Re: [lemonade] Re: WGLC compress-02 IANA registry

[Arnt Gulbrandsen <arnt at gulbrandsen.priv.no>]

Dave Cridland writes:
> On Thu Aug  3 02:08:26 2006, Tony Hansen wrote:
> > Using a separate registry, I was trying to make it harder to add 
> > future algorithms.
> >
> > If we modify comp-meth-tls, is that going to suddenly swamp us with 
> > lots of potential algorithms and make interop testing harder?
>
> I had assumed that if we allow for multiple algorithms by any 
> mechanism, then DEFLATE would become a Mandatory To Implement 
> algorithm.

I forgot to put that in. Sorry. There's also a wrong reference in -03, 
so a -04 is necessary. I'll do it once the chair have interpreted 
yesterday's hummish thingy. Possibly tomorrow, more likely Tuesday.

But Tony is right. To test interoperation, we need to test the algorithm 
each client-server pair chooses to use. If a given client and server 
both support LZS and DEFLATE, and the client prefers LZS (see 
comp-meth-ids), then we need to to test that they correctly choose LZS 
when talking to each other, and that their LZS implementations 
interoperate.

Arnt

_______________________________________________
lemonade mailing list
lemonade at ietf.org
https://www1.ietf.org/mailman/listinfo/lemonade