 roadmap? For example, are you intending
> PM> to do another spec revision before Java EE 7 is due? That would be
> PM> most helpful when evaluating the proposal.
> 
> Certainly.  We are indeed intending to do another revision before EE7 is
> due.  The intent of scoping this one for calendar year 2011 is to
> demonstrate a philosophy of smaller, but more frequent, releases of the
> JSF spec.
> 
> Ed
> 
> -- 
> | edward.burns at oracle.com | office: +1 407 458 0017
> | homepage:               | http://ridingthecrest.com/