A lot of questions have been posed about why version 1.1 of the Bluetooth specification was brought out and exactly what the changes were between that and version 1.0b. For all the fine details, the best place to look will always be the Bluetooth adopter's Web site, http://www.bluetooth.org, where you can find the specification itself and all the errata that led to the changes. However, it can be a slow process wading through errata, and if you're not very familiar with the specification, sometimes it's not obvious why and where changes have been made. This article looks at the differences between version 1.0b and version 1.1, and attempts to offer some explanations for the changes.
Why Change at All?
A prime objective for Bluetooth technology is interoperability. Any pair of devices carrying the Bluetooth brand should be capable of working together at a basic level. Any pair of devices supporting compatible Bluetooth profiles should be capable of connecting and using one another's services. This means that users can mix and match devices from any manufacturers, and the Bluetooth brand guarantees that they will work.
Well, that's the theory anyway, but to be able to do that, everybody has to interpret the specification the same way. Any areas of ambiguity could ruin interoperability. The working groups that wrote the Bluetooth specification thought they'd made it unambiguous, but it turned out that manufacturers were interpreting the specification in different ways—and that meant that devices couldn't always work together. Worse still, it turned out that some features couldn't even be relied on to work reliably between two devices from the same manufacturer. Something had to be done!
The Bluetooth Special Interest Group (SIG) has an errata process for fixing problems with the Bluetooth specifications. Any SIG member can submit errata through a Web page. Then an errata working group considers errata and votes on whether to accept them. By the summer of 2000, it was becoming obvious that there were so many errata on version 1.0b of the specification that it was worth going to a whole new version. However, there were concerns that the specification still might not be right. How could the SIG reassure everyone that the next version would be any better than the last?
The Bluetooth SIG organizes testing events called unplugfests, where many manufacturers get together and test their devices against one another. They decided to identify the errata that were most critical for interoperability and then ask everyone to test them at Unplugfest 4, held in November 2000. If version 1.0b plus the critical errata had no serious interoperability problems, then this could be used as the foundation of version 1.1. As it turned out, the unplugfest was a great success, so the SIG went ahead and produced a draft version of 1.1. After a few months, this was slightly modified, and version 1.1 was formally adopted.