This commit addresses two of the comments:
* renamed existing n-input git_buf_join to git_buf_join_n
* added new git_buf_join that always takes two inputs
* moved some parameter error checking to asserts
* extended unit tests to cover new version of git_buf_join

Add new functions to git_buf for:
* initializing a buffer from a string
* joining one or more strings onto a buffer with separators
* swapping two buffers in place
* extracting data from a git_buf (leaving it empty)

Also, make git_buf_free leave a git_buf back in its initted state,
and slightly tweak buffer allocation sizes and thresholds.

Finally, port unit tests to clay and extend with lots of new tests
for the various git_buf functions.

Don't fail when opening a new config file

reference_rename() cleanup

Fix typos in the README file

Never use the Qt trade mark with both letters capitalized, in the form
"QT".
(http://qt-project.org/trademarkpolicy.html)

Networking improvements

Update all stack allocations of git_filebuf to use GIT_FILEBUF_INIT
and make git_filebuf_open and git_filebuf_cleanup safe to be called
multiple times on the same buffer.

Signed-off-by: Vicent Marti <tanoku@gmail.com>

This uses the new Clay code. As you can see, the diff is minimal... It
works!

Fix typo in repository documentation

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

It can be useful to separate disconnecting from actually destroying
the object.

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

reference_rename used to delete an old reflog file when renaming a
reference to not confuse git.git. Don't do this anymore but let the user
take care of writing a reflog entry.

Signed-off-by: schu <schu-github@schulog.org>

Signed-off-by: schu <schu-github@schulog.org>

Actually rollback when we can't create the new reference. Mark the
rolled back reference as loose.

Signed-off-by: schu <schu-github@schulog.org>

0c49ec2d replaced git_futils_mv_atomic with p_rename without removing its
prototype.

Signed-off-by: schu <schu-github@schulog.org>

A simple branchname as refspec is valid and we shouldn't throw an
error when encountering one.

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

Add a next pointer to make it a linked list and add the 'pattern' and
'matching' flags.

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

Make it a bit more resilient.

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

There is no good reason to expose the negotiation as a different step
to downloading the packfile.

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

Thanks to Emeric.

It's redundant to do this (git doesn't) and Windows doesn't allow us
to overwrite a read-only file (which objects are).

Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>

Do not add the `git2` path to internal includes, or that will cause
an extra path dependency.

Comes with schu's stress tests for config files. Hopefully the diffs
will stay minimal from now on.

That makes more sense to me.