Remove Ruby 2.3ism

This release is only relevant to those who use suitably ancient Ruby versions that lack the safe navigation operator (&.), now your git versions will bump with style once more!

And lo! a new release appears

This release exists thanks to the contributions of bug reports and suggestions from many people.
My thanks to all of them for their time to make git-version-bump better.

• If you have a need to force git-version-bump to report a particular version number, rather than derive it from the tags in the git repo, you can now do that by setting the versionBump.versionOverride and/or versionBump.dateOverride git config variables.
  The README has example commands for doing so.
  It should be vanishingly rare for anyone to need to do this, but the sorts of situations in which it might be useful are explained in the issue.

  Thanks to Roy Warps (@RoyWarps) for the feature request (#23), and the discussion that allowed us to come to an agreement on the right approach.

• The website URL in the gemspec now points somewhere useful (#14; thanks to Chris Miller, @iamchrismiller, for the report)

• Certain corner cases of version resolution going wrong when a GVB-using gem is being used in a different git repo are now handled better.
  Thanks to Dmitry Fleytman (@dmitryfleytman) for his original bug report (#13), and to Roy Warps (@RoyWarps) for his slightly differently structured bug report (#22), which between them helped me to identify and fix the underlying issue.

• The command that opens the text editor that allows you to write release notes no longer captures the terminal's stdout/stderr, so that terminal-based text editors don't get hung up. (#24)
  Thanks to Geff (@p64) for the report, PR, and patience in getting it into optimum shape.

Detect and be friendly about all command failures

Inspired by #21, I've reworked the command execution logic to catch
command failures and provide suitably friendly error messages. For example:

Failed while pushing commits to the default remote repository (exit status: 128)
command output was:
----8<----
ssh: Could not resolve hostname jithub.com: Name or service not known
fatal: Could not read from remote repository.

Please make sure you have the correct access rights
and the repository exists.
---->8----

I did have to do surgery to some rather low-level bits of the system, so it's possible that I h0rked Windows support or something.
If so, my apologies, as I don't have a Windows system to test on.
Bug reports are welcomed and will be speedily dealt with.

Fix problems with signed commits when using GVB.commit_date_version

One very simple fix, which shows how rarely the commit_date_version method gets
used (which is as it should be).

Lite tags and Windows Wonderland

Thanks to Robert Clark (@rbclark), this release contains some Windows compatibility fixes,
as well as support for recognising "lite" (un-annotated) tags as version tags.

and I think it's gonna be a long, long time...

Various bugfixes that have accumulated over the years, along with some minor
features added by some helpful people.

Use the correct date

I always intended for the date to be the commit date, not the author
date; I just never noticed the mistake until now.

More 1.8 compatibility

v0.15.0

More 1.8 compatibility

GVB.commit_date_version

This allows you to use an alternate, non-tag-based, approach for your
package versioning, if semver doesn't make sense for your application. I
don't recommend it for general use.

Boneheaded bug fix release

Begone, evil Errno::ENOENT: No such file or directory @ realpath_rec!