How Can Operations Get the Applications
Software That They Want?

R. Bailey, SL Operations,
CERN

Will concentrate on two things based on 10 years experience in SPS and LEP division at CERN;
* What do 'operations' want?
* What is the best way to get it?


In SL division at CERN, any major development of application software to be used in the control room entails interaction and a division of responsibilities between people from operations, controls, accelerator physics and equipment groups. Providing the suite of applications for the operators to use for efficient exploitation of the machines is no exception to this. Indeed since routine operation accounts for the largest fraction of machine time, it should be regarded as the most important activity. We have found over 10 years that teams led by operations people, making pragmatic use of formal development methods, prove to be very effective in producing the required software. This has already worked in two large projects (the SPS development for multicycling and the LEP Sloppysoft system) and will hopefully do so in a third (the SL transfer lines) because:
* the methods provided coherence and direction - you need them for large projects
* operational experience ensured the correct requirements were specified - you need them
* the same people were involved in analysis, design and implementation - allows pragmatism
* some of the team were to use the software - gives important extra motivation
Experienced operations personnel are best placed to perform the overall coordination of the different groups of people involved in large projects because they have a vested interest in getting the end product. This paper examines the modus operandi of the above projects and contrasts this with less successful ventures.