The following is a snapshot of what we were working on during February 2010. There were 12 different customers where we put in more than a handful of hours during that month.

Technologies and Platforms

• 6 web 2.0 style (AJAX) web applications, developed with .Net/SQL and deployed on Windows.
• 1 web application with an applet, developed with Java/Jsp and SQL Server, deployed on Windows.
• 5 web applications developed with php and mysql, deployed on Linx.

Type of Product

• 5 are productivity tools for the business.
• 7 are the product that the business sells.

Status

• Of the seven products:
• Four are live, where one has been generating revenue for seven years, and three were launched within the last year. Of these, one generates revenue via ads, and the other three are pay-for-use.
• The other three have not been released yet.
• In general we implement software to improve productivity in small enough chunks so that it is in use soon after we start the project, so all of the five productivity tools are in use.

Development History

• 5 of the projects were started by Pacific Northwest Software
• 7 were started by other developers but our customer was not satisfied with them and found us to take it over.
• Of these, two had a pile of code that had significant problems in several areas, such that nothing worked and everything had to be thrown out.
• For the other five, we continued with the existing software and overhauled portions of it when it made sense.

Business Issues

• One that has not been released has a difficult business problem because the product is very useful after many customers are using it, but less so at the start. This makes the start up plan difficult, but if the business can gain enough customers it will take off and be unstoppable. The customers are a particular type of business so they can be targeted fairly easily. We can preload the system with enough customers to make it useful, but the difficulty is to give the customers enough functionality to make it useful, but hold back some functionality to encourage them to pay for it.
• One can be deployed with a number of different brands each with different behavior all based off a common database with configurable details for what specific information is shown to the user.

Technical Issues

• The application that generates ad revenue has significant performance requirements. It is deployed on many inexpensive computers to ensure uptime and responsiveness. The application is a heavily modified deployment of Wordpress. Most of the content is not unique to each viewer so the bulk of the hits are satisfied with a hit to a cache provided by nginx. The content of the pages have been optimized to make the browser load the different content, ads, images, and text such that it renders without resizing as the different items arrive, and such that it provides the content that the viewer is interested in first.
• One is used by blue collar workers which makes the user interface very challenging. The workers generally want to get the job done and go home. The software is used to collect information about the job they are doing instead of really helping them do the job, so the worker doesn't see much value in using it. If software tries to get too much detail or if the user interface is less than trivial to use, they will simply push buttons to satisfy the software without really providing the information. The end result will be useless data for the owner, and it is possible the owner won't recognize that the info is useless.