Re: definition of strict conformance

>>> is Lofton, earlier; >> is me, earlier; > is Lofton, last time

>So for example, in SVG Tiny, an implementation can choose whether to
>support embedded ECMAscript (within <script> tags), or not. And this
>disqualifies Tiny as a "strict conformance" definition?
>That seems fine to me.

Yes, that's what I meant. Glad you agree.

>>>Can there be "implementation dependent" features or behaviors?

>>If they fall under the guidelines for discretionary behaviors, which
>>would be broader than discretionary *choices*, then yes.

>This one I have more trouble with.  Perhaps because I consider
>"implementation dependent" to be the Ultimate Evil for interoperability.

A good example is choice of encodings. "Strict conformance" means that
the implementation must support, say, UTF-8 and UTF-16 but nothing else.
If they have the discretion to also support other encodings, then a
test suite can at least use UTF-8 with confidence. If the choice of
encodings is fully "implementation-dependent", then you can't count on
any particular encoding, so how do you write a test suite?

My hope is that we can combine discretionary CHOICES with discretionary
behavior (a.k.a. "implementation-dependent") for SpecGL at least, to
avoid having two separate Dimensions of Variability that are very
similar. Maybe the Glossary will need some adjustment.
.................David Marston

Received on Thursday, 1 August 2002 16:21:36 UTC