SchoolTool 2012 Report
Written for SchoolTool Project by Tom Hoffman on 2012-12-05
Development
-----------
This was the Year of CanDo. We completely re-wrote CanDo from the ground up. It had been written mostly by high school students, and it was easier to start over than try to fix it in place. It needed a new data model to allow tracking of score revisions over time, UI update to fit SchoolTool's new design, better support for entering and editing skills through the web, and many other improvements to make it useful globally as a core feature of SchoolTool.
We had to have this ready to be deployed in production in 20 counties across Virginia this fall, and through close collaboration with our partners at the Arlington Career Center and Virginia CTE met that deadline and released CanDo as a standard component of SchoolTool 2.3 in October.
In that process we also spent a lot of time revising core SchoolTool functionality in response to feedback from Virginia, in particular re-designing our import spreadsheets to make that process more accessible.
Another big project has been to add an asynchronous task queue for long-running and scheduled tasks: data imports, report generation, sending emails, DB maintenance and such. The groundwork and initial implementation is done; it will be fully implemented and released in 2013.04.
Douglas learned the D3 JavaScript library and implemented some data graphics I designed for CanDo. We will be adding infographics to other parts of SchoolTool in 2013.
Justas improved support for LDAP authentication. In particular SchoolTool can one-click configure itself to work with the standard LDAP configuration used with Zentyal server (and in turn Edubuntu, if it becomes based on Zentyal as planned).
There were also many small fixes, changes and refinements following up the complete re-design of the UI released in November 2011.
Through the end of 2012 we are now focusing on printed reports, bringing both the plumbing and presentation in line with the rest of SchoolTool and expanding the selection for teachers and administrators.
Schwadesign did two contracts for us this year, a design guide for printed reports and a UX review of the setup process.
Jeff Elkner personally funded development of a SchoolTool Quiz module by Douglas (in the evening and weekends, mostly). This will probably be made available in our PPA in 2013.
Deployments
-----------
CanDo continues to be the case study of how a collaboration between a team of open source developers can work with educators. It was adopted by several of the largest (and most open source doubting) counties in Virginia this year and is playing an increasingly high stakes role in student and teacher evaluation. Through much of the summer we held Google Hangouts between the developers and our lead contacts in VA several times a week. On the other hand, CanDo sucked up most of our bandwidth for most of the year, limiting outreach in other areas.
At this point we have two substantial commercial deployments of SchoolTool which are fairly opaque to us. There is the Critical Links Education Appliance. We sent Justas to Portugal to help them update the version they distribute to 2.1. New this year was the Intelli deployment in the Philippines, where they are using SchoolTool to manage a system to send SMS messages to parents based on attendance. PoV did customization work for them.
One small commercial deployment in Thailand, via JaanSi.com, has produced a lot of useful feedback from just one school, and we have managed to release several fixes to issues they have raised. Hopefully they will be able to grow to serve more schools.
We have not found the BIG developing world rollout. Longstanding relationships with people in Cambodia and Nigeria continue to be stuck, we are still in contact periodically. The promising OLE Nepal pilot conducted last year concluded that the rural schools where they conducted the tests simply did not yet have the capacity or need for an online management system. We have a number of contacts in other countries, but it is a slow and uncertain process. I would note that we are not losing out to other products in these cases, just that the projects fail to launch.
Based on user contacts, individual school installations have increased, but not dramatically.
Development Plan for 2013
-------
The main theme is to iterate on existing components and respond to user requests and bug reports quickly. There is plenty to do.
We are going to work quickly to get our asynchronous task queue architecture into the 13.04 repositories. I do not think we will have to build additional packages, but we will need to test against some already in Ubuntu.
We are looking at JuJu and trying to figure out if a cloud strategy based around it makes sense for SchoolTool. We have a basic working charm.
Marketing
---------
CanDo seems to be the most marketable part of SchoolTool at this point, particularly among schools that have the wherewithal to implement it and potentially pay for customization or support. Now that the code is settled, we need to focus on documenting and promoting CanDo for US career and technical education as well as systems using outcomes-based methods worldwide.
