Sorry, this is just a pet peeve of mine in general.
All the time I see people argue against “unnecessary” abstractions. But this almost always comes from the perspective of “I don’t personally understand it yet” which is just not a reasonable bar for anything to have to clear. Second most frequently it’s “less clear” which often just means “I haven’t internalized it yet” which is likewise a terrible evaluation method for something like functional iteration methods that have the possibility of being used virtually everywhere. And almost equally as often, the underlying objection is that it isn’t useful because the old way is just fine thanks, which is typically a perspective that has completely forgotten about all of the sharp edges and bugs that we all just grew to accept from the preexisting approach.
All of these types of objections are knee-jerk reactions. There are good arguments against bad abstractions, leaky abstractions, infrequently-used abstractions, overly-complex abstractions, and all sorts of other failures to abstract well. But people are so used to these that they reflexively oppose any new abstraction as overly complex or unnecessary simply because it’s new to them.
And that’s a terrible perspective to have, because quite literally all of the progress that has ever been made in the practice of software engineering has been due to abstraction.
All the time I see people argue against “unnecessary” abstractions. But this almost always comes from the perspective of “I don’t personally understand it yet” which is just not a reasonable bar for anything to have to clear. Second most frequently it’s “less clear” which often just means “I haven’t internalized it yet” which is likewise a terrible evaluation method for something like functional iteration methods that have the possibility of being used virtually everywhere. And almost equally as often, the underlying objection is that it isn’t useful because the old way is just fine thanks, which is typically a perspective that has completely forgotten about all of the sharp edges and bugs that we all just grew to accept from the preexisting approach.
All of these types of objections are knee-jerk reactions. There are good arguments against bad abstractions, leaky abstractions, infrequently-used abstractions, overly-complex abstractions, and all sorts of other failures to abstract well. But people are so used to these that they reflexively oppose any new abstraction as overly complex or unnecessary simply because it’s new to them.
And that’s a terrible perspective to have, because quite literally all of the progress that has ever been made in the practice of software engineering has been due to abstraction.