[net.internat] 7-bit vs. 8-bit international characters sets

tmb@talcott.UUCP (Thomas M. Breuel) (10/06/85)

In article <149@ecrcvax.UUCP>, pete@ecrcvax.UUCP (Pete Delaney) writes:
> Perhaps we could start with the issue of character sets.  It seems that ATT
> is intent on 7-bit ASCII whereas from my discussions with Eric Brunner,
> a major contributer to XOPEN, it sounds like 8-bit ascii if far more
> reasonable.  If 7 bit ASCII is used utilities must know how to deal with
> double byte character sequences. If these excaped characters are maped to
> 8 bit ACSII the utilities need not be changed. 

Neither representation of national characters (escaped vs. 8 bit) is
going to make a transition from an American ASCII to a different national
character set easy. Many programs utilise or mask the 7th bit routinely.
Even if the high bit of ASCII characters is preserved, the numerical
ordering is most definitely not going to agree with most national needs.
A change in computer character set is always going to cause major changes
in current operating and utility software, regardless of how the change
is implemented.

						Thomas.

steve@hcradm.UUCP (Steve Pozgaj) (10/08/85)

In article <518@talcott.UUCP> tmb@talcott.UUCP (Thomas M. Breuel) writes:
> ...
>A change in computer character set is always going to cause major changes
>in current operating and utility software, regardless of how the change
>is implemented.

That's true.  However, in order to attempt to achieve a truly internationally
applicable software bus, for which UNIX seems to have volunteered itself, an
effort must be made to anticipate solutions that are global enough to warrant
making a major change "only once".  I realize that that is a difficult goal,
but it is certainly the mood of the Internationalization effort sponsored by
/usr/group.  At least, that's my reading of what happened at the most recent
meeting (McLean Virginia, September 9, 1985).

It is also a reasonable goal.  To talk in terms of 7-bit or 8-bit is not too
meaningful for countries that have a recognized need for 16-bit characters.
Therefore, the fact that change will be necessary must, I believe, be accepted.
In order to minimize the frequency and impact of such change, a well thought
out standard must be decided explicitly, and not be allowed to happen by
accident.  I believe that that should be the intent of discussion in this
newsgroup.