[Graphhopper] start and final routing point

[Peter K]

Hi Johannes,

currently a "hot switch" is not possible. Probably there is still a
minor misunderstanding:
The location index is relative small and independent on the
shortest/fastest parameter, but the graph itself contains different
shortcut information if prepared with shortest or fastest. This is a
'limitation' of the algorithm (contraction hierarchies).

But if you have enough RAM (or you'r using MMap) you should be able to
load both graphs into two different GraphHopper instances. One could go
even the way of a mixed graph which contains both shortcut information -
that would save some RAM and disc space as the node data isn't
duplicated. But this is not implemented yet.

Regards,
Peter.

PS: thanks for the time to reformulate this in English and sorry, for my
confusion about mixing mailing list vs. private list!


> Thank you,
>
> if i prepare this two index repositories (shortest and fastest) next
> to each other, can i switch easily the repository by parameter, or do
> i need two graphhopper instances running?
>
> Greetings Johannes
>
>
>
> _______________________________________________
> GraphHopper mailing list
> GraphHopper at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/graphhopper


-- 
GraphHopper.com Your way is our destination!

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/graphhopper/attachments/20130312/92e14410/attachment.html>