17 Dec
2001
17 Dec
'01
10:33 p.m.
From: Fred Cisin (XenoSoft) <cisin(a)xenosoft.com>
There ARE many things that the WD will do that the NEC
won't, including
writing much sooner after the index pulse, IGNORING certain fields in
the
765A writes or read sonner than base 765 and the 37c65 even shorter.
It doesnt ignore fields in a multisector read/write however if the
sectors
are written with a interleave it will also keep things in numerical
order.
Add to that a full cylinder read does not bring all the non data crap
that formatted media requires to marks all the data spaces. The real
beauty
of that is if you have real DMA you can fire it up and read a whole
cylinder
and all your buffer has is neatly ordered data if the read was
successful.
That latter feature is nice if your doing a caching scheme.
Allison could probably gives us a more comprehensive
and authoritative
list.
Yes I can.
Allison
18 Dec
18 Dec
12:35 p.m.
From: Fred Cisin (XenoSoft)
<cisin(a)xenosoft.com>
>There ARE many things that the WD will do that the NEC won't, including
>writing much sooner after the index pulse, IGNORING certain fields in
On Mon, 17 Dec 2001, ajp166 wrote:
765A writes or read sonner than base 765 and the 37c65
even shorter.
Ah, Allison, where have you been all of my life?
If you would have told me that 15+ years ago, you would have saved me LOTS
of time and effort putting tape over index holes, making cables with
switches to interrupt the index signal, etc.
It doesnt ignore fields in a multisector read/write
however if the
sectors
are written with a interleave it will also keep things in numerical
order.
With a 765, if I want to read a sector from the second side of most Kaypro
disks, I need to feed it a value that matches the WRONG value that's in
the sector header for the H field. With WD, that field can be ignored.
Add to that a full cylinder read does not bring all
the non data crap
that formatted media requires to marks all the data spaces. The real
beauty
of that is if you have real DMA you can fire it up and read a whole
cylinder
and all your buffer has is neatly ordered data if the read was
successful.
That latter feature is nice if your doing a caching scheme.
Raw v formatted track read has advantges and disadvantages. The 765
approach is very handy for reading a bunch of "normal" sectors. The WD
approach makes it possible to read MFM that does not follow
"normal" sector header standards, such as Amiga, or some TRS-80 address
marks.
Personally, I would have preferred the "raw" approach of the WD, but I
make no claims to be representative of the marketplace.
Allison could
probably gives us a more comprehensive and authoritative
list.
Yes I can.
8424
days inactive
8425
days old
1 comments
2 participants
participants (2)
-
ajp166＠bellatlantic.net -
cisin＠xenosoft.com