Minor installation bugs (?)

[Allyn Dimock]

I have mlton-1999-3-19.1 -- if there is a later version then these bugs may
have been fixed.

1) mlton-lib.h refers to <huge_val.h> which is neither a standard .h file nor
is it included in the distribution.  This is presumably part of the standard
math.h header file.

2) In looking at mlton-lib.h, I noticed
#define MLTON_Real_maxFinite 1.79769313486231470e308
whereas limits.h on my systems have (I checked on SGI, and on Sparc systems)
#define	DBL_MAX	             1.7976931348623157E+308

3) the flags in src/Makefile: 
-m486 -malign-loops=2 -malign-jumps=2 -malign-functions=2
are hardware specific.  You might want to omit them.

-- Allyn