From: Keith Moore <moore(_at_)cs(_dot_)utk(_dot_)edu>
Is that concern consistent with the first 250 or 500 RFC's? Weren't
the first RFC's more like current Internet-Drafts?
it's not a valid analogy, because the ARPAnet community was much smaller
...
nowdays there is so much going on (and so many half-baked ideas) that ...
Just as in the free source community, it is better to let the marketplace
filter and sort than to trust any central body of self-appointed,
professional commissars. (I'm still not talking about the IAB, AD's or
IESG). Besides, people keep telling me that the only RFCs that matter
are those on the standards track, and no one is talking reducing the
filtering on them.
But that missed part of my point. Consider the RFC's below 500 that are
being gradually exhumed. Why do the authors of the last 10 years of IDs
merit more shielding than the authors of the previous 10 or 12? If an ID
archive is evil, then shouldn't the newly found, earlist RFC's be
immediately deleted from ietf.org and isi.edu? (And why are there
conflicts between those two on the early RFCs?)
I've seen only these objections to the I-D archive:
- it breaks a promise to authors to hide their mistakes after 6 months.
That might be true, but it's also a load of rubbish to imply that
any author worth reading honestly cares and/or believes that their
mistakes are hidden by the I-D expiration.
Me thinks those who really are worried about their old I-D's being
exhumed misunderstand the importance of their words to the cosmos.
- it's illegal
the words in 2026 say what they say, not what some wish they said.
Copyrights work as they do, not as some wish or claim.
- the bad ideas buried in the expired I-Ds would drown and poison the
world if released. However, the archive doesn't have to be as easily
downloaded as ftp.isi.edu:internet-drafts/*. The wailing about search
engines indexing dangerously wrong ancient IDs are red herrings. If
you don't trust robots.txt, then use a bureaucratic, human mother-may-I
shield. It's incredibly silly to claim that those who can't recognize
bad ideas will search old ID's for fodder. If such people could be
bothered to search an ID archive, they wouldn't be sources of noise!
The real objection hasn't been adverted. It is a matter of control. Some
old ....people close to retirement are hurt by the thought of losing
control that they know they never really had on words they've forgotten.
Also, it's hard for me to to believe that any of those arguing for any
position on this subject are so shy that they would be deterred by the
possibility that their words in an I-D will be slightly more easily found
than their far less carefully chosen words in mailing lists.
I'll concede that point. but we are talking about making I-Ds more
visible/accessible than mailing list archives.
I did say "slightly more easily".
Anyone who thinks that finding things among 10,000 ID's would be a lot
easier than finding the related words in the relevant WG mailing list
archives has not done much searching in ID's or RFC's.
And again, who says the archive needs to be wide open? I think it should
be wide open, but if it makes the avowedly timid feel better to have it
protected, then fine!
On the other hand, shielding I-D's from public criticism is part and parcel
of the steady evolution of the IETF into a professional standards body
criticizing I-Ds after they have expired hardly seems like something
that we want to encourage.
The greater danger lies in trying to discourage or prevent criticisms of
expired I-Ds.
I'm embarrassed by those who profess to shamed by their old IDs. If they
were so terrible, why did they inflict them us? If they were so terrible,
then there is a need to exhume them and get some humble pie eaten.
In fact, more people have commented about the I-D archive in the last
week or two then would use it in the next several years.
So, a modest proposal:
- Eliminate Internet-Drafts.
- Give RFC numbers to all submissions to the IESG. Any submissions
that don't make the standards track are marked Informational.
that would decrease the signal-to-noise ratio. we need to increase it.
The signal/noise ratio would be unchanged, because there would be either
the same number of documents (as I see ir) or fewer because all of those
shy expired I-D authors would stop making noise that might be criticized.
According to IETF dogma, the real work of the IETF is in the drafts, not
the finished RFC's. Whether you call those documents I-D's or RFC's, they
must be read, understood, and criticized.
You can't increase the signal/noise ratio with bureaucratic declarations
and procedural filters, as plenty of current drafts and RFCs demonstrate.
Vernon Schryver vjs(_at_)rhyolite(_dot_)com