Child packages [nn,pedo,incest,cons

Child packages [nn,pedo,incest,cons

[firth]

Subject: Child packages [nn,pedo,incest,cons]

In article <1994Oct14.094121@di.epfl.ch> Robb.Nebbe@di.epfl.ch (Robb Nebbe) writes:

>|>    The private part and body of any child can access the private part of
>|>    its parent (and grandparent...).

In principle, any mechanism that allows one to extend an abstraction
also allows one to break it.  That's a hard choice for the language
designer to make.

However, let's remember that there are tools other than compilers,
and you don't have to rely on the compiler to enforce *all* your
good programming habits.  It is still possible for an organsation
that prefers the Ada 83 style to prohibit the creation of child
packages, and this could be enforced by the development system,
just as specific local standards concerning Use clauses or Unchecked
Programming can be enforced.

Re: Child packages [nn,pedo,incest,cons

[eachus]

Subject: Re: Child packages [nn,pedo,incest,cons]


     A couple of years ago the issue of permitting or forbidding
illegitimate children was strongly debated (at the Portsmouth
meeting). It was a close call, and I think the sentiment was in favor
of forbidding.

     However, there were several other needed changes in the
visibility model involving child units (involving who could see whose
private parts).  Tucker's (elegent) solutions to those forced the
morality (and legality) of illegitimate children to become an
environmental issue, while mitigating the harm that they can do.

     As an early opponent of illegitimate children, I support the
current position.  Yes, you can shoot yourself in the foot by creating
an illegitimate child, but it is hard to shoot others feet as well.


--

					Robert I. Eachus

with Standard_Disclaimer;
use  Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...

Re: Child packages [nn,pedo,incest,cons

[Matt Kennel]

firth@sei.cmu.edu wrote:
: Subject: Child packages [nn,pedo,incest,cons]

: In article <1994Oct14.094121@di.epfl.ch> Robb.Nebbe@di.epfl.ch (Robb Nebbe) writes:

: >|>    The private part and body of any child can access the private part of
: >|>    its parent (and grandparent...).

: In principle, any mechanism that allows one to extend an abstraction
: also allows one to break it.  That's a hard choice for the language
: designer to make.

How?  You have an original thing A that is "not broken".

In most OO languages, you can make B which is "like A"
except for blah blah blah.  Of course, the new stuff that you
do might happen to "break" the original intent of A.

But if your original program used A, and just A, how could A now
be broken?

--
-Matt Kennel  		mbk@inls1.ucsd.edu
-Institute for Nonlinear Science, University of California, San Diego
-*** AD: Archive for nonlinear dynamics papers & programs: FTP to
-***     lyapunov.ucsd.edu, username "anonymous".