[OpenISO] Document types and labeling

[Norbert Bollow]

So far I can foresee OpenISO.org dealing with the following types of
documents that would be part of OpenISO.org's formal processes:

a) Procedural guidelines document
b) Legal document
c) Specification (developed in an OpenISO.org process or elsewhere)
d) OpenISO.org Endorsement of a specification
e) OpenISO.org Withdrawal of Endorsement (hopefully very rare)
f) OpenISO.org Problem Report about a specifications
g) Draft of any of the above
h) Comment about one or more documents of any of these types

The "legal documents" (such as covenants regarding patent rights) and
comments can simply be numbered sequentially (perhaps preferably in
different numbering domains).

For the others it would be helpful to have a somewhat sane labelling
system that allows to recognize immediately which documents are
closely related to each other.

I would suggest the following EBNF syntax for the label:

"OI" ( "STD-" | "NOSTD-" | "PR-" )? [A-Z] [1-9] [0-9]* [0-9] [0-9]
( ":" [1-9] [0-9] [0-9] [0-9] [a-z]? | "-draft" [1-9] [0-9]* )

with the following semantics:

* The "OI" part simply functions to differentiate the label space
  from those used by other standardization organizations

* The "STD-" prefix is used for endorsements

* The "NoSTD-" prefix is used for the withdrawal of an endorsement

* The "PR-" prefix is used for problem reports

* The upper case letter indicates a category of subject matter, e.g.
  "A" - OpenISO.org prodecural documents, endorsement criteria etc
  "B" - Nomenclature
  "C" - Physical layer and data link layer
  "D" - Network layer and transport layer
  "E" - Application layer protocols
  "F" - Textual data formats
  "G" - Other data formats
  ...

* The upper case letter will be followed by a numer that has at least
  three digits.  A number which is divisible by 100 will be allocated
  for any specification which is not closely related to one for which
  a number has already been allocated.  Otherwise if possible the
  number should differ only in the last two digits from the numbers
  of closely related specifications.

* The combination of upper case letter and the following number will
  be called the "TopicID" in the following.

* For published documents, the TopicID is followed by a colon and then
  the year number of the year (in the Gregorian calendar) in which
  OpenISO.org publishes the document.  If (in exceptional cases) more
  than one document with the same TopicID is published by OpenISO.org
  in the same year, a lowercase letter ("a", "b", "c",...) is appended.

* For draft documents, the TopicID is instead followed by the string
  "-draft" and a serial number that is simply incremented by one for
  each draft with the same TopicID.

Possible examples of documents that we might create this year:

OI A100-draft1 ----- first draft for the OpenISO.org Core Guidelines
OI A100:2007 ------- OpenISO.org Core Guidelines (first published version)
OI A101:2007 ------- OpenISO.org Openness Criteria (first published version)
OI A102:2007 ------- OpenISO.org Maturity Criteria (first published version)

OI B100:2007 ------- OpenISO.org Decsription of the ISO/OSI Reference Model

OI STD-D100:2007 --- OpenISO.org Endorsement of the IPv4 protocol
OI STD-D101:2007 --- OpenISO.org Endorsement of the TCP protocol

OI STD-D200:2007 --- OpenISO.org Endorsement of the IPv6 protocol

OI STD-F29300:2007 - OpenISO.org Endorsement of ISO/IEC 29300:2006

OI PR-F29500:2007 -- OpenISO.org Problem Report about OOXML

Greetings,
Norbert.


-- 
Norbert Bollow <nb at bollow.ch>                    http://Norbert.ch
President of the Swiss Internet User Group SIUG  http://SIUG.ch