Re: [Ietf-message-headers] HTTP header registration question

On Fri, 28 Sep 2007 12:52:45 +0200, Martin Duerst <duerst@it.aoyama.ac.jp>  
wrote:
> I have some comments about the document at
> http://dev.w3.org/2006/waf/access-control/Overview.html.
>
> First, the title is not easy to understand for people who
> aren't totally immersed in the Ajax business. Most Web
> Resources are, almost by definition, read accessible.
>
> I'd suggest a title like "Enabling Cross-Domain Read
> Access for Web Resources".

I think this would cause conflicts for people who want to use it in other  
contexts, but I personally agree with this. Maybe raise this on  
public-appformats@w3.org so other people can give input?

> I don't understand the sentence:
>>>>>
> XML resources may include an <?access-control?> processing instruction  
> within the XML Prolog to indicate in cases where the access control read  
> policy applies from which domains they can be fetched. [XML]
>>>>>
>
> First, the [XML] reference at the end looks totally out of context.
> I know that some style guides say that references should be at the
> end of a sentence, but this one is really unreadable. More important,
> "in cases where" doesn't make sense.

That should be replaced with "if" probably. And some punctation around it.  
I just made that change.

> The only way for me to understand this sentence is that the PI is
> interpreted on the server side, before the document is (or actually
> is not) sent, but experience with in-document information for server
> side has been detrimental (nobody wants to take that performance hit),
> so I highly advise against this. If the PI is supposed to apply
> client-side, the the above sentence needs heavy rewriting.
> Checking whether a document can be downloaded once it's already
> downloaded is of course nonsense.

Good point. Replaced "they can be fetched" with "their content can be  
accessed".

> Another point that I don't understand is that the TO-ASCII algorithm
> of RFC 3490 is mentioned several times, but the syntax only allows
> RFC 3986 URIs or RFC 1034 labels. TO-ASCII only makes sense if you
> start with an Unicode string, but RFC 3986 and RFC 1034 are pure
> ASCII already. I think it would be good to allow IDNs, but you actually
> have to say so.

I couldn't find definitions of "subdomain" and "label" elsewhere. And it  
seems that RFC 1034 doesn't actually restrict IDN usage, but I could be  
mistaken. If you have suggestions on how to improve this they are most  
welcome (on any list).

-- 
Anne van Kesteren
<http://annevankesteren.nl/>
<http://www.opera.com/>

_______________________________________________
Ietf-message-headers mailing list
Ietf-message-headers@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf-message-headers