Columbia Kermit/kermit.zip/newsgroups/misc.19980901-19981211/000387_news@newsmaster….columbia.edu _Thu Dec 3 10:30:36 1998.msg

Internet Message Format | 2020-01-01 | 6KB

Return-Path: <[e-mail]>
Received: from newsmaster.cc.columbia.edu (newsmaster.cc.columbia.edu [128.59.35.30])
    by watsun.cc.columbia.edu (8.8.5/8.8.5) with ESMTP id KAA29359
    for <[e-mail]>; Thu, 3 Dec 1998 10:30:36 -0500 (EST)
Received: (from news@localhost)
    by newsmaster.cc.columbia.edu (8.8.5/8.8.5) id KAA28907
    for kermit.misc@watsun; Thu, 3 Dec 1998 10:30:36 -0500 (EST)
Path: news.columbia.edu!watsun.cc.columbia.edu!fdc
From: [e-mail] (Frank da Cruz)
Newsgroups: comp.protocols.kermit.misc
Subject: Finnish character set
Date: 3 Dec 1998 15:30:34 GMT
Organization: Columbia University
Lines: 102
Message-ID: <746aqq$s98$[e-mail]>
NNTP-Posting-Host: watsun.cc.columbia.edu
Xref: news.columbia.edu comp.protocols.kermit.misc:9594


> Erkki Eerola wrote:
> > -----Original Message-----
> > From: Frank da Cruz [mailto:[e-mail]]
> > > >What application do you have that does not use any of Kermit 95's
> > > >built-in character sets?
> > >
> > > I need a character set which has the scandinavian characters
> > > and the same time the "backspace"-character "^".
>
> That was a typing error, Erkki meant "up-arrow"
>
> > What is the code of this character, and what do you expect it to do when it
> > is sent to the terminal screen?
>
> Up-arrow is decimal 94 and we would like to have it coming from key scancode
> \954 (the key below backspace, left of return. On finnish keyboard it has
> up-arrow, umlaut and tilde stamped on top of it). Old kermit produced up-
> arrow from that key by pressing shift-^-space, like MSDOS/Winxx pc's usually
> do. Win95-kermit beeps and a "?" appears.
>
> I tried "set terminal character-set german" and then "^" worked fine (only
> the swedish circumlex-a was missing). In Finland we need that character too.
>
> If we put character code \94 "^" to whatever key, result is same, in finnish
> character set it always comes out with a beep and "?". To me it seems that
> kermit-95 doesn't know what to do with character 94 in finnish character set.
>
> In good old times it was handled just like in us-ascii, with no conversion.
>
>     greetings,
>             Esa Aarnio, host/[e-mail]
>
> ps. Thank you for kermit, it has been a great terminal emulator and it is
> still used a lot here !
>
Thanks for saying so.

The situation with the 7-bit Finnish character set is somewhat confusing.

First of all, let us note that, in the ISO International Register of Coded
Characters Sets, there is a set called "NATS, Primary Set for Finland and
Sweden" (accompanied by a secondary set), ISO Registration Numbers 8-1 and
8-2, from the Scandinavian Newspaper Technical Cooperation Council.

These are the only 7-bit character sets for Finnish listed in the Registry.
These are not general-purpose character sets; they are only for use by
newspapers, and Kermit does not support them (nor has anybody ever suggested
that Kermit should support them).  They contain such characters as "Unit
Space A", "Unit Space B", "Solid" (slug), and "Short Dash", in place of the
letters and punctuation that most users prefer to have.

The 7-bit character-set that most people think of when they say "Finnish",
and the on used by all versions of Kermit when told to "set terminal (or
file) character-set finnish", is the Finnish National Replacement
Character-Set (NRC) used by the DEC VT220 and later terminals, which is
listed in the manuals for these terminals, and in "Using C-Kermit", 2nd
Edition, page 559.  In this set, the character in Column 5 / Row 14 (decimal
94) is Capital Letter U with Diaeresis (Umlaut).

When you say "up-arrow", you mean circumflex, right?  This is, indeed, the
character that is replaced by U-Diaeresis in this character-set.

Of course, this is the disadvantage of 7-bit national replacement character
sets: they replace characters from ASCII such as brackets, braces,
backslash, vertical bar, and sometimes also number-sign, at-sign,
circumflex, underscore, and accent grave, and sometimes people do not like
to give up certain of these characters, as in your case, the circumflex.

This is why the 8-bit ISO Latin Alphabets were developed.  (Of course these
have limitations of their own, and that is why Unicode and ISO 10646 were
developed).

When Kermit 95 was released, some people were upset that it did not include
the "set translate input" feature from MS-DOS Kermit.  This was a deliberate
decision, since Kermit 95 fully implements ISO 2022 character-set
designation and invocation, which allows many character sets to be used on
the same screen.  Since the host can switch the terminal among four (or more
character sets) on a per-character basis, the "set translate input" command
(which assumes there is only *one* character set) has no meaning in this
context.

The real problem is that MS-DOS Kermit's "set translate input" command
enabled (and encouraged) the creation of new, ad-hoc, nonstandard character
sets, to be used in lieu of standard character sets and proper ISO 2022
rules, e.g. for mixing accented letters and box-drawing characters.  This is
why it was such a bad mistake, and why it is so important not to repeat it.

MS-DOS Kermit and Kermit 95 both support a wide range of standard (and
corporate) terminal character sets that are also supported by real
terminals, plus the standards to mix them on the same screen.

It appears to me that your application is using a nonstandard character set,
and should be modified to use the regular Finnish NRC or, better, an 8-bit
ISO Latin Alphabet, in which you do not have to choose between accented
letters and brackets, braces, and circumflex.

However, it is possible that there is some Finnish or corporate standard I
do not know about, which defines the character set you are using.  In that
case, please let me know what it is.

- Frank