A bunch of programmers are upset with Facebook for pulling the rug out from under them

Last week, Facebook announced that on January 28, 2017, it's going to shut down its Parse service for developers, leaving many apps and the teams behind them scrambling to find a new home.

The Orbitz travel app is a Parse customer, as are Facebook's own Oculus Rift VR team and the Quip mobile word processor. So are lots of smaller developers, who relied on Parse rather than spend the cash to build the servers and software to do it themselves.

When Facebook bought Parse, it was already hosting 100,000 apps.

And so, the news of the shutdown meant that the rug was pulled out from under them.

"There goes my theory that Facebook's focus on developers and running services at scale would both improve Parse and make it stay around for a long time. Sigh. Yet another failed acquisition," said one Facebook comment.

Ali Mohsen, cofounder of Bahrain-based startup studio Level Z, told us in an email:

We still haven't decided what will be our migration plans, but definitely we are not going to use a backend as a server solution, as it may shutdown anytime like Parse did, which was a decision never expected to be taken by Facebook, as it really destroys FB relationship with developers.

What was Parse?

Parse, a Y Combinator-backed startup that came to Facebook in 2013 by way of a reported $85 million acquisition, provides a set of vital tools to help developers build and maintain slick apps for iPhone and Android phones.

Mainly, that means Parse provided a set of databases to keep track of your information across phones and operating systems, and the mechanisms for apps to send push notifications.

When Facebook bought Parse in 2013, it was facing an uncertain future and the aftershocks of a rocky IPO, and was looking to diversify its business look for the next big thing.

Now that Facebook has really nailed most of the challenges ahead and is growing like crazy, it doesn't really need that kind of broad scope. Parse could be a casualty of the social network's current success.

It's not a great move for Facebook, which spends a lot of time and energy reaching out to developers and getting them to integrate stuff like Facebook Login with their apps. Every app that integrates some kind of Facebook functionality also increases the chances that they'll see a Facebook ad, after all.

Parse isn't leaving customers totally high and dry, though. It's making available a whole set of free tools so you can host your own version of Parse on your own servers — a route that may be appealing to developers who are now skeptical of letting any tech company host any portion of their apps, since they could be shut down any day.

"It looks like they have some documentation on how to migrate, but I'm not totally sure what I'm going to do yet..!," says digital artist Jono Brandel, who currently hosts his app Typatone with Parse.

But many, if not most, of Parse's jilted customers turned to the service in the first place because they didn't want to run those services in-house, usually because they lack time, cash, or expertise.

Plus, using Parse on the backend and Facebook login on the front end had some additional benefits. It meant that the Parse database that contains user information automatically filled in if they logged in with Facebook. That doesn't happen with competitors.

Alternatives abound

Fortunately for developers, there are plenty of other services out there more than willing to pick up Parse's slack.

Earlier on Monday, Microsoft released a how-to guide for moving apps from Parse to the Microsoft Azure cloud, using its set of similar services. Google has Firebase, a popular Parse competitor that it acquired late last year. Amazon has its own mobile-app services. Apple is slowly but surely building its CloudKit for iPhone developers. And there are plenty of other startups providing alternatives, besides.

Judging from the email in my inbox and the buzz on Twitter, Google's Firebase looks to gain — or regain, from lapsed customers — the most ground in the market as a solid alternative, especially since, like Parse, it's free to use.

"The Firebase solution seems logical at this point until [Apple's] CloudKit cloud solution is more matured," Alvin Lawson, developer at fitness startup Sweat Society, told us in an email.

Still, it's resulted in a lot of existential angst for startups, who now have to choose between building their own infrastructure to handle the needs of their mobile apps, or going with another provider who could pull the rug out from under them at any moment.

And at least developers have a year to figure out their next move, since Parse is giving a nice, long window before it's "retired."

"I mean as far as things getting sunset ... Parse looks like it's taking a very diligent approach," Brandel says.

Read More »