I can confirm. This started failing today. Akamai is returning a 403 when retrieving the GPG key using curl or wget.

Make docker-composer specific environment variables overridable.

Update test dependencies.

Run tests using Behat in docker.

Speed up consecutive builds by caching cargo metadata.

Todo:

• [ ] Generate a .env file with a random SESSION_KEY and SECRET_KEY when running tests.

• [ ] Make the Mailgun mock server image a dependency for the Behat image.

• [ ] Not yet clear how to handle CSS + JS assets. At the moment the vendor folder is copied into the firetrack image when it is built. This means that:

  1. We need to execute composer install before building the image.
  2. It is slow when building, but possibly faster to run?
  3. If we make changes we need to rebuild.
  • Easiest is possibly to keep this for now since we are doing development without Docker. We only need it now for running tests on github.
  • Can we run composer install in the entrypoint? Possibly not since all images are built in parallel.
  • Look into multistage builds to download composer dependencies in an earlier stage.

• [ ] Introduce profiles for testing and initial scaffolding: https://docs.docker.com/compose/profiles/

• [ ] Speed up with cargo vendor and compiling dependencies first, refs:

  • https://benjamincongdon.me/blog/2019/12/04/Fast-Rust-Docker-Builds-with-cargo-vendor/
  • https://github.com/rust-lang/cargo/issues/2644

• [ ] We can make a separate docker-compose file for local development, where only postgres and Selenium are containerized.

• [ ] Make a slim production container.

• [ ] Images are very big (check with docker images). Find out how to remove needless cruft from them. https://alexbrand.dev/post/how-to-package-rust-applications-into-minimal-docker-containers/

On a project that uses the mink-selenium2-driver I am getting a fatal error when a Behat step fails before visiting a page:

Fatal error: Uncaught Error: Call to a member function source() on null in /app/vendor/behat/mink-selenium2-driver/src/Selenium2Driver.php:485
Stack trace:
#0 /app/vendor/behat/mink/src/Element/DocumentElement.php(37): Behat\Mink\Driver\Selenium2Driver->getContent()
#1 /app/vendor/lovers-of-behat/screenshot-extension/src/Photographer.php(48): Behat\Mink\Element\DocumentElement->getContent()
#2 /app/vendor/lovers-of-behat/screenshot-extension/src/Listener/FailedStepListener.php(63): LoversOfBehat\ScreenshotExtension\Photographer->snap('1679740540-expe...')

While this is a bug in Selenium2Driver - it should not call methods on NULL - we can prevent this from happening by checking that the Mink session is started before trying to take screenshots

Do not attempt to take a screenshot if the Mink session hasn't started yet.

Merge pull request #5 from LoversOfBehat/check-session-started

Do not attempt to take a screenshot if the Mink session hasn't started yet

Fixes #4

Fixes #4

On a project that uses the mink-selenium2-driver I am getting a fatal error when a Behat step fails before visiting a page:

Fatal error: Uncaught Error: Call to a member function source() on null in /app/vendor/behat/mink-selenium2-driver/src/Selenium2Driver.php:485
Stack trace:
#0 /app/vendor/behat/mink/src/Element/DocumentElement.php(37): Behat\Mink\Driver\Selenium2Driver->getContent()
#1 /app/vendor/lovers-of-behat/screenshot-extension/src/Photographer.php(48): Behat\Mink\Element\DocumentElement->getContent()
#2 /app/vendor/lovers-of-behat/screenshot-extension/src/Listener/FailedStepListener.php(63): LoversOfBehat\ScreenshotExtension\Photographer->snap('1679740540-expe...')

While this is a bug in Selenium2Driver - it should not call methods on NULL - we can prevent this from happening by checking that the Mink session is started before trying to take screenshots

You have xdebug: false in your configuration, this disables XDebug. Try setting it to xdebug: true. Ref https://docs.lando.dev/php/config.html#using-xdebug

PHP 8 compatibility.

CinemaCity is no longer using a separate server for their data API.

Update Drupal to 9.5.5.

Merge pull request #309 from pfrenssen/update-drupal-9.5.5

Update Drupal to 9.5.5.

Checklist

• [X] I have updated to the latest available Home Assistant version.
• [X] I have cleared the cache of my browser.
• [X] I have tried a different browser to see if it is related to my browser.

Describe the issue you are experiencing

The history graph cards and statistics graph cards perform a 300 ms height transition instead of appearing instantly in the right height. This can be visually disturbing on dashboards with a lot of graphs. This behavior seems to have been introduced recently.

history-graph-transition.webm

This happens in both Firefox and Chrome and is caused by a transition on the ha-chart-base .chartContainer element.

Describe the behavior you expected

The dashboard appears instantly without certain cards having to "settle" into their positions.

Steps to reproduce the issue

1. Add some history + statistics graph cards to a dashboard.
2. Open the dashboard.

What version of Home Assistant Core has the issue?

2023.3.1

What was the last working version of Home Assistant Core?

No response

In which browser are you experiencing the issue with?

Firefox 110.0.1, Chrome 110.0.5481.177

Which operating system are you using to run this browser?

Arch Linux 20230308

State of relevant entities

No response

Problem-relevant frontend configuration

No response

Javascript errors shown in your browser console/inspector

No response

Additional information

No response

The history graphs are now sliding open vertically when they first appear. There is a 300ms transition on the .chartContainer causing it. This looks really weird on dashboards that have a lot of history graphs. Is it possible that this merge has caused that behavior?

Add a script to generate random secrets for testing.

Place the mailgun mock server logs in a folder which can be shared as a volume in docker-compose.

Build the base firetrack image first so it is available for the multistage dockerfiles which depend on it.

Reuse the Firetrack binary from the main image.

This results in a small image and we skip a lengthy re-compilation.

Apply security updates.

Provide a simple script to run firetrack cli commands in docker.

Update packages to resolve a buildable tree on the most recent versions of Rust and OpenSSL.

The fix for this has been included in the latest stable release v2.20.1 which came out yesterday.