Decouple variable declarations so that each can have docstring.
In case of _relck() we also rename variable $relsrc to a more
context-specific term $oracle.

Where not required, they only add visual cruft.

Probably leftover from slightly different implementation.  There's
nothing to add to the error message now.

In this file we declare lot of globals and source lot of code, and
basically do only that, which makes this check less useful.

Overview of changes:

 *  Added `vbump` and `release` targets, defaulting to Z

Overview of changes:

 *  Changelog will be included in tag annotation message.

    The changelog generated by `vbump_*` target will be included in tag
    annotation message.  The annotation message is meant as release note
    of sorts, so it makes sense to have the changelog there.  (For
    example, GitHub will use the message as release note on Releases tab
    of a project.)

Annotation message is meant as official release note, so including it
there makes sense.

(For example, GitHub will use it to create release note under Releases tab.)

Overview of changes:

 *  Fixed bug when 'src' directory must have existed to build anything
 *  Improved README

Overview of changes:

 *  Added codename back to tag annotation message
 *  Added README.md

Overview of changes:

 *  Enabled overriding of bump size (MKIT_BUMPSIZE)

If for some reason, user wants to skip a version or two, this will
override version bumping in vbump_* and some checks new and tag
creation in release_*.

Overview of changes:

 *  Added (mandatory) mkit.ini version checks
 *  Revamped parts of mkit.ini structure

 *  Dry mode now implies debugging mode
 *  Fixed latest_tag git fact
 *  Made reldst optional
 *  Removed ini() debugging
 *  Removed unnecessary exports
 *  Code clean up and maintenance

although I haven't been able to put my finger on it, git-log must have
changed between Debian's 2.1.4 and Fedora's 2.5.something.  The previous
version worked on Fedora but not on Debian; hopefully this solution is
more robust.

This is the last version where we tolerate incompatible changes in
mkit.ini and from now on, all mkit.ini files must have the version line
(in first 3 or last 3 lines).

As per my interpretation of SemVer, if major version is 0, we may
introduce incompatible changes within minor, in that case we check that
as well.  Otherwise, we just check the major.

They are all about "distribution", rpmstuff and debstuff have basically
just one item, and having "lists" section was ugly and meaningless in
the first place.

Merge all files:group sections into one, where key is name of the group
and value is source name, optionally followed by destination path.

This breaks old mkit.ini's, but significantly simplifies future ones.

 1. Only one section is needed,

 2. group list is implied from the files section, removing the need
    of lists:group,

 3. destination path has been made optional (by default, basename of
    the source path is used).

Too verbose; makes it hard to read the rest.

It's hardly usable when silent

No need to bug everybody with relsrc/reldst

Overview of changes:

 *  Replaced 'vars' key with 'tokens'
 *  Improved version bumping (won't destroy spaces around '=')
 *  Improved robustness of token replacing
 *  Enabled more tokens on single line
 *  Added tokens: __MKIT_PROJ_VCS_BROWSER__ __MKIT_PROJ_MAINTAINER__ and
    __MKIT_PROJ_GIT_LASTHASH__
 *  Code maintenance and cleanup

It's easier to do quoting at least mediocre-right in sed than in perl;
eg. with Perl, even e-mail address was causing quoting breakage due to
the at sign.

Indeed a weird one...

Note that similar to __MKIT_PROJ_VERSION__, this value cannot be
reliably determined when outside git (e.g. distributor's build dir),
therefore must be kept inside dist source package.

Unlike version, though, user (ie. app developer, ie. mkit.ini
maintainer) most probably does not want to edit it, therefore we're
normally keeping it out of sight.

The value itself is normally just the last git commit's hash (SHA1 these
days), suffixed with ".dirty" mark if there were uncommitted changes
during build.  (Just the same as we do with semver())

If we're outside git and the hash could not be determined, that either
means the source package was mangled with or was built with older
version of mkit.  In either case, the value provided will be 'UNKNOWN'.

Some parts got a bit confusing over time, also there was repetition.

In this commit:

 *  the release fact gathering routines have been moved to separate
    module,

 *  some less useful routines have been removed,

 *  some internal interfaces have been changed to allow for more
    readable code.