[DotGNU]Re: Defining Web Services by example

[Norbert Bollow]

David Nicol <address@hidden> wrote:
> http://www.tipjar.com/nettoys.html
> Towards defining "web service" better, I invite all to consider the
> eight tipjar nettoys and decide if each is or is not deserving of
> the appelation.
> 
> 1: form tester -- use this as the action to debug your form.
> 
> 2: form handler -- form submission to e-mail.  Vulnerable.
> 
> 3: xl to html converter -- a cgi wrapper for Steve Grubb's xlHTML
> 
> 4: gnuplot interface -- a cgi wrapper for gnuplot
> 
> 5: redirector -- issues a "Location" header to first URL in form data
> 
> 6: playing cards -- 52 cards and four jokers in a directory
> 
> 7: mime decoder -- re-present a raw MIME attachment as a web page
> 
> 8: authenticated identity service -- general purpose SSO


> Which ones are "webservices?"  all of them?

According to the definition which I have recently proposed,
they'd all be "webservices".

> Not #6 because it is static?

This is the one where I have the greatest hesitations to call it
a webservice.  But then, maybe it actually makes sense to use
this as a webservice.  Suppose someone creates a computer game
that uses such playing card graphics.  That program could
maintain a local mirror of David's set of 56 cards (so that if
there are legal problem with one of the cards, and David
replaces the card for this reason, through David's webservice
the replaced card is made available automatically to the users
of the game which uses these cards).

Also, "Hello World", which has been implemented in just about
every programming language, has very static output, and it's
still considered to be a computer program.

So I think it's proper to call even this a webservice.

> Not #7 because performing the function it performs is
> trivially simpler than invoking it?

Simplicity is in the eye of the beholder - certainly there are
users who feel quite capable of invoking this webservice, even
though they may not be aware of any other way to decode a
base-64 encoded attachment :-).

> > But it still requires human intelligence to really determine
> > what the component does!  Even if the component advertises a
> > well-known standard service, you may want to check with someone
> > who reads Bugtraq before you believe it.  And it the component
> > does something that is at least a little innovative, it is
> > ridiculous to expect other software to be able to really
> > understand even what the component claims to be doing.
> 
> some kind of descriptive schema for cataloguing the available
> services is an interesting idea.

Like UDDI ?

> with standards like that, it would be possible to rig up
> a webservice shell that would job out every element in a
> command chain, for instance,
> 
>       du -ax /usr | sort -nr | head -n 500
> 
> to a separate remote webservice.  Although "sort" is the
> only one that would make any sense at all.

I agree that we definately should think about how to carry the
power of the unix shell to the world of webservices.  See my 
XML-shell proposal

http://archive.dotgnu.org/pipermail/proposals/2001-September/000031.html
http://archive.dotgnu.org/pipermail/proposals/2001-September/000032.html

Greetings, Norbert.

-- 
A member of FreeDevelopers and the DotGNU Steering Committee: dotgnu.org
Norbert Bollow, Weidlistr.18, CH-8624 Gruet   (near Zurich, Switzerland)
Tel +41 1 972 20 59       Fax +41 1 972 20 69      http://thinkcoach.com
Your own domain with all your Mailman lists: $15/month  http://cisto.com