On 02/02/2011 19:54, Bill Meier wrote: > Speaking of support [1], for the eventual Wireshark 1.6 release I suggest > consideration of ending support for older Windows compilers/SDK's. > (Wireshark 1.4 would continue to support compilers as now). > > For Wireshark 1.6: > > Support the following ? > > !ELSEIF "$(MSVC_VARIANT)" == "MSVC2005" || \ > "$(MSVC_VARIANT)" == "MSVC2005EE" || \ > "$(MSVC_VARIANT)" == "DOTNET20" || \ > "$(MSVC_VARIANT)" == "MSVC2008" || \ > "$(MSVC_VARIANT)" == "MSVC2008EE" || \ > "$(MSVC_VARIANT)" == "MSVC2010" || \ > "$(MSVC_VARIANT)" == "MSVC2010EE" > > and not these ? > > !IF "$(MSVC_VARIANT)" == "MSVC6" || \ > "$(MSVC_VARIANT)" == "MSVC2002" || \ > "$(MSVC_VARIANT)" == "DOTNET10" || \ > "$(MSVC_VARIANT)" == "MSVC2003" || \ > "$(MSVC_VARIANT)" == "DOTNET11" > > > Comments ? > > > Bill > > > [1] http://www.wireshark.org/lists/wireshark-dev/201102/msg00005.html > ___________________________________________________________________________ > > Possibly taking us off in a tangent, but I think the code detecting the compiler in use is actually wrong. IMHO hat is more important is the SDK that is being used. Each compiler ships with a version of the SDK, but users can easily download a newer SDK to use with an older compiler to gain access to newer OS features in the SDK header files. Oue current approach of tying available OS features to the compiler used is IMHO flawed.
If we went with this approach, the config.nmake and all the compile time conditionals could be cleaned up a lot. -- Regards, Graham Bloice
___________________________________________________________________________ Sent via: Wireshark-dev mailing list <wireshark-dev@wireshark.org> Archives: http://www.wireshark.org/lists/wireshark-dev Unsubscribe: https://wireshark.org/mailman/options/wireshark-dev mailto:wireshark-dev-requ...@wireshark.org?subject=unsubscribe
