-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "Rob" == Rob  <dickrp@wckn.dorm.clarkson.edu> writes:

Rob> I posted a few days ago about diffficulty printing when other
Rob> tasks were vying for processor time.  If I print on a clean
Rob> system, I get >4 pages per minute.  When I try to print the same
Rob> file with lpd at `nice = -20' and some other demanding task at
Rob> `nice = +20' it takes 4 minutes or more to print one page.

	[discussion of scheduling quantum deleted]

	I think that you'll find that this is more due to how the
parallel port driver works than the scheduling quantum (esp. by
evidence you present).  I don't have my Amiga online right now, but
the parallel port on the Amiga is a fairly dumb one, requiring a high
interupt load to service.

	This is one area where NetBSD is far slower than AmigaOS ---
interupt handling.  The AmigaOS can handle many thousands of interupts
per second (when I was driving my own printer under AmigaOS, the
system itself would get jerky, the interupts were going off so much).
NetBSD takes much longer to service an interupt.  In fact, it seems
that BSD functions much better on hardware with lots of buffers.

Dave.

- -- 
============================================================================
|David Gilbert, PCI, Richmond Hill, Ontario.  | Two things can only be     |
|Mail:      dgilbert@jaywon.pci.on.ca         |  equal if and only if they |
|http://www.pci.on.ca/~dgilbert               |   are precisely opposite.  |
=========================================================GLO================

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv
Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface

iQCVAwUBMqBvJofIfCnlwWwVAQEfbgQAoWOSJDUZd0n07qRrkXOJ6J1d0NtR+0kq
yJZt6OdcHkIB39k89POHsSp9EGNByH5GokMA8HHnZZoIIyf1TqSQh6eOIVLgv+JB
MV14IgPurKVtjbuIkYMSPz9cdNKuTmf5I2qooyzt+jpMnLf2fXzW4saQeTOlsmMC
cDUfRp3ZGDA=
=GRqR
-----END PGP SIGNATURE-----