There was no discussion under this agenda item.
It was decided that since the interests of the Single User Systems Steering Group overlapped with those of LANSC, PEB would communicate with the chairman of SUSSG and make our presence known.
A number of comments were raised with regard to paper LANSC/P3/82 by PEB. 10 Cambridge Pilot Rings
The meeting then turned to paper LANSC/P2/82 which, due to secretarial inefficiency, was not distributed until the meeting.
CSC then described the philosophy of the Universe papers which fall into two groups. There are Implementation Specifications which are almost cast in iron and are very difficult to change, and Outlines of Experiments which say in which areas the various Universe sites are interested.
Among some of the projects are:
It was noted with regard to paper LANSC/P23/82 (Local area networks at Daresbury Laboratory) that if an equivalent of BBP was designed for an Ethernet, then Universe bridges could be made to work over the Ethernet (as well as all the other existing services and protocols). As a result of this, the item ECMA protocols is to be added to the agenda as a standing item.
SNS reported that the PUNCH system will be made up of VAX computers. The initial order will consist of three 730s and one 750. The LAN technology to be used is still under consideration, and in order to provide experience, an experimental ring (made by Seel) has been purchased. They are currently waiting for the cable to be laid.
WPS reported on the state of the ring harmonisation proposals. Draft three had been issued and many comments were forthcoming. Certain comments (by CAMTEC) indicated that there were shortcomings in the document and a final version is expected soon.
The final specification will be sent to BSI at some future time. This will provide an authority for the specification.
The JNT ring protocol specification group expect to be close to the end of TSBSP. They expect to finalise the document in about 6 weeks time.
It was unclear precisely what a lightweight protocol was, but it was decided that they would be used in a Transaction environment, with large amounts of data, low intelligence of servers and where low latency was required.
It was also decided that connectionless protocols were a misnomer; there is nearly always some relationship between different transactions.
These would deal with data rates of over 100 Megabits per second. A possible candidate is a slotted ring. The impetus behind this activity is to try and define a standard before any serious implementations are produced.
The initial proposal is to get sufficient money to set up an experimental group to generate a specification and after two years to have some experimental networks.
It was suggested that a broadband experimental network be purchased and installed at some future time.
No protocol decisions have been taken, the initial phase is just to get the hardware working.
There are six possible sources for a ring gateway.
Gateway progress is depressing in that there is no product even on the horizon. Bridges are in the same state.
For PERQs it is possible to have an X25-ring gateway running on a PERQ UNIX system.
It would be desirable to initiate a product with a firm date on it.
The date of the next meeting is Friday, 8th October 1982 at the Rutherford Appleton Laboratory.
