Skip to Content.
Sympa Menu

shibboleth-dev - RE: WAYF talks (was WAYF cookie considered dubious)

Subject: Shibboleth Developers

List archive

RE: WAYF talks (was WAYF cookie considered dubious)


Chronological Thread 
  • From: "Scott Cantor" <>
  • To: "'Thomas Lenggenhager'" <>, <>
  • Subject: RE: WAYF talks (was WAYF cookie considered dubious)
  • Date: Thu, 21 Apr 2005 11:55:29 -0400
  • Organization: The Ohio State University

> How about a flexible and lightweight WAYF? It could be used either

I think we need to initially:

- standardize the cookie name and syntax (and I suggest we adopt the 2.0 CDC
for that) to enable somebody to write a plugin that could manage all
discovery cookies in a client easily

- let a thousand WAYFs bloom and see what takes root

> - as a local WAYF run by an SP with just a list of its partner IdPs.
>
> We are thinking about implementing such a WAYF in PHP (since
> we know it better than JSP) which could look like this:

This is the essential problem with "supplying" a WAYF to be used at an SP as
part of the project. PHP will work fine except for all the SPs for whom it
won't. That's why the Java SP can supply a Java WAYF and the C++ SP should
eventually supply an internal implementation of sorts that also requires no
additional software, however hard that is.

In paricular both PHP and Java make life harder for an SP that wants to run
solely on IIS. That's just an example of course. Everybody has their pets.

-- Scott




Archive powered by MHonArc 2.6.16.

Top of Page