Comments by Yossi Yaari

Yossi Yaari 10-Mar-13 11:42am View

I do agree with you in principal.

Yossi Yaari 10-Mar-13 11:40am View

:-)
thanks

Yossi Yaari 5-Mar-13 14:21pm View

that is of course correct in general.
and here the mathematical solution is very light weight. so I'm not sure how much it will improve performance. as you your self commented.

Yossi Yaari 5-Mar-13 14:19pm View

I think I fixed. thanks for the comment

Yossi Yaari 5-Mar-13 0:53am View

the best reason I think to do so is to minimize chance for error.
if the rectangles don't intersect - their not close enough. so very small chance for significant false positive answer on collision, if there is an error in the calculation.

Yossi Yaari 5-Mar-13 0:50am View

OK

Yossi Yaari 5-Mar-13 0:49am View

that would have very slow performance.
O(n) instead of O(1)

Yossi Yaari 5-Mar-13 0:48am View

what I mean is that you could add a method that accepts object of Type <t> and calculates the key by itself using the binding method. so that in theory, if the binding changed, you could leave the code the same.

MyClass a;
var data = new MemberBindingDictionary<myclass>(x => x.Name);
data.Add(a);
if(data.ContainsKey(a)) //<-- would stay true for any change of ^^^ binding method
{
...
}

Yossi Yaari 4-Mar-13 14:20pm View

sounds like a really forced question :-)

Yossi Yaari 4-Mar-13 13:40pm View

that's definitely the direction I was thinking.
just was wondering how bad the implementation would turn out. mostly in terms of performance.

and of course you could add a ContainsByKey using the same binding func. :-)

Yossi Yaari 4-Mar-13 13:21pm View

you meant IntersectsWith(...) of course

Yossi Yaari 4-Mar-13 13:13pm View

I think he means circles or 2D balls.
in any case your solution for calculating distance between centers is correct.

I wonder if it is worth testting Rectangle.Intersect in advance?

Yossi Yaari 4-Mar-13 12:34pm View

still not clear what kind of database solution your looking for

Comments by Yossi Yaari (Top 13 by date)