Xref: utzoo comp.sys.att:8118 unix-pc.general:4182
Path: utzoo!attcan!uunet!lll-winken!scooter!neoucom!wtm
From: wtm@neoucom.UUCP (Bill Mayhew)
Newsgroups: comp.sys.att,unix-pc.general
Subject: Re: hard disks for 3b1
Summary: Abiguous terminology of track and cylinder
Message-ID: <1840@neoucom.UUCP>
Date: 24 Nov 89 14:39:54 GMT
References: <1145@becker.UUCP> <11654@cbnews.ATT.COM>
Organization: Northeastern Ohio Universities College of Medicine
Lines: 21


I've run into the abiguous meaning of the term, "track," enough
that I find it rather aggrivating.  In the IBM PC world, "track"
often means the ring of data on one disk surface, while "head"
describes the device used to gain access to the data on different
disk surfaces.

In the Unix PC world (and seemingly most other places), tracks are
the disk surfaces, while cylinders are the concentric rings of data
on the tracks.  I think there are a few places where the Unix PC
manuals slip up and refer to cylinders as tracks.  And then there
are the 512 byte physical versus 1024 byte logical disk blocks just
to kee things fun too.  Some commands use logical blocks, while
others use physical blocks; the manual doesn't always tell which,
so guessing and experimentation is reuqired.

Where this mess gets confusing is where you run into smart disk
controllers that do virtual mapping so that the number of tracks
and cylinders is approximately equal; some ESDI controllers, for
instance.  The Adaptec controller that IBM uses in the model 80
comes to mind.