Reflex exercises: Getting startedPosted on September 19, 2017
Getting set up with Nix
This is all set up with Nix. If you don’t have Nix installed, you can get going with the
try-reflex script from the
> git clone https://github.com/reflex-frp/reflex-platform > cd reflex-platform > git checkout b365d0b3f5 > ./try-reflex > exit
and then you’ll be able to use the
work-on script later on, which is pretty handy.
(The specific commit is referenced there so that it matched the version of the
reflex-platform that I used in my Nix files)
You could also get Nix set up by doing:
> sudo mkdir /nix > sudo chown myuser /nix > curl https://nixos.org/nix/install | sh
as per our post on Getting started with Nix.
You’ll probably also want to run:
to set up your paths.
There are some platform-specific notes on the
reflex-platform page that imply that things might be broken on Arch Linux and things might need some tweaking on Linux Mint.
I’ve also had to fiddle around to get Nix working on OS X in the past, but runing
try-reflex and adding the binary caches worked for me when I tried it. I have vague memories of using homebrew to install things that were missing from the host environment when I last had trouble, so that might be a useful thing to try. Let me know if you have problems and I’ll try to find a way to help.
Binary cache setup
You probably want to get hold of some of the dependencies in binary form, rather than getting hold of them in source form and building them. It’ll save you a lot of time.
If you’re on NixOS, you can add the
reflex binary caches by adding the following to
nix.binaryCaches = [ "https://cache.nixos.org" "https://nixcache.reflex-frp.org" ]; nix.binaryCachePublicKeys = [ "ryantrinkle.com-1:JJiAKaRv9mWgpVAz8dwewnZe0AzzEAzPkagE9SP5NWI=" ];
sudo nixos-rebuild switch.
If you’re not on NixOS, you can add the following to
binary-caches = https://cache.nixos.org https://nixcache.reflex-frp.org binary-cache-public-keys = ryantrinkle.com-1:JJiAKaRv9mWgpVAz8dwewnZe0AzzEAzPkagE9SP5NWI=
and you should be good to go.
Getting set up with this tutorial
Once you have done that, or if you have Nix installed and the
reflex binary caches setup already, check out this repository and open a
nix-shell in the exercises directory:
> git clone https://github.com/qfpl/reflex-tutorial > cd reflex-tutorial/code/exercises > nix-shell
Poking around with the first exercise
There is a script in here for each of the exercises (and more will be added as the series continues).
For now, run:
exercises directory and then visit
http://localhost:8080 in your browser (although there is currently a bug in jsaddle-warp affecting Firefox).
This script is running
ghcid along with some magic to trigger a reload in the browser tab when the Haskell or CSS files change.
In this case the Haskell looks like this:
but you don’t need to understand any of the
reflex specific bits.
If you can spot the
Text literals in there then you are ready for this exercise.
The CSS looks like this:
You should open these files in an editor and tinker with the strings in the Haskell program and the color in the CSS class.
After you save you’ll either get errors in the terminal running the
ex00.sh script or the browser tab will reload with the new content.
Click around in the browser tab after each save to convince yourself that this is happening.
This is still fairly new technology, I’m new to using it, and I’m not sure if what I’m doing behind the scenes counts as abusing the technology. For any of the above reasons, you might have to restart the script or manually reload the tab from time to time. I’ll try to make this more reliable over time.
There might be some worrying looking output in the terminal that is running the script, along the lines of
threadWait: invalid argument (Bad file descriptor) or
ConnectionClosed, but those messages usually won’t cause trouble.
The goal of this exercise was to get you comfortable with the workflow, but some people might need more than that to feel a sense of accomplishment.
If that’s the case then I’m happy to accommodate you: your new goal is to edit the above files and to try to end up in a state where you have something in your browser that behaves like this: