Discussion:
XHTML in Usenet
(too old to reply)
Ivan Shmakov
2012-03-29 04:20:22 UTC
Permalink
[Cross-posting and setting Followup-To: news:news.misc and
news:comp.infosystems.www.misc.]
Test
Though generally frowned upon by the old time Usenet users, I
honestly believe that XHTML is the future of Usenet
Valid XHTML 1.0 Strict
I see, it didn't propagate to Google Groups.
It propagates just fine to giganews.
ACK, thanks. I also see it on Aioe, and, of course, on Eternal
September, where it was originally posted.
Looks nice.
As to the appearance, the XHTML contained within the message in
question didn't request a specific one. One of the most common
issues I've seen with "default" rendering of XHTML (and HTML)
files is that the renderer puts as many characters on a single
line as could fit the window width (given the font used.)
Especially on the now-popular 16:9 aspect ratio displays, it
becomes a complete disaster, with some 100 characters on line or
so. (Instead of the "traditional" 66 or so, IIRC, in
typesetting, or less than 80 for computer's "plain text.")

This is trivial to solve with CSS (like I did for, e. g.,
http://gray . am-1 . org/lhc/), but the problem is that these
new formatting facilities could easily be abused.

I guess, the only way to avoid such an abuse is to allow only a
subset (or "profile") of XHTML in Usenet. Also, it could help
to avoid various security issues, associated with XHTML.
I think we've come a long way since 300 baud modems and plain text.
Another issue with XHTML is that it's unreasonably hard to edit
in /both/ "visual" and "source" forms. It's possible to use,
e. g., Markdown (or another simplified markup language) to
author an XHTML document (as I did for the one I've posted), but
it's still a problem to quote it.

One possible work-around is to translate XHTML back to Markdown
(or the like) for the purposes of quoting, though this approach
is likely to have issues on its own.
What are you using for a NNTP client?
Gnus/Emacs [1], but the version I use doesn't allow for an easy
way to deal with XHTML articles.

[1] http://gnus.org/
--
FSF associate member #7257
Ivan Shmakov
2012-04-10 16:22:13 UTC
Permalink
[Cross-posting and setting Followup-To: to news:news.misc, for
the detailed explanation doesn't fit alt.testing.]
[-- text/plain, KODIROWKA 7bit, KODIROWKA: KOI8-U, 9 STROK --]
Test
Though generally frowned upon by the old time Usenet users, I
sincerely believe that XHTML is the future of Usenet
[Valid XHTML 1.0 Strict]
gOSPODI, SPASI MENQ OT \TOGO BUDU]EGO.
Failed. (Assuming that the intent was to check whether the
Cyrillic script is handled properly. I've cleared the 8'th bit
here, so that this reply itself uses a strict ASCII subset.)

The message in question lacks any information on what the
encoding (or charset, in MIME parlance) is used whatsoever
(while the encoding itself is not ASCII.)

Among other problems, this may require the reader of your
message to "guess" the encoding at his side, which may in turn
be considered a bad netiquette. Also, such messages may pose a
problem to automated processing tools, such as, e. g., Usenet
archives and search engines. (Surprisingly, Google has guessed
the encoding correctly.)

The standard way to include the information on the encoding
being used to a Usenet message is to use MIME headers, like:

Content-Type: text/plain; charset=koi8-r

(Of course, using software with proper MIME support will resolve
the issue just as well, and may have other benefits beyond
that.)
--
FSF associate member #7257
Loading...