W3C home > Mailing lists > Public > www-tag@w3.org > November 2010

Re: FW: "Is 303 Really Necessary?"

From: Jonathan Rees <jar@creativecommons.org>
Date: Fri, 5 Nov 2010 10:36:51 -0400
Message-ID: <AANLkTik5X8U+kkXQiCdE0RmK=okEsKX-9pROyFKV3P_f@mail.gmail.com>
To: nathan@webr3.org
Cc: Ian Davis <lists@iandavis.com>, Yves Lafon <ylafon@w3.org>, www-tag@w3.org
So what part of this syllogism do people dispute:

If GET U yields a 200 HTTP response, then the entity (sensu RFC 2616)
is a "representation" (sensu AWWW) of the resource "identified" (sensu
RFC 3986) by U.

Every GET of http://iand.posterous.com/is-303-really-necessary yields
a response with an entity whose dc:creator is Ian Davis.

Therefore every representation of
<http://iand.posterous.com/is-303-really-necessary> has dc:creator Ian

If every representation of a resource R has dc:creator Z, then R has
dc:creator Z.

Therefore, <http://iand.posterous.com/is-303-really-necessary> has
dc:creator Ian Davis.


On Fri, Nov 5, 2010 at 8:34 AM, Nathan <nathan@webr3.org> wrote:
> Jonathan Rees wrote:
>> On Fri, Nov 5, 2010 at 6:05 AM, Ian Davis <lists@iandavis.com> wrote:
>>> On Thu, Nov 4, 2010 at 7:27 PM, Jonathan Rees <jar@creativecommons.org>
>>> wrote:
>>>> This debate has been raging continuously since 2004 or maybe earlier,
>>>> so my first reaction is "not again".
>>> Well, me too. But I'm now of the opinion that 5 yrs of implementation
>>> experience of httpRange-14 is saying it's an unnecessary overhead and
>>> an impediment to linked data adoption by the mainstream. Talis is
>>> heavily invested in making linked data successful and has a great deal
>>> of implementation experience in infrastructure, publishing and
>>> consumption which is informing the arguments in my post.
>>>> If someone who is following the threads could post a summary here of
>>>> the arguments pro and con, or anything they've learned, when things
>>>> settle down a bit, I would be grateful.
>>> Actually my original blog post attempts to do that, listing out the
>>> current disadvantages of relying on 303 redirects and the principle
>>> advantage of doing it.
>>> http://iand.posterous.com/is-303-really-necessary
>>> Ian
>> Thanks, but this is not what I asked. Your post only presents one side
>> of the story and I was hoping to hear "pro and con". On www-tag we
>> have 6 years of impassioned defense of the 200-means-web-page story
>> and hash URIs - did no one come to their defense in the public-lod
>> thread? If not, how did the thread get to be so long?
> Many have said words to the effect of "you use slash URIs so do what you
> want, but we use fragments because 200-means-web-page", defended fragids and
> said "don't go telling people that /slash uris with 200 is a good idea", but
> I think general sentiment from all those who would and do fight for the use
> of fragments, is that it would be wasted breathe, they're not going to
> change URIs (cool URIs don't change) and never wanted the 303 in the first
> place, 303 does make life less-nice, so they want shot of it.
> It's a one way path here, it's clear the message from some is "we're
> sticking with non-frag URIs for things" the only room for change is to 200
> rather than 303, and this will happen.
> Here's the first demo (200 OK):
>  http://iandavis.com/2010/303/toucan
> argument being "does this break the web" - answer is obviously no, so excuse
> granted and on they go. Ignoring the effects on their data within the web
> scale graph.
> Best,
> Nathan
Received on Friday, 5 November 2010 14:37:24 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:08 UTC