comp.lang.smalltalk Google Group

Re: Parts should know their assembly from a whole - (sometimes, maybe)

j...@cruzio.com (cstb) — Sat, 06 Sep 2008 00:56:14 UT

Three cases:
1. "part"s of an integral "whole"
a. initial construction from parts results in an immutable whole.
b. removing any such part destroys/invalidates the whole.
2. "part"s is "part"s
a. the "whole" is merely short hand
for "all of the parts" of a group,
and adding/removing/replacing parts has no effect

Re: Parts should know their whole

rsarg...@5x5.on.ca (Richard Sargent) — Fri, 05 Sep 2008 23:46:15 UT

Ditto for a Child/Student is Enrolled in a School.

Re: Parts should know their whole

jimg1...@gmail.com (Emptist) — Fri, 05 Sep 2008 15:58:53 UT

Hi, Steven,
We may find that objects only live in relations, while a single object
might require a lot of relationship class to fulfill its functions.
Object-oriented might be better being relation-oriented. However, by
using abstract relations out of objects approach, the system can still
be very complicated.

Re: Parts should know their whole

jimg1...@gmail.com (Emptist) — Fri, 05 Sep 2008 15:21:23 UT

Thanks for your link, Eliot.
If I understand them right, it seems that their extension pay much
attention to the "HAS PARTS" side of the problem and parts won't know
much of the whole.

Re: Parts should know their whole

ste...@metacase.com (Steven Kelly) — Fri, 05 Sep 2008 08:36:28 UT

Sometimes you want to move the connection between "part" and "whole" out of
both of them, to become its own "relationship" class. Each relationship
points to a "whole" and a "part" (or "parts"). This also allows more
information to be stored about the relationship, e.g. an "employs"
relationship could specify the employer, the employee, and the start and end

Re: Any one who notices that there's been so many web views to this group?

ssefl...@somewhere.com (Malbs) — Fri, 05 Sep 2008 02:19:58 UT

I dont think it means anything. 10k isn't very many imo.. but if I was to
guess..
Maybe Chrome?

Re: Parts should know their whole

eli...@pacbell.net (Eliot Miranda) — Thu, 04 Sep 2008 18:21:41 UT

Edwin H. Blake, Steve Cook: On Including Part Hierarchies in
Object-Oriented Languages with an Implementation in Smalltalk.
ECOOP 1987: 41-50

Re: Parts should know their whole

jimg1...@gmail.com (Emptist) — Thu, 04 Sep 2008 11:05:32 UT

I agree. Tailors love these wise pockets, after all ;)
J G

Re: Parts should know their whole

jimg1...@gmail.com (Emptist) — Thu, 04 Sep 2008 11:02:14 UT

Hi, Richard,
There are so many people whose name are similar. In almost all
circumstances it's good for names to know who the guy it really stands
for.
And it's great that a kid finally remembers which school to go :)
Regards,
Jim G

Re: Parts should know their whole

nos...@nospam.com (geoff) — Thu, 04 Sep 2008 07:50:24 UT

. . . and the answer is yes if it is necessary to solve the prpblem and no
if one is doin it just do it.
--g

Re: Parts should know their whole

rsarg...@5x5.on.ca (Richard Sargent) — Wed, 03 Sep 2008 23:20:17 UT

The counter-argument goes like this:
Should a dollar bill know which wallet it is in? Oh wait, it's not in a
wallet. It's in my pocket. Should a dollar bill know which container it is
in?
Should a wallet know whether it is in a pocket or a purse? A front pocket, a
back pocket, a jacket pocket?
etc.

Parts should know their whole

jimg1...@gmail.com (Emptist) — Wed, 03 Sep 2008 15:00:25 UT

Hi,
I had a post related to the topic, see
[link]
Now after thoughts, I believe parts should know their whole, which
object they "belong to" (one to one or many to one relation).

Any one who notices that there's been so many web views to this group?

jimg1...@gmail.com (Emptist) — Wed, 03 Sep 2008 14:47:40 UT

Hi, Smalltalkers,
Google tells me that there are more than 10000 web views to this group
last week.
What does this mean?

Re: The Builder Pattern to rule them all Re: Smalltalk syntax can rule them all

rein...@soops.nl (Reinout Heeck) — Tue, 02 Sep 2008 09:23:34 UT

Been there, done that -- it's a nightmare.
It mostly boils down to templates being 'dead code', you cannot do
'senders of' or 'references to variable' etc -- it makes development slower.
Hard coding the html generation but leaving the RSS free to be tweaked
by a design department appears to give a much more productive

Reminder: VA Smalltalk Forum Europe 2008 (23. September 08)

jtuc...@objektfabrik.de (Joachim Tuchel) — Mon, 01 Sep 2008 04:13:20 UT

Instantiations and objektfabrik cordially invite VA Smalltalk Customers
and people interested in VA Smalltalk to the
VA Smalltalk Forum Europe 2008
This event will take place in Frankfurt am Main on September 23rd,
starting at 10:00 am. We ask you to regster for this complimentary
event. More Information and regsitration form can be found either at