Well I guess a cluster of quantum computers could translate much faster than a programmer.
FWIW if someone told me about modern android specs 20 years ago, I'd probably think they are mad.

2 members found this post helpful.

I actually find it a great boon for typing exactly because of its location. I don't have to move my fingers from the home row on the keyboard to click something. Mouse is the worst regarding that because I have to completely remove my hand from the keyboard to move it. Touchpad is closer, but it's not as precise as mouse AND trackpoint.

I believe the programmer years I quoted are actual for the current LoC count. And I doubt it would take less time to rewrite as a programmer must first read and understand the old code and only then write the new code.

Actually, any computer could and already can do it. Whenever you compile the source the compiler produces machine code. Any compiler can be instructed not to compile all the way there but to generate (optimized) assembly code instead.

In fact, that way the entire operating system could be automatically rewritten to assembly without much effort. The new assembly code would be bulkier than handcrafted one, but for most code it wouldn't matter. Someone could profile the code and optimize for size and speed where necessary, but I guess 80-90% of automatically generated code would be fine. New makefiles should be made to assemble everything and that would be it.

Therefore the task is actually easier than it looks, but the main question still remains, why? Assembly code is harder to maintain and it is incompatible with different CPU architectures.

I wouldn't consider myself a real programmer, but when I do create my own scripts, I find developing the logic and getting that in code takes the longest amount of time. Since that's already done, I would think the time to translate from one language to another wouldn't take nearly as long as it took to originally develop the logic and write the code (provided they have a strong understanding of both languages).

1 members found this post helpful.

Sure it can, but if you read recent news from ibm; few years from now this translation could be done in miliseconds.

People shouldn't be allowed to lie in Operating Systems forums or programming forums.

I think the same as you.

Thanks for posting.

Me thinking it would be a bit faster to rewrite code in a new language than writing it from scratch is by no means me believing that this project has any worth.

It would be a colossal waste of time for very little benefit. You're probably talking hundreds of thousands programmer years to rewrite the entirety of 15.0 to assembly... nobody is going to do that.

If this will happen, probably you will be quite soon banned for life - at least in this forum.

Honestly, I do not believe that you even have a vague idea of Assembler Programing. Not even a clue.

You are just amusing yourself showing us a crazy idea and/or you are a 10 years old. At least mentally.

3 members found this post helpful.

Having worked in a few programming languages - and having had to read and debug programs in a few that I do not typically write in - my opinion is that translating a program into a new language is often not the best approach.

The original program will have been written (hopefully!) to optimize the code using the constructs available in that language. A different language may have additional or different constructs that either 1) work better, or 2) make it much clearer what the code is trying to do. If you want maintainable code, then clear code is better than obfuscated code translated from a different language.

Ancient example: trying to translate matrix manipulation from arrays in FORTRAN to pointers in C. Or working with ancient FORTRAN code that only has DO loops, and realizing that much of it would be far better off in a language with for-while, or repeat-until or for-each-X-in structures. Better off understanding the original purpose of the code and re-writing it efficiently in the new language.

And compilers or interpreters try to optimize what they are working with as source. I expect they do a better job at optimizing code that is well-written in that language, rather than some bastardized attempt to make that language work the way that some other language works.

Of course, modern programs make extensive use of existing libraries, modules, widgets, etc. that are often coded in a completely different language.

My impression as a newbie here is that many of the participants speak more than one (verbal, not computer) language. When speaking in your second (or third, or fourth) language, do you speak more efficiently when you think in your first language and translate into the one you want to speak, or when you brain starts off thinking in that second language? My first language is English, and I am fairly fluent in French, but I speak much more coherently in French if my brain is thinking in French. If I think in English and then try to figure out how to say it in French, it's garbled, unclear, slow, etc. Not a pretty sight.

1 members found this post helpful.

You start XD

1 members found this post helpful.

Lets rewrite the os into assembler code! Oh wait, then we need to worry about which arch were building on, so lets make a program that reads a more generic language and builds arch specific code. Oh wait, we just invented a higher level language and a compiler again...

I'm surprised this thread has lived this long tbh

Looks like we are a bunch of kind grownups who love to amuse this kid.

I wonder if his mommy knows that he posts crazy things in LQ...

1 members found this post helpful.

OOPS! Wrong thread.

Hello human,

Sometimes I translate stuff from english to spanish, sometimes from spanish to english, and sometimes I understand what something means in english but I don't know how to translate that to spanish preserving the full meaning of the words in english.

Thank you very much for your comments on source code translation ideas.

Have you checked the C++20 course that is available in freecodecamp?

Back in 2020 C++20 wasn't still fully implemented for use in Visual Studio Code.

I'll use CLion aniways, not Visual Studio Code.

Kind regards.


I'm not sure if I have lied... Everyone start asking questions (make another forum thread in an appropiate subforum) and I will give answers.

Huh, what are you talking about?