I haven't gotten very far with the MIID list yet, this is mostly just compiler bugs. I'm going to be out for a few days, and I wanted to get this out first.

http://www.digitalmars.com/d/changelog.html

On Tue, 21 Sep 2004 19:19:15 -0700, Walter wrote:

> I haven't gotten very far with the MIID list yet, this is mostly just compiler bugs. I'm going to be out for a few days, and I wanted to get this out first.
> 
> http://www.digitalmars.com/d/changelog.html

DUI and tests compile and run
leds compiles but terminates with

"
Error: circular initialization dependency with module Leds
"

Anybody knows what's that?
can it be

 "Fixed private module construction order." ?

What should I look for?

Ant

"Ant" <duitoolkit@yahoo.ca> wrote in message news:pan.2004.09.22.03.15.15.767739@yahoo.ca...
> On Tue, 21 Sep 2004 19:19:15 -0700, Walter wrote:
>
> > I haven't gotten very far with the MIID list yet, this is mostly just compiler bugs. I'm going to be out for a few days, and I wanted to get
this
> > out first.
> >
> > http://www.digitalmars.com/d/changelog.html
>
> DUI and tests compile and run
> leds compiles but terminates with
>
> "
> Error: circular initialization dependency with module Leds
> "
>
> Anybody knows what's that?
> can it be
>
>  "Fixed private module construction order." ?
>
> What should I look for?

Two modules that import each other, and each has static initializers.

On Tue, 21 Sep 2004 20:24:58 -0700, Walter wrote:

> 
> "Ant" <duitoolkit@yahoo.ca> wrote in message
>> "
>> Error: circular initialization dependency with module Leds
>> "
>>
> Two modules that import each other, and each has static initializers.

thanks, I'll check that (might take awhile...)

Ant

On Tue, 21 Sep 2004 23:21:32 -0400, Ant wrote:

> On Tue, 21 Sep 2004 20:24:58 -0700, Walter wrote:
> 
>> 
>> "Ant" <duitoolkit@yahoo.ca> wrote in message
>>> "
>>> Error: circular initialization dependency with module Leds
>>> "
>>>
>> Two modules that import each other, and each has static initializers.
> 
> thanks, I'll check that (might take awhile...)
> 

It wasn't critical that my static vars/consts were static so I just made then instance variables and it pass that "circular init" constraint.

I know there was a lengthy discussion about this
I'll read it before complaining - but this doesn't smell very well :)

see ya when you're back, i guess.

Ant

DAMN YOU! :D

I just got D.NET caught up to 0.101 :)

Walter wrote:

> I haven't gotten very far with the MIID list yet, this is mostly just
> compiler bugs. I'm going to be out for a few days, and I wanted to get this
> out first.
> 
> http://www.digitalmars.com/d/changelog.html
> 
> 
> 

"Ant" <duitoolkit@yahoo.ca> wrote in message news:pan.2004.09.22.03.33.49.466003@yahoo.ca...
> On Tue, 21 Sep 2004 23:21:32 -0400, Ant wrote:
>
> > On Tue, 21 Sep 2004 20:24:58 -0700, Walter wrote:
> >
> >>
> >> "Ant" <duitoolkit@yahoo.ca> wrote in message
> >>> "
> >>> Error: circular initialization dependency with module Leds
> >>> "
> >>>
> >> Two modules that import each other, and each has static initializers.
> >
> > thanks, I'll check that (might take awhile...)
> >
>
> It wasn't critical that my static vars/consts were static so I just made then instance variables and it pass that "circular init" constraint.
>
> I know there was a lengthy discussion about this
> I'll read it before complaining - but this doesn't smell very well :)

The only way the runtime can tell who depends on what is the imports. The rule is that the imported modules must be initialized before the module that imports them. With two modules importing each other, this cannot be resolved.

If there are some changes I can make to the source that would make folding in the newer versions easier, I'd be happy to make them.

"Deja Augustine" <deja@scratch-ware.net> wrote in message news:cir09u$13ga$1@digitaldaemon.com...
> DAMN YOU! :D
>
> I just got D.NET caught up to 0.101 :)
>
> Walter wrote:
>
> > I haven't gotten very far with the MIID list yet, this is mostly just compiler bugs. I'm going to be out for a few days, and I wanted to get
this
> > out first.
> >
> > http://www.digitalmars.com/d/changelog.html
> >
> >
> >

Unfortunately, I've had to modify the front-end to a rather severe degree in order to move, store and interface with the various Managed C++ constructs necessary.

WinDiff is my friend :)

Thank you for the offer, though
-Deja


Walter wrote:

> If there are some changes I can make to the source that would make folding
> in the newer versions easier, I'd be happy to make them.
> 
> "Deja Augustine" <deja@scratch-ware.net> wrote in message
> news:cir09u$13ga$1@digitaldaemon.com...
> 
>>DAMN YOU! :D
>>
>>I just got D.NET caught up to 0.101 :)
>>
>>Walter wrote:
>>
>>
>>>I haven't gotten very far with the MIID list yet, this is mostly just
>>>compiler bugs. I'm going to be out for a few days, and I wanted to get
> 
> this
> 
>>>out first.
>>>
>>>http://www.digitalmars.com/d/changelog.html
>>>
>>>
>>>
> 
> 
> 

"Walter" <newshound@digitalmars.com> escribió en el mensaje
news:ciqo2o$i1p$1@digitaldaemon.com...
|I haven't gotten very far with the MIID list yet, this is mostly just
| compiler bugs. I'm going to be out for a few days, and I wanted to get this
| out first.
|
| http://www.digitalmars.com/d/changelog.html
|
|
|

Did something change regarding casting? I can't send a bug report because there're just so many things to reduce it (including the use of mixins), but I something like this used to compile:

void foo(void * x) { ... }
class A
{
    void bar () {
        foo( cast(void*) this );
    }
}

(this trimmed down example actually compiles, but the real one doesn't) Now I get "can't cast from A to void *". Did something change?

-----------------------
Carlos Santander Bernal