Review of Oct 1 webarch

Abstract: s/an/a/

Lose comma after "resources"

s/a provide/provide/

lose "and reason about" I have no idea what that means

"Web architecture encompasses both the definition of the information 
space in terms of identification and representation of its contents, and 
the protocols that define..."

"desired properties: efficiency, scalability..."

s/information media/media/

"This document's structure reflects the three dimensions..."

Status of this doc

"Where current practice conflicts with it..."

s/resolve their differences/address these issues/

might want to note that at this point there are no patent disclosures.



"is an information space in which objects, referred to as resources.."

"to illustrate typical behavior"

"software acting in the context of this information space.  Agents 

Can we come to consensus on the formatting or URIs, the quotes and 
monotype font look like dogshit in conjunction with each other

"to access weather information"

"Web agents exchange information via messages.  Protocols define..." the 
phrase about "these messages arise" is kind of contorted and doesn't help.

s/origin server/server/

3.Messages carry representations of resources.


XHTML representation data, which includes references to weather maps 
identified by URIs; the browser retreives and displays those maps.


"gives guidance"



s/includes some references/includes references/

"This document strives for brevity and precision while including 
illustrative examples"

s/some theoretical/theoretical/



"across communities motivates"

s/this represents a link/this constitutes a link/

Replace sentence beginning "A link defines..." with "Links play many 
roles and support many software behaviors."  For example, hypertext 
browsers automatically follow lins to images and render them in context, 
but follow links to other textual resourcs only on request, and replace 
the current display when doing so.

Do reasoning systems really focsu activity on namespace references?

s/named,shared, or linked to within/a part of/

Principle: Use URIs s/is intended/might reasonably be expected/

"An information resource is one that conveys information.." NO!  I think 
an information resource is one that is essentially "an online unit of 
information or service", i.e. not a car.


2.1  Good Praactice.

This is hard to get right.  What we're really saying is "don't do random 
escaping or re-coding, just preserve the characters you were given." 
Maybe "strive to avoid variation" or some such?

s/two URIs that are not equivalent/two URIs that are not identical/

is "authority agent" a well understood term?



s/; this is called URI opacity//

re the Editor's note: I am totally OK with using "Web agent" here.

s/that explains how to assign identifiers/that explains how identifiers 
are assigned/



do news: URIs only identify groups not messages?

I'd put the assertion beginning "Furthermore, designers should expect" 
in its own paragraph and lose the leading "Furthermore".  It's awfully 



3rd bullet point: "specifies the syntax and semantics of fragment 

Move the para beginning "The secondary resource" down one para, after 
the one beginning "Fragment identifier semantics.."  Then rewrite it to 
say "Existing URI schemes specifiy a variety of roles for fragment 
identifiers, including some portion of..."



s/encode the same visual map/provide the same visual map encoded/

s/their fragment identifier semantics (which are undefined)/define no 
fragment identifier semantics/



s/specification of/specification governing/



list item 1. s/, which//

We need to note that we're not suggesting that all these specs are 
consulted at run-time, rather each phase of the agent's actions is 
consistent with specs normative in that context.



s/persistence is always a matter/persistence is a matter/

Editor's note: I'm OK with the Moby Dick para.


2.7 Access control

Should this move into the "Interaction" section?

"Nadia and Dirk"... ooh, sounds like hot stuff



s/to state - or at least claim - formally/to assert/



what is this section talking about?  I have no idea?



In the <ol>, the second half of the 2nd <li> I think is important enough 
to be pulled out of the list and given its own paragraph near the top of 
this section.

s/This will include/This includes/



s/Thousands of independent/Large numbers of independent/



The "don't guess metadata" principle needs an example



s/have flourished/gorwn in number/



s/designers SHOULD/designers MUST/

Might be a good idea to say that HTTP code 404 is a good example of this.



s/are, in general, harmful/are in general harmful,/



s/ Textual formats are often/Textual formats are usually/

s/tools any more complex/tools more complex/



disagree with definition of "extensibility"



I don't agree with the last para.  I don't think the #1 problem with 
namespaces is DTDs.  First, the DTD problem isn't that bad, and 
secondly, the problems with poor understanding of interaction modes is 
much worse.



Principle on Qnames needs a little introductory text to set the stage.



Replace entire first para with "If all that is known about a 
representation of the resource is that it is 
encoded in XML, what is the correct interpretation of


in <ol>, in item 3, s/and\/or/and/



Sentence beginning "since XML documents", rewrite to: "Since XML's 
syntax is deigned to make documents self-describin..."

Received on Sunday, 5 October 2003 10:45:45 UTC