[LEAPSECS] stale leap second information

[Poul-Henning Kamp]

--------

>One question is that surely somebody else must have thought of this
>for conveying other sorts of non-DNS community metadata?  What
>lessons might be learned?

TXT records were popular for Rick-rolling at some point in the past.

The problem with doing a "service" like this is who you trust.

The reason I have been considering it at all, is to support my
Ntimed programs.  If you trust my source code you probably also
trust my DNS-skillz.

But would I point my source-code at some random persons service,
even if it were Tony or Tom ?  And would my users share my belief
that you can trust these fine gentlemen ?

I doubt it.

The ideal situation would be that IERS and the major timelabs agreed
to do this, so that a DNS lookup of leapsecond.iers.org,
leapsecond.nist.gov, leapsecond.ptb.de and so an all did the same
thing, and people could pick who they trust.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.