"Dan" == Dan Ingalls Dan.Ingalls@disney.com writes:
Dan> This is probably a general problem. There're are some 106 users Dan> of the global 'World' I'm too tired to check them all. Most of Dan> them should be tranlated into a test whether we're running a full Dan> screen world or an MVC world. The others might be candidates for Dan> #currentWorld. --sma"
Why isn't there a "do the right open" call? Why should everything that opens a top-level widget have to know *how* to open it?
Or maybe I'm missing something?
At 21:02 26.04.00 -0700, Randal L. Schwartz wrote:
Why isn't there a "do the right open" call? Why should everything that opens a top-level widget have to know *how* to open it?
It's the "that historically caused" reason. There was (and is) no real documentation so everybody hacked in some way, perhaps looking how somebody else it already did. Actually, there's I think an idiom (language pattern) to use when you want to start a morphic application
ApplicationMorph new openInWorld
(there're a some variants - some of them IMHO obsolete or at least badly named)
Or maybe I'm missing something?
The problem with free cell wasn't how to open morphs but not to use and interpret globals values. There're two kinds of morphic worlds available. Full screen worlds and worlds as MVC windows. (Actually, there're three, worlds in morphic worlds are also possible but life is difficult enough already so let's ignore that) The global World could be used to determine wether you're running as full screen or not. IMHO it shouldn't be used and instead of that, "self world isFullScreen" should be preferred. In a full screen world, there's only one and it's easy to refer to that world. With windowed worlds, you could have more than one and applications must be careful to open their child windows in their own worlds. Some applications don't honnor that I'm afraid.
bye
squeak-dev@lists.squeakfoundation.org