in addition to what mark indicates:
* there was a lot of push back on
the correlating of different hosts in the same DNS domain with
the same set of proxies (which may have more of a physical rather than
adminstrative correlation)
* The scope of what was to be searched for in DNS was pretty weak security wise
too.. no way to tell when administrative boundaries were crossed.
* It doesn't really solve the bootstrap problem well for all
environments
* There was no notion of dynamically establishing a set of
intermediaries.. discovering the description file was done
dynamically, but there was no protocol basis for individual
intermediaries to add or remove their entries/relationships in that
file.
of course its possible all these things are un-necessary and we'll end
up thinking WPAD is sufficient.. hopefully the requirements process
tells us that.
-P
[Mark Nottingham: Tue, Jan 30, 2001 at 01:16:31PM -0800]
>
> Some things that come to mind / I've heard;
>
> * WPAD discovers the javascript configuration format, which is not a
> standard.
>
> * WPAD has a laundry-list of discovery mechanisms, making client
> implementation as well as administrator configuration difficult
> (i.e., it's problematic to decide which WPAD mechanisms to deploy,
> because your clients may have different or unknown capabilities).
>
> Personally, I think the discovery portion of IDD may look very
> similar to parts of WPAD; it's the description that needs the greaest
> amount of work.
>
> Definately a good idea to include this in requirements.
>
>
> On Tue, Jan 30, 2001 at 12:55:04PM -0800, Phil Rzewski wrote:
> > At the meeting in San Diego, the stated motivation for the IDD work
> > included the statement that WPAD was ruled as an inadequate IDD protocol. I
> > remember at least two reasons cited as to why (reliance on JavaScript,
> > working only for HTTP), but there may have been more reasons. Can anyone
> > enumerate the exhaustive list? Would it be wise to have that list inside of
> > an IDD draft (say, an early section in the Requirements document) so that
> > there will be an authoritative source for that? I'm sure the question will
> > come up again, as WPAD exists in browsers, etc.
> >
> > --
> > Phil Rzewski - Senior Architect - Inktomi Corporation
> > 650-653-2487 (office) - 650-303-3790 (cell) - 650-653-1848 (fax)
>
> --
> Mark Nottingham, Research Scientist
> Akamai Technologies (San Mateo, CA)
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:22:56 MST