plip0
when using
lpt0
,
as
plip1
when using
lpt1
and
plip2
when using
lpt2
.
A parallel port can not drive a plip interface when configured without an irq. The port can be used to drive other devices when the corresponding plip interface is configured down.
The plip interface simulates Ethernet on a parallel cable with special wiring (see below), known as a parallel LapLink cable widely available from PC shops.
Its acronym is derived from SLIP and means Parallel IP Protocol, which, in fact, is wrong: it can handle any protocol available on Ethernet, not just IP.
Although a plip connection always is point-to-point, the interface is configured like every standard Ethernet interface and uses ARP to find its neighbor. This is inefficient, but provides interoperability with other operating systems.
Additional grounds are 18, 19, 20, 21, 22, 23 and 24. Do not connect D5(7), D6(8), D7(9), STROBE(1) and FEED(14).
ifconfig(8)
,
pppd(8)
,
slip(8)
Martin Husemann ported it to the NetBSD/i386.
The protocol is too much overhead, a bidirectional input/output routine switchable to a well-designed network line discipline with on-line compression would be far superior.
There are some parallel ports which can use all 8 data bits for input and output, you can make up a parallel handshake protocol and a cable with different wiring to get an 8 bit clean (instead of 4 bit now) data path out of this. This is currently not supported.
There is apparently work on the way (the parbus) which will integrate this and other parallel-port devices (zip drives e.a.) more smoothly.