Reported by briang1 on 2010-12-06 21:18
On one machine out of three all running the same version of XP etc, the snaps 970 will not allow sapi 4 to work, giving an error. If one reinstalls snap 946 the problem goes away, so it seems not to be corruption in the machine or anything like that.
I present the log fragmement below for a recent snap.

INFO - nvda (12:13:11):

Starting NVDA

INFO - core.main (12:13:11):

Config dir: C:\DOCUME1\brian\APPLIC1\nvda

DEBUG - core.main (12:13:11):

setting language to Windows

INFO - core.main (12:13:11):

NVDA version main-3971

INFO - core.main (12:13:11):

Using Windows version sys.getwindowsversion(major=5, minor=1, build=2600,
platform=2, service_pack='Service Pack 3')

INFO - core.main (12:13:11):

Using Python version 2.7 (r27:82525, Jul 4 2010, 09:01:59) v.1500 32
bit (Intel)

INFO - core.main (12:13:11):

Using comtypes version 0.6.2

DEBUG - core.main (12:13:11):

Creating wx application instance

DEBUGWARNING - inputCore.InputManager.loadLocaleGestureMap (12:13:14):

No locale gesture map for language en

DEBUGWARNING - inputCore.InputManager.loadUserGestureMap (12:13:14):

No user gesture map

.... So what is this all about by the way?...

DEBUG - core.main (12:13:14):

Speech Dictionary processing

DEBUG - speechDictHandler.SpeechDict.load (12:13:14):

Loading speech dictionary
'C:\DOCUME1\brian\APPLIC1\nvda\speechDicts\default.dic'...

DEBUG - speechDictHandler.SpeechDict.load (12:13:14):

file 'C:\DOCUME1\brian\APPLIC1\nvda\speechDicts\default.dic' not found.

DEBUG - speechDictHandler.SpeechDict.load (12:13:14):

Loading speech dictionary 'builtin.dic'...

DEBUG - speechDictHandler.SpeechDict.load (12:13:14):

4 loaded records.

DEBUG - core.main (12:13:14):

Initializing speech

WARNING - synthDriverHandler.SynthDriver.loadSettings (12:13:16):

No such voice: {8454410E-002C-0200-BC4F-5B1800D50059}

ERROR - synthDriverHandler.setSynth (12:13:16):

setSynth

Traceback (most recent call last):

File "synthDriverHandler.pyc", line 78, in setSynth

File "synthDriverHandler.pyc", line 416, in loadSettings

File "synthDriverHandler.pyc", line 31, in changeVoice

File "synthSettingsRing.pyc", line 73, in init

File "synthSettingsRing.pyc", line 124, in updateSupportedSettings

File "synthSettingsRing.pyc", line 40, in init

File "baseObject.pyc", line 21, in get

File "synthDriverHandler.pyc", line 295, in _get_availableVoices

File "synthDrivers\sapi4.pyc", line 213, in _getAvailableVoices

KeyError: 0

INFO - synthDrivers.espeak.SynthDriver.init (12:13:17):
This occured on running the installer which is why it was already set to sapi 4 on the machine. tthe voice was in fact an Orpheus one called Dave.

There was an associated error generated in Windows, though I'm not sure if this was a confused installer or the error, but here it is anyway.

Speaking u'nvda.exe - Application Error icon 1 of 5'

IO - speech.speakText (12:19:18):

Speaking u'nvda.exe - Application Error dialog The instruction at
"0x09577c42" referenced memory at "0x00000004". The memory could not be
"read".\r\n\nClick on OK to terminate the program'

IO - speech.speakText (12:19:18):

Speaking u'OK button'

All other machines have no problems, so I'm at a loss, but something has changed in the snapshots.