3 ms·
I have the C book... It is quite antiquated and has a lot of ugly pointer math in the tokenizer/lexer beginning parts. It does sound like the ML book is a bette
by Gracana 10y ago
I have the C book... It is quite antiquated and has a lot of ugly pointer math in the tokenizer/lexer beginning parts. It does sound like the ML book is a better choice.
- nickpsecurity 10y ago"It is quite antiquated and has a lot of ugly pointer math in the tokenizer/lexer beginning parts." It's possibly C code written by a guy that mainly does ML with a whole career kicking ass in ML languages and provers. That was my hypothesis when I saw both books available. I got a copy of the ML one instead. Appreciate confirmation it was a good choice. ;)
- mhh__ 10y agoYeah. AFAIK The C version only exists because ML is too niche for mainstream publishing. ML actually works as decent pseudocode(Or at least easier to mentally transpile to other langauges, unlike e.g. Haskell) as well, so the C/Java versions are basically pointless.
- nickpsecurity 10y agoThat's what I was guessing. Especially after reading the article below and seeing lots of compiler/prover authors go with ML's. https://news.ycombinator.com/item?id=14123100 https://news.ycombinator.com/item?id=14123100