Scrypto Concepts
Here is some orientation material for Scrypto programming. If you are super eager to start coding then you may skip this section for now. Just understand that Scrypto probably won't make a lot of sense without an appreciation of these concepts.
Scrypto and Rust are joined at the hip. You can't become a Scrypto programmer without learning a good bit about Rust in the process. Fortunately there is a lot of good training material for Rust. To save you time, this tutorial includes an appendix that points out which parts of Rust you need to understand. The rest you can pick up on an as-needed basis.
In the realm of programming languages Rust is considered medium level in complexity. However most of the advanced aspects of Rust do not come into play when coding with Scrypto. Therefore beginning programmers can start with Scrypto and expect to see real results from their efforts. Having said that, truly mastering Scrypto does take you deep into the Rust rabbit hole and so don't expect complete mastery to come about quickly.
Scrypto has a more sophisticated way to deploy logic to the ledger than first generation smart contract protocols. Let's dive in.
Blueprints are compiled source code that live on ledger in a form where it is available for anyone to reuse although there may be an author specified royalty involved with doing so. Blueprints never have state and so in that sense they are not active. Instead they provide one or more constructor functions that allow others to instantiate them. These constructors may have any number of arguments that can parameterize the operation of the blueprint's code. So while blueprints tend to be highly specialized in terms of their functionality, they may also be used to support many different use cases depending on exactly how they are instantiated.
In some cases blueprints are designed to work closely together with other blueprints and in this case they can be deployed together as a package. Deploying them together in this way means that a given blueprint can be sure that some other different blueprint exists and therefore its code may instantiate them as part of their operations. In addition it is reasonable to presume that deploying blueprints together as a package may save on deployment costs. Like blueprints, packages do not manage state information.
Note that the full details on the many ways that blueprints can be manipulated in order to make them work together on your behalf are not yet available.
To activate a blueprint you instantiate it by calling one of its' constructor functions. When this process completes you get the address of the newly created instance which is called a component. Components manage state and can gather, hold and distribute resources according to the logic provided in its' associated blueprint. Technically components never die however there may be logic within a blueprint that makes a component useless after certain conditions are triggered.
So in Scrypto components are the closest thing to what we think of when we think of a smart contract. However all of the smarts in a component derive from the logic that is defined in the blueprint that gave birth to that component via one of its constructor functions.
So can we say that Radix has smart contracts? In terms of capabilities, definitely yes. However the package/blueprint/component model is much more configurable. Therefore the alignment between components can be far tighter and safer. Ultimately a lot more functionality can be delivered with a lot less code than with traditional smart contracts.
Part of what makes Radix safer than other protocols is the design and implementation of the Radix Engine which helps drive the ledger. The core capabilities of all tokens, badges, NFTs and other resources on the Radix ledger are defined by the Radix Engine making their operations fast and predictable. As you will see, the decision to treat all assets and resources on the ledger similarly does not preclude third party Scrypto programmers from adding a lot of additional value to tokens and other resources by adding additional logic.
Another aspect of Radix that is crucial for DeFi is scalability. The Cerebus Consensus mechanism is the technology that allows Radix to take advantage of practically infinite sharding without giving up atomic composability. What this means for the Scrypto programmer is that they can create and deploy components that can leverage other deployed components in a variety of ways even if the requirement is to perform these operations atomically within a single transaction. You will only get a small taste of that aspect of Scrypto in this tutorial, but for advanced projects the ability to potentially leverage the entire ecosystem in this manner should create an unparalleled multiplier effect for everyone who participates.
Asset-oriented programming incorporates assets as first class types in the language. This allows the compiler and runtime to reason about asset handling using specially designed safety checks that protect programmers and end users alike. The resulting code is generally atomic and can entirely prevent certain types of bugs related to double spending, unauthorized access, misplaced assets, reentrancy and more.
Radix claims that Scrypto goes even further in this regards than earlier experimental languages such as 'flint' and 'Cadence'. For instance, in Radix the set of all tokens protected by the compiler and runtime extends to the core XRD protocol token as well. In addition there are access control entities known as badges that get the same added protections as tokens. Accordingly Radix refers to tokens, badges and similarly handled entities as 'resources' and they all get similar first class treatment in terms of safety and efficacy.
The early release of Scrypto implements tokens and badges. NFTs should also be supported either in the Alexandria release or soon thereafter. We will discuss the details about creating and managing these resources later in this tutorial, but here is the overriding rule: All resources must be held in a resource holder of some kind, at all times!
A Resource Holder is a container for resources. The Bucket is the workhorse holder. You pass resources around in buckets and manipulate their contents by moving them between buckets and vaults. However buckets are temporary. They must be emptied and/or burned by the end of a given public function or method.
If you need to hold resources more permanently within a component, you must use a Vault. For that reason you will see Vaults in blueprint structs, but never Buckets. As we shall see it is possible to do sophisticated things with Vaults such as add or drop them dynamically as needed. Just keep in mind that it will become prohibitively expensive if your component is designed in such a way as to require a huge number of Vaults.
It will pay off well to study all of the available functions and methods for Buckets and Vaults.
An Account is a resource holder that models a wallet within the simulator. Some of the details about interacting with and defining Accounts are not yet settled and so their eventual importance for Scrypto development is not well understood yet. We recommend avoiding them if at all possible since they are certainly going to change a lot going forward.
Taken as a whole the Radix stack, from the low level protocol to the highest level abstractions, is a lot to digest. No one should expect to master it all in a hurry all. The information on this page is just a brief introduction. Still we hope it helps orient you to the programming concepts that we are about to present to you.
Now let's get started with Scrypto!
Last modified 1yr ago