I got bitten by a Not Invented Here and decide to implement my own. I have no idea in fact how the previous two projects work exactly, since I only browsed the documentation and source for a few minutes (hopefully I'm not breaking any patents or licenses with my project :)
The basic idea is to create an Experiment, create the tests on templates and enjoy the reports.
I still haven't put it live on any of my sites, so it works on my development server. But I decided to release it anyway to see if others can install the app. And if you do, if it works as you want.
Some features:
- Experiment reports, accessible on Django's admin interface
- "Continuous experiment" (for the lack of a better name). It's an experiment which keeps recording goal hits past the first one. I have a forum on one of my sites, and was in doubt if adding a timestamp next to a list of topics would increase the pageviews on ^forum/(.*)
And hey, admittedly I don't have a lot of experience with split testing in general. Most of the stuff on Django MVT was done based on what I imagined a split test report to be and coding only enough to get there.
I appreciate any feedback, be on the app itself or testing in general :) Are there features you look for on this type of project?
There's django-ab http://github.com/johnboxall/django-ab and django-lean http://bitbucket.org/akoha/django-lean if you're interesting in A/B tests.
I got bitten by a Not Invented Here and decide to implement my own. I have no idea in fact how the previous two projects work exactly, since I only browsed the documentation and source for a few minutes (hopefully I'm not breaking any patents or licenses with my project :)
The basic idea is to create an Experiment, create the tests on templates and enjoy the reports.
I still haven't put it live on any of my sites, so it works on my development server. But I decided to release it anyway to see if others can install the app. And if you do, if it works as you want.
Some features:
- Experiment reports, accessible on Django's admin interface
- "Continuous experiment" (for the lack of a better name). It's an experiment which keeps recording goal hits past the first one. I have a forum on one of my sites, and was in doubt if adding a timestamp next to a list of topics would increase the pageviews on ^forum/(.*)
And hey, admittedly I don't have a lot of experience with split testing in general. Most of the stuff on Django MVT was done based on what I imagined a split test report to be and coding only enough to get there.
I appreciate any feedback, be on the app itself or testing in general :) Are there features you look for on this type of project?