Re: boundaries for the Web

If the W3C believes that the Web architecture can be documented and is 
ready to be documented now, the result should be complete and accurate. 
Constraining the Web with REST or markup or people-as-agents will make 
everyone's job easier in the short term, but it limits the W3C and the 
Web in the long term.

The Consortium is quite resource-constrained right now, and this could 
be one mechanism of addressing that; IMHO, however, it is an inferior 
one. I'd like to see issues addressed on their own merits, on a 
case-by-case basis, rather than having a constraining architecture 
making the decisions ahead of time.

Put another way; is there a good reason for putting up boundaries, other 
than as a means to allocate resources?



On Thursday, March 21, 2002, at 04:36  PM, Simon St.Laurent wrote:

> On Thu, 2002-03-21 at 18:13, Dan Connolly wrote:
>> I wonder what makes you say that... from what I can tell, this sort
>> of description of the Web has been used pretty consistently,
>> since before W3C was even formed:
>
> I think the W3C's own continuous growth into new areas is a good example
> of the lack of boundaries in its understanding of the Web.
>
>>> well beyond the common understanding of the
>>> Web among developers.
>>
>> Could you elaborate on that? What do developers think about the
>> Web that conflicts with the stuff in section 1?
>
> I'd suggest that most developers still think of the Web as a mechanism
> for exchanging information with a human being at the client end.
>
>> If they think the Web is just HTTP and HTML, then I think
>> that's an unfortunately limited view of the Web, and I hope
>> to persuade them to broaden their view a bit.
>
> I don't know that the W3C has ever attempted to reach (and certainly not
> include) most developers, so persuasion might be a good start.
>
>>>  The definition above applies to the Internet as
>>> well as the Web and to any number of other networked systems.
>>
>> The bit you excerpted does, but you clipped perhaps
>> the most relevant part:
>>
>>   Web Architecture is the set of rules that all agents in
>>   the system follow that result in the large-scale effect
>>   of a shared information space.
>
> Unfortunately that doesn't provide any constraints whatsoever on the
> definition of "Web".  Saying that the "Web" is whatever conforms to "Web
> architecture" is more or less an invitation to "Web architects" to do
> what they like.
>
> I suspect it's clear that I'm unhappy with a situation where "Web" means
> "whatever the W3C feels like doing".
>
>>> I would like to request that the TAG establish in this document a
>>> definition of "the Web" that includes clear boundaries for the Web -
>>> what is the Web, and what is not the Web.
>>
>> Any suggestions?
>>
>> The main things that I can think of that are not Web Architecture
>> are technologies that are local to one system; i.e. cut-and-paste
>> desktop standards, CGI and other inside-the-server stuff. Is that
>> the sort of boundary you had in mind?
>
> Heck, I see CGI as far more Web-oriented than Web Services or the
> Semantic Web.  I think we'd be better off looking at questions about
> what exactly the Web does to figure out what the Web is.
>
>> Another fairly clear boundary is that if URIs don't figure
>> in somehow, it's pretty disconnected from Web Archtecture. e.g.
>> TCP, french minitel, AOL, etc.
>
> That's hardly a boundary - it lets anything using URIs in, which can be
> pretty nearly _anything_.
>
> So what is the Web?
>
> Conservatively, I'd suggest that:
> The Web is a hypertext-based system built around the Hypertext Transfer
> Protocol which uses a combination of marked-up information and software
> to convey information to people.
>
> I'm quite aware that definition is constraining, but I'd be happy to
> hear others which provide constraints.
>
> --
> Simon St.Laurent
> Ring around the content, a pocket full of brackets
> Errors, errors, all fall down!
> http://simonstl.com
>
>
--
Mark Nottingham
http://www.mnot.net/

Received on Thursday, 21 March 2002 21:29:29 UTC