Monday, September 09, 2013

Cloud, HPC And Open Technologies Converge To Fuel Research, Innovation

When you put leaders of industry, research, and academia in a room for a day, what do you get? If you were at Argonne National Laboratory last month for the workshop on resource intensive open clouds, you got a taste of progress. The workshop was organized by leaders from Notre Dame University, Internet2, and Rackspace, in the interests of figuring out the next steps for the technical computing world, bridging old-world high performance computing with the new-world of cloud computing.

I was invited to this workshop on behalf of the PSF, and was excited by the prospect of being involved in an open and collaborative environment, tasked with figuring out how all of these sides of the story could come together. Many of the attendees are using OpenStack, an open cloud computing platform implemented in Python, and Python was a key technology for many of them in other ways. It looks like OpenStack-based and community-owned open clouds will likely become key points as the group progresses towards a better landscape to solve their computing needs, and if the past is any indication, Python will remain an important piece of the software powering it.

"The pace of innovation is accelerated and the diversity of solutions and approaches ensures that good solutions persist and not so good ones are quickly identified," said event organizer Paul Rad of Rackspace, on the topic of open and transparent workshops like this one.

My hope for this group is that future workshops can leverage some of our leaders in Python's large scientific community, many of whom are undoubtedly facing the challenges this workshop set out to improve on. Feel free to contact me at brian@python.org if you're interested in contributing to future efforts.

For more details on the workshop, see Paul's post on the subject at http://www.rackspace.com/blog/cloud-hpc-and-open-technologies-converge-to-fuel-research-innovation/. If you are interested in learning more about this initiative and/or in participating in future workshop sessions, please email OpenCloud@internet2.edu.

Thursday, September 05, 2013

Python Surpasses Standards, Reaches New Levels of Quality

Throughout Python's 20-plus year history, its quality has been in the hands of the volunteers around the world openly contributing to it. Thanks to Coverity, those volunteers have been pointed to many quality and security issues via Coverity Scan, a service which finds defects in C/C++ and Java projects at no cost.


As the CPython project includes over 370,000 lines of C code*, accounting for 42% of the codebase, a lot of it lies outside of the analysis tools our community has created to work with Python code. Since 2006, Coverity's scans of that code have found nearly 1,000 defects, 860 of which our contributors have fixed.

In an industry where the standard defect density is a rate of 1 per 1,000 lines of code, CPython has attained a rate of 0.005 defects per 1,000 lines, and "introduces a new level of quality for open source software," said Coverity.

“Python is the model citizen of good code quality practices, and we applaud their contributors and maintainers for their commitment to quality,” said Jennifer Johnson, chief marketing officer for Coverity.

The PSF and the rest of the community join Coverity in applauding all of those who have contributed their time and effort to make CPython a better project, along with the countless others who contribute to a powerful landscape of Python interpreters.


For more information, read Coverity's "Coverity Finds Python Sets New Level of Quality for Open Source Software" press release.


* generated using David A. Wheeler's 'SLOCCount'.