Re: [Devel] Recommended mapings: æ->ae, ø->oe, å->aa

[Knut Yrvin <knuty@xxxxxxxxxxxxx>]

lørdag 26. juli 2003, 14:27, skrev Petter Reinholdtsen:
> Retningslinjer for epostadresser basert på personnavn
>     <URL:http://www.uninett.no/publikasjoner/unot/unot-2003-001.html>
>
>   Anbefaling om bruk av tegnsett i Norge
>     <URL:http://www.uninett.no/publikasjoner/unot/95-009.html>

The problem area

The pupil are interested in writing their name correctly. Defects in
the e-mail address-standards prohibits them from writing their names
correctly. Is this defect (based on standards) only a e-mail-problem,
or dos it affects login (user name), and home-directory name? European
pupils should be able to login with Norwegian characters in their user
names, despite the e-mail defect. They should be able to save files
with national characters as recommended and intended in the note from
1995: http://www.uninett.no/publikasjoner/unot/95-009.html

The facts behind the discussion could be broadened

It's understandable that Ingrid Melve has recommended the «øæå->aoa»
solution. And Petter is right when he wrote: «There is no way you can
expect people to automatically guess a persons user name based on his
name. All the arguments above seem to assume that this should somehow
be possible.  It is not». I can also see that Melve and Gjerde uses
the ISO-standard X.408. But it seems to be a bit narrow.

The first argument is that Skolelinux gives linguistics precedence
over limitations in the technology enforced from (mostly) lack of
resources to do the right thing the first time (it never happens but
anyway ;-) -- or enforced limitations in vendor owned software with
user hostile licenses.

The other argument is the linguistic history that is not reflected in
the standards. I could argue that there is an unwanted barrier between
realists and humanist. The humanist don't like technology, and the
technicians don't like linguistics and the lack of scientific thinking
up to master level in the humanistic genre of the education.

The third is the linguistic findings

http://www.wikipedia.org/wiki/%D8

  The origin of the letter is a ligature for the diphthong "OE" that
  has become a letter in itself.

http://www.wikipedia.org/wiki/%C5

  Correct alphabetisation in Danish and Norwegian places "Aa" along
  with "Å" as the last letter in the alphabet.

http://www.wikipedia.org/wiki/%C6

  The origin of the letter is a ligature for AE.

Against this is the technological opportunities (in the e-mail
standards). It's no easy way to compensate for the flaws in the
standard, as this links will show: 

http://www.phon.ucl.ac.uk/home/sampa/home.htm
http://www.wikipedia.org/wiki/Norwegian_alphabet
http://www.stud.ntnu.no/studorg/ark/tech/fonalfabet.html

When there is no way to compensate for automaticly mappings back and
forth between «æøå» and some combinations of letters, in an human
readable form, 

-- should we be forced to choose, 
-- should it be the historical linguistic and precedence 
  for language-choices in Skolelinux which should prevail 
-- or should it be a technical standard where probably nobody 
  has asked the linguistic people what best for humans to read?

The big question (because it's to early to conclude)?

The goal should be that we allow user name with «øæå», home
directories with the Norwegian letters, and also in web-address
etc. When we are not doing this because of limitations in the
standards (or tools), should we take the linguistical approach, or
should we just follow the standard that is recommended?

Suggestion (solve it with usability :)

The simple user friendly soulution to this is to give the pupils an
updated addressbook connected to the LDAP-service, and there they can
search after correct names with Norwegian letters. The e-mail-address is
then of minor importance, because the pupils get the right adresses
easy from the database before sending the e-mail :-). When doing this 
you shadows the problem with øæå->aoa | oeaeaa ...

mvh Knut Yrvin