[wp-hackers] Google Web Accelerator

[ifelse]

> "change the returned content based on past user interactions..."
I think there's a misunderstanding of the spec. There's nothing wrong with
canging the returned content. That's a given. The key thing is that GET
requests are idempotent i.e. a single execution of the entire sequence
always yields a result that is not changed by a reexecution of all, or part,
of that sequence. Packaging a delete in a get request is clearly not
idempotent.

On 10/25/05, Denis de Bernardy <denis at semiologic.com> wrote:
>
> > specification, the GET method is defined as a Safe Method
> > which "SHOULD NOT have the significance of taking an action
> > other than retrieval."
>
> maybe the spec writers smoke pot? I mean... I can name no site that tracks
> visitors using the POST method. and many sites, such as amazon, change the
> returned content based on past user interactions... talk about a carnage
> if
> google preloads every possible GET request with fancy urls around. there
> would be no more machine learning except on evil google's servers, heh.
>
>
> > the question of whether we should actually comply with HTTP
> > 1.1 in this manner, and if os, how would we go about doing it?
>
> one way to comply without loosing the -- at times -- elegance of GET is to
> redirect using a GET request once the POST occurred successfully. This
> allows the user to click the back button as he normally would, at times
> where the bulkiness of a POST request gives the impression of being
> irrelevant (e.g. after switching themes). I recall WP does this already at
> some places, too.
>
> D.
>
> _______________________________________________
> wp-hackers mailing list
> wp-hackers at lists.automattic.com
> http://lists.automattic.com/mailman/listinfo/wp-hackers
>



--
Phu
http://ifelse.co.uk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://comox.textdrive.com/pipermail/wp-hackers/attachments/20051025/334ef0e8/attachment.htm