dougv.com « Doug Vanderweide

Tumblr Mangles Developer Relations

Last week I logged on to Tumblr and was confronted with this abomination:

Missing e notice from tumblr. Way to encourage API development, guys.

Needless to say, this is pretty disturbing, and I wonder what Tumblr is thinking by posting this.

Background

Some background: Tumblr is a blogging site, with social media overtones. Basically, you can easily follow other bloggers’ posts through a dashboard / search posts via tags, and it’s quite easy to repost material you find on other blogs.

Like most other major providers, Tumblr maintains an API. Until last year, it was mostly restricted to retrieving and submitting posts; it was recently expanded to allow some manipulation of blog settings and managing followers.

I like Tumblr a lot. I’ll go on a couple of times a day, and like most other Tumblr blogs, my blog is mostly reposts; it’s where I’ll dump links / reposts of things I see on the Web that I want to share.

Missing e has been around for a while. It’s a browser add-on for Webkit-enabled browsers; as its name implies, it leverages the API with some neat features that aren’t directly available through Tumblr itself.

For example, Tumblr has a lot of image posts. Missing e includes a magnifier feature that lets one see a full-sized image right from his dashboard, rather than having to engage in the several clicks it takes to see a full-sized image. Missing e also lets me more easily reblog items (including the automatic addition of tags to reblogged posts), manage my post queue, and otherwise make Tumblr easier to use.

I should note that I don’t know the developer of this plugin personally, nor have I spoken to him about this notice. (I have read his response to this outrage, however, and I find it remarkably calm, fair and responsible.) I don’t know if Tumblr has contacted him about its concerns or tried working with him on those issues (reading the developer’s responses, it sure sounds like they haven’t).

I also haven’t contacted Tumblr about this. I’m not interested in hearing whatever nonsense they intend to proffer as justification. I know what I read and I know how I feel about it as an API developer.

To Tumblr’s credit, they haven’t cut off API access to the plugin, which was certainly an option others might have pursued. It wouldn’t surprise me if a number of Tumblr users can’t tell where Tumblr ends and missing e begins, and thus they are swamped with support requests they can’t do much about. And it does make sense to me that missing e uses a lot of resources to accomplish its tasks.

A Completely Wrong-Headed Approach

That’s where my empathy for Tumblr’s plight ends.

First, Tumblr’s reliability, both in terms of its primary service and its API uptime, rivals Twitter for embarrassingly inadequate. (At least Twitter has the common sense to not blame third-party developers for their failure to stay up.)

That’s on Tumblr alone. It’s up to them to keep their service running.

I especially find odious the insinuations contained in this notice. While missing e is, in the base definition, a “hack” of Tumblr, the tone of this message suggests that the plugin isn’t well-written and may be up to no good.

Well, you can go to GitHub and look at the code yourself. Yes, it sends data to intermediary servers. Yes, it is technically possible for missing e to steal a user’s Tumblr credentials, to track Tumblr users’ activities, to obtain personally identifiable information, etc.

Let me be clear: I agree with another user that missing e in no way compromises user information right now. However, it could do so, by virtue of being a browser add-on; to that extent, the notice Tumblr posted is accurate, as they don’t directly accuse missing e of privacy violations, but do note it is possible for browser plugins to capture information a user never anticipated having captured.

Absent proof that there is an intention behind missing e to do that specifically, and to use such information for nefarious purposes — evidence Tumblr clearly could provide, if it existed — I find the tone of this note beyond insulting; it’s chilling.

My interpretation of this notice is, “We don’t like missing e. We’d just as soon ban it. But that’s not very Web 2.0 and it’s likely to generate PR static. So we’ll scare you, push you toward getting rid of it, then continue to serve those who want to use it.”

That’s being a dick. That’s being lazy. That’s being stupid.

A Proper Response

Were I in charge at Tumblr, we’d be going about this in an entirely different way.

• The first thing we would have done is offered the guy who wrote missing e a job.
• If not that, we would have offered to buy missing e outright.
• And if that didn’t pan out, we’d ask missing e users to rate its features, then build those into our platform.

Because what does Tumblr’s approach to this issue say? It says, “We aren’t interested in the reasons why missing e is a problem for us. We don’t care about our end users and why so many of them are using this plugin. It’s not that our product is inferior, and someone has made it better; it’s that we have what we have, and even though it can clearly be better, we’re more interested in the status quo.”

Or, as I’ll coin it, The MySpace Response: “Do what you like, so long it fits in our picture of our service.”

You saw how well that worked out for them.

I don’t suspect this blog post will cause any change whatsoever in Tumblr’s approach. I simply want to lament what is an absurd and insulting response to a relatively minor problem by a company that I expected knew better than that.

If Google brought you here because you’re worried about that notice, suffice it to say that I looked at the missing e code and, as of this writing, I see nothing there to be concerned about.

I do, however, see a lot to be concerned about in Tumblr’s handling of this matter.

All links in this post on delicious: http://delicious.com/dougvdotcom/tumblr-mangles-developer-relations