My point is that if we discuss about 'results of the experiment' and the IESG
plans to strongly encourage authors of Experimental RFCs to briefly describe
the results, they should also recommend to include in the RFCs in an explicit
manner the goals which the results of the experiments are to be measured
against.
This makes sense if you believe the Experimental RFCs are actually (with
respect to time, with respect to the people involved) coupled to any
experiments.
Otherwise it's just another layer of red tape, another hoop to jump through.
be limited to those actively involved with the experiment.
So the term 'experiment' already appears in 2026.
Yes (but note that this is about the "limited use" designation, not about
"experimental"). I've wondered about this, too, a long time ago when I started
to learn about the IETF and RFC 1310. More than once this has been explained
to me as a euphemism for "stuff should be used only by people that know the
stuff, know what they are doing and know what will break and how to fix this".
But of course I didn't record those hallway conversations and don't remember
who said that at the time.
Honestly, I would have no idea what I would have written into RFC 3940/3941,
except for "it is believed we haven't had enough deployment experience to make
this a PS, so please join us in the experiment of deploying this for the
applications where it is needed."
If that sentence is fine with the new procedure, I don't have a problem.
Of course, it would be easy to describe the result of the "experiment":
3940/3941 were upgraded to PS 5 and 4 years later, with some fixes learned from
the wider use that the Experimental status (and more time) brought with it.
But, again, I'm not sure what we'd have written if people had wanted us to
write up "the experiment" before going to PS. Maybe a list of fixes (section
10 of RFC 5740)?
I understand there may be grander things going on (like attempting to change
the way we use addresses in the Internet) that do justify jumping through those
hoops. But as a general rule, this procedure makes me unhappy. How about "The
IESG may ask for..."?
Grüße, Carsten