Marshall Rose (mrose@TWG.COM)
Fri, 08 Jul 88 20:49:22 -0700

     It is so tempting to simply flame at you two (Kastenholz and Smart)
     but rather than do so I will explain, calmly, the errors of your
     ways. Consider this a pronouncement of The Truth.

     It is a mistake to, as Robert Smart suggests,

> Anyway we don't want the standard ISO documents on-line. They're
> incomprehensible (except for the CLN documents which are nice). What
> we want is a description of the protocols that us ordinary folk can
> read.

     re-edit the ISO documents and then distribute them. The ISO
     documents use a consistent OSI terminology and are perfectly
     comprehensible from the basis of that framework. Three years ago I
     was unable to read an OSIfied document and make sense out of it.
     Today I am able to, and can tell you that the network layer
     documents aren't really any better or worse than any of the other
     OSI documents with respect to readability. In fact, if you want to
     read a truly outrageous document, get a copy of "The Internal
     Organization of the Network Layer" (the IONL), which will convince
     any thinking person that the TCP/IP architecture is vastly superior
     to the full-blown OSI network layer.

     The reason why the ISO documents are copyrighted is not so Omnicom
     can make a paltry sum of money on each standard (the ~$1000 figure
     is for their update service in which they filter the output of
     standards bodies for you and send you the things that you are
     interested in seeing). The real reason is so that


     The problem with having anyone edit them is that you lose meaning
     and misinform. OSI documents make sense once you learn the lingo.
     It is more formal than the language used in the DARPA/NSF Internet
     community (plain English) but it is not as bad as the way the Inca
     of Peru used to write things down (they stored information on
     strings, carefully knotted in a specific manner and with colored thread).

     Now, having told you The Truth, keep in mind that I think that it
     would be nice for someone to write a moderately lengthy explanation
     of what each OSI standard is saying, and to write that in a more
     easy to understand format. It would be less accurate, but would be
     useful for getting across the gist of things. However, editing
     standards and the like is simply the wrong way to do it.


ps: The word "OSIfy" is a new word that I invented earlier this year.
     The precise meaning is:

        "to obscure, to make unclear for no good reason"

     It is often used in the context of the output of standards
     committees, although its use is not limited to committees which
     produce OSI standards. In fact, some might claim that MILSTD 1778
     is an OSIfied version of RFC793.

This archive was generated by hypermail 2.0b3 on Thu Mar 09 2000 - 14:42:50 GMT