Re: Development Status Update 3

[John Fortin <fortinj at attglobal dot net>]

Thanks Harold.

I haven't had much time lately.  I am getting ready to install a production
system at work so I've been busy and will be till the end of the year.  I'll
monitor and add in as I can.

John

Harold Hunt wrote:

> Sometimes a problem can be daunting because you don't know the size of it;
> fortunately, the header issue turned out to be fairly simple.
>
> Identifier collisions between X and Win32 headers come down, so far, to
> three typedefs or #define'd data types, one function, and one macro and
> function collision.
>
> Data types:
> ATOM (Win32: WORD, X: unsigned long)
> BOOL (Win32: int, X: unsigned char)
> BYTE (same datatype, one #define'd, the other typedef'd, I think)
>
> Functions:
> FreeResource()  (Win32: deprecated, so it can be #ifdef'd out)
>
> Macro/function:
> CreateWindow (Win32: macro, can be #undef'd, X: part of screen structure,
> function prototype)
>
> For now I munged a couple header files to allow compilation, albeit with
> warnings; I'll have to go back later and thoughtfully address each of the
> five cases identified so far to determine a more appropriate resolution.
>
> I have removed all calls to functions residing in xf_dx.dll and replaced
> them with code in XWin.exe; I have not yet torn out all stray references to
> xf_dx.dll.
>
> My to-do list, slightly amended from yesterday, is as follows:
> 1)  Finalize header file fixes and produce diffs
> 2)  Finalize Imakefile changes and produce diffs
> 3)  Remove references to xf_dx.dll from XWin.exe code
> 4)  Standardize any identifier names that have slipped through
>         the cracks
> 5)  Try to eliminate as many build warnings as possible
> 6)  Remove deprecated files
> 7)  Add some comments/tips
> 8)  Tarball the source and diffs
> 9)  Release the framework for other developers to examine
>
> Keep in mind that we will still have to do some design work before we start
> implementing functions in the framework... though we may be able to throw
> together a machine independent implementation fairly quickly just to have
> something that works.
>
> I'll keep everyone posted,
>
> Harold