Fri May 19 00:27:16 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c) well, one last change: fixed bug in
	ring-from-rows/cols, that core dumped on 0 variables. 

	* (version.c) VERSION 3.0!! No more changes to the code allowed.
	without DAMN good reason!

	* (makefile.sun) changed CFALGS to -DSUN -O to optimize.

	* (ring.h) made NVARS depend on SUN vs MAC.  512 vars allowed on
	SUN's.

	* (shell.c cmdnames.c) changed "silent" to "shout", and changed
	semantics: it now always displays the result of whatever command
	it runs, no matter what the setting of prlevel is.

	* (qring.c) isQRing now checks for NULL value.  If NULL, returns
	FALSE. Fixes core dump problem in ring-sum when one of the
	rings has 0 variables.  

	* (standard.c input.c stash.c set.c) Made the following changes 
	from Dave: added "abort" set variable
	     fixed up showpairs set variable
	     branch labels now can have leading blanks before label.
	     newer subs_names

	* (cmdnames.c rmap.c gm_cmds.c) command name changes:
	  to-ring ===> fetch  (edit-map no longer called from this)
	  tr ===> transpose
	  sub ===> subtract

Tue May 16 19:51:46 1989  Mike Stillman  (mike at mssun8)

	* (betti.c set.c standard.c) added "showpairs" set variable a la
	Dave.

	* (gm_cmds.c) fixed random_cmd so it will take 0 rows or cols.

	* (qring.c) previously a non-monic ideal forced as a std basis
	into a qring didn't make the ideal monic: now it does.

Tue Apr 25 00:07:32 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c) fixed pres command so that it won't display index 0
	matrices either.

	* (betti.c) fixed betti command so that it won't display ".0"
	indexed matrices.

	* (gm_cmds.c) fixed iden command so that it allows 0 rows.

	* (monitor.c human_io.c) "what?" prompt for input now gets
	displayed even if prlevel is set positive.

	* (input.c hilb.c stash.c) made Dave's changes: fixed several
	portability problems, #0 now means number of parameters, #(i),
	#(@i), etc all work fine.  Added set variable "showmem" which
	causes memory usage to be displayed even if verbose isn't set.

Thu Apr 20 00:32:36 1989  Mike Stillman  (mike at mssun8)

	* (gm_poly.c) "contract" and "diff" had the major bug that the
	coef. of f in f.g was never considered!! Fixed the routine
	p_contract.  I can't believe we didn't notice this one before!

Thu Apr  6 17:09:10 1989  Mike Stillman  (mike at mssun8)

	* (many files) Dave removed extraneous local vars, etc. so that M
	will run under MPW 3.0.  Also various return values were inserted.
	I checked the ones in "my" part of the code, and they all look ok
	except for those in "monideal.c": they were previously true bugs:
	return values WERE expected: so I put them in.  "degree" should
	now work again.  Also, #pragma doesn't work on SUN's (at least
	mine), so I changed them to " #pragma".

Wed Apr  5 15:44:32 1989  Mike Stillman  (mike at mssun8)

	* (gm_cmds.c) fixed memory leak in "submat" command.

	* (ring.c) fixed memory leak in "ring" command.  Note however that
	every ring which is created does in fact lose some memory: the
	names of the variables and the size of the ring record itself.
	However, this is a small amount of memory compared to the loss of
	arrays.  Also fixed bug where each variable could not have degree
	larger than 512.  Changed the max degree to "maxdegree".

Tue Apr  4 16:35:43 1989  Mike Stillman  (mike at mssun8)

	* (koszul.c) changed koszul so that it installs the ring of the
	first argument if it is a matrix.

	* (qring.c cmds.c) routine "isQRing" added to qring.c  It's now
	used in cmds.c "ringsum_cmd" to determine if a quotient ring
	should be created.

	* (parse.c) fixed '&' (integer remainder) so that it works.

	* (rescmds.c) modulo command: finally found the memory trashing:
	various polynomials were being used without copying, and then
	killed. So I added p_copy in modulo routine.

Fri Mar 31 19:44:14 1989  Mike Stillman  (mike at mssun8)

	* (help.c) "help" command now works again, although the help file
	must be cleaned up alot yet.  "help" is now "menu-driven" by topic.

	* (all files) Added Dave's changes, which include copyright
	notices, and various changes to input.c to aid in debugging scripts.

Mon Mar 27 15:11:32 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c cmdnames.c) added command "is-zero" which tests whether
	a polynomial is zero.

	* (boxcmds.c gm_cmds.c) changed modlift_cmd to lift_cmd, and
	lift_cmd to mklift_cmd.

	* (parse.c parse.h, and input.c monitor.c) Added relational
	operators >, >=, <, <=, =, !=, and unary ! (NOT).  Fairly easy
	change.  I'm not sure I have a good precedence set for '!', but it
	seems ok...  The main problem is that '=' was used for an input
	prompt, creating all sorts of havoc.  At least for now, the input
	prompts have all been changed to '%' (PROMPT2 in input.c, and in
	the routine "prinput" in monitor.c.
	Note however, that this change is only for integers, not
	polynomials!! We need a command which tests whether a polynomial
	is zero. 

Fri Mar 24 12:40:55 1989  Mike Stillman  (mike at mssun8)

	* (qring.c cmds.c) added command "present-ring" which grabs the
	quotient ideal, putting it over the correct ring.
	NOTE: at this point all commands should be COMPLETELY consistent
	with quotient rings (at least homogeneous quotient rings).  If
	there are any inconsistencies, let me know.  The only problem is
	with the command "istd", which doesn'y even recoqnize quotient
	rings at all.  However, this command might be going bye bye soon
	anyway. 

	* (qring.c cmds.c) ring-sum now correctly merges quotient ideals.

	* (qring.c) qring now correctly will work with taking a quotient
	ring of a quotient ring.  Also, for consistency with the other
	ring commands, an ideal is returned consisting of all of the
	variables.

	* (hilb2.c hulb.c) changed "hilb", "hulb" to call the routine
	"monhilb" of monideal.c: this adds in Rideal if there is one, and
	in any case cleans up the code.

	* (monideal.c) changed "monmake" routine to take function argument
	AND use Rideal.  Therefore, codim, degree, k-basis, truncate all
	work correctly with quotient rings.  Remaining: hilb, hulb, hull.

Thu Mar 23 13:31:54 1989  Mike Stillman  (mike at mssun8)

	* (gmatrix.c) changed "iden", "random" so that they handle
	quotient rings where 1 = 0. (Is this a nit, or what!!!????)

	* (ed.c) changed gmR_mult to check for 0 element (found this bug
	while making everything consistent with qrings's).

	* (many files) modified the commands: cat, ideal, imap, int,
	k-basis, mat, poly, ring, ring-from-cols, ring-from-rows, rmap,
	sparse, submat, to-ring, truncate; the previous value of the
	"result", (whatever used the same name) can now always be used
	inside of EVERY command, e.g. "int i i-1".  Any case where this
	can't happen is now a bug, and should be brought to my attention.

	* (stdcmds.c) added command "std-minimal" which given a var list,
	finds the subset of a standard basis s.t. in_x(f) is a minimal
	generator of in_x(I).  Every such element is taken: this is diff.
	from "stdpart".  Useful in conjunction with "in" and "coef" for
	finding top(I), and locus where a projection isn't flat.
	(Actually, the locus where in(I-fiber) is constant).

	* (human_io.c) changed setdeggens so that prompt message wasn't so
	misleading.

	* (set.c cmdnames.c) added a new command: "set-value", which
	extracts the value of a "set" variable.

	* (monideal.c) Fixed "truncate" command so that it only inserts
	unit vectors if they are not in the submodule.  By the way, 
	using "truncate" on a zero ideal gives the same result as "power"
	applied to the variables of the ring, but is MUCH faster, for
	large powers.

	* (koszul.c) Fixed koszul command so the first argument can be :
	1. integer expression, or 2. matrix identifier.

Wed Mar 22 13:02:57 1989  Mike Stillman  (mike at mssun8)

	* (help.c cmdnames.c) added a new command called "commands", which
	nicely displays a list of ALL the commands of Macaulay: it gets
	the list internally; no help file is required.

	* (cmdnames.c shell.c set.c makefile) Added a new file
	"cmdnames.c" which is a list of all commands, and their routines,
	in alpha order.  command ABBREVIATIONS are now allowed.  This also
	works for "set" variables.  Note: several commands were removed 
	in the process: 
	   all mkres commands ...
	   new
	   attach, wait, detach, sighup

	* (human_io.c) changed mat_scan, sparse_scan so that if you give
	zero rows, it returns a 0 by n matrix, where n is now prompted for.

	* (set.c smallmon.c iring.h monideal.c types.h) Added a
	new "set" variable, "maxdegree", initially set to 512.  If this is
	changed, then larger (or smaller) degree monomials are allowed.
	Notice that this only is useful if the number of variables in the
	block is small: otherwise Macaulay sets the degree even smaller.
	However, it allows extremely high degree calculations at the
	expense of space.  Specifics: MAXD define removed from iring.h,
	"maxdegree" added to set.c, extra test put into "findBlock", 
	extra field "defaultmax" added to symmBlock structure (so that
	testing for equality of two symmBlocks would be easy), and all
	"set" variables are now defined in types.h.
	CURRENT PROBLEM: "gimmy" doesn't know how to give more than 64K or
	so, therefore, the current limit on degree is around 14000.

	* (gm_cmds.c gmatrix.c) Changed reduce_cmd, so it makes sure that
	the number of rows of matrix to reduce isn't too big: the
	resulting matrix was screwed up in this case.

	* (gmatrix.c) Changed random command, so that if a zero is found,
	it doesn't create a bad polynomial.

Mon Mar  6 15:19:48 1989  Mike Stillman  (mike at mssun8)

	* (makefile bugs.remaining) Added a new file of remaining bugs and
	things to add to Macaulay.  Feel free to add things to this list
	(mark it DB, and give more detail than I did!!).

	* (help.c) when displaying list of help, take into account
	"prcomment".

	* (input.c) get_cmd_line: if ac=0, loop screws up (Dave, is my
	change ok????).

Mon Feb 27 18:16:06 1989  Mike Stillman  (mike at mssun8)

	* (gmatrix.c) fixed yet another bug in jacob: column degrees were
	being added for every column, even if not every variable was
	selected. Simple one line fix.

Sun Feb 26 22:09:28 1989  Mike Stillman  (mike at mssun8)

	* (input.c shell.c monitor.c) Added Dave's new input.c file,
	made various other changes he had done:
	    prcomment works better with flushnum (gm_dlist.c human_io.c) 
	    if, jump commands added
	    topen fixed.

Thu Feb 23 19:18:34 1989  Mike Stillman  (mike at mssun8)
	* (cmds.c shell.c) renamed and fixed up "pdegs", "pcoldegs" to
	"row-degs" and "col-degs".  Now they take an optional argument
	which, if given, is set to a n by 0 matrix with row degrees = row
	(or column) degrees of the original matrix.

Tue Feb 21 01:13:55 1989  Mike Stillman  (mike at mssun8)

	* new commands added: "k-basis", and "truncate".  Truncate takes a
	a matrix as input, and a degree d, and returns a basis for the
	module in degree d and in higher dimensions.  The result is a
	matrix with the same number of rows as the orig. matrix: one must
	do "modulo" to actually present this matrix.
	"k-basis" takes a matrix, and a lo and high degree, and gives a
	k-basis for the k-vector space they generate.  Notes:
	(1) both commands expect a standard basis, and will just give a
	warning message if it's not (like all other such commands)
	(2) both commands take a set of variables: the output to both
	commands is a matrix with each column just a monomial in this set
	of variables.

	* (set.c) changed default value for prcomment to 1.

	* (monideal.c shell.c) adding various truncate and k-basis
	commands.  So far, "k-basis" has been added, and seems to work.

	* (rescmds.c shell.c) added "lift-std" command: creates a standard
	basis AND change of basis matrix, and throws away all syzygies.

	* (gmatrix.c) "jacob" didn't set degrees in homog. fashion.  That
	has now been corrected.

	* (gmatrix.c) "mult" didn't handle matrices with 0 rows, columns
	very well: now that is fixed (gm_mult).

	* (intersect.c) "quotient" command didn't work if I, J are both
	modules. One line fix in mkquotient().  Also, if degrees don't
	work out, the "computation" variable is set instead with the aux.
	matrix.  This will facilitate use in the inhomog. case.

	* (input.c) Lifted new path stuff from file input.dave.c on my
	mssun7 account. (new #includes's and one whole new file).

Mon Feb 20 20:40:21 1989  Mike Stillman  (mike at mssun8)

	* (shell.c) "continue" command added: this is now the way to
	continue from interrupted computations.  "exit" is the same as it
	used to be (so it is a little harder to make a mistake).

	* (monitor.c shell.c) Added a variable "flushnum" which is the
	number of chars (approx.) output to prflush (which is used for
	verbose output).  After it hits 50, it is reset, and a newline is
	displayed. routines changed: shell(), and prflush().

	* (makefile) added DAVECHANGES to "tar" creation.

	* (input.c) made changes that Dave made, in DAVECHANGES. 2 notes:
	In subs_names(), I had to remove a spurious line, which gave a
	syntax error, and seems to now work ok..., and I still have to
	make changes in space_cmd to add newline's for prcomment stuff.
	Otherwise all changes from DAVECHANGES as of 2/20/89 (6:00p) have
	been made.

	* (ring.h) made change for NVARS=100 for Mac version to compile ok.

	* (shell.c help.c makefile) added a new command, "help-file" which
	displays each line in the file beginning with 3 semicolons (the semi-s
	aren't displayed).  I created a new file for this, so CHANGE
	makefile for the MAC!!!

Fri Feb 17 18:07:51 1989  Mike Stillman  (mike at mssun8)

	* (gmatrix.c) made change to "jacob" suggested by Ian (good 
	change too!)

	* (stdcmds.c) made a change in force_std: a poly wasn't being
	copied, causing generally bad things to happen. Therefore, the
	"forcestd" command seems to now be working.

Thu Feb 16 12:00:42 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c) changed do_ring_vars() so it doe nothing if no current 
	ring.  (Fixed a bug, where if you typed 0 for #vars, you get a
	core dump.)

	* (spcmds.c) [made 2/15/89] changed and fixed "coef" command.  It
	now works, and also returns the monomial ideal of monomials it used.

Tue Feb 14 12:53:54 1989  Mike Stillman  (mike at mssun8)

	* (gm_cmds.c) changed iden_cmd so that it checks for a current
	ring.  If there isn't one, it complains, and exits.

Mon Feb 13 00:31:45 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c gm_dlist.c gmatrix. human_io.c) changed setdegs, and
	setcoldegs so that they always prompt for entire dlist: if return
	is typed, zero rows/columns are set to 0.

	* (vars.c) changed find_var so that when it sees an indexed
	variable, which it can't find, it looks for ring'var, where ring
	is the base ring of the computation.

	* (rescmds.c vars.c) Changed getIdentifier so that it tries to
	parse an integer after seeing a "." (unless it sees ".."). Changed
	the routines std, istd, syz, so they name their partial variables
	integers. (std, istd, syz: result is always "1".  syz, etc. aux.
	variable is "0").

	* (cmds.c gmatrix.c) Ring commands now create a 1 by numvars 
	matrix of their variables, as well as ring'zero matrix.

	* (vars.c) Fixed yet another bug: box routines call "make_var"
	with null name arguments: getIdentifier didn't like this at all,
	so a fix was needed.

	* (vars.c gm_dlist.c) Fixed two bugs: the ".." problem mentioned
	below: an identifier cannot have two .'s in a row.  Also, find_var now
	handles braces correctly.

	* (gm_dlist.c ring.c) Wherever a degree list is expected, you can
	now put a matrix name, and its row degrees will be used instead.
	Actually, such a matrix name is only one "item" in a degree list:
	you can tag on numbers after it, or before it, etc.  Also changed:
	in rings, if a non-positive degree is given to a variable, all the
	degrees are incremented by the same amount to make them positive.
	Also, if a weight vector has negative numbers in it, a multiple of
	the degrees of the variables is added to it, so that they will be
	non-negative. (Fixing a bug in this case).  PROBLEM REMAINING: if
	a variable is given in a degree list, such as a..b, then the
	getIdentifier routine currently eats the .'s so that this will
	appear to be one identifier: not exactly desired.

	* (cmds.c) Fixed bug in new "poly" command (it checked wrong
	number of arguments).

Sun Feb 12 22:52:58 1989  Mike Stillman  (mike at mssun8)

	* (vars.c parse.c parsepoly.c) a new and improved getIdentifier
	routine was added, which allows "braced" variables (vars enclosed
	in braces), quotes, @'s.  "Absolute" variables were removed.  All
	user defined variables must now be of this restricted form.

	* (input.c) strlen didn't check whether NULL string is given
	to it.  Anyway, this whole path crap doesn't work on the SUN.

	* (input.c h_grow.c h_tope.c h_interface.c hulb.c) Dave
	added path's (I think), and @ expansion inside scripts, for local
	variables.  I'm not sure what changes were made to the other files.

	* (shell.c) incorporated Dave's changes into my shell.c copy:
	New greeting, init_path now called, .Macaulay is now name of
	init file.

Fri Feb  3 18:16:30 1989  Mike Stillman  (mike at mssun8)

	* (cmds.c gm_dlist.c shell.c) added commands "min" and "max".

	* (cmds.c human_io.c shell.c) added command "poly": 
	   poly f a2-b2
	sets f to be the polynomial a2-b2. (Actually, its a 1 by 1 matrix).

	* (cmds.c shell.c) added command "characteristic", and changed
	command "nvars" so both need to have a ring as parameter (or
	something with a base ring).

	* (most files) in monitor.c: added routines newline, fnewline,
	printnew, fprintnew.  These control output line display.  A new
	set variable "prcomment" was added.  If >0 the above commands
	affix a semicolon to the beginning of each line.

	Also changed: a routine prerror (print error) was added to
	monitor.c which displays an error message, even if prlevel > 0.
	Hopefully, I changed all of the places where this is needed.

	Current problems: still need to check dlDisplay, p_pprint.  Also
	help, and greeting cause problems.  prmat is not set to use
	prcomment, since it seems more usual that this output will go to
	other systems.  Also "putmat", "putstd" check to se whether the
	file is the stdout.  If not, prcomment is not used.  h_ files were
	not touched at all.  input.c was not touched.  stash.c was.

Thu Feb  2 14:29:47 1989  Mike Stillman  (mike at mssun8)

	* (set.c, rescmds.c) changed autocalc to two set variables:
	autocalc and autodegree.  doAutocalc was changed to use this.

	* (parse.c) changed fprec[u-], [u+] to 5, so precedence of ^ would
	be higher than unary -, +.

Tue Jan 24 15:53:28 1989  Mike Stillman  (mike at mssun8)

	* (vars.c) changed garb_collect so that it would only set
	current ring to NULL if it is in fact "DEAD", not just "REMOVED".
	This should allow rings to be used completely anonymously: a ring
	can always be set to be the current ring, if there is a matrix
	over this ring.  Once there is not, and the name has been reused,
	then the ring is finally removed.

	* (cmds.c) Fixed "ring", "pring", "setring", "ring-sum",
	"ring-from-cols", "ring-from-rows" so that:
	1. each command always sets the current ring.
	2. whereever a ring is expected as a parameter, any user variable
	   can be used instead: it just takes the base ring.
	Left to do: when a ring has been "removed" (by reusing its name),
	currently there is no way to use that ring.  I would like to
	change this.
	Also: a) need to be able to get charac. of ring.
	      b) ring-from-cols, rows don't read variable names correctly.
	      (Actually-works ok on SUN, NOT on MAC).
	      c) wherever a d-list is expected, one should be able to put
	      in a matrix, and it should take the row degrees.

Tue Dec 13 12:43:28 1988  Mike Stillman  (mike at mssun8)

	* (rmap.c) Fixed a bug David E. found in "ev": in change1_vars a
	zero monomial found did not change the curent ring back as it was
	supposed to.  Hard to notice, since it gets reset "almost"
	immediately, so only if the component location is different in the
	two rings would it be a problem (which it is!).  Easy to fix.

Mon Dec 12 17:13:55 1988  Mike Stillman  (mike at mssun8)

	* (gm_cmds.c) Modified "cat" command so that it uses correct
	variable parsing.  It first looks for a variable.  If it finds
	one, ok.  If not, it supposes that the "variable" is an integer,
	and uses that. (So 0 is the default value, if something goes
	wrong).  Note that giving "w-z" as the parameter to "cat" has the
	same effect as using "w".

	* (rvar.c ring.c iring.h) Modified ring input, so it now asks for the
	monomial order last.  Added a couple of fields to "ring" record,
	in the file "iring.h".

	* (cmds.c ring.c shell.c)
	Added 3 new commands for rings:
	ring-sum -- (affine) tensor product of 2 rings
	ring-from-cols -- create a new ring with #vars = #columns of a
	matrix.  The weight of each variable is the degree of the corresp.
	column.
	ring-from-rows -- corresponding thing for rows.
	

Fri Dec  9 11:56:41 1988  Mike Stillman  (mike at mssun8)

	* (monideal.c) changed codim, degree so they optionally set an int
	var. as well.

	* (vars.h cmds.c shell.c) Added commands which extract integer
	quantities: "nrows", "ncols", "nvars", "row-degree", and
	"col-degree". Any int quantitiy that I'm missing?

	* (term.c) Changed tm_inEq so it now checks exactly the first "i"
	blocks of the monomial for equality.  Before, it skipped the
	component, but that is not desired here.  Question: should the
	corresponding subring routine check for components or not?  I
	left it the way it was: skips over components.

	* (input.c) made the change mentioned in para. below: added a
	variable array prlev2 = "incr-set prlevel 1", and in sniff_prlevel
	this is checked as well.

	* (set.c shell.c) Added incr-set command to incr. value of a set
	variable.  Now "set" actually sets the value to what was
	specified.  One more change needs to be made: in input.c we need
	to check "incr-set prlevel 1" as well.

	* (human_io.c) Removed "?" from printing field elements if the
	number can't be lifted: it is pretty obvious in context, and
	people get really bent out of shape when they see "?" in their
	polynomials.  (Can't really blame them either).  The set var
	"char0" still is there though, since this still can be useful.

	* (term.c gm_poly.c) Fixed bug in jacob, contract, diff:
	In positive char. the power is not then normalized, causing 
	zeros to be represented.  Fix: normalized exponent in p_diff,
	and in p_contract.  Also, changed tm_contract so it does NOT do
	the normalization.  This is because one wants "contract" to work
	ok in positive char.

	* (parse.c, gm_poly.c) Fixed bug in parsing:
	Division by 0: in both cases (int and poly) a/0 returns 0, and
	a&0 returns a.  Currently, an error message is given in int case
	but not in the poly case.  Should this change?

	* (vars.c) 
	Fixed two bugs:
	1. type <ring> would set current ring to NULL.  Now it
	doesn't.
	2. much worse: during compaction in gc, the b_alias field of
	variables wasn't being corrected, causing all sorts of bad 
	behavior.

Wed Dec  7 19:16:51 1988  Mike Stillman  (mike at mssun8)

	* (parse stuff) 
	Also changed: gm_poly.c
	p_divmod, p_div, p_mod added (or changed).  Polynomial 
	parsing now allows division and reduction by single poly. 
	(/ for division, & for reduction).

	* (rescmds.c) 
	Fixed bug in autocalc and garbage collection.
	Problem was: commands like "res a a", if stopped in the 
	middle, changed the computation location (!).  Fixed by
	having doAutocalc use names, and not hard addresses.

Tue Dec  6 19:42:44 1988  Mike Stillman  (mike at mssun8)

	* (parsepoly.c parse.h) also changed: (gmatrix.c poly.c vars.c parse.c
	  makefile)
	polynomial parsing has been added.  Still needs to be debugged
	though.  Poly input will probably be slower, but much much
	more versatile.

	* (rvars.c)
	Completely rewritten to allow integer expr as subscripts, also
	the notion of indeterminate has been cleaned up considerably.

Mon Nov 28 18:02:27 1988  Mike Stillman  (mike at mssun8)

	* (many files, including: human_io.c gm_dlist.c)
	Changed atoi's to getInt, and changed dlgetelem to use int
	parsing.  Changed get_int, get_defint as well.  Only place
	not yet changed is in poly. input.

	* (cmds.c, vars.c, generic.c, makefile, PARSE.C)
	Added file parse.c, which allows integer arithmetic via op. prec.
	So far, added "int" command, which creates integers using this 
	parsing.

Fri Nov 18 15:46:56 1988  Mike Stillman  (mike at mssun8)

	* (rmap.c)
	Completely rewritten, now there are 6 ring map commands.  This
	change also includes: shell.c -- commands changed.
	types.h, vars.h, vars.c, generic.c, boxprocs.c, hulb.c
	All references to ring_map are hopefully expunged.

	* (human_io.c)
	Added rdPoly, rdPolyStr -- these read poly's using continuation.
	Changed pl_scan, pl_sparseScan to use these.

Thu Nov 17 14:02:50 1988  Mike Stillman  (mike at mssun8)

	* This is the first entry since change.log


