I'm having a hard time trying to compile with gdc. I make use of windows functions from std.path, but it seams that the version is not set to Windows, and so I get an ugly assert(0) from path.d.

I've tried using -fversion=Windows, but then I get an error that this is a reserved word. Any ideas?

Thanks,
Mariano.

Marjancek wrote:
> I'm having a hard time trying to compile with gdc. I make use of
> windows functions from std.path, but it seams that the version is not
> set to Windows, and so I get an ugly assert(0) from path.d.
> 
> I've tried using -fversion=Windows, but then I get an error that this
> is a reserved word. Any ideas?
> 
> Thanks,
> Mariano.

I have encountered the same problem. This seems to be the fault of gdc, which should probably set version=Windows, not only version=Win32, as it does now (or at least allow us to set -fversion=Windows by ourselves).
In the meantime, an "acceptable" fix is to change all occurences of version(Windows) in the header files to version(Win32). Ugly, but there you go.

Downs wrote:
> Marjancek wrote:
> 
>> I'm having a hard time trying to compile with gdc. I make use of
>> windows functions from std.path, but it seams that the version is not
>> set to Windows, and so I get an ugly assert(0) from path.d.
>>
>> I've tried using -fversion=Windows, but then I get an error that this
>> is a reserved word. Any ideas?
>>
>> Thanks,
>> Mariano.
> 
> 
> I have encountered the same problem. This seems to be the fault of gdc, which should probably set version=Windows, not only version=Win32, as it does now (or at least allow us to set -fversion=Windows by ourselves).
> In the meantime, an "acceptable" fix is to change all occurences of version(Windows) in the header files to version(Win32). Ugly, but there you go.

Are you using Cygwin?  Enabling version(Windows) or version(Win32) would cause a binary incompatibility with Phobos.  I am thinking it would be better to just remove the restrictions from the Windows headers.

David

David Friedman schrieb:
> Downs wrote:
>> Marjancek wrote:
>>
>>> I'm having a hard time trying to compile with gdc. I make use of
>>> windows functions from std.path, but it seams that the version is not
>>> set to Windows, and so I get an ugly assert(0) from path.d.
>>>
>>> I've tried using -fversion=Windows, but then I get an error that this
>>> is a reserved word. Any ideas?
>>>
>>> Thanks,
>>> Mariano.
>>
>>
>> I have encountered the same problem. This seems to be the fault of gdc, which should probably set version=Windows, not only version=Win32, as it does now (or at least allow us to set -fversion=Windows by ourselves).
>> In the meantime, an "acceptable" fix is to change all occurences of version(Windows) in the header files to version(Win32). Ugly, but there you go.
> 
> Are you using Cygwin?  Enabling version(Windows) or version(Win32) would cause a binary incompatibility with Phobos.  I am thinking it would be better to just remove the restrictions from the Windows headers.
> 
> David
> 
I'm using MinGW32 GDC against GCC 4.0.3, built from SVN.
I made a version which has "Windows" removed from the list of protected identifiers.
So far, it seems to do the job. :happy:
Greetings

Downs wrote:
> David Friedman schrieb:
> 
>> Downs wrote:
>>
>>> Marjancek wrote:
>>>
>>>> I'm having a hard time trying to compile with gdc. I make use of
>>>> windows functions from std.path, but it seams that the version is not
>>>> set to Windows, and so I get an ugly assert(0) from path.d.
>>>>
>>>> I've tried using -fversion=Windows, but then I get an error that this
>>>> is a reserved word. Any ideas?
>>>>
>>>> Thanks,
>>>> Mariano.
>>>
>>>
>>>
>>> I have encountered the same problem. This seems to be the fault of gdc, which should probably set version=Windows, not only version=Win32, as it does now (or at least allow us to set -fversion=Windows by ourselves).
>>> In the meantime, an "acceptable" fix is to change all occurences of version(Windows) in the header files to version(Win32). Ugly, but there you go.
>>
>>
>> Are you using Cygwin?  Enabling version(Windows) or version(Win32) would cause a binary incompatibility with Phobos.  I am thinking it would be better to just remove the restrictions from the Windows headers.
>>
>> David
>>
> I'm using MinGW32 GDC against GCC 4.0.3, built from SVN.
> I made a version which has "Windows" removed from the list of protected identifiers.
> So far, it seems to do the job. :happy:
> Greetings

Both Win32 and Windows are supposed to be enabled for MinGW32.  Can you send me the contents of

<gdc prefix>/include/d/4.0.3/mingw32/phobos-ver-syms

(The 'mingw32' dir may be something like 'i686-pc-cygwin'.)

Thanks,

David

David Friedman schrieb:
> Downs wrote:
>> David Friedman schrieb:
>>
>>> Downs wrote:
>>>
>>>> Marjancek wrote:
>>>>
>>>>> I'm having a hard time trying to compile with gdc. I make use of
>>>>> windows functions from std.path, but it seams that the version is not
>>>>> set to Windows, and so I get an ugly assert(0) from path.d.
>>>>>
>>>>> I've tried using -fversion=Windows, but then I get an error that this
>>>>> is a reserved word. Any ideas?
>>>>>
>>>>> Thanks,
>>>>> Mariano.
>>>>
>>>>
>>>>
>>>> I have encountered the same problem. This seems to be the fault of gdc, which should probably set version=Windows, not only version=Win32, as it does now (or at least allow us to set -fversion=Windows by ourselves).
>>>> In the meantime, an "acceptable" fix is to change all occurences of version(Windows) in the header files to version(Win32). Ugly, but there you go.
>>>
>>>
>>> Are you using Cygwin?  Enabling version(Windows) or version(Win32) would cause a binary incompatibility with Phobos.  I am thinking it would be better to just remove the restrictions from the Windows headers.
>>>
>>> David
>>>
>> I'm using MinGW32 GDC against GCC 4.0.3, built from SVN.
>> I made a version which has "Windows" removed from the list of protected identifiers.
>> So far, it seems to do the job. :happy:
>> Greetings
> 
> Both Win32 and Windows are supposed to be enabled for MinGW32.  Can you send me the contents of
> 
> <gdc prefix>/include/d/4.0.3/mingw32/phobos-ver-syms
> 
> (The 'mingw32' dir may be something like 'i686-pc-cygwin'.)
> 
> Thanks,
> 
> David

I don't seem to have that folder. Temporary confusion ensues.
There is, however, an include/d/4.0.3/i686-pc-mingw32/phobos-ver-syms, so I'll append that instead.
It consists of
Windows
[some empty lines]
GNU_Have_fwide
GNU_Have_strtold
[some empty lines]

Hope that helps.