Re: Preliminary minutes from the Design Team Meeting: WPAD and Proxy.pac

Since proxy.pac tells you which hosts should be proxied and where the proxy
can be found, it is about discovery, just like WPAD. There was general
interest in the topic of discovery, including the possibility that in path
interception is actually the best form of it. In general, describing the
problems in the current discovery mechanisms should be part of the
informational doc on proxies that was discussed as a next work item for the
working group. It shouldn't be limited to WPAD or proxy.pac or whatever in
my opinion.

Peter


On Wed, Mar 12, 2014 at 8:54 PM, Mark Nottingham <mnot@mnot.net> wrote:

> Emile,
>
> On 11 Mar 2014, at 4:34 am, emile.stephan@orange.com wrote:
>
> > Hi Mark,
> >
> > The draft of the minutes says:
> >
> >> Discovery is hard. We encourage interception proxies through inaction.
> Not much interest in standardising WPAD (security concerns, deployment
> concerns), but strong interest in proxy.pac from implementers, due to
> considerable pain. Would be interested in clarifying the current format and
> normalising behaviour as much as possible, and potentially in extending /
> replacing the format. E.g., IPv6, secure proxy.
> >>
> > I had the feeling there was the same interest in refurbishing both of
> them.
>
> If by "both of them", you mean both proxy.pac and WPAD -- we had a
> discussion explicitly about WPAD, and many, many people expressed concern
> about security issues with the protocol; furthermore, there was very little
> interest in pursuing it, to my recollection. OTOH there was strong interest
> in proxy.pac.
>
> Cheers,
>
>
> --
> Mark Nottingham   http://www.mnot.net/
>
>
>
>
>