A question for OOP experts. - page 28

 
Anyway, I'll develop it and if there are any interesting developments, I'll share them. Everything is more interesting than discussing trivial matters. You could say it's an experiment.
 
Реter Konow:
Yes, it's simple. You and many others look at programming "from the inside". It's so big and you're all inside of it. It is the main thing and you obey its rules. I look at programming from the outside. It's just my creativity. I mean, I'm bigger than ONE, and I can do what I want. Well, it's everyone's choice.

The level of a child playing in the sandbox - everything is at his fingertips: castles, roads, pies. But to raise his eyes and look around - either his age is not right or fear prevents him from doing so.

 
Реter Konow:

My approach is also an OOP. Only in a different presentation and less "objectivity" at the moment. But the essence is the same.

1. encapsulation in the kernel is present.

2. Partial polymorphism. That is, the engine works with the Windows and the elements as such, but in each block there is an insertion of conditions that define work with specific types of elements.

3. At the moment there is no inheritance of the properties of objects, because the graphical core does not provide hierarchy. But this can be fixed.

I want to create a "lightweight" OOP based on kernel and engine, with minimal syntax, with digital representation of objects in the kernel, full hierarchy, classification, inheritance of properties and functions. That is, the same OOP, but mine.

"Lightweight" OOP - A meaningless phrase.

 
Реter Konow:
Can you imagine the advantage of a digital representation of objects? They can be multiplied and multiplied without stopping. Not references to classes, not instances of classes, but full-fledged objects described with all possible properties. You can't do that easily with standard OOP .

A false claim.

 
Реter Konow:
Anyway, I'll develop it and if there are any interesting developments, I'll share them. Everything is more interesting than discussing trivial matters. You could say it's an experiment)).
Look at "knowledge representation systems" for inspiration.
 
Very interesting, how are "complete" objects different from class instances?
 
Aliaksandr Hryshyn:
Check out 'knowledge representation systems' for inspiration.

Very useful article. Thank you.

https://habr.com/ru/post/346236/

 
Dmitry Fedoseev:

Read https://ru.wikipedia.org/wiki/Представление_знаний

This area of programming is experimental and there are no absolute canons. Everyone can invent their own approach to representing knowledge and creating artificial cognitive systems. The field is unconquered and there is much to seek in it. The standard approach may be of little use. There are only a couple of lines about using OOP in building expert systems.

Представление знаний — Википедия
  • ru.wikipedia.org
Представление знаний — вопрос, возникающий в когнитологии (науке о мышлении), в информатике и в исследованиях искусственного интеллекта. В когнитологии он связан с тем, как люди хранят и обрабатывают информацию. В информатике — с подбором представления конкретных и обобщённых знаний, сведений и фактов для накопления и обработки информации в...
 
Реter Konow:

Read https://ru.wikipedia.org/wiki/Представление_знаний

This area of programming is experimental and there are no absolute canons. Everyone can invent their own approach to representing knowledge and creating artificial cognitive systems. The field is unconquered and there is much to seek in it. The standard approach may be of little use. There are only a couple of lines about using OOP in building expert systems.

Read it and what?

 
Реter Konow:

Very useful article. Thank you.

https://habr.com/ru/post/346236/

An informative article.

Retag Konow:

Read https://ru.wikipedia.org/wiki/Представление_знаний

This field of programming is experimental and there are no absolute canons. Everyone is free to invent his or her own approach to representing knowledge and creating artificial cognitive systems. The field is unconquered and there is much to seek in it. The standard approach may be of little use. There are only a couple of lines about using OOP in building expert systems.

But this is how you want to do it:




SZZ: I wanted to make it easier for you and suggest you look at a ready-made solution, but what's the point? You can screw in your vision of knowledge there too, but just in case, google CLIPS .... just in case?

Reason: