CREN, EARN and L-Soft met yesterday in Copenhagen to discuss the future
of BITNET and what role L-Soft could play in all this. Here is a short
informal report, which does not necessarily represent EARN's or CREN's
view, I'll spare you the disclaimer.
The meeting was mostly an exchange of information where each party
described its future plans and tried to see how they could be made to fit
with the others'. As far as L-Soft was concerned, this was mostly a
presentation of L-Soft's strategic plans followed by a discussion on the
future of LISTSERV; there were a number of closed sessions without L-Soft
during which I assume that other relevant items were discussed.
To answer the question on everyone's lips, CREN was understandably not in
a position to make a decision regarding a global LISTSERV/LMail license
without getting back to the board. This is inevitable in spite of the
severity of the time constraints both parties are operating under, so
L-Soft agreed not to market the software to CREN members until the next
board meeting on October 7. Note that this does not mean we will refuse
to sell licenses to CREN members until then; we will just not be mailing
marketing material to CREN's members until the CREN board has made a
decision. In addition, due to the complexity of the situation and the
fact that some current LISTSERV users may already have plans to drop
their VM system, CREN will be conducting a survey in order to determine
what kind of license, if any, would be most appropriate. While I don't
know the details of the survey, I imagine that it will be sent at least
to the techrep of each LISTSERV site. So if your techrep is the person
managing the payroll system and you think that LISTSERV is crucial enough
for CREN to be acquiring a global license, you will want to make sure to
brief him as soon as possible so that he doesn't give the wrong answer
due to lack of information. Bear in mind that, even if charged back, a
volume license would be significantly cheaper than an individual one. If
all 170 sites were to be licensed, L-Soft would grant 50% of volume
discount, corresponding partly to administrative, marketing and customer
support savings. If the software has to be marketed individually to the
members, it is inevitable that money will take longer to flow in and
L-Soft programmers will end up having to spend some time answering
questions from prospective customers, as it is not realistic to hire
interim people and expect to be able to train them quickly enough. In the
long run L-Soft would collect more money due to the removal of the volume
discount, but the development of non-VM versions will be delayed.
Another major item on the agenda was decreasing the load of the current
core sites. L-Soft confirmed that DISTRIBUTE and the other peripheral
functions necessary to operate a DISTRIBUTE server would be available in
the first stage (code name "Patchwork"), since they are presently written
entirely in PASCAL. That is, the first non-VM version should be able to
operate a backbone server, and new servers could then be added at
universities that do not have VM. LISTSERV however only requires a small
amount of system resources; delivery of the message to the final
recipients is what is expensive. After removing my L-Soft hat I described
a number of things that could be done to improve the situation and, while
neither I nor L-Soft are likely to be involved in this and I can
obviously not speak for CREN, I got the feeling that very concrete steps
are going to be envisioned in the next few months. CREN is no longer
hoping that someone will find a "creative solution", but instead they are
going to work on a concrete list of proposals and ideas.
The last major item was the future of mailing list managers, and
unfortunately this did not conclude as positively as the rest of the
discussions. CREN is under intense pressure from people with a strong
following in the Internet community to basically throw away LISTSERV and
start something from scratch, in C and without NJE support or any other
"stone age legacy". While this was more or less common knowledge (or at
least not a surprise), given the contents of the RFP, I was not aware of
the seriousness of the problem. It appears that high level managers are
now making statements which one would ordinarily expect from young
technical people who think they know the answer to every problem in the
world, such as "it won't be useful unless it is written in C", "if it's
written in PASCAL it can't possibly be any good", "anything derived from
a mainframe-influenced design won't be useful to the Internet community",
and so on. As a computer professional I found this deeply depressing, but
of course it is a serious business problem and it needs to be addressed
as such. There is no solution within L-Soft because the only way L-Soft
could possibly produce a solution satisfying these people would be to
dump the current VM version and development plans, fire me, and hire unix
programmers to develop a totally new list manager from scratch, in C.
Apart from the fact that this would be like asking Sun to please dump
their current offering no matter how successful it is and start making
PC's instead, it simply doesn't fly from the business angle and I can
think of hundreds of other companies more qualified for this type of work
than L-Soft, whose name has probably been branded as a dinosaur
derivative anyway. So of course L-Soft will continue to support the VM
version and to develop VMS, unix and NT versions of LISTSERV according to
the plans that were already published. L-Soft works for the users, who
ultimately are the ones that decide what should or should not be bought.
While there will always be people arguing to their management that
LISTSERV should not be used because it is written in PASCAL, we are
confident that proven quality will prevail over religious arguments.
None of this solves the problem, unfortunately. I can imagine that this
course of action is temping for CREN as this would undoubtedly secure the
continued membership of unix-oriented members (including the non-bigoted
ones, which would still benefit from the availability of a unix list
manager with more functionality and a much bigger development budget than
the current free packages). Relying on L-Soft's politically incorrect
plans to provide a unix version of LISTSERV, however, would probably not
secure anything as the bigots would drop out in retaliation and the
situation would remain unchanged for the non-bigots (they might benefit
from L-Soft's unix development, but that is not tied to CREN membership).
I am confident, however, that a solution exists. The 170-odd members with
a VM system would receive actual benefits from a continued CREN
membership in the event of a global LISTSERV/LMail license. This should
be enough for them to renew their CREN membership another time, while
L-Soft proceeds with its development plans. The non-bigoted unix sites
should consider the implications of having L-Soft and CREN compete on the
provision of a unix mailing list manager, with L-Soft offering full
compatibility on 4 major systems by 4Q95 and having the advantage of a
proven, tried-out interface and design. If L-Soft delivers before CREN
does, it seems unlikely that the majority of unix sites would be willing
to wait for CREN's version just for the satisfaction of not running evil
PASCAL code, and CREN members will have funded an expensive development
for nothing, introducing more end-user confusion on the side. If CREN
somehow manages to produce an equivalent product in one year, there will
still be an L-Soft version coming out within the next 6 months, with the
advantage of compatibility with the installed user base and the level of
end-user supportiveness that comes from listening to what users want the
product to do, rather than focusing on making technical people
prematurely promoted to a management position happy. It will be a serious
fight and, given a suitably low entry-level price made possible by the
use of inexpensive hardware, it is quite likely that certain departments
will decide to use L-Soft's product on their own while others run the
politically approved version. Ultimately, the end users are going to get
what they want to have, and whether we like it or not we are living in a
world of PC users who are using the mail programs they are using, and
aren't going to switch just because they can't conveniently use the
politically correct list server with their current mail program.
At any rate, this survey would be the ideal time to tell CREN whether
your organization would rather see it compete with the current LISTSERV
or just stay in the background while it is being developed. I cannot see
any business reason for CREN to revive the RFP other than as a means of
securing continued income from a certain category of members. If a
majority of CREN members were to say that their decisions regarding
continued membership would not be affected, or would be negatively
affected, by a decision to develop a new unix list manager from scratch,
the problem will be solved.
Eric
|