Keyvan Nayyeri

Now it’s over four months of public existence of my Waegis spam filter service, and in this while it has grown smoothly. It’s also over three months of deploying the first stable version of the site and releasing it to the web. Since the beginning I had defined a roadmap for the next two major releases of the application and adding new features to the system. My initial plan was to release the second version by the end of November 2008 but to be honest, my recent business with personal stuff (that is the main reason to see the recent abeyance on my blog) hasn’t allowed me to accomplish this, and it looks like that I have to postpone the new release until the first quarter of 2009.

But I think that it’s worth sharing some parts of my plans publicly (and of course keep some exciting news private) because recently I experienced an acceleration in the number of clients, and some of them contact me to ask about the possibility for using free commercial services for their sites. Therefore, it would be interesting to know some parts of my plans for the future of the service.

I think that the most important concern for the clients of a spam filter service is the accuracy of its blocker and its capability to defeat spams, so I would say that Waegis has been able to do an excellent job in this area, and feedback have been very positive even though there has been a reasonable tolerance for the accuracy of the service for some sites in comparison with others, and I should work more on this. However, one of the most important factors in the success of such services is the amount of data that they have to act, and now I have much more information about the spam characteristics that I didn’t have at the start point, and I can use them to implement new spam rules. Thus, new spam rules that refine the quality are the first update on the service.

The second area of concerns is about the usability of the service and easing its usage. In fact Waegis was built in a simple way at the start point but there are some minor points that I can consider to simplify its usage. Still, the nature of this service requires some steps to be done by end users specifically if their site or engine doesn’t have a built-in support for the service, so I hope that the quality of the service can convince more engines to support it out of the box.

Limitations of the service based on commercial and free accounts is one of the main stuff that has been a question for clients in the past couple of weeks because it looks like that more clients are interested to have less limitations. Here I would say that my main plan is to move the solution to a completely free service but it cannot happen at once. The only reason to offer commercial services and put limitations is the costs of the site (i.e. hosting and maintenance) but fortunately I could get good experiences to handle it better, so you can expect a free service in the next upcoming months. In the short time, I’m going to reduce the limitations in the next three weeks.

Besides, there are some significant changes in the infrastructure of the service to apply some new technologies and scale up the software based on new requirements. I’d prefer to write about these changes after applying them because there would be some exciting news for everyone.

One question that has been asked from me is about the possibility of rewriting Waegis site with ASP.NET MVC, and my answer is certainly negative. It’s a wrong idea that some guys have to use MVC for everything, and changing a software that is already on production for no reason. Interestingly, such guys are those who just hear an announcement and start following that without knowing the principles. ASP.NET MVC has its applications and Waegis site is not a good place to use it. Who was complaining?!! Waegis is written with ASP.NET web forms and it has proven to have less issues (actually I never received any issue about the site after going to RTW). For Waegis I stick with web forms but most likely I’ll have some other projects that will be built on top of the ASP.NET MVC.

Moreover, I’m going to implement OpenID authentication on Waegis for users to be able to use their universal identity to log into the site and start using the service. Of course, the infrastructure for this feature is already a part of the codebase but I preferred to not include it in the first version of the application. My primary goal is to integrate the authentication with forthcoming Microsoft Live and Google IDs.

Furthermore, I have some long time plans to release the source code of the core system as an open source project while I keep the logic of spam rules private because this is the elixir of any service. However, this has the last priority in my queue for future changes and wouldn’t happen sooner than Summer 2009.

By the way, this was an outline of what is proposed for the future of Waegis but as always, I’m listening and you can drop me a line describing your comments, suggestions, and opinions for this service.