Fork me on GitHub
#clojure-dev
<
2017-05-15
>
matan05:05:39

just built clojure for the first time, entire build and packaging took just 3 minutes!

alexmiller05:05:25

you can skip the tests with mvn clean package -Dmaven.test.skip=true to cut that down some more

matan05:05:45

Is there any standard or special mechanism by which error message are issued by the compiler?

alexmiller05:05:00

well it throws an exception :)

matan05:05:00

I was considering playing with this aspect and/or with localization

matan05:05:10

No resources file or anything then...

alexmiller05:05:43

compiler exceptions in particular are typically wrapped in a Compiler.CompilerException

alexmiller05:05:04

oh, no resources file

matan05:05:19

okay 👍

matan05:05:30

sorry to bother all bussy developers here

matan18:05:52

What might be the place in the source, where parsing of user code takes place? Is it not this here? https://github.com/clojure/clojure/blob/f572a60262852af68cdb561784a517143a5847cf/src/jvm/clojure/lang/LispReader.java

matan18:05:15

I wonder why it's not been ported to clojure, not that there's anything wrong with Java. Maybe it simply works and no user value by changing it to clojure.

alexmiller19:05:46

that’s it and it has been ported to Clojure in the tools.reader project: https://github.com/clojure/tools.reader

alexmiller19:05:01

but the Java version is still noticeably faster

tbaldridge19:05:03

@matan there's also the bootstrap problem...how do you compile Clojure code if the reader is written in Clojure? There's tricks around that, but they almost all start with having to have a reader in Java, and porting it to Clojure later. So then the question is why port it to Clojure if it would only complicate the build process?

tbaldridge19:05:45

At least that's my view on it all, and the same applies to the compiler, immutable collections, etc.