Re: RFC 1524

[Keith Moore <[e-mail]>]

> > I'd like to know if a revision of RFC 1524 is planned, in conjunction
> > with the current revision of MIME ?
> 
> I certainly hope not. mailcap files today are like HOSTS.TXT 15 years
> ago. They don't scale. The number of media type definitions floating
> around is growing exponentially, and we're going to have to do
> something better than a flat text file to keep track of them.

I certainly hope *this* doesn't last.  We don't need nearly so many 
media types, at least not in common use.  The number of entries in
the mailcap file is the least of the problems.  We still have to
have a plugin for every media type (and given the proliferation of
proprietary nondocumented media types, it's damn near impossible to 
do the security analysis of the ones we have already).

Which is not to say that I don't like the idea of being able to
put each media type definition in a separate file or directory,
and making them accessible by a URL.  

But once we do that people are going to want to put the plugins 
themselves in that directory, and download them as needed.   We'd 
better be ready with a really safe language with which to interpret 
the plugin (no, it doesn't exist yet), AND a standard way to digitally 
sign the plugin along with an indication of "how much rope" it should 
have and have that signature verified by the client, before it gets 
executed (and no, that facility doesn't exist yet either).

Flat mailcap files could be a blessing after all.

Keith