Comment by mafribe
6 years ago
I understand the problem.
I teach compilers at a big university. And I would love to hire graduates with compiler skills for my startup, but find it difficult.
There are several structural problems that conspire to keep students from acquiring knowledge in low-level programming domain like compilation: a course needs to fit with the rest of the curriculum, and with student interest. I cannot get students interested in lexing and parsing, because they hear all day how exciting deep learning and big data are. Lexing and parsing are deemed solved in the 1960s, a sentiment I disagree with. In addition, classical theoretical computer science (e.g. automata and languages) is rarely taught with enough depths in many universities, so students lack the technical background. You can cover this in a compilers course (I do) but it takes time, and the students see it as an ordeal. Compilers as a subject is not helped by the most widely recommended book being the Dragon Book which is dated and way too long to be useful for beginners. Compare the Dragon Book with the material available as introduction to machine learning ... Many of the existing textbooks are also not covering modern compilation themes, in particular JITs, and compilation for multi-core and GPUs. I'd say there is currently no good undergraduate textbook on compilers. I could probably, with a lot of work, cobble something reasonable together from my lecture notes, but I don't believe in books, I'd like to do a good MOOC, but building suitable software infrastructure requires more work than I am currently willing to put in.
My department tried hard to remove the compilers course, as "no longer relevant", and "too hard". I threatened to resign if it was not kept as a mandatory course. For now this worked.
I never took a compilers class during university study and it is my greatest regret. I've since been self learning and it hasn't been easy.
Thanks for advocating for the importance of compilers. This lack of knowledge has been a limiter for me many times. I'd really like to fill a skill gap and be able to write a DSL or a little language to solve a problem.
From my experience in industry, while hardly any programmer will ever need to write a modern code generator, few things are more useful in practise than being able quickly to design a suitable DSL and implement a corresponding performant and correct lexer and parser. This is not something that can easily be picked up on the side.
In addition, understanding what happens under-the-hood when you program in a high-level language is an eye-opening experience for many students. This sentiment comes up in student evaluations year-in, year-out.
I never understood this. I took the compilers class at CMU, and I loved it. I went on to take a follow-up class implementing a compiler with higher-order types. Meanwhile, most of my classmates avoided compilers altogether.
I've worked in the industry since undergrad, and I've implemented several interpreters for DSLs. Sometimes I was asked to make it, but other times I had the flexibility to dream up a DSL on my own, and it ended up becoming indispensable to users. I've always loved building them, but to this day, parsing is still the most boring part to me.
I recently went through Bob Nystrom's Crafting Interpreters, and it was great fun. I implemented an interpreter in Rust. Material about compiling for GPUs would be interesting. But for me, personally, I really wish there were something updated for how IDEs use languages nowadays, in particular, walking through a good way to implement the language server protocol.
The calculator language that every compiler book starts with should be LSP-first.
A MOOC would be great, but I'd also be happy with a series of blog posts.
Does anyone know of good material bridging the gap between traditional compiler architecture and the LSP?
> Compilers as a subject is not helped by the most widely recommended book being the Dragon Book
Is it? Even I, an outsider [to the modern university courses], see that the baseline quite shifted to the TigerBook (aka Appel - Modern Compiler Implementation in [your choice]).
Is it at all possible that traditionally compilation is taught back to front? I am just a hobbyist, my CS degree did not include anything about compilation other than the theory of types of Grammar. I am the first to put my hand up and say I do not really know what I am talking about.
For me, compilers are interesting because of a. optimisations; at my level simple logic that is obviously advantageous to a CFG. b. interaction with the dependencies and operation of low level hardware; how does it REALLY work.
In the hobbyist pursuit of an understanding I have become belligerent to parsing. Parsing is interesting. But also not really. I would have benefited substantially from an abstract introduction via graphs, analysis and lowering. Both in terms of interest and fundamental understanding.
Thanks. I'm actually very interested in lexing and parsing, because that's probably 99% of the stuff a layman-programmer gets to do with compiler theory in job. I mean not everyone gets chance to write the backend part, but parsing skill is almost used universally.
A fun exercise is writing a lexer generator or a parser generator. It's probably easier starting out with writing a lexer generator like Flex for your favourite language. I recommend using the Flex input format, than you can test your creation using Flex as a testing oracle! This is not something you can do in a weekend.
Thanks mafribe, looking forward to your future book (if there is one) :D
Your HN profile is empty, so there seems to be no good way to check out your institution's program or to contact you.