Re: Draft finding - "Transitioning the Web to HTTPS"

On Tue, Dec 16, 2014 at 10:19 PM, Mark Nottingham <mnot@mnot.net> wrote:
>
> Hi Mark,
>
> On 16 Dec 2014, at 3:39 am, Mark Watson <watsonm@netflix.com> wrote:
>
> > ​It would be good to have some clearer discussion of caching in the main
> document. Presently there is a reference to "content optimization", but
> it's not very clear whether this includes transparent caching.​ I think the
> impact of HTTPS on ISP transparent caching should be clearly acknowledged
> and the TAG should explain their rationale for accepting this as a
> consequence of the proposed transition.
>
> I've added some text here:
>
> https://github.com/w3ctag/web-https/commit/f9c53a41accb892cad63358811e8d88d218f4d00
>
> Note that there's a distinction between "normal" shared caching proxies
> and "transparent" (more formally, interception) proxies.
>

​Thanks. I'm not sure it's obvious to those not familiar with the
terminology that "interception proxies" are not a subclass of "shared
proxies", since the former are in fact shared by many users ​in the English
sense of the word, albeit without the users' knowledge.

Could we say "shared HTTP caches explicitly requested / configured by
users" or something similar ?

...Mark



>
> The latter have long been condemned by the IETF, and are effectively a
> loophole that some networks have exploited. Closing that loophole is going
> to cause them some pain, but they can't say they're surprised by it (the
> writing has been on the wall for some time).


> Cheers,
>
> --
> Mark Nottingham   https://www.mnot.net/
>
>