In software engineering (and machine learning), we often need to name things to refer to them more easily in our discussions – be it projects, products or servers.
Naturally, we want to give descriptive names, so that the name expresses the nature of the thing. What makes this difficult is that the name also needs to distinguish it from other closely related things, for example, variations of the same ML model. In addition, the name should be short, readable and easy to remember.
The nature of these things also tends to evolve over time. What started out as a project with a specific purpose may develop into a more general product. A server initially deployed for a particular API, may end up running multiple services. And so, a name that once described the thing very accurately can quickly become meaningless and confusing.
In these cases, it’s preferable to follow the naming criteria laid out in this Request for Comment on how to name your computer.
Also see these posts: