Skew vs. interleave

[Mike Stein]

Same here.

I've spent many happy hours in 'the good old days' adjusting 'interleave'
of ST512/406 MFM hard disks to find the optimum setting for a particular
system/controller but had never even heard the term 'skew' until 5 or 6
years ago while playing with odd format diskettes, and then it was in the
track offset sense.

So I was surprised that some folks in the S100/CPM world use 'skew' in the
interleave sense, apparently because the CP/M documentation used it that
way. I'm always surprised how a field so dependent on rigid logical
concepts and definitions has so many inconsistencies.

Thanks everyone!

m

On Fri, Jul 30, 2021 at 8:47 PM Paul Koning <paulkoning at comcast.net> wrote:

>
>
> > On Jul 30, 2021, at 5:35 PM, Mike Stein via cctalk <
> cctalk at classiccmp.org> wrote:
> >
> > There's a small discussion on S100computers about the terms 'skew' and
> > 'interleave'.
> >
> > In CP/M documentation 'skew' refers to what's usually called interleave
> > these days, i.e. offsetting sectors on a track to compensate for the fact
> > that by the time the computer has processed a given sector the next one
> has
> > already passed by, so that the computer has to wait an entire revolution
> > for it to pass by the head again; in other documentation as in Chuck's
> > 22disk for example this is also called 'interleave'.
> >
> > However, in later documentation the meaning of 'skew' seems to have
> changed
> > to refer to the offset of sectors between adjacent tracks to compensate
> for
> > the time required to step the head.
>
> I've only ever seen the term "skew" with that second meaning.  The first
> thing you mentioned in my experience is always called "interleave".  For
> example, the DEC RX50 has 2:1 interleave and 3 sector skew.
>
> Interleave is normally written as the physical sector number difference of
> two logically adjacent sectors (so 2:1 means there is one other sector
> between logical sector 0 and logical sector 1).  In one place (David
> Gesswein's MFM emulator) I've seen it used the other way around, n:1
> meaning that logical sector n is physically immediately after logical
> sector 0.
>
>         paul
>
>
>