[jdev] Vapor (C# XMPP Server)

Jonathan Dickinson jonathanD at k2.com
Fri Apr 4 06:21:37 CDT 2008


Yes, I do build a DOM. I should be able to change the architecture to build objects instead of DOM objects later on: but I don't see that as high priority: everything is very loosely coupled so changes are quick and fluid.

Note that a 'protocol user' never interacts with the XML directly. The following applies:

Streaming xml -> Xml nodes -> Deserialized into objects -> Given to 'protocol user'

Optimally I want to do the following:

Streaming xml -> Deserialized into objects -> Given to 'protocol user'

The former is already in the framework, but I am at work (and behind a proxy) so I can't check the code in just yet.

I have to stress that agsXMPP won't facilitate this, everything MUST be done via Xml serialization attributes (XmlRoot, XmlElement etc.). This is to ensure maximum development speed for webservice developers as well as conformance to the current .Net webservices framework (which I will be proxying).

One thing I hope to achieve is to send the XSD over the wire to the jabber client and make the required proxy classes, just as Visual Studio does with the HTTP webservices binding. If I use the Xml Serialization attributes the facilities for this are built directly into the .Net BCL, so I can rely on the trustworthy and time-proven Microsoft implementations: which are deployed across thousands of machines.

jonathand at k2 dot com

-----Original Message-----
From: jdev-bounces at jabber.org [mailto:jdev-bounces at jabber.org] On Behalf Of Alexander Gnauck
Sent: 04 April 2008 01:08 PM
To: jdev at jabber.org
Subject: Re: [jdev] Vapor (C# XMPP Server)

Jonathan Dickinson schrieb:
> Is this okay, or will I have to roll my own XML writer (is the comment
> invalid)? (Note that the default xmlns is missing, but this is just an
> example). Should I face any problems with any less xml conformant clients?

thats fine, but also depends how you forward/route incoming stanzas.
Will you build a Dom from them?

Alex




More information about the JDev mailing list