[AFS3-std] Re: AFS and 'afs' URI scheme

[Mykyta Yevstifeyev]

Andrew,

2011/3/31, Andrew Deason <adeason@sinenomine.net>:
> On Thu, 31 Mar 2011 18:14:25 +0300
> Mykyta Yevstifeyev <evnikita2@gmail.com> wrote:
>
>> > Does moving the scheme to Historical impact our ability to use it or
>> > provide standardization on it in the future? As far as I know, nothing
>> > uses it right now, but (just speaking for myself) I am significantly
>> > less sure that it will continue to be unused in the future.
>>
>> Moving the scheme to Historical category does not restrict its usage,
>> but discourages it.  See http://tools.ietf.org/html/rfc4395#section-4
>>
>> This document also says nothing about specifying schemes currently
>> listed as Historical.  But there is an effort to revise RFC 4395
>> currently occurring in the IETF.  The Working Group doing this work on
>> the meeting right yesterday agreed that such action will be impossible
>> or strongly discouraged.
>
> Okay, then I would vote for keeping the afs URI scheme in its current
> status of provisional, and reserved for future standardization. I know
> it's been that way for quite some time, but if it makes it more
> palatable, we could probably come up with a more proper provisional URI
> submission without too much difficulty, given some time. Would the
> involved WG find it helpful if we did that?
>
The current WG does not work on particular schemes definitions.  It is
chartered to update procedures for their regsitrations.

However, the proposed updates to RFC 4395 says that Provisional status
is appropriate for schemes known to be used "in the wild".  IMO 'afs'
URIs is definitely this case.
>
> Adhering to the requirements of a permanent URI specification I expect
> will take much longer. I'm not entirely clear on how much the URI
> specification would need to involve the AFS protocol itself; there exist
> no published standards for a lot of the AFS protocol ("almost all of it"
> possibly depending on who you ask), so I'm not sure to what degree that
> makes this more difficult.
>
In order to give the scheme even Provisional regsitration, we should
decide on uniformed syntax, first of all.  Since there is no
standartized specification of AFS protocol, it will be almost
impossible.  So "as is" variant seems to suit current opinion of AFS
comunity.

Mykyta Yevstifeyev
>
> --
> Andrew Deason
> adeason@sinenomine.net
>
> _______________________________________________
> AFS3-standardization mailing list
> AFS3-standardization@openafs.org
> http://lists.openafs.org/mailman/listinfo/afs3-standardization
>