On 10/8/2020 11:58 PM, Brandon Long wrote:
So, I don't have extensive usage of these other systems or ones beyond
these, but the simple spec seems to satisfy
the majority of the examples. Hopefully others will speak up if other
systems do more than this.
Thanks for the review of existing practice. It's good to have something
like that in the list archive and even better to have it discussed.
There are a number of directions extensibility could take, here. Each
one will have some appeal. Maybe a lot.
The danger in any exercise like this is trying to guess which
capabilities are best to include or -- much worse -- trying to include
all of them -- or at least, someone's guess at what all of them are
(is?) Possibly demonstrating a bit of cultural bias, on my part, I tend
to call that the OSI approach...
However there is quite a bit of history suggesting that it is best to
start with something that is distinctly simple -- and even simplistic --
at first, to make it easy to develop and deploy, getting basic field
experience. Typically, that means starting with the smallest capability
that is deemed useful.
Best of all is to start with something that is already established,
while leaving room for /future/ extensions, based on actual demand. For
the current draft, I tossed in a simple, non-normative rule with a set
of emojis that already have an established base of experience.
If this gets deployed and used, and there is a community outcry -- or at
least a polite, cohesive request -- for one particular kind of
enhancement or another, that enhancement is likely to be easy.
ietf-822 mailing list