extradoc / pycon-advertisement / draft.3

Full commit
High-level languages are faster than low-level ones! - (Armin Rigo circa 1999)

For the past 8 years, the PyPy team has been working on this dream.  We
are delighted that we can now bring you a very fast, very flexible, and very
compliant Python, ready for production use.  We are very  proud of this 
result, and especially of our Just-In-Time compiler generator.

* most Python benchmarks run much faster than with CPython or Psyco
* tight loops run between 20 and 60 times faster than CPython
* our favourite large Python program, PyPy itself (200 KLocs) runs 2x as fast
* smaller total memory usage for large and long-running programs
* full (and JIT-ed) ctypes support for calling C libraries from Python
* a new "cpyext" layer which integrates existing CPython C extensions
* an experimental super-fast JIT-compilation of calls to C++ libraries
* support for X86 and amd64 -- we are very close to supporting ARM
* we are reliable, and with over 16,000 tests we can prove it.
* our generational GC is much faster than CPython's refcounting one.
* by the end of PyCon we will have full compatibility with  CPython 2.7

To see our speed results visit the PyPy Speed Center:

To find out more about PyPy:
   - visit
   - read our blog at 
   - or come by the #pypy channel on 

But we'd like your help. There are many ways you could support PyPy, an 
MIT-licensed Open Souce project, including:
   - contributing code  
   - coming to sprints
   - running your test suites under PyPy and reporting any errors you find
   - donating money to us through our non-profit foundation, via Flattr, 
     paypal, or Google Checkout. 

Making a donation is easy. It's just one click away on our blog site And because PyPy is a member of the
Software Freedom Conservancy, a 501(c)(3) charity, American donors
can often deduct the donation on their USA taxes.

Another way that you could help us is through finding us Paid Consulting 
Work.  We are the proven Python speed experts. And we know that companies and 
organizations exist out there for whom speed really is money.  They'd be 
happy to pay us to speed up their code, if only they knew that this service
was available.  Maybe you work for such a company, or maybe you know somebody
who does.  We'd like to get in touch.  Things we'd like to do for you include:

- speeding up your Python based project
- making your product run seemlessly on top of PyPy
- teaching you how to get the best out of PyPy
- modifying PyPy to better suit your needs
- on-site mentoring of developers, company sprints, teaching agile methodology
- commercial support contracts for PyPy
- porting your favourite C/C++ extension library to RPython, so that PyPy
  can jit it.
- integrating PyPy with Cython or NumPy

We could write a new backend so that PyPy can run on your favourite
VM or CPU archtecture, or even write an interpreter for your second favourite
dynamic programming language, so that it can be sped up as well!

Open End and merlinux along with the PyPy developers stand ready to help,
contact us at