21–30 of 38 posts
I did start to feel Racket’s “wordiness” towards the end - it started to feel a bit like COBOL. I’ve since moved onto Clojure and really appreciate the shorter keywords/function names/fewer parenthesis.
I still miss for/fold though - that thing is an absolute machine.
Earlier quoted context omitted.
I'm doing the learning in racket because the bottleneck is human understanding. That mnist takes 30 minutes per epoch isn't a worry when I don't even know what vector addition should look like.
> I don't even know what vector addition should look like. I think you're trying to imply you're inventing something new and racket enables you to explore... But what I read (as someone with a PhD in deep learning that has worked on sparsity) is you actually don't know the prior art and you're using racket as an excuse to reinvent a whole bunch of stuff that already exists in plenty of mature libraries in more mundan…
It took me 20 minutes.
Tell me how you'd do that in cpp?
I have been writing Elixir professionally for the bulk of my career. (It’s been pretty awesome!) Despite this, I consider Racket to be my native language because it’s so easy for me to think in Racket. It’s the perfect bridge between my brain and the problem domain.
I recently picked up Essentials of Compilation by Jeremy Siek, perhaps his book and lectures will make something click for me.
In the meantime Julia has been my happy medium for a “lisp”.
But it's really hard to give up the raw power of Common Lisp and ability to write low-level code. Even if the language and tooling is all more complicated.
Racket gets in your head like nothing else. Once you learn it, (Scheme does this too) you get x-ray vision to see every other language modulo syntax. I have been writing Elixir professionally for the bulk of my career. (It’s been pretty awesome!) Despite this, I consider Racket to be my native language because it’s so easy for me to think in Racket. It’s the perfect bridge between my brain and the problem domain.
Earlier quoted context omitted.
> I don't even know what vector addition should look like. I think you're trying to imply you're inventing something new and racket enables you to explore... But what I read (as someone with a PhD in deep learning that has worked on sparsity) is you actually don't know the prior art and you're using racket as an excuse to reinvent a whole bunch of stuff that already exists in plenty of mature libraries in more mundan…
I once replaced IEEE 754 floating point numbers in a model by balanced ternary floating point numbers. It took me 20 minutes. Tell me how you'd do that in cpp?
Earlier quoted context omitted.
I once replaced IEEE 754 floating point numbers in a model by balanced ternary floating point numbers. It took me 20 minutes. Tell me how you'd do that in cpp?
lol the same way we implement all of the reduced precision fp8, fp4 types today: by storing them in the corresponding uint: https://github.com/ggml-org/llama.cpp/discussions/15095
>Unlike their binary counterparts, posits and takums, tekums simultaneously accommodate both ∞ and NaR, while retaining the simplicity of negation by flipping the underlying trit string. Perhaps most strikingly, tekums enable rounding by truncation, a property that eradicates at a stroke some notorious problems of rounding in binary arithmetic: double rounding errors, cascading carries in hardware, and the attendant inefficiencies.
Earlier quoted context omitted.
lol the same way we implement all of the reduced precision fp8, fp4 types today: by storing them in the corresponding uint: https://github.com/ggml-org/llama.cpp/discussions/15095
Balanced ternary fp is not a reduced precision type of binary fp: https://arxiv.org/abs/2512.10964 >Unlike their binary counterparts, posits and takums, tekums simultaneously accommodate both ∞ and NaR, while retaining the simplicity of negation by flipping the underlying trit string. Perhaps most strikingly, tekums enable rounding by truncation, a property that eradicates at a stroke some notorious problems of round…
Yes I can read very well - can you?
> ... by storing them in the corresponding uint
I have spent a lot of time in the PL space and I just cannot wrap my head around the love of Racket. I absolutely love the innovation that comes out of the community though! The gradual types system work, nanopass, embeddable DSLs are all fabulous so I like to keep tabs on what they are cooking up. I recently picked up Essentials of Compilation by Jeremy Siek, perhaps his book and lectures will make something click f…
It was the first mainstream attempt to an ALGOL style Lisp.
https://opendylan.org/documentation/index.html
Pity it didn't went down well at Apple.