I need you, the list owners, to give input to a group of technical people
about the gravity of a LISTSERV problem. The message I am forwarding is
pretty technical and it is not possible to translate to layman's terms
without making some unfortunate simplifications, which tend towards
slightly overestimating the impact of the problem by generalizing it to
cases where it would exist anyway. But I feel I have to make this
description in order to allow the people who are actually impacted to
give their opinion, and not just people with a CS degree.
The change described in the lines marked '>' would considerably reduce
the amount of US recipients erroneously subscribed to european peers when
there are US peers as well (only peered lists are affected - well if the
list isn't peered there's only one place to subscribe people to :-) ).
Michael Gettes, whom I am copying on this message, is wondering how
serious a problem it really is. On the other hand, the proposal he is
mentioning is unlikely to be formally approved by CREN and the core sites
within a reasonable time frame, as there have been a number of
non-technical objections (such as "it shouldn't be the business of X to
decide Y" - there is a pretty good agreement about the technical aspects
of running a gateway). In order to avoid skewing the results, please do
not answer if you have never run a peered list. Again, this only affects
the owners and subscribers of peered lists, and I just want to make sure
people who answer have first-hand experience with the situation.
The Reply-To: field has been set to point to the list where this is being
discussed. Thank you.
Eric
*---------------------- Original message follows -----------------------*
On Sun, 07 Mar 93 01:05:47 +0100 [log in to unmask] said:
> Good. I had a long phone conversation with Michael on thursday and,
> barring any further political problem or unexpected objection from the
> core sites, we are going to do the same thing we did in EARN to solve the
> same problem - register all the core sites as "logical INTERBIT", in
> principle by VERS9304. Each core site will have the option of either
> implementing a genuine INTERBIT using suitable software (and in
> particular a Netdata-capable mailer), or simply routing the INTERBIT node
> to another core site willing to take the load (in practice, this means
> PUCC).
My delay in responding to this list is once again due to having responsibility
to projects and problems at my own site. I apologize if my lack of
participation has hindered any progress.
First, Eric seems to have misunderstood our conversation. I wanted to
confer with him about the problems and what it might take to resolve the
situation. I told Eric that I would speak with a couple of the core sites
that may be impacted if we do this. However, from my perspective, the
problem seems to not be as drastic as portrayed. The only INTERBIT traffic
that Europe is being hit with from the USA is from the peered lists.
Otherwise, PUCC is getting the rest of the USA INTERBIT traffic. I point
this out because there has yet to be any further movement on the proposals.
I will not rehash the problems thus far. I would prefer NOT go about
adding all these sites to INTERBIT without there being closure on the
issues of the proposals. Are there any technical objections? If there
are no technical objections, then I hereby formally ask that these
proposals 'be accepted or moved up a level'. Eric, I know you have no
appreciation for this request. You and I disagree here. I am not saying
I do not wish to resolve the INTERBIT situation or not do as the 'network'
wishes... but the bottom line is that proposals have been put forth, EARN
has agreed to them, and nobody within CREN seems to give a hoot. If it
is the opinion of people that the proposals should be completely scrapped...
then you should speak up. If nothing is said, and nothing else happens,
then, due to the lack of any procedures, the proposals will be considered
accepted and we will add sites to INTERBIT as quickly as possible, but
carefully.
Folks, I understand that we have accepted some degree of responsibility
in all this... and with that responsibility, along with John Wagner,
we are trying to act responsibly and carefully to make a highly visible
service of BITNET stable and as clean as possible. To then try and move it
forward and so on. I too am getting tired of this... but we alone cannot
make these decisions.
If certain systems in EARN are currently getting overloaded due to the
lack of action on this side of the pond, I would appreciate those sites
letting me know and hopefully we do something to alleviate your pain for
the time-being.
/mrg
|