Replies

Absolutely... All the work on increasing memory and CPU speeds were pointless because buffers and wait states had to be used to slow them back down due to slow bus speeds. Working on bus speed should have been the #1 priority.

Then to top it off we had MS doing their best to steal back any speed advantages with BS unneeded bloat ware. An older guru friend who was on the IBM program from the beginning was blowing his top about this each step of the way... Lol

Working on bus speed should have been the #1 priority.

Software, especially OS software was usually the biggest bottleneck. Oh, Windows MS support MCA well enough, and the local buses when they came out (usually), because the graphics and other other card companies wrote the drivers. DOS was so low level that each app needed its own drivers for a lot of the serious stuff (e.g. accessing special graphics and text modes, printer drivers).

Remember when all the new PCs had USB built in and neither Windows nor DOS could use them, so the peripheral companies didn't make USB interfaces. Two-way parallel (e.g. Zip Drives) or serial mainly for modems were good enough.

Apple did the world a favor when they released the USB 1.1-centric original iMac, dropping their own SCSI, RS-422 serial and ADB implementations, jump-starting the USB peripherals market and shaming Microsoft into putting real USB support in Windows 98se.

Even as late as 2000, there were VERY few PCI cards for modems, so that 56k modem either went external or on the obsolete ISA bus.