Re: svn commit: r550656 - /geronimo/server/trunk/modules/geronimo-transaction/src/main/java/org/apache/geronimo/transaction/manager/TransactionImpl.java

[Donald Woods]

Updated to log the stack trace as an error in revision 551929.


Donald Woods wrote:

Agree.  I'll update the log statement to include the stack trace.

Thanks for reviewing and commenting on the change.

-Donald

Kevan Miller wrote:

On Jun 28, 2007, at 12:11 AM, David Jencks wrote:

On Jun 26, 2007, at 7:08 PM, Donald Woods wrote:

Was just going on Kevan's response to YunFeng, that we shouldn't be using printStackTrace() in the code - http://www.nabble.com/Why-printStackTrace%28%29-in-the-source-codes-tf3975719s134.html

Heh. Sure... Blame it all on me... ;-)

I don't have a problem with logging the stack trace rather than printing it to the console, but even printStackTrace IMO is not a really big deal since the exception will only occur when someone has written a broken integration of something that needs xa. For instance I think openejb mdbs are currently broken this way.

--kevan

Right. I have no objections to logging stack traces, where we think they would be useful. My main point was that we should be thinking in terms of "logging" information, not "printing". The geronimo log should contain the information needed to identify/diagnose a problem (not a random mixture of logging and direct printing to STDOUT/STDERR).

smime.p7s
Description: S/MIME Cryptographic Signature