Is Comprehensive and Intuitively Usable Commissioning Software Feasible?

The commissioning period is challenging both for provider and user of accelerator operation software. With increasing knowledge of the specific accelerator new functionalities and methods become important. The variety of implemented features grows until certain standard procedures are settled and the programs can be simplified again for the all day tasks. It is not too difficult to provide the requested functionalities and certain handles for a graphical usage for expert users. For the general member of the commissioning crew, however, the GUI may consist of huge windows, context dependent presentations, hierarchically structured menu trees or compact screens dominated by (modifier) key and mouse click navigation. Understanding and usability of the programs tends to decrease with the increase of available features. By means of examples like the orbit display and control program the problem is outlined and some guidelines leading out of the dilemma are given. 1 GENERAL ORIENTATION The fulfillment of the apparent requirements for a graphical user interface (GUI) on a windowing system is the easy part of the design for an accelerator man-machine interface. Windows should not be too large. This would often hide other relevant information. They should not be too small to be readable. Behaviour and appearance should be uniform and consistent. Conflicts, inconsistant or harmful usage has to be prevented by protection mechanisms within the code. Configuration of action elements (e.g. menues) should be determined by the actual program context. In addition control software should obey some general rules that help to shuffle around the piles of windows. E.g. standardized head lines with a color coded title help to identify the scope of the application like accelerator part or device class involved. An ‘About’ button popping up author, version number and program status information helps to find the competent expert to get pointed support. 2 INFORMATION FINDING The first obvious approach to give access to program functionalities would be an open arrangement of controllers on a flat screen. As the windows get overloaded or too big grouping of entries according to common area of functionality becomes necessary. This may be done by hierarchies of sub-windows or menue trees. Often the unifying root Funded by the Bundesministerium für Bildung, Wissenschaft Forschung und Technologie and by the Land Berlin y Email: [email protected] entry is no more unique and meaningful. A not very precise search is required to find the desired function. One solution is to offer multiple ways to accept user entries simultaneously: menue, buttons on subwindows and keyboard shortcuts. The other way is to aim at very compact screens built from a few well designed and possibly multifunctional elements. A well designed example is the CERN wheel switch[1] (see Fig. 1). Figure 1: Wheel switch – example of a well designed widget It can be operated in a secure way with mouse clicks, keyboard arrow keys and by entering the numeric value. The input field accepts only numerical digits. The accessible MIN/MAX values are apparent. Since it scales nicely it fits into most panel layouts. The commercial XRT/graph[2] widget is another example. It supplies various convenient zooming, printing and data presentation options. The programming interface allows to install dedicated zooming grids, to attach pop-up labels to each data point, to (de)select points and to drag values. Figure 2: Clipping of a display showing the actually measured orbit data and the predicted effect of a 4-bump with angle and amplitude specified at 75.0 [m] Fig. 2 gives an impression how it can be used to display the measured orbit and the predicted effect of a closed bump. The target position of the bump (long vertical bar) can be freely placed with arbitrary precision by zooming into the lattice and selecting the location with a mouse click. Experts like it, regular crew members are reluctant 0-7803-5573-3/99/$10.00@1999 IEEE. 726 Proceedings of the 1999 Particle Accelerator Conference, New York, 1999