Re: [dnsext] draft-mohan-dns-query-xml-00.txt

[Wilmer van der Gaast <[e-mail]>]

On 28 September 2011 16:57, Mohan Parthasarathy <[e-mail]> wrote:
>
> If we can avoid hardcoding that would be nice. I am not sure I
> understand your proposal. Could you clarify ? The initial thought is
> that it would be nice to access it in a standard way but I do see the
> issues you noted above.
>
Basically what I'm suggesting is that instead of an IP address, the
user fills in a base URL. In your example on page 10 the base URL
would be "https://server_address/dns_service/". The client
implementation just glues query?name=.... to that base URL to get the
exact URL to query.

This lets me put the dns service under a different path, like for
example just /dns ... or I could have a /dns and a /dns-secure for
people who already want strict DNSSEC validation even though it may
make some names unresolvable.

>> B) Nit, but why not convert the booleans in responses from
>> <xx>value</xx> into <xx/> if it's 1 and <xx/> being absent if it's 0?
> Is this a standard approach ? Or it saves bytes ?
>
It saves bytes, IMHO it looks clearer and also it's (again IMHO)
easier to parse; you just check if the element is present or not,
instead of finding it and evaluating it (what to do if the text value
is not 0 or 1?).


Cheers,

-- 
Wilmer van der Gaast, Traffic SRE/Google Public DNS team.
Google Ireland.
_______________________________________________
dnsext mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/dnsext