| View previous topic :: View next topic |
| Author |
Message |
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54785
Location: 56N 3W
|
 Posted: Sun Jan 26, 2025 3:40 pm Post subject: |
 |
|
pingtoo,
There are lots of wee gems there, not just that.
| Fred Brooks wrote: | Plan to throw one away. You will anyway.
The bearing of a child takes nine months, no matter how many women are assigned to the task.
... |
I think that there is/was a fortune database of quotes.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
 |
wjb
l33t
Joined: 10 Jul 2005
Posts: 643
Location: Fife, Scotland
|
| Posted: Sun Jan 26, 2025 4:38 pm Post subject: |
|
|
re the last new bit because I don't understand what it's behaviour has to do with the operating system.
| Code: | | search_function(M[0-6]) |
Is this passing the whole array to search_function, and running it the way you'd expect in more traditional languages? In which case it's behaviour comes from it's implementation and it's interaction with the O/S.
Have you decided whether blocks are going to be allowed inside functions? |
|
| Back to top |
|
|
Goverp
Advocate
Joined: 07 Mar 2007
Posts: 2201
|
| Posted: Mon Jan 27, 2025 8:47 am Post subject: |
|
|
Gentoopc, you might like to investigate Occam, though I expect you will be able to see why it's not the same as your concept.
_________________
Greybeard |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 8:47 am Post subject: |
|
|
| wjb wrote: |
| Code: | | search_function(M[0-6]) |
|
Apparently I copied my text to add to the topic and didn't correct it. |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 8:55 am Post subject: |
|
|
| Goverp wrote: | | Gentoopc, you might like to investigate Occam, though I expect you will be able to see why it's not the same as your concept. |
My concept is simple. We must adapt what could work 50-60 years ago to the present. Earlier, when operating systems were not multitasking, perhaps line-by-line execution of code was correct. But now we need flexibility. What can we change? We do not have the strength to redo everything. Well, let's break everything into blocks. Inside the blocks, we will execute the code line by line, and we will be able to use the blocks themselves in any sequence with any priority. This will give flexibility. This will give the ability to run code in blocks simultaneously, in parallel. I suggest making at least a small step in 50 years in this direction. Many here refuse to do this. |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 9:04 am Post subject: |
|
|
| wjb wrote: |
Have you decided whether blocks are going to be allowed inside functions? |
we need to consider all this carefully. I think that the body of the function can be divided into blocks inside, but the priority of block execution should not be inconsistent with the priority of starting the function, if this happens there should be a warning, and the function inside its case, works in the normal mode without any priorities. that is, an error in priorities, cancels them all. this will not break the code, in the case when the programmer does not understand what he is doing. there are more important problems now, this is to develop a more flexible system of structures. I think that one
not enough. this needs to be enriched. I think that the methods of structures, if there are many structures, can be combined separately into structures. this will give something like a type class as in Haskell. |
|
| Back to top |
|
|
wjb
l33t
Joined: 10 Jul 2005
Posts: 643
Location: Fife, Scotland
|
| Posted: Mon Jan 27, 2025 12:00 pm Post subject: |
|
|
I don't think you've finished 'priority' yet.
If I understand where you're at, priority as a number that steers a block content to a core in low to high order? Ok, so there's effectively a queue of operations going to a core that's capable of being written in any order, potentially hidden in functions, and also interleaved with stuff for other cores. Consequently it's quite hard to work out what's supposed to happen even in small sections such as your example. |
|
| Back to top |
|
|
wjb
l33t
Joined: 10 Jul 2005
Posts: 643
Location: Fife, Scotland
|
| Posted: Mon Jan 27, 2025 12:05 pm Post subject: |
|
|
@Goverp
Yeah, and there's a proprietary descendant of Occam called "xc" with a more C-like syntax. It has 'tile's that get allocated resources, like specific cores, then you allocate tasks to tiles. |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 12:43 pm Post subject: |
|
|
| wjb wrote: | I don't think you've finished 'priority' yet.
If I understand where you're at, priority as a number that steers a block content to a core in low to high order? |
priority on the order of execution of a block of code. let's say we want to execute many blocks at once, we will give them priority, or otherwise the order of execution, like first, go, second go, third go..... in case of setting the priority, the kernel and processor schedulers should be guided by what the programmer did. you are programmers, not computer operators. you should program, not the compiler. if for some reason I need the code to be executed on certain processor cores, then it should be so. Linux kernel schedulers can be easily written in exactly these programming languages. well, this is an example. |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 12:53 pm Post subject: |
|
|
| wjb wrote: | | IOk, so there's effectively a queue of operations going to a core that's capable of being written in any order, potentially hidden in functions, and also interleaved with stuff for other cores. Consequently it's quite hard to work out what's supposed to happen even in small sections such as your example. |
this is already a deliberate pretending that you don't understand in order to say that the idea is bad. the idea has nothing to do with your inability to understand simple things. you are writing a function
| Code: |
void My_function(){
int a & 100;
int b & 300;
_block(0){ ++a; };
_block(1){++ b;};
};
|
what's so difficult about this? there will be a parallel increase in the value of variables on different cores. these variables are stack variables, otherwise they are local. but you will do actions with them simultaneously and this will not cause problems. if someone is a complete idiot and cannot program, then no programming language will be able to help him. anyone who programs must understand what he is doing. you cannot try to do and think through everything for him. a programmer must be free in his actions and understand their meaning. otherwise he is not a programmer. and not a person.
}; |
|
| Back to top |
|
|
Gentoopc
Guru
Joined: 25 Dec 2017
Posts: 406
|
| Posted: Mon Jan 27, 2025 1:01 pm Post subject: |
|
|
I will repeat once again that at the moment we need a programming language that will have a simple syntax conditioned by mathematics, logic or common sense. the core of the programming language is as simple as possible. but the language can be extended by type classes, as a combination of structure methods into separate structures, on the basis of which it will be possible to create new complex types that can be processed by many methods that we will group. this is for advanced. but the point is that everyone will be able to find in it what they need. you will be able to solve a lot of problems with parallel execution. maximum compatibility with C++ / C. this is just what you need. it can be yours, because you do it. no one will reproach you for using something non-standard. you have finalized, polished, and it is. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5348
Location: Bavaria
|
| Posted: Mon Jan 27, 2025 2:38 pm Post subject: |
|
|
Moderator note:
Our admin @NeddySeagoon has already warned in this post:
https://forums.gentoo.org/viewtopic-p-8852890.html#8852890
After that, serious and veteran users have written interesting posts. I had hoped that this thread would not take the same course as the other one - and that this thread would go to sleep peacefully if interest in the topic was lost.
Unfortunately, this - repeated - insinuation that other users are incapable of understanding something is no longer tolerable. I am therefore closing this thread.
_________________
https://wiki.gentoo.org/wiki/User:Pietinger |
|
| Back to top |
|
|
|
Portage & Programming
