Loading ...
Sorry, an error occurred while loading the content.

10775Re: [soapbuilders] [ANN]JAX-WS 2.1 FCS released

Expand Messages
  • Doug Kohlert
    Feb 9, 2007
      Steve,
      I am still poking around on the interop tests. I did discover that when
      doing WS-Policy interop testing about a year ago, the turned out to be a
      full-blown security interop test. Participants there were MS, IBM, BEA,
      SAP, Sun, Apache, Layer 7. The results can be view at:
      http://lists.w3.org/Archives/Public/public-ws-policy/2006Jun/att-0010/ws-policy-workshop-minutes-April-2006.doc

      Everyone was able to fully interop with everyone else except for one
      company that did not have a security implementation at that time. In
      the chart, Sun is P6.

      Steve Loughran wrote:
      >
      > On 2/8/07, Doug Kohlert <doug.kohlert@...
      > <mailto:doug.kohlert%40sun.com>> wrote:
      > > Oh, you are talking about the data corruption mentioned in Kohsuke's
      > > blog
      > >
      > (http://weblogs.java.net/blog/kohsuke/archive/2007/02/jaxws_ri_21_ben.html
      > <http://weblogs.java.net/blog/kohsuke/archive/2007/02/jaxws_ri_21_ben.html>).
      > > Yea Dims, we made that up; get real dude. Kohsuke even mentioned in
      > > his blog that we were not sure why it happened and that it might even be
      > > a Glassfish issue.
      >
      > Well I for one celebrate the release, and think the dev team should be
      > congratulated and given the rest of the week off to celebrate, even
      > though I personally have no intention of using JAX-WS or anything that
      > goes from Java to WSDL.
      >
      > o It is the SOAP client built in to Java 6, so any improvements in
      > performance are good, because they will help on the client.
      >
      > o Any improvements in interop can only be a good thing.
      >
      > o The async stuff looks interesting, though I fear that the Java
      > community is reading "Java Concurrency in Practise" and going
      > overboard with futures and callables the way Java1.1 and Swing did
      > with nested classes.
      >
      > As for the benchmarking, well, I don't know what you can conclude.
      > Probably comparing the new version against the old one is the best
      > metric, and here things are better. And the fact that the team has
      > managed to find a concurrency problem on a 4-way box with glassfish
      > means everyone is partway to fixing it -dims and dennis are now aware
      > of its existence.
      >
      > Doug :
      >
      > 1. any plans to slip this out into a Java 6 point release? Or is that
      > going to be stuck with the old codebase for its entire product
      > lifespan?
      >
      > 2. what's your current interop test process?
      >
      > -Steve
      >
      >
    • Show all 30 messages in this topic