fix missing hyphen from error messages

[avidrucker]

PR to resolve missing hyphen bug reported here: #275

As per this screenshot, this two character edit seems to fix the bug. Are there any tests I can run to confirm that there are no regressions?

[avidrucker]

@mhuebert Any advice on how best to move forward? I'm wondering if there is a process/step I need to take next for the PR to get reviewed

[daveliepmann]

hey, thanks for contributing!

I'm a little concerned that this might cause a regression related to 5404eaf — so if you can confirm that the examples in sample-error-messages still work properly, and maybe add an example for this use case (minimal repro seems to be (nth [:a] -1)?), then we should be good to go.

[avidrucker]

1. Is there a recommended way to confirm "proper" working of sample-error-messages? For example, just looking for errors in the console, or, running the bug catching notebook?

Some tests to ensure "proper" behavior would be ideal, and also, I'm not sure where what is "proper" is determined & written down (i.e. a design doc, specification, etc.) ...

2. Is there a design doc or specification that has been written down anywhere, or is it just in the minds of the maintainers?

[daveliepmann]

1. Is there a recommended way to confirm "proper" working of sample-error-messages? For example, just looking for errors in the console, or, running the bug catching notebook?

Just looking for errors manually, yeah. The last time I worked with this I got the CLJS REPL working and used (map (juxt identity tokenize) sample-error-messages) to make sure the result kept relevant info & wasn't misleading.

2. Is there a design doc or specification that has been written down anywhere, or is it just in the minds of the maintainers?

No spec, no. At this point I'm worried about the parts which have fallen out of my head :D