bullet_train 1.27.0 → 1.28.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5641e1bba44d5adf9523e1fd45e19d2eea108328bb5feb2b23443e20a8ea6113
-  data.tar.gz: 05e864b4627de4d59b9fbaa9b5471281bb6f23955404070032a7adcd40dfd161
+  metadata.gz: 8857c92fdfe34adf9b1ea9713b8706e0ae9ac89eb9ed16ec775f57066fb507eb
+  data.tar.gz: 3bb45989b17d76203fda0666a52c986f39bf6f0d4f045d7a730f36479ab5efed
 SHA512:
-  metadata.gz: 071b3ca1424d792da9af1f3a15991e5efc9d0e9ccf10db1bdc14a1eee53db63e59a8be250fce975f1beab86290d48ccd3e1f3b9e4201dd0e48a49af21e14a43a
-  data.tar.gz: 2fe6a3b4638c964af074eb0978b867e85e652fc25e0079dfcbdaef1176a38f1739d4cd6441739870173f7e4a5f306bd0176d20e6d95348cb9815c1f102ed9a01
+  metadata.gz: 05a88ce952824e3c9272f7f893356f9ca7c54654f713eb1f720ae6b5f0e1541e46ecb37ea542989964b73bae7b3938fde1deca5fb0893d9c7fb71f8b4937979b
+  data.tar.gz: 1484085ca14bed023f4f76b217eb7ee3b4319c754c682518427f258c1f46d54af553a98932d99e170c227218cf0ca539a82df3081acb4a8883d1ce1aa7320549

data/docs/docker.md

@@ -0,0 +1,94 @@
+# Using Docker with Bullet Train
+
+We publish a number of pre-built Docker images, and we include a couple of `Dockerfile`s in
+new (and upgraded) Bullet Train projects to help you get started.
+
+
+## Pre-built Images
+
+The images that we publish are layered to give us flexibility in how final images are produced.
+
+* `bullet_train/base` - This is the most basic image, and it contains only _runtime_ dependencies
+  that are needed for any/all reasonable use of a Bullet Train app. It contains things like `ruby`,
+  a postgres client, and image manipulation libraries.
+* `bullet_train/build` - This image builds upon `base` and included dependencies needed to build the
+  app, but not to run it. Specifically it adds `node` which is required for building assets, but
+  is not needed at runtime.
+* `bullet_train/dev` - This image starts with `build` and adds dependencies that are needed in
+  development mode, but that aren't needed in production.
+* `bullet_train` - This image contains the full starter repo and is ready to run. You generally
+  won't use this image, but if you wanted to give Bullet Train a quick test run it might be useful.
+
+The `Dockerfile`s for these building-block images are in the [`dockerfiles` directory of the `core` repo.](https://github.com/bullet-train-co/bullet_train-core/tree/main/dockerfiles)
+
+These images are versioned along with the gems, and they're built and published as part of our release process.
+
+You can find the packages here: <https://github.com/orgs/bullet-train-co/packages>
+
+You can install one of the images directly by doing something like this:
+
+```
+docker pull ghcr.io/bullet-train-co/bullet_train/base:1.27.0
+```
+
+And in a `Dockerfile` you can do this:
+
+```
+FROM ghcr.io/bullet-train-co/bullet_train/base:1.27.0
+```
+
+## Application Images
+
+We include two `Dockerfile`s in the starter repo:
+
+* `Dockerfile` is intended for building production images. Out of the box it should produce an image that is ready
+  to deploy based on the current version of your app.
+* `dev.Dockerfile` is intended to be run locally.
+
+## `Dockerfile`
+
+The main `Dockerfile` in a Bulle Train app is a two-stage file that's intented to produce a production-ready image.
+
+The first stage is based on `bullet_train/build`. In this stage there is a block where you can install any native
+**built-time** dependencies needed by your app. (Most apps probably won't need to install anything in this stage.)
+
+The second stage is based on `bullet_train/base` and is intended to be the smallest possible image that can be
+shipped to production. In this stage there is a block where you can install any custom **runtime** dependencies you need.
+
+## `dev.Dockerfile`
+
+`dev.Dockerfile` is intended to be used for running your app locally within Docker. This file uses `bullet_train/dev` as
+the base image and gives you a block where you can install any runtime and development dependencies.
+
+This file assumes that the local copy of the application will be mounted into the image. The convenience scripts below handle
+that for you, and then your local changes will be reflected into the Docker image immediately.
+
+To make easy use of `dev.Dockerfile` we've included a number of convenience scripts in `bin/docker-dev`
+
+```
+bin/docker-dev/
+├── build - Build the image. Usually only needed after changing dependencies in `dev.Dockerfile`.
+├── console - Run `rails console` inside Docker.
+├── create - Create your development application container.
+├── destroy - Destroy the development application container and image.
+├── entrypoint - Used by `dev.Dockerfile` to start the rails server in your container.
+├── shell - Get a bash shell inside your development application container.
+├── start - Launch your development application container.
+└── stop - Stop your development application container.
+```
+
+To get started with `dev.Dockerfile` you'd typically do this sequence:
+
+```
+bin/docker-dev/build
+bin/docker-dev/create
+bin/docker-dev/start
+```
+
+By default `bullet_train/dev` and `dev.Dockerfile` use the `host.docker.internal` virtual
+host to allow the app to connect to postgres and redis running on the host machine.
+
+In the future we plan to add a Docker Compose configuration that will allow you to run
+postgres and redis in their own containers.
+
+

data/docs/index.md

@@ -15,6 +15,7 @@
  - [JavaScript](/docs/javascript.md)
  - [Style Sheets](/docs/stylesheets.md)
  - [Internationalization](/docs/i18n.md)
+ - [Docker](/docs/docker.md)
 
 ## Developer Tools
  - [Super Scaffolding](/docs/super-scaffolding.md)

data/lib/bullet_train/version.rb

@@ -1,3 +1,3 @@
 module BulletTrain
-  VERSION = "1.27.0"
+  VERSION = "1.28.0"
 end

data/lib/bullet_train.rb

@@ -180,7 +180,15 @@ def silence_logs?
 end
 
 def openai_enabled?
-  ENV["OPENAI_ACCESS_TOKEN"].present?
+  if ENV["OPENAI_ACCESS_TOKEN"].present? && !defined?(OpenAI)
+    Rails.logger.warn "OpenAI access token is set, but the OpenAI gem is not loaded. Please add the 'ruby-openai' gem to your Gemfile to enable OpenAI features."
+  end
+
+  if !ENV["OPENAI_ACCESS_TOKEN"].present? && defined?(OpenAI)
+    Rails.logger.warn "OpenAI access token is not set, but the OpenAI gem is loaded. Please set the OPENAI_ACCESS_TOKEN environment variable to enable OpenAI features."
+  end
+
+  ENV["OPENAI_ACCESS_TOKEN"].present? && defined?(OpenAI)
 end
 
 def openai_organization_exists?

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: bullet_train
 version: !ruby/object:Gem::Version
-  version: 1.27.0
+  version: 1.28.0
 platform: ruby
 authors:
 - Andrew Culver
@@ -598,6 +598,7 @@ files:
 - docs/billing/stripe.md
 - docs/billing/usage.md
 - docs/desktop.md
+- docs/docker.md
 - docs/field-partials.md
 - docs/field-partials/address-field.md
 - docs/field-partials/buttons.md