On Fri, Aug 5, 2011 at 4:23 PM, Sean Turner <[email protected]> wrote:
> > I have no problem with adding something along the line of: > > OLD: > > The resulting solutions will support both JSON-encoded public keys and > X.509 public key certificates. > Works for me. > I purposely choose "support" and didn't specify which is the MTI. We can > have that debate when writing the spec. > > do others have an issue with this? > I would prefer not to specify an MTI at all. One of the neat features of C# is that you can now specify List <Foo> and get a strongly typed list of type Foo. Unlike in C where the list container is generic, you can specify a structure with a List <foo> and know that nothing else will be put in it. What I would like to end up with with Woes is scheme where a specification using Woes can specify Woes <PKIX> or Woes <Raw>. I think it is a mistake for every IETF security WG to get into discussions on which algorithms to make mandatory for the same reason. We should have one set of algorithms that is the preferred IETF set across the board and those should apply unless a particular WG has a very particular reason to do otherwise. In fact that is the reason I want WOES in the first place. I can write out a way to sign this stuff in a few hours and make it work for my stuff. I can even write up a spec. What I greatly dislike is a situation where we get into a committee and spend two years making a series of design decisions en-bank and then join another group and go through the whole set all over again, often with many of the same people. -- Website: http://hallambaker.com/
_______________________________________________ woes mailing list [email protected] https://www.ietf.org/mailman/listinfo/woes
