Edge case

Term popularized by Alan Cooper in his interaction design classic "Inmates are running the asylum".

Affiliate example

Let's say that you want us to add an affiliate feature to Bakesale. Let's dissect this:

• 3% of users may benefit
• but for 97% it is just in the way
• a huge amount of development time is used to build and maintain it
• more code, more problems (bugs, limitations etc, etc) to the entire application.
• a downward spiral of features (affiliate levels to users, commission levels to products etc, etc)

All this because a couple of merchants did not want to pay a little extra to join a hosted affiliate program or hire someone else to create the extension or did not want to take the time to build one themselves.

After adding a dozen of these features, Bakesale will not be very simple anymore. It would be just another me-too shopping cart offering everything to everyone and being nothing special for anyone.

This is the reason why we say "no" by default to new features, not (just) because we are lazy.

If you need something just for your shop that you believe would be useful to your customers or you, there is always the possibility of adding an extension. If no fitting extension is available, you can always hire the core develepment team, or another developer, to make it for you. Then you will have an affiliate system that works the way you want it to. This is the power of Open Source.

And when the cost (time or money) is yours to pay, you will think harder if this is really needed. I have seen too many features added to stores just because someone thought that someone *may* find them useful. In the end, these just get in the way of daily tasks