Reflections on For Your Information

As expected, my little throwaway protocol sees even less use than my previous design.  I nonetheless
find delight in the principles which had guided it, and choose to document some observations herein.

It generally behooves a protocol to be case-insensitive wherever possible, and I strongly considered
this for FYI; I only decided against it when the ugly face of Unicode and its bastard encodings came
into view.  The appeal of case-insensitivity couldn't compare to avoiding any interpretation of tags
whatsoever, in any case.  This enables tags to be a serialization of any data, not merely textual in
nature.  I later realized my sh client's interface enforces a minor restriction on tags, from how it
collects the tag to send, but this could be removed easily by a minor tweak, unnecessary regardless.

In computing, meaninglessness, even opaque, is usually preferable to the slightest symbolism at all.
This is true since meaning should approach at the higher rather than lower layers whenever possible.

A corollary to my design for the negative acknowledgment, the empty packet, is to consider the empty
tag designating an event of nothing to have happened always.  This is a particularly nice corollary,
and the one default meaning in the system, practically unavoidable, reminiscent of NIL within lists.

I'm pleased with how easy it is to treat the lack of any response as a negative acknowledgment in my
design, although this characteristic comes from most anything designed to give but one confirmation.
It's disgusting to me how so many systems could send a single response packet to serve their purpose
and instead establish a connection or few to send so many tens of thousands of packets in its place.

The utter complexity of implementation and computation caused by using RSS for such a simple task is
self-evident.  That last time I reflected on a throwaway protocol I'd made, although I'd yet to call
them by the name, I lightly remarked on an IETF RFC intended for the same task; said design was many
orders of magnitude more complicated, and yet perhaps only linearly more flexible.  Said design sees
no use in practice, however, and nowadays the IETF designs are even more complicated than they were.
It's a heavy error to disregard the sediment underneath supposedly-simple solutions.  The absolutely
shitty designs of the many underlying layers merely makes a questionable matter intolerable instead.

Advantages of bespoke simplicity are severely underestimated in present times for nefarious reasons.
.