Many Questions: Generics Variance

IL variance

Wesner Moise — Sat, 30 Jul 2005 00:31:44 GMT

Why would generic variance add too much complexity to C#, when IL already supports it.

re: Many Questions: Generics Variance

TAG — Sat, 30 Jul 2005 00:57:59 GMT

It's a Microsoft fault !!

You was able to add additional methods something like this:

class List<TItem> {
void AddRange<TOther>(List<TOther> list) where TOther: TItem {
}
}

re: Many Questions: Generics Variance

Mort — Sat, 30 Jul 2005 16:28:56 GMT

In my opinion C# team should keep C# as the static/ type safe language and avoid adding too much difficult to understand new concepts or ones that would cause even more run-time surprises. However I would like to see these more advanced features in C# like language also. My suggestion is that you allow developers to use advanced C# features only in files that have, for example, extension .csa or .cs3 etc. So if developer needed these features that might cause more dynamic behaviour for example, they could use them in C# project but if a second person (maintenance programmer?) came to take over the project, he would notice from the extension immediately that there may be advanced concepts used in the code that you need to learn first.

Blog link of the week 30

Daniel Moth — Mon, 01 Aug 2005 00:00:47 GMT

Blog link of the week 30

re: Many Questions: Generics Variance

TAG — Tue, 09 Aug 2005 14:57:03 GMT

Your argument type validation code "item is S" is flawed.
It return False for null-reference values - but this is legal for interface contracts you are defining !

re: Many Questions: Generics Variance

peterhal — Wed, 10 Aug 2005 23:37:00 GMT

Good catch TAG, you are correct. The checks for item is S, should also permit null.

Peter

Topics about Tour-eiffel | Many Questions: Generics Variance

Topics about Tour-eiffel | Many Questions: Generics Variance — Tue, 21 Apr 2009 18:54:06 GMT

PingBack from http://tour-eiffel.linkedz.info/2005/07/29/many-questions-generics-variance/