Consider removing the c_nonce_expires_in parameter

[tplooker]

The c_nonce_expires_in parameter is intended to signal to the client/wallet when the c_nonce expires and therefore when a client/wallet should obtain a new one. However, given a credential issuer can invalidate a c_nonce at any stage this parameter is a bit of an un-reliable signal to clients/wallets. Practically this means clients/wallets will have to account for an error at the credential response endpoint due to an invalid nonce even if it isn't expired. For that reason I think we should remove the c_nonce_expires_in parameter and therefore simplify the nonce handling.

[bc-pi]

Concur with the removal of c_nonce_expires_in for the reasons @tplooker mentions and others like just general aesthetics. I was wanting to drop it with the work on #381 but decided to try and keep the scope of changes there narrow in hopes of landing[1] the PR in a timely manner.

[1] I think that's what the cool kids say nowadays

[bc-pi]

Now that #404 has been merged, it seem like a good time to consider doing this. I think it's just the nonce endpoint now.

Would others agree?

I don't think the "has-PR" label is correct BTW.