RE: [emergency] Agenda: 7/29 Call

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Understood.  'Reference' implementation has a narrower meaning in ISO 
and other organizations I've worked with.  The term 'sample' implementation 
is preferred there.  The issue is normative binding of the implementation 
to the specification.  One should not have to read the code to determine 
how one must implement it.  One can review code if available to see a way 
and that gets into the value of open source.  On the other hand, binding 
the specification to any source could mean that only that source can 
be used.  Reference implementations (in the narrower sense) are sometimes 
applied to situations in which there is only one way to implement the 
spec and achieve its operational goals.  They are rarely necessary. 
Sample implementations are always useful.  

The IP problems are numerous but separable.

Thanks for the clarification.

len

-----Original Message-----
From: emtc@nc.rr.com [mailto:emtc@nc.rr.com]
Sent: Tuesday, July 29, 2003 10:10 AM
To: emergency@lists.oasis-open.org
Subject: RE: [emergency] Agenda: 7/29 Call


This is in response to Len's email
(http://www.oasis-open.org/archives/emergency/200307/msg00025.html)

We are saying the same thing. A "reference implementation" is nothing more
than an implementation of the spec that not only shows it works, but that
someone can look at to see a way to use it, but it does not mean nor imply
(unless we have a reason to do so) it is the only way. If the word
"reference" is what you are objecting to, then that is fine. It just what I
have used at W3C and OASIS before, but have no preference how we term it.

Here is the snippet from the OASIS TC Guidelines that detail what we must
do, which as you can see keeps "what" is built very loose:

"Certification by at least three OASIS member organizations that they are
successfully using the specification. (Despite numerous requests, the OASIS
TC Administrator feels it is not in the TC's best interests to further
define the meaning of "successfully using". The implementation could really
be anything from prototypes or proof of concept all the way up to
shrink-wrapped software. Defining this further would only restrict the
definition and make it harder for member organizations to say that they are
successfully using the specification.) This certification can be in the form
of a simple statement in email from a company representative, e.g. "I
certify that XYZ company is successfully using...." The implementers must
also certify that their implementations comply with known IP encumbrances
(see IPR below)."

Allen

You may leave a Technical Committee at any time by visiting
http://www.oasis-open.org/apps/org/workgroup/emergency/members/leave_workgro
up.php