[z-machine] [Inform mntce] The Z-machine and Informs 6 and 7

[fredrik@ramsberg.net]

Quoting samwyse <samwyse at gmail.com>:

> I will defer to those with experience in the matter.  However, most of
> the discussions that I've seen take the stance that the input of
> accented characters is unneeded, because users don't want to take the
> time to use proper accenting.  OTOH, this may be a chicken-and-egg
> problem, since a decent input method may encourage greater usage.

It's a matter of language. In German, ö (o with two dots over it)
is considered to be another version of o, ie ö is sorted as o in
dictionaries, and some words that contain an o change it to an
ö when put in plural etc. For Swedish, the situation in quite
different. Our alphabet has 29 characters - a-z, å, ä and ö (in
exactly that order). ö has nothing to do with o, except that they
look similar and that they both are vowels. In Swedish, we would
never refer to these characters as accented characters, since
these accents are considered an integral part of the character.
The accents are pretty much never voluntarily skipped by anyone
for any purpose (a few companies have removed their accents in
order to look less strange on the international market, or to
get their name closer to their domain name, but at the same
time their names lose their meaning and look silly in Swedish).

An example to show why we can't just skip the accents:

ara is a parrot
åra is an oar
ära is honor or glory

I would expect Finnish to be similar to Swedish in this respect.

I should also add that a lot of terps have an appalling support even
for output of accented characters, and some can't accept a
custom Z-character table (Zplet springs to mind).

/Fredrik