As we've seen, allowing Linux-only dependencies on GCC results in its
dependency tree growing out of control to the point of being extremely
painful to maintain.
Let's stop this situation from getting worse by:
- failing a `--strict` audit when there is a Linux-only GCC dependency.
This also prevents new formulae from having such a dependency.
- failing any audit whenever a formula that did not previously have a
Linux-only GCC dependency picks one up
If a future formula update causes a formula to fail to build on Linux
because it needs a newer GCC when it previously did not, then we should
not bottle it. We can bottle this hypothetical formula when our bottling
distribution includes a new enough version of GCC.
* Fix audit not being run at all due to always-false condition.
* Don't fail audit if versioned alias would clash with a versioned formula.
* Allow versioned aliases with the name of another alias.
Occasionally, we get PRs that attempt to add a dependency on a formula
in an external tap in homebrew/core (e.g. Homebrew/homebrew-core#94791).
We don't actually have an audit that disallows that. We should have one.
Logically new formulas should start at revision 0, but a
developer might use an existing formula as a template and
otherwise not realize they should remove this field.
Signed-off-by: Josh Soref <jsoref@users.noreply.github.com>
A resource can do `using: :homebrew_curl`, but, without this change, the
audit will still use system `curl`. This fails quite predictably.
Needed for Homebrew/homebrew-core#88330.
If the download url is using brewed curl, use that too for the homepage check.
Ideally one would introduce the "using: :homebrew_curl" stanza for the homepage
line in the DSL, but it is easier to implicitely use the same logic
for homepages and url's and define the "using:" only once.
- `download_strategy`: only request image index JSON for downloading
the manifest for the tab
- use a shared `OS` constant for the version of `glibc` we use in CI
- fix `skoepeo` typo
- ensure that blank hash values are deleted (again) rather than just
`nil` ones
- use a shared `Hardware::CPU` constant for oldest CPU we're
supporting/using on Intel 64-bit
- re-add comment to `software_spec`
