Re: [ALL] 48 Hours to Prioritize Graph Use Cases

On 06/10/11 11:09, Steve Harris wrote:
> On 2011-10-06, at 10:47, Andy Seaborne wrote:
>>
>> On 06/10/11 10:29, Steve Harris wrote:
>>> On 2011-10-06, at 08:35, Andy Seaborne wrote:
>>>>
>>>> On 05/10/11 23:16, Ian Davis wrote:
>>>>> I think 4.2 web crawling (or aggregation in general) is very
>>>>> important
>>>>>
>>>>> Also I don't see it on the existing list but being able to edit
>>>>> datasets at the granularity of the graph rather than the triple
>>>>> is very important. i.e. It is very convenient to be able to
>>>>> replace all the triples in one named graph with a new set when
>>>>> doing things like web crawling.
>>>>
>>>> Agreed.
>>>>
>>>> Some words added (section 1.7, to keep the numbering stable); Ian,
>>>> do feel free to refine them.
>>>>
>>>> This relates to 1.5 "Exchanging the contents of RDF stores".
>>>
>>> Agreed, on all counts.
>>>
>>> However I'd like it to be expanded in scope slightly to also cover
>>> information derived from non-RDF pages which are being crawled (pure
>>> self-interest, as this is where we get most of our data). It's
>>> slightly different from a modelling point of view, as foo.html
>>> doesn't typically contain any actual triples, just information which
>>> cause triples to be asserted.
>>
>> It's a wiki ... :-)
>
> Indeed, but it felt a little underhand to change the focus of a use-case after people had voted for it!
>
> If neither you or Ian feel it's an issue, I'll add a note to it.

It's better to capture it than loose it - how about adding a 
section/paragraph if you think it's does not warrant being "1.8"

	Andy

>
> - Steve
>
>>> That, in combination with “1.3 Graph Changes Over Time”, and “1.5
>>> Exchanging the contents of RDF stores” captures our most challenging
>>> data representation challenges.
>>>
>>> - Steve

Received on Thursday, 6 October 2011 10:59:40 UTC