Dear Alan Duncan.
I have only just seen the August issue of the Bulletin and therefore only just read your letter about the Atlas Laboratory. I don't know to what extent the Editor wishes to have his pages used for the discussion of the dismemberment of ACL and I feel, perhaps not unnaturally, somewhat inhibited from entering the discussion. But there is one remark in your letter which I cannot let go by.
You say that by 1972 we were running a little research work ... much of which could not be accommodated on the local machine. and that the reason for this was often said that the worker could not convince his own University that the time was justified. You choose your words carefully - as is to be expected from an Adviser to a Bank - and possibly you intended no implication. But it seems to me at least that anyone not knowing how we conducted our affairs will infer that we provided a soft option - any old rubbish turned down by a University computer centre Director could find a home at Atlas. When we started the Laboratory we did indeed encourage research workers to come to us, especially in the non-physical sciences and the humanities; there was then far less awareness of the scope and power of the computer than there is now, and I and my colleagues were quite clear that we must do far more than passively hand out machine time. But things changed, as you say, and with Computer Board funding the Universities became really very well equipped. We were not unaware of this, nor of the fact that what many research workers wanted was a guarantee of computational support for their projects, possibly stretching over a year or more. We therefore proposed to our masters, the Science Research Council, in 1971 that we should in future devote most of our resources - actually, 80% of the machine time we had available - to such guaranteed services and that all requests for such services should be made formally as in the case of all other applications for SRC grants, should be supported by the University and should be assessed by a competent SRC Committee or Panel. The Council approved, the system went into action and worked very well indeed. I find it hard to understand how you could have been unaware of this.
It was, by the way, because they had heard of this and thought it such an excellent procedure that the American National Academy of Science, when they were discussing the setting up of a National Resource for Computational Chemistry (their title), invited me to visit them in 1975 to tell them about the philosophy, practice and experience of the Atlas Laboratory.
And perhaps I may be allowed the luxury of a final comment, this time on your Point 3 - get another biggest machine and see what we can do with it. Times may have changed since I was facing SRC Committees and Boards, but I don't recall much enthusiasm for spending several millions of pounds (as like as not in this case in foreign currency - but did you have other ideas?) with no particular idea as to what it was for.
Sincerely yours,
Jack Howlett
