Re: winamp patch A0

[Vincent Béron]

Le sam 09/11/2002 à 11:22, Andreas Mohr a écrit :
> Hi all,
>
> IMHO we should now switch to Win98 SE as default Windows version.
> Too many programs expect at least Win98 to be running already.
>
> Winamp also checks the Version registry keys, so I added them to
> winedefault.reg.
> Index: winedefault.reg
>
> RCS file: /home/wine/wine/winedefault.reg,v
> retrieving revision 1.58
> diff -u -r1.58 winedefault.reg
> --- winedefault.reg	25 Oct 2002 19:05:19 -0000	1.58
> +++ winedefault.reg	9 Nov 2002 16:20:36 -0000
> @@ -72,6 +72,9 @@
>  "ProgramFilesDir"="C:\Program Files"
>  "RegisteredOrganization"="Change preferred organization in ~/.wine/system.reg"
>  "RegisteredOwner"="Change preferred owner in ~/.wine/system.reg"
> +"SubVersionNumber"=" A "
> +"Version"="Windows 98"
> +"VersionNumber"="4.10.2222"

This doesn't seem the place to do it, since if we specify something
different (in config) then the registry keys do not match, and some
other app could possibly have some problems.

Would it be possible to generate automatically those keys with a call to
GetVersionEx or something similar?

Vincent

winamp patch A0

[Andreas Mohr]

On Sat, Nov 09, 2002 at 11:43:32AM -0500, Vincent Béron wrote:
> This doesn't seem the place to do it, since if we specify something
> different (in config) then the registry keys do not match, and some
> other app could possibly have some problems.
>
> Would it be possible to generate automatically those keys with a call to
> GetVersionEx or something similar?

I *knew* that it'd receive this reply ! ;-\

I've also been thinking about doing it semi-dynamically after a call to
GetVersion*(), but I was a bit unsure about it.
Should we do it like that ?

Note that these registry keys are a *global* thing (wineserver !), whereas
the version value returned depends on the wine *instance* ("smart" detection
algorithm). This could spell trouble...

winamp patch A0

[Greg Turner]

On Saturday 09 November 2002 11:08 am, Andreas Mohr wrote:
> Note that these registry keys are a *global* thing (wineserver !),
> whereas the version value returned depends on the wine *instance*
> ("smart" detection algorithm). This could spell trouble...

indeed... why not just virtualize this registry key on the client side 
of the wineserver to return the right thing for a given process?