I've worked on tools that were slightly misnamed after 6 months, and completely misnamed after 2 years. At that point they were also usually just nearly useless due to feature bloat and/or lack of scalability, so deprecated or replaced with something better.
They didn't change names, but their successors would get a new one.
Isn't it even harder to re-define names in a company? There might be 3 people involved in re-definition, but it affects 15 people.
How are we going to notify those 15? Do we even know who those 15 are? Are we going to create a weekly redefinition newsletter?
I think in most cases new meaning deserves a new name. Everything else is just hacks.
How hard is it to change a name is a actually a really good metric for a company. If a simple rename takes several days, multiple approvals, rounds of QA, and a scheduled release next quarter, then you probably need those hacks.
I think that's very extreme. Products grow at a gradual pace. I don't think there are defining moments when a product no longer supports something, or is no longer used in a way that it was intended to.
I would argue it's easier to maintain peoples understanding of a product since that will also be done gradually. It's not easy to update naming inside of a code base without potentially breaking software significantly or causing unknown bugs elsewhere. I think most software would fail the renaming test. It's also generally not worth the money and time needed to make that change.
khazhoux|3 years ago
Half the article is like, "There was a component called YamlParser, which is now a browser-based stable-diffusion renderer!"
ThePadawan|3 years ago
They didn't change names, but their successors would get a new one.
badhombres|3 years ago
version_five|3 years ago
hbrn|3 years ago
How are we going to notify those 15? Do we even know who those 15 are? Are we going to create a weekly redefinition newsletter?
I think in most cases new meaning deserves a new name. Everything else is just hacks.
How hard is it to change a name is a actually a really good metric for a company. If a simple rename takes several days, multiple approvals, rounds of QA, and a scheduled release next quarter, then you probably need those hacks.
badhombres|3 years ago
I would argue it's easier to maintain peoples understanding of a product since that will also be done gradually. It's not easy to update naming inside of a code base without potentially breaking software significantly or causing unknown bugs elsewhere. I think most software would fail the renaming test. It's also generally not worth the money and time needed to make that change.