From mikec Mon Aug 19 13:10:47 1991
Received: by cbmvax.cbm.commodore.com (5.57/UUCP-Project/Commodore 2/8/91)
	id AA19754; Mon, 19 Aug 91 13:10:43 EDT
Date: Mon, 19 Aug 91 13:10:43 EDT
From: mikec (Michael Colligon - Product Assurance)
Message-Id: <9108191710.AA19754@cbmvax.cbm.commodore.com>
To: barker, drysdale, havemose, ian, jss, mario, ned
Subject: A2386 open issues
Cc: givler, jjszucs, mikec, rob
Status: RO


Please take a look at these issues.  We will get together to discuss them
soon.  Right now I'm shooting for tomorrow at 3:00 PM in the PA conf. room.

Mike


NOTE:  Some of the problems with the A2386 documented in this report may have
       been fixed by recent H/W hacks to the A2386 and more will be fixed on 
       the new revision board.  They are still on this list, because the PC 
       people do not have the time to retest their bugs.



OUTSTANDING A2386 HARDWARE/BIOS ISSUES
======================================

1)  Using an A:=MUX720 and B:=A1011 B: does not work at all.

2)  PC Ctrl-Alt-Del during CONFIG.SYS causes PC to lock up during subsequent 
    reboots.  The Amiga has to be rebooted to for the A2386 to be used.

3)  SETUP HDD types should allow PgUp as well as PgDn.

4)  SETUP "Shadow" line should be changed to "Shadow Janus."

5)  PC hangs when software warmboots are attempted.  If a program attempts to 
    reboot the system it may hang.  FDISK and PCMAG's WARMBOOT.COM show this 
    problem.  Only occurs if drive type 41(auto-configuring) is selected.

6)  IRQ 12 does not appear to function correctly.  Tested on both samples.

7)  The Bridgecard Setup screen should not refresh the drive type list
    when switching between HD1 and HD2.

8)  Maxtor 212M drive would not initialize, a "Hard Disk Controller"
    failure occurs.
        
9)  A "Pause/Break" key would be very useful for scrolling large text and
    aborting out of an unwanted file execution. 

10) In several places in the BIOS, the string "hit <PCHard>" has replaced 
    "hit reset".  I think a better string would be:  "use PCHard".

11) The BIOS floppy error messages, such as:  "Diskette <n> seek failure" refer 
    to A: as drive 0 and B: as drive 1.  However, in the SETUP screen, A: is 
    referred to as drive 1 and B: is referred to as drive 2.  I recommend 
    changing the BIOS to refer to drives 1 and 2, since this is the PC standard.

12) I've just gotten another "non-reproducible bug" that might ring a bell to 
    you guys.  I connected an IDE controller (with a Quantum 40AT) to my A2386 
    in my A3000. I configured the drive as type 41 (Auto-Config).  When I 
    re-started the system, I got a:  "No operating system" message from the 
    BIOS.  I turned everything off and then on again and it worked fine.

13) The BIOS SETUP program does not handle <Esc> properly.  This was tested 
    on two systems with different A2386's (one modified; one not).  PCPrefs' 
    video had both emulations enabled with monochrome as the default.  SETUP 
    had video set to MONO.  PC Mono was used as the video window.



OUTSTANDING MS-DOS COMPATIBILITY PROBLEMS
=========================================

1)  The Iomega Bernoulli Box II will not operate if the virtual/fake drive is 
    used, but works fine if setup as the primary drive (no virtual/fake drive).

2)  PC Technician 386 will not boot with the virtual/fake drive enabled.

3)  Boot-up problems with the QuadEGA+ video card.  If PCPrefs (on the Amiga) 
    has color video on, then no errors are seen.  This is directly opposite 
    from a VGA card.  Looking at port 80 (POST error codes) shows no errors.

4)  Boot-up problems with the Hercules Graphics Card Plus.

5)  MSDOS 5.0 EMM386.exe needs help in loading page frame for expanded memory.

6)  The 3COM Etherlink 16 network card has file transfer problems.
    Tested in fileserver configuration with Novell 386 3.11.

7)  The 3COM Etherlink 16 network card cannot be used in a workstation 
    configuration under Novell 3.11.  Due to conflicts between Novell's 
    IPX.COM, the bridgeboard's use of the D0000H memory area, and the 
    Etherlink's SETUP utility, the card cannot be setup as a workstation.

8)  Lotus 123 displays a screen of garbage immediately prior to graphics mode.

9)  Laplink III did not work. The parallel port would not connect with the 
    C386SX-LT for file transfers.

10) Mathmatica fails during installation.  After the second disk is copied, 
    the install program attempts to create a swap disk space.  If there is 
    insufficient space the program aborts with a notice of the space 
    restriction.  The Bridgecard hangs with an audible alarm and blank screen.

11) During Paintbrush IV setup the Numlock key functioned correctly only with 
    the 2,4,5,6,8 keys.  The dedicated cursor keys were somehow remapped as 
    numeric keys.  On exit, the state of the numeric keypad is altered.

12) Cursor keys will not function with F19 and DBase IV. 

13) Cursor keys lock at the third setup option on Software Carousel.

14) The ALT-SPACE key combination in Software Carousel which normally 
    produces a menu, does not function.

15) Balance of Power (Original -- Mindscape):  Windows 1.0-based; 
    no AMouse compatible mouse driver for Windows 1.0 is available, 
    preventing use of mouse; works in all other areas.
    Balance of Power (1990 Edition -- Mindscape):  Windows 2.x-based; 
    works in all areas in real mode of Windows 3.0

16) M1 Tank Platoon (Microprose):  
    "General failure reading drive C" during load from AutoBoot partition; 
    causes AutoBoot partition to become inaccessible; 
    the Amiga must be rebooted to regain access to the AutoBoot partition; 
    believed to be related to copy protection



A2386 RAM CARD TEST RESULTS
===========================

The matrix below states the pass/fail status of the RAM cards tested
with the A2386SX. Parenthesized numbers indicate references to notes
(located after the matrix) detailing the results.

Test procedure consisted of:

1) Installing hardware (and any associated software) per documentation.

2) Verify that extended memory is recognized is tested by POST.

3) Verify that extended/expanded memory is recognized by Manifest and CheckIt.

4)Run quick and full tests of extended and/or expanded using Checkit.

                                Expanded (EMS)          Extended

AST RamPage Plus 286            Fail (1)                Pass (2, 3)

Boca Research BocaRAM/AT        Pass (4)                Pass (2)

Boca Research BocaRAM/AT+       Pass                    Pass (5)

InterQuadram QuadRAM-AT         Pass                    Pass (2)

Monolothic JustRAM/ATZ          Not Tested (6)          Pass (2)

Everex RAM8000 AT               Not Tested (6)          Not Tested (6)

AST Advantage Premium           Not Tested (6)          Pass (2, 7)


Note 1: The EMS driver software failed with the message: "No EMS boards found"

Note 2: The extended memory was not configured in SETUP.  As a result, CheckIt 
failed to automatically recognized the memory.  Manifest and the BIOS POST both 
recognized (and tested, in the case of the POST) this memory.  CheckIt testing 
was performed by manually entering the memory address range of extended memory.

Note 3: The extended memory must be configured to address from 1664KB
($001A0000).  No explanation for this is currently apparent. Further testing 
will be performed to determine if this is due to the nature of the card.

Note 4: The expanded memory must be configured to place the frame at A000.

Note 5: The software for this card automatically configures the CMOS extended 
memory field, allowing it to be recognized by CheckIt and other software that 
queries this field.

Note 6: Not tested -- required software not available.

Note 7: The extended memory must be configured to address from 1408K
($00160000).  No explanation for this is currently apparent. Further testing 
will be performed to determine if this is due to the nature of the card.



OUTSTANDING A2386 UNIX COMPATIBILITY PROBLEMS
=============================================

1)  A "System Request" key must be added to the Amiga keyboard 
    layout to allow virtual terminal use in Interactive Unix.   

2)  The Bridgecard locks up after attempting to reboot out of FDISK on a 
    dedicated drive.

3)  "Hard drive controller failure" occurred twice during a PC boot.  
    Re-entering Setup and rebooting seemed to correct the problem.
    An IDE controller board was used during this occurrence. 

4)  UHC Unix fails to install and to boot on a previously installed drive.



NON-PROBLEM - SPEC.
===================

1)  SETUP allows you to specify memory above 640K as both extended memory and 
    shadow memory.  Apparently an attempt was made to fix this bug; however,
    there is now a new bug.  Now you can never specify that memory between
    640K and 1MB should be used as extended memory, even if all shadowing is
    disabled.



PROBLEMS SOLVED (BIOS 1.00.01)
==============================

1)  The Bridgeboard boots up in low speed.

2)  Escaping from SETUP will hang system.

3)  During boot-up, error messages indicated invalid configuration occurred.

 

PROBLEMS SOLVED (BIOS 1.00.02)
==============================

1)  A "warm reboot" failed after SCO Unix and SCO Xenix were shutdown.  A 
    "cold reboot" from the PC Hard icon resulted in a loss of keyboard use. 

2)  If no boot device is found, an error message of 
        "F1 to continue, F2 to enter SETUP".
    It should read <CTRL-ALT-ESC> instead of F2.  
    F2 will not allow entry into SETUP.


PARTIALLY FIXED
===============

1)  Some messages contain a string saying "...or hit reset switch to retry".
    The system has no reset switch.

From mikec Mon Aug 19 13:13:57 1991
Received: by cbmvax.cbm.commodore.com (5.57/UUCP-Project/Commodore 2/8/91)
	id AA19948; Mon, 19 Aug 91 13:13:42 EDT
Date: Mon, 19 Aug 91 13:13:42 EDT
From: mikec (Michael Colligon - Product Assurance)
Message-Id: <9108191713.AA19948@cbmvax.cbm.commodore.com>
To: barker, drysdale, havemose, ian, jss, mario, ned
Subject: Janus open issues
Cc: givler, jjszucs, mikec, rob
Status: RO


Please take a look at this too.  The same meeting tomorrow will discuss these
issues.  I plan to do H/W first, then S/W.

Mike

PS:  Unless you hear otherwise please plan on the meeting being at 3:00 PM
in the PA conf. room.  If you have a problem with that time please let me know.



PCWindow
========

Under certain circumstances, it is impossible to get two PC Mono
windows to appear on the same screen, except for the Workbench screen.
This also happens with "PC Color" windows if you select four colors.

The latest incarnation of PC Color has a default color 0 that is somewhat 
red.  Color 0 defaults to 0x02 (Red), 0x00 (Green), 0x00 (Blue).  
It should be:  0x00 (Red), 0x00 (Green), 0x00 (Blue) 

    SWENG - there is a historical reason for the reddish 'black' pcwindow 
    background color.  i don't know what it is - next time i find andy 
    i'll see if he knows.

The "About" and "Help" options of PCWindow cause the window borders to
be displayed. Why is this?

In PCWindow, only the active PCWindow window is refreshed.
Turning Window Freeze Off and On, has no effect.

The Color window of PC Window has excessive use of caps on the gadget labels.
"CANCEL"	-	Change to "Cancel"
"COPY"		-	Change to "Copy"

The CHECKIT/CHECKED menu item flags should be used for toggle-type items in 
the PC Window menus.  Examples include:  This Window's Screen, Colors, and 
Window Freeze.  This is per the Amiga user interface guidelines.

In PCWindow, <Right Amiga>-<.> is used as a shortcut for the Paste item of the 
Edit menu.  The Amiga user interface guidelines specify <Right Amiga>-<V> as 
the standard Paste shortcut.  In fact, this is one of the few standards given 
in the original Intuition documentation (from the AmigaDOS 1.0 days).

Throughout PC Window, the menu command key equivalents are given in
lower-case. This is non-standard.  This applies to ALL command key equivalents.

In PCWindow, the sub-menu items are placed to the left of the menu items.
This is non-standard and should be corrected (the sub-menu items should be
placed to the right of the menu items.)

With the "ClickToFront" commodity, a double-click on a PC Window to activate 
it and move it to the front, gets interpreted as a auto-resizing command.

Why does the Edit menu of PC Window have a Copy item when text is
automatically copied to clipboard as soon as the selection is completed?
This seems superfulous.

According to the A2386 end-user documentation, double-clicking in a PCWindow 
window to re-size it may not be compatible with all PC applications.  Why is 
this the case?  Obviously, this is not going to work if you're running AMouse 
and have the mouse on the PC, but how can a PC application interfere with 
mouse input on the Amiga?



PC Mono
=======

Video test results generated by CheckIt:

Video RAM
---------
Pass

Character Set
-------------
Pass

Text Attributes
---------------
Normal - uses color 1 of palette; pass
Blinking - uses color 2 of palette; does not blink
Bright - same as blinking
Bright Blinking - same as blinking
Reverse - pass
Reverse Blinking - same as blinking in inverse video; does not blink
Underlined - pass
Blinking underlined - same as blinking with underline; does not blink
Bright underlined - same blinking with underline
Bright blinking underlined - same as blinking with underline; does not blink

I understand that actually having blinking text is difficult on the Amiga, 
unless you use a larger palette, since if you just use the 4-color palette 
and blink color 1, EVERYTHING on the screen that uses color 1 will blink.  

How's this for a possible solution:

1) If the user chooses a custom screen (since it would be rude to blink 
   everything in a certain pen on his Workbench screen), and

2) the user chooses a >4-color palette (since you need at least 4
   colors for the menus, window dressing, etc.)

3) then we use:  color 4 for blinking and color 5 for bright

Obviously, the palette entry for color 4 would alternate between color 0 and 
color 1 and color 5 would be like color 1, but brighter.  Similar suggestions 
would apply to PC Color.  Unfortunately, as I intended in my bug reports 
yesterday, PC Color is not a happy camper on my machine.



INSTALL
=======

Install (BridgeInstall on 2.1g2 of the PC Install Disk) crashes with enforcer 
hits while clicking on a vertical scroll gadget during the "Delete files" mode.

If the user selects the "Cancel" gadget in the Janus Install program, the 
requester:  "Installation error:  User Abort from Install List Edit screen"
is displayed.  This is unnecessary; the user knows that he hit "Cancel".

The requester title bar says:  "File Blocks Selected To Install [ 432]".  This 
probably applies to all error requesters, although none of the others have 
been seen.  This title bar should be changed to something like "PC Install".

The BridgeBoard installation program is not up to the user interface and 
visual standards of Release 2.0 of the Amiga System Software and does not 
meet the Amiga user interface guidelines.  It needs to be replaced.

The Janus install scroll gadgets do not auto-repeat.  Since each action moves 
one line, it takes a long time to move through a hard disk partition.

During the actual disk access involved in creating an autoboot partition, 
<Ctrl>-<C> cannot be used to abort MakeAB.

The BridgeBoard documentation indicates that C:MakeAB will fail with a
"Volume <volume> is full" requester, followed by a disk full error, if the 
volume does not have sufficent free space for the autoboot partition.
C:MakeAB should be modified to check the free space on the volume and fail 
with an error message indicating that there is insufficent free space for 
the autoboot partition, before starting the process.



PC INSTALL DISK
===============

Change the name of the disk to something without PC in the name.

The following filenames contain leading spaces and embedded spaces for
no apparent reason:  "SYS:PC/ Flipper .info", "SYS:PC/ LPT1 .info",
"SYS:PC/ PCHard .info", "SYS:PC/ PCPrefs .info".

    SWENG gave some convoluted excuse, that is really silly.

Recommend renaming the PCHard program to PCReset, or something similar.
PCHard implies, PC hard disk, or something related.

    SWENG agreed to fix but has not done so.

Change AMOUSE to refer to first and second ports instead of right and left, 
so that it won't be A2000 specific.

    SWENG agreed to fix but has not done so.

LPT1 is incompatible with CMD (or any other parallel.device re-director).



PCPrefs
=======

The "Use" option of PCPrefs appears to be irreversible.  To reproduce:
Use PCPrefs to change the "SAVE" settings and hit Use.  When PCPrefs is run
again, it comes up with the "SAVE" configuration (not the "USE" one).  This, 
alone, is a violation of the Prefs editor guidelines.  The configuration on 
screen cannot be used or saved because the gadgets are ghosted (since nothing 
has been changed).  So to get back to my saved configuration, I have to reboot.

Some of the gadgets should be checkboxes, not cycle gadgets.  Checkboxes are 
for Boolean choices, while cycle gadgets are to select one of <n> non-Boolean 
choices.  Here are the gadgets I think should be checkboxes:  Janus Handler 
Shadowed, Monochrome Display Adapter Emulation, Color Graphics Adapter 
Emulation, and Bridgeboard Speaker.  This will require a little layout change, 
since checkboxes and cycle gadgets placed close together look ugly.

PCPrefs doesn't support multiple preset files.

PCPrefs doesn't have the standard Prefs editor menus.



FLIPPER
=======

If A: is a shared Amiga/PC drive, inserting an Amiga floppy in DF0:, waiting 
for Flipper to recognize diskette as Amiga format, and rebooting the PC 
causes the PC side to display the message: "Diskette read failure - Press 
<Ctrl>-<Alt>-<Esc> to run SETUP, F1 to retry boot, " to be displayed.  This 
is incorrect.  It should ignore the Amiga disk and boot from the AutoBoot or 
IDE hard disk drive.  The trailing "," should be removed from the message.

    SWENG - there is no way for the pc to know if the amiga has 
    control of the shared floppy, so it tries to read it anyway.  
    you'll get the read error message, and that can't be fixed.

Flipper does not display a template when the "?" argument is given.
In fact, it appears to do absolutely nothing -- no error messages, no
window, no nothing.  This is in violation of the Amiga user interface 
guidelines.  As with MakeAB, Flipper should use ReadArgs under >=V36, 
or emulate the behavior of ReadArgs under <V36.

If Flipper is run with an invalid number of arguments (or no arguments), it 
outputs the message:  "Argument count".  The correct error message form is:  
<program name>: <error>.  In this case, <program name> is "Flipper".  The 
<error> should be a standard AmigaDOS error message.  So, if Flipper is run 
with an invalid number of arguments (or no arguments), it should output the 
message:  "Flipper: Required argument missing" (DOS error message 116) or
"Flipper: Wrong number of arguments" (DOS error message 118).

Flipper has a "MODE=QUIT" tool type.  This seems sort of unnecessary.  It 
also has QUIT command line option.  This also seems sort of unnecessary.  
Why not just let the user select the close gadget?



KEYMAP
======

The numeric keypad <Enter> key maps to the PC side numeric keypad <9> key.  
This was discovered using the QAPlus PC diagnostic software's keyboard test.  
The keymap specified was XT, which most closely matches the Amiga keyboard.
An attempt was made to fix this bug, but now the <Enter> does nothing.



DUAL PORT MEMORY
================

Dual port memory is corrupted when one of our services test programs is run.  
The program displays an Amiga screen on a VGA card.  The evidence of 
corruption is visual.  The corruption only takes place when Janus is shadowed.

    SWENG observed this problem by running DiskMasher on a DJMounted 
    drive while scrolling the PC screen.  SWENG observed that the problem 
    does not occur when the 2386 is in a 68000 based system.  However, 
    they believe that shadowing or not shadowing Janus has no effect.



AMIGA SCSI AND PC IDE CONFLICT
==============================

Start an Amiga HDD test, and while it is running, run the hard disk test from 
the CheckIt diagnostic package on an attached IDE drive.  The first time a 
"Controller error: 0Ah" message was displayed by CheckIt.  NOT REPRODUCABLE



AUTOBOOT PARTITION
==================

PC Technician diagnostic software fails with an AutoBoot partition on an 
A2386, but works fine from an IDE drive.

The A2286 hangs after running an application either Bridge or Starfleet II 
and doing a warm reboot of the bridgeboard.  Still hangs if PCHard is used.



ARead
=====

The default destination for ARead should be the current directory.  
This would make it compatibility with the MS-DOS Copy command.



PC COMPATIBILITY
================

Attempted to install Windows 3.0 on A2386 and noted two problems:

1) AMOUSE.COM was running.  Since this emulates the Microsoft Mouse, it was
   assumed that the Windows setup should recognize it and make that driver 
   the default.  However, the Windows setup did not recognize it at all 
   (and made the default:  "No mouse or other pointing device").

2) In the installation process, Windows attempts to detect what display 
   driver it should use.  It decided that the Bridgeboard's CGA emulation 
   was:  "Olivetti/AT&T Monochrome or PVC Display"

SWENG REPLY
-----------
that was the problem - if you don't use the new VDDCGA.386 display driver,
you won't be able to run the DOS shell from Windows 3.0 in 386 mode, or
any other DOS program for that matter.  Windows applications work just
fine.  once you install the new VDDCGA.386, you should be able to run the
DOS shell and DOS applications.

windows 3.0 won't use DOS mouse drivers, so having AMOUSE.COM running when
you try to install windows means nothing.  you must explicitly tell windows
which mouse driver to use.

windows does indeed think we have an olivetti/at&t display controller - the
reason is that we provide 32K of video memory on our "CGA card" but CGA is 
supposed to have only 16K.  the olivetti CGA has 32K, and that's how they 
"automatically" sense your display type.



DJMOUNT
=======

If the PC is rebooted or crashed during Amiga access to an ADisk/DJMount 
partition, the Amiga DJMount handler locks up.  However, if the Amiga is 
not accessing the DJMount partition, the Amiga DJMount handler continues 
to work after the PC restarts.

Janus may lose synchronization between the PC and Amiga if the Amiga 
runs DJMount while the PC is in the process of resetting.  The loss of 
synchronization is characterized by either the Amiga DJMount locking up 
or the PC AMouse displaying the message "Waiting for AMouse on Amiga
Press a key to abandom" and not returning (until a key is pressed).



JLINK
=====

The "JLink" section of the BridgeBoard/Janus documentation indicates that if 
the user specifies the /C option and the Amiga JLink file already exists, the 
user will be prompted with the option of replacing the file or terminating 
the operation.  In actuality, the operation is terminated without prompting.

In the "JLink" section of the Janus end-user documentation, the message:
"JLink Error: Drive is already linked, unlink first" is described as 
being displayed if the user attempts to link two JLink files to these 
same MS-DOS drive designation (such as D:).  The actual error message is:  
"Drive D: is already linked"  Which is correct? I prefer the first one 
(from the docs), since it provides a recommendation on how to solve the 
problem, instead of just telling the user that the error occured.

The header "JLink Error:" is indicated for some of the JLink error 
messages listed in the Janus end-user documentation. For example:
       JLink Error: Close error on virtual volume
The actual error messages in JLink do not have the "JLink Error:" 
header.  I have a slight preference for a header (just "JLink", 
though, not "JLink Error" -- obviously it's an error).  In any 
case, I'd like to see it consistent -- either all messages should 
have the header or none of them should.

In the JLink documentaton, the message:  "JLink Error: Nothing linked" is 
described as being displayed when the user attempts to unlink a drive that is 
not linked.  The actual message given by JLink is:  "Drive D: is not linked"
I prefer the form used by JLink.  However, I'd like a confirmation from
you that this what it's supposed to be.

In the Janus end-user documentation, the message:  "JLink Error: Close error 
on virtual volume" is described as appearing if a JLink operation is attempted 
and the Amiga volume containing the JLink file is removed with the link active.
The actual effect of this is the system requester:  "Please insert volume
<volume name> in any drive."

The A2386/Janus end-user documentation indicates that you should unlink any 
JLink drives before re-booting the Amiga.  This brings to mind the obvious 
question:  What happens if you don't?

SWENG REPLY
-----------
well, the rumor on rebooting without unlinking JLINK drives is that you will
lose some of the data (ie, the stuff most recently written won't get flushed
to the amiga disk).  i personally have never observed this phenomenon, but
it's probably a good idea.  it'll at least close the JLINK file so that
you could (for instance) delete it.



BUG DESCRIPTION
===============

The JLink mechanism causes a lock to remain open on the JLink file if
the PC-side is rebooted or crashes while a JLink drive is linked.


OUR SUGGESTION
==============

Our idea is to have something like a PC reset flag in the dual-ported memory 
that the Amiga services could watch.  This could be a memory location in 
dual-ported memory that contains a magic number indicating that the PC has 
reset.  One of the first things that the PC BIOS would do is set this flag.

When the Amiga saw this, it could close all locks associated with JLink 
devices and get everything (JLink, AutoBoot, AMouse, etc.) into a known 
startup state. Then, it would clear the flag and be ready to receive service 
requests from the PC.

I'm not sure about this, but isn't it possible for an Amiga program to 
receive an interrupt when dual-ported memory is written to by the PC?  
If this is the case, we could even avoid polling this memory location.

This sounds fairly easy to implement, and provides a good deal of
protection from havoc caused by resetting the PC.


SWENG'S NOT SO USEFUL RESPONSE
==============================

BUZZZZ!!!!  WRONG!!!!!

"EASY to implement", he says.  heh heh heh.

if it was so EASY, it would have been DONE a long time ago.  step 1: make 
the underlying garbage actually FUNCTION!  THEN you can start mucking 
around.  personally, i'd rather nuke the whole thing and start over.

patching things like resource tracking into janus is not impossible, but
i'm only one guy and there are a lot of more useful things for me to do
than figure out how to put in huge kludges without breaking everything
else.  wait for janus 3.0 (if there ever is one).

:) for the humor impaired.  but seriously, guys, stop speculating about how to 
do these things.  i *KNOW* how to do these things - but the original design 
didn't anticipate these things, so it's real hard to stick them in afterwards.



COMMUNICATIONS PROBLEMS - NOT BUG
=================================

Mike Brim and John Szucs attempted to reproduce the problem (initially 
reported by Engineering) of Windows 3.0 not allowing the use of a CGA 
display and 386 enhanced mode together.  We tried this and were unable 
to reproduce the problem.



ENHANCEMENT DESCRIPTION
=======================

Under Release 2.0 (or better) of the Amiga System Software, ARead should 
support 2.0-style wildcards (using the dos.library ParsePattern, ExAll, 
MatchFirst, MatchNext, and/or MatchPattern functions).



FIXED
=====

The idea of a 4-color (only) Workbench screen is hard-coded into PCWindow.

If you switch Flipper to manual mode, you cannot return to automatic mode 
unless there is a disk in the shared drive.

Flipper does not correctly indicate the MS-DOS drive label of the shared 
drive.  For example, if you have the shared drive configured as B: for the 
MS-DOS side, the Flipper window still shows it as A:.  Flipper has not been
tested with a drive other than DF0: shared.

Selecting the "RAM=A000" option in PCPrefs and selecting the "Save" or "Use" 
option in PCPrefs prevents AutoBoot hard disk drives from working.  Attempting 
to access the AutoBoot hard disk drive cause the PC to lock-up and the Amiga 
to run very slow (as if Forbid() and/or Disable() are being repeatedly called 
for long periods of time).  drysdale mentioned that other "bad things" happen 
when this option is selected.



PARTIALLY FIXED
===============

Since many of our users are not familiar with text editors, modify C:MakeAB 
to automatically update SYS:PC/System/aboot.ctrl.  In addition, a C:RemoveAB 
or C:DeleteAB command would remove the need for the user to manually delete 
the AutoBoot drive file and edit the SYS:PC/System/aboot.ctrl file.

MakeAB does not supply a template and prompt for parameters if the "?" 
parameter is given.  This is in violation of the Amiga user interface 
guidelines.

From mikec Mon Aug 19 13:16:20 1991
Received: by cbmvax.cbm.commodore.com (5.57/UUCP-Project/Commodore 2/8/91)
	id AA20223; Mon, 19 Aug 91 13:16:17 EDT
Date: Mon, 19 Aug 91 13:16:17 EDT
From: mikec (Michael Colligon - Product Assurance)
Message-Id: <9108191716.AA20223@cbmvax.cbm.commodore.com>
To: barker, drysdale, havemose, ian, jss, mario, ned
Subject: Meeting
Cc: givler, jjszucs, mikec, rob
Status: RO


Disregard previous references to a meeting.  I will have to find a new time.

Mike

From mikec Mon Aug 19 13:38:43 1991
Received: by cbmvax.cbm.commodore.com (5.57/UUCP-Project/Commodore 2/8/91)
	id AA21512; Mon, 19 Aug 91 13:38:39 EDT
Date: Mon, 19 Aug 91 13:38:39 EDT
From: mikec (Michael Colligon - Product Assurance)
Message-Id: <9108191738.AA21512@cbmvax.cbm.commodore.com>
To: barker, drysdale, havemose, ian, ned
Subject: A2386 and Janus meeting
Cc: mikec, phil
Status: RO


Let's try to have the meeting tomorrow at 1:00 PM in the PA conference room.
We'll start with the A2386 and move on to the Janus software.  I would still
like everyone to be at both parts to express their opinions.  If you can't
make it please let me know and we can try to reschedule if necessary.

Mike

