« first day    last day (660 days later) » 

12:14 AM
So is this a golfing language or not?
 
@ThomasKwa It is not.
 
Why is it tokenized then?
 
@ThomasKwa It isn't, that's just something that he was thinking about adding
 
Hmm.
 
12:17 AM
Pytek will be a full language. However, there will be a "compiler" that does the tokenization for golf-related purposes.
There will be a "version" that could be written with CP437 or whatever.
 
 
3 hours later…
3:44 AM
I think I'll use ;; for closing a bunch of parens/brackets and ;n; for specifying how many. ; alone will separate statements, as usual.
 
oooh, that's good
 
Great, that's settled!
room topic changed to Pytek: Discussion for this to-be-developed language (no tags)
I'm considering \func[named args](args) syntax.
 
4:28 AM
@El'endiaStarman you better start making a parser for this before your imagination makes promises you can't keep :P
 
So how will this work? Compiled or interpreted? JIT? Transpiled to Python?
Whatever you do I highy encourage you to consider making a transpiler to asm.js or some way of interacting with PNaCl. I've been looking into both of these technologies and they seem like promising ways of making language that works in and out of the browser
*a language
Anyways do tell me when you have some codez I'd love to contribute
 
4:48 AM
I plan on using PyPy.
 
So transpiling to Python?
 
I guess so.
My main concern isn't program speed, but rather programmer speed.
 
5:02 AM
@quartata It'll probably be two weeks before I start any real coding. I want to finish projects this week, and then the next week will be N Battleship. The user "cat" is really eager to see it up and running.
In the meantime, I'll continue figuring stuff out about the language. :D
Oh, I'll go ahead and note: we must do unit tests for this language.
Plus, I'll think about auto-generated documentation.
 
 
12 hours later…
5:31 PM
@El'endiaStarman PyPy?
 
Okay.
 
6:26 PM
@El'endiaStarman I did a little more research and it looks like if you use PyPy to make a language it'll be JIT compiled
So that's cool
 
 
3 hours later…
9:13 PM
Idea I had while getting ready to shower: an expression with _ in it is automatically a function, and they can be identified like _1, _2, _a, _ab, _b, that sort of thing.
The reasoning is that it's kinda annoying to have to type lambda x:x+2 in Python, for example. Why not just _+2?
Trying to think of an example where this would be useful.
Ah, also, regarding \track, I think I could do it this way:
Keeping in mind that \track needs a sequence, a condition, and an action.
\track[condition](sequence){action}
\track[_>\max($)](\collatz(200)){print(_)}
I think $ could mean "itself".
So like x = [0,1]+\addByDiffs($) would generate the Fibonacci sequence.
And again, like _, identifiers can be appended to distinguish them.
Ah-ha, okay, I have an example for _'d identifiers.
\map( _1**2+_2, \zip( [0,1,2,3], [6,7,8,9] ;; => [6,8,12,18]
$ identifiers would be useful when doing something like a,b = $2,$1+$2.
...is it me or did I just write the Fibonacci sequence?
I must make this possible.
Oh, regarding how the interpreter will know which is which, identifiers will be sorted lexicographically and the first "thing" will go to the first one, second to second, etc.
 
 
1 hour later…
10:36 PM
I'm really gonna have to deliberate and figure out details of how a,b = $2,$1+$2 works, because $1+$2 is integer addition but I would expect to be able to loop through a. Like, a would be the infinite list [0,1,1,2,3,...].
 

« first day    last day (660 days later) »