Re: Applications Messaging Standard

Reading though the list, there seems to be a fair bit of disagreement about the scope of this standard - why not take the approach that has been taken in other areas and first define a "simple" version of the protocol, which is essentially has exactly the same scope as PLASTIC as it currently stands - while being careful to avoid non-PC acronyms ;-).. Then have a discussion about some all-conquering application messaging standard as a separate thread.

There are several simplifying aspects to the current PLASTIC scope and design that have made it a success;

• it works for a single user being the only user on a single desktop
  • do not need to worry about security - only respond to messages that originate from localhost.
  • do not have to worry about complicated topologies for delivery.
• it does define the message format and the transport (ignore the java RMI aspect - everyone agrees that should be dropped)
  • XMLRPC is widely supported even in simple scripting languages (no socket programming necessary)
• it is designed for interactive messages
  • guaranteed delivery etc. not so important - it is obvious to the user when something is not working - it should just happen now....

I think that a decision to make a slightly cleaned up PLASTIC *the* Simple application messaging standard would be a great benefit to the VO.

Paul Harrison
ESO Garching
www.eso.org Received on 2007-02-13Z09:05:11