From: greg andruk (gja@meowing.net)
Date: Wed Oct 02 2002 - 21:43:57 CDT
Charles Lindsey wrote:
> In <3D99F9A2.5070200@meowing.net> greg andruk <gja@meowing.net> writes:
>>Charles Lindsey wrote:
>>>In <3D951384.3010102@alex.blilly.com> Bruce Lilly <blilly@erols.com> writes:
>>>
>>>>Considering that newsgroup components are really tags that
>>>>identify a part of the path from the root of the newsgroup
>>>>hierarchy, there is a possible approach that avoids the
>>>>various problems with the alternatives discussed to date. I
>>>>envision something like the following:
>>[something not too far from what was introduced back in 1997 or so]
>>>So you want to introduce a protocol that does not exist anywhere yet,
>>It's documented, exists on the Microsoft and Sun news server products,
>>and is trivial to add to most servers.
> I think you had better explain what you are talking about.
At this point I'm sure it's all just pissing in the wind, but I'll
summarize, if only as a reminder that there are ways to do things that
don't involve breaking backward compatibility with anything.
A copy of the draft implemented by Microsoft and Netscape (now Sun) is
archived at <http://www.karlsruhe.org/rfc/draft-hernacki-nntplist-02.txt>
Advantages:
- 100% backward compatible, existing clients and servers can
simply ignore it.
- Charset religious issues are left up to hierarchy maintainers
and server operators.
- Simple implementation, really no harder than supporting
LIST NEWSGROUPS in NNTP (the mechanism is virtually identical,
with the only major change being that B/Q-encoding is allowed).
- It's been working just fine on corporate intranets for years.
Disadvantages:
- No currrent support in newgroup controls (there was still a
lot of back-and-forth over how to do application/news-groupinfo
at the time).
- No current support in the NNTP standard (again, waiting
on a base document with an extension mechanism).
- Not invented here.
I know that there have been concerns about confusing the two kinds of
newsgroup names, but in real life this doesn't turn out to be any harder
than understanding the difference between "chl@clw.cs.man.ac.uk" and
"Charles Lindsey." One is a mishmash of computerese, the other is an
actual human readable name.