4 ms·
What does BrowserID do that every existing OpenID provider cannot? The BrowserID FAQ says the architecture is "decentralized", but login relies on Mozilla's bro
by cpeterso 15y ago
What does BrowserID do that every existing OpenID provider cannot? The BrowserID FAQ says the architecture is "decentralized", but login relies on Mozilla's browserid.org server.
- jmathai 15y agoThe goal is that others support the verification portion (hence it being decentralized). The browserid.org site is required today because others haven't implemented support on their backends. Hope that changes sooner than later.
- decentralized 15y agoOn the linked page it says: "NOTE: You may choose to validate assertions on your own server. While a bit more complicated you can reduce your dependencies on others. Refer to the specification and the source for the reference validator." So while it's easier to use their server, it appears it is entirely possible to implement without it. Perhaps in a few weeks, we will see some good implementations. As I understand, the main advantage over OpenID is that instead of needing a <username>@<someopenidprovider> login, people can simply use an email address as their login. If all email providers supported OpenID, there would be little difference, but that has not happened (yet).
- resnamen 15y agoIt provides a canonical identifier for your account without compromising your identity to BrowserID-enabled services. The OpenID spec divulges who you are.
- nodata 15y agoHow does it compare to Microsoft Passport? How does it compare to client-side certificates?
- wmf 15y agoIt's decentralized and the UX is usable, respectively.
- icebraining 15y agoThe OpenID spec divulges who you are. How so? As far as I know, the only mandatory information is the OpenID identifier (the URL you need to input to authenticate). While more information can be exchanged, that's completely optional and up to the user. For example, MyOpenID always asks what "persona" - if any - you want to send when you authenticate. BrowserID is no better: you need to input your email address.
- laconian 15y agoThe OpenID identifier can act as a relational key across services.
- icebraining 15y agoSo can the email address in BrowserID, or the username in the typical user/pass combination. If you don't want to be connected, use different IDs for different services. Nothing in OpenID prevents a person from having multiple identifiers; in fact, Gmail generates different identifiers for each domain you authenticate on: http://blog.stackoverflow.com/2009/04/googles-openids-are-unique-per-domain/ http://blog.stackoverflow.com/2009/04/googles-openids-are-un...
- thristian 15y agoMozilla's browserid.org server is used for two things, currently: - validating the signed authentication token you get back from calling the authentication function. As user "decentralized" points out, this is just a convenience; your service can do this itself if you don't mind getting your hands dirty with crypto code. - Serving the JS shim for user-agents that don't implement navigator.id.getVerifiedEmail() natively. Once browsers implement this function themselves, and webmail systems can tell the browser that the user owns a particular email address, this won't be needed either.
- ozten 15y agoThe design of BrowserID is more aligned with real user behavior in the wild. An example is users remember an email instead of a URL. The API/Protocol for an implementing website is much easier, based on implementer feedback.