RE: Chunking driver (was RE: [Simple] MSRP Boundary header)

[Eric Burger <eburger at brooktrout.com>]

Actually, -07 clearly explains why chunking is needed (finally!).  It is for
when you have a single pipe with multiple sessions, and you want to ensure
some sort of "fairness" algorithm.

The concept really looks good when we introduce Relays.

Thus, I am on board, and 2K or 8K is quite livable.

> -----Original Message-----
> From: Sean Olson [mailto:seancolson at yahoo.com]
> Sent: Sunday, July 18, 2004 12:48 PM
> To: 'Eric Burger'
> Cc: simple at ietf.org
> Subject: RE: Chunking driver (was RE: [Simple] MSRP Boundary header)
> 
> 
> Perfect. Move that number to 8K. That just makes the argument 
> stronger. The
> point is that you can make an intelligent decision about the 
> dividing line
> between where to chunk and where not to. Of course in that 8K 
> packet, you
> are likely to put quite a few small IMs (for the relay and 
> MCU scnearios).
> Which is even more reason not mess with scanning for a boundary.
> 
> -----Original Message-----
> From: Eric Burger [mailto:eburger at brooktrout.com] 
> Sent: Saturday, July 17, 2004 11:33 PM
> To: seancolson at yahoo.com
> Cc: simple at ietf.org
> Subject: Chunking driver (was RE: [Simple] MSRP Boundary header)
> 
> Ummm.  Isn't the transport TCP?  If you are on a LAN, aren't 
> you going to be
> using 8K packets, anyway?  E.g., I really don't get the 
> chunking drive.
> 
> Can't claim "small device": are you saying that these devices won't be
> surfing the web, getting e-mail, etc.?
> 
> > -----Original Message-----
> > From: Sean Olson [mailto:seancolson at yahoo.com]
> > Sent: Tuesday, June 29, 2004 10:48 PM
> > To: 'Vijay Hampapur Hampapur Parthasarathy'; 'Ben Campbell'
> > Cc: simple at ietf.org
> > Subject: RE: [Simple] MSRP Boundary header
> > 
> [snip]
> > 
> > The choice of which to use will typically be made based on content 
> > size and is a simple implementation decision.
> > For example, for messages larger than 2K, it probably makes 
> sense to 
> > chunk up the content. Otherwise, why not use a simpler approach.
> 

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