axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Doug Davis" <...@us.ibm.com>
Subject Re: cvs commit: xml-axis/java/test/functional TestEchoSample.java TestStockSample.java TestTransportSample.java
Date Sun, 02 Sep 2001 17:38:58 GMT
Removing it would be ok as long as it is replaced with something that
doesn't use sockets/tcpip as the transport.  I helps keep us honest about
not being transport specific.
-Dug


Sam Ruby/Raleigh/IBM@IBMUS on 09/02/2001 10:58:11 AM

Please respond to axis-dev@xml.apache.org

To:   axis-dev@xml.apache.org
cc:
Subject:  Re: cvs commit: xml-axis/java/test/functional TestEchoSample.java
      TestStockSample.java TestTransportSample.java



Glen Daniels wrote:
>
> OK, that code indeed notices the timeout for doTestIBM.  The
transportsample
> test was still hanging though, I assume because the thread waiting on the
> URL data was still wedged.  While in there I noticed that the req file
> wasn't getting deleted in this case because invoke() never came back in
> FileReader, so I switched to getting the Message as bytes to read in the
> whole file and then deleting it before calling invoke().  Also fixed the
> FileSender to increment its index even if a timeout occurs.
>
> The latter fix enabled the second FileReader to find the xml2.req file
and
> complete the XXX test successfully.
>
> The synchronization in this sample still seems inadequate and potentially
> trouble-causing.  It does work now, so I'll put the test back.  And yes,
it
> was in fact much easier to remove it.

I've taken a look at what it would take to make the requests and responses
"pretty" again.  Unfortunately, testArgAsDOM fails.

It seems that "isDirty" is not being completely maintained.  If something
is marked dirty, everything above it should be marked dirty.  And after a
clean parse, the entire envelope should be marked clean.

It occurs to me that it would be much easier to simply remove the test...

- - -

If there isn't interest in maintaining the File transport sample, then I'm
willing to go along with a consensus resolution to remove it.  As long as
it remains, I'm opposed to removing tests simply because they expose
problems in that sample.

- Sam Ruby




Mime
View raw message