We're definitely considering that, but we don't want to be just another enterprise LDAP vendor. What would be the unique selling point, in that case? Elasticity?
The Platform as a Service angle is very interesting, and we're planning to release an addon for the major platforms such as Heroku, GAE, OpenShift, and CloudStack. CloudStack is already setting a precedent with their new LDAP authentication functionality. We'll see if it catches on.
Thanks Steve, appreciate your feedback. You're right, data and system availability are primary concerns, and ldap.io is designed with those factors in mind.
In the event of networking or connectivity problems -- for example, if your Internet link goes down -- we're considering providing an on-premise virtual appliance configured to automatically replicate with the hosted servers. It would be very low maintenance and a good option for BCP.
For the authentication issue in general, consider the current trend of Identity as a Service and Single Sign-on as a Service -- Okta, OneLogin, and Ping, to name a few. If you use ldap.io for authentication, it would be in the same category of services, and you would also have more control of the data layer.
Funny, I wrote a plugin earlier this year and came up with the exact same name. I never had time to polish and release it, but if you're interested, here's the source on GitHub:
Weasel words may be used to detract from an uncomfortable fact, such as the act of firing staff. By replacing "firing staff" with "headcount reduction", one may soften meaning."
Interesting choice of words: "sunset" instead of, say, "shut down". It's clearly a weasel word, and using a noun as a verb makes it all the more jarring.
At the moment, ldap.io is intended as more of a general purpose infrastructure service, rather than strictly identity or authentication.