|
Re: Request for review of draft-yevstifeyev-genarea-historic-03
2011-03-03 09:42:18
No, I do not agree with you.
Our current definition for historic, and our current choices for when to
move things, have worked sufficiently well.
I have not seen any evidence that there is a problem that needs solving.
I have also not seen any evidence that the changes you propose to the
definitions would help anything.
Yes, there are problems with our current definitions.
Those problems serve to keep us from declaring something historic when
we shouldn't.
They do, as a side effect, make it harder to declare things historic
even if they actually are?
I will even agree that there is a small cost to that. Users of our
standards may sometimes think some things are relevant that aren't.
But changing the definitions to make it easier to change the status does
not actually change the problem. The problem is that it is hard for
anyone (the IETF, vendors, operators) to reliably know what is actually
in use.
Yours,
Joel
PS: I have actually noticed with several specifications I worked with
that it can easily take 5 or 10 years for the actual usage (often not
the initially expected usage) for a standards track or experimental
protocol to emerge. So premature declaration taht something is historic
can do actual damage.
On 3/3/2011 10:28 AM, Mykyta Yevstifeyev wrote:
Joel,
I agree with you that it is really hard and even impossible to determine
what is going on in the Internet regarding some technology, protocol,
etc. If we set the impossible criterion for the Historic documents, this
will really make very few sense. So, as I have been pointed out, I find
removing the regulation about 'out-of-use' technology at all quite
acceptable. Do you agree?
And as for the comment from Dave.
03.03.2011 17:18, Dave CROCKER wrote:
On 3/3/2011 7:11 AM, Joel M. Halpern wrote:
The first point, to echo Andrew Sullivan, is that even if a protocol
is in use
on the public Internet, it is not always easy to detect.
...
The second point is that enterprise uses and other private network
uses are
still valid uses. The fact that a protocol may be used only inside a
virtual
private network, or only inside a corporate data center, or in only
within a
military facility, does NOT mean that it is not used. Such limited
use is still
valid use and should not result in our declaring something obsolete.
+1
Declaration of historic needs to be based on affirmative data. The
declaration is actually only important to make for protocols that are
known to be problematic.
This is already covered by the 'deprecated' criteria in my draft.
All the best,
Mykyta Yevstifeyev
Issuing a declaration for mere non-use is a matter of convenience, not
need, IMO.
d/
03.03.2011 17:11, Joel M. Halpern wrote:
There are, in my opinion, two problems with Mr. Yevstifeyev's
assertion below that it is easy to determine when things are out of use.
The first point, to echo Andrew Sullivan, is that even if a protocol
is in use on the public Internet, it is not always easy to detect. It
may be used in only some portions of the net. It may be used inside
some other protocol that makies detection ahrder.
The second point is that enterprise uses and other private network
uses are still valid uses. The fact that a protocol may be used only
inside a virtual private network, or only inside a corporate data
center, or in only within a military facility, does NOT mean that it
is not used. Such limited use is still valid use and should not result
in our declaring something obsolete.
Yours,
Joel
On 3/3/2011 4:27 AM, Mykyta Yevstifeyev wrote:
Hello Eliot,
Thank you for reading the document. Pleas efind some comments in-line.
2011/3/2, Eliot Lear<lear(_at_)cisco(_dot_)com>:
...
I bring to your attention RFC-4450, in which we made a bulk status
change of a bunch of PS to Historic precisely because we couldn't find
anyone using those protocols. However, such observations are
imprecise. For one, it is hard to observe what is going on on the
Internet, and those who do don't usually share their data (there is
some, but it is often regionally based, like the GINORMOUS store at
ETHZ). Another issue is that a protocol that is not detectable on the
Internet might be in use on private networks.
When we say 'out-of-use' we consider the usage of something in the
overall Internet. It is mostly not very difficault to find this out
via those people who take part in the IETF.
...
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf
|
|