With BuildWorld() we get more results back (now we return the whole
model, including the false assertions).
Besides, now solving with BuildWorld() detects an invalid case:
when we supply a provided, the definitionDB shouldn't explictly supply
also the package that has to be provided. This would cause to 'shadow'
packages between repositories.
The test was invalid before, and shouldn't have contained A1. Moved the
test to Pending to inspect it further in subsequent dev iterations
This is currently not a valid use case. Virtuals are empty packages and
if the `build.yaml` is completely empty, nothing could depend on them.
Let's try to not be too smart and build the package image if a source
image is supplied, and fail hardly when we depend on a virtual in build
time.
We don't generate anymore images if packages are empty - those are now
virtuals which just generates empty artifacts.
Virtuals are not meant to be required by other packages in build time,
because it would violate the virtual packages purpose (they are just
useful for runtime).
This test was used to verify version selection of the best match during build
time, not to actually test any build process. Inject steps so images are
actually generated, and they can depend on each others.
When we are swapping packages, we do not run the solver to gather things
to install, but we trust the given list when calling computeInstall. In this case, the assertion
returned by computeInstall is empty, as we force l.Options.NoDeps.
This change generates the assertion list while calling computeSwap so
it's available later when we call ExecuteFinalizer.
When we cycle, we don't necessarly have all the packages into the DB
yet.
With this change, luet annotates the reverse dependency without any version, and we try to
update revdeps table when new items gets added, by checking the version
required in the selector.
Thanks to @joostruis for noticing the issue
