west@ru-cs44.UUCP (west) (06/07/84)
This is a follow up to Geoff Collyer (geoff@utcsstat.UUCP)
and his comments on Dave Anderson's ``big is beautiful''.
As a /44 user, (and I'm not necessarily limited to
using ``just'' a /44) I could not agree more with Geoff. It is
very dangerous to start assuming that efficiency is no
longer a design criterion for complex software (i.e. news). It
is also arrogance of the worst order to assume that just because
one site has money enough for the newest hardware that sites
who are constrained to work with ``lesser'' machines must
suffer. Unfortunately, not all sites have the leverage
or the resources to provide their users with all the latest
VAXes or whatever the flavour of the month is. It's not a case
of not wanting VAXes (or what-have-you), it's a case of simply
not being able to afford it. Such sites are already having to
pay twice over for being the ``poor relations'' - they are
stuck with the abilities of their current machines AND are
being penalised for something beyond their control. Even if
sites can afford VAXes, they may have many very good reasons
for not wishing to upgrade their present facilities; a new news
system is not sufficient reason to make them change their
minds.
To introduce a de-facto split in the USENET community
by introducing another news version effectively requiring a
minimum hardware configuration goes against the spirit of any
network such as USENET, and will in the end be self-defeating
as those with the ability to impose their ``minimums'' on
everyone else will drive out all those who fail to conform.
The same applies to the current argument over newsgroups.
In certain circumstances, of course, it is entirely
wrong to move at the pace of the slowest; if you can't stay the
pace, you must drop out of the race. This is NOT such a circumstance.
There is undoubtedly a need for news to be revamped (volunteers
take one step forward...) but the news system, of all systems,
must be totally compatible in a way that few other systems need
be. Total compatibility is often best achieved through total
portability.
The upshot of all this is that although news may be in
need of an overhall, and I personally am in favour of
implementing the time-bomb proposal to solve the unused-group
problem, the new system must be able to run on all current
sites, and preferably on any system running Unix, now or in the
future. There are already certain minimum requirements to be
able to run news, let's not add any more.
Jerry West
Computer Science
University of Reading
United Kingdom.