
CHANGES BETWEEN 6.12 and 6.11
James Dugal
	- added format.c to SRCS in Makefile
	- noted RETURN didn't enter insert mode
Peter King
	- pointed out iscntrl is broken on some other systems as well
	- sent some lint cleanups
Michael Richardson
	- patch to stop format looping when scientific notation was selected
Glenn T. Barry
	- code to turn on hardware scrolling and added 'slow speed display'
		speedups, default for SYSV3 or see -DSUNOS41SYSV in Makefile.
Tom Tkacik
	- fixes to make sure J and K move same amount, and re-added H code
Jeff Buhrt
	- fixed a possible xfree(NULL) in getent() (found when adding K_VAL)
	- merged compiler cleanups
*	- added $(name)qref to print a Quick Reference card
	- got rid of INVALIDS that may have been left around
*	- pressing return on a empty line puts you into insert mode
		(like in <=Sc6.1). When entering you can also press ESC
		to go into the editor (no change); this is also documented
		now so it might stay around this time.

CHANGES BETWEEN 6.11 and 6.10

Jonathan I. Kamens
	- sc.doc now mentions the tutorial file in the FILES section
Andy Fyfe
	- pointed out 3 locations where a NULL should have been '\0'
Robert Bond
	- pointed out the ERROR could hide a cellerror
Piercarlo Grandi
	- H,J,I,K now move 1/2 screen
Ulf Noren
	- changes for AIX V3.1
		- defined CHTYPE and NLS for the preprocessor. CHTYPE is
		the type of every character in a curses window.
		- Added KEY_BACKSPACE to nmgetch
		- strtof ifdef
	- Iteration change: when Sc says: "Still changing after 9 iterations"
	Sc at that point will have eval'd 9 times
Chris Metcalf
	- pointed out I broke setlist when adding 'goto {error,invalid}'
James P. Dugal
	- iscntrl() wasn't handling TABS though CRs under Pyramid OSx4.1
Peter King
	- BROKENCURSES patch for nl()/nonl() bug on some BSD systems
	- backups, tutorial file, and man references now depend on $name
	- DFLTPAGER to DFLT_PAGER fix

CHANGES BETWEEN 6.10 and 6.9

Tom Tkacik
	- when moving off the current table (resizing) now move the cursor
		on 'l' or 'k'.
	- patches to sc.doc to correctly format the vi-mode notes
Jim Clausing
	- made sure / doesn't try to divide by zero.
Tom Kloos
	- correction to substr() example in help.c
Piercarlo "Peter" Grandi
	- Disable non-constant expressions while loading
	- Added extra code in dealing w/ floating point exceptions
	- #ifdef'd SAVENAME (vs hardcoded SC.SAVE) to allowing changing the
		emergency save name.
Casey Leedom
	- Makefile changes: man extension, RINT note, make values should
		never be left undefined and then referenced, don't leave
		around *.old's
Tom Anderson
	- patches to add type of column format (note format now has 3 args)
Jeff Buhrt
	- xmalloc/xfree fatal() will now call diesave()
		(MAKE SURE the saved file is ok if this were to happen)
	- history[] is now a circular queue, this will cut down on the
		number of data moves and also xmalloc/xfree calls
		(idea from Keith Bostic)
	- cells with an error (ex: divide by 0) will show 'ERROR'
	- you can 'goto error' (or 'goto') to find an ERROR (for next ERROR)
Robert Bond
	- When in numeric mode the ^B, ^F, ^N, ^P key will end a numeric entry.

CHANGES BETWEEN 6.9 and 6.8

Jim Richardson
	- pointed out vi mode was not documented in sc.doc
	- found a nasty buffer limit bug in savedot()
	- a side effect was ^D could cause a core dump (-Jeff)
Tim Wilson
	- Hints on compiling on Ultrix
Eric Putz
	-patch for printfile() (sc died on huge # of columns in a W)
Jeffrey C Honig
	-patch for lex.c which bombed on SunOS 4.1 if $TERM was not set
Tom Kloos
	-psc now calls [+-.] strings vs numbers.
	-also pointed out a format reversal problem
Jack Goral
	-changes to Makefile to compile under SCO Unix V rel 3.2.0
Mark Nagel
	-changes to allow arbitrarily complex formatting of cells 
Kim Sanders
	-^W generated an incorrect equation (line was not started at beginning)
Mike Schwartz
	-a put command will use the same encryption key as when the
	file was read.
	>I have a suggestion for making the encyrption option of "sc" more
	>usable:  Right now, if you use the -x option when you start up sc, it
	>prompts you for the key (just like "vi -x" does).  But when you try to
	>write the file out using the Put command, it asks for the key again
	>each time.  Why not make it use the same key you used before (as "vi
	>-x" does)?  That would really help, because as it is, each time you try
	>to save the file you run the risk of mistyping the key.
	>
	>You might think this causes a security problem, since the key is then
	>an argument to crypt, and hence is visible from ps.  But when crypt
	>runs, the first thing it does is to copy the key to an internal buffer
	>and then zero out the argv copy, so the window of vulnerability is
	>vanishingly small.
Adri Verhoef
	- pointed out a ^D caused a core dump (fixed)
Gene H. Olson
	- format now grows the spreadsheet before setting the column format.
	- removed an extra ';' that caused a possible column number trashing
Paul Eggert
	-sc now also has round-to-even, also known as ``banker's rounding''.
	>With round-to-even, a number exactly halfway between two values is
	>rounded to whichever is even; e.g. rnd(0.5)=0, rnd(1.5)=2,
	>rnd(2.5)=2, rnd(3.5)=4.  This is the default rounding mode for
	>IEEE floating point, for good reason: it has better numeric
	>properties.  For example, if X+Y is an integer,
	>then X+Y = rnd(X)+rnd(Y) with round-to-even,
	>but not always with sc's rounding (which is
	>round-to-positive-infinity).  I ran into this problem when trying to
	>split interest in an account to two people fairly.
	-While we're on the subject, @round(X,Y) should also work when Y
	>is negative. For example, @round(123,-2) should yield 100.


CHANGES BETWEEN 6.8 and 6.7

Jeff Buhrt (with help from some beta testers-Thank you)
	  1) added a per row memory allocation
		-runs in about 1/2 run time and 1/3 the space of 6.6vm.1
		-insert/delete row now just moves pointers (# == maxrow+1-currow)
			and blanks one row (of columns (maxcol))
		-as the number of cells grows the size is more linear
			(no more ##Meg images except for 100,000's of rows....)
		-row to column pointer translation is done by a macro (ATBL)
			that returns a pointer to the cell pointer.
			*ATBL would be a pointer to a *ent (cell).
		-the maximum # of columns is limited by ABSMAXCOLS or
			sizeof(struct ent *)*maxcols (whichever is smaller)
			(702 * 4 = 2808 is no real limit even for 286 large model)
		-the maximum # of rows is limited by the virtual memory limit or
			sizeof(struct ent **)*maxrows (whichever is smaller)
			(4*X=64k, X=16384 rows (excluding malloc overhead) on
				a '286 large model. Even w/ 3.25Meg and 10Mhz)
			(plus of course any memory used for cells)
	2) dolookup (int vs double)
	3) dolookup calling eval w/ ent * not enode *
		 (dolookup called w/ ent * not enode *)
	4) cleaned up a lot of .... *x = 0 to  (.... *)0 (cmds, interp)
	5) psc: fwidth/precision were reversed on the output
	6) Backup copy (on save) using same mode to [path/]#file~
		 (will prompt if a backup fails)
	7) put y/n prompt function into yn_ask(mesg)
	8) found a move(x,y) in sc -> move(y,x) and only move when needed
	9) we use FullUpdate || changed (to see if ANY cells changed) 
		before trying to redraw the screen in update
		(now we don't try to redraw every time a key is hit)
		-if we are stand[ing]out we do not create a cell just to force a
		 standout inside the repaint section of update()
		-only draw blank cells if we cleared it or it is standing out
		reason: the less work (what to update) curses has to do, the faster
			a screen update will be (less cpu required)
	14) {insert, delete}col replaced w/ {open,close}col(currow, numcol_to_insert)
		(limits looping)
	6.7.1.1
	15) goto nonexistant cell may loop
	16) make sure that startup size will at least fill the screen w/ cells.
	17) added version.c
	6.7.1.2
	18) When we would normally die w/o saving (SIGQUIT, etc), we now ask
		if people would like to save the current spreadsheet.
		If 'y', saves to the current file name, otherwise ~/SC.SAVE,
		then /tmp/SC.SAVE if all else fails.
	6.7.1.3
	19) don't use malloc.c for production code
	20) progname is now truncated to just the basename (systems w/ long paths
		caused problems)

CHANGES BETWEEN 6.1 and 6.7

Dave Lewis - 
	Found and fixed a null pointer derefrece in the 'R' command.

Rob McMahon -
	Changed the ctl() macro to work with ANSI style compilers.
	Cleaned up some non-readonly text problems.

Rick Linck -
	Fixed a bug in lex.c - Ann Arbor Ambassadors have long ks and ke
	termcap entries.

Sam Drake -
	A fix for undefined C_* symbols in AIX.

Peter Brower -
	Cleaned up the INTERNATIONAL ifdefs with more portable code.

Glen Ditchfield
	Cleaned up a problem in crypt.c when the encrypted file shrank.

Bob Bond -
	Vi style editing for the command line.
	A bug in range name aliases.

Jeff Buhrt -
	-Added "~" filename expansion.
	-702 columns (A-ZZ) and unlimited rows/cells based on max. memory
	-fixed a few bugs
	-slightly decreased CPU usage
	-MAKES backup copies of files
	-understands ~$HOME stuff

CHANGES BETWEEN 5.1 and 6.1:

Andy Valencia -
	xmalloc aligns data to a double boundary.

Lawrence Cipriani -
	Fixed a bug in the "do you want to save this" sequence.

Soren Lundsgaard -
	A null pointer derefrence.

Rick Perry -
	Cleaned up a problem with modchk() in sc.c.

Gregory Bond -
	Added code for multi argument versions of @min and @max.

Tad Mannes -
	Added code to save/restore hidden rows and columns when the
	data base is saved or restored.

Marius Olafsson -
	INTERNATIONAL changes.  Allows full 8 bit characters (if
	curses supports them.)

Kurt Horton -
	Added support for @pv, @fv and @pmt financial functins.
	Tested lots of different systems, linting.

John Campbell -
	Support for VMS.  See VMS_NOTES.

Peter King -
	 User selection of row or column order for recalculation.
		Also affects order of traversing regions in /f and /r
	 User setting of automatic or manual recalculation.
	 User setting of number of times to try recalculation.
	 + and - commands when in non-numeric mode to do 
		increment and decrement operations.
	@index, @stindex, @atan2, @lookup  functions.
	Save/restore options.
	Support for TeX, LaTeX, and better support for tbl in "T" cmd.
	Provision of a copyent function to copy entries (same code repeated
		in several locations)
	Forwrow, backrow, forwcol, backcol functions to replace
		repeated code
	Correct interpretation of ESCAPE or ^G as an abort when in a 
		two character command such as 'ar' or 'ac'
	Cleanup in eval() - catches non-trap function errors.

Bob Bond - 
       Added search options to "g".
       Added supression of hidden columns to "W"
       Added the mod operator "%"
       New help functions.
       Constant prescale "$"
       Added string matching to @lookup.
       Some more bug fixes.
       Testing, integration, documentation.

Alan Silverstein-
	Greatly revised the manual entry.
	Added menus for ^E command and row/column commands, which
	involved a bunch of code cleanup.

	Changed top row display to clearly indicate string labels
	versus number parts, and to distinguish string functions from
	constant labels.

	When the character cursor is on a cell (not topline), ^H
	(backspace) is like ^B (move back one cell), rather than being
	ignored.

	When the character cursor is on a cell (not topline), ^I (tab)
	is like ^F (move forward one cell), rather than being ignored.
	^R is no longer identical with ^L.  Now ^R highlights all cells
	which should be entered by a user because they contain constant
	numeric values (not the result of a numeric expression).

	Added a ^X command, similar to ^R, which highlights cells which
	have expressions.  It also displays the expressions in the
	highlighted cells as left-justified strings, instead of the
	label and/or value of the cell.

	Added indirection functions (@nval() and @sval()) for simple
	table lookups.  Given a column name and row number, they return
	the numeric or string value of the selected cell.

	Added external functions (@ext()) for non-trivial
	computations.  Given a command name and argument, it calls the
	command and reads back one output line.

	Added a ^T,e command to toggle enabling of external functions.

	Changed ^T,t to only control the top line display, and added
	^T,c to control current cell highlighting.  (Separated the
	functions.)

	"!" (shell escape) gives a vi-style warning if there were any
	changes since the last write.  (No change to manual entry.)

	Fixed some startup, error, and prompt messages to be cleaner
	and/or more consistent.  (No changes to manual entry.)

	Fixed a bug:  If @substr() upper bound (third parameter) is
	past the end of the string operand, return the substring
	through the end of the string, rather than returning a null
	string.

	Fixed a bug:  Reset SIGINT to default after forking before
	calling shell escape program and before starting pipeline (for
	commands which support this).  Didn't reset SIGINT before
	calling crypt and external functions because in both cases it
	should be irrelevant.  (No change to manual entry.)

CHANGES BETWEEN 6.1 and 6.2:


Chris Cole-
	Compatibility with Lotus 1-2-3
		a) @hlookup(expr,range,expr)
		b) @vlookup(expr,range,expr)
		c) @round(expr,expr)
		d) @if(expr,expr,expr)
		e) @abs(expr)
