>Does anyone know of a bug or problem in the irq-driven parallel
>driver? The polling version seems to work, even though it complains
>about numerous things, such as "printer offline" or "out of paper". I
>have even tried a different printer and a different cable, all with
>the same result.

After messing with the interrupt-drive parallel port driver for a while last
year, I am convinced that it is generally a lost cause.  While I didn't have
the problem you were having, I did have some serious data corruption problems.

After looking around for a while, I discovered the following things:

- Nearly all other popular operating systems (MS-DOS, NT, etc etc), use
  a polling driver.

- The same sort of data corruption bugs exist in the Linux and FreeBSD
  parallel port drivers (Linux you can tune some delay parameters, however).

- Some parallel port manufacturers don't make the IRQ code work, because
  all of the commercial OS's use polling.

So personally, I'd just use the polling driver :-)  Is there a reason you
really want to get the interrupt driven one working?

(can't help you on the messages you get from the polling driver, however,
but I suspect your hardware is slightly out-of-spec.  You could always
comment those messages out in the driver source).

--Ken