[ANSI-Smalltalk] A list of interesting methods, not present in the ANSI standard

Andres Valloud andres.valloud at gmail.com
Sun Mar 16 21:17:57 GMT 2008


Bruce,

The intent of going through this list is as expressed before: to propose the
inclusion of de-facto standard messages into the actual standard, as long as
we all agree of course.

Just in case... out of the list for which you ask questions (e.g. on deep
copy, broadcasting etc), I pruned out most of the methods and boiled it down
to the considerable smaller list which I think has a larger chance of
gaining consensus.  The shorter list does not have deepCopy, displayOn:,
dependencies, broadcasting, and so on.

Perhaps become: is problematic, and this is something I'd like to figure
out.  Isn't become: two-way in all Smalltalks by now?

The message identityHash is necessary for (customary) implementations of
things like MethodDictionary, IdentityDictionary, and IdentitySet.  I do not
know of any Smalltalk that does not use identityHash for these purposes.

Andres.

On Sun, Mar 16, 2008 at 2:03 PM, Bruce Badger <bwbadger at gmail.com> wrote:

> Some questions occur to me ...
>
> o is become: one way or two way?
> o What is broadcast?
> o is there an agreed way to do a "deep copy"?
> o ditto for dependencies
> o is allOwners as obvious as it appears?  Does GC interfere?
> o what does displayOn: do differently from printOn:?
> o is there an agreed way to handle immutability of objects (assuming
> that's what isReadOnly is about)
> o what does storeLiteralOn: do?
> o When do we need identityHash isn't == based on object identity?
>
> In general, I wonder if big lists of proposed selectors are a good
> starting point.  They seem to be an expression of a solution (or part
> of one, or of more than one) without saying what they are trying to
> solve.
>
> Perhaps we can express these things as (sort of, kind of) aspects.
> So, the identity aspect of an Object includes == and perhaps identity
> hash?  The dependency aspect etc etc
>
> Perhaps breaking things down in this way will let us express an
> *intent* before we go ahead and propose a solution?
>
> All the best,
>    Bruce
> --
> Make the most of your skills - with OpenSkills
> http://www.openskills.org/
>
> _______________________________________________
> ANSI-Smalltalk mailing list
> ANSI-Smalltalk at lists.openskills.org
> http://lists.openskills.org/cgi-bin/mailman/listinfo/ansi-smalltalk
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openskills.org/pipermail/ansi-smalltalk/attachments/20080316/9103de06/attachment.htm


More information about the ANSI-Smalltalk mailing list