Subject: From: Date:

Re: [ECS] ECSW & Win98 LEC1964.aol.com Sun, 9 Aug 1998 16:03:25 EDT

---

       I have been running ECSW under Windows 98 for the past five or six
weeks with generally no problems attributable to the 98 version. My opinion is
that Windows 98 is Windows 95 "Plus". Nothing much different to speak of, just
a few more bells and whistles.
       I elected to use Windows 98 because of Microsoft's moronic decision to
restrict access to Windows 95 OSR2. For those of you that don't know, the
retail version of Windows 95 does not support "Ultra-DMA" and the IDE Bus
Master features along with a number of fixes and revisions. The OSR2 version
was available if you purchased a new computer recently or supposedly a new
motherboard or hard drive or if you begged some one to sell it to you. Very
irritating since I had upgraded by purchasing a motherboard with the Intel
chipset that supports these features. I did try Intels separate drivers. That
was a riot. With Windows 98 I believe all of the OSR2 features are available
and everything seems to work as advertised.
       As far as FAT32 goes, personally, I see no reason to use it. With the
costs of very large hard drives so reasonable, if I need more disk space, I'll
buy a new one. I am not sure that the authorization would be seen as valid
code during a FAT32 conversion so to be on the safe side I would not leave it
there. There is also a very great risk it will not work at all under FAT32. I
would just be very careful. As it applies to me, I try to use the "KISS"
principle as often as possible just to keep me out of trouble.
       A caveat here, (No disrespect Mark G.) but I don't trust upgrades. I
don't care to have manipulation of my carefully installed software by anyone
and there has to be a very good reason to accept the risks that are involved.
There also is a very good reason your always told to backup your hard drive
during these episodes. So I did not "upgrade" to Windows 98. It was loaded on
to a bare drive followed by ECS and Microsoft's SDK30.
       I did have some difficulty getting IRQ assignments to my liking. If
your BIOS is PnP ready you do have some control over IRQ assignments. Once I
sorted that out, they fell in exactly as I wanted.
       For those of you that might be interested, I have for many years run
ECS from a RAM drive. With ECS 3.2 it was very easy. Of course a RAM drive of
adaquate size is created and all necessary files transferred to appropriate
directories. A UPS is an necessity. Comspec should define the location for
command.com. The hard drive's 12v and 5v potentials were switched off manually
leaving ECS entirely running from the fastest memory available (RAM). I also
felt with the hard drive "off line" there was a special protection for the
authorization and the ecs.cfg. I could do editing with the RAM drive version
and testing prior to making it permanent on the hard drive side which was kind
of neat. Some would say that turning your hard drive on and off  (thermal
shock) is more damaging than leaving it on all the time. To me ECS is a
"working" system and once I'm satisfied I don't make a lot of changes. There
was one period of 14 months where the hard drive was not turned on at all. One
other nice feature is that I don't have an "F-16 at at mach 2" as a permanent
visitor in my den. The high pitch whine of a hard drive is very aggrivating. I
also have purchased power supplies designed to be ultra quiet (large cooling
fans turning less RPM). Some visitors actually have to ask if the computer is
on. To say the least this definitely pleased my wife.
       I am currently running ECSW (under Windows 98) from a 30 mb RAM drive
using the BIOS (not Windows) to "depower" the hard drive after 3 minutes of
idle time. Although testing is not complete, everything so far looks very
good. Only ECS has been relocated to the RAM drive. I start ECS using the run
command ,obviously not the ECS ICON. I have picked up some additional cooling
fan noise since Intel can't design a processor for a desktop without putting a
fan on it. As far as how often hard drive access is required I really don't
know. It's not often. Once the system settles down I don't detect it at all
unless I minimize ECS and use Windows.
       Since the subject came up a while back about performance, specifically
event processing, I might mention that with my 2700 event lines in 67 events I
run generally 13 event passes/sec and communicating with the Micro REDAC is
basically the defining performance factor. I really don't know if 13
passes/sec is  good or bad, but I am running much faster than my old system.
One thing I did do years ago was to add testing at the beginning of most
events. The few added event lines test to see if I really need to go down
through the entire code in that event. The test lines end with an entry: "Else
Event Exit  Set  True", which I assume would limit most of my processing to
essential events and to others on demand.
       I would like to end by posing a few questions. (1.) If I elected to use
the USB (Universal Serial Bus) say for the keyboard, the mouse and possibly
for the printer would their original IRQ's be available for reassignment? In
my case that would be 1, 12, and 7. I sure could use them. What would be their
status? (2.) In the past I was able to "share" IRQ 7. Windows 98 (device
manager) had a real problem with this. Is it because one of the newer
"enhanced" parallel port was defined? My printer and most that I know of don't
use interrupts anyway. In fact I recently set up a friends older computer
under Windows 98 and Windows didn't assign any interrupts to lpt1 and his
printer works fine. (3.) Has anyone used the new dual    comport card that
works off the PCI bus? Does Windows Device Manager accept shared interrupts
for Comports on the PCI bus?

Larry C.

---