Re: White Spaces in cert:modulus

On 14 Jan 2012, at 21:47, Dominik Tomaszuk wrote:

> On 14.01.2012 19:52, Henry Story wrote:
>> As I see it this means that all spaces in the sequence are going to be reduced down to 1 space. Which means that the canonical representation will be different between strings with no space and strings with space.
> That's right. This was the reason, why I suggested [1] to build own datatype, based on xsd:string (or xsd:normalizedString or xsd:token) with restrictions, based on regular expressions, which allowed whitespaces between numbers a few weeks ago.
> But my idea was lost in the vote.

Your idea was noticed, but did not gather much support since you were proposing a variant of what we already had. Ie we already had a flexible datatype called cert:hex. You were proposing the same but with a different datatype, which meant there was less flexibility for not much gain. So it did not make things clearer whilst reducing flexibility and forcing us to make decisions without the support of the larger RDF community.

We decided instead to trying to be closer to the standards so that we could gain support of Datatype inferencing from SPARQL and other tools. Your proposal would not have give us that.

So given that that is a question we deal with a few months ago, and xsd:hexBinary had very large support, we are now looking to see if there is something about xsd:hexBinary that we don't know or that is unclear. This is easy to work on because it is a standard so we don't have to make things up here. We could perhaps even ask that this be clarified and improved in the next version of RDF. It seems that Virtuoso for example parses xsd:hexBinary very much the way cert:hex did.

Henry

> 
> [1] http://lists.w3.org/Archives/Public/public-xg-webid/2011Nov/0110.html

Can't access this at this time.

> 
> Best regards,
> Dominik 'domel' Tomaszuk

Social Web Architect
http://bblfish.net/

Received on Saturday, 14 January 2012 21:12:53 UTC