7526 Code History

7526 MICROCODE PROBLEMS/FIXES

This file describes the bug fixes incorporated into each
successive release of the 7526 Data Collection Terminal's
ROM microcode.

Code levels marked as BETA are not completely tested
by IBM, and should be used only if the work-around is
not sufficient and the situation is critical.

The latest entry (highest level number) includes all prior fixes.

------------------------------------------------------------------
CHANGE HISTORY
03/25/98 Added B4.34, B3.31
12/17/97 Added B4.30 fix and enhancement, B3.25 enhancement
10/22/97 Added B4.29, B3.21, B3.23 fixes
07/14/97 Added B4.27 fix
08/29/97 Added B3.19 , B4.29 fixes
06/10/97 Added B3.17 fix
05/30/97 Added B4.25 / B3.15 enhancments
05/29/97 Added B4.24 fixes and enhancement
05/27/97 Added B4.23, B3.13 fixes
05/12/97 Added B4.22, B3.11 fixes
04/03/97 Added B4.21 fixes
03/17/97 Added V4.20 fixes
11/21/96 Added B4.12 fixes
11/12/96 Added B4.10 fixes
10/07/96 Added B3.06, B4.05 & B4.06 fixes/enhancements
09/10/96 Added B4.04 fixes
08/19/96 Added B3.04 fix
08/13/96 Added B3.03 and V4.03 fixes/enhancements
07/30/96 Added B3.02 and V4.02 fixes
05/14/96 Added V3.00 changes
02/12/96 Added V2.50 fixes
12/14/95 Added V2.40 fixes
10/09/95 Added B2.34 fixes
09/25/95 Added B2.32 fixes
08/11/95 Added B2.30 fixes
03/28/95 Added description of B2.29 code fixes
03/15/95 Added description of B2.28 code fixes
06/27/94 Added description of B2.26 code fixes and enhancements.
05/23/94 Added description of B2.25 code fixes.
04/14/94 Added description of B2.24 code change.
03/07/94 Added description of B2.23 code change.
02/16/94 Added description of B2.22 code fixes.
12/21/93 Added description of B2.21 code fix and added problem
descriptions of some known bugs.
10/07/93 Added description of V2.10 code fixes and enhancements.
09/24/93 Added description of B2.13 code fixes.
09/07/93 Added description of B2.12 code fixes.
08/06/93 Added description of B2.11 code fixes.
06/16/93 Added description of 2.00 code fixes and enhancements.
03/24/93 Added description of 1.03g BETA code fixes.
03/12/93 Added description of 1.03f BETA code fixes.
01/07/93 See change bars for additions or modifications.
01/05/93 Added description of 1.03e BETA code fixes.
12/22/92 Added description of 1.03c BETA code fixes, moving many
of the unfixed bugs to this listing.
12/21/92 Added more known, unfixed bugs to B) section.
-------------------------------------------------------------------

CODE VERSIONS

The bugs described were fixed in the subject code level.

FLASH FOR 7526 ETHERNET CARD
----------------------------

Version B4.34 03/25/98 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------

1 Ethernet driver trace routine could write outside its memory area,
resulting in hung terminal and other bad symptoms. This happens more
with very busy Ethernet networks.

2 Terminal may hang as it is being put off-line by the controller. This
happens more frequently when the DCT is executing a CFR (doing IdleManager
calls) when the off-line command A0 is received.

3 Back-to-back SEND then remote verification transactions could confuse
the terminal, causing it to ignore the response for the remote validation
and/or send some data twice.

New feature
-----------

1 DCT no longer reprograms the bar code decode coprocessor when changing from
one fixed length Interleaved 2/5 size on the idle screen to a different fixed
length read in a transaction program. This can save 1-2 seconds of delay before
the DCT is ready for use by the worker.

Version B4.30 12/17/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------

1 Compacted numeric files were not loading properly -- including that used
to store fast-clocking transition times (terminal would never go into
fast-clocking mode).

New feature
-----------

1 Menu system now includes a 'Lock Adjustments' option which,
when set to 'YES', prevents the terminal user from adjusting
the beeper level or the LCD contrast.

Version B4.29 10/21/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------

1 Transaction was being sent at end of txn program with AUTO set to OFF when
worker would press Enter more than once on last input.

Version B4.28 8/29/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------

1 Fix for E03 error when paused in a ONKEY goto/gosub command AND a
digital-input or RS-232-input initiated event happens.

2 Fix for using the F28 (space) key in an ONKEY command

3 Reduced or eliminating RS-232 inbound overruns when used in busy Ethernet
network.

Version B4.27 7/14/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Fixes:
------

1 7526 may have trouble properly handling a time-out of its ARP table
entry for the server.

Version B4.25 5/30/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Enhancments:
-----------_

1 Added TimerGetTicks and TimerResetTicks CFR APIs

Version B4.24 5/29/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------

1 Serial port reads could get garbled data

2 Command 01 did not result in file download request

Enhancments:
-----------_

1 Added GetDCTInfo CFR API

Version B4.23 5/27/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Fix for various bad symptoms when a power voltage to the terminal fell or rose slowly
through the Non-maskable Interrupt power failure voltage level.
(as when the Operational Battery Backup feature is installed and is allowed
to run all the way out of power).
Symptoms could be: Loss of files, corrupted data,
corrupted idle prompt, terminal lock-up upon re-start of the 7526.

2 Fix for terminal response to host commands 7, 9A, and 9F -- they were coming back
with truncated data, or too much data.

3 Serial port reads could get garbled data

4 Command 01 did not result in file download request

Version B4.22 5/12/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Printing was prone to hanging up the machine.

2 :J FORMAT transaction programming command did not work
for variable length output -- target UV was always
empty when finished.

3 Additional fixes for time-of-day driven alarms -- there was a
possibility of an alarm set-point being skipped during the day.

4 Printer spooling timeout was increased to accomodate printers
which suspend accepting new data while form-feeding.

Version B4.21 4/03/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Print from CFR and transaction program was broken if you
didn't happen to have called PrtInit() from a CFR.
Workaround is to do this PrtInit() call in the Function 0/2
CFR code.

Version V4.20 3/17/97 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Various bugs which could cause terminal lack
of communications response in a heavily-loaded Ethernet
environment. Enabled hardware fixes for this fix.
NOTE: THIS LEVEL REQUIRES UPDATED 7526 PROCESSOR CARD
HARDWARE. LOOK FOR A YELLOW WIRE BETWEEN U31 AND U7
AS AN INDICATION, OR TRY THIS FLASH -- IF TERMINAL LOCKS-UP
DURING MEMORY POST COUNT, THEN YOU DON'T HAVE THE
REQUIRED HARDWARE LEVEL. BACK-OFF TO B4.12 UNTIL YOU
CAN GET YOUR 7526 PROCESSOR CARD UPGRADED.

Version B4.12 11/12/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Various bugs which could cause terminal reset, lock-up, or lack
of communications response in a heavily-loaded Ethernet
environment.

Version B4.06 10/7/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Printer port would not print on next power-up if the 7526 was
turned off while data was still in its out-bound printer queue

Enhancement:
-----------
1 Transaction program printer command now does automatic
spooling of data to the printer port, allowing you to do SEND to
Printer commands repeatedly without intervening DELAY statements
to avoid printer buffer full errors.

Version B4.05 9/16/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 A scanner sensor port used for the input on a Read with Verify
would be locked out of usage if a remote validation response
timed out.

2 Potential fix for occasional loss of communcations to terminal.
(Sorry, can't be more definite than that.... )

Version B4.04 9/10/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 Interactive mode -- 7526 will now wait up to 5 minutes for
activity from the server before dropping out of service. It
was using Timer A, which cannot be set high enough to match
the RUTHERE polling frequency of DCC/2r.

2 Failure to respond to commands -- when DCC/2r is restarted,
it sets the sequence number on commands back to 1. The 7526
was failing to always notice this, and was not responding to
commands which it considered to be duplicates spawned by the
TCP/IP network.

Version B4.03 8/13/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 The terminal was sending multiple ACKs to command IM (set
multiple user variables in one command). This seems to be
harmless, but was not right....

2 TimerAlarmAsynch set-points (digital output points driven
automatically at particular times of day) could fail to
execute when more than 2-3 hours passed without the 7526
being set off-line/on-line (e.g. a time synch or download).

Enhancements:
-------------
1 Added display of user variables in idle prompt capability.
See http://wwprodsoln.bocaraton.ibm.com/dc/uvdisp.htm for
details.

Version V4.02 7/26/96 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
1 With certain file sizes and transaction buffer sizes, the
storage allocated for transaction program user variables may
be non-functional, resulting in strange data being read from
the user variables.

2 The presentation of the 'Low Power - Stop' message when input
power to the terminal drops to brown-out levels was sometimes
causing corruption of memory. This message is no longer presented
(the terminal now will simply appear locked up if power drops
below the threshold but does not then drop low enough to cause
a hardware reset).

Version V4.00 (ETHERNET CARD FLASH LOAD **ONLY** )

Bugs fixed:
----------
NONE

Enhancements:
-------------
1 This is the first code level which is resident in the
7526 Ethernet-Serial-Parallel card's FLASH memory module.
It includes the ETHERNET and TCP/IP support.

EPROM FOR BASE 7526
-------------------
Note: If you are using the 7526 Ethernet-Serial-Parallel card,
you must have version 3.0 or higher. If you have version 3.0
already and are using the Ethernet card, then there is NO need
to upgrade your EPROM beyond 3.0 -- the flash file on the Ethernet
card is used instead of the EPROM for all but the first stages
of machine boot-up.

Note: Begining with V3.10, all even-numbered EPROMs require
7526 processor cards with the fixes described under FLASH V4.20
above. Even-numbered EPROMS are available only factory-installed.
Upgrade EPROMs from 3.11 on will run on any processor card level.

Version B3.31 03/25/98

Bugs fixed:
----------

1 Back-to-back SEND then remote verification transactions could confuse
the terminal, causing it to ignore the response for the remote validation
and/or send some data twice.

2 Terminal may hang as it is being put off-line by the controller. This
happens more frequently when the DCT is executing a CFR (doing IdleManager
calls) when the off-line command A0 is received.

New feature
-----------

Version B3.25 11/03/97

New feature
-----------

1 Menu system now includes a 'Lock Adjustments' option which,
when set to 'YES', prevents the terminal user from adjusting
the beeper level or the LCD contrast.

Version B3.23 10/21/97

Bugs fixed:
----------

1 Transaction was being sent at end of txn program with AUTO set to OFF when
worker would press Enter more than once on last input.

Version B3.21 9/15/97

Bugs fixed:
----------

1 Send data on serial port was broken -- would send endlessly, going
past end of output string.

Version B3.19 8/29/97

Bugs fixed:
----------

1 Fix for E03 error when paused in a ONKEY goto/gosub command AND a
digital-input or RS-232-input initiated event happens.

2 Fix for using the F28 (space) key in an ONKEY command

Version B3.17 6/10/97

Bugs fixed:
----------

1 Fix for breaking printer function when there is no serial-parallel
card installed. This was broken in 3.11-3.15 (not including 3.12).
Units with serial-parallel card would work just fine -- units without
won't print nuthin...

Version B3.15 5/30/97

Enhancments:
-----------_

1 Added TimerGetTicks and TimerResetTicks CFR APIs

Version B3.13 5/27/97

Bugs fixed:
----------
1 Fix for various bad symptoms when a power voltage to the terminal fell or rose slowly
through the Non-maskable Interrupt power failure voltage level.
(as when the Operational Battery Backup feature is installed and is allowed
to run all the way out of power). Symptoms could be: Loss of files, corrupted data,
corrupted idle prompt, terminal lock-up upon re-start of the 7526.

2 Serial port reads could get garbled data

Version B3.11 5/12/97

Bugs fixed:
----------

1 Printing was prone to hanging up the machine ONLY WHEN the
Serial-Parallel adapter was installed.

2 :J FORMAT transaction programming command did not work
for variable length output -- target UV was always
empty when finished.

3 Additional fixes for time-of-day driven alarms -- there was a
possibility of an alarm set-point being skipped during the day.

4 Printer spooling timeout was increased to accomodate printers
which suspend accepting new data while form-feeding.

Version V3.06 09/28/96

Bugs fixed:
----------
1 Printer port would not print on next power-up if the 7526 was
turned off while data was still in its out-bound printer queue

Enhancment:
-----------
1 Transaction program printer command now does automatic
spooling of data to the printer port, allowing you to do SEND to
Printer commands repeatedly without intervening DELAY statements
to avoid printer buffer full errors.

Version V3.05 9/16/96

Bugs fixed:
----------
1 A scanner sensor port used for the input on a Read with Verify
would be locked out of usage if a remote validation response
timed out.

Version B3.04 08/19/96

Enhancements:
-------------
1 Fixed E09 error when running a fast clocking transaction
after a CMD D transaction program had aborted prior to
sending a transaction.

Version B3.03 08/13/96

Enhancements:
-------------
1 Added display of user variables in idle prompt capability.
See http://wwprodsoln.bocaraton.ibm.com/dc/uvdisp.htm for
details.

Version B3.02 07/26/96

Bugs fixed:
----------
1 The presentation of the 'Low Power - Stop' message when input
power to the terminal drops to brown-out levels was sometimes
causing corruption of memory. This message is no longer presented
(the terminal now will simply appear locked up if power drops
below the threshold but does not then drop low enough to cause
a hardware reset).

Version V3.00 05/14/96 (THE LATEST EPROM CODE FACTORY-INSTALLED)

Bugs fixed:
----------
1 A Call CFR (CCFR) parameter string which used a
'+' character would get corrupted by the microcode as
it scanned for ONKEY labels to do address-fixup.

2 The processor board DI/DO port now comes up with power
to pin 20 off until a valid file 0 configuration record
is received.

3 Intialized DI point states to prevent false triggering
after download.

Enhancements:
-------------
1 Support for Ethernet-Serial-Parallel and Serial-Parallel
optional communications adapters.

Version V2.50 02/12/96

Bugs fixed:
----------
1 The subroutine 'stack pointer' was not being reset to
0 upon starting a transaction program. Thus, if you cancelled
(shift-clear) a program while it was in a subroutine the
stack would not be released. Cancelling 25 times would cause
an E08 stack error ultimately. Fixed in this version.

2 Timer alarms (both async and direct) now properly calculate
the day of the week and will execute on the day(s) desired.

Version V2.40 12/14/95

Bugs fixed:
----------
1 A ;N (branch based on user variable comparison) command
which is done after any kind of subroutine call command
(e.g. Gosub, OnSub ) will accidentally treat the branch
(goto) to the label as subroutine call. 25 times through
this loop will cause an E08 stack error. V2.40 fixes this.
Workaround: Put a 'do nothing' Goto command/label set
right before execution of the ;N.

Version B2.34 10/09/95

Bugs fixed:
-----------

1 Time interval-initiated events (i.e. TimerSetInterval)
would run even when the terminal was in Keylock (set-up) mode,
often overwriting the set-up display. Interval events now
will not run when the terminal is in set-up mode.

2 Transaction programs would encounter an E03 or E06 error
when a SKIP return code from a CFR would cause an attempt
to skip over another CCFR statement which had one of various
ASCII characters below DEC 048 in the free-form parameters
string. The SKIP function will now properly handle all
characters in the parameters EXCEPT a colon (:) or semi-colon
(;) which are special delimeters in the transaction programs.
Colon and semi-colon should NOT be used in CCFR parameters.

3 The microcode now constantly re-enables the time-of-day clock
in order to keep it from getting stuck in a mode where it
would not give its 1 second interrupts. This should prevent
the idle screen and transaction timestamp clock from getting
stuck.

4 Fixed low-power/power-off interrupt code to properly set
the data segment before trying to show 'Low Power. Stop.'.
Symptom was having the 'Low Power. Stop' message show up
on split onto two line on a 7526-150/155/200.

Version B2.32 09/25/95

Bugs fixed:
-----------

1 If an on-line/off-line command sequence (often as part of a
time synchronization command) was received while the 7526 was
in a CFR, the terminal may come back on-line with both the
idle screen showing AND the CFR running. Ultimately, the
7526 will reset (when the CFR did a return).

NOTE: The fix involved a change in operation which may
help those who suffer from having the clock stop on the 7526
display.

Version B2.30 08/01/95

Bugs fixed:
-----------

1 A bar code sensor port can get locked out (non-functional)
when both ports are used simultaneously.

Version B2.29 03/28/95

Bugs fixed:
-----------

1 Fixed bug introduced in B2.26-B2.28 which made time-of-day clock
stop IF the processor card in the 7526 was printer-capable
board (surface mount chips) WITH a Benchmarq brand real time
clock instead of the older Motorola clock chip. (The clock chip
is located underneath the beeper on the board ).

Version B2.28 03/15/95

Bugs fixed:
-----------

1 The debouncing algorithm for both DI-initiated transactions and
debounced DI counters was not working properly for bouncing
which lasted over 4-5 milliseconds, resulting in missed DI
activations.

Version B2.26 06/27/94

Bugs fixed:
-----------

1 If lower case hex characters are used for the ;A key command,
the transactions sent to the host contain that same lower case
letter.

Workaround: Change the characters to be upper case.

2 If an alpha key is pressed as the first character when using
Read (:R) with numeric only from the keyboard and sensor ports,
the sensor port is locked out, and the timeout never happens.

Workaround: The keyboard still works, enter the badge by hand
or press Clear to abort the transaction.

3 Validation does not work in the second part of a link key.

Workaround: Extend the length of the transaction. Link keys are
no longer needed to make long transactions.

4 The Model 200 acts like a 150/155 if the Space key is pressed
immediately after entering setup at any time.

Workaround: None known.

Enhancements:
-------------

1 Alphabetic keys are now ignored when using numeric only keypad
entry with sensors under Read (:R). The sensor stays enabled
until the first numeric key is pressed.

2 Added code for confirmation on whether or not the terminal is
a Model 150/155 or a 200.

Version B2.25 05/23/94

Bugs fixed:
-----------

1 Alarms in the Timer* CFR APIs don't trigger at midnight.

Workaround: Move the timers to be a minute later

2 The clock stops as shown on the display and in transactions.

Workaround: None. You need to reset the terminal to
continue processing either through pressing Clear,
or power off/on.

Version B2.24 04/14/94

Bugs fixed:
-----------

1 Fast-clocking transactions always try to do a remote validation
after a cold start, but before a validation is performed through
a key-initiated transaction.

Workaround: Perform a key-initiated transaction with validation
after a cold start to make everything work correctly.

2 Cursor positioning via escape characters embedded in Files 2
through 8 does not work.

Workaround: Use :U (cursor position) to place the cursor or use
:E (show text) to place the text at the desired
location.
NOTE: The 7526 requires two escapes to come to the terminal for
one escape to be placed in the data file. If only one
escape character is in the file, it will be discarded.

Version B2.23 03/07/94

Bugs fixed:
-----------

1 Removed some partial changes that were causing downloading
problems in B2.22.

Version B2.22 02/16/94

Bugs fixed:
-----------

1 Fixed a problem of not setting the return code properly for
TimerClear* functions.

2 Fixed an incorrect calculation of the length of beeps for good
and bad indicators. It would not work properly for anything
over 1.25 seconds.

3 Fixed two incorrect calculations in the TimerSet* CFR APIs that
were causing timers to not work correctly. The timers are still
not 100%, but they are getting better.

Version B2.21 12/21/93

Bugs fixed:
-----------

1 Fixed a problem with getting a watchdog reset when attempting to
decode a UPC or EAN in reverse direction.

Version V2.10 10/07/93

Bugs fixed:
-----------

Fixes in all 2.1x beta levels are included in V2.10. Other than
the version level, B2.13 is identical to V2.10.

Enhancements:
-------------
Model 150 support, which is a 7526 with 32K of RAM, and a 20-key
keypad with a two-line display.

Version B2.13 09/24/93

Bugs fixed:
-----------

1 Fixed problem of not seeing the toggles on the DI points from the
DidoRead CFR API.

2 Fixed problem of not using the correct day in the TimerSetAlarm
and TimerSetAlarmAsynch CFR APIs.

3 Now kill latch key transactions when setup is entered. In the
past, the latch key would take the display from setup.

Version B2.12 08/31/93

Bugs fixed:
-----------

1 Fixed problem of not giving good beep in diagnostics for successful
magnetic or bar code scan.

2 Fixed problem of accepting user variable 0 as a target for :F
Append to user variable command.

3 Fixed problem of accepting user variable 0 as a target for ;7
Append string literal to user variable command.

4 Fixed problem of the reader not working on a CFR call following a
DELAY.

Workaround: Display a message, and then immediately go into the
CFR. The message content is unimportant, and the
CFR can clear it right away.

Instead of doing:
DELAY
CCFR
...

Do this...
DELAY
SHOW
CCFR
...

Version B2.11 08/06/93

Bugs fixed:
-----------

1 Fixed problem of code page 857 characters on the upper half of
the code page not displaying properly.

2 The terminal no longer hangs if Enter is pressed rapidly before
display of the copyright statement.

3 Model 100 no longer loses the display after doing the memory test
from the diagnostics.

4 The printer option no longer shows in the diagnostics menu on the
original level system boards.

Version V2.00 06/11/93

Bugs fixed:
-----------
Fixes in all 1.03x beta levels are included in V2.00.

1 Updated CFR library to allow access to user variables > 255.

2 Changed the communications handler so that an ACK which comes
immediately is still recognized.

3 Fixed problem of user variables being read into too small
of a buffer.

4 Fixed race condition on :W that could cause the file 0 pointer
to go past the end of the transaction.

5 Added code to prevent reading of user variables before File 0
has been downloaded.

6 No longer allow going into setup again via Shift-Keylock if
already in setup.

7 No longer leave the good transaction message on the screen after
heavy transaction input.

8 Changed the Port A interrupt watcher to run at 50ms instead of
100ms.

9 No longer allow writing to user variable 0 via CMD I.

10 No longer assume Enhanced mode for AliasTransID CFR API.

11 The terminal now rejects CMD 9D2 if no data is provided.

12 CMD 9E0 is now rejected if the terminal is not loaded.

13 Fixed problem of calling a CFR twice in a row locking the
keyboard.

14 An error is now returned if a SendTransaction is done before a
key has been pressed or AliasTransID has been called.

15 Host command Q now gives an error if a RS is not at the end of
the user variable number.

16 Trying to display messages beyond the end of File 8 now give E02.

17 CMD F now returns the right value after a download.

18 No longer reset when many transactions are initiated without pause
in a single transaction program.

19 No longer give double transactions when setup is entered and exited
while in the midst of a download.

20 No longer watchdog reset on the first read after cold start if a
transaction has not been initiated, and fast-clocking or badge-
initiated transactions are not defined.

21 No longer allow loading of files while the terminal is online.

22 No longer show the "Updating the clock..." message unless all tests
have passed.

23 Port B no longer drops off if a lot of transactions are entered
using both ports simultaneously.

24 Now allocates memory for user variables on small file 0 download.

25 Fixed several different problems that were caused by errors in the
user variable memory allocation.

26 Fast-clocking is now removed when a new download with it absent is
used after it was previously active.

27 Fixed some problems with File 9 that caused the buffer to wrap in
cases where a lot of transactions were entered quickly.

28 Now give BUFFER FULL and abort the transaction when the buffer is
full.

29 DI counters are no longer cleared on download.

30 ClearUserVariable was changed to clear the transaction record
instead of user variable 100.

31 The BUFFER FULL message no longer leaves before there is enough
space for the next transaction.

32 The last port read value is no longer affected by reads that
occur within a CFR.

33 Fixed some interaction problems with CMD D and latch key.

34 Fixed a problem of CMD Q returning too much data when in
compatibility mode.

35 Made the timings for asynchronous DO setting to be more accurate.

36 Now put the result of Read DI in user variable 0.

37 Offline/Online while within a CFR no longer hangs the keyboard.

38 Fixed problem of display showing wrong info when SEND followed
by a SHOW command.

Enhancements/new features:
--------------------------

1 Added :G (Goto), :GS (Goto), :Z (Gosub), :ZS (Gosub), ;R
(Return), ;N (Conditional execution), and (Label) key commands.

2 Added GetTerminalAddress CFR API (Miscellaneous API).

3 Added direction of last read info to SenRead and SenQueryPort APIs.

4 Changed CMD J9 and CMD 6 to show printer selection and the
board level.

5 Added printer support. Includes CFR APIs, :W support, and
file 0 record 0. Also has \x?? support and diagnostics.

6 Added SKIP/NOSKIP support.

7 Added remote validation support and CMD V.

8 Added the Format User Variable command. :J.

9 Added optional positioning information for CMD B and CMD C.

10 Added file number and spinning character to indicate packets
received.

11 Now allow the red LED on port A to be used even if the HP
barcode is present.

12 The user variables are now all cleared when a new download is
received.

Version V1.03h (BETA) 04/03/93

1 Fixed the problem of not updating the seconds on the time if a
transaction program was written such that it did not return to the
idle screen between successive transactions.

Workaround: None.

Version V1.03g (BETA) 03/24/93

1 Fixed a problem where the magnetic reader on Port A would be
lost if a read was done before pressing a key, and no definition
for fast-clocking or badge-initiated transactions exists.

Workaround: Press the Clear key (shift-clear on M200) to get the
reader working again.

Version V1.03f (BETA) 03/12/93

1 Fixed power-down code. It would, on rare occasions, cause the
terminal memory to be corrupted, and either cool or cold start.

Workaround: None.

Version V1.03e (BETA) 01/05/93

1 The first I2of5 read after a warm start, shift-clear, or power off
then on is always unsuccessful.

Workaround: Scan again. Later reads should work fine.

2 Fixed out of service bug where power-off/on caused the
terminal to come on-line.
This problem still exists in V1.03. It is not fixed as previously
reported.

Version V1.03c (BETA)

1 Interleaved 2/5: The general purpose READ (:R) command does not
appear to properly set up the required fixed length of the I2/5
symbol for the decoding algorithm.

Workaround: If you are using the READ command,
try converting to the FB (fixed badge) command instead.

2 Interleaved 2/5: We have found that if
a) You are using I2/5 WITH CHECK DIGIT,
and b) You are using port A
and c) A bad check digit is decoded

then port A will not work again until you cycle power on the 7526.

Workaround: Use port B, if not already in use.
Don't do check digit verification.

3 Fixed a bug with :5 (set Time Of Day transaction).

Workaround: None known.

4 Fixed bug of hanging when CMD I is sent and the terminal is not
downloaded.

Workaround: Do not try to use the user variables before the
terminal is downloaded.

5 Shift-clear now breaks loops where no task is watching the
keyboard.

Workaround: None known.

6 Fixed the problem of not giving an E02 when trying to access a
message number higher than the last one in the file.

Workaround: None known.

7 Disabled the ALT key light while in diagnostics.

Workaround: None known.

8 Fixed a bug of allowing adjustments in a CFR or access to setup
when either setup or adjustments are selected, and the other is
turned off.

Workaround: Use a password to protect access to setup. No way
to prevent access to the adjustments other than turn
off both setup and adjustments from within the CFR.

9 Fixed bugs in I2of5 decoding that were preventing good read rates
with the I2of5 barcode symbology.

Workaround: None known other than switching to another barcode
type.

Enhancement in 1.03c:

1 Added a message to show on the display when the power drops enough
to trigger the power down sequence. This is a diagnostic message.

Version V1.03b (BETA)

1 Special characters in numeric keypad input modes (e.g. NK command,
READ with only numeric keys enabled) cannot be input (e.g. decimal
point, percent, or any other shifted keys of the numeric area.

Workaround: Use the READ command and enable all keys (alphanumeric)
for input, or use the input CFR in the 7526PANL PACKAGE on CIMTOOLS.

2 Shows a "Power Fail" message when bulk DC power failure is detected
(for informational purposes), but 5V power has not yet failed.

Workaround: If terminal appears to be 'hung', cycle power on the
unit and re-try.

3 The date and time would not re-display after issuing a CMD 2 with
a DC2 character.

Workaround: None known.

Version V1.03 10/23/92

Bugs fixed:
-----------

1 Fixed bug of enabling the DIs when the clock interrupt was
handled.

2 If multiple beeps with a long duration occur in rapid
succession, the beeper would stick on.

3 Fixed bug of not giving E02 message for some cases when the
message was not in File 8.

4 No longer accept period into compacted file 9. It was being
treated as a RS, which was breaking the record count.

5 Problem now fixed. See 1.03e beta code.

6 Fixed bug in sub_pointer that caused 64K boundary problems.

7 Added code to disable RS-485 transmitter in power-down NMI
interrupt so that terminal does not interfere with communications
line when operational battery backup drained to below 8.5V.
Also turn off the DO ports when in power-failed mode.

8 Fixed CFR validation bug. 1/256 chance of the wrong
validation result being returned.

9 Fixed bug in keyboard device driver of not giving space character
properly to CFRs. Also, the shifted keys in a CFR were not
being mapped into lower case characters.

10 CMD B and CMD C are now ignored in buffered mode.

11 Fixed error in calculating when to go into fast-clocking.

12 Fixed bug in allocating memory for user variables when there
is memory available that has not been allocated to files.

Version 1.03 includes fixes as described below for all previous
releases

Version V1.02 07/28/92

Bugs fixed:
-----------

1 File 8 now gives an error if it is initialized as a numeric
file.

2 Added checking in SenSetDiscrimGroup for I2of5 being even and
not length of 0.

3 Fixed problem of wrapping back to the first line of the
display on Model 200 when writing beyond the end of the
second line.

4 Fixed problem of accepting T instead of 2 for CMD E (write
DO) to toggle the DO point. T is now illegal.

5 Added HP barcode defines and fixed bar code type in CFRs.
(Refers to CFRAPI26.H file on programming support diskette).

6 Fixed error status not being set for CMD E0 when the port
is set for DI.

7 Changed CMD 9B to reject size of 0 for File A and File E.
Allow File 9 to have a size above 65535. Changed the file
size maximum to be 65535 instead of 65500. Corrected an
error for CMD 9Bz (broadcast).

8 Successive SpkBeeps interrupted by a key beep now work.

9 Files are no longer erased when changing a communication
parameter.

10 Fixed several problems in compacted mode for several 9x
commands.

11 Fixed compacted mode validation and special cases of
non-compacted validation.

12 Password now is erased when the pins are jumpered and a cool or
warm start occurs.

13 Fixed problem of password getting set if Clear is pressed
on the password prompt in setup to abort entering a password,
then re-enter setup and go past the password prompt with
ENTER.

14 Fixed CMD F to return the correct value for memory remaining.

15 Fixed problem of not being able to validate > 40 bytes
against files 2-7 via CFR or key command.

16 Now ignore line 2 on Mod 100 if specified on via File 0
Record 0.

17 Time now displays after power off/on if DC2 delimited
message was on the display.

18 Fixed characters representing STX and ETX in datascope
mode. They were wrong due to changes in the display
code page.

19 Closed two windows in CFR exiting that could prevent the
keyboard or reader from working.

20 Fixed problem of not re-enabling a laser scanner after power off/on.

21 Corrected loss of the last character of the transaction
buffer due to incorrect size calculation.

22 When the transaction buffer was filled, going into setup
and then returning via shift-clear was setting the clock
to all 000s.

Enhancements/new features:
--------------------------

A Changed default for CMD H from 50ms to 5ms.
B Added CMD Q (Query User Variable)
C Rewrote the keyboard code to enable adjustments and entering
setup at all times, except when specifically prohibited.
D Added code to cold-start the terminal when a new version
EPROM is put in the system.
E Added code to do more extensive testing of memory on a warm
or cool start to make sure that memory is not corrupted due
to power being off.
F Added ability to configure the number of user variables via
bytes 140-142, and the size of them through bytes 143-145.
G Note that this can cause an error if the user requests the
exact amount of memory for file 9, instead of telling it to
use the rest of memory (size 0).
H Added ability to kill a CMD 2 with a DC2 by sending a DC2
with no data behind it.
I Now start the RAM count on a cold start with 000K.
J Added ability to set up variable length fast clocking reads.

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

B) CONFIRMED BUGS: FIXES NOT YET AVAILABLE

REPORTING
SYSTEM
NUMBER DESCRIPTION

1 If setup is entered while in Latch mode, the latch mode will take
the display.

Workaround: None known.

2 If I2of5 is defined, variable length reads can not be performed
with the magnetic reader.

Workaround: Use fixed length reads.

3 When using the laser to read bar codes in diagnostics, if the
terminal has been downloaded, only the bar codes specified in the
download can be used by the laser. If the terminal has not been
downloaded, all bar code types are available. This is not
consistent.

Workaround: Cold start the terminal if you need to test all types
of bar codes.

4 If file A is downloaded at a size smaller than 161 bytes, no
error is given.

Workaround: None should be needed, since this is an error case
that is not flagged.

5 When performing validations against large files, the terminal
allows the data that was entered to be cleared and re-entered, or
to even abort the transaction, but does not affect the state of
the validation.

Workaround: Do not change the data that was entered, or abort the
transaction while the validation is in progress.

6 The beep length for good and bad beeps is not correct for values
longer than 2.5 seconds.

Workaround: Use beeps of less than or equal to 2.5 seconds.

7 The number of beeps is not always correct for good and bad beeps.

Workaround: None known.