On Wed, 21 Apr 2010, Clinton Jeffery wrote: > One obvious thing to consider is whether it is time to remove all the VMS > code. Can anything > give any reason not to? :-)
I suppose you could ask at http://www.openvms.org/ which still seems to be alive (mentioning activity this month). You could frame the question along the lines of "If none of your community are able/willing to help us support you, then we have insufficient resources to do so, therefore does anyone want to help?", I suppose. Getting help from people experienced in obscure[?] portability issues could be useful more widely. > > Clint Hugh, probably being too idealistically inclusive... > > On Tue, Apr 20, 2010 at 11:44 AM, James Albert > <[email protected]>wrote: > > > I saw the same problem building 11.7 on my amd64 debian box. > > The problem is the in the sys.h, where around line 373 the include of > > zlib.h is > > surrounded by an undef of VMS, then a re-define. This is because tests > > of the type #ifdef VMS are present in the zlib includes. However, well > > before > > this at line 293, there is an include of png.h, which in turn includes > > zlib.h, but > > without the undef'ing, which means in zlibdefs.h there is an unhappy > > attempt > > to include unixio.h. This seems to be the only place where the VMS stuff > > causes > > an issue; I shamelessly put a #undef VMS directly into zlibdefs.h rather > > than figure > > it out at the time, but you could put the protective undef/define wrapper > > around the > > include of png.h if you wanted to be more correct. I hope this helps, late > > though > > it is. > > > ------------------------------------------------------------------------------ _______________________________________________ Unicon-group mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/unicon-group
