Bonnie,

If you are not ignoring his questions, why I did not read an straight answer? What you guys ignore is that his question is one that people wrote books about and are not in agreement yet, but nobody aknowledges it here and we need to surrender our thougths for otherwise we are personally attacked or laughed at? Let me tell you this, all the examples proposed, as far as I can see, had nothing to do with the need to use interfaces to simulate MI, nor with MI directly, which is the only thing that Doru complained, but instead of answering that, the issue was sidetracked to show how good are the interfaces for other things that were never questioned. Of course I might be wrong, I am well known for that, but I am willing to accept that.

About the collections thing, laugh all you want, I can laugh at the thought of thinking that interfaces is the only solution to a class design problem, I can't imagine what the poor souls condemned to use a pure OO language like Smalltalk would do.


>Hugo,
>
>>you are an arrogant person who obviously do not understand Doru's point and you hide in personal attacks instead of answering his questions straight, which are not only valid but true, even if I disagree with his example, but you, not very graciously, ignored to answer his questions.
>
>I'm sure Perry can stick up for himself, but I'd just like to say that we are not ignoring Doru's questions. The problem is that it looks like, no matter what we say to Doru, he won't agree with it and comes up with pretty bad solutions (>>"there are better ways, like using collections to store which class is edible for which class, etc."). I mean ... come on!!! Use collections for this?!?!?!
>
>Perry's just getting tired of arguing with him ... and he's not the only one.
>
>~~Bonnie
>
>
>
>
>
>>>I wasn't going to bother replying to you anymore, as I know other's feel this way also. But this last comment is flat out wrong, laughable at best.
>>>
>>>"there are better ways, like using collections to store which class is edible for which class, etc."
>>>
>>>You've obviously not coded in dotnet or java. As everyone has said, Interfaces are a great way to enforce good coding standards. I've not ever, ever seen a situation of spaghetti code arising out of use of interfaces.
>>
>>
>>Well, I think this will be my only reply to you too, as is a waste of time for you are an arrogant person who obviously do not understand Doru's point and you hide in personal attacks instead of answering his questions straight, which are not only valid but true, even if I disagree with his example, but you, not very graciously, ignored to answer his questions. Ah, in fact, I have also a hard time imagining why you need an interface in that example (other than the need for them in typed languages maybe?) or even why this is a good example of using interfaces to simulate MI, which, again, is the key of the matter, so I would love to see the person who proposed it to show how would he do it, at least Doru is trying to answer.
>>
>>First, dotnet is not a language, for you can not "code" in dotnet and you can make use of MI directly with some of the "dot net languages" without the need of using interfaces.
>>
>>Second, the discussion is about concepts, so the languages are irrelevant. You claim that interfaces are a way to enforce good coding standards, and I also happen to think they are great, but they hardly enforce a good design, that is the key of the question, in fact many of the great advantages of the interfaces have nothing to do with good design, but to overcome "limitations" (in quotes for I know is not the right word) of the language, i.e to allow polymorphism in non-dynamic languages (for example you do not need to use them, in fact the concept of interfaces as we are discussing here is newer than the language itself, in a pure OO language like Smalltalk, so in your view I guess it is impossible to have a good design in said language, of course you pay a price too, like Rick said) and ah! or to simulate MI, forgot what started this all
>>
>>Third, it seems that you still not understand that Doru is not arguing about interfaces but about MI but of course he is not discussing any of the other great advantages of Interfaces, did you read him saying anything about being worthless to enforce contracts or please can you point me where did he said that the use of interfaces leads to spaghetti code? He said, or at least is what I understood, that the use of interfaces to simulate MI CAN lead to spaghetti code, which is a far cry from what you are saying he said.