Can you use smart contract 'instances' rather than deploying each new contract?

Question

My apologies in advance for this noob question, I think a scenario might be the best way to ask:

I have a simple escrow smart contract that facilitates the buying of products. For simplicity sake, the contract is deployed at 0x01... and has 3 methods:

deposit() payable - buyer only

shippedItem() - seller only

finalize() - buyer only

and a simple web interface acting as an online store.

User 1 comes to the store and buys product by depositing in the smart contract. The seller marks the item shipped and the deployed contract state becomes 'awaiting finalization'.

User 2 comes to the store and buys product by...Im not 100% sure how? Would a new contract need to be deployed for this user? Or is there a way to instead have contract 'instances' that could be created and interact as independent objects? Ex:

--> Deployed contract at 0x01...

         --> Instance 1   State: awaiting finalization
         --> Instance 2   State: awaiting deposit 

I did see a question similar to this that said you could use what are called libraries to get this kind of behavior, but I'm still unsure if contract libraries fit this scenario. I ask this because from a practical business standpoint, having the buyer or seller deploy a new contract each time does seem like it could end up being quite expensive on the mainnet. Thanks for viewing my question.

Answer 1

I think this is better handle with better contract design. Instead of handling buy/sell transaction variables one-by-one, you can have a struct defining an escrow transaction. Something like:

struct escrowTx {
  uint256 TransactionID;
  address Seller;
  address Buyer;
  bool Executed;
  ...
}

Then each escrow transaction is just a new instance of this object and the smart contract manages it through its lifecycle. Then only one contract is needed.