Parse is dead. Long live Parse!

This is one of the news you do not want to hear. Never. After a couple of years, Facebook has suddenly decided to kill one of its most appreciated tool: Parse.

The understanding of the shutting hit me a few hours ago, as soon an ill-fated (and unexpected) post was published on Parse official blog.

Parse has been an amazing tool, years ahead of any other competitor, for its capability to empower developers to create backend-less app. They started in 2011 and moved on fast, releasing a client SDK supporting a wide set of languages (from Android to iOS, JS, C# just to name a few) and had a considerable adoption from the community, that perceived this technology as a possible solution to the start-up need for speed in product development.

One of the most notable capabilities of Parse was, of course, PFObject: a base class implementing Active Record pattern to allow for user creation and persistence. After some time, many other competitors (such as PaasBox) came out, but they could never reach Parse.

Parse was always one step ahead, releasing many critical improvements: in a few months we saw object relations and queries, followed by client data storage and synchronization (also known as pinning).

The adoption of Parse spreaded in 2013 when Facebook acquired the company, thus empowering the team with their skilled developers and releasing support for Push notifications, analytics and the first version of Parse Cloud console.

Then competition came: Google App engine (released in 2008, out of beta in 2012), Microsoft with Azure and finally AWS with Lambda and DynamoDB persistence. All of these were (and are) interesting solutions for serverless development but still remained behind Parse - at least regarding the support - because of its capability to manage the whole infrastructure directly from a client app, and provide a structured object to do that.

The only other technological alternative to Parse is Salesforce Platform (actually Salesesforce App Cloud) which provides through SObjects many capabilities, even though only for Enterprise B2C apps. So, Parse was unique in its own way.

Moreover, it was all but a dying project ...

Last fall in Palo Alto, attending at an annual Meetup at Facebook, Parse team explained to developers a detailed roadmap towards platform support extensions and a new Parse portal, which was released by the end of the year, improving user experience and object editing.

A few weeks ago Parse team published an update to their iOS SDK supporting Apple Watch and AppleTV, again acclaimed by developers community as a welcome new.

Suddenly a lightning in the sky ...

Today they have announced, for still unknown reasons, that Parse will be discontinued and is going to end its life one year from now. This is an astonishingly bad announcement, because a lot of projects currently rely on Parse in production (some side news talks about 500,000 apps) and many more (such as Stripe) spent a lot of time building tools to integrate Parse with their products.

Here at Neosperience we love server-less development and embraced it more than a year ago, trying to figure out which would have been the most promising choice. After many projects using both AWS and Parse we decided to adopt the former, mainly due to Parse lack of support for multi-tiered-customers (i.e. channel partners), reduced flexibility, app creation through apis and strong vendor lock-in.

Today it is revealed to have been a wise choice and our customers can sleep soundly under Neosperience Cloud. Unfortunately this is not true for a lot of developers out there, because for yet-to-be-explained reasons Parse has been found dead. Long live Parse!