[Top] [All Lists]

Re: Q+A: Mnemonic to Proposed Standard

1992-04-08 05:04:47
(1) Could we create a separate list for this, or take it offline?  If
the thing is going to be either "informational" or ""experimental", I
don't see the WG having to take a position on anything but whether it
ight interfere with anything else we are doing as a WG.  The answer to
that is pretty clearly "no".

Two specific comments on the latest set of transactions...

(2) The canonical behavior of other standardizers.
That is true, some of the characters do not have the same names
as in 10646, some even does not have names (some vendor sets).
So I had to judge what it was. And that could be a source of errors.

Actually, in that case I wouldn't call it an "error", I would call it
"your opinion". So people using your tables to convert between
character sets would be basing their conversions on your opinion. The
people who actually designed or actually use the character sets may
have a different opinion. That is also one of my concerns.
   On a DIS vote, ISO insists on near-unamimity, with a rule that two
(!) unresolvable negative votes from P-members triggers a special
problem resolution procedure involving the Secretary-General.  So
everyone gets to look at something like this, compare it to their own
practice, and say "no" if anything is seriously wrong.  And the voting
procedure makes it hard for member bodies to sit one of these things out
and then be surprised by it.
   On the other hand, this particular safety measure is one of the
reasons "their" standards sometimes take such a horribly long time.

(3) The misbehavior of standardizers.
But it may not be a great help, as some standardisation bodies
are reluctant to standardise something which is already in use,
and thus gives somebody an advantage.

This is probably the most ridiculous thing I've heard this week (well,
maybe today). If there are standards bodies that behave in this
fashion, they deserve to...
  I've never heard of this one being pulled in the character set area,
but, indeed, in some others, saying "if we chose that solution, we would
be giving manufacturer A an advantage in the marketplace, hence we have
to choose something completely new which we will go invent" is the
classic ISO model for saying "not invented here".  It is also  a
behavior prohibited by the ISO Directives, at least as I read them: they
contain words about standardizing the consensus of existing practice,
not about inventing.  But that piece of knowledge plus some small amount
of an appropriate international currency will buy the proverbial cup of
   On the other hand, this is a non-issue wrt the proposal at hand,
since the main claim used to push it forward in IETF has been that it is
already in active use.

<Prev in Thread] Current Thread [Next in Thread>