12 releases
0.3.8 | Oct 5, 2023 |
---|---|
0.3.7 | Oct 5, 2023 |
0.3.6 | Sep 28, 2023 |
0.3.0 | Aug 25, 2023 |
0.1.1 | Aug 7, 2023 |
#9 in #watchexec
250KB
5.5K
SLoC
Note: ghcid-ng
has been renamed to ghciwatch
Please use the ghciwatch
crate instead.
ghcid-ng
The next generation of ghcid
, a ghci
-based file watcher
and recompiler. ghcid-ng
watches your modules for changes and reloads them in
a ghci
session, displaying any errors.
Why a reimplementation?
When we started working on ghcid-ng
, ghcid
suffered from some significant
limitations. In particular, ghcid
couldn't deal with moved or deleted
modules, and wouldn't detect new directories because it can't easily update
the set of files being watched at runtime. We've also seen memory
leaks requiring multiple restarts per day. Due to the ghcid
codebase's
relatively small size, a reimplementation seemed like a more efficient path
forward than making wide-spanning changes to an unfamiliar codebase.
Why Rust?
Rust makes it easy to ship static binaries. Rust also shares many features with
Haskell: a Hindley-Milner type system with inference, pattern matching,
and immutability by default. Rust can also interoperate with
Haskell, so in the future we'll be able to ship ghcid-ng
as a
Hackage package natively. Finally, Rust is home to the excellent cross-platform
and battle-tested watchexec
library, used to implement the
watchexec
binary and cargo-watch
, which solves a lot of the thorny problems
of watching files for us.
Why not just use watchexec
or similar?
Recompiling a project when files change is a fairly common development task, so
there's a bunch of tools with this same rough goal. In particular,
watchexec
is a nice off-the-shelf solution. Why not just run
watchexec -e hs cabal build
? In truth, ghcid-ng
doesn't just recompile the
project when it detects changes. It instead manages an interactive ghci
session, instructing it to reload modules when relevant. This involves a fairly
complex dance of communicating to ghci
over stdin and parsing its stdout, so
a bespoke tool is useful here.
Dependencies
~24–56MB
~1M SLoC