HEP Experiments Live Longer than Graphics Concepts

• Lifespan of HEP experiments such as SLD = about 10 years
• Lifespan of most graphics concepts = 3-6 years

Even the most basic questions, such as which programming language to use, change quickly. SLD software was founded so long ago (1986) that it began with discussions over whether to use

• ADA
• Pascal
• or Mortran (a SLAC Fortran pre-compiler)
• C was not yet a standard

Just the command input part of the SLD Event Display has gone from

• 1986 - Draw your own menus
• 1995 - Motif (or some other Widget set) does it for you

Buy Insurance: Insulate Yourself from Your Graphics Package

• 1987 - SLAC Unified Graphics was not going to become a world standard for 3D graphics
• We didn't know what was
• 1995 - Still waiting for a 3D graphics standard (is it GL?)

Because of this uncertainty, SLD bought an insurance policy.

Referring back to the block diagram of the SLD Event Display, you can see that we carefully separated our display functions:

• Read the Physics data, create an internal display list (the Build phase)
• Read that internal display list, call the underlying graphics package (the Transmit phase)

We never did change our underlying graphics package, but the insurance paid off anyway

• allowed us to run at least simple tests on other graphics packages (Phigs)
• allowed us to create flat file ASCII encodings of our display lists for use in vendor's benchmarking tests (and a VR test, never completed for other reasons)
• allowed us to add various features not supported by SLAC UG (such as object pick on low-end graphics devices)

Unless you have personal control over your underlying graphics package, buy insurance.