RE: [sca-assembly] FW: [sca-assembly-comment] Microsoft's responsefor Public Review of SCA Assembly Model v1.1 - 15 day review

["Estefan, Jeff A (3100)" <jeffrey.a.estefan@jpl.nasa.gov>]

Mike and Martin (and SCA-Assembly TC Members),

It is unacceptable in my opinion that this issue (Issue #132) and its cousin (Issue #149) simply be closed with a resolution of "No Action."  The points made by Michael Champion of Microsoft and by Philipp Knoradi of Siemens are very valid ones and we should not take an OASIS-only parochial perspective on these closely related issues.  Included below is the Don Ferguson response to Jim Marino on this topic from back in June 2009 who's words in his last sentence should not be taken lightly.

-----------------

From: Ferguson, Donald F [mailto:Donald.Ferguson@ca.com] 
Sent: Tuesday, June 23, 2009 10:40 AM
To: Jim Marino; OASIS Assembly; sca-assembly-comment@lists.oasis-open.org
Subject: RE: [sca-assembly] [ISSUE 132] Rebuttal: Against the use of portability and functions as reasons for requiring one of the existing 4 languages

I do not see any reason to believe that we cannot define a conformance test for SCA Assembly. If we are able to do so, we would not need to require support for the core languages. I assert that 1) We should not assume that we cannot define a compliance test until we have tried. 2) An assembly spec that places requirements on internal implementation languages is flawed.

Finally, I was present in IBM when we started work on SCA. Language independence and encapsulation of implementation was an explicit objective.  The IBM architecture leadership would have vetoed any assembly spec with this requirement.

Dr. Donald F. Ferguson
Distinguished Engineer
Corporate Senior Vice President
Chief Architect Products and Technology
donald.ferguson@ca.com
donff2@aol.com
www.donald-ferguson.net/blog

-----------------

If we want to insure SCA-Assembly portability to the maximum extent possible, is it not possible to create a Java EE-like Compatibility Test Suite (CTS) for SCA runtimes?  If such an effort is too daunting to accomplish in the next several months, then we can't simply punt on these very important points regarding implementation neutrality when it comes to the SCA-Assembly spec.  We must either remove the conformance clause that mandates use of one of the current SCA implementation languages, or consider some acceptable variant to that language, e.g., Sanjay's proposal.

Recognizing that it requires a 2/3rds supermajority to re-open a closed issue, I would still like to propose that Issues #132 and #149 be reopened with acceptable resolution to the concerned community other than "No Action." 

Regards...

 - Jeff, NASA/JPL