[Info-vax] Open Source on OpenVMS - A Progress Report

[MetaEd]

I wrote:

> if the message can be represented with Latin-1 (ISO-8859-1) codes, Google Groups does so
[snip]
> But if the message cannot be represented with Latin-1 codes, Google Groups uses UTF-8 codes

Also, if the message cannot be represented with Latin-1 but can be
represented with Windows-1252, Google Groups will prefer that to
UTF-8. This is another nod to newsreaders that lack MIME support,
because Windows-1252 is another extended ASCII.

Maybe the euro symbol is the one most likely to show up in this
newsgroup that would throw the message into Windows-1252.

Phillip makes the point that it can be practical, as an alternative to
MIME, to send messages without encoding, because the network has been
8-bit cleansed.

For older newsreader software, I agree this causes no more difficulty
in comprehension, provided that the endpoints both use an extended
ASCII.

But please do not make it impossible for people with modern tools to
receive the message correctly and display it using the same characters
used at the origin. Using 8-bit code as an alternative to MIME is
still encoding but without indicating what code was used. If the MIME
headers were present, the receiver could recognize the code and
display the message correctly. Without them, it can only guess. Which
extended ASCII is it? Latin-1? Windows-1252? ISO-8859-15? DEC-MCS?

If you are going to use 8-bit code, use MIME headers to tell the
receiver what you used:

    Mime-Version: 1.0
    Content-Type: text/plain; charset=ISO-8859-1 (or DEC-MCS, whatever
you used*)
    Content-Transfer-Encoding: 8bit

But good luck talking Google into 8bit encoding for Usenet. It
requires 8-bit transport, and 8-bit transport violates standard. If it
were me, I would make the same choice. Violating standard is not to be
done lightly, especially when there are alternatives.

Practically speaking, It's all well and good to argue that Usenet is 8-
bit clean. But there are obviously still a few systems without MIME,
so why bet against there still being standards compliant 7-bit news
systems.

I wonder if anybody is still working on Son of 1036.

* see http://www.iana.org/assignments/character-sets