Present computer language, satisfied?

[alpha2cen]

There are many languages in the computer world.

Recently developed advanced language is difficult to classify whether a language or an application package.

What will be next language? What will be needed for next language?

[khaled]

This is my guess .. My assumption is 30 %

 

 

 

The next programming\scripting language will be related to Web 2.0,

 

It will give the properties of a virtual Distributed Operating System,

 

End of Assumption

 

 

[kurtis]

I'm hoping it's just a self-programming, programming language. I'm sick of all of the mindless, repetitive tasks that go into Computer Programming.

[alpha2cen]

How about one of the future language is the one which is based on mathematics?

Today's language is too programing dependent.

[alan2here]

Totally weekly typed OO language where.

 

• Any function or object instance in scope can be used on the fly stand alone, as an operator or as a member.
  
• Never specifiy type anywhere such as when passing and returning values.
  
• No inheritance although similar things can easily be done.
  
• Functions can code up other new functions.
  
• Like Java in that everything is like a pointer to an object, where there is automatic cleanup.
  
• Dissimilar types can be compared with == and where there are no comparison members for them in scope or in either object then the result is false.
  
• Individual members can be removed from or added to an object instance at any time.
  
• Instead of 'book b = book()' then it would be 'obj b = book' where book is the name of a function that composits a "book" object from other members, each of which are created by functions. Buit-in types (at the bottom level) also look like functions.
  
• Code can be passed around as well, for example 'obj c = {some code...}' allowing you to roll your own flow controll structures such as various sorts of loops, case statements etc...
  
• Everything that is typed from the keyboard is effectively a string, so number(42) that returns the number 42 would be writtain as number("42").
  
• A built-in type called 'set'. In this example there are 3 elements: obj s = {number("42") some code... "cheese" member(key {some code...})}. Notice the similarity between entities conventionally thought of as sets, functions and objects.
  
• There is a libary of construction functions that is synchronized to a central online one.
  
• Add and remove things from scope as you need them from the libary so the number of symbols in scope stays small.
  
• Store various ways of achieving the same result in functions so that other code can request the most appropriate version based on its requirements.
  
• The library includes things like strings that can be spellchecked in specified langueges and polygons where details can be specified such as lengths and angles and other details can be quiried such as what other lengths and angles are.
  
• A member of the number built-in type would allow you to specify the maximum precision the number could if required go up to instead of there being various number types of differnt precisions.
  
• A built-in type for storage of data which is just binary information.
  
• Open a file and easily store instances of anything that is in scope inside it.
  
• Semi-defined values, for example it's less than 4, it's a value valid in this equation, it's an item in this set, it's got multiple values all of which must be true, or(and(1 2 3), and(4 5 6)) etc...
  
• In all 3 of thease cases result occurs if n is 2 or 3, if(n == 2 || n == 3, result), if(n == {2 || 3}, result), if(n == (2 || 3), result)
  
• Usefull default loops such as for("5") to occur 5 times, for(bool) where bool is a function that creates a default bool, the loop would then occur twice, once for true and once for false, a more difficult but not insurmountable obstacle for compound types given the very weakly typed nature of this language, for(set_name) would loop once for every item in a set.
  
• Use the 'param' function to get a parameter, for example left\right for when used as an operator or a number or key to get a specific parameter.
  
• gate("2") == gate({bool, bool}, {bool}, "2") == and. gate("8") == gate({bool, bool}, {bool}, "8") == and. I have an abstraction for generic mathematical operations that occur between integers as well.
  
• inputs.keyboards.0 to get the first connected keyboard to the device that the program is being run on. Whatever is actually the case this will appear to return Unicode charecters. Similarly, inputs.key("a"), the device might have no keyboards but a few individual keys instead.
  
• set_name.every.increment() to perform increment on every member of a set.
  
• Tasks are like functions but where the function that called them continues to execute immediately instead of waiting for the function that has been called to finish, like threads. 'obj n = {some code...} [new line] n.task' or 'obj n = task({some code...})'
  
• Objects from the libary wouldn't contain tasks that outlived there member functions unless they were named accordingly.
  
• Objects from the libary would have a names member which was a set containing strings, this would be referenced when refering to high level abbilites of the object where 'does it have a member with thease properties' and similar requests would become tedious and unreliable.
  

Edited January 30, 2011 by alan2here

[alpha2cen]

How about one of the future language is the one which is based on mathematics?

 

The language based on mathematics is easy to learn.

Every educated people can easily use it without any more education.

And, through time pass by, the programming language base form is not easily changed.

There are many programming languages, but it is too difficult for novice.

[Xittenn]

The language based on mathematics is easy to learn.

Every educated people can easily use it without any more education.

And, through time pass by, the programming language base form is not easily changed.

There are many programming languages, but it is too difficult for novice.

 

I don't concur! Learning math is much harder than learning a programming language. I find learning math far more rewarding but learning programming far more entertaining and this may have some influence on my statement. I'm getting pretty good at knowing the ins and outs of programming but I have to revisit even the most basic of math concepts whenever I need to use it.

 

There are languages that focus on developing around a mathematical context already. I think the most appropriate one to refer to would be Haskell and the general family of functional programming languages; although there are other much more mathematically founded languages!

 

I wish to, in my lifetime, write a language that is mathematically founded and categorize it as a Grand Unified Parallel Processing Language. It will be a harmonious blend of logic and analogue processing dedicated to the virtualization of reality on a machine. I wish to name it, and forgive me for making this statement, Kermit-On-Krack for no other reason than I loved the fact that the Kermit protocol was named as such and it will be a rather insane language. Just sayin' .... This would however require some rather fundamental changes to the way a processing unit would be designed as differential equations on a digital processor are slow even when using the most optimized numerical methods ....

Edited January 29, 2011 by Xittenn

[alpha2cen]

I don't concur! Learning math is much harder than learning a programming language. I find learning math far more rewarding but learning programming far more entertaining and this may have some influence on my statement. I'm getting pretty good at knowing the ins and outs of programming but I have to revisit even the most basic of math concepts whenever I need to use it.

 

There are languages that focus on developing around a mathematical context already. I think the most appropriate one to refer to would be Haskell and the general family of functional programming languages; although there are other much more mathematically founded languages!

 

I wish to, in my lifetime, write a language that is mathematically founded and categorize it as a Grand Unified Parallel Processing Language. It will be a harmonious blend of logic and analogue processing dedicated to the virtualization of reality on a machine. I wish to name it, and forgive me for making this statement, Kermit-On-Krack for no other reason than I loved the fact that the Kermit protocol was named as such and it will be a rather insane language. Just sayin' .... This would however require some rather fundamental changes to the way a processing unit would be designed as differential equations on a digital processor are slow even when using the most optimized numerical methods ....

 

We can learn mathematics during the school course, but can not learn many programming languages.

And more developed thinking is comes from mathematics.

If programming language were more mathematically developed, we could use it more developed work easily.

But present program language is developed for special programmer, i.e., highly educated programmer.

[Xittenn]

Well I don't know about all languages and whether or not the entirety of them satisfy the needs of all applications. I do know the one language I spend my time in and that is C++; well sort of maybe know! If I could change anything it would be the following:

 

1) Name mangling would have some new universal cross platform specification that all compilers would abide by.

 

2) All function pointers should be the maximum size required of a polymorphic member function pointer no expense spared. I think 32 Bytes is the value of choice here. This would allow for callback member functions at a standardized system level. It would also allow for a standardized C++ implementation of delegates. All classes get message pumps without the overhead of managing a pool and when combined with 1) we have convenient event sink manipulation.

 

3) The new decltype keyword and trailing return type should allow for a much more developed compile time assessment of what the return should be. I'm not saying however that the language should cease to be strong typed as alan2here suggested, just that the language should strongly support Meta Programming as well; even more so than it already does.

 

4) The protected keyword should have polymorphic behavior and a base class should have access to another base class of the derived or the complete object without having to declare friends. Friends looks kind of metro or something and I hate when my code isn't looking quite the way I want it to. CRTP!

 

5) A template specialization should be accessible from its own translation unit, I do not like sticking all my code at the bottom of a header.

 

6) There should be a universal Interface Definition Language that extends to all platforms and to all languages.

 

7) A static struct should be initialized member-wise, initializer lists don't look right.

 

8) Temporaries should be considered a defined method of initialization for large static const constructs and arrays and the life of the temporaries extended indefinitely to the lifetime of the object being initialized. This is already fact but is questionable in ways and should be better defined.

 

9) Static and Dynamic code libraries should have a standardized definition, again to all platforms.

 

.... there are more but I can't think of them right now .... aside from the above I like C++ and hope it maintains it's priority in the languages used for application development. C# and Java may have their place and I guess I appreciate what it is they do but I haven't really used them yet. I think I will be adding a layer of abstraction to my projects by developing a coding environment that sits on top of my libraries and it will be in C# and/or Java.

Edited January 29, 2011 by Xittenn

[alan2here]

Id like to see number 6 that works with C++, your C++, my languege, functional languages and really low level languages. I agree that interfaces are verry important, everything else is just detail.

Edited January 29, 2011 by alan2here

[sneha123]

Hello friends,,,,this is sneha here,,,,I'm hoping it's just a self-programming, programming language. I'm sick of all of the mindless, repetitive tasks that go into Computer Programming..........thanks for nice share with all people..........

[khaled]

Creation of a new programming language has a reason, either by Vendors, or in essence to a new technology,

for example Quantum Programming Language will probably appear in the future if Quantum Computer

came to reality ...

[verno]

The creation of a new language....hmm....

 

Well, I'm not sure exactly what it would be, but my bet is if they're making a new one it should be more efficient, powerful, or have a completely different purpose than the ones already made,

other wise it would be completely pointless.

 

I'd go with what khaled said, about it being related with web 2.0.

[Svetoslav Pavlov]

You are right. I hope it will come soon...

 

This is my guess .. My assumption is 30 %

 

 

 

The next programming\scripting language will be related to Web 2.0,

 

It will give the properties of a virtual Distributed Operating System,

 

End of Assumption

 

 

[Xittenn]

You are right. I hope it will come soon...

 

ASP.NET?

[khaled]

ASP.NET?

 

Xittenn, ASP.Net is already made and deployed, although it's rarely used,

JSP is used by a variety, but PHP is used by the most, due to server costs ...

 

I expected a new computer language, with Web2.0, HTML5, and CSS3,

a new language that works between the browser & the server, combines

everything from ASP,JSP, and PHP deploying new Web2.0 technologies ...

[Andreas]

One possibility is that programming languages as we know today will be reduced to a minimum.

 

Back in the early days a lot of work was needed just to display a simple button or edit box, compare that to

today where you use ready made building blocks pretty much all gui elements such as buttons, list boxes and dialogs etc.

 

Creating an application for registering customers for example could in theory be reduced to dropping a "person object" on

a form. If you need to customize it, adding fields and such could be done without coding.

 

I think accessing the "code" and modify it would still be possible but most would be written in a simple script language.

 

More and more programs will most likely run completely independent of platform, compiler output will consist of two parts "appearance descriptor" and "code behind", where it's up to the target platform to render the GUI according to its own standard, and either the platform itself or a remote server will execute the proper code behind, depending on type of application.

 

 

This is just a possibility, and seeing as I like the "dirty" code, nothing I'd look forward to, except for the platform independence.

 

Cheers