In a proprietary project if your boss says "do this" you either do it or find another job.
No, what's much more common in the proprietary world is that you tell your boss "Say, we really need to fix this code before it costs us customers or becomes a security risk" and your boss replies, "Listen, Ace, as soon as you become CFO you can initiate that little project and go explain to our stockholders why we blew $54 million retasking 1,200 of our people to rewrite something we should have coded properly 10 years ago . But until the time you're named CFO, Ace, the Marketing and Accounting Departments dictate the shots. And they're happy with the new-customer-to-lost-customer ratio. Now, go get me coffee."
Real-world examples:
Microsoft's ActiveX (security experts first pointed out how dangerous it was in a networked environment way back in '94; it remains unfixed 15 years later despite an average of one new major ActiveX exploit-in-the-wild per month since then).
IBM's Lotus Notes (which IBM cheerfully refused to do anything about until 2007 despite it creating an entire cottage industry of Websites chronicling its many failures; finally God Himself intervened and threatened Sam Palmisano with enternal damnation unless he did something about it)
Adobe Creative Suite (which, in each subsequent Windows version, becomes less compatible with SMB and spends more and more CPU cycles attempting to find unauthorized copies of itself on your network and fewer cycles focusing on the task you just gave it.
* * * * *
"All my life, I always wanted to be somebody. Now I see that I should have been more specific."
—Jane Wagner
