Some more possible requirements

[Andrew Draper <ADRAPER@altera.com>]

> Me too! SOMEONE apart from Marc, James, Rob and me must have opinions!

OK, here are my thoughts on requirements, my background is that at in
previous jobs I a) was internal DNS administrator for a medium sized company
and b) wrote a DNS resolver library.


* Zone files should remain easily editable.
* I should be able to perform offline DNSSEC signing if I desire.

For ease of debugging the offline signed file should be sufficiently similar
to the unsigned file that I can tell, for example, whether my offline
signing script has broken.

This does not restrict the character set used in zone files, it just implies
that the character set of a signed file should be capable of being the same
as the character set of the unsigned file.


* I should be able to mix different languages in the same zone.

For example: I should be able to have the name of the UK office in English,
the French office in French, the Norwegian office in Norwegian and the Thai
office in Thai, all followed by .mycompany.com.

One way I might do this is to have the real name of each system in French
(to make life easier for my French speaking IS staff), and add CNAMES/DNAMES
to alias things in different languages (to make things easier for users), or
vice versa.  Or something else.


* The spec should not make me pay registrars many times to register the same
name.

Ie if I want banos.com I should not also have to pay for BANOS.com (and all
the other permutations of upper and lower case if there is more than one
non-ASCII character).  This implies that I should be able to register some
sort of canonical form.

[In case my mail gets garbled, the third character of my example is N-tilde]

This also holds when I am editing my internal zone files, having a canonical
form will make my files smaller and easier to maintain.


* An IDN capable resolver should not generate any more traffic (for both IDN
and pure ASCII names) than a non IDN capable resolver.

This implies that there should be some sort of canonical form which the
resolver can use for requests.


* No DNS element should be required to guess anything.

Any guessing should be outside the protocol specified here.  If a web client
wants to guess whether some characters are Chinese or Japanese it should be
allowed to do so, but the DNS protocol (resolver - network - server - zone
file) should remain deterministic.

If we allow multiple character sets then they should be clearly identified.


* The same request should generate the same response, regardless of the
location (or localisation settings) of the resolver, the master server and
any slave or caching servers involved.  

Again, there may need to be some localisation within the software which is
using the resolver library, but that is not the business of the DNS
protocol.


* It should be possible to build a caching server which does not understand
the language in which a request (or response) is encoded, and which works as
well for IDNs as in the ASCII-only case.


That's all I can think of for the moment.

Regards,

    Andrew Draper.