From: greg andruk (gja@meowing.net)
Date: Wed Oct 09 2002 - 23:49:31 CDT
Charles Lindsey wrote:
> Yes, when you introduce a brand new feature into a protocol, you do not
> expect it to interoperate with the existing base.
If it's been done right, a new feature does interoperate, providing
graceful degradation. That involves reducing functionality for the
older components rather than outright breaking it.
DNS can offer lots of things the old hosts files could't do, but hosts
and literal addresses still work fine for those who want to use them.
CSS offers all sorts of neato formatting features for browsers that
understand CSS, but an older or simpler browser can still make use of
the same Web pages, with reduced functionality. Was the "best when
viewed with browser XXX at YYY x ZZZ resolution" phenomenon only a local
problem?
MIME works fine with crufty old mail servers, and primitive MUAs are
limited in what they can do about displaying messages using MIME, but
they can still send and receive the messages without changes. Would
breaking the entire installed base of sendmail to support arbitrary
8-bit headers be only a local problem?
XOVER lets newsreaders that know about it get threading information and
useful article sizes. Older clients written to RFC 977 can't get those
article sizes and have to do a lot more work to get threading
information, but they can still read and post to all the newsgroups.
Would changing the HEAD command to print XOVER contents been only a
local problem for the clients who weren't rewritten to use the new
version of the protocol?