[jsr-314-open] [ADMIN] JSF 2.0 Spec Snapshot

[Simon Lessard]

Hi,
 
Please disregard my comment about DataModel and thread safety since they're not, by definition, thread sensible as one implementation can easily use a ThreadLocal to store the current rowIndex. However, I think my comment would be interesting for the concrete classes provided by the spec since those are not thread safe so a thread safety notice could be useful there.
 
 
Regards,
 
~ Simon

________________________________

From: JSR 314 Open Mailing list on behalf of Ed Burns
Sent: Wed 3/11/2009 10:51 AM
To: JSR-314-OPEN at JCP.ORG
Subject: Re: [jsr-314-open] [ADMIN] JSF 2.0 Spec Snapshot



>>>>> On Thu, 05 Mar 2009 11:00:20 +0000, Pete Muir <pmuir at REDHAT.COM> said:

PM> Ed,
PM> Here are some questions:

PM> * is delta state saving in this revision? The state saving stuff is
PM> still in an appendix, and I find it hard to check.

It's not in the revision.  The stuff in the appendix is now quite stale
due to the lifecycle changes introduced to accomodate Andy Schwartz and
Dan Allen's concerns over the performance of view metada gathering.
Ryan is currently reworking the state saving branch, including Martin's
recent patches, but it's tricky.  The state management revision may have
to be slipped to 2.1.

PM> * Issue 450 (https://javaserverfaces-spec-public.dev.java.net/issues/show_bug.cgi?id=450
PM> ) still hasn't been merged into the spec - I've asked about this one
PM> multiple times. Can we get this merged in?

THANKS for your persistence.  This email will include an action plan for
Roger and I to follow, and this will be on it.

PM> * the big changes due to view parameters don't appear to be merged
PM> in?

See action plan below.

PM> Otherwise, the areas I have been involved in look good:

PM> * exception handling
PM> * bean validation

Thanks for the oversight.

>>>>> On Thu, 05 Mar 2009 16:32:18 +0100, Martin Marinschek <mmarinschek at APACHE.ORG> said:

MM> Hi Ed,
MM> my feedback below. I already (wrongly) sent these using my Gmail-address,
MM> but I guess (hope) these mails didn't make it...

I've put these into the action plan.

MM> also, should I send out the patch-proposal that I made and the
MM> follow up discussions I had with Leonardo to the EG? Pete and Andy
MM> seemed to be interested...

AS> Is this regarding delta state saving?  I am definitely curious to hear
AS> where we ended up with this.

As mentioned above, late breaking view parameter changes have impacted
our ability to include the state saving changes into 2.0.

>>>>> On Thu, 05 Mar 2009 17:26:44 +0000, Pete Muir <pmuir at REDHAT.COM> said:

PM> Another couple more after talking to Dan:

In action plan

PM> RegexValidator

ACTION PLAN for Roger and Ed.

These need to be done to the spec.

rogerk * 10.3.1.1: in table 10.3, the f:ajax attribute is called event -
  in several paragraphs around this, events - which one is true?

  It is now "event".  Roger should be taking care of this soon.

edburns * Merge spec issue [SelectItemsCollection-450] into spec.

edburns * Merge all view parameter changes into spec.

edburns * 3.1.8: should the tree-visitor be mentioned here?

edburns * 4.1.18.4: shouldn't this reference AfterAddToViewEvent?

edburns * 5.4.1: seems to have old spec language (language on error
  handler says that only predestroy does not call through to it)

edburns * 6.1.14: managed-bean-scope can now be view on top of none,
  request, session, application - can it also be flash? I don't see this
  mentioned anywhere.

edburns * the spec language for the BeanValidator is light compared to
  other validators

  (see 3.5.5 - for example RegexValidator's description vs. BeanValidator

Thanks for your feedback!

Ed

--
| ed.burns at sun.com  | office: 408 884 9519 OR x31640
| homepage:         | http://ridingthecrest.com/