Excerpts from TODO: 13-May-92 Re: Second Draft of "Implic.. Einar
Stefferud(_at_)nma(_dot_)com (2589)
Now, if you want to back up and define these mutant transformer things
as gateways, and write gateway standards, I will go along with the
effort to develop the gateway standards.
Hmm. Care to offer a definition of "gateway", for this purpose?
I think that the value of these transformations is so large that there
will be people doing it no matter WHAT we say. I'd like to offer them
as much guidance as possible. I tend to agree that this is basically a
gatewaying phenomenon -- I've often likened it to ASCII-EBCDIC
translations, which are also a gateway function -- but I do think that
this will increase the number of sites that want to think of themselves
as gateways. (For example, when I was at CMU, we had a very centralized
connection point between the large "andrew.cmu.edu" community and the
rest of the world, but we thought of our delivery software as "just an
MTA". However, in that community there would be a strong argument to
be made for some of the kinds of translation described in my document,
because it is a community of thousands of people with a single (huge)
shared file system, which changes a lot of the tradeoffs in terms of
external vs. included, GIF vs JPEG, and so on. If I were still working
there, I might argue that this is what we want to do, so now we are a
gateway. That's fine, but what does it mean? It would be nice if we
could offer some guidance. Just saying "now you are a gateway" doesn't
really mean a lot....