Run the product setup wizards from the plugin JARs by pathob · Pull Request #498 · deftdevs/bootstrapi

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z

Quality Gate failed

Failed conditions
62.2% Coverage on New Code (required ≥ 80%)
11.0% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

pathob · 2026-07-14T11:59:39Z

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables.

The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH).

Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

The plugin JARs now double as setup tools: java -jar drives the product's setup wizard over HTTP so a freshly installed instance can be set up unattended, e.g. from a deployment hook job running next to the application. The CLIs are built on the JDK HTTP client alone, so the plain plugin JAR runs without any further dependencies, and they are configured through BOOTSTRAPI_SETUP_* environment variables. The Crowd wizard is driven completely, including the database step. For Jira and Confluence the database connection must already be configured, as the official container images do; their wizards are driven from the application properties and data step onwards. All CLIs detect an already set up instance and exit successfully, resume at the current wizard step where the wizard supports it, and verify the resulting application state at the end: the wizards answer invalid input with an error page instead of an error status, so only the resulting state proves the setup went through. The Crowd wizard flow is based on the crowd-init script of the ldap-crowd-adapter project (Apache License 2.0, ASERVO Software GmbH). Every CLI is unit tested against a stubbed wizard, and integration tests set up pristine instances from the official container images backed by a PostgreSQL container. The integration tests are excluded from normal builds and run explicitly, see CONTRIBUTING.md.

sonarqubecloud · 2026-08-10T17:44:34Z