|
|
|
| Authored by: BitOBear on Sunday, May 06 2012 @ 04:49 AM EDT |
Imagine again that we are moving from one single family house to another.
An organizational genious labeled a box "white powders" and put
everything that is white and powdery in the one box.
Would it work? Sure. When I was looking for the sugar, or flower, or baking
powder or salt I would know exactly what box to go to.
But when I was looking for the concrete, the "rose bush dust"
(pesticide), rat poison, and plaster of Paris I would be looking in that same
box.
Now I could separate the baking supplies and the home maintenance supplies into
smaller interior boxes to keep the poison and the plaster out of dinner.
The system would work.
So now, if I am making chocolate cupcakes I would need to get a bunch of stuff
from the white_powders.baking_supplies box. -however- I would also need to go to
black_powders.baking_supplies for the cocoa powder as well.
Now over the centuries we have learned that keeping the poisons and the foods
separate is a good thing, so we wouldn't do this. We have also learned that
keeping all the baking supplies in one place, like the kitchen, and keeping all
the household construction supplies somewhere else like the garage, are good
things to do.
So while systems of organization -can- be created in many ways, there are
"more natural" and "less natural" organizations. There are
also some organizational choices that are fraught with danger. You just don't
store gasoline next to the water heater inside a house.
We also know what it is like to be in a strangers house. If you go into their
kitchen you have certain expectations on what cabinets and drawers you would
need to open to make a glass of chocolate milk. Glasses tend to be up high.
Spoons in a drawer near the counter. That sort of thing.
We know the kitchen API.
Now Harmony -could- have put the flatware under the sink and the garbage can on
top of the fridge, but why on earth -would- they when everybody already knows
more or less where things go?
Now computers are stupider than people because they can not think at all, so
they cannot tolerate even simple changes like putting the drawer full of
flat-ware out in the china hutch in the dining room. Sure its a change that
might make sense, particularly if you have "good silver" and
"daily flatware" like when I was growing up.
Nobody would change an already known API just because they -could-, at least
nobody sane would. That's the computer science equivalent of slashing your own
tires right before work just so everybody in human resources knows not to mess
with your sick-days. It just -doesn't- happen in a sane world.[ Reply to This | Parent | # ]
|
| |
| Authored by: Ian Al on Sunday, May 06 2012 @ 05:47 AM EDT |
From your 'White powder' thread, it is political correctness gone mad to put
chocolate coloured powder into the 'black' package. That aside, a few, more
appropriate, observations.
One point I take away from the comment is that computer packages are not, by and
large, modelled on the real world. They are man-conceived boxes and sub-boxes
and the set of properties for each box is not inherited, but classified by the
platform designer. The boxes are unlike anything in the real world. You can
create new objects in the box. The box gives the new object the properties that
the designer of the box printed on the outside of the box.
Objects created in the boxes at the beginning of the process are all given their
properties by the object designer and not by inheritance from the higher level
boxes. The designer gives the objects their properties according to his
inheritance design principles.
Once the platform is largely complete, all the properties allocated to the box
by the platform designer are given to any new objects created and put in the
box. The inheritance is by design rather than a process deriving properties from
other boxes using inheritance principles.
The interface is just a modification of the designer's inheritance model to
introduce multiple inheritance. It does not introduce a new inheritance process.
It just modifies the 'inheritance' model so that more freedom is given to the
allocation of property sets given to objects created in a specific box.
To return to our sheep, the box designer uses skill and judgement and creative
expression to decide which properties each box will impart. The decision to put
one box inside another so that a box inside another box will impart all the
properties of the outer box is part of that creative expression.
There are one or two rules about whether you can protect that creative
expression with copyright law. First and foremost, the creative expression must
be fixed in a medium. If the concept of boxes in boxes representing an
inheritance concept are not so fixed in a medium - no protection, whatsoever.
So, is the Java API inheritance model fixed in a medium or is it only visible as
a concept when someone points out that it is packages in an inheritance model?
It is the latter. The packages, themselves, are a concept because they are not
fixed in a medium. The concept of the packages containing sub-packages on an
inheritance model is a concept related to the concept of packages. Neither the
packages nor the inheritance model can be protected by copyright.
Oh, yes, before I forget, if the use of the detail in the package concept is
essential for compatibility and functionality for programmers to use the
language platform then, for that reason alone, it is not protectable by
copyright law - not in the inheritance model designers wildest imaginings.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
| |
| Authored by: Anonymous on Sunday, May 06 2012 @ 03:42 PM EDT |
Interfaces are the primary means of interaction with Windows COM, which would
have been designed right around the same time as Java (considering it came out
the same year as part of Windows 95).
[ Reply to This | Parent | # ]
|
|
|
|
|
