W3C home > Mailing lists > Public > www-qa@w3.org > September 2003

Re: [qaframe-spec] What is an implementation?

From: Alex Rousskov <rousskov@measurement-factory.com>
Date: Fri, 5 Sep 2003 08:56:38 -0600 (MDT)
To: Bjoern Hoehrmann <derhoermi@gmx.net>
Cc: www-qa@w3.org
Message-ID: <Pine.BSF.4.53.0309050847390.47207@measurement-factory.com>


On Fri, 5 Sep 2003, Bjoern Hoehrmann wrote:

> * Alex Rousskov wrote:
> >> You'd need a term for the process where a product attempts to do or
> >> does what the specification requires (currently "implements")
> >> though.
> >
> >or "based on"?
>
> I thought about that, but it is probably a bad choice if you want to
> express that anyone is, well, implementing a technology in a software
> product for example.
>
> >> I do not necessarily think of e.g. a HTTP request as a class of
> >> product but it is more a product than an implementation.
> >
> >HTTP request is not a conformance subject :-). HTTP agent creating,
> >forwarding, or receiving an HTTP request is.
>
> Why? I understand the concept of a HTTP request as a "class of product"
> and a product as an instance of a "class of product".

Good point. It may be a matter of opinion because HTTP specification
is not very precise. RFC 2616 conformance policy defines conformance
for HTTP "implementations". In RFC 2616 context, I interpret
"implementation"  as "agent" (client, server, or intermediary).
However, one could interpret "implementation" in a broad sense and
include HTTP messages.

Fortunately, the practical difference is negligible because a
standard-violating HTTP message virtually always implies a
standard-violating agent.  The reverse is not necessarily true because
some common agent behavior is not expressed via a single HTTP message
(e.g. caching).

Alex.
Received on Friday, 5 September 2003 10:56:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 12:14:00 GMT