On Thu, 2008-01-03 at 00:56 +1030, Tim Ansell wrote:
> > Another thing is that I feel that the difference list is a little bit
> > underspecified. Are duplicate ids allowed in the difference list? Is there
> > a
> > limit to the list?
>
> No, duplicate ID's do not make any sense in my understanding of
> difference lists.
>
> My idea with difference lists where that they should return objects
> which have changed since X (not necessarily in the order they changed).
>
> For example, on tpserver-py I would just do a query something like
> "SELECT id FROM blah WHERE modtime > diftime". This even works for dead
> objects in my case, as once it's dead, it's mod time never changes
> again.
>
> This means you shouldn't need to store anything extra for difference
> support. (You may need to store extra information for history stuff, if
> in the future we need to figure out which turns an object changed).
>
> In the worst case this is just returning a list of all object IDs
> (including all objects which have died).
Yep.
> In the best case you get no
> object IDs. My guess is that around 20%-30% of objects in many universes
> will be static (and all dead objects should be static).
>
> I will reply to the "objects are always changing" grip in another email.
>
> Does that make sense? Nash, is that what you where thinking?
The only thing I immediately see is the list _is_ providing too much
information. So I know someone has a colony (or remote mining, or has
fleets in orbit) around a planet because I keep seeing it change.
Obviously this is a continuing issue I keep raising ;-)
Regards,
nash
_______________________________________________
tp-devel mailing list
[email protected]
http://www.thousandparsec.net/tp/mailman.php/listinfo/tp-devel
