Re: [DNSOP] WGLC for draft-ietf-dnsop-let-localhost-be-localhost-02

[Åke Nordin <[e-mail]>]

On 2018-02-10 23:36, Viktor Dukhovni wrote:
> On Sat, Feb 10, 2018 at 08:21:14PM +0000, Warren Kumari wrote:
> 
>> <with no hats at all>
>> Interestingly enough, Steve Sheng and I wrote just such a document a
>> number of years ago (around the time of the initial name-collisions
>> drama). Even though I'm 95% sure it included the phrase "tilting at
>> windmills" my search foo fails me at the moment... but it basically
>> deprecated search list processing in the DNS.
>> There are many things which would be safer, less complex, and safer if
>> search lists didn't exist -- would people be interested in discussing
>> the idea, or is it just too out there?
>> </hats>
> 
> FWIW, I am having no success at all convincing anyone when grumbling
> about all the truncated names in URLs (that rely on search lists)
> at current place of work.  And, sadly, I don't think it is especially
> atypical...

Not atypical at all. While URL usage isn't much anylonger
where I try navigating chaos during the days, way too many
internal applications depend on short names resolving. At
least most storage mount points are specified using FQDNs
nowadays, but those names has always been controlled by
comparatively few technicians who should have known better
from the outset.

+1 to anything I can wave with when arguing against search
lists. People doesn't seem to believe me when I use
performance arguments against it.

-- 
Åke Nordin <[e-mail]>, resident Net/Lunix/telecom geek.
Netia Data AB, Stockholm SWEDEN [phone]