Hi,

show the exact following code to dmd 0.102 on Linux and it drops dead instantly. Believe it or not. (Have not tried <0.102.) Swap the definition of Foo(T1) and Bar and it works fine.


class Foo()
{
}

class Bar : Foo!()
{
}

class Foo(T1)
{
}

void main()
{
}


Bastiaan.

on my linux machine (Mepis linux) I just get a segfault :)

Bastiaan Veelo wrote:
> Hi,
> 
> show the exact following code to dmd 0.102 on Linux and it drops dead instantly. Believe it or not. (Have not tried <0.102.) Swap the definition of Foo(T1) and Bar and it works fine.
> 
> 
> class Foo()
> {
> }
> 
> class Bar : Foo!()
> {
> }
> 
> class Foo(T1)
> {
> }
> 
> void main()
> {
> }
> 
> 
> Bastiaan.

clayasaurus wrote:
> on my linux machine (Mepis linux) I just get a segfault :)

:-) That is what I mean. But it seems to happen very early (during lexical analysis?) because it is not preceded by an error message and changing the order of definitions has a say.

How is Windows doing in this regard?

Bastiaan.

Bastiaan Veelo wrote:
> clayasaurus wrote:
> 
>> on my linux machine (Mepis linux) I just get a segfault :)
> 
> 
> :-) That is what I mean. But it seems to happen very early (during lexical analysis?) because it is not preceded by an error message and changing the order of definitions has a say.
> 
> How is Windows doing in this regard?

No better. Results in GPF on WinXP.

> Bastiaan.