Hacker News Reader

Comment by reikonomusha | Hacker News Reader

Super excited to see this! Activity around Common Lisp has been effervescent as of late. The Amsterdam Lisp group just met, the Bay Area Lisp group is active, CL is kicking in Japan, and now this.

Common Lisp may be old, and may not have adopted the fads it couldn’t have foreseen, but it’s still a modern, efficient, high-productivity language that has proven decade after decade its ability to cut through some of the most difficult programming problems, from OS’s on custom hardware to quantum computers.

View on HN

I have alot of projects that I want to work on coming up, and every time I start them, I always want to pick a lisp, but never end up doing it because I just can't find the support that I expect and hope to be present. This might be a bit of a chicken and egg problem, here are some of the things that put me off:

1. The JVM (in the case of clojure) -- the apps I want to write I want to statically compile easily, and have very light footprints (at least at the lower limit). I always wince when trying to deploy a JVM application (elastic beanstalk, graylog, etc).

2. Support for new emerging tech/libraries like gRPC and swagger -- yes I could always write my own software for this, but it just isn't a smart use of my time. Even haskell has easy support for stuff like grpc. Haskell is also really easy for me to trust due to just how safe the language makes me feel.

3. Established and "boring" choice for web server frameworks

4. Just choosing between SBCL (which seems to be the best choice for CL), racket, and scheme (I've done projects in clojure before and it was very enjoyable but I just don't want to bring in the JVM/get involved in that ecosystem)

5. Easy cross-platform binary generation (I saw buildapp and get that I can make FASl binaries, but what if I wanted to compile for armv6 easily from a non-arm system?)

The only language I can imagine that has the performance, simplicity, and deployability benefits that I want seems to be Golang these days. Rust would theoretically also be an option (big fan of increase in expressiveness and de-factor memory safety), but it still seems

Can you tell me why I'm wrong? I really do love lisps, and I know that it's got the expressiveness that fits like a glove (that you made) and with dialects like SBCL the native speed that I want.

The answer may very well be that seeming lack of libraries/resources is my "fault", in the chickend-and-egg sort of sense (someone like me who's interested in the libraries that don't exist in the form I want yet ISN'T investing into the tooling).

[EDIT] - Yeah, just writing this post has made me feel bad I think I'm going to get back into writing common lisp on my next project.

Thank you for the comment! It's definitely not your fault, but more a community fault and general lack of momentum.

Common lisp implementations indeed lag behind mainstream languages in many occasions, and it's really hard to justify it as a choice for something like straightforward web app compared to other combinations like python + django that provide a lot to get up and running very easily + lots of robust libraries.

What's necessary on my opinion is to create a community effort to pinpoint all missing pieces in ecosystem regarding practical applications like webservices and address them by creating corresponding libraries event if it means direct port at the moment.

How bout a lisp that produces go assembly? Wish I had time to work on something like that. :(

Thanks for mentioning Amsterdam lisp meetup. We're very small at the moment but constantly looking for new members and speakers to make it more interesting and attract more people to explore lisp languages.

Reply on news.ycombinator.com