Coding da Vinci is the first German open cultural data hackathon. Founded in Berlin in 2014, Coding da Vinci is a project that encourages and supports GLAM institutions (galleries, libraries, archives, museums) to share the digital records of their collections online under open licenses. It achieves this by making an extended hackathon-style event that brings those datasets (and the curators that manage them) into contact with the creative tech and open data communities, showing what can be possible when this kind of data is made open.
The website codingdavinci.de was launched in early 2014 in advance of the first Coding da Vinci event in Berlin, 26-27th April 2014. At that time, the website was built to communicate a single event, with the possibility that it might become a yearly event in Berlin. Since these beginnings, Coding da Vinci has grown from a single central annual event to a bigger project with multiple regional events, often running in parallel, managed by regional event teams, with growing collections of open datasets and the creative projects they have inspired. High time for a relaunch!
Task management: Tasks are broken down as issues and prioritised/tracked in the "Relaunch 2018" project kanban.
The project was set up using the Composer template for Drupal projects. Installation requires Composer for managing PHP packages.
composer install --no-dev
to download required components.web
subdirectory.../config/sync
. (Setup will cause an error due to Drupal issue #2756331.)<domain>/admin/config/media/file-system
afterwards. Additionally, fill the web/sites/default/files
directory with the files corresponding to the database dump. (Initially, database dump and file directory contents will be provided by exporting those from staging.)<domain>/admin/config/system/matomo
.--no-dev
flag when running composer install
.npm install -g yarn
), run yarn
to download required packages.npm install -g gulp-cli
), you can run gulp
every time SCSS files need to be compiled.yarn build:js
. (Watch Drupal issue #2957390 for removing the package.json
file duplicated in the root at some point in the future.)Drupal nudges administrators about available security updates. When updating Drupal core or when applying a Drupal module’s new major version, a backup of the website’s files should be created. The database may be dumped to a file by running vendor/bin/drush sql-dump
using the drush command line shell. Updating is easiest using composer in combination with drush.
Updates should be applied in a development environment first: Run composer update
for minor version updates or composer require <component>:<new version constraint>
for major version updates. After ensuring that the update works properly, push the changes to the composer.lock
and composer.json
file to the repository.
While just being required for development, updating yarn packages is similar: Run yarn upgrade
for minor version updates or yarn upgrade <package name>@<version constraint>
for major version updates and push the changes to yarn.lock
and the package.json
file to the repository.
In the production environment, pull the updated repository and run:
composer install --no-dev
to synchronize composer components.vendor/bin/drush updatedb
to run potential database updates.vendor/bin/drush cr
to rebuild the cache.Never run composer commands other than composer install --no-dev
in production environment!
The site’s configuration is stored in the repository for having it applied when installing the project. Eventually, site maintainers will change configuration in the production environment. In order for plain development installations to be configured properly, it is recommended to sync back configuration changes made in production to the repository from time to time by exporting the configuration and pushing it to the repository. Configuration may be exported by running vendor/bin/drush cex
, which will overwrite the files in ../config/sync
. Those files would need to be pushed to the repository.
Code should generally adhere to the Drupal coding standards.
Yarn scripts are set up to lint ES6 JavaScript files and SCSS files:
yarn lint:js
to check ES6 JavaScript files using ESLint.yarn lint:scss
to check SCSS files using stylelint.This Dupal project is available as Docker container from GitHub: https://github.com/codingdavinci/relaunch2018/pkgs/container/relaunch2018%2Fcdv
To execute the pre-compiled Docker container run the following command with the variables set for your environment. An example for Docker Composer can be found in docker-compose.yml.
docker run -d -p 8080:80 -P \
--env "MYSQL_HOSTNAME=cdv.example.com" \
--env "MYSQL_DATABASE=mycdvdatabase" \
--env "MYSQL_USER=mycdvuser" \
--env "MYSQL_PASSWORD=mycdvpassword" \
--env "MYSQL_PORT=3306" \
--env "HASH_SALT=myverysecretcdvhashsalt" \
--env "UPDATE_FREE_ACCESS=FALSE" \
--env "FILE_PUBLIC_PATH=sites/default/files" \
--env "TRUSTED_HOST_PATTERNS=\"^localhost\$, ^127.0.0.1\$\"" \
ghcr.io/codingdavinci/relaunch2018/cdv:latest
Please also see web/sites/default/settings.php for additional information.
Environment variable | Description | Example |
---|---|---|
MYSQL_HOSTNAME | Database connection. Server host name. | cdv.example.com |
MYSQL_DATABASE | Database connection. Name of the database. | mycdvdatabase |
MYSQL_USER | Database connection. Login name for database. | mycdvuser |
MYSQL_PASSWORD | Database connection. Password for database. | mycdvpassword |
MYSQL_PORT | Database connection. Server Connection Port. | 3306 |
HASH_SALT | Salt for Drupal's one-time login links, cancel links, form tokens, etc. | myverysecretcdvhashsalt |
UPDATE_FREE_ACCESS | Access control for update.php script. | FALSE |
FILE_PUBLIC_PATH | Public file path. | sites/default/files |
TRUSTED_HOST_PATTERNS | Trusted host configuration. | "^localhost\$, ^127.0.0.1\$\" |
HTPASSWD_USER | If HTPASSWD_USER and HTPASSWD_PWD are set, ... |
user |
HTPASSWD_PWD | the user will get a Basic HTTP Auth request when accessing the website . | mypassword |
HTPASSWD_GREETING | Customized HTTP Auth greeting | Hello, welcome! |
UPDATEDB_ON_STARTUP | If set to yes , drush updatedb will be executed once on container start. Default: no |
yes or no |
CACHEREBUILD_ON_STARTUP | If set to yes , drush cache-rebuild will be executed once on container start. Default: no |
yes or no |
USE_REDIS | Use Redis as memory cache. Default: no |
yes or no |
REDIS_MAXMEMORY | Set a memory usage limit for Redis to the specified amount of bytes. Default: 1gb |
1gb or 100mb |
Run in the folder with Dockerfile
:
docker build -t cdv .
And start Docker container with:
docker run -d -p 8080:80 -P \
--env "MYSQL_HOSTNAME=cdv.example.com" \
--env "MYSQL_DATABASE=mycdvdatabase" \
--env "MYSQL_USER=mycdvuser" \
--env "MYSQL_PASSWORD=mycdvpassword" \
--env "MYSQL_PORT=3306" \
--env "HASH_SALT=myverysecretcdvhashsalt" \
--env "UPDATE_FREE_ACCESS=FALSE" \
--env "FILE_PUBLIC_PATH=sites/default/files" \
--env "TRUSTED_HOST_PATTERNS=\"^localhost\$, ^127.0.0.1\$\"" \
cdv