Blockers for using upstream debos

[evelikov]

Here's a list of blockers for us to use upstream debos instead of our fork:

• proper exitcode handling - any of Reverse the exitcode initialization value #198 Reverse the exitcode initialization value #278 or Refactor error handling #303 - please add tests
• add option to install extra packages in the pacstrap stage - currently we implicitly install only "base" - pacstrap: add option to install extra packages #481
• relative partition sizes - image_partition: add ability to skip "start" and specify relative offsets #285

We've had a fork for 4 years, hope we get the above sorted before we hit 5 🤞

[obbardc]

add option to install extra packages in the pacstrap stage - currently we implicitly install only "base" is the only thing which hasn't got a pull request outstanding, perhaps you could open one ? I assume it should be very quick to get in.

[evelikov]

I am rather hesitant to open new MRs since getting stuff is repeatedly 3+ months of regular rebasing and pokes.

As debos/fakemachine becomes a more actively maintained project and starts churning through it's MR I would be more than happy to submit some. Looking through the list there is a MR from 2018 without a single comment from the maintainers ☹️

Sorry if I sound like a downer - I simply cannot afford to spend an extra 4x time rebasing and poking.

[obbardc]

You want to add new things but don't want to create pull requests to add the new things? Seems like you're only blocking yourself here.

[evelikov]

Not sure what you mean, sorry.

I am happy to open a PR for the final piece, once any of the other blockers (or any PRs really) get some traction.

[obbardc]

Still not sure why you can't open a PR for what appears to be some trivial change to an action you wrote ? :-)

Simply if some PR is ready to be included, it will be merged.

[obbardc]

In the meantime, I have created #481 for the missing part.

[obbardc]

This isn't a constructive issue - it's just a list of other issues. Closing.