Ok, I send a question on the list.
2007/8/21, Rainer Jung <[EMAIL PROTECTED]>:
> Henri Gomez wrote:
> > The define told us that we shouldn't use flush on AS400 V5R2 and previous.
> >
> > On V5R4 (UTF8 mode), we could use it.
> >
> > This code (no flush) is pretty old and was provided by i5/OS IBM
Henri Gomez wrote:
The define told us that we shouldn't use flush on AS400 V5R2 and previous.
On V5R4 (UTF8 mode), we could use it.
This code (no flush) is pretty old and was provided by i5/OS IBM Labs,
so better stick with it. On V5R3 and previous release, we didn't see
the send_bodyct was pro
The define told us that we shouldn't use flush on AS400 V5R2 and previous.
On V5R4 (UTF8 mode), we could use it.
This code (no flush) is pretty old and was provided by i5/OS IBM Labs,
so better stick with it. On V5R3 and previous release, we didn't see
the send_bodyct was probably set another way
Hi Henri,
so now i5/OS should use send_bodyct like the other platforms. What I
don't understand is:
[EMAIL PROTECTED] wrote:
@@ -375,7 +369,7 @@
static void JK_METHOD ws_flush(jk_ws_service_t *s)
{
-#ifndef AS400
+#if ! (defined(AS400) && !defined(AS400_UTF8))
if (s && s->ws_private
Author: hgomez
Date: Mon Aug 20 06:16:13 2007
New Revision: 567689
URL: http://svn.apache.org/viewvc?rev=567689&view=rev
Log:
Fix for send_bodyct on i5/OS. The problem was in jk (not flushing) and not on
IBM implementation (my bad)
Modified:
tomcat/connectors/trunk/jk/native/apache-2.0/mod_j
