CompuServe Thread

Forum unknown · Programming

#WB1.2

11 messages in this thread

#45888From: Dave Haynie/C128 LandDec 28, 1986 1:41 PM

Maybe the driver itself. But what if a runaway program trashes a floppy's track buffer just before it gets written out. Its already assumed to be in MFM format; its just sitting there, in the 30K or so track buffer, waiting for a Paula DMA to dump it out to disk. With a few floppies, you've got a pretty big target for a runaway program to hit. But yea, it would be really weird for anything to successfully change the driver program to become destructive — kind of like the chance of a lifeform living successfully with a major mutation. -Dave

#45894From: Larry Phillips/ICUGDec 28, 1986 1:53 PM

Ahh… never thought about the track buffers themselves. That's 5K per drive, average system being a 10K target. Now that can definitely be a problem. Now where did I put that MMU?

#45898From: Brian NiessenDec 28, 1986 2:16 PM

Larry, aren't track buffers 11K/track for 3.5 drives? (That is the full track buffer) plus the actual data buffers.

#45956From: Larry Phillips/ICUGDec 28, 1986 7:46 PM

I guess they are Brian, though if so, they should properly be called "cylinder buffers", as there are two tracks per cylinder. If they are true track buffers, then they will be just over 5K. Anyone know for sure?

#46040From: Dave Haynie/C128 LandDec 29, 1986 3:48 AM

The trackdisk buffers are 512 x 11 = 5632 bytes in length.

#46042From: Larry Phillips/ICUGDec 29, 1986 4:01 AM

Thanks Rick. So they are true "track" buffers and not "cylinder" buffers.