add "end of" section comments to avoid confusion for skipped

Several times I and others have mistakenly added packages to
expected-haddock-failures instead of expected-test-failures
for example - hopefully these section end comments will be
agreeable and help avoid such problems.

I added for all the long sections but not yet for the shortest
ones (ie skipped-builds and skipped-profiling) though maybe
they should have them too for consistency.
This commit is contained in:
Jens Petersen 2015-10-08 10:50:56 +09:00
parent 0909d667dd
commit 8e0e01ec8f

View File

@ -1552,6 +1552,9 @@ packages:
# https://github.com/fpco/stackage/issues/880
- http-types < 0.9
# end of packages
# Package flags are applied to individual packages, and override the values of
# global-flags
package-flags:
@ -1608,12 +1611,16 @@ package-flags:
nix-paths:
allow-relative-paths: true
# end of package-flags
# Used for packages that cannot be built on Linux
skipped-builds:
- hfsevents
- Win32
- Win32-notify
# By skipping a test suite, we do not pull in the build dependencies
skipped-tests:
- dbmigrations
@ -1699,6 +1706,9 @@ skipped-tests:
- uuid
- uuid-types
# end of skipped-tests
# Tests which we should build and run, but which are expected to fail. We
# should not fail a build based on a test failure for one of these packages.
expected-test-failures:
@ -1996,6 +2006,9 @@ expected-test-failures:
# https://github.com/alcidesv/second-transfer/issues/3
- second-transfer
# end of expected-test-failures
# Haddocks which are expected to fail. Same concept as expected test failures.
expected-haddock-failures:
# https://github.com/acw/bytestring-progress/issues/4
@ -2021,6 +2034,9 @@ expected-haddock-failures:
# fedora/haskell/cabal-rpm/f23/
- persistent-mysql
# end of expected-haddock-failures
# Benchmarks which should not be built. Note that Stackage does *not* generally
# build benchmarks. The difference here will be whether dependencies for these
# benchmarks are included or not.
@ -2064,10 +2080,14 @@ skipped-benchmarks:
# https://github.com/liyang/true-name/issues/2
- fast-builder
# end of skipped-benchmarks
skipped-profiling:
# https://github.com/nomeata/ghc-heap-view/commit/8d198eb8fbbad2ce0c4527c781659f35b8909c04#diff-8288955e209cfbead5b318a8598be9c0R10
- ghc-heap-view
# Mapping from Github account holding a package to the Github users who should
# be pinged on failure. If no value is specified here, then the owning account
# will be pinged.
@ -2104,3 +2124,5 @@ github-users:
lambdabot:
- DanBurton
- mokus0
# end of github-users