« first day (890 days earlier)      last day (777 days later) » 

1:27 AM
!!/status
@lyxal I am doing spectacularly.
Guess you don't report review dismissals
That's fine, just good to know
1:51 AM
Gamers, it's with a little sadness that I have to announce that apl won the recent week golf language vote
It was a good run of trolling while it lasted lads
Any gamers stationed to troll apl on WeekGolf will now be re-deployed on other copious trolling efforts
2:49 AM
Get hyper to swap the vyxal bot with the APL bot
Savage
 
1 hour later…
4:00 AM
@lyxal what are we going to use for v3
either Scala or Go
why not rust
4:41 AM
too strict
I like a little flexibility with types
 
3 hours later…
7:52 AM
Alright, important announcement: version 3 will not be written in go. It will still be written in scala
4
I did some initial experimenting with lazylists and go just makes me want to commit frick
Therefore, scala will be used
But we'll have to find a way to do it without needing an overload of directories
For a golfing language that aims for simplicity, the file structure should also be simple
There's no more uncertainty about the language that will be used. I'm firm on Scala because of its functional programming support
That is all
I'll be looking into how to make a scala project with minimal nested directories
8:14 AM
It looks like the minimum needed is
.
└── src/
    ├── main/
    │   └── scala/
    └── test/
        └── scala/
I wonder if src can be renamed to vyxal and if you can just have scala in main without going deeper into a scala directory
8:44 AM
While I still remember - I also have an action plan
An action plan on how to kick-start v3 development for good
1) create a simple txt file of what each element is
Like the first ever style of txt file
2) just get started on the tokeniser
Don't worry about finer details or the yaml until needed
3) less reviews required for v3 PRs
We're in a quiet period, so less will be needed
1 approval should be fine
4) no timelines for completion
We're already "5 months behind schedule" according to the timeline I made in November last year
Which goes to show just how useful it was
(it wasn't useful at all)
That should be enough to get things going
After that, more detailed planning can happen when it needs to happen
We'll also need a way to create actually helpful documentation
More on this later
 
2 hours later…
10:24 AM
Alright about the helpful documentation
If you've ever tried to figure out how the lexer and parser and transpiler work, you'll know that it isn't exactly easy
You have to read through a whole lot of text and then go through the source code to piece together what's happening
I'm thinking for version 3 (before things get too complex) a sort of diagrammatic and simple English guide to the different components of the interpreter could help
Something updated gradually as things are added
That way, it'd be written while the person doing the implementation still knows what's happening
Plus it'd help future contributors know how to approach any code
At this stage it's just an idea I'm thinking of
Maybe I'll work on something for the current interpreter as a proof of concept
10:45 AM
Actually,
CMQ: what would best help you understand how the lexer/parser works?
4
E.g diagrams/written explanation/comments in the source code
 
7 hours later…
6:05 PM
@lyxal those are minimum. you can have no package to not go deeper into directories
@lyxal write it out in antlr or smth

« first day (890 days earlier)      last day (777 days later) »