Lambda the Ultimate

I've wrote a few of these kind of systems, and while the code is elegant, debugging that code is a pain, even printf is of limited use. Linguistic constructs that can't be supported well in an environment will just languish, data binding is a good example where it could go both ways.

A membrane metaphor generally pervades much of my approach to computing problems: separated spaces that talk to each other while holding inconsistent world views. I tend to mistakenly assume others hold the same unspoken basic perspective. My view of actors and concurrent processes is mainly one of interacting cells that might not share resources while being able to message one another.

(Lately my day job involves adding incremental config for daemons managing key exchange for encryption, so there's a membrane metaphor at multiple levels, in IPsec endpoints as well as tunnel managing intermediaries. A common static world view is not the norm.)

What do you think of them as an abstraction?

Papers about membranes I look at now go almost directly into math I don't care about. So I don't know if you mean just the basic metaphor, or also various intricate scaffolds built atop the general idea.

I guess P Systems are named after Gheorghe Păun who developed them in 1998, since generalized into membrane systems using the biology inspired cell metaphor.

There's an earlier similar effort called boundary logic based on Spencer Brown's Laws of Form, axiomatized separately by Bricken and Kauffman. It looks like Wikipedia talk pages are trying to decide where boundary logic goes in math articles. In the early 90's there was some cross disciplinary talk about using boundary logic to apply biology metaphors to things like virtual reality designs -- see the 1993 VEOS project (William Bricken and Geoffrey Coco). Seems to me this is an earlier incarnation of the same sort of membrane metaphor applied to computing. I spent some time in 1994 talking to a guy about virtual reality design and boundary logic, because I was still a little obsessive about universal spatial metaphors at the time.

The main fruitful takeaway is that local differs from global, and that local has a nested cell structure when there is order as opposed to chaos. But most people are not good at allowing more than one global context to exist in mental models. So theory is almost totally undeveloped, except to the extent some folks like actor oriented systems when they intuitively understand something is going on with membranes and local contexts.