When standards aren't really standard
Far from being an academic exercise, standards work can involve as many special interests and commercial considerations as your average political convention.
I have to admit that when I first came to Lightwave, I believed most telecommunications standards were cooked up by a bunch of academics meeting in obscure “working groups” in Geneva, home of the International Telecommunication Union. One of the first stories I followed for this magazine involved a standards-making effort that expanded my view. First, I realized that standards also could be created through a series of road shows here in the States sponsored by such organizations as the Telecommunications Industry Association (TIA), American National Standards Institute (ANSI), and IEEE. Second, I discovered that far from being an academic exercise, standards work could involve as many special interests and commercial considerations as your average political convention.
That fateful standards process was the TIA/ANSI/EIA-568B premises cabling effort, particularly the attempt to enshrine a standard small-form-factor (SFF) connector. The standards committee (TIA 42.8) had the right idea-come up with a standard fiber-optic-cable connector that would be the same size as copper’s ubiquitous eight-pin modular jack (known to most people as an RJ-45). Such a standard connector would enable optical equipment to match the footprint and port density of competing copper-based offerings while making it easier for network managers, installers, and integrators to specify and deploy fiber in the enterprise. There was no shortage of available connector designs, which made it likely the committee could come up with a connector that would meet just about any application the industry could conceive.
Unfortunately, that panoply of choices proved to be the standard’s undoing. The stakes for each connector maker were considerable-if your connector became the standard, you had it made. If not, well, better luck next time. Each manufacturer made sure that it had representation within the various working groups and subgroups, and the standards process quickly devolved into a market-driven exercise designed to ensure that if your connector wasn’t going to become the standard, then no one else’s would either. As a result, the 568B standard contained only a list of specifications a connector had to meet to be standards-compliant. Several connectors eventually met those specifications, and the market was flooded with incompatible “standard” connectors.
I was reminded of this somewhat sordid tale recently when the Fiber Optics Association (FOA) announced it would take another run at developing a true-and single-standard SFF connector. According to the FOA’s Jim Hayes, the proposed connector will be manufactured cheaply using proven parts and technology (the FOA currently proposes 1.25-mm ferrules in a plastic body with a spring latch); easily terminated using standard technology; testable using standard processes; easily applied to transceivers, patch panels, and other equipment; and available to anyone on an open source basis, without licenses. While it would seem that the FOA has its hands full for the same reason that the TIA did-too many connector vendors with too much to lose if the status quo falls apart-Hayes says his group will attempt to out-flank the connector vendors by pitching the idea directly to their customers, who have had to deal with the mess 568B created. As Hayes told me in an email, “Manufacturers will do what the market wants.”
So, does the market want a standard connector? When 568B failed to settle on a single connector, vendors hailed the TIA’s willingness to “let the market decide.” So far, most transceiver vendors have settled on the LC for their SFF and pluggable devices for carrier equipment, although NTT continues to insist on the MU for its networks. In the premises, however, the LC and MU battle such competitors as the MT-RJ, VF-45, and E-2000, which has to be a headache for cable suppliers, equipment providers, and network designers alike.
So, I can see cable suppliers that didn’t have a vested interest in supporting a connector of their own design as being in favor of the FOA’s move. Transceiver vendors have already settled on the LC or MU; if they would hate anything more than supporting multiple connectors, it would be switching over to a completely different connector. So wouldn’t it make sense if the LC or MU could become the connector of choice in both service-provider networks and the premises?
If the legacy of 568B was to let the market decide which connector best met its requirements, I believe that decision has been made outside the premises. In that context, I applaud the FOA’s attempt to bring connector standards and the market they are supposed to facilitate into a more rational state, even if I think standardizing on an existing connector would be easier to implement than working with a new design. If you’d like more information on what the FOA has up its sleeve or want to offer support or a suggestion, find it on the Web at www.thefoa.org.
Stephen M. Hardy
Editorial Director & Associate Publisher