This release improves the reliability of the Rhino build environment for large projects, and upgrades all the bundled requirejs libraries.
Mohawk Surveys now features a programmable reporting system, powered by a functional scripting language developed specifically for the platform.
The scripting language was designed for sophisticated analysis of response data, with strict guarantees to ensure that scripts from untrusted sources could be run safely in a sandboxed runtime.
django-watson 1.1.3 is now available on GitHub and PyPi. It features better support for punctuation in queries, and the ability to rebuild search indices for specific models.
Model historylinks are no longer updated on fixture loading, fixing edge-case issues with multi-table inheritance.
django-reversion was my first open-source contribution to Django, and has become the de-facto solution for model data version control. It's good to give something back to the ecosystem that has taught me so much.
Even better, much of it's recent development has been carried out by the community. Congrats to everyone who has helped carry it through six major releases of Django. It's been an awesome ride!
The latest release of django-herokuapp provides a Django 1.6-compatible project template, as well as a new heroku_audit management command for diagnosing common heroku deployment bugs.
With Christmas over, the Tesco Festive Feast Planner has now been shut down, having spent two high-traffic months on the Tesco Christmas portal.
It is expected to appear again over the coming year, repurposed for other national holidays. Look out for it again next Christmas when you're planning your feast!
The latest release of django-require is now available on GitHub and PyPi. This release upgrades the bundles RequireJS to 2.1.8, as well as fixing the tests to work in Django 1.6.
This release adds compatibility for Django 1.6, as well as removing the type flag on Version models.
django-watson 1.1.2 has just landed! Key features include Python 3.3 compatibility and a more memory-efficient buildwatson command.
A new bulk mailing system has been added, allowing email newsletters to be created, scheduled, and dispatched to high-volume mailing lists. The system is designed to handle millions of emails per month, at a rate of 100K emails per hour.
The system makes heavy use of RabbitMQ and Celery to parallelize sends amongst a large number of background workers, and is horizontally scalable should future volumes increase.
Other key technologies contributing to the send performance include PostgreSQL full text indexing and Amazon Elasticache.