Re: New dhclient broke multiple domains in domain-name

On Thu, Jun 16, 2005 at 04:44:40PM -0400, Garrett Wollman wrote:
> <<On Thu, 16 Jun 2005 13:16:46 -0700, Brooks Davis <brooks_at_one-eyed-alien.net> said:
> 
> > The RFC is shockingly lacking in this area.  It's rather odd that they
> > send a value for domain, but not for search (in resolv.conf).  It's not
> > suprising that people ended up abusing this to set search.
> 
> There is a standard (unless it didn't make it out of I-D) for doing
> this.  The shocking thing is that isc-dhcp(d) has never supported it.
> I used to have something like this in my dhcpd.conf:
> 
> #option domain-search-order code 119 = string;
> #
> # This was generated using the following command:
> #       perl -e 'print "\3lcs\3mit\3edu\0\2ai\xc0\4\xc0\4\2w3\3org\0"' | hd
> # ...and represents the search-list lcs.mit.edu, ai.mit.edu, mit.edu, w3.org
> # in DNS compressed encoding.
> #
> #option domain-search-order 03:6c:63:73:03:6d:69:74:03:65:64:75:00:02:61:69:c0:0
> 4:c0:04:02:77:33:03:6f:72:67:00;
> 
> AFAIK not even Microsoft clients bother to implement this.

It doesn't look like this made it out of the working group, at least I
can't find anything like it in RFC2132.

-- Brooks

-- 
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529  9BF0 5D8E 8BE9 F238 1AD4