Adecade ago, the Optical Fibers in Communications conference (OFC) could always be counted on as the ultimate in sandbox experiences. Academia's brightest lights offered the latest research results on solitons and wave-division multiplexing, although little practical was expected to come of the results. All that changed as Ciena Corp., Pirelli Fiber and others launched a dedicated WDM market in the mid-1990s. By the end of the decade, even the most far-out exploratory programs, such as Hewlett-Packard/Agilent's bubble-jet photonic switch, became the subject of instant financial-analyst speculation.

Has the recession set back the clock 10 years? On the eve of the optical industry's two big fall shows-the European Conference on Optical Communication (ECOC) and the National Fiber Optic Engineers Conference (NFOEC)-all signs point to an optical industry that's licking its wounds as it retreats to its research roots.

In late August, Ciena reported a quarterly loss that spurred many analysts to release "underperform" ratings for the company's stock. A few weeks earlier, Agere Technologies Inc. had followed up the closure of its Breinigsville, Pa., automated photonics plant with the announcement that it would get out of all optoelectronic

component manufacturing, apparently for good.

It's obvious that a near-term market for many photonic components is minuscule at best, causing developers to rethink programs for 2-D and 3-D switches, collimators, tunable lasers, splitters, combiners and higher-integration devices. But is the industry really ready to revert to an era of only coarse, analog manipulation of wavelengths? Must all work on furthering the Boolean manipulation of light return to its spawning grounds, the university labs?

It must if the manufacturing company in question still puts an overriding premium on "shareholder value." But I believe a communications component or OEM company, if it wishes to be built to last rather than built to flip, must favor the notion of performance for stakeholders over performance for shareholders. A viable communications industry preparing for an inevitable upturn must fend off the impulse to please shareholders quarter by quarter, and those investors who claim to be truly interested in communications must sacrifice short-term returns for long-term architectural viability.

Otherwise, we may as well leave OFC, ECOC, and NFOEC to the academicians and be done with the notion of a fiber-based broadband infrastructure.