On Saturday, February 1, 2014, Hector Santos wrote:
Hi Chris,
How does the IETF feel about having "multiple standards" of the same thing, in other words, redundancy?
How many? It depends on the use case, so for one use case the IETF SHOULD have a one decision proposal.
I don't think we generally want that, and I don't think we want "bad" ideas becoming standards because it becomes harder and costly to undue.
There may be bad ideas backed by experts and evaluated good just because presented by experts or known persons.
However, we do need to document them too as well to allow people to try them out. So if the authors don't wish to make it, EXPERIMENTAL, BCP or INFORMATIONAL, a better option is to further it as a Proposed Method rather than get bounce around, wasting so much time as a Proposed Standard where Rough Consensus decisions is more problematic than ever before.
IMHO making it start as method directly may be allowed only in the general area. For other areas any new method should start as informational and passed then updated to method proposal. Any informational method need an approved standard pointed to.
Keep in mind there is also the "politics" part of it too of getting things pushed and supported. Once a proposed standard starts, you might as well "forget" about other possible solutions. They might not be allowed. However, I am seen them also used to weaken and even destroy PS efforts - raises doubt, etc, which is good when it shouldn't had been a PS in the first place.
We SHOULD solve the politics problem in IETF because it is a BAD behaviour as you mentioned its influences. The general area should solve that by new BCPs.
A PM would be easier I feel to pursue with less "politics," IMO. A lesser peer review process. Experimental and informational are too weak.
IMHO, They are not weak, that is part of politics may be as well.
If we took "Best" out of BCP, then perhaps a CP would be closer to a PM or Internet Method (IM). It just isn't the only, close-ended standard.
Agree but it needs some new process procedure.
Perhaps it should be made clearer that a BCP or Proposed Standard is not absolute in the IETF as far as solutions are concern.
Disagree.
The suggestion is to provide a way to document strong multiple solutions which may not necessary the "standard" or "best" solution.
You may mean strong possible multiple solutions related to possible applicability statements.
no solution method proposed only if it is related to an IETF standard or BCP. Each method should need informational document related passed the updated to method.
Personally, I would write more documents, get involved more with PM documents. PS are too hard. Too stressful and time consuming and the others, like BCP is just the wrong category. Maybe take away the B from CP and we got it. :)
Yes agree,
Thanks for your feedback.
--
HLS
On 1/29/2014 5:29 AM, Dearlove, Christopher (UK) wrote:
I'm not sure more categories is what we want.
I have observed all of the following (not at the same time, though some were) from people not familiar with the IETF and RFCs, although invested in or using RFCs:
- Assuming that all RFCs are standards.
- Assuming that only Standards matter, that Proposed Standards and Draft standards aren't important.
- Assuming that Draft Standard is a category below Proposed Standard.
- Recognising that there are standards and not-standards, but not distinguishing in any way between Standard, Draft Standard and Proposed Standard (calling all of them Standards).
- Just plain confused.
(And the concept of BCP is probably completely unknown to anyone thinking any of the above.)
Another category isn't going to help here.