Re: VOTE: JNDI patch (was RE: I sent in a patch that fixed the EJB issues.)

[Berin Loritsch <bl...@apache.org>]

Glen Daniels wrote:
> 
> Hi Berin!

Hi Glen!

I forgot completely about the Axis IRC chat today!  I would have been there
otherwise.

> > -----Original Message-----
> > From: Berin Loritsch [mailto:bloritsch@apache.org]
> > Sent: Tuesday, October 02, 2001 12:05 PM
> > To: Axis Development
> > Subject: I sent in a patch that fixed the EJB issues.
> >
> >
> > There are actually a couple of ways that patch can be
> > applied.  The getInstance() method can be used as is
> > (in otherwords, always bound to a JNDI context) or, we
> > can return the AxisClient instance directly if no JNDI
> > context is found.
> 
> I think I like this a little better as a first step.  This means we'll need
> JNDI jars to build axis, but won't need to bulk up with another jar for a
> minimal JNDI in non-EJB cases.

This is true.  I implemented it the way I did mainly because I wanted to test
that it can be done that way.

> We need to be able to pass a ConfigurationSupplier for each engine into the
> system somehow - perhaps by mirroring log4j's initialization pattern (look
> for a *config.xml file in the classpath if we don't find a configurator
> system property).  This should get merged in here somewhere.  I'll attempt
> to do this.

The FileProvider does essentially do this.  If the file does not exist in
the expected location, it pulls it from the jar.

> Other than that, I think this looks good.  Can we get a vote on committing
> it?
> 
> +1 from me.
> 
> > Please give me some feedback.  Theorhetically, it should
> > wipe out four of the five bugs in BugZilla!
> 
> A nit: I think it was premature to mark all the Bugzilla bugs as "FIXED".
> IMHO, the team should decide that a solution is appropriate and functional,
> and get it integrated with the source tree, *before* the status says FIXED.
> 
> --Glen

I thought that was what CONFIRMED was for.  Sorry.  Which status should I
use instead?