Re: help me with WIN95 value of version16()

[Gerard Patel]

At 08:34 PM 26/08/2001 -0700, you wrote:
> > The return value of  GetVersion16() for win95 changed recently from 0x5f03 to
> > 0x304 in misc/version.c (patch 1.43 -> 1.44).  The version of "real"

win95 I use
> > (Czech, win95 2nd ed) gives 0x5f03 while "wine --winver win95"  gives 0x304. 
> > As I do not have the US version of win95 I am not able to determine the right
> > value.
>
> I would like to help you and I have a Win95 system around. But I
> don't have a Win16 compiler as these things are getting quite rare.

I can compile and run 16 bits apps, on win3.1 if needed (it's a pain
because it's old hardware I hardly use otherwise)

GetVersion 16 bits must be a special case, returns 3.95, while 32 bits apps
get 4.00

Gerard

help me with WIN95 value of version16()

[Francois Gouget]

On Mon, 27 Aug 2001, Gerard Patel wrote:
[...]
> I can compile and run 16 bits apps, on win3.1 if needed (it's a pain
> because it's old hardware I hardly use otherwise)

:-)
> GetVersion 16 bits must be a special case, returns 3.95, while 32 bits apps
> get 4.00

Yes, I think I read something about that at the time.
   IIRC, MS messed up with GetVersion16. They documented
HIWORD(GetVersion16()) as returning the major version number when it was
in fact in the lower version number or vice-versa. Because of that when
they switched to Win95 they wanted the new version number to be
greater than the old, no matter how you read it, hence the 3.95:
  0x5f03: 3.95 > 3.1 and 95.3 > 1.3
  0x0304: 4.3 > 3.1 and 3.4 > 1.3
  0x0004: 4.0 would not do since 0.4 < 1.3

Now, why would we have a version with 3.95 and another with 3.4? Is
it a Win95 vs Win95 OSR2 issue? Localization should not affect this
(though everything is possible).

--
Francois Gouget         [e-mail]        http://fgouget.free.fr/
    I haven't lost my mind, it's backed up on tape around here somewhere...