From cahudson@gate.net  Mon Dec 19 18:12:24 1994
Return-Path: <cahudson@gate.net>
Received: from tequesta.gate.net (root@tequesta.gate.net [198.80.2.38])
          by mailhost.WI.LeidenUniv.NL (8.6.9/FHM-1.20-M) with ESMTP id SAA09780
          for <aria@wi.leidenuniv.nl>;  Mon, 19 Dec 1994 18:12:22 +0100 (MET)
Received: from Chris      .gate.net (orfl-7.gate.net [199.227.3.7]) by tequesta.gate.net (8.6.9/8.6.9) with SMTP id MAA28506 for <aria@wi.leidenuniv.nl>; Mon, 19 Dec 1994 12:13:15 -0500
Message-Id: <199412191713.MAA28506@tequesta.gate.net>
X-Sender: cahudson@mailhost.gate.net (Unverified)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Mon, 19 Dec 1994 13:16:28 -0500
To: aria@WI.LeidenUniv.NL
From: cahudson@gate.net (C. A. Hudson)
Subject: Re: Aria LISTENER Alt?
X-Mailer: <Windows Eudora Version 1.4.2b16>

Daniel,
Thank you for your very detailed response to my Aria LISTENER questions.

With no resolution yet on my GPF's, and my Mom getting further frustrated by
not being able to call for a Nurse, I had to suspend my attempts to get
LISTENER to work unattended, at least for now. In order to get something
going, I ended up with a Sound Blaster and its Voice Assist
speaker-dependent PC-based recognition software. While this is apparently
robust and has not GPF'ed at all, it does appear not as accurate as
LISTENER, and much more sensitive to ambient noise, and the level of the
spoken commands. Its also "half duplex", compared to LISTENER which didn't
occupy the PC's WAVE channel as VA does. But it works.

Let me try to respond to the questions you posed on how I was using LISTENER:

On the GPF's, yes, I only started to get them after I had installled my
work-around for "pushing" the cancel button when the "ambiguous" window
appears. Without my workaround, I would cancel the screen manually, perhaps
two or three times, and then give up and kill the whole thing when I
couldn't get rid of it appearing.

My workaround uses a Windows scripting language called WINBATCH from Wilson
WindowWare. I got it off Compuserve - the file there is WB5Q-N.EXE. I do not
know if they have an IP address or FTP, if you wish to give me a location to
ftp it to you, I would be glad to do so. WINBATCH looks to me to be a mature
product. It allows you to sort of do in Windows the things you would do in a
DOS .BAT file - only more. It allows you to check for the existence of a
named Window, and pass it "keystrokes". I used this to check for the
"ambiguous" LISTENER Window every 10sec., and then kill it with a "c". After
anywhere from perhaps 10 to 30 cycles of this, LISTENER would GPF. Here is a
partial DRWATSON printout - they are all the same:

Dr. Watson 0.80 Failure Report - Wed Dec 14 18:53:45 1994
LISTENER had a 'Exceed Segment Bounds (Read)' fault at KERNEL 1:8ee8
$tag$LISTENER$Exceed Segment Bounds (Read)$KERNEL 1:8ee8$rep     movsw$Wed Dec
14 18:53:45 1994
$param$, Last param error was: Invalid handle passed to KERNEL 1:05c9: 0x68a2

CPU Registers (regs)
ax=73d8  bx=73c0  cx=4680  dx=0088  si=ffff  di=73ae
ip=8ee8  sp=6010  bp=603e  O+ D+ I+ S- Z- A+ P- C+
cs = 0117      d440:a95f Code Ex/R
ss = 148f     94260:8b7f Data R/W
ds = 148f     94260:8b7f Data R/W
es = 148f     94260:8b7f Data R/W

CPU 32 bit Registers (32bit)
eax = 000073d8  ebx = 000073c0  ecx = 00004680  edx = 00000088
esi = 0000ffff  edi = 000073ae  ebp = 0000603e  esp = 80016000
fs = 0000         0:0000 Null Ptr
gs = 0000         0:0000 Null Ptr
eflag = 00000602

System Info (info)
Windows version 3.10
Retail build
Windows Build 3.11
Username C. A. Hudson
Organization
System Free Space 5531744
Stack base 15836, top 25686, lowest 22858, size 9850
System resources:  USER: 90% free, seg 07af  GDI: 83% free, seg 0607
LargestFree 9678848, MaxPagesAvail 2363, MaxPagesLockable 717
TotalLinear 2940, TotalUnlockedPages 724, FreePages 331
TotalPages 973, FreeLinearSpace 2384, SwapFilePages 682
Page Size 4096
4 tasks executing.
WinFlags -
  Math coprocessor
  80386 or 80386 SX
  Enhanced mode
  Protect mode

Stack Dump (stack)
Stack Frame 0 is KERNEL 1:8ee8                ss:bp 148f:603e
0117:8ee3  d1 e9                 shr	cx, 1
0117:8ee5  1e                    push	ds
0117:8ee6  07                    pop	es
0117:8ee7  fd                    std
(KERNEL:1:8ee8)
0117:8ee8  f3 a5                 rep	movsw
0117:8eea  fc                    cld
0117:8eeb  46                    inc	si
0117:8eec  46                    inc	si

Stack Frame 1 is LISTENER 1:0995              ss:bp 148f:605a
1497:0986  c7 46 fa 0000         mov	word ptr [bp+fa], 0000
1497:098b  6a 42                 push	42
1497:098d  ff 76 f8              push	word ptr [bp+f8]
1497:0990  9a 0543 0117          callf	0117:0543
(LISTENER:1:0995)
1497:0995  89 46 fc              mov	[bp+fc], ax
1497:0998  83 7e fc 00           cmp	word ptr [bp+fc], 00
1497:099c  74 03                 jz	short 09a1
1497:099e  e9 0045               jmp	near 09e6

Stack Frame 2 is LISTENER 1:0243              ss:bp 148f:606a
Stack Frame 3 is XWI320 42:0ecb               ss:bp 148f:607a
Stack Frame 4 is XWI320 9:0132                ss:bp 148f:608e
Stack Frame 5 is XWI320 9:02cf                ss:bp 148f:60b0
Stack Frame 6 is XWI320 9:16b6                ss:bp 148f:6122
Stack Frame 7 is XWI320 9:1896                ss:bp 148f:6132
Stack Frame 8 is XWI320 10:0034               ss:bp 148f:6142
Stack Frame 9 is XWI320 22:01ad               ss:bp 148f:616c
Stack Frame 10 is XWI320 55:012e               ss:bp 148f:6192
Stack Frame 11 is LISTENER 1:8005              ss:bp 148f:61a8
Stack Frame 12 is LISTENER 1:7ebf              ss:bp 148f:61c8
Stack Frame 13 is LISTENER 1:005e              ss:bp 148f:61de
Stack Frame 14 is XWI320 13:00bc               ss:bp 148f:620c
Stack Frame 15 is XWI320 63:0bc5               ss:bp 148f:6226
Stack Frame 16 is XWI320 38:10d4               ss:bp 148f:625c
Stack Frame 17 is XWI320 24:0b6b               ss:bp 148f:626e
Stack Frame 18 is XWI320 24:11d8               ss:bp 148f:6284
Stack Frame 19 is USER 1:27bb                  ss:bp 148f:629e
Stack Frame 20 is XWI320 42:0c46               ss:bp 148f:62e6
Stack Frame 21 is XWI320 42:0e71               ss:bp 148f:62fe
Stack Frame 22 is LISTENER 1:0502              ss:bp 148f:6328
Stack Frame 23 is LISTENER 1:5795              ss:bp 148f:634a
Stack Frame 24 is LISTENER 1:0475              ss:bp 148f:6430
Stack Frame 25 is LISTENER 1:2d43              ss:bp 148f:6442
Stack Frame 26 is LISTENER 1:2d10              ss:bp 148f:6454
System Tasks (tasks)
Task DRWATSON, Handle 113f, Flags 0001, Info   26864 03-10-92  3:10
  FileName C:\WINDOWS\DRWATSON.EXE
Task  PROGMAN, Handle 063f, Flags 0001, Info  115312 11-01-93  3:11
  FileName C:\WINDOWS\PROGMAN.EXE
Task  WBAT16I, Handle 10e7, Flags 0001, Info   15200 10-11-94 22:04
  FileName C:\WINBATCH\WBAT16I.EXE
Task LISTENER, Handle 14b7, Flags 0001, Info   69776 10-28-93 13:56
  FileName C:\ARIA\LIST\BIN\VOCMACRO.EXE

Stop Dr. Watson 0.80 - Wed Dec 14 18:54:30 1994



On your question on adding info on LISTENER to the docs;, to me LISTENER is
why I bought the Aria 16se. In my view there are better "playback" cards
(GUS), and certainly better SB emulators. The promise of the Aria was its
DSP programmability, and on-board speech recognition potential. Too bad that
for me this remains right now a promise rather than a reality.

I don't expect one board to be best at everything - its more likely to be
"Jack of all trades - master of none". To me it makes more sense to use a
real cheap SB when needed for games, and skip the hassels of flakey
emulation. I finally did this with my GUS when I got fed up trying to make
its very poor emulation work. Likewise with the Aria, its potential is where
it is relatively unique - on-board DSP-based speech recognition. At least
this is my view. A long answer supporting your adding a section on LISTENER
to the docs.

Finally, thanks for your advice on SSDK and LSDK. Like wise because of time
availability, and my "working" SB/Voice Assist alternate, I'm going to defer
getting further into LISTENER for now. I'll advise the group when and if I
pick it up again.

Thanks again for your help.
Chris

