Re: [tcpm] I-D Action: draft-ietf-tcpm-1323bis-07.txt

>    RFC 1323 seems to deserve a path to Historic...

You sure?

That's make WindowScale historic too.... (and yes, the mingling of WS, TS and SACK in one RFC1072 two decades ago was not the wisest approach).

I think 1323bis should finally be published, simply to fix the omissions and bugs of the original text in 1323.

And for the time being, PLEASE let us not discuss the merits of using RTTM/PAWS/TSopt; 

Most stacks leave that as a global option, and from what I can tell, many disable this by default. And that is fine. But if one chooses to use TS, the mechanisms should be described properly, without any significant issues in the RFC for someone how is not into hunting down bug fixes to an RFC in running code. 

And WS, as described in 1323, does have its share of issues (implementers not considering window retractions properly). 

I'll post a -08 shortly, for a consolidated text. I would urge anyone who has a major bone to pick with the TSopt, to come up with improved text (thanks Joe for suggesting improvements). So far, the existing text describes the typical TSopt use case / state machine. 

Best regards,

Richard Scheffenegger

> -----Original Message-----
> From: tcpm-bounces@ietf.org [mailto:tcpm-bounces@ietf.org] On Behalf Of
> John Leslie
> Sent: Freitag, 29. März 2013 14:32
> To: Joe Touch
> Cc: tcpm@ietf.org
> Subject: Re: [tcpm] I-D Action: draft-ietf-tcpm-1323bis-07.txt
> 
> Joe Touch <touch@isi.edu> wrote:
> >
> > On Mar 27, 2013, at 8:25 PM, Mark Allman <mallman@icir.org> wrote:
> >>
> >> I think this is just absurd.  This is about PAWS.
> >
> > If you want absurd, see PAWS. That mechanism can be accomplished with
> > a very short option that encodes the top byte(s) of an extended
> > sequence number space - in much fewer than 10 bytes.
> 
>    I've got to agree with Joe here. PAWS is complicated, non-intutitive,
> and drains entirely too much TCP option space. If the function is
> important, we should design a better way to do it.
> 
>    If, OTOH, the funtion isn't worth designing a better way, I'd let the
> sleeping dogs lie until we can deprecate it.
> 
>    RFC 1323 seems to deserve a path to Historic...
> 
> --
> John Leslie <john@jlc.net>
> _______________________________________________
> tcpm mailing list
> tcpm@ietf.org
> https://www.ietf.org/mailman/listinfo/tcpm