[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Transition, libghc-xmonad-dev



  Hi Clint,


Thanks for the suggestions!

Clint writes:

> * Review https://release.debian.org/transitions/html/haskell.html for
>   packages that are broken on a subset of architectures, and investigate
>   the build failures on those architectures.  You can also get a different
>   view for a particular architecture by going to (example for amd64)
>   https://buildd.debian.org/status/architecture.php?a=amd64&suite=sid

I started at the top: The first package failing to build is
haskell-bytestring-show. Looking at the logs, it is failing because the
Setup.lhs file contains a reference to runTests:

  main = defaultMainWithHooks (simpleUserHooks { runTests = tests })

which was removed in Cabal 3.0.0.0, having had the notice "DEPRECATED
runTests "Please use the new testing interface instead!"" since Cabal
1.18.0.

I can make the package build by simply removing the "{ runTests = tests
}", but that might be too much of a shortcut - or maybe not, as I can
see no tests in the package.

So I guess this patch would fix haskell-bytestring-show:

Description: remove deprecated runTests
 Deprecated since Cabal 1.18.0, removed in Cabal 3.0.0.0.
 The package contains no tests.

---

--- haskell-bytestring-show-0.3.5.6.orig/Setup.lhs
+++ haskell-bytestring-show-0.3.5.6/Setup.lhs
@@ -2,6 +2,6 @@
 > import Distribution.Simple
 > import System.Cmd (system)
 
-> main = defaultMainWithHooks (simpleUserHooks { runTests = tests })
+> main = defaultMainWithHooks simpleUserHooks
 
 > tests _ _ _ _ = system "runhaskell Tests/Properties.hs" >> return ()


I know, it's not a lot, just a beginning.

I should probably have started by looking at pandoc. It won't build
because it depends on an older version of libghc-cmark-gfm-dev - so what
I should be doing to solve that is to find a way to help with getting
pandoc updated at least 2.8, which depends on the version of
libghc-cmark-gfm-dev in unstable.

> * Look into RC bugs at
>   https://bugs.debian.org/pkg-haskell-maintainers@lists.alioth.debian.org

Ah, ok, I should have looked here first - as I find:

  #962890 [G| | ] [src:haskell-bytestring-show] haskell-bytestring-show:
   Removal notice: broken and unmaintained

which I guess I makes my little "investigation" and patch above moot.

> * Look at the output of the package-plan build at, for example,
>   https://jenkins.debian.net/job/haskell-package-plan/ and help package
>   anything with an "Added:" line that isn't already in the NEW queue

Ok, so looking at #3130's console output, I get this list of Added:

  Added:    atomic-write is 0.2.0.7 here but does not exist in the archive.
  Added:    brittany is 0.12.1.1 here but does not exist in the archive.
  Added:    butcher is 1.3.3.1 here but does not exist in the archive.
  Added:    casa-client is 0.0.1 here but does not exist in the archive.
  Added:    casa-types is 0.0.1 here but does not exist in the archive.
  Added:    contravariant-extras is 0.3.5.1 here but does not exist in the archive.
  Added:    czipwith is 1.0.1.2 here but does not exist in the archive.
  Added:    data-tree-print is 0.1.0.2 here but does not exist in the archive.
  Added:    deque is 0.4.3 here but does not exist in the archive.
  Added:    doclayout is 0.2.0.1 here but does not exist in the archive.
  Added:    emojis is 0.1 here but does not exist in the archive.
  Added:    ghc-exactprint is 0.6.2 here but does not exist in the archive.
  Added:    ghc-lib-parser is 8.8.3.20200412.1 here but does not exist in the archive.
  Added:    hi-file-parser is 0.1.0.0 here but does not exist in the archive.
  Added:    hslua-module-system is 0.2.1 here but does not exist in the archive.
  Added:    http-download is 0.2.0.0 here but does not exist in the archive.
  Added:    jira-wiki-markup is 1.0.0 here but does not exist in the archive.
  Added:    js-dgtable is 0.5.2 here but does not exist in the archive.
  Added:    keysafe is 0.20200214 here but does not exist in the archive.
  Added:    microlens-aeson is 2.3.0.4 here but does not exist in the archive.
  Added:    monad-memo is 0.5.1 here but does not exist in the archive.
  Added:    multistate is 0.8.0.2 here but does not exist in the archive.
  Added:    one-liner is 1.0 here but does not exist in the archive.
  Added:    ordered-containers is 0.2.2 here but does not exist in the archive.
  Added:    ormolu is 0.0.3.1 here but does not exist in the archive.
  Added:    pantry is 0.4.0.1 here but does not exist in the archive.
  Added:    rebase is 1.4.1 here but does not exist in the archive.
  Added:    relational-record is 0.2.2.0 here but does not exist in the archive.
  Added:    rerebase is 1.4.1 here but does not exist in the archive.
  Added:    rio-orphans is 0.1.1.0 here but does not exist in the archive.
  Added:    rio-prettyprint is 0.1.0.0 here but does not exist in the archive.
  Added:    selective is 0.3 here but does not exist in the archive.
  Added:    semialign is 1.1 here but does not exist in the archive.
  Added:    strict-list is 0.1.5 here but does not exist in the archive.
  Added:    tar-conduit is 0.3.2 here but does not exist in the archive.
  Added:    tasty-lua is 0.2.2 here but does not exist in the archive.
  Added:    template-haskell-compat-v0208 is 0.1.2.1 here but does not exist in the archive.
  Added:    tz is 0.1.3.4 here but does not exist in the archive.
  Added:    tzdata is 0.1.20190911.0 here but does not exist in the archive.
  Added:    unsafe is 0.0 here but does not exist in the archive.
  Added:    wai-cors is 0.2.7 here but does not exist in the archive.

I'm not sure where to start, or how I help most efficiently with getting
pandoc updated, so xmonad can be.

Maybe I should try to get rid of the dependency on Text.Pandoc in
xmonad, it is only used to generate the man-page from a markdown file
with some keybindings inserted.

Thanks for helping me take a tiny step forward in understanding how this
all fits together!


  Best regards,

    Adam

-- 
 "Scientific developments can always be made logical        Adam Sjøgren
  and rational with sufficient hindsight. It is        asjo@koldfront.dk
  amazing, however, how clouded the crystal ball
  looks beforehand."


Reply to: