On Jul 12, 2006, at 12:14 PM, Henrik Østerlund Gram wrote:
On Jul 12, 2006, at 10:43 AM, Henrik Østerlund Gram wrote:
Took me forever to find this too by the way; I think it's a
terribly bad idea to not propagate exceptions back up the callstack.

In this case, I agree. Exceptions that occur synchronously should be
propagated. There's also a good argument for the unified place to
receive all exceptions though :)

As in being able to do something like System.setExceptionTarget("/ dev/null") in any application just to avoid handling exceptions or declaring them thrown where appropriately? :) Maybe there is a valid argument for it in some cases, although I honestly cannot think of one right now.

In a server application, the "caller" would be the MINA core that received the request, but its not a good place to handle the exception, so it gets sent to IoHandler.exceptionCaught(). Also for things happen asynchronously, its a way to have a common place to handle exceptions from such processes w/o having to duplicate the handling code each time.

But for cases where code is executing synchronously from user-level code, I 100% agree with you.

java.lang.NoSuchMethodError:
org.apache.mina.common.TransportType.getEnvelopeType()Ljava/lang/
Class;

       at org.apache.mina.common.support.AbstractIoFilterChain
$1.filterWrite(AbstractIoFilterChain.java:122)

       at
org.apache.mina.common.support.AbstractIoFilterChain.callPreviousFil te
rWrite(AbstractIoFilterChain.java:583)

Do you have an older MINA version on your classpath?

Nope.  This is my first face-off with MINA and she's killing me.

Where did you get your MINA jar? And can you post your sample program so I can reproduce? (Wondering if there's a bad jar on a distribution site somewhere)

-pete


--
[EMAIL PROTECTED] - http://fotap.org/~osi



Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to