Jump to content

  • Log In with Google      Sign In   
  • Create Account


Just another thought (designing a fast but powerful language)


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
77 replies to this topic

#61 King Mir   Members   -  Reputation: 1945

Like
3Likes
Like

Posted 19 February 2014 - 05:05 PM

I wanted to ask this in a previous thread but I thought it needed it's topic.
Would it be possible to make a compiled language as fast as c and c++, multi paradigm like c++, less complex than them both and can be used like in really complex situations like in making an OS/game engine/IDE etc.
Just wondering (may have mistakes above).

Thanks.

That's a really broad mission statement. Do you have experience making OSs, game engines, or IDEs? If not, how do you hope to design a language for it? What are the problems with existing languages in these domains? Are they the same accross those domains?

 

I just want to say that designing a good language involves a lot of decisions that aren't even being remotely considered here.

 

Reducing LOC count is a ridiculous metric for the strength (or quality) of a language. You guys are all arguing about standard library contents and not even anything pertaining to actual language design.

 

 

I predict that any attempt at producing a language in this thread will end in tears.

QFT. A lot of the discussion has been about the size of the standard library, which is not something a new language is not going to have any way.

 

@apochPiQ: that's why i started the thread. You could just talk about some things that are more important.
Thanks smile.png

I can recommend two objectives you should add to your language design:

1) Facilitate features as a library. Boost show the benefits of having features that can be the implemented without a new language. But broadly, the ability to do this in existing languages is limited.

2) High level abstractions that you can peal away. This is the aim of  the STL and other libaries. You describe the function, not the implemenation, but the implementation is in the same language, so it can be fine tuned/swapped out when needed.

 

 


I need to think more.

There is an old saying I've been told for years that I think applies nicely to that: "Think less, code more."

I disagree with this. Design is an important, especially with a giant project like a compiler. Code is important, to be able to test out features, but Nathan2222 is no where near ready to write any code.

 

 

You still don't even know what problem(s) you are trying to solve. Unless and until you can name a concrete area in which your language will be different, you're just scrabbling around in the dark. And unless you can argue convincingly that that language difference is an advantage, nobody will ever care.

 

 

I started building a new programming language eight years ago. It still is far from finished and lacks even the most basic tools like a debugger. Granted, I don't spend huge chunks of time on it except for rare occasions, but that should give you some idea of what you're up against. Oh, and I started with a very clear list of things I wanted to solve and why my approaches were advantageous.

QFT.

 

I too have a compiler project on the back burner, and believe me it's no small task.

 

You need to be able to answer the question succinctly: Why should I use your language? You haven't done that in this thread.


Edited by King Mir, 19 February 2014 - 05:20 PM.


Sponsor:

#62 King Mir   Members   -  Reputation: 1945

Like
0Likes
Like

Posted 19 February 2014 - 05:17 PM

 



ultramailman, on 18 Feb 2014 - 4:32 PM, said:

Are you planning to write a compiler for this language you have in mind?

Yeah. An ide.
Won't i have to?

 

You won't have to. But it might be good to try writing a compiler for a simple and existing language first. After that, you should have more insight.

 

Chances are, a language will have enough basic features from an existing language, that you could just start on those. It's not that hard to chanrge syntax as you write a compiler -- the hardest challenge with doing that is updating tests. So you can start writing the language with what syntax you think it'll have, implementing those simple features you know your language will have in one fom or another. And it allows you to play with your language as early as possible.



#63 Nathan2222_old   Members   -  Reputation: -400

Like
0Likes
Like

Posted 19 February 2014 - 05:37 PM

ultramailman, on 18 Feb 2014 - 4:32 PM, said:
Are you planning to write a compiler for this language you have in mind?
Yeah. An ide.
Won't i have to?


You won't have to. But it might be good to try writing a compiler for a simple and existing language first. After that, you should have more insight.
Chances are, a language will have enough basic features from an existing language, that you could just start on those. It's not that hard to chanrge syntax as you write a compiler -- the hardest challenge with doing that is updating tests. So you can start writing the language with what syntax you think it'll have, implementing those simple features you know your language will have in one fom or another. And it allows you to play with your language as early as possible
the purpose of designing the language will be to enable anyone willing to spend a few hours learning a few things be able to create whatever they want.

I've been searching google throughout for things like AI programming languages (by AI, i don't mean languages used to program AI), smart languages, intelligent languages which led me to text vs visual programming (i really didn't know if there was a thing like visual programming till today) and then i started noticing the c vs c++ vs python vs other languages and started searching for things like has programming changed in 10 years. Some of these things may not be directly related but i did see several of my ideas spread over all of them.

No, i haven't made an OS, game engine, ide etc. but i want to do that before even trying to do make this language. My old signature showed that.
After making the language, i'll probably redo the projects in my language as a test.
I haven't programmed in two days probably because i think a lot and new ideas pop out and i leave my learning but i like coding.

UNREAL ENGINE 4:
Total LOC: ~3M Lines
Total Languages: ~32
smile.png
--
GREAT QUOTES:
I can do ALL things through Christ - Jesus Christ
--
Logic will get you from A-Z, imagination gets you everywhere - Albert Einstein
--
The problems of the world cannot be solved by skeptics or cynics whose horizons are limited by the obvious realities. - John F. Kennedy


#64 TheChubu   Crossbones+   -  Reputation: 4062

Like
6Likes
Like

Posted 19 February 2014 - 05:48 PM

In my thoughts, i've created a google chrome style browser with less than 300 lines of code.

In my thoughts I'm like a magical flying unicorn dude.


"I AM ZE EMPRAH OPENGL 3.3 THE CORE, I DEMAND FROM THEE ZE SHADERZ AND MATRIXEZ"

 

My journals: dustArtemis ECS framework and Making a Terrain Generator


#65 fir   Members   -  Reputation: -452

Like
1Likes
Like

Posted 19 February 2014 - 05:56 PM

After making the language, 

 

you will be probably at age of 90 then (you doing many assumptions about your superpowers)



#66 King Mir   Members   -  Reputation: 1945

Like
1Likes
Like

Posted 19 February 2014 - 09:33 PM

the purpose of designing the language will be to enable anyone willing to spend a few hours learning a few things be able to create whatever they want.

So you're making easy to learn as an objective. That's concrete, but then why is your language better than other languages with a similar goal, like python and Go? It also goes against some of your other objectives.

I've been searching google throughout for things like AI programming languages (by AI, i don't mean languages used to program AI), smart languages, intelligent languages which led me to text vs visual programming (i really didn't know if there was a thing like visual programming till today) and then i started noticing the c vs c++ vs python vs other languages and started searching for things like has programming changed in 10 years. Some of these things may not be directly related but i did see several of my ideas spread over all of them.

There is indeed quite a body of knowledge out there about language design. But you need to design to what you know, so IMO you should have a concrete niche for your language before really getting into research.

No, i haven't made an OS, game engine, ide etc. but i want to do that before even trying to do make this language. My old signature showed that.
After making the language, i'll probably redo the projects in my language as a test.
I haven't programmed in two days probably because i think a lot and new ideas pop out and i leave my learning but i like coding.

Making an OS, a game engine, or an IDE are big projects on their own. So is making a compiler for your language. Working on your own, part time, I wouldn't aspire to do more than one of them. You can dream, but planning things that far down the road is futile.

But my point was that you can't design a good language for these task if you don't know what the challenges in them are.

#67 BHXSpecter   Members   -  Reputation: 1411

Like
0Likes
Like

Posted 19 February 2014 - 09:40 PM

I disagree with this. Design is an important, especially with a giant project like a compiler. Code is important, to be able to test out features, but Nathan2222 is no where near ready to write any code.

I know design is important, but Nathan2222 is so lost in the theory of making a language and compiler that he is going to spend more time designing it, finding problems with it, redesign it, find more problems, etc and never get to the coding. Even Bjarne Stroustrup, in I believe it was his C++ Programming Language Special Edition, said it is better to get a foundation down quickly and then slowly add features and build on it rather than being stuck in an infinite design loop. I believe I've seen the same concept applied to video game programming too. That is why I stick to code more and think less, because I, myself, have stopped production on some projects or games because I would start finding things I didn't like and want to redesign them and then get lost in the design process.


Edited by BHXSpecter, 19 February 2014 - 09:41 PM.

"Through vengence I was born.Through war I was trained.Through love I was found. Through death I was released. Through release I was given a purpose."


#68 King Mir   Members   -  Reputation: 1945

Like
0Likes
Like

Posted 19 February 2014 - 10:11 PM

 

I disagree with this. Design is an important, especially with a giant project like a compiler. Code is important, to be able to test out features, but Nathan2222 is no where near ready to write any code.

I know design is important, but Nathan2222 is so lost in the theory of making a language and compiler that he is going to spend more time designing it, finding problems with it, redesign it, find more problems, etc and never get to the coding. Even Bjarne Stroustrup, in I believe it was his C++ Programming Language Special Edition, said it is better to get a foundation down quickly and then slowly add features and build on it rather than being stuck in an infinite design loop. I believe I've seen the same concept applied to video game programming too. That is why I stick to code more and think less, because I, myself, have stopped production on some projects or games because I would start finding things I didn't like and want to redesign them and then get lost in the design process.

That may be a sound aproach, but Nathan2222 currently hasn't been able name the design goals of his language. That needs to come first. He can start learning compiler design and the tools and libraries that are available for that, but coding novel language features is woefully premature.



#69 SunDog   Members   -  Reputation: 232

Like
0Likes
Like

Posted 19 February 2014 - 10:21 PM

 

In my thoughts, i've created a google chrome style browser with less than 300 lines of code.

In my thoughts I'm like a magical flying unicorn dude.

 

 

In my thoughts I've convinced Salma Hayek to be my personal pleasure slave for life



#70 Nathan2222_old   Members   -  Reputation: -400

Like
0Likes
Like

Posted 19 February 2014 - 11:10 PM


After making the language,


you will be probably at age of 90 then (you doing many assumptions about your superpowers)
this isn't something i'll do on my own

UNREAL ENGINE 4:
Total LOC: ~3M Lines
Total Languages: ~32
smile.png
--
GREAT QUOTES:
I can do ALL things through Christ - Jesus Christ
--
Logic will get you from A-Z, imagination gets you everywhere - Albert Einstein
--
The problems of the world cannot be solved by skeptics or cynics whose horizons are limited by the obvious realities. - John F. Kennedy


#71 ApochPiQ   Moderators   -  Reputation: 14890

Like
2Likes
Like

Posted 20 February 2014 - 12:03 AM

Then you have a secondary problem to solve: how will you convince anyone else to join your project?

#72 fir   Members   -  Reputation: -452

Like
0Likes
Like

Posted 20 February 2014 - 01:51 AM

 

 

After making the language,


you will be probably at age of 90 then (you doing many assumptions about your superpowers)
this isn't something i'll do on my own

 

as to making many things on your own, i personaly belive that today it is more hard than before (in 80'ties etc) (see some thread here on this forum if making games is harder today,) some people belive this is easier, but personally i belive this is much harder, and this is a problem imo (everyting growed 10 or 20 times in complexity or so, )



#73 Aardvajk   Crossbones+   -  Reputation: 5935

Like
0Likes
Like

Posted 20 February 2014 - 03:55 AM

Domain specificity can be taken to extremes in order to simplify languages, but then you lose flexibility:

 

main

{

    play_my_mmorpg()

}

 

is a language, assuming that all of the details of the mmorpg are already written and linked in via the compiler's standard inclusion mechanisms. But you can only write one game with it. As you extend the language to allow you to modify the output, you add complexity.

 

So you trade off. If you want your language to be simpler than existing solutions, you have to lose flexibility.

 

Maybe there is a magic, silver-bullet approach nobody has thought of yet. Maybe it is sitting out there waiting to be discovered. What you have to ask is, given that many thousands of extremely experienced and educated computer scientists with years of work and research behind them have been studying these topics for decades, how likely is it that you will be the one to invent it?



#74 fir   Members   -  Reputation: -452

Like
0Likes
Like

Posted 20 February 2014 - 04:41 AM

Domain specificity can be taken to extremes in order to simplify languages, but then you lose flexibility:

 

main

{

    play_my_mmorpg()

}

 

is a language, assuming that all of the details of the mmorpg are already written and linked in via the compiler's standard inclusion mechanisms. But you can only write one game with it. As you extend the language to allow you to modify the output, you add complexity.

 

So you trade off. If you want your language to be simpler than existing solutions, you have to lose flexibility.

 

Maybe there is a magic, silver-bullet approach nobody has thought of yet. Maybe it is sitting out there waiting to be discovered. What you have to ask is, given that many thousands of extremely experienced and educated computer scientists with years of work and research behind them have been studying these topics for decades, how likely is it that you will be the one to invent it?

The silver bullet here could be working 25 times faster than other specialist (doing in one day the things they do in 25 days) but very soon

i think you would die because of exhaustion 

 

Other way would be to live 400 years and doing tings at normal speed (even resting when tired to get back the motivation) but then you will die becouse of groving old

 

that is the situation 

 

(my personal approach is closer to the second way mentioned, Im trying to be patient and do only some small kind of things but aimed

at some kind of base quality, yet it is all (even simple thing) very hard)



#75 Nathan2222_old   Members   -  Reputation: -400

Like
0Likes
Like

Posted 20 February 2014 - 04:43 AM

Domain specificity can be taken to extremes in order to simplify languages, but then you lose flexibility:

main
{
play_my_mmorpg()
}

is a language, assuming that all of the details of the mmorpg are already written and linked in via the compiler's standard inclusion mechanisms. But you can only write one game with it. As you extend the language to allow you to modify the output, you add complexity.

So you trade off. If you want your language to be simpler than existing solutions, you have to lose flexibility.

Maybe there is a magic, silver-bullet approach nobody has thought of yet. Maybe it is sitting out there waiting to be discovered. What you have to ask is, given that many thousands of extremely experienced and educated computer scientists with years of work and research behind them have been studying these topics for decades, how likely is it that you will be the one to invent it?

i don't like mmorpg's.
The possibility of me inventing a language like that, i'll say 50:50.
I've also wondered the same thing. What could i make that hasn't been done. Why would i want to make an OS when there's Microsoft and Apple, why would i would i want to make a game engine when there's Unreal and cryengine?
There's also why is it that anytime i check google to demotivate me from doing these i see people making exactly what i'm intending to make. Why are people making languages, why are people making OS's, why are people making engines, why would Apple think of making a search engine? Also, why aren't there 4 programming languages, why didn't they leave it at 4? It's all readily available but yet it's still being done.
About 3 months ago, the question was why would i want to program? Why don't i just sit down and wait for numerous companies to churn out consumer products for me to use and not bother about anything? Well, about 3 months later i'm asking for this.
It's a lot of why's.

UNREAL ENGINE 4:
Total LOC: ~3M Lines
Total Languages: ~32
smile.png
--
GREAT QUOTES:
I can do ALL things through Christ - Jesus Christ
--
Logic will get you from A-Z, imagination gets you everywhere - Albert Einstein
--
The problems of the world cannot be solved by skeptics or cynics whose horizons are limited by the obvious realities. - John F. Kennedy


#76 fir   Members   -  Reputation: -452

Like
0Likes
Like

Posted 20 February 2014 - 05:01 AM

 

Domain specificity can be taken to extremes in order to simplify languages, but then you lose flexibility:

main
{
play_my_mmorpg()
}

is a language, assuming that all of the details of the mmorpg are already written and linked in via the compiler's standard inclusion mechanisms. But you can only write one game with it. As you extend the language to allow you to modify the output, you add complexity.

So you trade off. If you want your language to be simpler than existing solutions, you have to lose flexibility.

Maybe there is a magic, silver-bullet approach nobody has thought of yet. Maybe it is sitting out there waiting to be discovered. What you have to ask is, given that many thousands of extremely experienced and educated computer scientists with years of work and research behind them have been studying these topics for decades, how likely is it that you will be the one to invent it?

i don't like mmorpg's.
The possibility of me inventing a language like that, i'll say 50:50.
I've also wondered the same thing. What could i make that hasn't been done. Why would i want to make an OS when there's Microsoft and Apple, why would i would i want to make a game engine when there's Unreal and cryengine?
There's also why is it that anytime i check google to demotivate me from doing these i see people making exactly what i'm intending to make. Why are people making languages, why are people making OS's, why are people making engines, why would Apple think of making a search engine? Also, why aren't there 4 programming languages, why didn't they leave it at 4? It's all readily available but yet it's still being done.
About 3 months ago, the question was why would i want to program? Why don't i just sit down and wait for numerous companies to churn out consumer products for me to use and not bother about anything? Well, about 3 months later i'm asking for this.
It's a lot of why's.

 

 

the difference is they do it and youre talking, try it yourself ..

 

I know it is fun to walk ant talk "i will do everything" (like someone says you will get an electronic hat that will magically make it) but this is

a blit like watching a crazy man making a fool of people here..

(and this is a bit sad of you)


Edited by fir, 20 February 2014 - 05:03 AM.


#77 Olof Hedman   Crossbones+   -  Reputation: 2738

Like
1Likes
Like

Posted 20 February 2014 - 05:17 AM

There is always room for improvement, if you have some specific task in mind, with well defined requirements.

 

Perfection is unfortunately out of reach for us mere humans, with limited lifespan and resources. We are not gods (yet).

 

Engineering is all about making the right trade-offs to get something done in reasonable time, with reasonable resources.



#78 Aardvajk   Crossbones+   -  Reputation: 5935

Like
2Likes
Like

Posted 20 February 2014 - 09:38 AM

 

Domain specificity can be taken to extremes in order to simplify languages, but then you lose flexibility:

main
{
play_my_mmorpg()
}

is a language, assuming that all of the details of the mmorpg are already written and linked in via the compiler's standard inclusion mechanisms. But you can only write one game with it. As you extend the language to allow you to modify the output, you add complexity.

So you trade off. If you want your language to be simpler than existing solutions, you have to lose flexibility.

Maybe there is a magic, silver-bullet approach nobody has thought of yet. Maybe it is sitting out there waiting to be discovered. What you have to ask is, given that many thousands of extremely experienced and educated computer scientists with years of work and research behind them have been studying these topics for decades, how likely is it that you will be the one to invent it?

i don't like mmorpg's.
The possibility of me inventing a language like that, i'll say 50:50.
I've also wondered the same thing. What could i make that hasn't been done. Why would i want to make an OS when there's Microsoft and Apple, why would i would i want to make a game engine when there's Unreal and cryengine?
There's also why is it that anytime i check google to demotivate me from doing these i see people making exactly what i'm intending to make. Why are people making languages, why are people making OS's, why are people making engines, why would Apple think of making a search engine? Also, why aren't there 4 programming languages, why didn't they leave it at 4? It's all readily available but yet it's still being done.
About 3 months ago, the question was why would i want to program? Why don't i just sit down and wait for numerous companies to churn out consumer products for me to use and not bother about anything? Well, about 3 months later i'm asking for this.
It's a lot of why's.

 

 

I've been programming for about thirty years now, and have yet to find something I could do that hasn't been done. If you have a "great idea", the first thing you should assume is a zillion others have already had the same idea, some of whom were in a position to assess its feasibility.

 

By all means, please go implement your idea. You'll find once you have more than a "great idea" people will start taking you seriously. As it is at the moment, people just want to give you a reality check to stop you wasting time. Ours and yours.






Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS