Title: RE: [ebxml-msg] Have you implemented Version 2.0 ? David, I supported the testing process because it was the right thing to do, but I cannot agree with you that we can or should attached any idea of conformance/compliance to the process. The process lacks discipline that would be required for such a validation. I agree that you attempted to ensure conformance to the specification. The issue is not the testing steps. It is the execution. True conformance testing requires that the tests be completed by an independent third party--that was not done. We did not ship our product to your lab for testing, we did it behind the closed doors of our institution, just as all the participants did. What did we do during the course of the testing to our product? Who really knows? CEAL and other accredited testing facilities use a more sterile process that ensures that the certification is for the product they have in hand, not something that is promised or attested to by the one who built it. I insist there is a difference. Your argument about cost might seem to be compelling, but i t is a dangerous thing to ascribe test validity to cost. But, at the same time, I will say that comparative compliance testing is available at a similar cost. We are paying less than the DGI cost for FIPS 140 compliance testing through one of the three certified labs in North America. They use rigorous processes within their NIST certified spaces to ensure that our products do exactly what we say they do, and exactly what the standards require. I state again, DGI had the right idea. I believe that the testing accomplished positive things. Most important, it provided some traction for ebXML when there was a perception that momentum was being lost--it brought ebXML transport directly into the view of the community. Then, the testing demonstrated that different implementation of the same specification were interoperable. We have lamented the specification's immature state, but the truth is vendors were able to use it to build real product. I submit that the various vendors used bright people to build their products, but the opportunity for misinterpretation and error was huge. I do not recall any serious discrepancies during the dry run phase. This would seem to say the the specification can be used by ebXML adopters. Last, the testing revealed a few discrepancies in the specification. There were some issues that were reflected back to the OASIS TC for resolution. The testing helped build more quality into the V 2.0 specification. With that on the table, I still firmly hold to the idea that we need a different model for compliance testing and I am looking to the IIC TC for direction in that area. Ralph Berwanger Ambassador to Standards bTrade,Inc