Re: [foaf-dev] Re: updated FOAF spec

Shadi Abou-Zahra wrote:
>   * foaf:mbox_sha1sum - I thought this was recommended over foaf:mbox in 
> one of the earlier FOAF specs.

Some of us found this ambiguous as to casing of the source mailto: URI. 
It wasn't clear to us whether any real systems allowed for case to make 
a distinction in the username portion of a mailto: URI (which would make 
jon@hackcraft.net and Jon@hackcraft.net distinct) and as such we weren't 
sure whether we could force a URI to lowercase before applying SHA-1 so 
that each address would have one possible mbox_sha1sum rather than 2^n 
where n is the number of alphabetic characters in the address.

If we could be sure it would be safe to do so, then enforcing a 
lowercase operation first would be advantageous for this way.

Better still would perhaps be a standard Unicode case-folding operation 
to better deal with cases like ß -> SS -> ss. However since Unicode 
case-folding of Latin characters uppercases them, and most existing 
sha1sums will be based on lowercase or mostly lowercase addresses this 
would need to be a new property to deal with the incompatibility (before 
we even begin to look at implementation availability of Unicode 
case-folding).

Received on Wednesday, 13 June 2007 13:00:37 UTC