|Jack Hendren, President and Chief
Ashlar, Inc., Sunnyvale, CA
Before joining Ashlar, Hendren was general
manager/vice president of IBM's CADAM EDA Division, with responsibility
for development, marketing, and distribution of P-CAD EDA software
products for PCs and workstations worldwide. He also spent 12 years at
Motorola Computer Systems, where his last assignment was as senior
marketing and manufacturing executive. He has a BA in math and physics
from Hanover College and an MS in operations research from the Naval
Postgraduate School in Monterey, CA.
Computer platform changes have several impacts on developers of CAD software, says Hendren. Among the impacts: continual efforts to make sure user interfaces accommodate the platforms.
Design News: What do you see as the key technology issues facing software vendors now?
Hendren: Naturally, there are several. The largest is the rapid change in platforms. That constant change makes the environment unstable. In any year, there could be two platform changes. This year the changes include the Power PC. Windows '95 and NT are also changes, although NT hasn't made the jump we expected. All the changes affect the way suppliers program user interfaces, and the way customers use them, or the way they share data with others. We have to keep a team of people expert in each platform to follow the changes.
Software companies can't always pick the winning platforms, so Ashlar's products run on Macs and on Intel-based systems. We know that the ability to exchange data reliably is under the surface of all the issues software vendors deal with. Data sharing is vitally important to corporate America, where managers want purchasing and design, as well as other functions, on the same data base. CAD vendors have to open up their systems to reading and writing data from other vendors' systems, and the shared geometry in systems must be manufacturable. We try to identify key data formats that support the work of multiple users and back them up as standards. As a front-end supplier, we have to be able to export data to multiple environments.
What will be the next major breakthrough in engineering software?
A: The breakthroughs will be in usability of the software. In word processing, when software allowed users to work on a page with free forms, the industry went from a character-based system to a shape-based system. That change opened up the possibilities for desk-top publishing. Today, high-end CAD, in essence, requires users to have between a bachelor's and a master's degree. At Ashlar, we have given users context-sensitive feedback, and we're going to add that beyond 3-D wireframe to surfacing and solid modeling levels. We're trying to show that CAD can be open to the masses, too, and it's that kind of ease of use that will make it happen.
Q: What effect will massively parallel computing have on engineering?
A: Software developers have moved to C++ and object-oriented programming. The latter allows you to spread out what you're doing to multiple processors so operations will be quicker. That also enables software to answer questions with incomplete input by running what-if scenarios. Massively parallel processing can help bring math-intensive, specific-function tools to the general engineer. The result will be renaissance engineers who can do much more than they can now.
Do you plan to expand Ashlar's product offerings into such areas as solid modeling?
A: Our first areas of extension are surfaces, shading, and rendering. People think of things as surfaces, and it's natural to want to adjust a design by manipulating the surfaces. It's easier to use surfacing than Boolean operations or solids. So, our direction is to apply the Vellum paradigm to surface design, then continue into the solids area. In each case, we have to ask, "How do we make it as easy to use as Vellum is today?" Interestingly enough, our overseas customers are telling us to put more 2-D functionality in, while our domestic customers are telling us to make it more exciting in 3-D. We will balance the two.
Is seamless integration among software products possible?
A: It's almost possible, if you follow one platform or vendor's conventions. Microsoft, Sun, and Apple are all pushing their conventions. Customers would certainly like seamless integration to happen, especially if they don't have to pay for it. But, if there is no additional income from it, vendors would have difficulty providing for seamless integration. Nevertheless, it's inevitable that more and more process and database integration will happen.