Keyvan Nayyeri

Back in May 2008 I introduced a new project that I was working on for almost three months before then and was planning for it for almost three years, and that project was Waegis, an online spam filter service that targets modern spam types and is written completely as an enterprise .NET solution based on .NET Framework 3.5 and SQL Server 2008. In July and after some private tests, I released the public Beta version of the site and service and finally Waegis went to the stable stage in August.

Since the infancy stages of planning and development, I had to seriously consider several aspects of development, maintenance, marketing, promotion and business in the project and make sure that the project survives and scales up in the long run especially in regard to the existence of a few competitors that had big companies in their back such as Akismet, TypePad anti-spam and Defensio.

While I developed something that I personally liked it, my initial decision for the business of Waegis was to have an oldie but goodie plan consisting of cheap commercial accounts for sites with high usage and offering free services to smaller sites, a plan that was tested successfully in many circumstances even by some competitors!

After almost three months of public existence of Waegis and smooth refinement of its features, quality and accuracy rate and of course, collecting a good number of users and sites, recently I’ve been refactoring some parts of the codebase to compact the project size and make it easier for maintenance and extension. Simultaneously I was also thinking about the business and marketing to foresee the future of the service and prepare for what is going to happen in this field.

From a technical point of view, Waegis is in an excellent position where it’s smoothly getting more clients and is becoming better day after day. A closer look at the progress of its accuracy rate displays this smooth progress and persuades me that this service is going to act much better than competitors. Those who have used the site and service can leave their comments on this but Waegis is comparatively faster and has better features to offer.

After development and implementation of Waegis, now I think that at the moment such a solution can’t accomplish what it deserves on our web because there isn’t a good balance between the cost of development, maintenance and administration and the income whether as money or as the acclaim and support by the community. While I’m sure that in the longer run, things turn out to be much better but there is an acme for this lifecycle in every period of time that still wouldn’t be something deserved for such a project.

Building an online anti-spam required much effort especially in the knowledge and work because such a solution is not a regular data in/data out system with well-known documented patterns and requirements, and having an effective service obliges more effort and examinations. You can’t build a software that simply catches all spams without careful studies and analysis on the behavior of spammers and adapting your mathematical methods with their changes.

On the other hand, there isn’t a widespread awareness of such online services and how they can be the best solution for filtering spams. Normal users can not distinguish the advantages of these services, so they wouldn’t pay enough attention to these solutions.

I just make it shorter and summarize my recent analysis here. After some careful examinations on Waegis, I ended up with some perceptions and in the future I may make some decisions regarding these conclusions to help the progress of this project.

The first and most important finding is the fact that the future of online web services ends in free services because the progress of the web proves that new solutions break the market and force vendors to reduce their commercial restrictions or eliminate them completely. Same is true in the field of online spam filter and the progress represents this fact, so in the upcoming months I may reduce the limitations of the service to move toward an open free service. Besides, like any other project, Waegis requires promotion and advertisement and I would consider this as well.

The second perception is the inability of the .NET community to be a good target field for Waegis and I have to extend my target field to a more general solution. I should leave my emphasis on the .NET community to be able to keep the positive progress of the service. As I have stated before, whole Microsoft thing is slow in adeptness with something new and the .NET community is not an exception! To be honest, I found that many of the members of the community can’t understand the role of Waegis!

The last conclusion is the repetition of something that I stated above: there is not a good balance between efforts of development and maintenance and the returning income and this is a stop point for putting further efforts into the project. Primarily I started this project to train my skills and knowledge in various fields (software architecture, software development and mathematics), and I already have gotten what I expected to get from the end result, so it’s better to get more energy from other people on this thread. I may accomplish this by turning the project into an open source project and releasing the whole source code (or the core part) under an open source license. As an alternative option, I may turn the project into an academic project but both options wouldn’t happen in the short time.

I’m 100% open to your thoughts and feedback, so feel free to drop me a line and let me know how I can make Waegis better.