Games people play

[Georg Wrede]

Lutger wrote:
> Georg Wrede wrote:
> 
>> Walter Bright wrote:
>>> But on the other hand, I don't think this is a reason to be hesitant 
>>> to use D. If those more unusual combinations cause a problem, just 
>>> don't use them that way. It isn't necessary to remove the features 
>>> from the language!
>>
>> The C++ guys certainly shouldn't complain, they already do it every day.
>>
>> And not using the entire feature set in each source file might be a 
>> good idea with any language. :-)
> 
> But the C++ guys do complain. For every thread on C++ at gamedev, there 
> is no end to the complaints of experienced developers (I'm not counting 
> fanboys). Unexpected interactions between features and undefined 
> behavior is one source of frustration, what makes C++ complex.

Ah, my point exactly: they shouldn't complain about a few isolated 
"feature clash problems" in D. "Don't complain about eating bean stew 
when what you've eaten for years is just pig fodder." Except D is more 
like ice cream, IMHO. ;-)

> If game developers are to move to a different language than C++, this 
> has to be better, not the same. 

I'd like to see the person who says D is not *much* better in this respect!

> In this light, C# may look more 
> favorably to some (indie at least, who don't need to port to all 
> platforms). 

You can't be serious! Maybe for an indie writing a new tic-tac-toe, but 
for Serious Development, I sure hope not.

> More lightweight, more tools, libraries, company support. I 
> think if you want to attract game developers, you will have to compete 
> with C# not C++. 

That's another thing. Tools and libraries we lack, no denying.

> I don't know if not using features is a valid option, because it's based 
> on the premise that it's understood already how exactly they will 
> interact with other features, and that 3rd party libraries will not use 
> them. Even if this is so, it does increase complexity of development.

The C++ guys should be used to it.

> Is this viable in a project with, say, half a million lines of code?

I'd say not. But to start the next project in a new language (after 
suitably trying it out with a couple of prototyping projects), then yes.