Chris Muller uploaded a new version of ToolBuilder-Kernel to project The Trunk:
http://source.squeak.org/trunk/ToolBuilder-Kernel-cmm.48.mcz ==================== Summary ==================== Name: ToolBuilder-Kernel-cmm.48 Author: cmm Time: 15 June 2011, 11:23:16.939 am UUID: b0bc3b05-7fe2-8847-8af3-5831ca1777aa Ancestors: ToolBuilder-Kernel-ul.47 Added UIManager>>screen to provide access to the Display. ToolBuilder subclasses already have several dependencies on Graphics, so this does not, in a practical sense, add a new dependency on Graphics. =============== Diff against ToolBuilder-Kernel-ul.47 =============== Item was added: + ----- Method: UIManager>>screen (in category 'accessing') ----- + screen + ^ Display! |
On 15.06.2011, at 16:23, [hidden email] wrote: > Chris Muller uploaded a new version of ToolBuilder-Kernel to project The Trunk: > http://source.squeak.org/trunk/ToolBuilder-Kernel-cmm.48.mcz > > ==================== Summary ==================== > > Name: ToolBuilder-Kernel-cmm.48 > Author: cmm > Time: 15 June 2011, 11:23:16.939 am > UUID: b0bc3b05-7fe2-8847-8af3-5831ca1777aa > Ancestors: ToolBuilder-Kernel-ul.47 > > Added UIManager>>screen to provide access to the Display. ToolBuilder subclasses already have several dependencies on Graphics, so this does not, in a practical sense, add a new dependency on Graphics. > > =============== Diff against ToolBuilder-Kernel-ul.47 =============== > > Item was added: > + ----- Method: UIManager>>screen (in category 'accessing') ----- > + screen > + ^ Display! I could see the usefulness of a #screenSize accessor. But exposing a raw bitmap form seems not very useful for many kinds of a UI framework. - Bert - |
I presume you mean #screenBounds (since we need the points). Usually
I shy away from adding delegating methods for simply accessing the individual "primitive" attributes (like, a Point's X, a Date's month, a Color's luminance, or a DisplayScreen's bounds) of a contained object, to try to avoid too much method-proliferation. But I definitely agree with your statement. What does anyone else think? Remove #screen in favor of #screenBounds? On Wed, Jun 15, 2011 at 11:30 AM, Bert Freudenberg <[hidden email]> wrote: > > On 15.06.2011, at 16:23, [hidden email] wrote: > >> Chris Muller uploaded a new version of ToolBuilder-Kernel to project The Trunk: >> http://source.squeak.org/trunk/ToolBuilder-Kernel-cmm.48.mcz >> >> ==================== Summary ==================== >> >> Name: ToolBuilder-Kernel-cmm.48 >> Author: cmm >> Time: 15 June 2011, 11:23:16.939 am >> UUID: b0bc3b05-7fe2-8847-8af3-5831ca1777aa >> Ancestors: ToolBuilder-Kernel-ul.47 >> >> Added UIManager>>screen to provide access to the Display. ToolBuilder subclasses already have several dependencies on Graphics, so this does not, in a practical sense, add a new dependency on Graphics. >> >> =============== Diff against ToolBuilder-Kernel-ul.47 =============== >> >> Item was added: >> + ----- Method: UIManager>>screen (in category 'accessing') ----- >> + screen >> + ^ Display! > > > I could see the usefulness of a #screenSize accessor. But exposing a raw bitmap form seems not very useful for many kinds of a UI framework. > > - Bert - > > > |
On Wed, Jun 15, 2011 at 9:49 AM, Chris Muller <[hidden email]> wrote:
> I presume you mean #screenBounds (since we need the points). Usually > I shy away from adding delegating methods for simply accessing the > individual "primitive" attributes (like, a Point's X, a Date's month, > a Color's luminance, or a DisplayScreen's bounds) of a contained > object, to try to avoid too much method-proliferation. > > But I definitely agree with your statement. > > What does anyone else think? Remove #screen in favor of #screenBounds? +1 Morphic and Tweak have a display bitmap, but other UI frameworks like native widgets or HTML don't. They probably do have some notion of the screen area, though. Colin |
Free forum by Nabble | Edit this page |