Could "data plumbing" be a unifying thread in PL/systems design?
A large part of the value of {type,category} theory, monads & folds &c, is simply in capturing plumbing: the boring, repetitive patterns that get data from point A to point B.
3
6
18
Conversation
In general I would love more systems-design thinking in PL/type system research. This could link in nicely. Ie. how do we support the challenge of upgrading and migrating of code on existing, large, stateful systems.
1
Could also help in terms of deploying and iterating on language designs as well, in the presence of existing code. For example you can think of source code as a distributed, hard to upgrade distributed system.