Thank you to all the ones that have commented upon my question about
personal names in databases.
The consensus I see is UCS, UTF-8, 7-bit fallback with double
representation of the names (space not being a problem), (8-bit fallback),
the possibility of double forms for other alphabets and writing systems.
This is all very good, and please go on if there is more to be said.
I still must confess to you that I am not at all too optimistic when it
comes to whether I am able to convince the authorities that anything
concerning our 7-bit encoding should be changed at all (the never change a
winning team-philosophy). The "ignore the small additional marks" thinking
is deeply rooted, as I also have problems in making them demand the
Norwegian letters in passports (the current OCR-format allows ascii only,
and OCR-B does not get the support it deserves).
Thus, any report on how other countries do this would be valuable, and any
report giving utf-8 (or even 8-bit) as working systems for census
information will be valuable.
Is #any# country coding these databases with more than 128 characters
(national ascii dialects), say with 8-bit encodings like Latin 1, or Latin
2?
Is #any# country coding these databases with more than 256 characters, that
is with some procedure like multiple 8859-x-shifting of the right side of
the code table (did we call it 2202?)?
Is #any# country using utf-8 or other ucs technologies?
On this list, we can consider utf-8, etc., but when I turn to the
department they will be #very# reluctant to turn to a system (UCS encoding,
utf-8, etc) that they never have heard about.
-------------------------------------------------------------------
Trond Trosterud t +47 7764 4763
Lingvistisk institutt, Det humanistiske fakultet h +47 7767 3639
N-9037 Universitetet i Tromsø, Noreg f +47 7764 4239
Trond.Trosterud@hum.uit.no http://www2.isl.uit.no/trond/index.html
Test string-please ignore:ᄘ¹š¼¿-Á‚‰¸Šº¾-â¡¥³†ˆ-™¢²…‡-æøåäö-ÆØÅÄÖ
-------------------------------------------------------------------
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:45 EDT