There were many problems to solve as there were few off-the-shelf solutions. In the absence of standardised formats and procedures, international (and national) exchange of data on magnetic tape was a matter of responding afresh to each request for similar data sets.

Lack of experience with air-conditioning in Australian conditions led to problems, such as those encountered when using CSIRO equipment at Monash University. During the summer of 1965 the system could not cope with the combined atmospheric and machine-generated heat and it was a matter at times of 'run for 10 minutes and turn off for 20'.

This added to the nightmare of transferring the Bureau's library of climatological data contained on about 20 million punch cards. First, the condition of some of the very old cards was poor; they tended to stick together when it was humid, and to separate when it was dry, and jams in the card readers were frequent—at times averaging about once every 200 cards added to that were inadequate error-detection systems on the magnetic tape units.

Floor loading requirements and computer room layout generally were also unfamiliar problems. On one occasion the fire alarm was set off by a contractor's engineer smoking while working on the smoke-detection system.

It also was not realised immediately that the shocks female operators experienced at the console resulted from undischarged static electricity due to the interaction of the friction of card and paper trolleys, the condition of the false flooring, and the operators nylon underwear.

Then there were problems of distrust—one very zealous officer went to a lot of trouble to re-create card files from the magnetic tapes onto which the card data had been transcribed.

Expectations—too little, or too much

There were two general types of user—those who expected too little from the computer and those who expected too much. The latter were much harder to handle as they became disenchanted or frustrated he others, pleased with what they got, were quick to ask for more. On the other hand, frustrated enthusiasts kept up the pressure for development of facilities.

Building of confidence in the system was hampered by the general attitude of distrust in the community—'Can computers think ?' Will they take over?' Nor was it helped by a fear of job loss or job degradation—'I don't want to see a computer sitting in my chair': A common approach was to concede that computers 'might be alright for some sections, but it won't work here'.

The contrast of attitudes to computers then and now may well be more important than the difference in catalogue-rated capacities. Then it was distrust, now they are widely welcomed. Whereas the first configuration was specified mainly by computer specialists, now highly expert users know the capabilities of the facilities they seek and exert a highly significant influence on the specifications.

Bureau of Meteorology