4 ms·
It might have been an effective thesis if it didn't start off by defending coding academies. I think it's a given today that your ability to do the job trumps
by critthinker42 11y ago
It might have been an effective thesis if it didn't start off by defending coding academies. I think it's a given today that your ability to do the job trumps your ability to generate credentials for the job, but stripping specialized developers is setting yourself up for failure. Also, in many cases it makes sense to have at least someone on the team with some academic exposure to specific problem domains. Most people can write some code to accomplish the task, some people can be clever about it, but sometimes you also need people who can refer back to their <data structures, algorithms, math, whatever> classes so they can either write some performance critical code, solve a hard math problem, or assess the engineering design decisions from a more trained perspective -- especially in an industry dominated by egos, is ego + lack of specialized knowledge always the best approach? Certainly it gets stuff done. Also, as devs get older (years more experience picking up specialized knowledge along the way), this problem can obviously be mitigated.
Edit: to be fair, I mean stratification is appropriate at a project level of organization level, clearly you should be pigeonholed.