Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 15 Dec 1998 12:41:29 -0000
From:      "Scott, Barry [MDN05:7E26:EXCH]" <barry.scott.tsbarry@nortel.co.uk>
To:        freebsd-isdn@FreeBSD.ORG
Subject:   RE: streams implementations
Message-ID:  <81C8165DD2A7D111AD700000F81F29CB025049E7@nwcwi19.europe.nortel.com>

next in thread | raw e-mail | index | archive | help
Jonathan M. Bresler [jmb@FreeBSD.ORG] said.

> the idea of streams is wonderful, the realization is costly.  each
> layer added (or module pushed) slows down processing and hurts
> roughput.  ritchie developed streams for serial, if i remember
> correctly.  streams was then applied to networks.  there is an RFC
> about layering being bad for networking and the relative performance
> of NIT vs BPF prove the case.

	At DEC we used a comms exec called LES (Layered environment
services) that
allowed a web of nodes to be connected. The VAX PSI and DEC NIS routers and
parts of DECnet Phase V where LES implemented comms. LES ran inside the
VAX/VMS
OS and in embeded systems like the routers. For design and debug LES ran in
user
mode simulation.

	The performance was better then direct call between layers. And
having
clean OO like seperatation of layers and (literally) message passing made
implementation fast.

	Stream (the concept) can be implemented as a very light weight
mechanism.
Of course we passed buffers around not individual bytes. And used very smart
buffer management.

			BArry

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-isdn" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?81C8165DD2A7D111AD700000F81F29CB025049E7>