[Am-info] Microslop - beauty and elegance -- &nbsp

[Marcus de Geus]

In response to a message from "Erick Andrews" <eandrews@star.net>:

Erick,

> E-mail should be simply human readable.

I couldn't agree more. This is why the MIME Quoted-Printable encoding schem=
e was designed with readability in mind. If you can decode it, you get to s=
ee all the characters as they were intended to be seen. Read the encoded te=
xt, and you can still get a pretty good idea what it's all about, even if t=
he non-ASCII characters show up as metacodes -- and the metacodes, being hu=
man-readable, are easy to look up if you should be so inclined.

> I think it's extremely inconsiderate of people to send e-mail in HTML.

Very diplomatically put. Other words than inconsiderate come to mind.

> If someone needs to send special characters (or some markup language)
> beyond the national language ASCII code pages and character sets,
> ask first, or use MIME, Multipurpose Internet Mail Extensions.

Almost, but not quite. Change that to:

"If someone needs to send special characters beyond ASCII, they should use =
MIME. If anyone considers sending some markup language, > NUL".

Note that ASCII does not include any national language code pages or specia=
l character sets (i.e. IBM character sets). Pure ASCII (which is what we sh=
ould be dealing with, see below) is the 7-bit ASCII character set (00h-7Fh)=
, period.

RFC 2822 stipulates that an Internet e-mail message may contain only 7-bit =
(pure ASCII) code. Anything beyond that (i.e. 8-bit code -- usually represe=
nting diacritical characters -- which would otherwise be open to various in=
terpretations depending on the local code page regime) requires some encodi=
ng mechanism, e.g. MIME Quoted-Printable, which provides a highly robust en=
coding mechanism for non-ASCII characters.

Of course, most mail relays (Microsoft and other) these days consider anyth=
ing attached to them (including the Internet) their local domain, so they w=
ill insist on sending out 8-bit text messages, which may or may not get int=
erpreted correctly at the receiving end, depending on the local code page m=
echanism. But then, why bother with standards?

Marcus de Geus
-- 
marcus@degeus.com
www.degeus.com