Feature: --keep-going for task to continue on error

[sofusalbertsen]

I'm missing the --keep-going feature of make in Task.
From the documentation of make:

--keep-going

Continue as much as possible after an error. While the target that failed, and those that depend on it, cannot be remade, the other prerequisites of these targets can be processed all the same. See Testing the Compilation of a Program.

The scenario I have at the moment is this:

• I use task to create infrastructure as code (eksctl based).
• I have a CI system that spins up an infrastructure on PR's via my task deploy task.
• If something goes wrong, I would like task to run my task destroy task, disregarding that all the sub tasks might fail (in the case that the task deploy did fail at some point).
  At the moment, the only way I can do this is through https://taskfile.dev/usage/#ignore-errors , but I do not want to set this as a permanent configuration setting, only at runtime.

[grumpper]

Did someone do any work on this?

@sofusalbertsen did you find some workaround for this?

I am currently eating a similar pickle where I must run a task with two other tasks as dependencies and even if one of them fail all execution stops...

[andreynering]

For some use cases defer can be used as a workaround to ensure something is ran.

[trulede]

@sofusalbertsen you can use defer along with EXIT_CODE to trigger your destroy task(s) in case the deploy task fails.

https://taskfile.dev/docs/guide#doing-task-cleanup-with-defer

Otherwise you can control ignore_error via a template variable (which you can set at the CLI, for example). Something like this:

ignore_error: {{.KEEP_GOING}}