Some debug build changes

[Nutomic]

Not sure why the dev dockerfile builds a release, makes more sense like this. The rm call doesnt really do anything that I can tell (it definitely wont save space like this).

For the update file, I added sudo because of this. But I can remove that and run sudo ./docker_update.sh if you prefer that way. I also removed the -d flag so logs are easier to see. The more I think about this, the more I feel we should just get rid of this file and put the command into the readme...

Last change is adding the intellij folder to .gitignore. Btw, I'm curious what setup/IDE you use for development. Intellij + Docker seems cool, but not perfect.

[dessalines]

This is the dev docker, but doing a --release still makes sense because this is the dockerfile that builds the prod images. It only takes a few extra seconds, and all the deps are cached anyway. Once it builds once, rebuilds from code changes only take a few seconds (as long as docker is properly saying "using cache" for each step.

The building a fake project with the Cargo.toml and the rm are necessary for some weird reason for the rust caching, which people had to do in a strange way because rust has no --dependencies only build step for docker's cache. Here's where ppl show how to do that, and where this is from: rust-lang/cargo#2644

[Nutomic]

Ah okay, it would be nice to add that link as a comment in the Dockerfile then. I would have thought that release mode would be much slower to build, and also make debugging more difficult. But you obviously know this better than me, so I will just close this.

[dessalines]

We can reopen this, and I can make an additional one that keeps the --release for the deploy.

[Nutomic]

Oh, I didnt even realize that the same Dockerfile is used for debug and release. I'm not sure if a seperate one is needed, but if not, lets at least move the Dockerfiles outside the dev folder ;)