freenode/#sicl - IRC Chatlog
Search
10:45:38
no-defun-allowed
There are no hyphens in the other IRC user's name, so I will have to rebut with Hungarian notation. Hello perSplttst.
10:56:45
heisig
no-defun-allowed: It is time to demote splittist to the venerable FORTRAN naming convention. Hello SPLTST :)
11:00:39
beach
Are Fortran and C++ the only languages whose name are not valid identifiers in the language?
11:03:57
no-defun-allowed
There are some languages which enforce snake_casing of variables, so their names would be right out (for variable identifiers).
11:04:18
beach
Better grammar: We say a language is IMPOVERISHED is its name is not a valid identifier in th language. Are Fortran and C++ the only impoverished languages?
11:08:19
no-defun-allowed
I thought I had some: it is bad style in Python to not use snake_case for variables, so a variable named Python would not be idiomatic, and a linter program would probably mention it. Rust gives you a compile-time warning for that. I also hoped that there was a similar situation for Go (but lowercase go is a keyword used for starting a thread), but it's probably more like Python, where another tool tells you off.
11:09:22
no-defun-allowed
However, the names would be idiomatically written if they named classes or types.
11:20:19
ck_
.. like Visual Basic, for example. Also, similar but different: Brainfuck, Whitespace, Mondrian
11:22:58
no-defun-allowed
λ λ . e is probably invalid syntax. But then people who aren't mathematicians write λ actual . λ names . because they understand why you'd want to do that, so no, not really.
11:26:20
no-defun-allowed
You also can't name a slot Self in Self (it expects capitalised words to continue longer message sends).
17:32:49
scymtym
maybe i'm missing something, but wouldn't this be simpler and more versatile: https://github.com/scymtym/Concrete-Syntax-Tree/commit/79078e77cd2b2c1c345f4a38301c594795cf3c52 ?
18:01:08
heisig
scymtym: The old version only cstifies the car, the new version cstifies both the car and the cdr of a cl:cons. Is that intentional?
18:01:37
heisig
Apart from that, the changes look good. Especially since an atom is literally defined as "an object that is not a cons".
18:18:37
scymtym
heisig: thanks for having a look. i think (cst:first CST) should itself be a CST, yes. probably something like (samep (cst:raw (cst:first CST)) (first (cst:raw CST))) should hold
18:21:45
scymtym
hm, i thought CSTIFY wasn't currently used anywhere, but now i found uses in CST-TO-AST. i should make sure that the change does not break those uses