You should be optimizing for value produced per programmer-hour spent. If an open-source solution lets you do this better, then go with it; if it doesn't, then

You should be optimizing for value produced per programmer-hour spent. If an open-source solution lets you do this better, then go with it; if it doesn't, then don't. Note that "value" may be different for different situations. * If you're writing code for a bootstrapped one-person startup, i.e. this is a software product, website, or service that you're going to sell, then you would probably measure value as expected profit. * If you're writing code as part of your job, you should probably be trying to maximize the things that the software's supervisor(s) / user(s) want the software to do. This is usually (but not always) aligned with expected future profit/revenue for the business. * If the project is a learning experience to improve your own understanding and capabilities, then you should seek out unfamiliar technologies and interesting problems, and not worry so much about polishing the rough edges.