🚀 go-pugleaf

Some stuff:

    1)  The bug mentioned concerning garbage returned from 
        fread() has been fixed.  The fixed version is now
        being shipped and provided as a FREE revisionary update; 
        send disks and a SASE.

    2)  I/O redirection does in fact work, with one caveat:
        the file name MUST follow the angle bracket with
        no intervening spaces.  In other words,

            blah > blah < blah

        will NOT work, while 

            blah >blah <blah
        
        will work.  This will be tended to at some future date.

    3)  I know that everyone (well, everyone except me) HATES the
        text editor because the arrow keys don't do "the right 
        thing" (I suppose according to the ANSI arrow key standard :-)).
        I've heard comments to the effect that the editor is 
        utterly useless because of this "bug".  This is not a bug,
        but rather a design decision which seems to be unpopular.
        This will be changed in the next major release.  Allow me
        to point out that I still believe that for movement of
        over a few lines, the mouse is probably faster once you
        get used to it.  I know, I know, the mouse button is sort
        of wimpy, but your finger gets stronger after a while.
        Anyway, anyone who knows of other flaws which render the
        editor a waste of disk space can send me mail.  Some old
        bugs (specifically, problems with search/replace and with
        the hard disk driver) have been fixed already.

    4)  We have noticed a prevailing belief that in order to use
        the floating point system, programs must be linked to
        the "double.o" code.  This is FALSE.  The "double.o"
        code is only needed for 64-bit precision.  If it is not
        specified, the linker uses a 32-bit library which is MUCH
        faster.  In either precision, double values are stored
        as 64 bits, but the 32-bit libraries only compute with
        (guess how many...) 32 bits.  For casual floating point
        computations, DON'T link with "double.o".  As to the 
        various problems with floating point, all  have been
        fixed and are available with the current revision (the
        free revisionary update mentioned above).  Along with
        the fixes are library routines for absolute value.

PLEASE don't beg me for dates.  I don't know.  Once again, the
current revision is available for no charge.  Call for info.

Mike McNally    Digital Lynx Inc.
                hanger-about at Megamax, and the idiot who likes mice

                ...convex!ctvax!megamax!m5