But it's counterproductive for the company, and some experiments a developer may want to quickly do will never be done at all. It kills a lot of creativity. The same holds, of course, for restricting internet access so a dev might never find out whether an interesting looking github tool really is good (for doing productive work!). Also, the company probably wants their products to run on a diverse zoo of machines and OS versions prior to release.
So why a company would take away such access on dev machines is a mystery to me -- probably misguided (and futile) attempts to keep in 'control'.
What do you mean?
Emergencies can happen where you need to stay an hour or two after EOD, but if it’s more than once a year I’m out, personally, local admin permissions or not.
The actual question I’d be interested in is how do developer convenience and security stance balance against each other, and how are the inevitable conflicts resolved.