Today is our first day back to work after nearly a week long retreat for all Rails Dog employees. Everyone had a good time working together in person and a lot of great ideas came out of the meeting. We also had a chance to talk about our future plans for Spree. The following is a list of some of the goals that we hope to achieve in the next six months.
Improved Test Coverage
Nobody likes bugs in their code and the Spree developers are no exception. Even more frustrating than a bug is a bug that is introduced as a result of fixing another one. The only real way to solve this is through better test coverage. Until recently the test coverage for Spree has been fairly modest due to a lack of programming resources but we’ve recently taken steps to address that.
We recently made a big push for more comprehensive test coverage with new RSpec tests. We also hired two new full time programmers (John Brien Dilts and Neeraj Singh.) Neeraj is an active contributor to Rails itself and is already hard at work on identifying gaps in the test coverage which he will then address. We’ve also hired Piotr Sarnacki as a contractor to help with cucumber tests and to work on documentation (starting with the promotions stuff.) Piotr is a committer on the Rails project and will also be helping with the Engines improvements that are coming in Rails 3.1.
Many of our users are also fairly capable programmers and they can handle most of their programming needs. These users are asking for a paid support option to fallback on during those occasional instances where they need some additional assistance from a Spree expert. Other developers are comfortable handling the programming but would like a little bit of help setting up their development and/or production server environments. We are actively planning a paid support option to help address these needs. We’re still working on the details (including the cost) but there will soon be a paid support plan available and staffed by Spree core team members and other experts hired by Rails Dog.
Its been almost two years since I started thinking about the idea of a "fancy" demo for Spree. The idea is to build something that goes beyond the intentionally plain look and feel of a standard Spree install and show how a real world Spree site can look with a nice design and the proper extensions. This demo will also serve as a test bed for upgrading to newer versions of Spree (we can test out the migration process on a complex example store before we do a new release.)
We started work on the demo last week during our company retreat. We made good progress and we’re hoping to launch the demo by the end of this month. We’re also going to make the source code completely free and publicly available. This means that users can choose to fork our fancy demo as a starting point if they like most of the features and customizations it provides.
We’re also very interested in making Spree more accessible to those outside of the Rails development scene. One way of helping people to evaluate Spree is to have a one click install that would allow you to "try before you buy." Keep in mind we use the term "buy" to refer to the cost in terms of set up or development costs to get a real world Spree instance up and running. Spree itself is completely free and licensed as open source under the New BSD License.
We’re planning on developing a way to have the one click install process set you up automatically on a free Heroku instance. This way you are literally up and running on a free server within minutes. Keep in mind that this free instance would not really be suitable for large scale production use and would be intended for evaluation purposes only. Still, it would be possible to run a very small store on the free instance and no other ecommerce solution out there offers this (to our knowledge.)
Stable 1.0 Release
In addition to improved test coverage, we’re working to finish our overhaul of the source code and address any lingering areas that require attention. We’re hoping to stabilize the API and provide an easy upgrade path for future releases. In addition to a stable 1.0 release, we’ll be working on a more comprehensive set of documentation.
Improved View Customization
Finally, Brian Quinn has been working hard on a solution to the current limitations we have for customizing the default Spree views. Brian is on the front lines doing a lot of real world Spree deployments for paying clients. The biggest pain point is with view customization (and maintaining those customizations) and last week we got a sneak peak at his proposed solution. We’ll let Brian present his ideas in a separate blog post but I wanted to mention it here as well. Look for some dramatic improvements to view customization in the not too distant future.