shindman@utcs.uucp (Paul Shindman) (01/06/86)
Unfortunately the guidelines governing the use of NETNORTH/BITNET
preclude this type of feed. I would be *very* upset if this started
for the following reasons:
-the rules say so. To quote for the usage guidelines "All usage of
BITNET must be consistent with its goal to facilitate the exchange
of non-commercial information in support of a member institution's
mission of education and research". As far as I know, NETNORTH has
adopted the BITNET guidelines.
-even if you could convince me that net.*.*.* was "consistent with" etc,
NETNORTH/BITNET is a 'store and forward' type network with only one
unique path to each node. USENET garbage would quickly grind the
network, as well as everyone's spool, to a halt. For example, if
York crashed, the news for Humber College, McMaster, and Laurentian
(as well as several York nodes) would pile up at Guelph. So multiply
net.sources.mac and net.sources by the number of receiving sites #
downstream from York and the spool at Guelph quickly fills.
-NETNORTH/BITNET already has several moderated discussion groups.
At present these are restricted to technical issues dealing with
the network. There is nothing preventing individuals from having
their own private mailings on topics of their choice.
So that's my opinion, of course.
--
-----------------
Paul Shindman, U of T Computing Services, Toronto (416) 978-6878
USENET: {ihnp4|decvax}!utcs!shindman
BITNET: paulie at utoronto IP SHARP MAIL: uofthenry@utzoo.UUCP (Henry Spencer) (01/06/86)
> -the rules say so. To quote for the usage guidelines "All usage of > BITNET must be consistent with its goal to facilitate the exchange > of non-commercial information in support of a member institution's > mission of education and research"... This sounds remarkably like how most of us justify Usenet as it stands. The wording is consistent with a large fraction of the Usenet traffic, even if the way it is interpreted isn't. > ...NETNORTH/BITNET is a 'store and forward' type network with only one > unique path to each node. USENET garbage would quickly grind the > network, as well as everyone's spool, to a halt. Usenet is also a store-and-forward network with basically one unique path to each node. I recognize that IBM boxes are not as robust and capable as our Unix machines :-), but this problem does not seem insuperable. We cope with it every day. > For example, if > York crashed, the news for Humber College, McMaster, and Laurentian > (as well as several York nodes) would pile up at Guelph. So multiply > net.sources.mac and net.sources by the number of receiving sites # > downstream from York and the spool at Guelph quickly fills. No, because the stuff isn't (shouldn't be) send from one central point to everybody; it would go to York and the generation of multiple copies for Humber etc. would occur there. This is the way it works now. And yes, having a site go down for a while does tend to produce pileups, although there are ways of controlling that. Again, I realize that IBM hardware and software cannot cope as well as Unix :-), but the problem is neither new nor intractable. > -NETNORTH/BITNET already has several moderated discussion groups.... Good, so there is precedent for this sort of activity. -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
mason@utcsri.UUCP (Dave Mason) (01/07/86)
(This is partially in response to private communications)
There seem to be 2 comments relevant here:
1) RYERSON is directly connected to GUELPH (CANADA1) via a 9600 bps line,
which is connected to (I believe) Cornell via another 9600 bps line. The
people I have talked to here don't seem even slightly concerned about 1Mb/day
being transferred, particularily if it is overnight, after all that is only
about 20 minutes at that speed. All the Netnorth links (and I am fairly sure
the bitnet links as well) are dedicated lines, so there is no incremental cost.
(rayan@utai has some doubts about this, I will verify it tomorrow and follow-up)
It is still not clear to me where the news feed onto bitnet would happen, but
I think this is all worth exploring.
2) There is the problem of replies. I would suggest that our news forwarder
could run through uupath (a path optimizer) to find the best uucp-only
return path from here (or from utzoo) and include that as the return address
so a reply would travel a fairly good uucp-only route by default.
I think this addresses rayan's concerns.
It appears to me that the Bitnet charter & the Usenet charter are almost
identical: ....non-commercial...technical...research...education...
even if the Bitnet charter is more formal.
It seems to me that potentially a lot of money could be saved this way.
../Dave
--
Usenet: {dalcs dciem garfield musocs qucis sask titan trigraph ubc-vision
utzoo watmath allegra cornell decvax decwrl ihnp4 uw-beaver}
!utcsri!mason Dave Mason, U. Toronto CSRI
ryesone!mason Dave Mason, Ryerson Polytechnical Institute
CSNET: mason@Toronto
ARPA: mason%Toronto@CSNet-Relay
BITNET: FCTY7053@RYERSON.BITNET