I had been at Bell-Northern Research (BNR) for just shy of ten months when a new project, Next Generation Node, was started up. The motivation was the terrible realization that implementing support for ISDN, a relatively new telephony communication standard at the time, was going to take 325 person years. This was considered to be completely unacceptable.

The president of BNR pointed out to ISDN management that there was a new group, called the TELOS project, that was designing a new, graphical way to write software for telephony switches. He suggested we investigate whether using TELOS could reduce the amount of effort needed to support ISDN.

Some very senior people were assigned to the project. I was the only newcomer on the team, chosen because I didn’t know the current software very well, so I was unbiased and would presumably question preconceived notions of how telephony software was written. That is, I’d ask the stupid questions. We decided that we could, in fact, use TELOS, to support ISDN provided that we could use loop-around trunks to interwork with the various feature phones BNR supported.

BNR had designed many different types of phones with a variety of features and capabilities. If an organization had different types of these phones, it was expected that they would all work together with the various features the phones supported, within reason. So, if you put a call on hold on one phone, you could pick up that call on a different phone, even if the phone was a different type. However, a call coming into the organization from outside had restrictions on what you could do with that call because it was of an unknown type. For us, we were allowed to treat ISDN phones within the organization the same way as if they were outside the organization. So, we would support limited feature integration with the other phones in the organization as if the ISDN phones were outside the organization. We accomplished this by connecting them all to trunks that would go outside the office, hit the telephone switch, and be sent right back, hence the term “loop-around trunks”. That way, calls from ISDN phones within the organization would look like calls from outside, and so the expectations of feature integration with the other phones in the organization would be much lower..

Since we could support ISDN with reasonable effort under these conditions, we planned to proceed with supporting ISDN via Telos. I still have a floor plan of the new offices we were going to occupy as we kicked off the project.

However, at the last minute, it was decided that we couldn’t interwork with other telephones over loop-around trunks and would, instead, have to support all the features integrations with other telephone types in the organization. With this change, we couldn’t do the ISDN software in reasonable time using TELOS. We said good-bye to our friends in TELOS and began searching for another way to support ISDN, which is a whole other story worth telling one day. We embarked on what became the Generic Services Framework project, or GSF for short. The TELOS team continued without us, but on October 1, 1992, they were spun out of BNR to become a new startup company, ObjecTime Limited. Little did I know that I’d become an employee of ObjecTime on April 25th, 1994.