On Fri, Jan 30, 2015 at 11:54 AM, Rob Stradling <[email protected]>
wrote:

> Thanks again, DKG.  Option 2 it is then.
>
> 2. "?." matching =1 redacted label.
>
>
> Do you think we should support redacting (for example) "
> public.secret.example.com" to "public.?.example.com" ?
>
> Or, to keep it simple, shall we disallow unredacted labels to the left of
> a redacted label?
>

>From an operations viewpoint I would say to disallow them:

* It will at best be rarely used, with few, if any, non-contrived use cases.

* It is begging humans or software to miss seeing/correctly processing the
unusual "?" buried in the middle of what is easily mistaken for a normal
FQDN.

* Additional (if only slightly) code complexity and testing will be
required but rarely used and unimportant to most customers. Often such low
priority features get poorly implemented/tested.

* The preceding two points suggest some implementations and deployments
could end up with vulnerabilities from a little known feature..

* (Remembering non-contiguous subnet masks) It potentially could prevent
some sort of optimization for the other 99+% of cases.

* An attacker (or mistake) could exploit the increased chance of human
misreading. Detecting/fixing this mistake may be difficult for those who
made it in the first place.


Without a really strong use case it seems the negatives easily win over the
minor increase in flexibility.

-Scott-
_______________________________________________
Trans mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/trans

Reply via email to