Re: ldp-patch Requirements

On 11/09/13 00:22, Sandro Hawke wrote:
> I took a pass at enumerating possible requirements for ldp-patch. I
> included all the ones I think are important or I've heard someone else
> express as important, plus two or three others that are reasonable but
> not necessarily warranted.
>
> http://www.w3.org/2001/sw/wiki/PatchRequirements
>
> If there are any for which you can't see why anyone would want that, let
> me know and I'll explain it better.
>
>      -- Sandro

I appreciate the effort you've put in to the requirements page but I'm 
worried that simply gathering requirements will lead to a "swiss army 
penknife" proposal, something that that is in fact trying to do everything.

I think we need to define the problem to stop an explosion of features. 
  A strawpoll on requirements does not build a shared sense of the what 
end goal is.

A couple of examples:

At what point does patch stop and SPARQL Update start to be the better 
approach?   e.g. section 3.

What's the use case for calculating the patch-diff of two graph? (and 
why is part of patch and not a client-side operation?)

I see many of the requirements as competing with existing technology, 
leading to something that is too ambitious.  It is hard but the real 
question to me is how little can we get away with?

	Andy

Received on Wednesday, 11 September 2013 10:57:11 UTC