RE: Improving the state machine

I just gave a try at the send/recv clarifications.

https://github.com/http2/http2-spec/pull/659


Hervé.

> -----Original Message-----
> From: phluid61@gmail.com [mailto:phluid61@gmail.com] On Behalf Of
> Matthew Kerwin
> Sent: jeudi 27 novembre 2014 00:51
> To: Greg Wilkins
> Cc: Martin Thomson; HTTP Working Group
> Subject: Re: Improving the state machine
> 
> On 27 November 2014 at 09:28, Greg Wilkins <gregw@intalio.com> wrote:
> 
> 
> 
>  <snip>
> 
>  But other than than old lost battle.... I think the send/recv clarifications
> are good ones.    Not such a big fan of the use of the word "either", which looks
> like it applies to the frame type and not to the send or recv.   Maybe
> "send/recv" would be better than "either".
> 
>  ​
> 
> 
> 
> I agree, the word "either" isn't great. Since the space in the diagram is so
> limited, I'd be tempted to suggest symbols for send/recv/both. However that
> would get messy fast.
> 
> Why not leave out the "either", and reorder the two ambigious cases: "R / send
> ES", "R / recv ES" ?
> 
> If that's too finicky, this PR is still an improvement.
> 
> 
> --
> 
>   Matthew Kerwin
>   http://matthew.kerwin.net.au/

Received on Thursday, 27 November 2014 08:00:40 UTC