In my travels, a common question that’s posed is something along the lines of “is it safe to buy pre-standard Wi-Fi technology?” While the question comes up most often in relation to new physical layer standards (such as 802.11g, 802.11n, or 802.11ac), it is a fair question for any emerging technology within the Wi-Fi arena.
The short answer is that yes, it is safe to buy products before the ratification of an underlying IEEE standard. The longer answer takes a bit too long to explain in a webinar Q&A session, which is why I wrote this post.
As a business, networking is about interoperability. When you buy an access point (AP), whether from us at Aerohive or down at the local electronics store, you do so with the expectation that you can use it with your favorite computing devices. The standard is a set of rules about how to put together a packet that can be sent from an AP over to a smartphone, but all rules are not created equal.
The fundamentals of a physical layer standard remain fixed throughout development. Putting bits on the radio link is the essence of what is done in hardware, and in the networking business, we try to push packets through hardware as quickly as possible. Standards and hardware are co-developed on parallel tracks at the same time because once the basic characteristics of the packet format are worked out, the rest of the work is fairly straightforward. In the 802.11n days, the big work of picking packet formats took place in the merger of the original “big three” proposals to become the first draft of 802.11n. With 802.11ac, the initial process was simpler because the underlying technology was evolutionary instead of revolutionary.
As a specification marches along the standardization process, it is theoretically possible that technical changes can be made. In practice, it is quite rare. Hardware design based on the draft standard has started, and early release products (the hardware equivalent of software’s beta cycle) are available. As the standard begins to emerge, product vendors are already creating interoperable products – in this case, typically aided by the Wi-Fi Alliance certification programs to guide what we are making.
Altering the specification to require a hardware change can be done, but would need to be approved by the 802.11 working group, composed of working engineers who are building products as they work on the standards. Making technical changes to 802.11 standards requires a 75% supermajority, and that high bar implies that changes require an overwhelming consensus. In the absence of a serious problem uncovered as the standard emerges, it falls into the category of “Things That Just Won’t Happen.”
As a participant in the standards process, my interests were in creating a standard that increased speed for our users and ensuring that everything we sold in the “emerging standard” phase of the market would be software upgradable to meet the requirements of the final ratified version of the standard. My interests were quite common, and create in effect a self-fulfilling prophecy. If the developers of the standard want the changes between the “interoperability-certified draft standard” and the “final standard” to be small, they will be.
Or, the shorter version: We standardize what we are building. The standard reflects the technical consensus that created interoperability at the beginning of the cycle, so there will not be major changes during the standards process. Buy new products when you need new capabilities, not when a publication milestone is reached.