Details
- Reviewers
- None
- Group Reviewers
portmgr - Commits
- R11:07679900322e: Mk: add a BUILD_RUN_DEPENDS variable
Diff Detail
- Repository
- R11 FreeBSD ports repository
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
I'm guessing you want to implement this ASAP to get it in use.
I will amend FreshPorts to extract this variable. If I got that work done before this change goes into use, it would capture all future uses. If I don't, it's just a one-off script to refresh the database from the ports tree.
I don't think freshports needs to extract this, you can keep using build/run, it is only to be used as some kind of syntaxic sugar, and it's mostly to avoid people doing RUN_DEPENDS=BUILD_DEPENDS.
FYI, I have code changes in place everywhere but prod. I'm letting them run before updating prod tomorrow. If this could hold off until after that, I would appreciate it as it would avoid work for me. Thanks.
I don't think freshports needs to extract this, you can keep using build/run, it is only to be used as some kind of syntaxic sugar, and it's mostly to avoid people doing RUN_DEPENDS=BUILD_DEPENDS.
oops. I just saw this.
Are you saying that make -V RUN_DEPENDS will still pull back the usual stuff? Same with make -V BUILD_DEPENDS ... hmm.
It's exactly what the patch does, put BUILD_RUN_DEPENDS into BUILD_DEPENDS and RUN_DEPENDS.
What happens if a Makefile contains:
BUILD_RUN_DEPENDS=... BUILD_DEPENDS=... RUN_DPENDS=...
What also comes to mind is:
- portfmt
- portlint
- Order of Variables in Port Makefiles: https://docs.freebsd.org/en/books/porters-handbook/order/