Re: [TF-PP] Possible starting points

On Tue, Sep 29, 2009 at 10:35 AM, Luke Wilson-Mawer
<luke.wilson-mawer@garlik.com> wrote:
> Hi Paul,
>
> One question...
>
> Paul Gearon wrote:
>>>
>>> 2/ Property paths with variables and IRIs or prefixed names.
>>> (issues include restriction of what can be asked a la ?p* discussion)
>>>
>>
>> +1. Takes memory, but is not hard (I've implemented this in the past).
>>
>
> H did your implementation behave when binding multiple ?ps ?

It depended. If the query plan bound the predicate first, then it was fine.

If not, then any non-trivial graph needed to remember every node in
the graph in RAM as it went (I remember these to detect loops). But in
general it was OK. It's a pretty simple technique. (I'll explain it if
that helps)

Regards,
Paul