Michael Kukat writes:

>But we are talking about the ksh builtin echo command, and the echo(1) was
>referenced by the ksh manpage, so this is the syntax for the builtin echo of

You're right of course in this special case, since we're speaking about
pdksh on NetBSD, which doesn't seem to attempt to emulate the system's
version of echo(1).  Here's what I get with the real ksh (1993-12-28 k+)
on a FreeBSD system, where ksh's echo seems to use the BSD style of things:

$ echo '\0100\c'
\0100\c
$ print '\0100\c'
@$ 

>ksh. And my tests also were in ksh. If one sais "ksh", the OS and style of
>executable doesn't matter, as long as you use the builtins.

No, the builtins often cannot be relied up either, unless it's a builtin
such as "print", which is not available as an external command and as
such needs not retain any backwards compatibility.
A shell at compilation often configures itself to use the system's style
of doing things (so that it doesn't break scripts, isn't Unix a mess...)

--mkb