.github
datadog
doc
examples
foreman
src
t
Config
Controller
Helper
input-types
jobs
basic.nix
build-output-as-input.nix
build-product-simple.sh
build-product-with-spaces.sh
build-products.nix
bzr-checkout-input.nix
bzr-checkout-update.sh
bzr-input.nix
bzr-update.sh
config.nix.in
darcs-input.nix
darcs-update.sh
deepgit-builder.sh
deepgit-input.nix
empty-dir-builder.sh
fail.sh
git-input.nix
git-rev-input.nix
git-rev-update.sh
git-update.sh
hg-input.nix
hg-update.sh
nested-attributes.nix
runcommand.nix
scm-builder.sh
succeed-with-failed.sh
svn-checkout-input.nix
svn-checkout-update.sh
svn-input.nix
svn-update.sh
lib
plugins
scripts
Makefile.am
api-test.nix
api-test.pl
build-products.t
evaluate-basic.t
evaluate-dependent-jobsets.t
s3-backup-test.config
s3-backup-test.pl
setup-notifications-jobset.pl
test.pl
.gitignore
.yath.rc
COPYING
INSTALL
Makefile.am
Procfile
README.md
bootstrap
configure.ac
default.nix
flake.lock
flake.nix
hydra-api.yaml
hydra-module.nix
shell.nix
version
This is a breaking change. Previously, packages named `packageset.foo` would be exposed in the fake derivation channel as `packageset-foo`. Presumably this was done to avoid needing to track attribute sets, and to avoid the complexity. I think this now correctly handles the complexity and properly mirrors the input expressions layout.