wide streams in EC++

["P.J. Plauger" <pjp@dinkumware.com>]

Date: Sun, 27 Sep 1998 05:38:38 +0200
To: c++-embedded@cygnus.com
From: bonnardv@pratique.fr (Valentin Bonnard)
Subject: wide streams in EC++

>[pjp]  Bet you didn't know that a fully conforming Standard C++ Library defines
>*four* wide-stream objects in <iostream>,

Well, I know that, but I also know that you weren't talking to me.

[pjp]  I thought I was.

>along with the four better-known
>char-based stream objects. And given the bizarre initialization requirements
>for these eight objects, it's rather difficult to eliminate all code for these objects
>*even if they're not used in the program.*

Well, it's certainly possible for the compiler to deal with that but
I trust you that w/o any compiler support it might be quite difficult
for the library writer.

[pjp]  It is.

 Why didn't you relaxed the 'bizarre' initialisation rules instead
of removing wios and all (ok, I know, you didn't _removed_ anything,
you didn't made wios mandatory) ?

[pjp]  Because it wasn't for me to decide. I implemented a specification
worked out by the Embedded C++ Technical Committee, over many
months and after considerable thought. If it were left to each of us to
develop the subset we thought best, we would have mayhem, not a
coherent specification that appears to meet many needs.

It seems to me that:
- many (not all of course) embeded devices can be/are localised
- wchar_t makes it possible or at least easier to write portable
  localised applications

[pjp]  That may be, but people with more experience on this topic than
I decided to omit nearly all wchar_t support from the EC++ library.

P.J. Plauger