Put them together either too fast or too slow and you'll just get a fizzle, not a BOOM. Likewise, if the shapes aren't precise, your odds for a fizzle increase, and shaping Uranium into a proper core is non-trivial.
Nor can your gun-type barrel permit the gasses to break away a piece of the detonator bullet/core, or even erode any sizable part of the core, lest that radioactive material reach the intended center of the globe/donut far enough in advance to fizzle your device.
If you've ever paid close attention to a firearm firing on the range, you've probably noticed that some gasses actually leave the barrel *prior* to the bullet exiting your rifle. With an open lead ball, some tiny part of that sort of bullet is actually eroded externally into lead gas that likewise escapes your muzzle-loader, for instance, prior to the bullet itself escaping.
Now picture that same erosion occurring with an open Uranium or Deuterium bullet, sending radioactive gasses into your core prior to the critical mass being assembled.
These are non-trivial issues, and they only scratch the surface of the problems faced by getting a working nuke to become operational and reliable.
Yes, it can be done, and having the fissionable material is a big step toward getting there, but the rest of the equation is still non-trivial is my point.
As an aside, which would you find more unnerving: a detonated nuclear bomb...or a dud? If the former, would you volunteer to defuse the latter?
This question was posed to me once, years ago. I'd be interested in your answer.