There are many things that become so innocuous that not only is their presence not particularly noticed, but should they disappear, their absence is not obvious either. This leads to difficult domestic moments, usually starting with lines such as "Say, what happened to the chair we used to keep in that corner?"
We perhaps should broaden the category of unnoted vanishings to include another item: web-based design. Not so long ago, we were all being assured by the pundits of the new world order that all design activity would soon be conducted via the web. The vision was one of data sets and tools residing on massive third-party server farms, with design teams only paying per-use fees for an astonishing array of tools and for access to infinite computing power.
Like that chair, we pretty much got used to having these capabilities around, at least in theory, without actually ever sitting on them. And now, it appears, they may be quietly disappearing . At least one significant EDA vendor is dropping the web-based tool model entirely, and others are definitely de-emphasizing it. Public exchanges for selecting components and establishing contracts appear to be drifting away as well.
Before dismissing this as just another unjustified casualty of the Internet bubble, it's worthwhile to take a closer look. There were several specific problems with the web-based tool model from the beginning that, ignored in the web feeding frenzy, gradually conspired toward the model's demise.
One of these issues was the reflexive desire for security among design managers, however little this longing might be shared by the engineers actually doing the work. A lot of managers -- and even more, corporate legal folks -- never became comfortable sending the intellectual property for a design over the all-to-fallible web, to journey through an unknown itinerary of servers. It might come to rest, perchance to take root, on someone else's disk farm. No amount of reassurance and no number of lectures on encryption seem to have put this unease to rest. Certainly the high-profile antics of hackers against very public, and presumably very protected, sites didn't help the image much.
Another issue may have been equally important. Outside academia, a design manager's responsibility doesn't end with hand-off to manufacturing. It is customary to document and archive the design so that it can be recreated if necessary. In many contract situations, you can substitute "mandatory" for "customary." But it was never clear how this could be achieved with web-based design services. Who would do the archiving? Did the service provider even have the legal right to archive a complete tool set? To whom did the rights, and more important the physical possession of the data, pass should the service vendor suffer the fate of most start-ups? These issues also tended to uproot web-based activities before they could take hold, at least in larger organizations.
Finally, there was a less concrete issue, one that has plagued many attempts to create value for intangibles via the web. The perception of engineers, raised on the Internet in college, often in the days before web commerce, is that stuff you get on the web is free, and darn well worth it. This perception doesn't rear an ugly head in the midst of negotiations, scattering the well-intentioned attendees. Rather, it manifests itself as a lurking feeling, always just beyond immediate recognition but always near enough to influence thought, that any product delivered via the web is competing against something equally good, but free.
This spirit has animated many admirable things, including, arguably, the birth of the EDA industry as we know it from the cradle of a few university FTP sites. It has nurtured and watched over the growth of Linux as a viable alternative to the Unmentionable Empire. But it has also choked out a lot of potential cash crops carefully planted by entrepreneurs and investors at considerable expense.
So for a variety of reasons, some quite tangible and some quite hard to quantify, web-based design has been an uphill battle, and is showing signs of ending badly. But as irony would have it, as one model declines on the web another is prospering rather too well.That would be the use of intranets, and in some cases the Internet, as a medium for design participation among geographically dispersed teams. Just how to do this -- where to put the servers and tool licenses, how to manage revisions and even how to get the data moved around -- is emerging as a primary topic in design management.
For starters, design teams are reporting that sharing a design between multiple sites means moving data sets that can take hours to transmit at T1 rates. With the seemingly geometric growth in data files, we can imagine a scenario in which a block design physically can't be moved around the earth fast enough to keep up with the daylight. Nonetheless, if web-based design may be waning, network-based design is booming through a healthy inf ancy. And the increasing globalization of design expertise promises a boisterous, if bandwidth-poor, childhood. Those web-based tool vendors may find that their hard-won experiences in handling design via the web will be valuable in a quite different endeavor.
Ron Wilson is editorial director of ISD Magazine and a contributor to EE Times. He has covered chip-related matters for 15 years for various industry publications, and was once, in the distant past, a designer himself.