Re: hardware etc

["Arlen P. Walker" <arlen -dot- walker -at- JCI -dot- COM>]

Chuck, you're almost right:

>        EISA vs ISA vs VESA -- These are designations for IBM-clone
>bus structures.  ISA is the old tried and true IBM PC/AT bus.  EISA
>is a newer, purportedly faster, bus structure championed by
>Hewlett-Packard and others.  VESA is the latest, and fastest, bus
>found in the newer "local bus" machines.

You imply here a relationship which isn't present. EISA slots will accept
ISA cards, and both were intended for all types of PC expansion cards (and
the EISA bus *is* faster; purportedly carries a connotation which implies
it's not really so). VESA is a bus specification designed for video cards
only, although other card makers have tried to make it fit their purposes.
ISA cards will not work in VESA slots.

>        Flash BIOS -- I'm not sure.  If I'm right, it refers to
>BIOS in FROM (Flash ROM) which gives you "instant-on" BIOS.  It
>just means the PC starts running sooner after power up or reboot.

Flat-out wrong, I'm afraid. "Flash" memory refers to technology devloped by
Intel which allows ROM to be changed by programs knowing the appropriate
key. Under normal circumstances it behaves exactly like ROM, but it makes
it possible for the computer company to send you a disc to update your ROM
BIOS without making you replace the chips (or take it to a service station
for them to do the work). All ROM is "instant-on" in that it's available
for use as soon as power is applied to it. If a machine with Flash BIOS
"starts running sooner" (by this I think you meant starts loading DOS
sooner) it simply means the manufacturer has modified the power-on self
test that all machines perform on startup, perhaps even bypassing it
completely.

>        Whenever I'm shopping for a new hardware or software item,
>I refer to the various personal computer magazine buyer's guides
>and performance comparisons.  In the IBM-clone magazines, most
>comparisons are made against benchmark programs and MS/Windows
>performance.  MS/Windows and applications that run on it are
>enormous memory and CPU cycle hogs.  Multitasking under MS/Windows
>multiplies the demands on the PC.  The PC magazines offer the
>best guidance I've found on what to purchase.

And many hardware manufacturers "cheat" these benchmarks. Under the
pressure to achieve better WinMark performance, some companies have been so
blatant as to put special code in the video ROM which checks for particular
bechmarks programs and shifts into a mode designed to run that benchmark at
blazing speed. Others have written caching algorithms whose sole purpose is
to run a particular benchmark faster. In neither case do the real-world
speeds measure up to the benchmarks.

I can't say this often enough or strong enough: DO NOT USE MAGAZINE
PERFORMANCE BENCHMARK PROGRAMS AS AN EVALUATOR OF SYSTEM PERFORMANCE!!! All
benchmark programs can be fooled. Instead trust performance benchmarks
which are based on real-world applications (how fast did Word scroll
through this document, how fast did Pagemaker print this document, etc.).
These are the best indicators of system performance, and the only
meaningful basis for system comparison. If you can't find a review which
uses this kind of benchmarking, find someone wanting to sell the product
who'll let you test it yourself.

Have Fun,
Arlen

arlen -dot- walker -at- jci -dot- com
---------------------------------------------------
This mail message contains 100% recycled electrons
---------------------------------------------------