LibF77 News - most recent first August 9, 1985: add d_dprod() for dprod() when -r8 flag is used. July 6, 1985: switch r_lg10 and d_lg10 to use new log10 in -lm. switch functions of form r_*.c to be declared as float instead of double. use -f flag for most single precision routines to get around C bug; also put extra stores in 4 routines to bypass bug (See Usef in Makefile & file Test_float). June 26, 1985: move f77_abort() from libF77/main.c to libI77/f77_abort.c so C progs can call f77 subroutines. fix pow_ii to divide by zero for i**j where i=0 and j is < 0 fix pow_hh by applying all fixes from pow_ii to it. June 6, 1985, 4.3BSD beta version: all error exits now go through f77_abort(). Normally get dump only if loaded with -lg (via f77 -g). Can use f77_dump_flag envir. var. to override that (set it to 'y' or 'n'). new routine: random() delete sinh.o, tanh.o - identical to C routines in libm cabs.o, d_sqrt.o, z_abs.o - now in assembler in libm bit.o: not documented, man 3f bit now shows better ways. routines in bit.o also had inverse arg. order to built in funct. complex division by zero - now get error comment.