Except that "pseudo" sounds disparaging, and I actually like an object model which has multiple inheritance of interfaces and single inheritance of implementations (same as Java, and maybe C# too?). Multiple inheritance of implementation results in ambiguity. Composition/delegation is a better approach in my view. Lots of literature about this that the original poster can read.
Having said that, I did have one occasion in my career when I followed an example from Bertrand Meyer's book and implemented a tree node in C++ as a link and a list. But in general I think implementation inheritance is over-used and multiple inheritance can get you into trouble. On Fri, Oct 10, 2008 at 7:57 AM, Ryan Gravener <[EMAIL PROTECTED]>wrote: > Pseudo multiple inheritance. > > Ryan Gravener > http://twitter.com/ryangravener > > > > On Fri, Oct 10, 2008 at 10:23 AM, Richard Rodseth <[EMAIL PROTECTED]>wrote: > >> And sometimes you even have interfaces with no methods. In this case >> it's a "marker" (often a parent of other interfaces) and when used in method >> signatures you get type checking. >> >> >> On Fri, Oct 10, 2008 at 6:26 AM, Michael Schmalle < >> [EMAIL PROTECTED]> wrote: >> >>> Hi, >>> It's ICommand. >>> >>> The reason is you can stack interfaces on top of each other allowing more >>> decoupling to the implementing concrete classes. >>> >>> This interface is obvious. Any class that implements it needs eval() and >>> only eval. It's like a singleton declaration of implementation. >>> >>> If you jammed this evel() method into IUIComponent, maybe all components >>> don't need eval. Make sense? >>> >>> Also another good example of this type of interface in the flex framework >>> is IDataRenderer, it's only declared property is 'data'. >>> >>> Mike >>> >>> On Fri, Oct 10, 2008 at 9:19 AM, flexaustin <[EMAIL PROTECTED]>wrote: >>> >>>> I was wondering if someone can explain why you would need an >>>> interface >>>> so short? >>>> >>>> INTERFACE: >>>> >>>> package my.package.area >>>> { >>>> /** >>>> * Interface for methods that evaluate an object and return a result. >>>> */ >>>> public interface IEval >>>> { >>>> /** >>>> * Evaluates the input object >>>> * @o the object to evaluate >>>> * @return the computed result value >>>> */ >>>> function eval(o:Object=null):*; >>>> >>>> } // end of interface IEval >>>> } >>>> >>>> USAGE OF INTERFACE: >>>> >>>> if (value is IEval) { value = IEval(value).eval(o) }; >>>> >>>> Cairngorm has a short interface like this as well, though I cannot >>>> remember what it is. In Cairngorm they say its for naming or to make >>>> the code easier to understand? I am just not sure why you would do >>>> this? Help me see the light! >>>> >>>> TIA >>>> >>>> >>> >>> >>> -- >>> Teoti Graphix, LLC >>> http://www.teotigraphix.com >>> >>> Teoti Graphix Blog >>> http://www.blog.teotigraphix.com >>> >>> You can find more by solving the problem then by 'asking the question'. >>> >> >> > >
