Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
Content-Type: text/plain; charset=utf-8
From: Tim Bruijnzeels <tim@ripe.net>
In-Reply-To: <58E87CF4-2694-474F-BD48-DAB6E2BA78E4@zdns.cn>
Date: Tue, 9 Aug 2016 13:50:55 +0200
Content-Transfer-Encoding: quoted-printable
Message-Id: <2338C59A-F77B-4D7C-8B60-B6D577207F71@ripe.net>
References: <58E87CF4-2694-474F-BD48-DAB6E2BA78E4@zdns.cn>
To: Declan Ma <madi@zdns.cn>
Archived-At: <https://mailarchive.ietf.org/arch/msg/sidr/0WlU7adsPB1aowWT9_4sNiYEIEk>
Cc: sidr <sidr@ietf.org>
Subject: Re: [sidr] Time frame for changing RPSTIR to the new validation
 algorithm
Precedence: list

Hi,

I am still working on an update to the document.

It's encouraging to hear that another group is looking at maintaining =
RPSTIR. When I discussed the 6 months with Rob we felt that this should =
be doable (he can correct me if I am wrong). In our case we have had a =
working implementation of the algorithm for quite some time, and those =
changes by themselves are not hard. With the new OIDs we do need to =
change some things around - making this dependent on OIDs rather than a =
local configuration setting, but it's not a big deal.

I think it's reasonable that the ability to update implementations from =
both RP implementations and CA implementations should factor in to a =
mandatory to implement time frame for RPs, and later CAs. But I also =
believe that we should be somewhat ambitious, especially w.r.t. RPs =
because their implementation is on the critical path. CAs can only start =
to publish when RPs support this.

Regards
Tim


> On 08 Aug 2016, at 16:54, Declan Ma <madi@zdns.cn> wrote:
>=20
> Folks,
>=20
> I recall Tim has suggested a schedule for transition to the new =
validation algorithm during Berlin meeting. His proposal is to mandate =
RP support for the new all 6 months after the doc is published as an =
RFC.
>=20
> My team is now responsible for RPSTIR update.=20
>=20
> I am afraid that the proposal is sorta aggressive. IMHO, six-months =
after publication of the RFC may be too soon. Although this WG has been =
through discussions of the validation-reconsidered, but the conclusion =
was just reached.=20
>=20
> This is really a fundamental change to both RP and CA, I anticipate we =
need more time to have RP software code changed and thoroughly tested, =
to accommodate the new validation procedures.=20
>=20
> Anyway, we are evaluating the time frame,trying to give absolute dates =
as soon as possible.
>=20
>=20
> BTW, OID issue has not been yet with IANA feedbacks. We can=E2=80=99t =
be sure how long it will take to go through the IESG approval process, =
incurring uncertainty. I am not reassured whether it is prudent to get =
started with RPSTIR update right away.  =20
>=20
> I am therefore looking forwards to seeing more discussions and =
comments in this thread, especially from other RPKI implementers.
>=20
>=20
> Di
>=20
>=20
>=20
> _______________________________________________
> sidr mailing list
> sidr@ietf.org
> https://www.ietf.org/mailman/listinfo/sidr