Some README tweaks as proposed by gorakhargosh

README.rst

@@ -1,25 +1,19 @@
-bjoern: Fast And Ultra Lightweight Asynchronous WSGI Server
+bjoern: Fast And Ultra-Lightweight Asynchronous WSGI Server
 ===========================================================
 
-A screamingly fast, ultra-lightweight asynchronous `WSGI`_ server for CPython,
-written in C using Marc Lehmann's high performance `libev`_ event loop and
-Ryan Dahl's `http_parser`_.
+A screamingly fast, ultra-lightweight asynchronous WSGI_ server for CPython,
+written in C using Marc Lehmann's high performance libev_ event loop and
+Ryan Dahl's http_parser_.
 
 Why It's Cool
 ~~~~~~~~~~~~~
 bjoern is the *fastest*, *smallest* and *most lightweight* WSGI server out there,
 featuring
 
-* < 1000 Lines of Code
+* < 1000 lines of code
 * Memory footprint ~ 600KB
 * Single-threaded and without coroutines or other crap
-* 100% WSGI compliancy (except for the `write callback design mistake`_)
-
-.. _WSGI:         http://www.python.org/dev/peps/pep-0333/
-.. _libev:        http://software.schmorp.de/pkg/libev.html
-.. _http_parser:  http://github.com/ry/http-parser
-.. _write callback design mistake:
-                  http://www.python.org/dev/peps/pep-0333/#the-write-callable
+* 100% WSGI compliance (except for the `write callback design mistake`_)
 
 What's Not So Cool
 ------------------
@@ -32,7 +26,9 @@ libev
 Arch Linux
    ``pacman -S libev``
 Ubuntu
-   ``apt-get install libev``
+   ``apt-get install libev-ev``
+Mac OS X (using homebrew_)
+   ``brew install libev``
 Your Contribution Here
    Fork me and send a pull request
 
@@ -48,33 +44,46 @@ Usage
    
    bjoern.run(wsgi_application, host, port)
 
-   # alternatively, the mainloop can be run separately:
+Alternatively, the mainloop can be run separately::
+
    bjoern.listen(wsgi_application, host, port)
    bjoern.run()
 
 But What About...
 ~~~~~~~~~~~~~~~~~
-Fapws3?
+Fapws3_?
 -------
 Sucks. Really, the code is an incredible mess. It likes to segfault.
 
 I tried to patch Fapws so that it would support threading,
 but after about two hours of brain slime feeling, I decided
 to write my own WSGI server -- "Fapws done right".
 
-gunicorn?
+gunicorn_?
 ---------
 bjoern is about 5 times faster. Plus gunicorn is too much bloated.
 
-uWSGI?
+uWSGI_?
 ------
 Awesome project, but way too much bloat.
 
-meinheld?
+meinheld_?
 ---------
 Unfortunately now bloated with gunicorn and coroutine/greenlet crap,
 seemed to be a very nice server at first.
 
 $SERVER?
 --------
 Probably too much bloat, too slow, does not scale, buggy, ...
+
+
+.. _WSGI:         http://www.python.org/dev/peps/pep-0333/
+.. _libev:        http://software.schmorp.de/pkg/libev.html
+.. _http_parser:  http://github.com/ry/http-parser
+.. _write callback design mistake:
+                  http://www.python.org/dev/peps/pep-0333/#the-write-callable
+.. _homebrew: https://mxcl.github.com/homebrew/
+.. _meinheld: https://github.com/mopemope/meinheld/
+.. _uWSGI: http://projects.unbit.it/uwsgi/
+.. _gunicorn: http://gunicorn.org
+.. _Fapws3: http://fapws.org