I agree with Keith Elliston that comparing to what other's have is
a top argument for getting better computing facilities, but I draw
attention to an essential factor missing in most discussions about
facilities: the quality of service.
Quality is defined and built in practice by management principles.
Wazzat? Just so we know what we are talking about,
good management IS -identifying problems, their solutions, and
ensuring they are solved. It's an ongoing process, attempting
to deal with new problems in the most efficient manner as they arrive.
A computing facility usually lets other people use it apart
from the people who run the computers, so it offers a SERVICE.
Maybe that's obvious, but some people haven't thought through what
that really means.
You have to ask and answer questions like:
How do you offer a good computing service?
How do you tell how good the service is?
How well does it help biologists with their research?
Going back to Keith's goal, to get better computing facilities, it is then
essential to look at how effective other facilities are, so you can point
your finger and say "Geez, look how well their research is going with
that kinda service, we need that quality".
Getting bigger and better computers and more staff
doesn't mean a lot if the service doesn't deliver what's needed.
As I've said, offering an effective service is a management problem.
You need to apply professional management skills to help deal with it, to
define what you need, get the resources you need, get it running,
and keep it running, and get the service updated to keep up
with what the biologists need. And all that in the most efficient way.
Of course their are constraints, like the budget, so you aim for
a defined level of service that you can deal with.
Power for Change in Joining Forces:
If biologists need more from the computing service than it can give,
then it's time for computing management and the biologists using
the service to get together to go for more.
Getting in 'Da Management':
Decent sized facilities of course do have some kind of
management structure, but not all are as effective as they might be.
Academic institutions are normally not too hot on management,
especially time-based management which is important for a multi-task
environnment like a computer service. So we can learn from our business
friends who know all about it. You might even consider head-hunting
a manager or two from an appropriate part of the commercial sector,
as some have done, to improve effectiveness at a site.
However, moving commercial practices into the academic environment
without adapting them for the new environment could be a bad move.
The academic world has different priorities to the business world,
for a start, academia relies on having time for creative analytical
thought, that has no place in all, or almost all, parts of a major
An initial stumbling block to bringing in management, as well as
support bods and equipment, could well be at the decision making level.
Few senior staff members at many biology institutions or departments
are computer literate, beyond typing a letter on a PC. It can be difficult to
persuade decision makers of the importance of something they don't understand,
so educate them!
Existing support bod(s) at a small facility might simply be managing
themselves. In an upgrade they might not initially go for the idea of a
professional manager, as they like their independence. Also, because they
are working flat out they could well think they're doing an effective job.
However, they may also complain that the department doesn't
take their requests for more equipment and people seriously.
This is a prime situation for calling in 'da management'.
The hard working support bod may not take kindly to instruction from
above, but it's a two-way street. A manager can set priorities
on the many different tasks involved in support, to maximize efficient
use of the support person's time. And it is part of their job
to argue for and help get the necessary equipment and extra staff.
Manager are placed closer in the heirarchy to the decision makers,
so they are in a better position to influence the decisions, naturally.
Conclusion: Go for quality service, get in 'da management'.
The role of management and other factors in getting better biocomputing
facilities is discussed by me and the other guys in the SEQNET User
Documentation crew, U.K., in a previous posting on bio-soft, "Improving
Front-line User-Support: Proposals for Greater Efficiency in
Computational Molecular Biology." This has now appeared in BINARY,
Birmingham University, U.K.
Disclaimer: I am not nor have been a manager in this life, or any previous one.