Should a Netflix or Twitter-style web service use REST or SOAP?

I've implemented two REST services: Twitter and Netflix. Both times, I struggled to find the use and logic involved in the decision to expose these services as REST instead of SOAP. I hope somebody can clue me in to what I'm missing and explain why REST was used as the service implementation for services such as these.

1. Implementing a REST service takes infinitely longer than implementing a SOAP service. Tools exist for all modern languages/frameworks/platforms to read in a WSDL and output proxy classes and clients. Implementing a REST service is done by hand and - get this - by reading documentation. Furthermore, while implementing these two services, you have to make "guesses" as to what will come back across the pipe as there is no real schema or reference document.

2. Why write a REST service that returns XML anyway? The only difference is that with REST you don't know the types each element/attribute represents - you are on your own to implement it and hope that one day a string doesn't come across in a field you thought was always an int. SOAP defines the data structure using the WSDL so this is a no-brainer.

3. I've heard the complaint that with SOAP you have the "overhead" of the SOAP Envelope. In this day and age, do we really need to worry about a handful of bytes?

4. I've heard the argument that with REST you can just pop the URL into the browser and see the data. Sure, if your REST service is using simple or no authentication. The Netflix service, for instance, uses OAuth which requires you to sign things and encode things before you can even submit your request.

5. Why do we need a "readable" URL for each resource? If we were using a tool to implement the service, do we really care about the actual URL?

A canary in a coal mine.

I have been waiting for a question like this for close to a year now. It was inevitable that this day would come and I am sure we are going to see many more questions like this in the coming months.

The warning signs

You are absolutely correct, it does take longer to build RESTful clients than SOAP clients. The SOAP toolkits take away lots of boilerplate code and make client proxy objects available with almost no effort. With a tool like Visual Studio and a server URL I can be accessing remote objects of arbitrary complexity, locally in under five minutes.

Services that return application/xml and application/json are so annoying for client developers. What are we supposed to do with that blob of data?

Fortunately, lots of sites that provide REST services also provide a bunch of client libraries so that we can use those libraries to get access to a bunch of strongly typed objects. Seems kind of dumb though. If they had used SOAP we could have code-gen’d those proxy classes ourselves.

SOAP overhead, ha. It’s latency that kills. If people are really concerned about the number of excess bytes going across the wire then maybe HTTP is not the right choice. Have you seen how many bytes are used by the user-agent header?

Yeah, have you ever tried using a web browser as debugging tool for anything other than HTML and javascript. Trust me it sucks. You can only use two of the verbs, the caching is constantly getting in the way, the error handling swallows so much information, it’s constantly looking for a goddamn favicon.ico. Just shoot me.

Readable URL. Only nouns, no verbs. Yeah, that’s easy as long as we are only doing CRUD operations and we only need to access a hierarchy of objects in one way. Unfortunately most applications need a wee bit more functionality than that.

The impending disaster

There are a metric boatload of developers currently developing applications that integrate with REST services who are in the process of coming to the same set of conclusions that you have. They were promised simplicity, flexibility, scalability, evolvabilty and the holy grail of serendipitous reuse. The characteristics of the web itself, how can things go wrong.

However, they are finding that versioning is just as much of a problem, but the compiler doesn’t help detect issues. The hand written client code is a pain to maintain as the data structures evolve and URLs get refactored. Designing APIs around just nouns and four verbs can be really hard, especially with RESTful Url zealots telling you when you can and cannot use query strings.

Developers are going to start asking why are we wasting our effort on support both Json formats and Xml formats, why not just focus our efforts on one and do it well?

How did things go so wrong

I’ll tell you what went wrong. We as developers let the marketing departments take advantage of our primary weakness. Our eternal search for the silver bullet blinded us to the reality of what REST really is. On the surface REST seems so easy and simple. Name your resources with Urls and use GET, PUT, POST and DELETE. Hell, us devs already know how to do that, we have been dealing with databases for years that have tables and columns and SQL statements that have SELECT, INSERT, UPDATE and DELETE. It should have been a piece of cake.

There are other parts of REST that some people discuss, such as self-descriptiveness, and the hypermedia constraint, but these constraints are not so simple as resource identification and the uniform interface. The seem to add complexity where the desired goal is simplicity.

This watered down version of REST became validated in developer culture in many ways. Server frameworks were created that encouraged Resource Identification and the uniform interface, but did nothing to support the other constraints. Terms started to float around differentiating the approaches, (HI-REST vs LO-REST, Corporate REST vs Academic REST, REST vs RESTful).

A few people scream out that if you don’t apply all of the constraints it’s not REST. You will not get the benefits. There is no half REST. But those voices were labelled as religious zealots who were upset that their precious term had been stolen from obscurity and made mainstream. Jealous people who try to make REST sound more difficult than it is.

REST, the term, has definitely become mainstream. Almost every major web property that has an API supports "REST". Twitter and Netflix are two very high profile ones. The scary thing is that I can only think of one public API that is self-descriptive and there are a handful that truly implement the hypermedia constraint. Sure some sites like StackOverflow and Gowalla support links in their responses, but there are huge gaping holes in their links. The StackOverflow API has no root page. Imagine how successful the web site would have been if there was no home page for the web site!

You were misled I’m afraid

If you have made it this far, the short answer to your question is those APIs (Netflix and Twitter) do not conform to all of the constraints and therefore you will not get the benefits that REST apis are supposed to bring.

REST clients do take longer to build than SOAP clients but they are not tied to one specific service, so you should be able to re-use them across services. Take the classic example, of a web browser. How many services can a web browser access? What about a Feed Reader? Now how many different services can the average Twitter client access? Yes, just one.

REST clients are not supposed to be built to interface with a single service, they are supposed to be built to handle specific media types that could be served by any service. The obvious question to that is, how can you build a REST client for a service that delivers application/json or application/xml. Well you can’t. That’s because those formats are completely useless to a REST client. You said it yourself,

> you have to make "guesses" as to what > will come back across the pipe as > there is no real schema or reference > document

You are absolutely correct for services like Twitter. However, the self-descriptive constraint in REST says that the HTTP content type header should describe exactly the content that is being transmitted across the wire. Delivering application/json and application/xml tells you nothing about the content.

When it comes to considering the performance of REST based systems it is necessary look at the bigger picture. Talking about envelope bytes is like talking about loop unwinding when comparing a quick-sort to a shell-sort. There are scenarios where SOAP can perform better, and there are scenarios where REST can perform better. Context is everything.

REST gains much of its performance advantage by being very flexible about what media types it supports and by having sophisticated support for caching. For caching to work well though nearly all of the constraints must be adhered to.

Your last point about readable urls is by far the most ironic. If you truly commit to the hypermedia constraint, then every URL could be a GUID and the client developer would lose nothing in readability.

The fact that URIs should be opaque to the client is one of the most key things when developing REST systems. Readable URLs are convenient for the server developer and well structured URLs make it easier for the server framework to dispatch requests, but those are implementation details that should have no impact on the developers consuming the API.

The Twitter API is not even close to being RESTful and that is why you are unable to see any benefit to using it over SOAP. The Netflix API is much closer but it’s use of generic media types demonstrates that failing to adhere to even a single constraint can have a profound impact on the benefits derived from the service.

It may not be all their fault

I’ve done a whole lot of dumping on the service providers, but it takes two to dance RESTfully. A service may follow all of the constraints religiously and a client can still easily undo all of the benefits.

If a client hard codes urls to access certain types of resources then it is preventing the server from changing those urls. Any kind URL construction based on implicit knowledge of how the service structures its urls is a violation.

Making assumptions about what type of representation will be returned from a link can lead to problems. Making assumptions about the content of the representation based on knowledge that is not explicitly stated in the HTTP headers is definitely going to create coupling that will cause pain in the future.

Should they have used SOAP?

Personally, I don’t think so. REST done right allows a distributed system to evolve over the long term. If you are building distributed systems that have components that are developed by different people and need to last for many years, then REST is a pretty good option.

SOAP is an object-oriented, remote procedure call technology stack. It works by building a new abstraction on top of an existing protocol (HTTP).

REST is a document oriented approach, that simply uses the features of an existing protocol (HTTP). "REST" is just a buzzword -- the concept is this: Just use the web the way it was designed to work!

In response to edits to question:

1. "Implementing a REST service takes infinitely longer than implementing a SOAP service."

Um, no, it can't be infinitely longer. And in cases where what you are trying to retrieve is already a document or file, it's actually much faster. For example, the OGC spec for WMS (Web Mapping Service) defines both a SOAP and REST version of the protocol, and there's a reason why almost nobody implements the SOAP version -- it's because if you're trying to get a map, it's a lot easier to just build a URL and fetch image bytes from that URL than it is to bother with encapsulating it into a SOAP message. But yes, I will agree that if the point of the web service is to transfer some strongly-typed object in a domain object model, SOAP is better suited for that use.

2. "Why write a REST service that returns XML anyway?"

Well, yes, that can be silly. But it depends on what the XML is. If there's a clearly defined schema for it somewhere, then there's no ambiguity. For example, you can think of WSDL URLs as being a kind of RESTful web service for retrieving information about a web service. In this case, adding the overhead of another SOAP request would be pointless.

In general, REST wins when the content that is being transferred can be thought of as a file, as a single unit. SOAP wins when the content needs to be treated as an object with members.

3. "I've heard the complaint that with SOAP you have the "overhead" of the SOAP Envelope. In this day and age, do we really need to worry about a handful of bytes?"

Yes. Not in every circumstance, but there are sites with a great deal of traffic where it makes a difference. Is it enough of a difference to outweigh the semantic differences of using SOAP instead of REST? I doubt it. If you're doing an object remoting protocol and the number of bytes is making a difference, SOAP is probably not the tool for you anyway -- maybe you should be using CORBA or DCOM instead.

4. "I've heard the argument that with REST you can just pop the URL into the browser and see the data."

Yes, and this is a large argument in favor of REST if it makes sense to view the data in a browser. For example, with image data, it's an easy way to debug the service -- just paste the URL into your browser's address bar and see what the image looks like. Or if the data returned is in XML, and you have a referenced XML stylesheet that renders into readable HTML in the browser, then you get the benefit of semantic markup and easy visualization all in one package. But you are correct, this benefit mostly evaporates when working with more complex authentication schemes. If you can't encode all your authentication information into each HTTP request, then I would argue that it doesn't count as REST at all.

5. "Why do we need a "readable" URL for each resource? If we were using a tool to implement the service, do we really care about the actual URL?"

Well, it depends. Why do we need readable URLs for any resource on the web? You can read Tim Berners-Lee's essay Cool URIs Don't Change for the rationale, but basically, as long as the resource may still be useful in the future, the URI for that resource should stay the same.

Obviously, for transient resources (like the "today's Money" link in the essay) there is no need for it, since the need to reference the resource goes away if the corresponding resource goes away. But for more permanent resources (like StackOverflow questions, for example, or movies on IMDB), you want to have a URL that will work forever. When you're designing a web service, you need to decide if the resources themselves could outlive your service, and if so, then REST is probably the right way to go.

For the record, yes, I've been developing web pages since well before NetFlix or Twitter existed. And no, I've not yet had any need or opportunity to implement a client to either NetFlix or Twitter's services. But even if their services are atrociously difficult to work with, that doesn't mean the technology they implemented their services on top of is bad -- only that those two implementations are bad.

To make a long story short: REST and SOAP are just tools. They each have strengths and weaknesses. If the only tool you have is a hammer, then every problem looks like a nail. So get to know both tools, and learn how to use them correctly, and then choose the right tool for each job.

An honest question deserves an honest answer. But first, why did you use the text of this question as an https://stackoverflow.com/questions/76595/soap-or-rest/3322449#3322449">answer to another question if you did not think it was rhetorical in nature?

Anyway:

1. "Tools exist for all modern languages/frameworks/platforms to read in a WSDL and output proxy classes and clients. Implementing a REST service is done by hand by reading documentation."

Just like browser vendors have read and re-read the HTML 4.01 specification up and down to try to implement a consistent browsing experience. Have you reflected on the fact that browsers were invented long before internet banking and stackoverflow, and yet, you can use a browser to do just those things. This is made possible because of the sole reason that everybody agrees to use HTML (and related formats like CSS, JS, JPEG etc).

Blogging is actually not that new, and someone came up with AtomPub, which allows any blogging software to access and update posts in a blog, much like any web browser can access any web page. That's pretty neat, and works because of the RESTful constraints imposed by the protocol.

But for Twitter and Netflix, there is no universal agreement that "all microblogs in existence shall use the media type application/tweet", mainly because microblogging is so new. Maybe in a few years time a few microblogging services settle on the same API so that Twitter, Facebook, Identica and can interoperate. None of their existing APIs are anywhere near RESTful, however much they claim, so I don't expect it to happen real soon.

"Furthermore, while implementing these two services, you have to make "guesses" as to what will come back across the pipe as there is no real schema or reference document."

You've hit the nail on the head. REST is all about distributed and hypermedia, and that pretty much sums it up. A browser looks at what it gets from a request and shows it to the user. A HTML page usually spawns a lot more GET requests, for example CSS, scripts and images. An image is typically only rendered to the screen, JavaScript is executed, and so on. Each time, the browser does what it does because it found the link in an <img> or <style> tag and the response media type was image/jpeg or text/css.

If Twitter makes a hypermedia based API, it will probably always return an application/tweet every time you follow a link to a tweet, but the client should never assume it, and always check what it gets before acting on it.

2. "Why write a REST service that returns XML anyway?"

This all boils down to media types. Like HTML, if you see an element

Designing media types for generic problem domains (as in the HTML media type for the rich text problem domain) is very hard. Making media types for very narrow problem domains is probably a lot easier (like a tweet). But it's always a good idea to design for extensibility and specify how clients (and servers) are supposed to react when they see elements or data items that don't match the spec. JPEG, for example has an Application-specific record type (e.g. APP1) which is used to contain all sorts of meta data.

3. "I've heard the complaint that with SOAP you have the "overhead" of the SOAP Envelope. In this day and age, do we really need to worry about a handful of bytes?"

No, we don't. REST is absolutely not about being efficient over the wire, it's actually trading wire efficiency in. REST's efficiency comes from the possibilities of caching enabled by all the other constraints: http://www.ics.uci.edu/~fielding/pubs/dissertation/rest_arch_style.htm#sec_5_1_5">Fielding's dissertation notes: The trade-off, though, is that a uniform interface degrades efficiency, since information is transferred in a standardized form rather than one which is specific to an application's needs. The REST interface is designed to be efficient for large-grain hypermedia data transfer, optimizing for the common case of the Web, but resulting in an interface that is not optimal for other forms of architectural interaction. I don't think that the SOAP Envelope byte count overhead is a valid concern.

4. "I've heard the argument that with REST you can just pop the URL into the browser and see the data."

Yes, that's also an invalid argument. It doesn't work that way. Even if it did work, most narrow REST APIs out there use media types that browsers have no idea about and it still won't work.

But there are a lot more possibilities than a browser to test a HTTP based API, like command line utilities or browser extensions that allow you to control almost any aspect of a HTTP request, inspect response headers and discover links for you to follow. But even so, this is nowhere near as easy as generating WSDL stubs and making a three line program to call the function anyway.

5. "Why do we need a "readable" URL for each resource? If we were using a tool to implement the service, do we really care about the actual URL?"

If you look at how the web works, I'm pretty sure that humans are by and large glad that the URI for a wikipedia page looks like this, http://en.wikipedia.org/wiki/Stack_overflow instead of http://en.wikipedia.org/wiki/?oldid=376349090. But it actually is not important to REST. The important thing to try to get right is to choose to place relevant data in the URI that is not likely to change. You might think that the database ID will never change, but what happens when two data sets need to be merged? All your primary keys change. The page title (Stack_overflow) will not change.

Sorry for the long response, but I believe this question is valid, and hasn't been addressed before here on SO. I'm sure Darrel Miller will add his answer once he's back too.

Edit: formatting

Martin Fowler has a post on the Richardson Maturity Model which does a great job explaining the difference between SOAP and REST.

WSDL and other document level protocols are redundant. The HTTP protocol supports a much richer set of operations besides just serving documents and submitting forms.

Supporters of REST are uncomfortable with that redundancy.