.bp
.H 1 "Future directions"
.sp
.P
Some plans for knotEd include:
.sp
1)  Performance enhancement.  The routine find_cross checks edges that
are already known not to cross every time it is called.  It could
keep track of this and reduce the time complexity of the routine.
Many other calculations could be improved.
.sp
2)  Kirby Calculus and other difficult operations should be done by the
editor.  This way the user benefits in (hopefully) three ways where he
uses knotEd:  basic operations are easy, operations are all checked for
legality (isotopy lock) and complex operations are entirely automatic.
.sp
3)  The ability to merge two stored knots into one.  The data structures 
rely heavily on the absolute location of records in an array (this
made find_cross INFINITELY easier) so a little code would be needed here.
.sp
4)  Machine independent storage.  Current method of storing knots uses
fwrite to write out records this is machine dependent and unreadable
to both humans and other programs.  Some simple grammar would be sufficient
for this task.
.sp
5)  More methods of getting hard copy.  Currently hardcopy comes only through
Tektronics format or saving the screen to a file.  The ability to
draw in PostScript, MetaFont, or some grammar would be nice.
.sp
6)  Smarter redraw.  Currently we update the whole screen.  The ability
to redraw portions would improve performance and cut down on annoying 
flicker.
.sp
7)  More knot theory.  Actually have the program try to reduce knots into
a simpler form.
.sp
8)  More calculations available.  Calculating software (ala Goldschmidt, Walker,
and Baxter) make knotEd more useful (and vise versa).
.sp
9)  More knots.  Somebody (I don't have the time at the moment) should
sit down with Rolfsen's knot theory and draw all the knots in the
appendix into the knot library.
.sp
10) knotEd ported to other machines.  This should be easy as the program
I based knotEd on ported easily.
.sp
12) Some decent documents.
