我目前正在研究OAuth 2.0和OpenID连接规范。以下是我的理解: 之前他们是:

OpenID was proprietary implementation of Google allowing third party applications like for newspaper websites you can login using google and comment on an article and so on other usecases. So essentially, no password sharing to newspaper website. Let me put up a definition here, this approach in enterprise approach is called Federation. In Federation, You have a server where you authenticate and authorize (called IDP, Identity Provider) and generally the keeper of User credentials. the client application where you have business is called SP or Service Provider. If we go back to same newspaper website example then newspaper website is SP here and Google is IDP. In enterprise this problem was earlier solved using SAML. that time XML used to rule the software industry. So from webservices to configuration, everything used to go to XML so we have SAML, a complete Federation protocol OAuth: OAuth saw it's emergence as an standard looking at all these kind of proprietary approaches and so we had OAuth 1.o as standard but addressing only authorization. Not many people noticed but it kind of started picking up. Then we had OAuth 2.0 in 2012. CTOs, Architects really started paying attention as world is moving towards Cloud computing and with computing devices moving towards mobile and other such devices. OAuth kind of looked upon as solving major problem where software customers might give IDP Service to one company and have many services from different vendors like salesforce, SAP, etc. So integration here really looks like federation scenario bit one big problem, using SAML is costly so let's explore OAuth 2.o. Ohh, missed one important point that during this time, Google sensed that OAuth actually doesn't address Authentication, how will IDP give user data to SP (which is actually wonderfully addressed in SAML) and with other loose ends like: a. OAuth 2.o doesn't clearly say, how client registration will happen b. it doesn't mention anything about the interaction between SP (Resource Server) and client application (like Analytics Server providing data is Resource Server and application displaying that data is Client)

从技术上讲，这里已经给出了很好的答案，我想到了给出简要的进化观点

OpenId -仅用于身份验证。

OAuth—用于身份验证和授权。授权依赖于access_token，它是JWT令牌的一部分。它可以包含用户权限的详细信息或任何有用的信息。

两者都可以依赖第三方认证提供商来维护他们的帐户。例如，OKTA身份提供者，User在OKTA登录页面上提供凭据，在成功登录时，用户被重定向到消费者应用程序，头部有JWT令牌。

现在OpenID连接是最相关的，所以我将解释OpenID连接和OAuth 2之间的区别。

OpenID connect指定IDToken标准:https://openid.net/specs/openid-connect-core-1_0.html#IDToken

这是OpenID连接的主要贡献。因此，它指定了身份验证完成后响应中应该包含的内容。

IDToken需要是JWT令牌，并包含用户的信息，如用户id、用户名等。返回的信息取决于授权时传递的请求。它还包含令牌的过期日期，并且应该包含令牌的数字签名。此签名用于使用公钥验证令牌。

第二大差异与公钥有关。OpenID连接使用所谓的发现或众所周知的端点。它是一个公开开放的端点，只返回一个带有公钥和授权端点等值的JSON。

https://openid.net/specs/openid-connect-core-1_0.html#SelfIssuedDiscovery

因此OpenID本质上是与身份验证相关的，因为它指定了IDToken，这是通过检查数字签名和IDToken的过期日期来验证用户身份所必需的。

OAuth处理授权，特别是与作用域和验证资源服务器上的访问令牌相关的授权。

但是，正如这里所写的，OpenID使用OAuth 2授权进行身份验证。

https://openid.net/specs/openid-connect-core-1_0.html#AuthRequest

身份验证请求是OAuth 2.0授权请求，它请求授权服务器对最终用户进行身份验证。

简而言之，尝试将OpenID视为使用JWT令牌的身份验证，将OAuth视为具有作用域的授权。

OpenID是关于身份验证的。证明你是谁)，OAuth是关于授权(即。授予对功能/数据等的访问权。而不必处理原始的身份验证)。

OAuth可以在外部合作伙伴站点中使用，允许访问受保护的数据，而无需重新对用户进行身份验证。

博客文章“从用户的角度看OpenID与OAuth”从用户的角度对两者进行了简单的比较，而“OAuth-OpenID:如果你认为它们是同一件事，你就找错了对象”有更多的信息。

更多的是对问题的延伸而不是答案，但它可能会为上面伟大的技术答案增加一些视角。我是一个在很多领域都很有经验的程序员，但是在网页编程方面完全是个新手。现在尝试使用Zend框架构建一个基于web的应用程序。

Definitely will implement an application-specific basic username/password authentication interface, but recognize that for a growing number of users the thought of yet another username and password is a deterrent. While not exactly social networking, I know that a very large percentage of the application's potential users already have facebook or twitter accounts. The application doesn't really want or need to access information about the user's account from those sites, it just wants to offer the convenience of not requiring the user to set up new account credentials if they don't want to. From a functionality point of view, that would seem a poster child for OpenID. But it seems that neither facebook nor twitter are OpenID providers as such, though they do support OAuth authentication to access their user's data.

在我读过的所有关于这两者及其区别的文章中，直到我看到上面Karl Anderson的观察，“OAuth可以用于身份验证，这可以被认为是一种无操作授权”，我才看到任何明确的确认OAuth足以满足我想要做的事情。

In fact, when I went to post this "answer", not being a member at the time, I looked long and hard at the bottom of this page at the options for identifying myself. The option for using an OpenID login or obtaining one if I didn't have one, but nothing about twitter or facebook, seemed to suggest that OAuth wasn't adequate for the job. But then I opened another window and looked for the general signup process for stackoverflow - and lo and behold there's a slew of 3rd-party authentication options including facebook and twitter. In the end I decided to use my google id (which is an OpenID) for exactly the reason that I didn't want to grant stackoverflow access to my friends list and anything else facebook likes to share about its users - but at least it's a proof point that OAuth is adequate for the use I had in mind.

It would really be great if someone could either post info or pointers to info about supporting this kind of multiple 3rd-part authorization setup, and how you deal with users that revoke authorization or lose access to their 3rd party site. I also get the impression that my username here identifies a unique stackoverflow account that I could access with basic authentication if I wanted to set it up, and also access this same account through other 3rd-party authenticators (e.g. so that I would be considered logged in to stackoverflow if I was logged in to any of google, facebook, or twitter...). Since this site is doing it, somebody here probably has some pretty good insight on the subject. :-)

很抱歉这篇文章写了这么长时间，而且更多的是一个问题而不是一个答案——但是Karl的评论似乎是在OAuth和OpenID上大量的帖子中最合适的地方。如果我没有找到更好的地方，我提前道歉，我确实试过了。

我想谈谈这个问题的一个特定方面，如以下评论所述:

OAuth:在授予某些特性的访问权限之前，必须进行身份验证，对吗?所以OAuth =什么OpenId +授予访问某些功能?- Hassan Makarov 6月21日1:57

是的……也没有。答案很微妙，所以请耐心听我说。

当OAuth流将您重定向到目标服务(即OAuth提供者)时，您很可能需要在将令牌交还给客户机应用程序/服务之前使用该服务进行身份验证。然后，生成的令牌允许客户端应用程序代表给定用户发出请求。

注意最后一句话的一般性:具体来说，我写的是“代表给定用户”，而不是“代表您”。一个常见的错误是假设“拥有与给定用户拥有的资源交互的能力”意味着“您和目标资源的所有者是同一人”。

不要犯这样的错误。

虽然您确实使用OAuth提供者进行身份验证(例如，通过用户名和密码，或者SSL客户端证书或其他方式)，但客户端获得的回报不应该被视为身份证明。例如，在一个流中，对另一个用户的资源的访问被委托给您(通过代理，OAuth客户端)。授权并不意味着身份验证。

要处理身份验证，您可能需要研究OpenID Connect，它本质上是OAuth 2.0设置的基础之上的另一层。以下是关于OpenID Connect(在我看来)最突出的一点(来自https://oauth.net/articles/authentication/):)

OpenID Connect is an open standard published in early 2014 that defines an interoperable way to use OAuth 2.0 to perform user authentication. In essence, it is a widely published recipe for chocolate fudge that has been tried and tested by a wide number and variety of experts. Instead of building a different protocol to each potential identity provider, an application can speak one protocol to as many providers as they want to work with. Since it's an open standard, OpenID Connect can be implemented by anyone without restriction or intellectual property concerns. OpenID Connect is built directly on OAuth 2.0 and in most cases is deployed right along with (or on top of) an OAuth infrastructure. OpenID Connect also uses the JSON Object Signing And Encryption (JOSE) suite of specifications for carrying signed and encrypted information around in different places. In fact, an OAuth 2.0 deployment with JOSE capabilities is already a long way to defining a fully compliant OpenID Connect system, and the delta between the two is relatively small. But that delta makes a big difference, and OpenID Connect manages to avoid many of the pitfalls discussed above by adding several key components to the OAuth base: [...]

The document then goes on to describe (among other things) token IDs and a UserInfo endpoint. The former provides a set of claims (who you are, when the token was issued, etc, and possibly a signature to verify the authenticity of the token via a published public key without having to ask the upstream service), and the latter provides a means of e.g. asking for the user's first/last name, email, and similar bits of info, all in a standardized way (as opposed to the ad-hoc extensions to OAuth that people used before OpenID Connect standardized things).