Re: Java "interface" vs. OO interface
From: Justin Farley (null_at_void.com)
Date: 11/23/03
- Next message: Joe \: "Re: Java "interface" vs. OO interface"
- Previous message: Isaac Gouy: "Re: [Cincom Smalltalk NC] - VW 7.2 and OS 6.9 available"
- In reply to: Dave Harris: "Re: Java "interface" vs. OO interface"
- Next in thread: Joe \: "Re: Java "interface" vs. OO interface"
- Reply: Joe \: "Re: Java "interface" vs. OO interface"
- Reply: Dave Harris: "Re: Java "interface" vs. OO interface"
- Reply: Uncle Bob (Robert C. Martin): "Re: Java "interface" vs. OO interface"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sat, 22 Nov 2003 23:45:31 GMT
Dave Harris wrote:
> null@void.com (Justin Farley) wrote (abridged):
>>> I don't think impure interfaces should be banned by the language.
>>
>> "impure interfaces" seems to me a contradiction in terms. Isn't the
>> whole point behind the constraint of Interfaces, is that they are,
>> at the very least, purely abstract?
>
> I think Java uses interfaces instead of classes because the designers
> couldn't figure out a good implementation of MI of classes. The
> ideology was a post-hoc justification.
The simplest thing that works, works well.
To think that the designers of Java and .NET couldn't figure out a good
implementation of MI, when there were plenty of well known good
implementations already, I think is a bit presumptuous. From what little
I've read from James Gosling (Java) and Anders Hejlsberg (C#), I'm convinced
that they *chose* the simplicity of constrained single
implementation-inheritance.
The same with DCOM/CORBA, and the subsequent simpler Service-Oriented
Architecture of Web Services. It is an intentional desire for constraints,
rather than the lack of a good implementation of something more complex,
that has driven the design IMO.
And again with the dumbing down of GUIs into the Web Browser single-click
stuff. The simplest thing that works is intentional design, not any lack of
a "good implementation".
>> There are many design patterns where pure abstract and instanceless
>> classes are useful. Enough for a language to adopt a new keyword
>> to identify them as such.
>
> I don't think so. Keywords need more justification than that, in my
> view.
Surely not the keyword alone?!? Humans are highly linguistic. I was being
somewhat sarcastic in suggesting that a new keyword was significant. What
is important is the explicit identification (and significance) of purely
abstract interfaces in designs. If all it takes is a keyword, it is well
worth it.
-- Justin
- Next message: Joe \: "Re: Java "interface" vs. OO interface"
- Previous message: Isaac Gouy: "Re: [Cincom Smalltalk NC] - VW 7.2 and OS 6.9 available"
- In reply to: Dave Harris: "Re: Java "interface" vs. OO interface"
- Next in thread: Joe \: "Re: Java "interface" vs. OO interface"
- Reply: Joe \: "Re: Java "interface" vs. OO interface"
- Reply: Dave Harris: "Re: Java "interface" vs. OO interface"
- Reply: Uncle Bob (Robert C. Martin): "Re: Java "interface" vs. OO interface"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|