From: Seth Breidbart (sethb@panix.com)
Date: Thu Apr 01 2004 - 21:37:34 CST
John Stanley <stanley@peak.org> wrote:
> Since the Subject header is not defined to carry such information in a
> netnews environment, software SHOULD NOT attempt to interpret it. It's
> defined to be "unstructured", that's how software should interpret it.
You weren't the first to bring this up (at least not in the quoted
message), but I'm finally sufficiently annoyed to pick on the point.
What do you mean by "interpret"? I don't know of any software that
attempts to treat a header as high-level code and execute it without
compiling.
Next, in two consecutive sentences you first state that software
SHOULD NOT attempt to interpret something, then state *how* software
should interpret it. I see an inherent contradiction.
There's nothing inherently wrong with software displaying things based
on some function of a header, even an "unstructured" one; for
instance, shrinking the font size of a long header to make it fit
seems like a reasonable thing to do (sometimes). Likewise, coloring a
Subject header based on a _guess_ as to whether or not the article is
spam is a reasonable thing to do. (Some users might want that, some
might not want that; but there's nothing wrong with software that does
it, is there?) Given that, how can changing the order of disply of
articles based on the Subject header be _wrong_? It may be desirable
to some and undesirable to others; that's fine.
The major effect of deprecating the use of "Re: " will be to
_increase_ the use of other such prefixes. That's a bad result.
Seth