My point is that classifying corporate:bad, startup:good is just unsafe thinking. If you look at the company on its own without stereotyping its demographics, you'll come out better in the end.
This past weekend I managed to pick up a book on Go and hoping to deploy something through appengine by the end of week. If you want something, you find the time.
And I work 48 hours a week minimum.
Like most people here, I'm often learning something new, or new aspects of the familiar. Despite the inertia, large comps need people to lead the way, whether it's training how to properly use source control, teaching new languages or platforms, setting up application standards to make troubleshooting possible and aid comprehension for maintenance.
The problem is actually having a large amount of existing code, and having to live with it for a few years, rather than a blank slate to be written with whatever Apache (or others) just released last week. Don't think that code won't feel old in 2 years.
DHH did a wonderful talk about "legacy code" a couple years back, I think it's on the IT Conversations site somewhere.