0.5.0

Incidents now say when the next news comes, and what happened once they are over. A restarting server no longer floods your channels, and each service waits as long as you choose before it shows an outage.

What's new

• Next update time. When declaring an incident or posting an update, write a delay ("30 min", "1 h") or a time ("14:30"). The incident's page and the status banner say when the next update is due, then how late it is once the time has passed.
• A bell for editors. The masthead counts unread notices: an update that is 5 minutes late (also by email, once, with a mail server set), a service the checks find down, a destination that refuses its messages.
• One message for a wave of outages. Services that fall in the same minute, such as those of a restarting server, are told in one message per destination, and their return in one more. Thanks to @mafsi for the report (#13).
• A delay per service. "Declare an outage after [3] min without an answer", from 1 to 60 minutes, so a critical server alerts after a minute and a printer after fifteen (#13).
• A report after the incident. Once an incident is resolved, an editor writes what happened, why, and what changes so it does not happen again. It heads the incident's page, the history marks the incident, the feed carries it, and the destinations that receive incidents get it once.

Fixes and security

• "Back to …" in an event's details now shows only to those who may use it.
• The library that cleans the text of updates moves to its latest version, which fixes two flaws reported on 13 October in options Statup never turns on.
• The library that encrypts outgoing connections (notifications, checks) moves to its latest patch, which fixes four flaws reported on 14 October.

For webhook users

Grouped outages add a services list, and service keeps naming a service alone. A new happening, reviewed, carries the report of a resolved incident. See the body sent to any other tool.

Upgrade

Back up your data first (see the self-hosting guide), then, next to your docker-compose.yml:

docker compose pull && docker compose up -d

The database is upgraded on start. Every change is listed in the changelog.