-------------------------------------------------------------------------------
-------------------------------------------------------------------------------

MET/CAL V7.20A

Mar 19 2008

File name "readme720A.txt"

-------------------------------------------------------------------------------
Section 1:  Table of Contents -------------------------------------------------
-------------------------------------------------------------------------------

	1. Table of Contents

	2. MET/CAL Versions

	3. Introduction

	4. Problem Reports

                Corrections made after V7.20 SP1.

	5. Enhancements

		Enhancements added after V7.20 SP1.

-------------------------------------------------------------------------------
Section 2:  MET/CAL Versions --------------------------------------------------
-------------------------------------------------------------------------------

	Interim versions of MET/CAL (Run Time and Editor) are designated by
	a single letter after the primary version number.

	Upper-case letters designate later builds than lower-case letters.

	The version sequence for V7.20 is:

		7.20  - 15-Nov-2006
		7.20a - 12-Dec-2006
		7.20b - 16-Feb-2007
		7.20c - 13-Mar-2007
		7.20d - 10-Apr-2007
		7.20e - 23-Apr-2007
		7.20f - 08-May-2007
		7.20g - 14-May-2007
		7.20h - 29-May-2007
		7.20i - 06-Jun-2007
		7.20j - 08-Jun-2007
		7.20k - 14-Jun-2007
		7.20l - skipped
		7.20m - 29-Jun-2007
		7.20n - 05-Jul-2007
		7.20o - skipped
		7.20p - 17-Jul-2007
		7.20q - 20-Jul-2007
		7.20r - 08-Aug-2007
		7.20s - 16-Aug-2007
		7.20t - 24-Sep-2007
		7.20u - 27-Sep-2007
		7.20v - 30-Oct-2007
		7.20w - 20-Nov-2007
		7.20x - 04-Dec-2008 (V7.20 SP1)
		7.20y - 11-Jan-2008
		7.20z - 06-Mar-2008
		7.20A - 19-Mar-2008

-------------------------------------------------------------------------------
Section 3:  Introduction ------------------------------------------------------
-------------------------------------------------------------------------------

	MET/CAL V7.20A is an engineering build that corrects a number of
	defects in V7.20 SP1.  In addition, this distribution implements
	a number of product enhancements.  Refer to sections 3 and 4 below
	for details.

	All enhancements and bug fixes released in V7.20 SP1 are included
	in this distribution.

	The reason this distribution is designated as an "engineering build"
	rather than a service pack is that, in general, the software changes
	implemented after the V7.20 SP1 release have undergone less testing
	than would be done before a formal factory release (such as V7.20)
	or a service pack release (such as V7.20 SP1).

	MET/CAL users are advised to make back up copies of the MET/CAL
	executables, DLLs, and accuracy files before installing the files
	included in this distribution.

	Refer to "readme720.txt" for a description of enhancements and
	corrections in V7.20.

	Refer to "readme720sp1.txt" for a description of enhancements and
	corrections in V7.20 SP1.

	You must have MET/CAL V7.20 (or later) installed before installing
	this distribution.

	This distribution includes:

		(1) Installation instructions:

			MC720A_Instructions.txt

		(2) A readme file that describes all of the changes since
		    V7.20 SP1:

			readme720A.txt

		    (This is the file you are currently reading.)

		(3) New MET/CAL Editor executable:

			mce32.exe

		    The version is 7.20A.

		(4) New MET/CAL Run Time executable:

			mcr32.exe

		    The version is 7.20A.

		(5) MET/TRACK executable:

			mettrack.exe

		    The build #444.  MET/TRACK has not been updated since
		    V7.20 SP1 was released.

		(6) Flexible standards initialization file.

			user_config_instr.ini

		    "user_config_instr.ini" has not been updated since the
		    V7.20 SP1 release.

		(7) Dynamic link libraries:

		    Dynamic link libraries are contained in the included
		    zip file "mcdll.zip".

				dblicenz.dll
				genrpt.dll
				metgendb.dll
				rpt32.dll
				text.dll

		    Some of the DLLs are unchanged from the V7.20 SP1 release.

		(8) Updated MET/CAL message file:

			mctext.mdb

		    New messages have been added.

		(9) On-line help files.

		    This distribution includes the most recent versions of
		    all MET/CAL on-line help files.

		    The files are contained in "mchelp.zip".

		(10) Guardbanding documentation:

			Guardbanding.pdf

		(11) Accuracy files:

			All accuracy files that are new or updated since
			the V7.20 release are included.

			The accuracy files are contained in "mcacc.zip".

-------------------------------------------------------------------------------
Section 4:  Problem Reports  --------------------------------------------------
-------------------------------------------------------------------------------

	---->  List Box Termination Fails With More Than Two User-Configured
	       Buttons

		Report Number:  	46148
		Resolved:       	Mar 18 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		Using the MATH FSC list box functions introduced in V7.20 SP1,
		when non-default buttons are configured, and there are more
		than two buttons, choosing the 2nd button from the left causes
		MET/CAL to operate as if the window has been user-terminated.

		The problem does not arise when the default buttons ("Advance"
		and "Terminate") are used.

		The problem also does not arise when the "+ok" configuration
		option to configure "Ok" and "Cancel" buttons is used.

		The distribution corrects the implementation so that it
		operates as follows:

			(1) If "+term" is configured, and there are two
			    or more buttons, the list box window is
			    operator-terminated when the right-most button,
			    other than "Search", is chosen.

			(2) If "-term" is configured, or there's
			    only one button, the list box is never
			    operator-terminated.

	----> Fluke 8845A and 8846A Cannot Be Configured with 5500/CAL

		Report Number:  	46110
		Resolved:       	Mar 14 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		In V7.20 SP1, and earlier versions that support the Fluke 8845A
		and Fluke 8846A, 5500/CAL does not allow the 8845A or 8846A to
		be configured.  This prevents the 8845 and 8846 FSCs from being
		used with 5500/CAL.

		The problem does not affect MET/CAL.

		The problem is resolved in this distribution.

	----> 6100 Reference Accuracy Calculation Does Not Handle "NA" Correctly

		Report Number:  	46004
		Resolved:       	Mar 06 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		In MET/CAL V7.20 SP1, and prior versions of MET/CAL that
		support the "6100" FSC, the calculation of the reference
		accuracy is not handled correctly in cases where the required
		specifications (floor and/or gain) are set to "NA" in the
		6100A accuracy file.

		The underlying problem is that "NA" is not handled as a
		special case.  That causes the calculated reference accuracy
		to be zero.  The error was essentially hidden prior to the
		release of V7.20 SP1, because both the test uncertainty ratio
		and measurement uncertainty calculations effectively ignore
		a reference accuracy of zero.

		However, with the addition of the MATH FSC "RefAcc" function
		in V7.20 SP1, the case where the reference accuracy is zero
		is specifically handled as an error condition.	That change
		caused the problem in the 6100 FSC to be noticed.

		In this distribution the 6100 FSC has been updated to properly
		handle the case where required 6100A specifications are set to
		"NA" in the MET/CAL accuracy file for the Fluke 6100A.

	----> 9100 Test with CO or HV in MOD3 and Nominal Units = ohms
	      Incorrectly Viewed As Measurement

		Report Number:  	45970
		Resolved:       	Mar 04 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		In order to distinguish the variable quantity in a test
		(VARQ) from the fixed quantity (FIXQ) MET/CAL uses an internal
		function that analyzes an instrument statement to determine
		if the reference instrument is being used as measurement
		device or a source device.

		For many reference instruments the decision is simple,
		because the instrument can only be used in one mode (i.e.,
		it's always a measurement device or always a source device).

		For certain instruments, however, some tests involve using
		the instrument in a measurement mode and other tests involve
		using the instrument in a source mode.

		For the Fluke 9100, the algorithm has been this:

			If the MOD3 field specifies CO (continuity) or HV
			(insulation), the instrument is being used as a
			measurement device.

			For all other 9100 statements the instrument is
			being used as a source device.

		The problem with this algorithm is that it doesn't correctly
		handle all cases where the MOD3 field is "CO".

		In particular, when MOD3 is "CO", and the NOMINAL units
		are "Z" (ohms), the 9100 is being used as a source device,
		not a measurement device.

		For example, in a test like this:

			1.001  ASK- R N P F W
			1.002  ASK+ K
			1.003  VSET RSM = Last
			1.004  9100 10.0Z 1.0U CO 4W

		MET/CAL incorrectly concludes that the 9100 is being used as
		a measurement device.  This causes the VARQ and FIXQ result
		quantities to be incorrectly swapped.

		The same problem is also present in insulation mode
		(MOD3 = "HV").	In insulation mode, when NOMINAL units are
		ohms ("Z") the 9100 is being used as a source device, not
		a measurement device.

		This distribution resolves the problem.

	----> 8508 FSC Incorrectly Allows MOD3 = D4 for DC Voltage

		Report Number:  	MC-3465
		Resolved:       	Mar 04 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		The 8508 FSC help file documents the MOD3 field as follows:

			The MOD3 field Specifies the measurement resolution
			or input zero.

				o    D4       4.5 digit resolution
				o    D5       5.5 digit resolution
				o    D6       6.5 digit resolution
				o    D7       7.5 digit resolution
				o    D8       8.5 digit resolution
				o    ZR       Perform input zero

			Rules:
			Function           D4  D5  D6  D7  D8
			----------------------------------------------
			DC Voltage                x     x      x     x
			AC Voltage                 x     x
			DC Current                x     x      x
			AC Current                 x     x
			Normal Ohms           x     x      x     x
			True Ohms                x     x      x     x
			HV Ohms                   x     x      x     x
			Temperature             x     x      x     x
			Frequency          x             x

		However, the following statement compiles without
		error:

			8508  190.00mV  D4  N  2W

		This distribution resolves the problem.  It is no longer
		possible to compile the unsupported statement shown above.

	----> MATH Functions "Strip", "StripL", and "StripT" Corrupt Memory
	      When Argument is Empty or NULL

		Report Number:  	45917
		Resolved:       	Mar 04 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		The MATH FSC functions "Strip", "StripL", and "StripT",
		which were included in V7.20 SP1, cause an internal memory
		problem when the argument passed in is a NULL or empty string.

		The procedure-level work-around is to test to make sure the
		length of the argument is greater than one before calling
		"Strip", "StripL", or "StripT".

		This distribution resolves the problem.

	----> Error in 100 Ohm 385 RTD Specs in Fluke 8846A Accuracy File

		Report Number:  	MC-3463
		Resolved:       	Feb 22 2008
		Application:    	MET/CAL Accuracy File
		Required Version:	7.20z

		In MET/CAL V7.20 SP1, and prior versions that support the
		Fluke 8846A, there are errors in the gain specification in
		the accuracy file sections that apply to the 100 ohm 385 RTD.

		The error affects the following sections of the accuracy file:

			(1) RTD Pt 385 100Ohm NPLC 100
			(2) RTD Pt 385 100Ohm NPLC 10
			(3) RTD Pt 385 100Ohm NPLC 1
			(4) RTD Pt 385 100Ohm NPLC 0.2
			(5) RTD Pt 385 100Ohm NPLC 0.02

		The gain specification is incorrectly stated as 1.000 on lines
		3, 4, 5, and 6 in each section listed above.  The correct gain
		specification value is 0.000.

		This distribution includes an updated 8846A accuracy file
		("8846.acc") that corrects this problem.

	----> MATH FSC Does Not Trap Illegal Floating-Point Values In
	      Some Cases

		Report Number:  	45603
		Resolved:       	Jan 25 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In V7.20 SP1 and prior versions the MATH FSC does not
		properly handle illegal double-precision floating-point
		values in some cases.

		The following case is handled in an acceptable manner:

			MATH MEM = 1e307 * 1e3

		The product of 1e307 and 1e3 is a number larger than the
		maximum double-precision floating-point value (per IEEE 754).
		When MET/CAL executes the statement above a floating-point
		exception (FPE) occurs, terminating the MATH statement.

		However, two cases that are not handled properly are:

		(1) MATH MEM = 1e400

			This statement directly assigns a value greater
			than the maximum double-precision floating-point
			value to a numeric register.  This assignment
			should generate an error message but, in fact,
			the bit pattern representing "infinity" (printed as
			"1.#INF") is assigned to MEM.

			Allowing this bit pattern to be assigned may have
			unacceptable consequences when the value is used in
			subsequent expressions.  For example, a relational
			expression like:

				IF (MEM == 1)

			always succeeds when MEM is set to 1.#INF.  (This is
			a side-effect of MET/CAL's built-in epsilon-based
			comparison designed to protect users from the
			side-effects of floating-point rounding.)

			In this case MET/CAL should be able to catch the
			error at compile time.

		(2) MEMI Enter the value 1e400
		    MATH MEM1 = MEM

			This is a similar problem, but the 1.#INF value is
			entered by a different means.  I.e., rather than direct
			assignment of a literal value, the illegal value is
			entered indirectly.  In this example MEMI is used,
			but it could as well be IEEE, or any other procedure
			statement that assigns a value to a numeric register
			that is then directly accessed by the MATH FSC parser.

			In this example, the error should be detected at
			run time.

		The problems described above are resolved in this distribution.

		The MATH parser has been modified to check for illegal literal
		floating-point numeric values in procedure statements, and
		also prevent pushing an illegal floating-point numeric value
		onto the run time stack.

	----> MATH Function "lbText" Fails to Force Inclusion of Static Text
	      in List Box Window

		Report Number:  	45553
		Resolved:       	Jan 22 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		MET/CAL V7.20 SP1 included a family of new MATH FSC functions
		used to configure and display list boxes.

		A list box window optionally includes a scrollable static
		text area above the list box.  The optional static text area
		may be enabled using the "lbConf" function with the argument
		"+text".  The default is "-text" (i.e., not enabled).

		However, per the on-line help file for the MATH FSC:

			"+text" configures the list box window to include
			a scrollable, read-only text area above the list box.

			The default is "-text".  However, adding text lines to
			the list box window automatically configures "+text".

		I.e, making a call to "lbText" is supposed to override
		the default and make it unnecessary to also include an
		'lbConf("+text")' configuration call.

		In V7.20 SP1 this does not work as intended.  A call to
		"lbText" does not, by itself, enable the read-only text
		area above the list box.  It is necessary to include an
		'lbConf("+text")' call to display the text lines.

		This distribution resolves the problem.

	----> MATH Function Call "Uncv(0)" Fails When NMEAS = 1

		Report Number:  	45551
		Resolved:       	Jan 22 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		The MATH FSC function "UNCV" is designed to retrieve individual
		measurements in a sequence of one or more measurements taken
		when measurement uncertainty is enabled.

		The argument passed to "UNCV" specifies the index of the
		measurement.  The index is 1-based, so "UNCV(1)" gets the
		1st measurement, "UNCV(2)" gets the 2nd measurement, and so on.

		As a special case, "UNCV(0)" returns the average of the
		measurements.

		However, when the number of measurements is 1 (NMEAS = 1),
		"UNCV(0)" always returns zero.	It should return the average
		of the measurements.  I.e., since there's only one measurement,
		it should return the same value that "UNCV(1)" returns.

		This problem is present in V7.20 SP1 and earlier versions
		that support the "UNCV" function.

		This distribution resolves the problem.

	----> MATH FSC List Box Function "lbShow" Fails When Window Has
	      Only Buttons

		Report Number:  	45549
		Resolved:       	Jan 22 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		MET/CAL V7.20 SP1 included a family of new MATH FSC functions
		used to configure and display list boxes:

			lbAdd	   - add item to list box
			lbAddList  - add list of items to list box
			lbAddV	   - add item to list box, with assoc. variable
			lbButton   - add button to list box window
			lbClose    - close list box window
			lbConf	   - configure list box window
			lbGetSel   - get name of n-th selected list box item
			lbGetSelX  - get index of n-th selected list box item
			lbNew	   - create new list box window
			lbResp	   - get user's text response
			lbSel	   - initially select list box item (by name)
			lbSelCount - get count of selected list box items
			lbSelX	   - initially select list box item (by index)
			lbShow	   - display list box window
			lbText	   - add text line to list box window
			lbUnSel    - remove initial selection of item (by name)
			lbUnSelX   - remove initial selection of item (by index)

		In  V7.20 SP1 "lbShow" fails when an attempt is made to
		configure a list box that contains only a title and buttons
		(i.e., does not contain read-only static text, does not contain
		a list box, and does not contain a single line edit control).

		The problem is resolved in this distribution.

	----> Error in MATH FSC Functions "Insert", "Replace", "Write", and
	      "Delete"

		Report Number:  	45548
		Resolved:       	Jan 22 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		MET/CAL V7.20 SP1 includes a family of new line-oriented
		file functions:

			Append	  - append line to file
			Delete	  - delete line from file
			EOF	  - determine if file is at end of file
			Insert	  - insert line in file before specified line
			LineCount - count lines of specified file
			Read	  - read specified line from file
			Replace   - replace line in file
			Write	  - write line to file after specified line

		Four of the functions in this group (Append, Delete, Insert,
		and Replace), depend on the creation and subsequent renaming
		of a temporary file.

		These four functions do not work correctly in V7.20 SP1
		because the temporary file is removed before it can be renamed.

		Any attempt to use one of these four functions results in
		an error message indicating that the temporary file does
		not exist.  This aborts the intended change, and leaves the
		original file as it was prior to the function call.

		The problem is resolved in this distribution.

	----> 4950 Test Function Fails in Demo Mode

		Report Number:  	45526
		Resolved:       	Jan 17 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In MET/CAL V7.20 and V7.20 SP1 the 4950 FSC test function
		fails in demo mode.

		This error is present in V7.20 and V7.20 SP1, but is not
		present in earlier versions.

		The underlying problem is that the test function parses
		and checks the response to the STANDBY? query, but fails to
		pre-load the buffer in demo mode with a valid response.

		This problem does not affect the 4950A in real mode, and
		does not affect the 1271A, 1281A, or 8508A in demo mode or
		real mode.

		A valid 4950A STANDBY? response is:
	
			"00:00"

		(The STANDBY? query is used to determine the time remaining
		until the 4950A has stabilized and is ready for use.)

		The problem is resolved in this distribution.

	----> Second HEAD Statement Not Cleared When Procedure User-Terminated

		Report Number:  	45521
		Resolved:       	Jan 17 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		V7.20 SP1 includes support for a 2nd HEAD line on the Run
		Time and Test Run main windows.  The syntax used to specify
		a 2nd HEAD line is:

			HEAD -2 <text>

		In V7.20 SP1, when the user terminates the procedure abnormally
		by choosing "Terminate" in a prompt window (for example a
		DISP statement), then choosing "Terminate" again in the Post
		Test window, MET/CAL fails to clear the 2nd HEAD line.

		The problem is resolved in this distribution.

-------------------------------------------------------------------------------
Section 5:  Enhancements ------------------------------------------------------
-------------------------------------------------------------------------------

	----> Function to Close a File

		Report Number:  	46164
		Resolved:       	Mar 19 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		MET/CAL V7.20 SP1 included a group of line-oriented file
		I/O functions.	Files are opened automatically when certain
		functions are called to read or write lines to files.
		For example, "append", "insert", "replace", "read", and
		"write" all trigger the opening of a file that's not yet open.

		Because these functions use buffered file I/O, file contents
		may not be written to disk until the file is closed.  In the
		V7.20 SP1 implementation, all open files are automatically
		closed when the procedure terminates.

		This approach is sufficient in most cases, but in order to
		provide a means for the procedure writer to force the closing
		of a file before procedure termination this distribution
		includes a new MATH FSC function named "close".

		By calling "close" the procedure writer may allow other
		processes, external to MET/CAL, to access a file before the
		procedure that created or updated the file has terminated.

		Refer to on-line help for the MATH FSC for details.

	----> Function to Check for Required MET/CAL Version

		Report Number:  	46126
		Resolved:       	Mar 19 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		In this distribution there's a new MATH function that allows
		a procedure writer to check the currently running version of
		MET/CAL (Run Time or Editor Test Run) against a specified
		version to make sure that the running version is the same
		as, or more recent than, the specified version.

		The name of the function is "VersionCheck".

		Prior to the implementation of "VersionCheck" this operation
		was done by calling a subprocedure.  That approach had
		several disadvantages:

		(1) It takes more lines of procedure code to set up and make
		    the subprocedure call.

		(2) It requires that the version check subprocedure be
		    included in all procedure packages.

		(3) It encodes the MET/CAL version numbering scheme into
		    a subprocedure external to the MET/CAL application.
		    This is the most significant disadvantage because it
		    constrains MET/CAL's ability to revise the version
		    numbering scheme.

		The new function is "VersionCheck".  If the specified version
		is later than the executing version, an error message is
		generated, and the system returns to the
		"Procedure Step Terminated" PostTest window.

	----> File-Based Accuracy Confidence Lookup Function

		Report Number:  	46112
		Resolved:       	Mar 14 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		File-based accuracy lookup functions were recently added
		to MET/CAL.

		The new file-based functions are:

			ACCF   - 1-param accuracy spec lookup
			ACCF2  - 2-param accuracy spec lookup
			ACCFF  - 1-param accuracy floor spec lookup
			ACCFF2 - 2-param accuracy floor spec lookup
			ACCFG  - 1-param accuracy gain spec lookup
			ACCFG2 - 2-param accuracy gain spec lookup

		Refer to report #45731 for a description of the functions
		listed above.

		There is now an additional file-based function that returns the
		confidence value from the header of a specified accuracy file.

		The name of the new function is "CONFF".

		Refer to on-line help for the MATH FSC for details.

	----> Capability to Change List Box Item Names

		Report Number:  	46090
		Resolved:       	Mar 12 2008
		Application:    	MET/CAL
		Required Version:	7.20A

		MET/CAL V7.20 SP1 includes a family of new MATH functions
		that allow a procedure writer to configure and display a
		window containing one or more of the following:

			(a) multi-line readonly text control
			(b) list box
			(c) single-line edit control
			(d) button row

		In the V7.20 SP1 implementation list box windows can be
		modified and subsequently re-used, but supported list box
		modifications are limited to adding new list box items.

		Therefore, in V7.20 SP1 the only way to change or remove an
		existing list box item is to close the existing list box window
		and then create a new list box window with the changed item (or
		without the item at all if the intent is to remove the item).

		To make it easier to modify or remove list box item, there
		are now two new functions, "lbRepl" and "lbReplV".

		"lbRepl" changes a specified list box item.  If the replacement
		string is an empty string, the specified item is removed
		from the list box.

		"lbReplV" is similar to "lbRepl", with the difference
		that the item to be changed is specified not by the item
		name, but rather by the name of an associated variable.
		The association must initially have been made by calling
		"lbAddV".  As with "lbRepl", if the replacement string is
		empty, the list box item is removed.

		Refer to on-line help for the MATH FSC for details.

	----> MATH Functions to Convert Formatted Dates to Elapsed Time
	      Since Jan 01 1970

		Report Number:  	45963
		Resolved:       	Mar 04 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		In the MET/CAL procedure language it is difficult to compare
		formatted dates to determine the proper date order.

		This distribution includes two new MATH functions designed
		to make it easier to compare dates:

			ugdate - convert formatted date (UTC) to elapsed
			         seconds since Jan 01 1970 (UTC)

			uldate - convert formatted date (local) to elapsed
			         seconds since Jan 01 1970 (UTC)

		Refer to the MATH FSC help file for details.

	----> File-Based Accuracy Lookup Functions

		Report Number:  	45731
		Resolved:       	Feb 08 2008
		Application:    	MET/CAL
		Required Version:	7.20z

		In MET/CAL V7.20 SP1, and earlier versions, the ACCF FSC
		cannot be used to lookup values in accuracy files associated
		with user-configured instruments.  This is because ACCF is
		tied directly to MET/CAL's built-in instrument FSCs.

		The MATH functions "ACCV", "ACCV2", "ACCVF", "ACCVF2", "ACCVG",
		and "ACCVG2" exist for looking up accuracy values in accuracy
		files for configured instruments, including user-configured,
		but the lookup algorithm automatically associates an accuracy
		file with a configured device by matching the file to the
		device name and the cal interval for that device in the
		inventory record in the database.  This works well in most
		cases, however, there are cases where a special accuracy
		file with a different cal interval is needed.

		To provide this capability this distribution includes a
		group of six new MATH FSC functions that operate in a manner
		similar to the "ACCV", "ACCV2", "ACCVF", "ACCVF2", "ACCVG",
		and "ACCVG2" functions, but take the accuracy file name,
		rather than the name of a configured instrument, as the
		first argument.

		The new functions are:

			ACCF   - 1-param accuracy spec lookup
			ACCF2  - 2-param accuracy spec lookup
			ACCFF  - 1-param accuracy floor spec lookup
			ACCFF2 - 2-param accuracy floor spec lookup
			ACCFG  - 1-param accuracy gain spec lookup
			ACCFG2 - 2-param accuracy gain spec lookup

		There's also a new, optional initialization file parameter,
		"AccfDir", that can be used to specify the directory where the
		ACCF-functions look for accuracy files.  (If the procedure
		specifies the full path name of a file, that specification
		takes precedence.)

		If the procedure does not specify the full path name, and
		"AccfDir" is not specified in the initialization file,
		MET/CAL uses the normal MET/CAL accuracy directory (as
		specified by the "accdir" initialization file parameter).

		Refer to the MATH FSC help file for details.

	----> MATH FSC Functions to Access Fluke 792 Data

		Report Number:  	45705
		Resolved:       	Feb 06 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In this distribution ease of use of the Fluke 792A Thermal
		Transfer Standard by MET/CAL procedure writers has been
		improved by the addition of a group of MATH FSC functions
		design to retrieve information from Fluke 792A AC-DC difference
		data files.

		The supported file format is the format used by the Fluke
		Standards Lab.

		The following functions have been implemented:

			Open792    - open, read, and check specified Fluke
			             792 data file

			Get792	   - get interpolated AC-DC difference value
			             from Fluke 792 data file

			CalDate792 - get calibration date from Fluke 792
			             data file

			SN792	   - get serial number from Fluke 792 data file

		Refer to the MATH FSC help file for details.

	----> Modify MATH Function "RefAcc" to Suppress Error Message When
	      Accuracy is Zero

		Report Number:  	4571
		Resolved:       	Feb 06 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		MET/CAL V7.20 SP1 introduced a new MATH function named
		"RefAcc".

		The purpose of "RefAcc" is to retrieve the reference accuracy
		determined in a preceding setup statement in the current test.

		Prior to the introduction of "RefAcc" MET/CAL was forgiving
		about the situation where the reference accuracy is zero.

		The TSR calculation ignores a reference accuracy of zero.
		Since the reference accuracy is the denominator of the
		TSR, the rationale is that the TSR must be greater than 4,
		because an "infinitely accurate" reference has to be much
		more accurate than the UUT.

		The measurement uncertainty calculation doesn't ignore
		the reference accuracy, but it's inconsequential in the
		sense that it simply contributes nothing to the overall
		measurement uncertainty calculation.  That is,
		RSS(0, v1, v2, ...) is the same as RSS(v1, v2, ...).

		However, "RefAcc" regards a zero reference accuracy as an
		error condition.  This error condition is detected regardless
		of whether the procedure includes calls to the MATH "RefAcc"
		function.

		This can be problematic when looking up the accuracy in a
		MET/CAL accuracy file where the reference instrument is a
		sensor (meter), the NOMINAL value is zero, and the instrument's
		specification does not include a floor component.

		An example of this problem occurs when using an Agilent 4418B
		power meter.  A 4418 statement with "0pct" in the NOMINAL
		field causes an error message to be generated when MET/CAL
		attempts to lookup the accuracy of the 4418 at 0%.

		In this distribution there's a new initialization file
		parameter called "RefAccAllowZero" that can be set to "yes"
		to cause MET/CAL to suppress the error message generated
		when the reference accuracy is zero.

		"RefAccAllowZero" defaults to "no".

	----> Inter-Address Delay for IEEE-488 Port 0 "Find Listeners" Search

		Report Number:  	45628
		Resolved:       	Jan 30 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In MET/CAL V7.20 SP1, and earlier versions, MET/CAL searches
		the IEEE-488 bus (port 0) before initial procedure execution
		to determine the addresses of instruments on the bus, and
		to cross-check against the instrument configuration for
		the workstation.

		For some devices, the IEEE-488 address search does not work
		correctly, because the instrument is too slow to correctly
		handle the IEEE-488.2 "Find Listeners" protocol.

		In particular, the Data Proof 160B Scanner answers at multiple
		IEEE-488 addresses if the "Find Listeners" check is allowed
		to run too quickly.

		In MET/CAL this distribution there's a new MET/CAL
		initialization file parameter named "FindListenDelay".

		Legal values are 0 to 10000 milliseconds.

		The parameter is optional.

		The default value is zero.

		Example:

        		FindListenDelay = 150

	----> 8901 FSC Help File Does Not State That an HP 8901A Must Be
	      Configured As An 8901B

		Report Number:  	MC-3454
		Resolved:       	Jan 28 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In MET/CAL V7.20 SP1, and earlier versions, the 8901 FSC
		help file states that the FSC supports the 8901A, but fails
		to state that the instrument must be configured as an 8901B.

		This distribution includes an updated 8901 FSC help file
		that clarifies this point.

	----> Add Ability to Specify Button with Input Focus in List Box Window

		Report Number:  	45554
		Resolved:       	Jan 22 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		MET/CAL V7.20 SP1 includes a family of MATH FSC functions
		used to configure and display list box window.

		Most components of a list box window are optional.  It is
		therefore possible to construct a list box window that does
		not include an actual list box, and also lacks a line where
		the user can enter text.

		A list box window configured that way is limited to an
		optional title, optional scrollable read-only text, and a
		group of one to eight buttons.

		For a window of that type it would be useful to be able to
		configure which button becomes the default button, receiving
		the input focus when the window is first displayed.

		In this distribution there's a new MATH function named
		"lbButtonFocus" that can be used to control the default
		button in a list box window.

		"lbButtonFocus" works only in a list box window that contain
		neither a list box nor a text-entry line.  In other words,
		it works in list boxes limited to one or more of the following:

			(a) a window title
			(b) a read-only text line
			(c) buttons

		Refer to on-line help for the MATH FSC for details.

	----> 39A and 195 FSCs Now Control Fluke 280 Series Arbitrary
	      Waveform Generators

		Report Number:  	MC-3443
		Resolved:       	Jan 11 2008
		Application:    	MET/CAL
		Required Version:	7.20y

		In MET/CAL V7.20 SP1, and earlier versions, the 39A and 195
		FSCs cannot be used to control Fluke 280 Series arbs.

		The following table shows the compatibility relationship:

			Fluke 281 <--> Wavetek 39A
			Fluke 282 <--> Wavetek 195 without Option 001
			Fluke 284 <--> Wavetek 195 with Option 001

		In this distribution the 39A FSC may be used to control
		channel 1 of the following:

			Fluke 281 1 Channel 40 MS/s Arbitrary Waveform Generator
			Fluke 282 2 Channel 40 MS/s Arbitrary Waveform Generator
			Fluke 284 4 Channel 40 MS/s Arbitrary Waveform Generator

		In this distribution the 195 FSC may be used to control
		the following:

			Fluke 282 2 Channel 40 MS/s Arbitrary Waveform Generator
			Fluke 284 4 Channel 40 MS/s Arbitrary Waveform Generator

		Since the Fluke 280 Series instruments are controlled using
		FSCs developed for the older models, it is necessary to
		configure the older model (using the MET/CAL Config Editor).
		In the database Inventory Table, however, the correct name
		may be used for reporting purposes.

		More specifically, the Fluke 281 must be configured as a
		Wavetek 39A, and the Fluke 282 and 284 must be configured
		as a Wavetek 195.

		A run time error is generated if a procedure incorrectly
		attempts to control channel 3 or 4 of a Fluke 282 (because
		the 282 has only 2 channels).

-------------------------------------------------------------------------------
End "readme720A.txt"
-------------------------------------------------------------------------------
