Theatre Design & Technology - Spring 1989 - 62

...and much more. FULLY ILLUSmATED (B & W) 200·PAGE CATALOG.

in 3'009 binder

$/5.00 includes one-year supplement sUbscription
ORtlAI.lENTAL RESOURCES,ltlC

BOX 3010-n

Idaho Springs, CO 80452

(303) 567·4987188

Scenic Paints, Lighting, Tools, Special effects.
Draperies on quotation
Large Stocks

Immediate Shipment

Write On Your Letterhead For Complete Catalog

MUTUAL HARDWARE CORP.
5-45 49th AVE., L.I.C., NY. 11101

TD&T·

SPRING 1989

(718) 361-2480

expected to be any easier.
Unfortunately, most of the types of lighting control
standards being proposed would be vastly more complicated than DMXS12 or AMXl92. The amount of work
required to draft them, to implement them, and then to
debug them would be considerable- which suggests we're
going to have to think very carefully about how we approach developing them.
Because of the work involved, these new standards, to
find support, must offer powerful practical advantages and
be easily applied across a wide range of products. Without
the prospect of that support, little motivation exists to
even draft one. Few people are interested in spending six
months on the lighting industry's version of "Esperanto."
The effort required to implement and to debug a communications protocol also suggests it may be time to look
at the methods we use to define and distribute them.
Like most standards, the protocols we already have reflect several different approaches. The "construction" approach dictates how a product is to be built to meet the
standard. Where it is not possible or desirable to dictate
how something is built, it may be more effective to set
performance standards; that is, to permit the manufacturer to use any construction so long as it achieves the
specified results. DMXS12 is largely a performance standard. But, if no ready method of testing products for its
performance/conformance is available, the value of a performance standard is limited. Today, years after DMXS12
was defined, we still don't have a handheld tester that
allows a user, consultant, field technician, or manufacturer to determine whether a product is truly DMXS12.
What's more, the written standard itself requires testing. Without the most careful drafting, a standard contains
ambiguities and omissions. Frequently its drafters don't
recognize those weaknesses because of their own assumptions and conventions that make the published standard
"clear" to them. Unfortunately, we discovered with
DMXS12 that those assumptions and conventions aren't
always shared-even by the drafters themselves. Because
formal standards construct (for better and for worse) an
abstract model by which products are designed, products
are engineered to meet the standard (rather than to be
compatible with other products). When the same standard
is open to different interpretations by different people,
products are no longer engineered to meet the same
standard.
A certain amount of this is inevitable. But simply publishing a standard and walking away from it isn't going to
work very well when standardized protocols are complex.
We need a method of learning from the "mistakes" people make in interpreting and implementing a standard, and
we must then share that knowledge, whether by revising
the standard or by generating "How To and How Not To"
advisories for reference.
Because a published standard. generates inertia that
makes it difficult or impossible to change, we should insist
that any lighting control standards process must require a
multi-manufacturer beta test before publication. Hand the
proposed standard to several engineers with no prior exposure to it. Ask them to build a compatible product or at
least outline one. If their prototypes aren't compatible, it's
the standard that needs fixing.
Supporting a standard protocol requires a duplication of
engineering effort by each manufacturer. In some cases,
new standards could be more readily adopted and less

Table of Contents for the Digital Edition of Theatre Design & Technology - Spring 1989