Driver:	TWFUTCAM, Version 4.10.00 January 30, 1993
	Copyright (c) 1990-1993, Emeritus Technologies.
	All rights reserved.

	TapeWare Future CAM Device Driver Module

	Date		Mod	Description
	-----------	---	-----------------------------------------------
	05-Aug-1992	RAC	Copied from TWASPI and started conversion to
				Future Domains Future CAM specification

	06-Aug-1992	RAC	Due to the fact Future CAM does not support
				I/O from above 1Mb like ASPI, changed all
				physical pointer passing to virtual ptrs and
				removed all *_AT flags from the CCB. Also, had
				to add TapeWares Memory strategy handler to
				inform the kernel not to use extended or XMS
				memory, only standard.

	24-Aug-1992	RAC	Under DOS, Future/CAM is apparently not saving
				the registers as indicated in the Future/CAM
				documentation. The DS segment registers was
				getting fouled up and pointing into the 
				TapeWare data segment. Also, in the callback
				function SignalCallback, added a CLD 
				instruction to make sure the direction flag
				on the CPU is cleared.

	03-Sep-1992	RAC	Corrected problem in GetControllerInformation
				using sprintf to return the status info. This
				is not legal under DOS.

	05-Sep-1992	RAC	Future/CAM for DOS is not allowing a 
				disconnect to occur without a Save Data Ptrs 
				message. It gives a "Unexpected Reselect"
				error. ANSI spec specifically states that it
				is legal for a device to disconnect WITHOUT
				sending a prior Save Data Ptrs message if
				no data has been sent or received. 
				Unfortunately, ADIC followed the ANSI spec
				and does not work correctly with CAM since
				it does not send a Save Data Ptrs message
				prior to loading an element.

	06-Sep-1992	RAC	Removed DeviceTimeout feature. This was 
				causing serious problems with Wangtek QIC 
				drives and the ADIC autoloader. It seems that
				Future/CAM for DOS simply ignores the timeout
				value that we send, so it has to be disabled..
				Yet another problem to report to FutDom...

	06-Sep-1992	RAC	We have a problem here... We cannot enable
				disconnects for the reason previously stated
				on Sept 5. Apparently, if we do not allow
				disconnects, Future/CAM for DOS will 
				automatically timeout after 5 seconds 
				regardless of what the sc_timeout field is
				set to. In order to support the majority of
				devices, doing normal things like reten or
				format, which take longer than 5 seconds to
				complete, we must re-enable disconnect.
				This will not work with some devices, but
				unfortunately, theres not a thing we can
				do about it except yell very loudly at
				Future Domain... The good news is that 
				Future/CAM for NetWare/386 seems to be working
				just fine without any of the messing around
				required for DOS.

	30-Jan-1993	RAC	Removed processing of DEVLOAD.

