Commits on Apr 25, 2017
  1. Remove obsolete

    skids committed Apr 25, 2017
Commits on Dec 11, 2016
  1. Merge pull request #4 from zoffixznet/patch-1

    skids committed Dec 11, 2016
    Add mandatory "perl" META field
Commits on Oct 26, 2016
  1. Add mandatory "perl" META field

    zoffixznet committed Oct 26, 2016
    The `perl` field specifies the minimal perl version for which this distribution can be installed and is a mandatory field. The value of `"6.*"` indicates any version suffices.
    It is recommended to use [Test::META]( module as an author test, to catch any issues with the META file.
Commits on Feb 4, 2016
Commits on Dec 31, 2015
Commits on Dec 25, 2015
Commits on Dec 24, 2015
Commits on Nov 6, 2015
Commits on Sep 11, 2015
Commits on Sep 5, 2015
  1. GLRify everything, and bump the release version.

    skids committed Sep 5, 2015
    Eventually will have to give all the submodules versioning.
Commits on Jul 14, 2015
  1. After a long and constructive conversation with ugexe++ in IRC

    skids committed Jul 14, 2015
    this is IMO the least-surprise-for-the-most-people balance
    point as to what to do about test files finding their libs.
    1) Naive users can just perl6 t/test.t
    2) For that use case blib/lib is first, if you have one, like
    from inside a .panda_work directory.
    3) Automated utilities should not care -- they will either
    build from within the package directory or know better
    than to build with preexisting system ./lib or ./blib in their CWD.
    4) The case where people are expecting precomps to be
    someplace other than blib/lib should be rare
    5) The case where people are simultaneously reaching out from
    inside a directory that has a different lib/Sum or blib/Sum
    and typing a full path to our test files should also be rare.
    6) Data files can reasonably be assumed to stay alongside us in
    t/ where they started, so use $?FILE to find them.
    Now to see if panda likes it.
Commits on Jul 13, 2015
  1. Oh, OK I get it now. Someone thought it was a good idea to flag

    skids committed Jul 13, 2015
    any Failure object ever created that does not get handled, even
    if it never gets thrown?  I think that is rather pedantic and
    limits the usefulness of Failures.
    Anyway, hopefully this time I will fix the tests.
  2. Keep some injected failures from reaching DESTROY. I think.

    skids committed Jul 13, 2015
    The errors only seem to show up when panda does the tests
    so my visibility is limited here.
Commits on Jun 23, 2015
  1. Merge pull request #2 from ugexe/more-provides

    skids committed Jun 23, 2015
    add Sum::HAS and Sum::Digest::Marshal to provides
Commits on Jun 19, 2015
  1. Temporarily put in some workarounds for spread of RT#123623

    skids committed Jun 19, 2015
       looks like it my be fixed soon, but kept for older installs for a while
Commits on Jun 18, 2015
Commits on Jun 13, 2015
  1. Trying for a quick-fix for panda install woes

    skids committed Jun 13, 2015
       hoping order preserving sort will do it.
Commits on May 25, 2015
  1. hyphenate ALL TEH!!!

    skids committed May 25, 2015
Commits on May 16, 2015
  1. Hyphenate throws_like

    skids committed May 16, 2015
  2. Fix a test for Failures that left them unhandled.

    skids committed May 16, 2015
    Type the exception while I'm at it.
    Throw the exception rather than just produce a failure, because Proxy weirdness
Commits on May 9, 2015
  1. Adapt to some more GLR flattening/Nil/() changes

    skids committed May 9, 2015
    Improve an exception message
Commits on Apr 29, 2015
Commits on Apr 28, 2015
  1. isa_ok --> isa-ok

    skids committed Apr 28, 2015
Commits on Apr 25, 2015
Commits on Apr 20, 2015
Commits on Apr 19, 2015