[Uri’el-2012]

Were you at Piscataway or Morristown?

I spent the eighties at Bell Labs in Denver.

We built PBXs up to the size of a CO.

[Myrddin]

Sounds like fun. I was offered a D55 promotion and transfer to Bellcore in Piscataway if I wanted to stay with the PacBell. I opted to leave for green pastures with a local DoD contractor. That turned out to be a good move. I really didn't want to move to New Jersey. My new employer has been very good over the last 18 years.

There was a cousin to COSNIX call MERT that was used in the ACD devices. That was simply amusing reading. I never had to work on it. When I first started in Toll Equipment Engineering, I read an interesting paper by Jay Goldstein on Directed Hypergraph databases. That was in 1980. By 1986, I was writing specialized database routines for LFACS that were implemented as directed hypergraph databases. It appears to have been implemented using a CODASYL style database grabbing keyed binary blobs. Once the blob was in memory, you effectively did a dynamic "cast" of the blob to the datatype. That provided the necessary offset information to access the fields of the body properties, edges and hyperedges. It was a neat design, but always left me a bit nervous about writing correct code. Thankfully, I managed to have 100% successes with it. I really wasn't enjoying have to be "perfect" every time. It would have been very easy for a simple error to do some serious damage to the database.