[cweill]

Exactly. One of the biggest things I keep in mind these days is that most people, myself included, are bad at knowing what's most important to get done. It may seems that more code is more work is good, but the banter with colleagues improves communication and cohesion which can have a much higher impact than writing that additional code. If a higher up tells you to do something, you can be pretty sure it's important (at least to your report chain), otherwise it can be tough to identify relative importance of work activities.

[cherryturnover]

Even letting a higher up down is expected. You have to feel it out though if the certain thing is one of those. But if you know it can't be done or you got time, one of the best things you can do is abandon ship. It's really absurd and kind of a bad thing to lace your coworkers with, but if you've got references elsewhere and people within whom you're on good relations with, it won't affect your career one bit. Not only will you not get the blame for it when things aren't done presently, but you don't have to deal with the impossible task management gives you! Which in the grand scheme, if you can stick it out long enough before they fire you and get a job elsewhere, on paper, you never failed! You always achieved your goals.

The entire Gervais Principle goes into detail on this and it's a fun read. A lot of it sounds like just general office politics written in a funny way, but there are some nuggets that a lot of people who aren't in the "sociopath" category (the authors words), aren't willing to do. It's kind of like "the best time to be looking for a new job is now" philosophy in work.