On 03/22/2010 11:15 AM, Mark Thomas wrote:
On 22/03/2010 17:05, Filip Hanik - Dev Lists wrote:
here is what I posted to the JSR list
This is in relation to
http://markmail.org/message/gafapyibbowo4jdb
There are two problems that we encounter with the test code below
1. This servlet does support Async, so it should call complete(), not
rely on the container to do so
You and I disagree on this one. We'll see what the EG says. The TCK
expects the container to call complete() and my reading of the spec was
also that the container should call complete().
Doesn't this create a chicken or the egg problem?
In this specific case,
ThreadA invokes /Test
ThreadA calls dispatch
ThreadA dispatches to ThreadB (to dispatch to /Test)
When should the container call complete(), when ThreadA exists, when
ThreadB exists, or does the container have to keep tabs and call it when
the last one exists?
Filip
2. Since it calls dispatch() before it calls request.setAttribute, it
ends up in a loop
If the dispatch returns immediately like the spec says it should, the
attribute will be set (hopefully) before the thread started by the call
to dispatch reads the attribute. I don't like the TCK dpending on the
timing but I think that it what it is trying to do.
Mark
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
