[Sois-tcons] RE: TCONS

Fri Apr 1 05:10:02 EST 2005

Steve,

This plan of work is all very commendable, but I feel it is missing the
point somewhat. Right now TCONS, and indeed SOIS, are under threat of
extermination One of the items tabled for discussion during the CESG meeting
next Thursday is:

2. Dissolving SOIS and moving its work into SIS and SLS. The TCOA and TCONS
work could become time-constrained "flavors" of SIS application services and
SIS networking services, respectively. The Bus/LAN work would be just
another "flavor" of a space link service, as would the Wireless BOF.

So this is not a threat that we can ignore, the more so since this
discussion item has been proposed by Adrian himself.

I personally am absolutely convinced that SOIS has a very important role in
the CCSDS and will argue vigorously for its continued place within that
organization. Furthermore I am equally persuaded that TCON belongs in SOIS
and will argue for that just as strenuously. However, we must have material
with which to fight our case.

As you know, the world is full of inappropriate and technically inferior
protocols (and I include many of the internet protocols among these), so it
is clear that technical excellence is not a guarantee of success. If we look
at the case of ISO standards, the reason behind this is very simple - ISO
spent far too long polishing, refining, and supposedly perfecting their
offerings. By the time they put them out on the street, other standards had
already achieved an unassailable deployment base and most of the ISO
communication standards have languished on the shelves ever since, despite
their technical superiority!

I would therefore earnestly request that you focus your energies over the
next few days on making the case for TCONS. This must be a rational based on
WHY, not on the whats and hows of TCONS, and it should not be diluted by
focusing on technical details like whether and how it supports IP for
instance. Could you please come up with some structured text that makes this
case in terms of the real onboard needs and constraints?

If TCONS is to survive the next couple of weeks, we all need to pull
together on this one. The sooner you can give us something that we can
challenge, clarify, and refine, the stronger will be our case.

Cheers,

            Chris.

  _____  

From: Steve Parkes [mailto:sparkes at computing.dundee.ac.uk] 
Sent: 01 April 2005 10:43
To: 'Dai Stanton'; 'sois-tcons at mailman.ccsds.org'
Cc: Chris Plummer; 'Patrick.Plancke at esa.int'
Subject: RE: TCONS

Dear TCONS members,

Dai asked me to fill him in on the results of yesterday's teleconference.

Once I had started to write a reply I thought that it would be useful for
everyone.

Please let me know if I have got anything wrong.

The current plan is to complete the work on the Guaranteed Service prior to
the meeting in Athens. Then whatever is agreed there we will have finished
that work so that when we return to it in two years time at least we will
not have to start again from scratch. I want to spend one day in Athens
briefly going over this and making sure that there are no major issues and
that the various sub sections for the Best Effort and Guaranteed Delivery
services are consistent. We will then move on to work on the scheduled
service, which everyone agrees is needed. At the same time we will ask those
people that have proposed TCP/IP for TCONS to work on a section for the red
book that we can review once complete. We will also review the idea of using
the Network and look at the reliability model for spacecraft onboard
networks.

Chris and Patrick plan to rigorously defend TCONS at the CESG meeting based
on the constraints and needs of the onboard network environment.

The proposed agenda for the TCONS group work at Athens is:

1.	Blood letting and hot air generation (Athens is a good choice of
venue, although somewhere in Italy would have been better) - half a day

a.	This will hopefully result in an action on some TCP/IP experts to
write a section for the TCONS red book for review by us at the Fall meeting

2.	Requirements from Messaging Service - half a day

a.	Particularly on traffic classes (QoS)
b.	Are there any time critical requirements that cannot be met by
TCP/IP?
c.	Is it envisaged to run Messaging service over TCP/IP or UDP, what
would the constraints be?

2.	Review of Best Effort and Guaranteed Delivery services - no more
than one day

a.	Aim is to tidy up what we have done so far so that we can push the
scheduled service forward

3.	Consideration of IP header for TCONS network layer pros and cons -
no more than half a day

a.	What fields do we need?
b.	Pros and cons of header compression?
c.	How are different traffic classes handled?
d.	Outline recommendation for a compressed version of IP header

4.	Reliability model for SOIS - half a day to one day

a.	Reliability on a spacecraft
b.	Development of a reliability model
c.	This is important to prepare for the work on the scheduled service

5.	Scheduled Delivery service - two days + Saturday after the TCONS
outing
6.	Odds and Sods (those other meetings that we cannot get out of)

TCONS members comments on this agenda please.

PS Dai - Thanks for the update to the red book.

I will send out an updated red book on Thursday next week.

Regards

Steve

------------------------------------------

Steve Parkes

Space Systems Research Group

Applied Computing, University of Dundee

Dundee, DD1 4HN, Scotland, UK

Tel: +44 1382 345194

Fax: +44 1382 348838

Mobile: +44 784 138 3779

  _____  

From: Dai Stanton [mailto:dstanton at keltik.co.uk] 
Sent: 31 March 2005 21:59
To: sparkes at computing.dundee.ac.uk
Subject: TCONS

Steve,

Sorry I had to leave the TCONS call telecom today. What was the conclusion?

Is TCONS going ahead in its current form?

Dai

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/sois-tcons/attachments/20050401/4b1f2d7a/attachment.html