knight@unipas.fmi.uni-passau.de (Alan Knight) (07/18/90)
It seems to be pretty much a common assumption that object-oriented programming works well with graphics, and that it is a lot easier to do graphics, particularly with a windowing environment, using OOP. On the other hand, I havent really found a good reference where somebody actually looked at the relative difficulty. Does anybody know of such a reference? Thanks, Alan Knight knight@unipas.fmi.uni-passau.de
wissk (Peter Wisskirchen) (07/18/90)
In article <1990Jul17.180931.16042@forwiss.uni-passau.de> knight@unipas.fmi.uni-passau.de (Alan Knight) writes: > > It seems to be pretty much a common assumption that object-oriented > programming works well with graphics, and that it is a lot easier to > do graphics, particularly with a windowing environment, using OOP. On > the other hand, I havent really found a good reference where somebody > actually looked at the relative difficulty. Does anybody know of such > a reference? In my book "Object-Oriented Graphics--From GKS and PHIGS to OO-Systems" (Springer Verlag, 1990, to appear in August 1990, ISBN 0-387-52859-8) I compared especially PHIGS with an OO alternative, showing the practical advantages which can be achieved with an OO approach. Besides inheritance, the oo naming concept and the way how graphics primitives can be edited are quite useful. In addition my (Smalltalk-80) examples tend to the conclusion that a prototype-delegation model is better suited to the needs of interactive graphics than class-inheritance. I'am interested in discussing this (delegation or inheritance for interactive graphics?) Peter W.
knight@unipas.fmi.uni-passau.de (Alan Knight) (07/18/90)
In article <3127@gmdzi.UUCP> wissk (Peter Wisskirchen) writes: >In article <1990Jul17.180931.16042@forwiss.uni-passau.de> >knight@unipas.fmi.uni-passau.de (Alan Knight) writes: >> It seems to be pretty much a common assumption that object-oriented >> programming works well with graphics, and that it is a lot easier <stuff omitted> > >In my book "Object-Oriented Graphics--From GKS and PHIGS to OO-Systems" >(Springer Verlag, 1990, to appear in August 1990, ISBN 0-387-52859-8) >I compared especially PHIGS with an OO alternative, showing the practical >advantages which can be achieved with an OO approach. >Besides inheritance, the oo naming concept and the way how >graphics primitives can be edited are quite useful. In addition my >(Smalltalk-80) examples tend to the conclusion that a prototype-delegation >model is better suited to the needs of >interactive graphics than class-inheritance. >I'am interested in discussing this (delegation or inheritance for >interactive graphics?) >Peter W. This is quite interesting. We have been doing work with a system for computational geometry, which has a lot of similarities to graphics. What is it that makes you believe that a prototype/delegation approach is more suitable? We havent really looked at the possibility of using such a model, but it might be helpful in overcoming some of our problems. Alan Knight knight@unipas.fmi.uni-passau.de (until July 25th) after that use Alan_Knight@carleton.ca