Re: [hybi] Comments on draft-ietf-hybi-permessage-compression-05

[SM <[e-mail]>]

Hi Takeshi,
At 05:13 12-03-2013, Takeshi Yoshino wrote:

>It's commonly used in RFC 6455. Maybe you meant that I should cite 
>RFC 6455 and define the term in this document. Done.

Yes.  What I meant was what the implementer would have to do as 
"fail" does not specify specific behavior, e.g. return a specific 
error condition.

>Not sure which is the most supported format...

It is a matter of personal preference. :-)  Pick something you find 
better and people can review and comment.

>Basically I feel like it's too much. But yes, it's good to show an 
>example of a diagram with the RSV1 is set once.

Ok.

>Sorry I couldn't get what you mean.

What I meant was not to get into all the details of how to 
implement.  It is better in my opinion to explain what is being 
sent.  Please let me know if I did not explain it well.

>Added one concern (citing CRIME attack).

Ok.

>Which part of the sentence is bad for the RFC Editor?

I don't really know how to explain this.  I would say that the 
sentence looks more like pseudo code instead of one written in English.


>Could you please point a sentence which is considered to be making 
>liberal use of the key worlds? I tried to remove inappropriate use 
>but couldn't find so many.

I'll comment using an example from draft-ietf-hybi-permessage-compression-07:

   'A server MAY attach the "c2s_no_context_takeover" extension parameter
    to disallow the client to use the LZ77 sliding window used to build
    frames for the last message the client sent to build frames for the
    next message to send.  The "c2s_no_context_takeover" extension
    parameter has no value.  Clients SHOULD be able to accept the
    "c2s_no_context_takeover" parameter.  A client that received this
    parameter MUST reset its LZ77 sliding window for sending to empty for
    each message.'

There are three RFC 2119 key words in the above, i.e. MAY, SHOULD, 
MUST.  The client "SHOULD be able to accept" is an assumption.  The 
"MUST reset" is what the client must do.   I could write it in terms 
of what "c2s_no_context_takeover" means and what the client must do 
when it receives the parameter.

Regards,
-sm