Re: Ben Campbell's Yes on draft-ietf-httpbis-alt-svc-13: (with COMMENT)

On 2 Mar 2016, at 12:36, Mike Bishop wrote:

> 2.2 - Hypothetically, a more complex client might cache per network 
> location and revive the cached entries when it returns to the network 
> where it received them.
>

That seems reasonable. It might be worth mentioning that in the text as 
an example of why one might not follow the SHOULD.

> I think your reading of MAY/SHOULD is correct.  Using Alt-Svc itself 
> is totally optional -- but if you choose to, this is when you SHOULD 
> switch to a given alternative.

I suspect the "if you choose to" part was intended to be implied by "if 
a client becomes aware of an alternative service...", with the idea that 
only clients that implement this would "become aware". But I think it 
would be more clear to say something like "If a supporting client 
becomes aware..."

Thanks!

Ben.

Received on Wednesday, 2 March 2016 18:42:26 UTC