axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Shellman, Joel" <>
Subject RE: when is trunk going to be available for new features?
Date Mon, 05 May 2003 15:57:55 GMT
Can I please request very strongly that there should be a "no-config" mode of use. I tried
out GLUE a little while ago, and it did what I expected. I could make a service available
with no config information beyond passing in the instance and endpoint.

In other words, I don't want to have to specify bean mappings (they're all beans, why wouldn't
they be?), and I don't want to have to put in special tags in my code. I don't think anyone
should have to.



-----Original Message-----
From: Glen Daniels 
Sent: Monday, May 05, 2003 8:18 AM
To: ''
Subject: RE: when is trunk going to be available for new features?

Hi Thomas!

> I would like to merge the new features and bug fixes I have
> committed to 
> the dynamic_deserilaization_branch into the trunk within the 
> next couple of 
> weeks due to our internal release schedule. For your 
> recollection these 
> features were voted to be branched off not to disturb the 1.1 
> release about 
> 2 months ago. Is there any chance we could branch off the 1.1 
> final release 
> code within that time frame to open up the cvs trunk/head 
> again, or is the 
> release going to happen sooner?

I am hoping to get 1.1 out the door this week if possible, pending a couple more bug fixes
and getting the TCK results to Sun (in process).

I haven't reviewed your diffs on the branch yet but I'd like to make sure that your stuff
and the current autoTyping, both of which exist for pretty much the same reason, are merged.
 In other words I'd rather not have two separate ways that Axis auto-serializes java types
with no explicit typemappings.  Your stuff seems to rely on putting metadata in the java classes
themselves (very cool, and dovetails nicely with future xdoclet work), whereas the autoTyping
stuff assumes no mapping at all is there.  Does your code (I haven't read through it yet)
allow you to specify the XML type Qname in the metadata, or does it make one up?  If the former,
you need to have some way of registering the class on the deserializing side, right?  If the
latter, we need to make sure we're consistent about the names we use.


View raw message