C# "dynamic"

The Future of C#, Part Two

Fabulous Adventures In Coding — Tue, 28 Oct 2008 19:25:04 GMT

Well, I intended to spend the last three weeks blogging about C# design process in anticipation of our

C# "dynamic," Part II

Chris Burrows' Blog — Tue, 28 Oct 2008 20:58:48 GMT

Yesterday, I made an attempt to introduce the C# dynamic feature by describing what change there has

re: C# "dynamic"

configurator — Wed, 29 Oct 2008 02:45:47 GMT

(This has been cross-posted from http://www.codethinked.com/post/2008/10/28/C-40-New-Features-Part-1-dynamic-keyword.aspx )

I wonder what would happen if I do this:

dynamic test = new TestClass();

test.TestMethod1();

test = new TestClass2();

test.TestMethod1();

where both classes have TestMethod1 defined.

Then, <Main>o__SiteContainer0.<>p__Site1 would be defined after the first call, and would be calling a wrong function after setting the dynamic variable to a different value.

Of course, I haven't seen or installed the CTP so I wouldn't know the exact implications.

ozone, the difference is that var is still bound at compile time. the line

var test = new TestClass1();

is translated into

TestClass1 test = new TestClass1();

You can see that using reflector.

However, dynamic is translated to object, and changes every usage of that variable, as seen in the example given by Justin.

re: C# "dynamic"

Steve — Wed, 29 Oct 2008 05:51:56 GMT

Curious about some implications.

Can you nest dynamics as in dynamic x; x.Prop.Func()?

Can you make a List<dynamic>?

Thanks!

C# "dynamic," Part III

Chris Burrows' Blog — Wed, 29 Oct 2008 15:00:15 GMT

Here are a few good resources that you ought to look at for information about dynamic from PDC: Anders

Dynamic in C#

Sam Ng's Blog — Thu, 30 Oct 2008 01:01:04 GMT

The other day I was playing around with some office code, and I found myself writing a lot of code much

re: C# "dynamic"

Tanveer Badar — Thu, 30 Oct 2008 18:52:01 GMT

VB with a syntax of C. Or C# without option strict.

I really can't make up my mind.

re: C# "dynamic"

cburrows — Thu, 30 Oct 2008 21:02:11 GMT

Steve,

Yes and yes. The result of dynamic operations are typed as dynamic, so you can layer them (except conversions, which are typed appropriately). And you can build generic types with dynamic too. I'll talk more about this later.

Tanveer,

You're right that this is something that VB has done, in a way, for a while. At least one difference is that you get to control it at a finer granularity than "option strict off," which affects your whole compilation unit as I understand it. Another difference is that we target the DLR, not the VB runtime.

chris

re: C# "dynamic"

Anonymous — Thu, 30 Oct 2008 21:50:09 GMT

Seems to me C# is really heading in the wrong direction. Up to 2.0 and some of 3.0 everything was amazing and perfect. Everything was designed perfectly, to where it constrained developers to also design accordingly and to follow rules. This in turn made errors much easier to find especially at compile time (Generics). Now, it seems as if we're just throwing all this out the window and trying to just add every little convenient feature anybody can think up. It's making the framework very messy and it's going to result in less compile time errors caught and more run time errors...

re: C# "dynamic"

David Nelson — Fri, 31 Oct 2008 18:32:57 GMT

I am trying really hard not to pre-judge, but I have to say that so far I haven't seen anything that makes me think that the "dynamic" type is a good idea. I understand that dynamic languagues are becoming more popular, but that doesn't mean that every language in the world needs to support dynamic typing. Different languages exist for different reasons and cater to different needs. It is not possible for one language to do everything, no matter how much the C# team might want to believe that. There are already projects underway to bring existing dynamic languages into the .NET world, and for that matter VB.NET is already capable of late binding when it is necessary for things like COM compatability. Did the C# team somehow feel the need to compete with these languages? Did they forget the core principles which have made C# so successful? It would have been a far better choice to make the language MORE strict, using the great features being developed in Spec#, rather than trying to build in this dynamic capability which adds little value and carries with it the very real possibility of seriously diluting the language.

C# is based on static typing for a reason; tossing that out the window for the sake of the latest fad is irresponsbile at best.

re: C# "dynamic"

int19h — Fri, 31 Oct 2008 22:02:06 GMT

> Did the C# team somehow feel the need to compete with these languages?

I think you're missing the point. C# is not trying to compete with these languages, it's trying to interoperate with them. With C# 4.0, I can write a class library in Python or Ruby, and then use it from C#, whereas before, it could only be the other way around. And this is cool.

re: C# "dynamic"

evildude — Mon, 03 Nov 2008 18:55:46 GMT

Why anybody in right mind would write class library in Python or Ruby?

re: C# "dynamic"

Mark Rendle — Mon, 03 Nov 2008 19:12:45 GMT

Does this new dynamic stuff include an "eval" of any sort?

re: C# "dynamic"

Dolgov V. — Mon, 03 Nov 2008 21:47:24 GMT

It's a very bad idea.

C# 1.0 aim to more strict and safe code. Even unboxing (if variable's type was "object") can't be made without explicit type convertion. But now, in C# 4.0, it's can.

And such simple use of Python or/and Ruby libraries in C# is the way to minefield of errors.

Community Convergence XLVII

Charlie Calvert's Community Blog — Tue, 04 Nov 2008 00:11:42 GMT

Welcome to the 47th Community Convergence. We had a very successful trip to PDC this year. In this post

re: C# "dynamic"

Mark Rendle — Tue, 04 Nov 2008 14:59:20 GMT

It's quite possible to create a minefield of errors in C# already, if you're a bad enough programmer. If you're a good programmer, and you cover your code with unit tests, then there's no reason why dynamic code should be any more error-prone than static code.

If C# 4.0 was becoming a fully dynamic language, I'd be at Microsoft's gates (NPI) with pitchforks and torches. But all it's doing is adding a new type and some compiler support, and hugely simplifying a multitude of tasks, not least Office interop. That's got to be a good thing.

re: C# "dynamic"

David Nelson — Wed, 05 Nov 2008 00:36:18 GMT

@Dolgov

Explicit casting is required any time you cast down the type hierarchy, which is what you're doing when you cast from "object" to a value type; boxing and unboxing was a side effect that was handled automatically by the compiler.

@Mark

"If you're a good programmer, and you cover your code with unit tests, then there's no reason why dynamic code should be any more error-prone than static code."

You're right, tests are great. And compile-time type checking is just another kind of test, one that the compiler won't let you turn off. Why would I voluntarily eliminate an entire class of tests that is available to me, and hope that the tests that I have written myself cover all the possible problems that the compiler could have caught for me?

"It's quite possible to create a minefield of errors in C# already,"

Its quite possible for an idiot to create a minefield out of anything, and that applies to every area of life, not just programming. How to deal with idiots is not germane to the discussion. The question is, how do we help competent, well-intentioned developers "fall into the pit of success"? Making things more dynamic doesn't move us any closer to that goal, it pushes us farther away.

re: C# "dynamic"

int19h — Wed, 05 Nov 2008 14:07:28 GMT

Thankfully, C# these days is not similar to Java in that it does not try to be as idiot-proof as possible by severely limiting its features. Yes, "dynamic" can be misused. So can lambdas, generics, partial classes/methods, and God knows what else. But in the right hands they're all great and useful tools, and that's what is important.

re: C# "dynamic"

Mark — Wed, 05 Nov 2008 18:21:20 GMT

David, repeat after me: COMPILATION IS NOT TESTING.

re: C# "dynamic"

David Nelson — Thu, 06 Nov 2008 00:24:50 GMT

Mark,

Semantic compile-time errors such as type checking serve the exact same purpose as unit testing typically does in dynamic languages. I just choose to use the C# compiler to do that first step in the testing for me, rather than throwing out all of those checks that are already built in and then trying to recreate them (DRY anyone?).

re: C# "dynamic"

Dolgov V. — Thu, 06 Nov 2008 09:52:05 GMT

@David

"Explicit casting is required any time you cast down the type hierarchy, ..."

Yes, of cause. When we use static typing.

But with dynamic I can write

dynamic a = new Animal();

Duck d = a;

with no compile errors.

C# "dynamic," Part IV

Chris Burrows' Blog — Thu, 06 Nov 2008 20:45:32 GMT

Today, let's geek out about the language design regarding the dynamic type. Type in the language vs.

re: C# "dynamic"

odahan — Thu, 06 Nov 2008 21:25:07 GMT

I'm just hoping (dreaming?) that C#4 dynamic feature will be OFF by default. And to switch it ON I propose to hide the parameter very deep in a crypted xml config file with a bizzare name. Of course a very long password will be necessary to decrypt the config file and change the switch to ON. The very long password will only be accessible opening an account on a very expensive MS portail.

If all those conditions are ok for you, then I'm ok for the dynamic implementation in C# 4 !

re: C# "dynamic"

Alex G — Sun, 09 Nov 2008 01:43:21 GMT

Funny how people complain at every new C# feature. Remember back in the pre 3.0 days when people complained of lambda/closure...

There's no motivation for a developer to use dynamic unless there is a real need, so I don't see a problem. It's like some of you are affraid us C# programmers will think we're supposed to scrap all type declarations and use dynamic everywhere...

Well truly, it is you sunday morning theorists that scare me.

Thanks to the C# team for making our lives simpler when it comes to interoperability.

Next time I hear office integration, I might want to hurt someone for suggesting we install office in the server farm, but at least not for suggesting we interop with it from C# :)

re: C# "dynamic"

David Nelson — Mon, 10 Nov 2008 22:27:48 GMT

@Alex,

I don't remember there being a significant uproar about lambdas or closures. They were somewhat misunderstood, but I don't remember seeing a lot of people arguing that they should not be added to the language.

The problem with dynamic is that there are many C# developers, especially those reluctantly converted from VB, who WILL think they are supposed to scrap all type declarations and use dynamic everywhere! I am NOT saying that simply having a potential for misuse is a reason not to add a feature to a language; I am sick of having my objections cast in that light. What I AM saying is that I don't believe that the feature as it is currently implemented brings enough added value to the language to outweigh the negative effects it will have. I also believe that there were other options for implementing late binding which would fit better into a statically typed language.

C# "dynamic," Part V

Chris Burrows' Blog — Tue, 11 Nov 2008 05:04:27 GMT

Let's look at this: dynamic d = null ; object o = d; // not an implicit conversion Last time , I said

C# "dynamic," Part VI

Chris Burrows' Blog — Fri, 14 Nov 2008 17:20:59 GMT

We left off last time with this piece of code public class C { public static void M( int i) { } public

C# Dynamic - CSharp's new feature of the coming version 4.0

Journal of Abu Sayed Mohammad Ismail — Mon, 15 Dec 2008 21:01:29 GMT

Very good resources for the coming version... Sam Ng Dynamic in C# Part One Dynamic in C# Part Two Chris

re: C# "dynamic"

Bob Reselman — Tue, 23 Dec 2008 22:27:20 GMT

This sounds really cool.

Tell me please, what is the real world benefit? I am afraid that this is a reintroduction of the VB Variant. But, somehow I think that it is not. I wish that understood this better.

re: C# "dynamic"

dracoon — Sun, 28 Dec 2008 05:00:14 GMT

Instead dynamic,

can't u make a lot better performance of unboxing operation ?

Simulating INumeric with dynamic in C# 4.0

Luca Bolognese's WebLog — Thu, 05 Feb 2009 19:47:48 GMT

When I wrote my Excel financial library I agonized over the decision of which numeric type to use to

ASP.NET MVC for RoR developers: do as locals do

CodeClimber — Tue, 24 Mar 2009 00:47:13 GMT

ASP.NET MVC for RoR developers: do as locals do

C# “dynamic,” Part VII

Chris Burrows' Blog — Wed, 25 Mar 2009 03:36:12 GMT

It has been a while since I posted anything here, and I have the same excuse everyone else does. Work,

A quick analyze of the .NET Fx v4.0 Beta1

Patrick Smacchia [MVP C#] — Thu, 21 May 2009 11:29:46 GMT

Normal 0 21 false false false FR X-NONE X-NONE Let’s have a look to see what the tool NDepend shows