Are people here generally aware of Cell lang? I se...
# thinking-togetherd
Deklan Webster
07/02/2019, 12:38 AMAre people here generally aware of Cell lang? I searched on here and saw only a couple comments about it (in the history that I can see). I found it last night while searching for langs with reactivity built-in. It references stuff that I'm only familiar with from this community: Out of the Tarpit paper, importance of relational model outside SQL. http://cell-lang.net/
👍 4
😍 1
s
Sébastien
07/02/2019, 6:31 PMAnd the Cell compiler is implemented in cell! https://github.com/cell-lang/compiler/tree/master/src
d
Doug Moen
07/02/2019, 7:24 PM"Out of the tar pit" is an essential paper: http://curtclifton.net/papers/MoseleyMarks06a.pdf
Doug Moen
07/02/2019, 7:29 PMThe Cell language (and the two presentations I just linked to) fascinate me for what they suggest about how to make programming better, and for what they imply about programming language design.
Doug Moen
07/02/2019, 7:37 PMI'm trying to use these insights to design the Curv programming language. An unsolved issue for me is, what mechanisms should be provided for data abstraction?
Doug Moen
07/02/2019, 7:44 PMI've been using OOP for many years now. I now see OOP practice as a primary source of unnecessary software complexity. The Tar Pit paper provides a lot of backing for my intuition: it mentions State, and Unnecessary Data Abstraction (which is most data abstraction), as common sources of complexity.
Doug Moen
07/02/2019, 7:48 PMThe Tar Pit paper makes this amazing assertion: Power Corrupts. When I was young, I thought that I wanted programming languages to be as powerful as possible. When I first encountered Continuations in Scheme, I thought "Wow, this is really powerful!". Of course, continuations are generalized gotos, and today, I think they are a terrible idea.
Doug Moen
07/02/2019, 7:57 PMSo now, look at the Cell data model (http://cell-lang.net/data.html). Note what's missing. First, there are no mutable objects, no object references, no pointers. Pointers are the GOTOs of data structures (and remember, Power Corrupts). In the Cell data model, all data is hierarchical. There are no cycles, no aliasing. This limits complexity.
d
Deklan Webster
07/02/2019, 8:28 PMI ought to read that paper so I can understand all this better. One thing I want to say: the documentation for Cell is excellent. How did you all find out about Cell? It seems very obscure. I can't even find the creator's name
d
Doug Moen
07/02/2019, 8:56 PMThe Cell source code is "Copyright (c) 2017-Present cell-lang", so I think the author wants to remain anonymous.
🤐 1
2 Views