MarsEdit’s history with Tumblr has been long and bittersweet. Starting with a December, 2008 post in which I encouraged Tumblr to add fundamental support for apps like MarsEdit to their API:
There are relatively simple changes that Tumblr can make to their API to facilitate an acceptable user experience not only from MarsEdit, but from other client applications that might choose to embrace the service’s API.
It feels great to finally have an official release out there that supports Tumblr. I would like to thank Marco Arment for his patience in working with me to improve the Tumblr API, so that it will work better with MarsEdit.
Although I was satisfied with the initial changes made to the API, there were areas that still needed work. The most significant shortcoming in Tumblr’s API remains to this day: lack of support for uploading and embedding images independently from “photo” style posts. This means that MarsEdit users who write traditional “text” blog entries on Tumblr cannot add images to those posts, even though the same functionality is available in the Tumblr web editor.
Over the years, other issues have interfered with MarsEdit’s integration with Tumblr, including reliability problems, unexpected api outages and overhauls, among other surprises. In short, it’s been a bumpy road.
In recent years, things have leveled out. Stability seems to have improved after Yahoo! acquired the company in 2013, and the Tumblr API has, for better and for worse, appeared to remain unchanged for many years now. Unfortunately, this lack of change also means that the decade-old lack of support for uploading images has not improved.
The latest chapter in this saga began last year when Verizon acquired Yahoo’s major products, including Tumblr. They established a new company called Oath, which is an umbrella for many sites including Tumblr and, until a few days ago, Flickr. The sale of Flickr makes me wonder where Tumblr fits into Verizon’s overall vision for Oath’s brands.
Most of Oath’s services, including major tech press outlets such as TechCrunch, revolve around providing consumers with content, as opposed to hosting user-contributed content. Do they want out of the blog-hosting business, the way they evidently wanted out of the photo-hosting business? Or did they merely see Flickr as being redundant with Tumblr?
Amidst the uncertainty about Tumblr’s future, the Tumblr Engineering team at least seems to be very optimistic. In an April 10 post titled “A Big New Beautiful Future for the Web at Tumblr“, they write about significant plans to renovate their whole approach to developing the website:
It was time to step back, survey the world of web technology, and clean house in a big way. That we could finally test some of the new tech we’ve been itching to use was just a little bonus.
What’s most interesting, and most encouraging, to me is the list of priorities they cite in rethinking their approach. At the top of the list:
A web client codebase fully separated from the PHP codebase that gets its data from the API in the same way our mobile apps do.
In case it’s not obvious, this is very encouraging for developers of 3rd party apps such as MarsEdit. Why? Because encoded in the ambition to implement their web client using the “API in the same way our mobile apps do” is an implicit promise (though not an Oath™) that they will be modifying the existing API to supply all the same functionality that is currently available on the web. In short: anything they can do on Tumblr.com should soon be doable by apps such as MarsEdit.
This would be a very welcome change, and I’ll be eager to see how it unfolds. It’s not a sure thing, though. WordPress has made a similar shift over the past few years with the development of their WP-REST API. Theoretically, they intend to reimplement the WordPress web interface entirely in terms of that API. What I didn’t anticipate though was that the API would not have a viable authentication model for 3rd party clients. Currently the WP-REST API can only be used on standard WordPress installations with “cookie authentication,” meaning by other web software running in the same browser. It’s effectively unavailable to apps like MarsEdit, and even to WordPress own mobile app.
Tumblr’s ambitions to re-orient development around the API may come with similar gotchas. I know enough after ten years supporting the service to know that I can’t count on anything until I see it. Still, this news leaves me more cautiously optimistic about Tumblr than I have been in years! If Tumblr sticks around, and if they do expand their API to support more functionality in MarsEdit and other apps, I’ll be eager to update the app to support it. Let’s hope for the best!