[3607] in java-interest
Re: protected is not?
daemon@ATHENA.MIT.EDU (Michael Daconta)
Fri Nov 17 10:28:20 1995
Date: Fri, 17 Nov 1995 06:35:18 -0700 (MST)
To: "D'Arcy Smith" <darcy@arcs.bcit.bc.ca>, java-interest@java.sun.com
From: daconta@PrimeNet.Com (Michael Daconta)
At 08:11 AM 11/16/95 -0800, D'Arcy Smith wrote:
Hi Darcy,
Thanks for responding. I agree with you that it sucks. I have not
done much with packages but do not really see why they should change
such a fundamental concept like data hiding.
The Java team has done alot of smart things. I don't see why they
broke compatibility on simple things like the access specifier and
the command line arguments (primarily no arg 0 as prog name).
I expect compatibility to be broken on major philosophical differences
(i.e. operator overloading).
- Mike
>Michael Daconta wrote:
>>
>> I hear what you are saying but that does not track with the meaning
>> of a Protected variable. Essentially what you said was - it is
>> only "protected" outside of the package. That would be a serious
>> change to the idea of access specifiers.
>>
>> Also, packages are primarily a mechanism to separate namespaces.
>> Being in the same package should have nothing to do with permission
>> to access protected variables. If that is the default, I think the
>> default should be changed to the same meaning as in C++.
>>
>
>Consider all classes in a package as being friends.
>
>> Changing the meaning of an access specifier even in the same package
>> does serious damage to the concept of encapsulation and data hiding.
>>
>> I'd rather hope this was a bug. Java team? (And the survey says...)
>>
>
>It appears to be intentional ... I've been modifying the Beta javac
>to try out my idea (see below) ... and the code is quite deliberate.
>Fortunatly the change is small (add a new keyword, check the access
>in a slightly different manner (1 'if' stmt). The change should only
>take me about 1 hour (Arthur ... 5 mins :-)
>
>Yes this sucks ... I have proposed to this group & to Sun to add
>'shared' access that allows package-wide access thus protected would
>have it's C++ meaning - also the implicit friending would be gone.
>
>..darcy
>--
> D'Arcy Smith, Systems Analyst
> Applied Research in Computer Systems (ARCS) Laboratory
> British Columbia Institute of Technology (BCIT), Burnaby, BC, Canada
> E-Mail: darcy@arcs.bcit.bc.ca URL http://www.arcs.bcit.bc.ca
> Tel: (604) 432-8281 Fax: (604) 436-1297
>
>
-
This message was sent to the java-interest mailing list
Info: send 'help' to java-interest-request@java.sun.com
