🚀 go-pugleaf

In article <9102011902.AA26871@excalibur.lcs.mit.edu> gildea@EXPO.LCS.MIT.EDU (Stephen Gildea) writes:
>LOAD_AVE_TYPE and LOAD_AVE_CVT are defined by all the m- files, so s-
>files should not define them or the compiler may complain.  These
>definitions should be removed from s-usg5-4.h of 18.56.

Isn't LOAD_AVE_TYPE and LOAD_AVE_CVT more an operating system issue than
hardware issue?  For example, Xenix, SysVr3, SysVr4, and MACH all run on
PC-architecture i386 boxes; but there isn't a set of values that is
appropriate to all of these systems than can be placed in m-intel386.h.

I just looked and seen that m-intel386.h defines the load average 
parameters for Xenix if XENIX is defined.  Certainly the load average
information for MACH and SysV.4 could be similarly defined here.  But
what is gained?  This makes m- files complex, and there is a good chance
that it will get out of sync with the corresponding s- file. 

Perhaps a reasonable comprimise would be for the m-files to check to
see if LOAD_AVE_TYPE, etc. has been defined and only define the values
if they haven't.  In this way, A s- file would be able to overide the
defaults selected in any m- file.

-- 
J.T. Conklin    UniFax Communications, Inc.

                CADnet Inc, San Ramon California
                jtc@motcad.portal.com, ...!portal!motcad!jtc