by Darius Kazemi, August 30 2019
A proposal from IBM
RFC-242 is titled “Data Descriptive Language for Shared Data”. It's authored by Lois Haibt and Alvin P. Mullery of IBM's Thomas J. Watson Research Center.
The technical content
This is a paper about shared data on the network, and it's in the context of the IBM experimental project Network/440, which is the topic of RFC-187.
The authors divide data representation problems into three categories:
- bit-level representation and encodings
- representation of collections of data
- representation of high-level structures like indexes, cross-references, and so on
IBM's developed a “
descriptive language” for representing these kinds of data problems. They claim that their language simplifies programming when compared to the RAND Form Machine.
Their system not only describes data types (like floating point, string, boolean) but also groupings of data into tree-like structures, and tagging of data with arbitrary properties over arbitrary fields. So a chunk of data may have properties that overlap with other ranges of the data.
This language seems more complex than the RAND Form Machine, and is coming coming outside of the established NWG community. IBM, for being a gigantic computer and software company, hasn't attended many NWG meetings to date and I think their proposals suffer for it. While the Network Working Group is open to all, it seems like you can't just lob proposals over the wall and expect people to adopt them or even read them if you haven't first made some human connections.
Looking forward in the record, this RFC doesn't seem to make much impact on the NWG community.
Lois Haibt began working at IBM in 1955 and was the only woman on the FORTRAN team at IBM in the 1970s. There is a brief biography and a very long interview with her at the Engineering and Technology History Wiki.
I couldn't find much on Alvin Mullery, but I liked this note he wrote to his alumni magazine at Brown University chastising them for calling the 1970s the “prehistoric” age of computing.
How to follow this blog
I'm Darius Kazemi. I'm an independent technologist and artist. I do a lot of work on the decentralized web with ActivityPub, including a Node.js reference implementation, an RSS-to-ActivityPub converter, and a fork of Mastodon, called Hometown. You can support my work via my Patreon.