What you do matters more than what you intended
This isn’t an “actions speak louder than words” sort of post, although I suppose they’re related thoughts. It’s more about the idea that whatever you support trumps what you intended to support. It’s more of an XKCD #1172 post.
On a mailing list I’m on, there was recently a thread about NOAA preparing to retire a satellite-based data transmission service called NOAAport. One person wrote:
But what I think people are missing is NOAAPort was never truly designed to serve anyone outside of NOAA. It is a service by NOAA, for NOAA, that we were lucky enough to be able to tap into because it was clear-air transmission.
Another participant responded:
Whatever its original design motivation, NOAA has knowingly operated NOAAPORT for decades as a primary public dissemination mechanism. Entire academic, commercial, emergency-management, broadcast, and value-added ecosystems were built around that availability with NOAA’s knowledge and participation. After decades, its actual role matters significantly more than the original design intent.
This is an important lesson for maintaining an open source project. Whether it’s a feature in the software or a service you provide, once it’s there, people will expect it to remain there.
If you say a particular feature or configuration is unsupported, but then you start supporting it, it’s now supported. You’re allowed to change your mind, of course. You can start out not intending to support something and then later decide to do it. But if you claim you’re not and you do, then you are.
This means that if you decide you’re not going to support something, then you have to continue to say “no” to it whenever it comes up. Saying “no” is one of the harder things we have to do, so it’s good to remind ourselves sometimes.
This post’s featured photo by Maayan Nemanov on Unsplash.
