What does the DNS Client service actually do? [Wireless Internet]

Prev: BEWARE OF AT&T!
Next: Mis-using DNS for solving WWW browser problems.

From: Jonathan de Boyne Pollard on 30 Mar 2010 04:35

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<blockquote cite="mid:hopm7v$s9d$1(a)tioat.net" type="cite">
The strange thing is that almost everyone on the net recommends we
turn OFF the DNS Client (aka DNS Caching) services, [...]
</blockquote>
It's usually folk wisdom, that almost certainly has been passed on
from each to the other.  It's often not based upon knowledge of what
the DNS Client does or is.  (One person in your list there thinks that
the DNS Client is a DNS server, for example.)  It's instead based upon
an Animal-Farm-like simplistic notion of  "service enabled bad,
service disabled good", which is of course wrong.  A second person in
your list even blames the DNS Client service for the facts that xyr
several ISPs are not presenting the same views of the DNS namespace as
each other, and that sometimes DNS lookups produce (gasp!) answers that
say that a particular domain name doesn't exist.
Interestingly, one of the items in your list is someone posting this
piece of folk wisdom in a discussion forum and having it debunked by
other people.  As you note, what you'll find written by Microsoft
doesn't support this folk wisdom, either.  As M. Fekay says, Microsoft
is right about its own software, here.  There are some instances
where Microsoft gets things wrong about its own products, usually
resulting from the fact that it's a big company and in such companies
the people who write the user documentation are sometimes not the
people who develop the software, or from the fact that even people
within Microsoft aren't immune from believing Internet/WWW-garnered
erroneous received wisdom from time to time, but this particular
instance isn't one of them. 

The most noteworthy item in your list is the documentation for
Simple DNS Plus.  That's the only one that you present that actually
gives a sensible reason for not having the DNS Client service
enabled: namely that Simple DNS Plus is a fully-fledged <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FGA/dns-server-roles.html#ResolvingProxy">caching
resolving proxy DNS server</a>, and if one has one of those
locally, having the extra caching in the DNS Client service on the same
machine makes no sense.
Further sensible advice is the advice that you're ignoring, but that
you'll find equally widely disseminated: Don't use DNS for this task at
all.  It's the wrong tool.  The DNS is not a tool for meeting WWW
browsing customization needs. <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FGA/verisign-internet-coup.html">That's
a lesson that the world learned in 2003</a>.  Use an
advertisment-blocking HTTP proxy server;
use <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FGA/web-browser-auto-proxy-configuration.html">a
PAC
script</a>, or use one of the many WWW browser plug-ins that do
what you want to do.
There two lessons here:
<ul>
<li>
Folk wisdom is often based upon people using magic incantations
and not really understanding what their computers do.
</li>
<li>
Abusing the DNS to solve an HTTP problem is wrongheaded.
</li>
</ul>
</body>
</html>

From: DevilsPGD on 30 Mar 2010 12:29

In message <24v3r51l7cng7aih00joajrrtesa4i96sm(a)4ax.com> Joel
<Joel(a)NoSpam.com> was claimed to have wrote:

>Jonathan de Boyne Pollard <J.deBoynePollard-newsgroups(a)NTLWorld.COM> wrote:
>
>> There two lessons here:
>
> Well, the first lesson should be quoting some part of the original message
>so other knows what you are talking about.

Not to call you out on missing something obvious, but the previous
poster *did* quote, but it was done using indents rather than quote
marks.

From: DanS on 30 Mar 2010 12:42

DevilsPGD <Still-Just-A-Rat-In-A-Cage(a)crazyhat.net> wrote in
news:6m94r519vcqc5ngldtv7mh8cjms771p8gr(a)4ax.com:

> In message <24v3r51l7cng7aih00joajrrtesa4i96sm(a)4ax.com>
> Joel <Joel(a)NoSpam.com> was claimed to have wrote:
>
>>Jonathan de Boyne Pollard
>><J.deBoynePollard-newsgroups(a)NTLWorld.COM> wrote:
>>
>>> There two lessons here:
>>
>> Well, the first lesson should be quoting some part of
>> the original message
>>so other knows what you are talking about.
>
> Not to call you out on missing something obvious, but the
> previous poster *did* quote, but it was done using indents
> rather than quote marks.

It's kind of hard to see *anything* in the posters HTML.

Here's what I actually see on my screen......I didn't even
attempt to read it...............

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<blockquote cite="mid:hopm7v$s9d$1(a)tioat.net" type="cite">
The strange thing is that almost everyone on the net
recommends we turn OFF the DNS Client (aka DNS Caching)
services, [...]
</blockquote>
It's usually folk wisdom, that almost certainly has been
passed on from each to the other.  It's often not based
upon knowledge of what the DNS Client does or is.  (One
person in your list there thinks that the DNS Client is a DNS
server, for example.)  It's instead based upon an 
Animal-Farm-like simplistic notion of  "service
enabled bad, service disabled good", which is of course wrong.
  A second person in your list even blames the DNS Client
service for the facts that xyr several ISPs are not presenting
the same views of the DNS namespace as each other, and that
sometimes DNS lookups produce (gasp!) answers that
say that a particular domain name doesn't exist.
Interestingly, one of the items in your list is someone
posting this piece of folk wisdom in a discussion forum and
having it debunked by other people.  As you note, what
you'll find written by Microsoft doesn't support this folk
wisdom, either.  As M. Fekay says, Microsoft
is right about its own software, here.  There are
some instances where Microsoft gets things wrong about its own
products, usually resulting from the fact that it's a big
company and in such companies the people who write the user
documentation are sometimes not the people who develop the
software, or from the fact that even people within Microsoft
aren't immune from believing Internet/WWW-garnered erroneous
received wisdom from time to time, but this particular
instance isn't one of them. 

The most noteworthy item in your list is the documentation
for Simple DNS Plus.  That's the only one that you
present that actually gives a sensible reason for not
having the DNS Client service enabled: namely that Simple DNS
Plus is a fully-fledged <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FG
A/dns-server-roles.html#ResolvingProxy">caching
resolving proxy DNS server</a>, and if one has one of 
those locally, having the extra caching in the DNS Client
service on the same machine makes no sense.
Further sensible advice is the advice that you're ignoring,
but that you'll find equally widely disseminated: Don't use
DNS for this task at all.  It's the wrong tool.  The
DNS is not a tool for meeting WWW browsing customization
needs. <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FG
A/verisign-internet-coup.html">That's
a lesson that the world learned in 2003</a>.  Use an
advertisment-blocking HTTP proxy server;
use <a
href="http://homepage.ntlworld.com./jonathan.deboynepollard/FG
A/web-browser-auto-proxy-configuration.html">a
PAC
script</a>, or use one of the many WWW browser plug-ins that
do
what you want to do.
There two lessons here:
<ul>
<li>
Folk wisdom is often based upon people using magic
incantations
and not really understanding what their computers do.
</li>
<li>
Abusing the DNS to solve an HTTP problem is
wrongheaded.
</li>
</ul>
</body>
Attachment decoded: untitled-1.htm
</html>

From: Jonathan de Boyne Pollard on 30 Mar 2010 15:53

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<blockquote cite="mid:6m94r519vcqc5ngldtv7mh8cjms771p8gr(a)4ax.com"
type="cite">
Not to call you out on missing something obvious, but the
previous poster did quote, but it was done using indents
rather than quote marks. 

</blockquote>
Actually, it wasn't done with indents.  It was done with <code><blockquote></code>,
which is, as the name states, a quote.  The usage here was directly in
line with what the element is actually for.  Using it to
indent is in fact a mis-use, <a
href="http://w3.org/TR/html401/struct/text.html#h-9.2.2.1">deprecated
in the HTML 4.01 specification</a>. 

</body>
</html>

From: Mike Easter on 30 Mar 2010 19:20

posted to alt.internet.wireless only, where I read it; disrespecting
and disregarding the original crossposted group and also the f/up to a
another group I don't read

Jonathan de Boyne Pollard wrote:
Content-Type: text/html;

> Actually, it wasn't done with indents. It was done with |<blockquote>|,

You shouldn't be posting anything other than plaintext into general
usenet discussion groups; (and) your plaintext should be using the
accepted > quote marks with that plaintext (and) you should only post
html into groups where it is acceptable, which isn't very many.

And -- you should get a shorter handle (while you are at it, reforming).

--
Mike Easter
I don't post into
groups I don't read

First | Prev | Next | Last
Pages: 1 2 3
Prev: BEWARE OF AT&T!
Next: Mis-using DNS for solving WWW browser problems.