Parts should know their whole

Hi,

I had a post related to the topic, see
http://groups.google.com/group/comp....ow+whole&fwc=2

Now after thoughts, I believe parts should know their whole, which
object they "belong to" (one to one or many to one relation).

Say team members are parts, and team is the whole. If the team knows
members but members don't know the team, then communication has to be
done through a super powered third part.

This is not effective. And is not democracy.
Program complicated at once.

So the smalls need to know the big.

Now I add instance variables to parts giving access to the whole while
I believe this should be done deeper in the language itself.

Regards,

Jim G

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.


Reductio ad absurdum is a wonderful tool for reasoning.

--
Richard Sargent
5x5 Computing Solutions, Inc.
rsargent@5x5.on.ca
http://www.pendragonfarm.com/
"Emptist" <jimg1968@gmail.com> wrote in message
news:a8b67b14-ad91-44be-925e-abf1b2e8c87c@2g2000hsn.googlegroups.com...
> Hi,
>
> I had a post related to the topic, see
> http://groups.google.com/group/comp....ow+whole&fwc=2
>
> Now after thoughts, I believe parts should know their whole, which
> object they "belong to" (one to one or many to one relation).
>
> Say team members are parts, and team is the whole. If the team knows
> members but members don't know the team, then communication has to be
> done through a super powered third part.
>
> This is not effective. And is not democracy.
> Program complicated at once.
>
> So the smalls need to know the big.
>
> Now I add instance variables to parts giving access to the whole while
> I believe this should be done deeper in the language itself.
>
> Regards,
>
> Jim G
>

> Should a wallet know whether it is in a pocket or a purse? A front pocket,
> a back pocket, a jacket pocket?

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

--g

On Sep 4, 7:20*am, "Richard Sargent" <rsarg...@5x5.on.ca> wrote:
> 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.
>
> Reductio ad absurdum is a wonderful tool for reasoning.
>
> --
> Richard Sargent
> 5x5 Computing Solutions, Inc.
> rsarg...@5x5.on.cahttp://www.pendragonfarm.com/"Emptist" <jimg1...@gmail.com> wrote in message
>
> news:a8b67b14-ad91-44be-925e-abf1b2e8c87c@2g2000hsn.googlegroups.com...
>
> > Hi,
>
> > I had a post related to the topic, see
> >http://groups.google.com/group/comp....se_thread/thre...
>
> > Now after thoughts, I believe parts should know their whole, which
> > object they "belong to" (one to one or many to one relation).
>
> > Say team members are parts, and team is the whole. If the team knows
> > members but members don't know the team, then communication has to be
> > done through a super powered third part.
>
> > This is not effective. And is not democracy.
> > Program complicated at once.
>
> > So the smalls need to know the big.
>
> > Now I add instance variables to parts giving access to the whole while
> > I believe this should be done deeper in the language itself.
>
> > Regards,
>
> > Jim G

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

On Sep 4, 3:50*pm, "geoff" <nos...@nospam.com> wrote:
> > Should a wallet know whether it is in a pocket or a purse? A front pocket,
> > a back pocket, a jacket pocket?
>
> . . . and the answer is yes if it is necessary to solve the prpblem and no
> if one is doin it just do it.
>
> --g

I agree. Tailors love these wise pockets, after all

J G

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

Emptist wrote:
> Hi,
>
> I had a post related to the topic, see
> http://groups.google.com/group/comp....ow+whole&fwc=2
>
> Now after thoughts, I believe parts should know their whole, which
> object they "belong to" (one to one or many to one relation).
>
> Say team members are parts, and team is the whole. If the team knows
> members but members don't know the team, then communication has to be
> done through a super powered third part.
>
> This is not effective. And is not democracy.
> Program complicated at once.
>
> So the smalls need to know the big.
>
> Now I add instance variables to parts giving access to the whole while
> I believe this should be done deeper in the language itself.
>
> Regards,
>
> Jim G
>


--
The surest sign that intelligent life exists elsewhere in Calvin &
the universe is that none of it has tried to contact us. Hobbes.
--
Eliot ,,,^..^,,, Smalltalk - scene not herd

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
dates of the employment. With multiple people working for multiple
companies, maybe returning to the same company several times, a simple
whole->part or whole<->part solution wouldn't work.

James E. Rumbaugh: Relations as Semantic Constructs in an Object-Oriented
Language. OOPSLA 1987: 466-481

Steve

"Eliot Miranda" <eliotm@pacbell.net> wrote in message
news:VGVvk.20557$mh5.6971@nlpi067.nbdc.sbc.com...
> Edwin H. Blake, Steve Cook: On Including Part Hierarchies in
> Object-Oriented Languages with an Implementation in Smalltalk.
> ECOOP 1987: 41-50
>
> Emptist wrote:
>> Hi,
>>
>> I had a post related to the topic, see
>> http://groups.google.com/group/comp....ow+whole&fwc=2
>>
>> Now after thoughts, I believe parts should know their whole, which
>> object they "belong to" (one to one or many to one relation).
>>
>> Say team members are parts, and team is the whole. If the team knows
>> members but members don't know the team, then communication has to be
>> done through a super powered third part.
>>
>> This is not effective. And is not democracy.
>> Program complicated at once.
>>
>> So the smalls need to know the big.
>>
>> Now I add instance variables to parts giving access to the whole while
>> I believe this should be done deeper in the language itself.
>>
>> Regards,
>>
>> Jim G
>>
>
>
> --
> The surest sign that intelligent life exists elsewhere in Calvin &
> the universe is that none of it has tried to contact us. Hobbes.
> --
> Eliot ,,,^..^,,, Smalltalk - scene not herd

On Sep 5, 2:21*am, Eliot Miranda <eli...@pacbell.net> wrote:
> Edwin H. Blake, Steve Cook: On Including Part Hierarchies in
> Object-Oriented Languages with an Implementation in Smalltalk.
> ECOOP 1987: 41-50
>
>
>
> Emptist wrote:
> > Hi,
>
> > I had a post related to the topic, see
> >http://groups.google.com/group/comp....se_thread/thre...
>
> > Now after thoughts, I believe parts should know their whole, which
> > object they "belong to" (one to one or many to one relation).
>
> > Say team members are parts, and team is the whole. If the team knows
> > members but members don't know the team, then communication has to be
> > done through a super powered third part.
>
> > This is not effective. And is not democracy.
> > Program complicated at once.
>
> > So the smalls need to know the big.
>
> > Now I add instance variables to parts giving access to the whole while
> > I believe this should be done deeper in the language itself.
>
> > Regards,
>
> > Jim G
>
> --
> The surest sign that intelligent life exists elsewhere in * * *Calvin &
> the universe is that none of it has tried to contact us. * * * Hobbes.
> --
> Eliot * * ,,,^..^,,, * *Smalltalk - scene not herd

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.

On Sep 5, 4:36*pm, "Steven Kelly" <ste...@metacase.com> wrote:
> 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
> dates of the employment. With multiple people working for multiple
> companies, maybe returning to the same company several times, a simple
> whole->part or whole<->part solution wouldn't work.
>
> James E. Rumbaugh: Relations as Semantic Constructs in an Object-Oriented
> Language. OOPSLA 1987: 466-481
>
> Steve
>
> "Eliot Miranda" <eli...@pacbell.net> wrote in message
>
> news:VGVvk.20557$mh5.6971@nlpi067.nbdc.sbc.com...
>
> > Edwin H. Blake, Steve Cook: On Including Part Hierarchies in
> > Object-Oriented Languages with an Implementation in Smalltalk.
> > ECOOP 1987: 41-50
>
> > Emptist wrote:
> >> Hi,
>
> >> I had a post related to the topic, see
> >>http://groups.google.com/group/comp....se_thread/thre....
>
> >> Now after thoughts, I believe parts should know their whole, which
> >> object they "belong to" (one to one or many to one relation).
>
> >> Say team members are parts, and team is the whole. If the team knows
> >> members but members don't know the team, then communication has to be
> >> done through a super powered third part.
>
> >> This is not effective. And is not democracy.
> >> Program complicated at once.
>
> >> So the smalls need to know the big.
>
> >> Now I add instance variables to parts giving access to the whole while
> >> I believe this should be done deeper in the language itself.
>
> >> Regards,
>
> >> Jim G
>
> > --
> > The surest sign that intelligent life exists elsewhere in * * *Calvin &
> > the universe is that none of it has tried to contact us. * * * Hobbes.
> > --
> > Eliot * * ,,,^..^,,, * *Smalltalk - scene not herd

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.
The mechanism of the language, such as the "variable" mechanism, may
need to be changed a bit to enable two-way awareness of relationship.
Or, maybe some further adjustments. But I am not able to describe it
clearer, since I know little of languages at so deep a level.

Best Regards,

Jim G

"Steven Kelly" <stevek@metacase.com> wrote in message
news:gc6wk.51$ig.31@read4.inet.fi...
> 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

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

--
Richard Sargent
5x5 Computing Solutions, Inc.
rsargent@5x5.on.ca
http://www.pendragonfarm.com/