Re: Review of "SPARQL 1.1 Update"

On 11 Jan 2010, at 11:57, Andy Seaborne wrote:
>>> I think that introducing ";" for all operations because this one  
>>> short
>>> form needs it is not balanced so I'm keen to find a way to avoid  
>>> that
>>> necessity.
>>
>> Ah, I was thinking that the last ; would be optional, like . and  
>> triples.
>
> I was assuming that also.  It's that every (multi-operation)  
> sequence now needs to have ";"s when the syntax issue is confined to  
> the abbreviated short form of DELETE that strikes me as not ideal.

Well, the cost is that you have to type a ";", but the advantage it  
should be clearer to users what the expression means. I don't really  
see that as a significant cost.

Alternatives involving scoping brackets or similar require more  
complex syntactic structures, and affect even single expressions.

> I also think that multiple operations in one request will not be  
> uncommon.  Basic data loading might be commonly one operation  
> although surely much of the need for SPARQL Update Language is for  
> those operations not done by the HTTP update style.  It will include  
> things like ensuring graphs exist before other operations happen.

I agree, I'd expect multiple operations to be fairly common.

- Steve