On 12/20/06, Steve Jones <[EMAIL PROTECTED]> wrote:
> REST is a name put on a software implementation style that sits on
> HTTP.
No, it is not; it is a name put on something that already exsisted,
kinda like the name AJAX
for something I had done for years already.
> > Of course you did. What you're talking about is the full REST thing,
> > with PUT and DELETE and the semantics where the URL is a resource;
> > that's the part you haven't done. And it's true that people haven't
> > done this, and that this isn't the part of the success of the web. But
> > the first part *certainly* is.
>
> So what you are saying is that people have done REST but without
> resources... you do see the issue with that statement surely?
Please read what I write; "the semantics where the URL is a resource"
which is far fetched from what you're saying. What a lot of people
have done is "domain/something/getItem?id=xxx" in which the semantics
of the *URL* (and not parameters) means nothing in terms of resources,
while "domain/something/xxx" does.
> I used HTTP, I used parts of the HTTP spec (though far from all) and
> used that to ship HTML to browsers. I know its comforting to claim
> that this is REST but in reality it isn't, it is building a thin
> client server->human publishing system. This is vastly different to a
> server->server system.
Why? You're talking nonsense; of couse a computer can talk to another
computer using HTTP, and in fact we do this by the bucketloads every
day. I think you need to stop thinking of HTTP in terms of what a
browser does with it.
> > A question might be; how much of the WS-* stack needs to be successful
> > before WS-* as a stack is deemed successful? Surely there are
> > principles and foundations in place that can be highly successful even
> > if you don't use the full stack?
>
> Agreed, but I wouldn't claim I was using WS-Security when I wasn't or
> that I'd used WS-* when all I had done is put the word "SOAP" in a
> messages or used XML-RPC.
Not sure I follow; since the comparison is on the fundamental levels,
as in REST vs. WS-*, if you're using some of the WS-* parts you're
using the WS-* stack, just like if I'm using parts of REST I'm being
RESTful. What irks you so in this part of the debate?
> > All browsers are REST clients, some more than others, but REST
> > nevertheless. Not sure why you're saying otherwise.
>
> 1) Did Microsoft, Netscape, NCSA, Mozilla and Opera read and obey the
> REST architectural guidelines before implementing their browsers
No, it didn't exsist as *that* paper under *that* name. Does that then
equate to there being no such thing as REST before someone named it
such?
> 2) Is the goal of browser based systems computer->human or computer->computer
Depends on the browser. Some of them have all sorts of blog-readers,
newsgroup stuff, plugins for semantic web stuff, automatic updates and
so forth, but notice the name "browser" which hasn't been valid for
the last 8 years or so; the focus of that app *is* to be a human
interface to the web. Why do you think the potential of REST fails
just because a *browser* is RESTful? A browser is something people use
to take benefit from the web, and I can tell you that I do lots of
things with my browser which falls way outside of the simplistic idea
of computer->human thing you're trying to push. And I don't even know
why the hell you're trying to retrofit 30 years of technology into
black and white like this; it makes no sense.
Maybe you're trying to do the whole "well, REST might be ok for
browsers, but *real* software can't use that sort of toy". Please
explain.
> That is why. To me it looks like a desperate attempt to get a
> successful reference for REST by using systems that were never built
> based on the REST paper.
...
> I built Web Sites well before I knew of the REST paper, they went live
> and they worked and I can 100% say I didn't think of it in the terms
> outlined in the REST paper.
You've got some mixed up idea that AJAX, REST, SQL, IDE, HTML, MPEG
and bucketloads of other technologies didn't exsist before someone sat
down and gave it a name and some guidelines.
Alex
--
"Ultimately, all things are known because you want to believe you know."
- Frank Herbert
__ http://shelter.nu/ __________________________________________________
