A few weeks ago, a coworker asked about adding and additional library to parse some data. The current library is at EoL, and we've had to make a few in-house pa

A few weeks ago, a coworker asked about adding and additional library to parse some data. The current library is at EoL, and we've had to make a few in-house patches plug the holes. We don't currently have time to do a complete replace and refactor (though we're working it in the schedule). I was asked, "Why can't we just use them in tandem? Does it matter if we parse the data twice in this instance?" No. In this instance, it doesn't matter for this specific data to be parsed twice. The overhead would be minimal. But what if we stopped refactoring code and appended new feature onto existing code? What if we didn't bother to remove defunct API calls, and instead, we discard the old responses and just made new ones with new data? That's why we don't just launch satellites. 4 (I think it was 4) additional objects 10cm across won't have much of an affect. The problem is when everyone starts launching satellites of various sizes no matter if they can be tracked or not.