Tim,

> On a more practical point, the idea of using self-like messages to 
> define instance variables would be acceptable IFF those messages were 
> properly private. In order to provide this privacy one would need to 
> implement some mechanism to allow properly private methods 
> and the same mechanism would (very likely) solve the worries about 
> overly public methods already in the system. I'd be quite happy to 
> see such a privacy mechanism if anyone has good ideas.

Well, I'm working on such an extension of Smalltalk as a (quite
independent) part of the traits-v2 project. The whole thing already
exists in theory, but I haven't gotten around to implement it yet. The
idea is to make something that is much simpler, more flexible and more
intuitive than the C++/Java approach. It does not require any keywords
(such as public, protected or private), and at least the basic version
will be downwards compatible to existing ST. Furthermore, it will be
transparent in the sense that people who do not want to deal with it can
just "turn it off".

Oh, and just to make this clear: As traits-v1, the whole thing will be
just a publicly available research project that we are persuing in
collaboration with Andreas Raab. Whoever wants to use it can do so. But
there will be no pressure...

More to come,
Nathanael




> -----Original Message-----
> From: squeak-dev-bounces at lists.squeakfoundation.org 
> [mailto:squeak-dev-bounces at lists.squeakfoundation.org] On 
> Behalf Of Tim Rowledge
> Sent: Freitag, 11. April 2003 21:23
> To: The general-purpose Squeak developers list
> Cc: Tim Rowledge
> Subject: Re: Need feedback on simple idea
> 
> 
> 
> On Friday, April 11, 2003, at 12:01  PM, Swan, Dean wrote:
> > 	You also must consider that as a steward of the 
> "kernel", inquiries
> > about such interesting research topics coming from you can 
> no longer 
> > be "innocent inquiries".
> Exactly. Any form of 'authority' brings with it a sometimes onerous 
> burden of having to be very careful what you say.
> 
> On a more practical point, the idea of using self-like messages to 
> define instance variables would be acceptable IFF those messages were 
> properly private. In order to provide this privacy one would need to 
> implement some mechanism to allow properly private methods 
> and the same 
> mechanism would (very likely) solve the worries about overly public 
> methods already in the system. I'd be quite happy to see such 
> a privacy 
> mechanism if anyone has good ideas.
> 
> tim
> -- 
> tim at sumeru.stanford.edu
> 
>