Re: [Pipmak-Users] This code show the image-cache-bug...

[Christian W. <cwa...@gm...>]

[For those who only read -users: This is a continuation of a thread=20
started on pipmak-devel. I'm moving it here because I'm interested in=20
feedback from everyone. Read the beginning here:=20
<http://thread.gmane.org/gmane.games.devel.pipmak.devel/77>.]

I wrote:
> What trips you up is this (from section 3.12 "Image Manipulation" of th=
e=20
> reference):
> "pipmak.loadimage(filename) ... Calling this function multiple times=20
> with the same path returns the same object, and the image is only=20
> reloaded from disk if it isn=92t already in memory."
>=20
> ...
>=20
> I'm not sure what the best solution to this is. I don't think giving up=
=20
> the returns-the-same-object behavior of pipmak.loadimage is a good idea=
,=20
> as I rely on it in many places. I'll try to think of something better

I'm still pondering the copy-on-write thing, but here's another solution:=


- Rename pipmak.loadimage to pipmak.getimage to emphasize its behavior=20
(keeping the old name around for backwards compatibility).

- Disallow modification of images gotten from pipmak.getimage - if you=20
want to draw on an image, you must create it using pipmak.newimage.

- Allow pipmak.newimage to be called with a path instead of width and=20
height, to initialize the new image from a file.

What do you think?

This would be simpler to implement, but would complicate things from the =

point of view of the user, by dividing image objects into two classes.

  -Christian