New top-level domain names are coming [Telecom]

The Internet Corporation for Assigned Names and Numbers (ICANN) has announced more progress toward the issuance of Generic Top-Level Domain name (gTLD) domains. These new top-level domains will give applicants the chance to have a corporate trademark or community name as the _last_ item in an Internet address: for example, instead of sears.com or fsf.org, the new gTLD program would allow addresses such as sales.sears or emacs.fsf.

Details are at

formatting link
.

I'm curious what others think about how, or if, this will affect telecom.

-- Bill Horne

(Filter QRM from my address for direct replies.)

Reply to
Bill Horne
Loading thread data ...

Or not, of course. Administrators will generally want to disable IDN on their own systems (although of course users are expected to want them) so that they can handle and respond to attacks coming from sites using foreign character repertoires.

-GAWollman

Reply to
Garrett Wollman

Telecom Digest Moderator wrote in :

Way back when I first learned about this Internet thing and Internet e-mail, using user@name with no dots in the name ment: the *local* mailserver with that name, so one would use that to send mail from userA@server1 to userB@server2 without having to type snipped-for-privacy@server1.subdomain.topleveldomain. A lot of software still assumes that names without dots can be affixed with the local domain name or search domain.

For example, when I try your ideal name in a search for a mailserver at home:

$ host -vt mx horne Trying "horne.idefix.net" Trying "horne.koos.koffie.dot" Trying "horne" Host horne not found: 3(NXDOMAIN) Received 98 bytes from 2001:888:1011::694#53 in 0 ms

Now when any toplevel domain can be in use and when people start to use e-mail addresses like you write your ideal address, software will be fixed to assume e-mail addresses are always fully qualified, even when there is no . in the righthandside. But expect interesting problems when this is first implemented.

On the other hand, spammers have broken e-mail enough that this will probably be just a minor problem.

Koos van den Hout

Reply to
Koos van den Hout

Just qualify the domain name with a trailing dot, e.g.

nslookup wildbill@horne.

Works OK will all my antique DNS implementations.

Michael

Reply to
Michael Grigoni

wrote in :

Hmmm, interesting. Looking just now at some old email from 29 years ago (1980), we had ARPANET email addresses rms@mit-ai, thad@sri-kl, etc.

It wasn't until later that real TLDs (.net, .mil, .edu, etc) appeared.

Precisely. I don't expect the transition to be without grief.

And what about domain squatters? Bill, better register "horne" ASAP. :-)

As I wrote in another post, I'm glad I'm retiring this year. :-)

Reply to
Thad Floryan

I thought the idea was people and corporations could create their own TLD at an enormous cost (6 figures). If I were to create mayson I would still need something in front of it. Perhaps snipped-for-privacy@mail.mayson.

Reply to
John Mayson

Yowee! Somehow I missed what the cost will be -- probably won't be too many domain squatters. :-)

Not really. Depends how the system(s) and internal routing is/are setup. Many of us don't require, for example, "www." to reach the webserver (or "ftp." for the ftp server, etc.) at our domains since the port(s) (80 for (most) HTTP) is/are known.

The fine folks at MIT had a neat idea years ago that somehow never took off: "web." instead of "www." -- two syllables ("web dot") instead of ten syllables ("double-you double-you double-you dot") and actually more explanatory and intuitive.

Reply to
Thad Floryan

They're asking for trouble. They assume that there is ONE trademark per name, and that's totally TOTALLY wrong.

For instance, many companies share the Johnson's trademark -- a foot soap company, a shampoo company, and a wax company, to name just three. All kinds of companies are like that. GE, NBC, Apple, Zenith, ABC, etc.

Reply to
David Kaye

Not to put too fine a point on it, I pronounce "www." in only five syllables: "tri-ple duh-bya dot". Still way more than just two, of course ... .

Cheers, -- tlvp -- Avant de repondre, jeter la poubelle, SVP

***** Moderator's Note *****

I don't think anyone worries about "www" anymore. Most browsers try adding the subdomain automagically if they get a 404 on just the domain name: if "billhorne.com" doesn't work, they'll try "

formatting link
" without the user needing to do anything.

Anyway, I just say "dub dub dub", and everyone knows what I mean.

Bill Horne Moderator

Reply to
tlvp

AFAIK "www" is just a DNS convention to indicate a web server - it can be anything you like and still work exactly the same.

It is the "http" that tells things to use the (default) Hypertext TransporT Protocol for the connection as a web page - and default to TCP Port 80 if there is no alternate port specified.

People get blinkered by convention in these things.

-- Regards, David.

David Clayton Melbourne, Victoria, Australia. Knowledge is a measure of how many answers you have, intelligence is a measure of how many questions you have.

***** Moderator's Note *****

"www" is a subdomain. At the start of the web, http servers were usually separate machines, and putting "www" in front of a domain name made it easy to divert web traffic to a different server than a company's regular email/gopher/finger/ machine.

In like manner, "ftp" is commonly used to point file-transer-protocol traffic to a separate server, mostly for security reasons: there's less chance of mixing private and public files if ftp requests go to a separate machine.

It's almost unimportant now: most companies with a web presence will either divert web (i.e., tcp port 80) traffic to their web servers at their firewall, or declare a CNAME in DNS that causes DNS to return the "web" address when someone asks for the domain without "www." in front of it (it's "supposed" to be the other way around, but c'est la vie). Since DNS isn't able to distinguish TCP port numbers, it's common to provide an MX record which will divert email traffic away from the "web" address.

Clear as mud, right? ;-)

Bill Horne Moderator

Reply to
David Clayton

Only rarely, if ever. Normally it's just a leaf.

That server being the one whose name was "www", or more fully, "

formatting link
".

You can't have a CNAME record at the apex of a zone.

With hindsight, it's clear that Tim B-L should have used SRV records. The only problem is that SRV records hadn't been invented yet. The earliest proposal for SRV that I can find, RFC 2052, is from 1996, when the Web was already seven years old; its "introductory example" starts with a URL. (And I don't fault Tim for not having invented SRV records himself; it would have taken at least that long for enough name servers to be replaced to make them usable.) SRV didn't become a Proposed Standard until 2000, long after the Web took off to the point that it was no longer feasible to make a wholesale change in the way URIs were interpreted.

-GAWollman

Reply to
Garrett Wollman

Granted I'm not an Internet expert. I've just used it for nearly 25 years. But the point I was trying to make was if I registered the top-level domain (TLD) "mayson" I think I would still need a hostname in an email address. john@mayson would be analogous to john@com. Granted if I controlled "mayson" I could say nothing existed below it, but I would think this would cause problems because mail servers wouldn't know what to do with it. Did I mean mayson.com? mayson.us? What if I was able to snag, say, "cnn" as a TLD (unlikely, but what if I did)? If someone accidently sent an email to wolf.blitzer@cnn would I get it?

John

Reply to
John Mayson

Why would you have to say nothing existed below it? You can have an MX record on "mayson", and an A record on "www.mayson". This is analogous to having an MX record on "mayson.us" and an A record on "

formatting link
".

Mail servers don't try to figure out what you mean. They just take the name after the "@", and look up the MX record to see where mail should be delivered (if there's no MX record they'll look for an A record).

Yes. But good luck winning the trademark dispute with CNN so you can keep the domain.

Reply to
Barry Margolin

You might want to ask the guy whose address has been n@ai for many years. (Yes, his name is Ian.) In some mail programs it works, in some it doesn't. The current spec says the domain is supposed to have at least two components, but the older versions were ambigious so feel free to pay your $185,000 (plus $100 registration fee) and take your chances.

R's, John

Reply to
John Levine

Are you absolutely sure that MX records are not required?

My understanding (and experience) is that they are mandatory, and the RFC seems to say so:

formatting link

-- Regards, David.

David Clayton Melbourne, Victoria, Australia. Knowledge is a measure of how many answers you have, intelligence is a measure of how many questions you have.

Reply to
David Clayton

N It is possible that the list of MXs in the response to the query will be empty. This is a special case. If the list is empty, mailers should treat it as if it contained one RR, an MX RR with a preference value of 0, and a host name of REMOTE. (I.e., REMOTE is its only MX). In addition, the mailer should do no further processing on the list, but should attempt to deliver the message to REMOTE.

That's also a pretty ancient RFC. RFC 2821 is the current SMTP RFC, it says more clearly:

If no MX records are found, but an A RR is found, the A RR is treated as if it was associated with an implicit MX RR, with a preference of 0, pointing to that host.

-- Barry Margolin, snipped-for-privacy@alum.mit.edu Arlington, MA

*** PLEASE post questions in newsgroups, not directly to me *** *** PLEASE don't copy me on replies, I'll read them in the group *** ***** Moderator's Note *****

ISTR seeing an "MX" query returned from a DNS server with the "A" record included: is this done to prevent a second DNS lookup if the MX is empty?

Bill Horne Moderator

Reply to
Barry Margolin
+--------------- | Barry Margolin wrote: | > Mail servers don't try to figure out what you mean. They just take the | > name after the "@", and look up the MX record to see where mail should be | > delivered (if there's no MX record they'll look for an A record). | ........ | Are you absolutely sure that MX records are not required? | My understanding (and experience) is that they are mandatory, | and the RFC seems to say so: | |
formatting link
Well, RFC 974 was obsoleted by RFC 2821, and now has "Status: HISTORIC". RFC 2821 [the "rewrite" of RFC 821] quite explicitly agrees with what Barry said above:

formatting link
Simple Mail Transfer Protocol Obsoletes: 821, 974, 1869 April 2001 Updates: 1123 ... 5. Address Resolution and Mail Handling ... a DNS lookup MUST be performed to resolve the domain name. ... The lookup first attempts to locate an MX record associated with the name. If a CNAME record is found instead, the resulting name is processed as if it were the initial name. If no MX records are found, but an A RR is found, the A RR is treated as if it was associated with an implicit MX RR, with a preference of 0, pointing to that host.

Perhaps you were thinking of the next few sentences, which *don't* contradict the previous but do require failure if an explicit MX RR is present but unusable [regardless of any A RR]?

If one or more MX RRs are found for a given name, SMTP systems MUST NOT utilize any A RRs associated with that name unless they are located using the MX RRs; the "implicit MX" rule above applies only if there are no MX records present. If MX records are present, but none of them are usable, this situation MUST be reported as an error.

In any case, it is certainly the case that an A RR *without* an MX RR for the same domain is quite legal, and is supported by every major piece of mail relay software I know of. I have administered numerous machines which lacked MX RRs, and they had no trouble at all receiving mail.

-Rob

----- Rob Warnock

627 26th Avenue San Mateo, CA 94403 (650)572-2607
Reply to
Rob Warnock

Each MX record points to an A record. If the server happens to have that A record, it sends it along so the client doesn't have to do a separate lookup.

R's, John

Reply to
John Levine

In article ,

There's actually an RFC -- i'm too lazy to go look up the number -- that addresses that. automagic pre-pending, _and_, appending, of certain common prefixes (specifically, "www."), and suffixes (a list, which includes '.com', and '.net' [or the local 'national' equivalents, e.g. '.co.uk' for england], and the 'commonly used' national suffixes in the locale -- e.g. .us, .ca, .fr, etc.)

Reply to
Robert Bonomi

In article ,

"Not Exactly". DNS servers try to be 'smart' and anticipate the things you would 'usually' proceed to ask for, based on the current request. They then supply those things 'in advance' (in the "additional records" part of the reply packet), so that that information is cached in the downstream (i.e., 'closer to you') DNS servers. This makes for much faster response time on the subsequent requests.

a MX query will return the _name_ of the mail server(s) in the 'answer' section of the response, *AND* the address (A and/or AAAA records) in the 'additional records' section of the response, _if_ the mailserver is in the same domain 'zone' as the hostname being queried for.

To wit:

If you query for the mailserver for 'foo.bar.baz', and mail for that host is handled by 'mail.bar.baz', you'll get a response consisting of a PTR to 'mail.bar.baz', _and_ an 'additional' section containing the A record for mail.bar.baz'.

OTOH, mail for foo.bar.baz is handled by 'mail.yahoo.com', then you will only get the PTR record in the reply section, and nothing in the additional records section. [to get the A record for 'mail.yahoo.com', one has to consult a different zone, and probably a different 'authoritative' DNS server, hence the 'bar.baz' DNS server cannot supply authoritative data]

Reply to
Robert Bonomi

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.