It has been forecast that by 1980, international and intercontinental links between networks will be routine, and most major computer users in industralized countries will have reasonably easy access to a worldwide communications utility.

On the other hand, the development of effective standards for data network interconnection may continue to be hampered by the large number of organizations participating. Coordination between all of these groups is not always as good as it might be, and the lines dividing areas of responsibility are not always clearly drawn. The dangers in such a situation are that so many non-standard and incompatible systems will be constructed that adoption of an eventual standard would be difficult and costly; or that a single-company standard would be imposed on the industry through market power rather than technical merit.If future standards develop in an evolutionary fashion from the framework that has already been built, then these dangers should be avoided. (ira w. cotton, international standards for data communications status report, 1977)

no standard is perfect, they say in the report. so they identified some criticism that should be discussed. well, it was a long discussion after all. with the end that datagrams, which were included in tcp/ip won. not an official, but a de facto standard at that time.

At level three, however, the controversies have been most heated, The X.25 Recommendation has been criticized on the basis that it is too complex, fails to support datagrams, and duplicates functions at several levels [PO76a,PO76b].

It is indeed true that the X.25 Recommendation is a complex specification, intended for implementation in a host computer or intelligent front end. For simpler, single access terminals (such as remote job entry terminals), a simpler interface, perhaps a subset of X.25 would be more appropriate. This subject has been raised in the appropriate standards bodies, and is expected to be addressed in the next CCITT study period.

hubert zimmermann, who was working on cyclades, where the datagrams were developed (group pouzin+support of the arpanet crew) and later on on the osi reference model, said, that the arpa people didn't show interest to join this discussion that accompanied the iso-osi standardization process. vint cerf's answer to this:

The simple answer is that the DARPA effort had been under way since 1973. By the time the first papers on OSI architecture emerged,1978, we were 5 years into the development and testing of TCP/IP and it was time to implement and deploy. As the DARPA program manager for the effort, I was not willing to spend another 5 years in negotiations on a new suite of protocols when we had something that looked pretty solid.

wise decision as the osi-discussion went on for years. in the meanwhile on the other side: with unix came an open implementation of tcp/ip, it became a de facto standard within local area networks with ethernet - no question about circuits involved, as zimmermann says, and finally intel was producing a chip ...