> On Aug 26, 2015:10:51 AM, at 10:51 AM, Kent Watsen <[email protected]>
> wrote:
>
>
>
> > Another approach is to not rely so heavily on one giant uber-tree
> > that MUST be correct on the first try and never change.
>
> I agree that an uber tree stands to make things worse.
This seems to be the case the more we think through this.
> Distinct modules have distinct namespaces and no collisions concerns. But
> even better than that, distinct modules promote competition.
Or simply multiple versions of the same modules at the same time. ODL
lets you do this, for example, and happily works. But the other more
IETF-related situation is as you say, if there are two draft models for the
same features. One could and should
be able to run them together, for at least experimental purposes.
> I have no issues with there existing modules with overlapping concerns, even
> when implemented simultaneously by the same server. I'm projecting, but it
> seems that the uber tree approach would put a freeze to such experimentation,
> which is fine for a specific project to hoist onto itself, but seems
> inappropriate for a standards organization. Again, I like the idea of
> relocatable modules, as it seems to allow coexistence of both options.
I agree.
Tom (as individual).
>
> Kent // as a contributor
>
>
>
_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg
