Re: [Uri-review] Initial inquiry into URI proposal (nym)

[David Booth <[e-mail]>]

On 06/27/2013 04:32 PM, DataPacRat wrote:
> On Thu, Jun 27, 2013 at 3:06 PM, David Booth <[e-mail]> wrote:
>> On 06/27/2013 02:43 PM, DataPacRat wrote:
>
>>> What I'm trying to build with nym (one way or another, and with or
>>> without other complications thrown in), is to have some sort of
>>> structure that uses such indirect addressing: to take any given URI,
>>> such as a webpage or mailto: address, and instead of referring
>>> directly to whatever is pointed at by that URI, having the nym use
>>> indirect addressing to be a reference to whatever that URI describes.
>>
>> That sounds very much like
>> http://thing-described-by.org/
>> (or the shorter version http://t-d-b.org/ ), except that
>> http://thing-described-by.org/ also allows the URI to be dereferenceable
>> without requiring clients to implement a new URI scheme or protocol.
>>
>> You might want to look at "Converting New URI Schemes or URN Sub-Schemes to
>> HTTP" also:
>> http://dbooth.org/2006/urn2http/
>
> Thank you for making the suggestion; but it's an approach I already
> thought of, and rejected, for a number of reasons, at least two of
> them major.
>
> For one, both of them involve establishing or using a particular
> domain to use as a storer or redirector; which introduces a single
> point of failure, which goes rather against much of the point of
> having a distributed protocol/format/whatever.

That can be addressed using a .well-known URI suffix instead:
http://tools.ietf.org/html/rfc5785

David

>
> The other is a classic bit: "There are fourteen separate sites which
> store personal identity information in different ways: Let's make a
> new one that puts them all together!" -> "There are fifteen separate
> sites which store personal identity information in different ways...".
>
>
>
> Thank you for your time,
> --
> DataPacRat
> "Then again, I could be wrong."
>
>
>