On 18-jan-04, at 23:17, grenville armitage wrote:
Actually it's pretty much the same topic, as there needs to be a way
to
preserve drafts that are important in some way or another.
If it is important, it'll progress the work of some group in the
IETF and be archived as an RFC.
Really. What's the number for the GSE RFC again? Even current work such
as draft-ietf-idr-as4bytes-07.txt stays in draft limbo for years.
If it (the I-D) doesn't capture work
well enough to be archived as an RFC then it ought to fade from IETF
I-D storage.
I agree it would be nice if things worked this way but they don't.
If the authors (or someone else) feels strongly that there's
still material in their I-D of value, 'archiving' it on a personal
website
for google to find.
Again, this simply doesn't work well enough in practice. Very little
web content stays in the same place for many years, and for a variety
of reasons, some things can be incredibly hard to search for in search
engines.
One way to
do this would be to make all drafts that are worth preserving an RFC.
This means drastically lowering the standards for what can be
published
as an RFC.
Is the standard for Informational currently that onerous?
I guess I'll have to find out...
Another way to do it would be to simply archive all
drafts. I agree this has the unpleasant side effect that all those
drafts that become obsolete (or are so from inception) stay around
forever.
Yes, that's a major problem. Organizations need to clean out their
clutter on a regular basis just like individuals do.
This argument is bogus as long as mailing list archives for stuff like
draft announcements are kept.
Then perhaps we encourage WG members to post a summary of their non-RFC
ideas to the WG mailing list around the time when work is wrapping up.
The ideas will thus survive (to be found through Google) as long as
the mailing list archive is around.
Searching in such an archive is only possible if you know the search
terms in advance. For instance, the draft I mentioned earlier isn't
easily found when searching for "32 bit as number".