2009/2/9 Marcus Denker <denker@iam.unibe.ch>:
On 08.02.2009, at 18:13, Igor Stasenko wrote:
Hi Markus,
i have some comments about morphs :)
To my personal feel, a drop shadow feature (and all what is connected with it) is quite controversial, i don't think its worth promoting hasDropShadow shadowOffset to ivars. Instead, i think its worth demoting these ones to a void! :)
What I wanted to say is that now certain state that is accessed on *every redraw* is kept in a dictionary in MorphExtension, while at the same time there are instance variables in MorphExtension that are never used at all.
So I think we would get some speedup by just storing the state that morhic accesses now on every redraw into iVars (while cleaning up, e.g. removing etoys related state or checking if it really makes sense to have both balloonText amd balloonTextSelector as iVars...).
This would not be a real cleanup. It's exactly some small thing of the kind that is easy to not do "because it can be done even better"...
I was mistaken, i thought you were talking about putting into Morph ivars , not MorphExtension. Sure, putting ivars into MorphExtension does not hurt anyone, because it doesn't makes it better or worse :) I still doubt that placing them in ivars will affect performance that you will notice it. Mainly because there are many other things in event handling and drawing that take a times more cycles.
Can anyone tell me why morph (and moreover - canvas) should be aware of a shadows effect?
No idea... I am not a specialist for graphics stuff. To me, morphic looks like an experiment abandoned while it was heavily hacked to meet the daily needs of etoy-development. A lot of stuff seems to be hacked into it in a hurry, and many things where never improved "because one could do even better". Which of course lead to nothing beeing done at all.
I think this happens because there is no dedicated people, like Gary who constantly paying attention on improving this part of Squeak. When we develop in jump-in-then-jump-out style, we really don't care about improving - just writing a code and make sure it works, and then leaving it rot there, because this part is not interesting and not exciting at all :)
Marcus
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- Best regards, Igor Stasenko AKA sig.