would you say that Perl is not a strong type language since we do not need to define the type of the variable?

Actually, you must define the type of the variable in Perl by using the right sigil. It's a question of specificity.

Questions about whether a language is strongly typed or not don't lead to anything productive. It's usually used to apply against preconceptions about code robustness, code size, coding speed, etc. Those are definitely interesting areas of study. Whether a language is strongly typed or not isn't. In short, your question suffers from the XY Problem.

I recommend that you read the results of Dominus's attempt to define "strongly typed".

Dynamic programming language - execute at runtime many common behaviors that other languages might perform during compilation

Given how common dynamic programming languages have become, your definition doesn't match any languages. Your definition is a contradiction.

Questions about whether a language is strongly typed or not don't lead to anything productive. It's usually used to apply against preconceptions about code robustness, code size, coding speed, etc. Those are definitely interesting areas of study. Whether a language is strongly typed or not isn't. In short, your question suffers from the XY Problem.

ikegami ++!





---

---

I am trying to improve my English skills, if you see a mistake please feel free to reply or /msg me a correction

C++ has run time vtables and I think Perl has an equivalent concept for OO. So in that sense it is dynamic (can figure out what exact code to execute for a method call at run time - enables polymorphism among other things.

In Perl you can actually write a program "on the fly" and then execute it - how much more dynamic could it be?

Sure that other Monks will have lots to say about all of this.

Update: So I am basically going to duck 'n cover for now.

would you say that Perl is not a strong type language since we do not need to define the type of the variable?

Strong typing has no much to do with variable types, it has to do with conversions from one type to another. Strongly typed languages typically disallow implicit conversions between unrelated types. What you are referring to is static typing

Questions about whether a language is strongly typed or not don't lead to anything productive.

I do believe the same but at the same time think that it could get important when you want use more than one language in a project or you want to learn a new language, and can potentially face issues that can arise from the variation in their type systems

For example if one is accustomed in writing in VB.net and wants to write in C# as well, he could find the transition not so smooth, not because of that they are different languages per se, but because their type systems allow or disallow certain operations which could interfere with existing experience.

For example in VB.net the Object type behaves dynamic language style(when option strict off is turned off) while in C# the behaviour is purely statical (until lately one would have to use the Reflection services to get dynamic style behaviour) or as far as weak typing goes, the compiler would allow implicit conversions between unrelated types in an expression evaluation i.e treat a string as a number (like VB.net and Perl do) while in C# would not.

Certainly, one cannot label a language as purely strongly typed or purely weakly typed because even Perl notoriously labeled as weakly typed displays strong typing properties ,in some cases, but I think that if one has a grasp on the principles would enjoy a smoother transition and/or improved production

can potentially face issues that can arise from the variation in their type systems

Then talk about the variations. Using a term that both applies and doesn't apply to just about every language doesn't help.

Based on these definitions, I would generally agree.

As to the second point, it is worth mentioning that many language implementations today (and language designs today) are far more “dynamic” in nature than many earlier languages literally could afford to be.   Processor speeds and memory sizes are orders of magnitude larger than before, making many strategies practical which could not be considered in earlier times.   Languages quickly followed suit.   So, this distinction, in practical terms today, is just not that useful a distinction anymore.   It has become “academic.”

Q:   Where do you draw the line?
A:   Only where the line is useful.

You don't need to have to specify the type for the language to be strong typed. Ask Google about "type inference". There are languages that do not expect you to declare the types (most of the time), because they can infer the type and test the consistency of a variable's/function's use. Actually for the most simple cases you no longer have to specify the type even in C# or VB.Net.

However it should be pointed out that type inference is a compile time afair and not a runtime afair;C# did not turn dynamic because of the use of the 'var' keyword.

In the most simple cases such as with primitive data types, type inference provides a shortcut instead of explicitly denoting the type of the variable :

var a=10; #the compiler infers that a is an int

but its real value comes when having to assign the return value of an expression's evaluation to a variable and is difficult to work out the variable's type so you must cast or re-cast;cases like that occur frequently with Linq queries