From Newsgroup: comp.lang.c++

https://sourceforge.net/projects/cscall/files/MisFiles/ClassGuidelines.txt/download
... OO can have many meaning. I took the object to mean the basic entity of the programming model. The concept of object, as a foundmental concept, has to be solid, practical and easily usable. Otherwise, more codes and efforts will be needed latter to fix it making the original goal, practically, a lie.
IOW, (nearly) a flawless model is all the basics...
After 25 years, I feel the idea of OO is lost. Something not right with OO?
I'd rather believe the idea of OO being not followed or addressed is the culprit.
--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.c++

wij wrote:

https://sourceforge.net/projects/cscall/files/MisFiles/ClassGuidelines.txt/download
... OO can have many meaning. I took the object to mean the basic entity of the
programming model. The concept of object, as a foundmental concept, has to be solid, practical and easily usable. Otherwise, more codes and efforts will be needed latter to fix it making the original goal, practically, a lie.
IOW, (nearly) a flawless model is all the basics...

After 25 years, I feel the idea of OO is lost. Something not right with OO? I'd rather believe the idea of OO being not followed or addressed is the culprit.

its lost from the begining...same as so called polymorphism etc

the problem in opp is those pointers you make and carry (pas in and out)

the right idea i call "hipermodularity" or "miclomodularity" and i wrote
a lot on this mainly on c.l.c

here if you have a

module Smth
{
int x,y;
int foo() {}
int goo() {}

}

its just liek structure but it als has code in it
the instances are visible right away just like in normal structures,
you also can pout it in array

Smth table[10]

and so on (by so on i mean like definig operators also)

simply no new and no pointer passing..there is more to it but
its a longer stream of insights and details

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.c++

On Wed, 2024-10-30 at 12:29 +0800, wij wrote:

https://sourceforge.net/projects/cscall/files/MisFiles/ClassGuidelines.txt/download
... OO can have many meaning. I took the object to mean the basic entity of the
programming model. The concept of object, as a foundmental concept, has to be solid, practical and easily usable. Otherwise, more codes and efforts will be needed latter to fix it making the original goal, practically, a lie.
IOW, (nearly) a flawless model is all the basics...

After 25 years, I feel the idea of OO is lost. Something not right with OO? I'd rather believe the idea of OO being not followed or addressed is the culprit.

General rule of the element of a programming language:
If too much effort is spent in figuring out what keywords/pseudo-directive should be in the program. Something is wrong. E.g.
T* a=b;
The target is just "mov a,b", anything more may be considered redundant and inefficient. Now we have const, constexpr, volatile/atomic,promotion/narrowing,mutable,modifier/attr..., and maybe isnull, ownership,...) to consider in programming "mov a,b" for "T a=b".
Why not using assembly? Well, after not much tries, we found what we wanted was not much different from C, but with something compromised (e.g, stack,...),
and mostly, some classified as hardware 'specifics' (ironically, differences  are the reason different hardwares exist. we use new hardware because it has something new).
Does it have to be C, afterall hardware can be redesigned to align with high level languages? Answer: At least we don't have idea of any machine that can exceed Turing Machine (quantum computing is just faster in some aspect). Assembly,C is pretty much close to the description of TM...
In this stage, I would say C is desirable is because it is close to human natural language of which the feature is recursion,modifiers in 'sentence' (not list of instructions):
https://study.com/academy/lesson/noam-chomsky-on-language-theories-lesson-quiz.html
... all languages have formal universals and principles in common, with specific options and limits
for variation in grammar and features between languages...
But too close to natural language should not be the goal because uncertainty should and must exist in human language... So the language of a model.
I had a initial test: https://sourceforge.net/projects/symbcomp/
I feel programming language (C/C++) can do more, and can even be better targeting as the foundation of math/logic. Reason: it models the real thing, better than math. and logic.
--- Synchronet 3.20a-Linux NewsLink 1.114