hc-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Becke <be...@u.washington.edu>
Subject Re: [PATCH] Correct handling of close and additional test cases
Date Tue, 03 Dec 2002 15:31:00 GMT

On Tuesday, December 3, 2002, at 10:13 AM, Eric Johnson wrote:

> I applied my changes to the latest in CVS, and ran all the test cases. 
> The patch is attached.

To apply this patch the committers will need it to be in the unidiff 

> Since I'm running with a local tomcat, I did not see the four failures 
> referenced in earlier emails.  I ran the latest from within our 
> applications as well, and it performed exactly as expected. 
> Unfortunately, I am seeing a new failure on the test-local target:

This happens for me as well.  It seems that the connection is made 
before the timeout can occur.  Looks like the response timeout system 
is not sensitive at the 1 millisecond level.

> As near as I can tell, the new HttpMethodBase.wrapResponseStream() 
> function, and its corresponding inner class _could_ be made obsolete 
> if the three other stream wrappers would take on the responsibility.  
> On the other hand, the current solution is clean, and isolates 
> portions of the code appropriately.  Anyone else have any take on > this?

I'll have to take a look at this.  I think it is still needed to 
release the connection, but I'll take a look once I can apply the patch.


View raw message