On 24/1/21 01:52, Joe Touch wrote:
On Jan 23, 2021, at 8:37 PM, Fernando Gont <fgont@xxxxxxxxxxxxxxx> wrote:
On 23/1/21 14:12, Joe Touch wrote:
On Jan 22, 2021, at 11:37 PM, Fernando Gont <fgont@xxxxxxxxxxxxxxx> wrote:
One *internet-draft* certainly doesn't undermine E2E. However, I guess
that an *RFC* published as a "Proposed Standard" probably does (undermine) E2E?
Not when it doesn’t update the hundreds of other standards that don’t.
It doesn't formally update them, true. But it is a de-facto update: behavior that goes against such other standards has been approved as PS.
There is no such thing as de facto updates.
defacto update = a document that is doing something that goes against
existing specs and documents. It fails to formally update existing
documents, as well as having existing "Update" tags -- still, it changes
the assumtions and behaviors of said specs.
(Do I like that? -- Of course not)
You’re confusing the “let a thousand flowers bloom” approach with a change in architecture.
"Let a thousand flowers bloom" would possible for *internet-drafts*,
*or* for standards that can exist without harming harming each other.
When the "mangling EHs on the fly" flower blooms, the IPv6 flower fades.
I could ask the question: can IPv6 routers remove EHs on the fly?
The IPv6 architecture implies that it doesn't. But there's a PS (not
just Info or Experimental) about that -- so yes, they do.
The whole point of having an architecture is that it guides the
development of specs. When it doesn't, the "architecture" is just words
on paper.
--
Fernando Gont
SI6 Networks
e-mail: fgont@xxxxxxxxxxxxxxx
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
