Tim:
>>> If I were configuring my DHCP server to hand it out to clients, that
>>> would be the following in the dhcpd.conf file:
>>>
>>>   option domain-name  "internal.";
>>>
>>> It's going by proper standards that a domain name ends with a dot.

& again, Tim:
>> That may not be needed, now.  But apparently was when I set things up
>> around 20 years ago, and still works fine that way.
>>
>> In network configurations, ending with a dot indicates that it *is* the
>> top of the chain, and nothing else should be appended to it.


Mike Wright:
> Thanks for pointing that out.  DNS is very simple in concept and very 
> complicated in implementation.  A lot of people just take for granted 
> that it works.  Mess it up and *nothing* works ;/
> 
> As to the dot.  DNS is uniform.  *Every* domain name, top-level or 
> otherwise, and every subdomain name, no matter how deep, is followed by 
> a dot.

It certainly is, as far as DNS is concerned.

I'm not sure how well DHCP follows the rules (both client and server),
I'm not even sure you have to pass a domain name along to the client. 
There's every chance that a client just takes the words sent to it, and
decides for itself what's a <hostname> <subdomain> <domain> <top-level-
domain> simply by counting the dots, rather than accept words into
certain parameters as described to it.

When I set the network up 2 decades ago I followed the rules and best
practice, to set up the network properly.  But when I look at the
dhcpd.conf man file, now, I see they've just done something like:

  option domain-name "example.com";

Now a client could easily make some assumptions about that, and
probably be correct.  But it could easily make some presumptions, and
get it wrong.  Particularly if you used something unusual.

Inside of a LAN, my PCs' hostname and domain could easily be
lounge.house and study.house and work quite fine for most things.  But
a few things won't like that...

If I ran an internal webserver on each PC, and wanted to browse to it
by name with Firefox, and if the FQDN doesn't meet Firefox's
preconceived notions it treats it as a Google (or other) search query
instead of trying to load it as an address.  Even if there are DNS
records for it.  It doesn't try loading it, then play guessing games. 
It just does its own thing straight off the bat.

The number of times I've tried to access my printer's configuration
server, and had to fight with Firefox to stop searching the internet
for a shop selling that model printer, is a right pain.  Even when I've
typed http://laserprinter/ as a fully complete *hostname* address with
protocol prefix and ending with a slash, it'll ignore that and do a
google search for a laserprinter.  Or it will change the http:// to
https:// (which won't work, because my printer only has a port 80
server).

So, yes, using proper domain names certainly helps (and Firefox has
some coding to recognise .internal as special).  I've also killed off
an awful lot of automation in Firefox so it stops making stupid
assumptions.

When working on computers and testing things, sometimes you have to
keep hammering away at the same address until it loads.  You certainly
don't need the browser changing the address on you all the damn time,
turning it into a keyboard fight.

-- 
 
uname -rsvp
Linux 3.10.0-1160.119.1.el7.x86_64 #1 SMP Tue Jun 4 14:43:51 UTC 2024 x86_64
 
Boilerplate:  All unexpected mail to my mailbox is automatically deleted.
I will only get to see the messages that are posted to the mailing list.
 

-- 
_______________________________________________
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue

Reply via email to