Re: SECS & VSECS: Small European Character Sets

From: Markus Kuhn (
Date: Sun Aug 16 1998 - 05:22:38 EDT

Michael Everson wrote on 1998-08-15 11:18 UTC:
> As the editor of a Commission-funded project to develop one or more subsets
> of the UCS, I must protest again, that Markus Kuhn, who has been
> participating in the CEN Workship WS-MES, has published the "SECS" and the
> "VSECS" without the consensus of European and other experts actively
> working in this area.
> It may be the case that, with or without modifications, the subsets Markus
> has described will be adopted by the Workshop.
> Despite the fact that Markus has named these as "European", they do not,
> however, reflect European consensus and may not meet European requirements.

I apologize for using the term "European [TM]", which according to the
tone of this posting apparently is a registered trademark of Michael
Everson and the CEN Workship on Multilingual European Subsets, and which
therefore must not be used under any circumstances in any context by
private mere mortal inhabitants of the Continent of Europe [TM] without
prior written permission by Michael Everson.


I *never* ever implied in any of my postings about SECS/VSECS that they
represent anything agreed upon by the CEN WS-MES, and therefore, your
"protest" is, sorry, a bit childish. The SECS/VSECS proposal is
completely my own (unfunded) volunteer work, the major input were
existing published ISO standards and published literature on European
character set requirements. I even used completely new names for my
proposal to make clear that they are my personal projects and not
CEN-MES work.

I might have implied that I personally consider the drafts the CEN-MES
workshop has produced so far is not yet very satisfactory (and I do not
have the impression that I am completely alone with this opinion). Of
course I will continue to represent my technical opinion in various
mailing lists.

I do propose to the CEN-MES workshop my SECS/VSECS proposals in the
presented or modified form as a candidates for standardization. I
equally offer them to any other interested organization, including
Unicode Inc., JTC1, ECMA, IETF, etc., for further discussion and
potential adoption. I believe that wide discussion of the various subset
proposals on the usual mailing lists will make it more likely that the
end results will not only have the formal consensus of some
Commission-funded committee but will also meet real user requirements.

Why do I want to have some independent discussions about European [TM]
Unicode subsets going on outside CEN-MES?

I do think that some influential CEN-MES linguistic experts occasionally
represent a bit extreme or at least very academic requirements for
extremely narrow linguistic applications without realising all the
consequences for implementors and users. I also got the impression that
the CEN-MES results were already a priory fixed politically to conform
to certain standards that I think are of close-to-zero practical
relevance (namely ISO 6937 and ENV 1973). This led to IMHO somewhat
bizarre results such as that the current MES-1 draft does not contain
the Euro sign, the directed quotation marks, and is not even a superset
of ISO 8859-1, and that MES-2 contains many digraph characters that make
fixed-width font implementations extremely ugly and cause encoding
ambiguities. Therefore, I was not entirely convinced that I will like
the final outcome of the CEN-MES (but I will be delighted if I shall be
proved wrong here). The current CEN-MES requirements might be quite far
from the requirements of most other computer users, and some independent
competition I think is very healthy. Therefore I want to get some
feedback from a broader user community than from just a few committee
linguists (some of which have in my purely personal opinion repeatedly
demonstrated a disturbing lack of experience in broader practical
computing issues). Is that really inappropriate?


Markus G. Kuhn, Security Group, Computer Lab, Cambridge University, UK
email: mkuhn at,  home page: <>

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:40 EDT