On Wed, Dec 22, 2010 at 11:09 AM, Alexandre Bergel <alexandre@bergel.eu> wrote:
Raster image = bitmap. Â As you know, bitmaps have the ability, no matter whether you render 1 node or 1-billion nodes and edges on it, it does not take up much more space.
It depends on the size of the bitmap :-)
Not really; if you start with an empty (say, all white), uncompressed bitmap, painting various graphics on it does not change its uncompressed size.
I am not sure to understand. Suppose you have a domain (e.g., cars, persons and ownership). You can render this domain the way you want. Dragging nodes does not changes your models. It changes the graph built by Mondrian (i.e., instances of MONode and MOEdes).
Right. I am not talking about the cars and persons, I am talking about "Mondrians domain model", "(i.e., instances of MONode and MOEdes)". If I have 1 million cars and persons and I want Mondrian to render a visualization involving all of them, then it has to create and *keep* 1 million MONodes. I want to do it without keeping them since that cannot scale. Assume, perhaps, that I have my 1 million Cars in a Magma database, and I don't want to persist 1M MONodes just to render a quick picture.
What is exactly the problem you're trying to solve?
The (imaginary) problem I'm interested to solve is I have a huge model persisted in Magma, I want to make a visualization based on one million+ objects without creating and keeping 1M MONodes in memory.... Instead, I would like to create each MONode, render it on the raster image, but don't keep it so it can be GC'd. But its representation is already preserved on the bitmap.