🚀 go-pugleaf

In article <1a1imaINNbdo@huon.itd.adelaide.edu.au> dclunie@sirius.ucs.adelaide.edu.au (David Clunie) writes:
   ...I thought I'd get the ppp one going and later hack it do both
   ppp and slip depending on some configuartion parameter.

   Before I start, I thought I'd just ask if anyone else had done this
   before and had any success !

When we released v1.0 of Morning Star PPP, it did only PPP.  We stuck
our noses in the air and proclaimed that we were such purists that we
wouldn't bother supporting the ancient, ill-defined, less-functional,
nonstandard SLIP.  But then we kept hearing from customers who had to
talk to hosts that talked only SLIP, and the protocol purity party
line seemed to make less and less capitalistic sense.

So I spent a morning experimenting, and found that once the on-demand
dialing, idle-line hangup, packet filtering, IP interface, VJ
compression, and all that other work has been done, it's really pretty
simple to short-circuit the FCS generation and checking, the asyncmap
handling, and all of PPP's other useful features.  I hot-wired the
finite state machine to "Opened", since SLIP has no negotiation phase.

So our PPP daemon quickly had a "slip" command line option.  Since it
was already there, it was easier to delta it into our SCCS'd sources
than to rip it out again.  Maybe another half-day total since then has
been spent debugging and documenting it, and adding a convenient
peer-CSLIP-capability-detection feature.  So it shipped in v1.1, and
now the marketing folks have another feature for their bullet list.

Our source is probably significantly different from source of the
Purdue on-demand dialing implementation you said you were about to
try, but I don't know how different because I haven't looked at the
Purdue code.  I know for sure that our product is quite a bit
different from the other popular free Sun-based PPP code.  But you
still shouldn't have too much trouble making any existing PPP
implementation do SLIP, since SLIP is such a simple protocol.

In article <1a1imaINNbdo@huon.itd.adelaide.edu.au> dclunie@sirius.ucs.adelaide.edu.au (David Clunie) writes:
   I am about to set up a dial-on-demand style ppp link (you know - when
   a tcp/ip packet or whatever comes along it gets queued until the
   phone link is dial, which later times out if not needed). I am
   planning on using "dp.2.0.tar.Z" to do the job, which seems easy
   enough to install.
   [...]

The copy of this I have doesn't have any documentation on how to
actually use the programs that get made (the README says what the
various directories contain in a very general way). While I can
probably figure it out from the etc/* samples, I thought I'd ask for
documentation first.

Brian
--
       Brian Glendenning - National Radio Astronomy Observatory
bglenden@nrao.edu          Charlottesville Va.           (804) 296-0286

Err...  get dp-2.2-beta.tar.Z, there are only two swall typos in the
Makefile (I can send you the patches), and it's much better documented
and usable than -2.0.


--
Paper mail: Ignatios Souvatzis, Radioastronomisches Institut der 
            Universitaet Bonn, Auf dem Huegel 71, D-5300 Bonn 1, FRG
Internet:   souva@babsy.mpifr-bonn.mpg.de ignatios@cs.uni-bonn.de

Err...  get dp-2.2-beta.tar.Z, there are only two swall typos in the
Makefile (I can send you the patches), and it's much better documented
and usable than -2.0.


--
Paper mail: Ignatios Souvatzis, Radioastronomisches Institut der 
            Universitaet Bonn, Auf dem Huegel 71, D-5300 Bonn 1, FRG
Internet:   souva@babsy.mpifr-bonn.mpg.de ignatios@cs.uni-bonn.de

> The copy of this I have doesn't have any documentation ...

No. A pity ... but I have found lots of useful and relevant
advice by reviewing the documentation with dialupip - the slip
predeccessor of dp.2.0. This is particlularly true of the scripts,
though there are small "enhancements" !

Maybe if I learn enough about the program while I am hacking at it
I will write some docs ! No promises.

david
-- 
David Clunie (dclunie@itd.adelaide.edu.au)