Submission of SML to W3C - verification of CAM approach!

This was made yesterday to the W3C:

  http://www.w3.org/Submission/sml/

 

On brief initial review of this - seems like the team has followed very
much the CAM approach here!

 

Thoughts / differenciators:

 

1) SML creates a lot of extension syntax whereas CAM uses XPath
directly

2) CAM assertions are business-user friendly syntax - whereas SML uses
arcane logic syntax

    e.g. in CAM 

 

       setLength(urn:IPAddress,
16)

 

    in SML 

                 <sch:ns prefix="tns" uri="urn:IPAddress" />
                   <sch:pattern id="Length">
                      <sch:rule context=".">
                          <sch:assert test="count(tns:address) = 16">
                          </sch:assert>
                      </sch:rule>                      

                   </sch:pattern>
                </sch:schema>            

3) Intriguingly - CAM processor could easily incorporate a SML step as
part of post-processing

 

4) Does not seem like SML can do anything that CAM cannot

 

5) CAM can handle multiple structures and structure variances better I
believe - and also context driven parsing.

 

Anyway - good to have the W3C validating our use case!

 

DW

"The way to be is to do" - Confucius (551-472 B.C.)