Re: [freesci-develop] State of Glutton on Win32

[Hugues Valois]

Hi Walter/all,

I'm not sure if by "switch to mingw" you are actually suggesting deprecating 
support for Visual C++. It sounds like you are, and if that's the case, I 
would have to strongly disagree. Of course I have no problems with "adding 
support for mingw", or any other compiler/tool.

If I misunderstood you, then you can skip the following rant/opinion :)

As a Visual C++ user (for about 10 years), I certainly wouldn't be 
interested in switching to a different product that makes ME less 
productive.  I know I would never have bothered compiling FreeSCI myself if 
there was no VC++ project, much less implement a game selection screen.

Now the fact that Hugues Valois doesn't want to switch from VC++ to gcc 
probably doesn't mean anything to the FreeSCI team.  After all, I'm not 
contributing anything right now (except this long and boring email).  My 
main interest at the moment is to be able to get the latest code from CVS, 
build using VC++, and test games.  I plan on doing that when a glutton 
release is near.

My opinion is that dropping VC++ support will not help motivate and attract 
typical windows C++ programmers (a majority of which use VC++ and like the 
product) to help out the FreeSCI team.  Adding gcc support may help in 
attracting students that want to learn for free.  Dropping VC++ support 
would detract more experienced windows programmers that use VC++ at work.

That's my own prediction.  If it turns out I'm wrong, I'll gladly admit it.  
I will be very happy if more people are working on the win32 port!

It's true that Visual Studio (the suite) can be expensive.  But to compile 
FreeSCI you only need Visual C++.  You can purchase Visual C++ 2003 Standard 
Edition for 89$US at amazon.com.  It's not free, but I wouldn't qualify that 
as "very expensive software".

What's the cost/disadvantages of keeping the Visual C++ support?
- Project files need to be updated separately from the makefiles
- Can't use gcc extensions*
- Can't use latest C standard*

* Compiling with VC++ and gcc ensures that the C code is more "generic" than 
it would be if only gcc was used.  This seems to me like a good thing, worth 
the occasional annoyances.  Not sure about this, but can gcc be used to 
build for Windows Mobile platform?  If not, Embedded VC++ or VC++ 2005 would 
be required.

In summary, the points I am trying to make are that
- Support for Visual C++ builds isn't so costly that it should be abandoned
- Visual C++ is not that expensive and it's widely used
- Dropping support for Visual C++ in favor of mingw won't attract that much 
more qualified developers to help with the win32 port
- Adding support for mingw could be worthwhile, but not at the expense of 
VC++

That's my 2 cents,

hugues


> I know I've said this before, but I think it would be better to switch to 
> mingw for the windows port. That way we could integrate it into the 
> automake/autoconf system. Another point is that Visual Studio is very 
> expensive software and this severely limits the number of people that might 
> be able to help out with the windows port.
>
> Regards,
>
> Walter van Niftrik