Re: [DotGNU]Definition of "webservice"

[Mario D. Santana]

> Rikard Linde <address@hidden> wrote:
> 
> > Hmm. That's interesting. I perceive web services as
> > any service that's offered on the web regardless of
> > technology. [...]

So anything that delivers HTML over a network is a web service?

I think of web services a little more narrowly -- when their usefulness
depends enough on technologies far enough outside the mainstream of web
standards. How much dependency, how far, is enough? The distinction is
arbitrary in the first place, a handy shortcut. So if the distinction
isn't useful, it's not a web service.

I think Norbert's points (function and use must be described, and
services delivered, via open protocols) point to the upper limits of how
far is enough. At the low end, web services start looking like html
files. At the high end, they start looking like application frameworks or
something.

A web service could be a particular feature (e.g., on a web site,) or its
enabling technology, or both together.

So the term is useful to refer to things that make sense if you look at 
them a certain way. These things often (always?) make sense in other 
ways, too. 

Just my 1/4 byte.

mds