🚀 go-pugleaf

In article <8494@charon.cwi.nl> dik@cwi.nl (Dik T. Winter) writes:
>As somebody else mentioned already: should a spelling-checker allow a
>German A-umlaut in a Swedish word?

No, it should not. As well as it should not allow similarly-looking
hierogliphs of mumbo-jumbo tribe. Think of that as of different letters
which happened to look remarkably alike -- like I and l. Moreover indroducing
foreighn letgter in a middle of a word requires deliberate action and in practice
is real seldom occurence (i have a keyboard with Cyrillic with a lot
of similar-looking letters - o, e, k, x, c, m, E, T, O, p, P, A, D, H, K,
X, C, B, M with TWO codes in KOI-8) -- ever saw it in my postings here?
And i do not use spell-checkers.

>If so, what are the consequences,

The good spell-checker will suggest to replace the letter with the
correct one.

>Moreover, one question: how would you encode the German A-umlaut such that
>it sorts properly (i.e. as if it is the letter combination AE)?

The sorting order should be strict -- if you have two identical words
with a-umlaut and ae in the middle is it the same word? If it is then
ae IS a variation of a-umlaut and should always be treated as a signle
letter.

>Even in
>a single language (German) you can not come up with a coding that gives
>proper sorting. 

You can come with a reasonable approximation anyway.

>And also how to code the German eszet such that when
>uppercased it becomes the *two* letters 'SS' (and how do you lowercase
>that again?  Things are not as simple as you appear to think.

Then there is an uppercase eszet which looks like SS. An input program
will map the keyboard accordingly.

Forget about "traditions" -- users do not care which code is inside if
it looks like their usual stuff.

Basically it is a purely mathematical problem -- you've got a number
of orderings and map it into single (partial) ordering by merging
as much as possible elements. The number of meta-alphabets is generally
determined by the criteria of minimality of duplication.

--vadim