Tumblr has always been a great place to put the things you’ve made, but we also want it to be the place you go to make great things.
Recently we released our own gif maker tool to arm the masses with an easy way to turn the treasure trove of videos and bursts on their camera roll into something perfectly fit for Tumblr.
A few things that were really important for our GIF maker:
It should be baked into the Tumblr experience and the existing post flow (not a separate app!).
It shouldn’t be limited to using the in-app camera in that particular moment—you already have great material on your camera roll.
It’s fast and it just works. Never worry about hitting the file size limit.
But we also wanted to give people the opportunity to tweak their creations, if they wanted to. So we added in some additional controls. You can trim or crop your GIF. You can speed it up or slow it down. You can loop it, rebound it, or reverse it.
In the original designs, each of these things were broken out into separate steps (and the image resolution was manually controlled). Take a look:
10,000 years ago, before the advent of written language, the earliest app developers relied entirely on symbols and animation to convey a message to users. Of course, we still rely on non-lingual cues as shorthand for specific actions and to help people navigate the app, but in the last few hundred years we have also started to include copy.
Copy makes it possible to communicate complex concepts, but in its ubiquity it can get overlooked. Not so in the Tumblr app. Like those early app developers who tracked antelope migration patterns on cave walls and first used magnifying glasses to represent search, we take a very personal approach to copy. In this post we want to share a few places we’ve tried to lift Tumblr out of the drab standardization of many modern day apps.
Error messages
When we’re rewriting error messages we try to keep two things in mind: Make the error as comprehensible as possible and, if possible, suggest a solution. Here’s a long-retired string from our Android app:
<string name=“photo_capture_exception”>Unable to open camera to capture photo.</string>
This error occurs not because of anything the app did, but because—incredibly—on Android, you can delete your camera. But the user wouldn’t know that from the error. They’d just think the app sucks, give it a bad rating, and probably stop using Tumblr all together. That’s why it now reads:
Believe it or not, you don’t have a camera app installed.
The new version tells the user 1) why they can’t take a picture (no camera app), and 2) how to resolve it (go get one).
Conversations on Tumblr aren’t like conversations anywhere else. You can’t just leave a comment on someone else’s post. You have to bring that post, wholesale, over to your own blog, before adding your piece to the bottom of it. Good reblog commentary, in any context, sets off a whole new cycle of reblogging, sending a whole new set of conversations rolling around Tumblr, all of them picking up new comments as they go. It’s like a Katamari.
Transformative commentary in a reblog can completely change the meaning of a post. Old posts are rediscovered, already-weird posts get transcendentally weird.
But there were oddities with the way this commentary was presented.
These blocks of content get smaller as the conversation grows, which in turn increases the distance between the names of the users and whatever they said. It eventually makes the whole thing tricky to figure out—especially in the mobile app, where most people use Tumblr. If you enjoy someone’s commentary midway down a deep reblog tree, you probably aren’t going to follow that thin gray line back to the top of the post to find out who said it. Worse, deep conversations were being pushed completely off the post, effectively shutting them down.
There are charms to that chaos, but ultimately it can make Tumblr hard to approach for new people, and leaves posts vulnerable to being consumed by their own popularity.
So, how do you bring clarity to the reading experience while maintaining the constructive sensibility of reblog conversations? Turns out, a solution to this “who said what?” problem has existed since the creation of online chat. Simply listing the comments one after another was the first way it was done on the Plato Talkomatic, the grandfather of online chat, and it remains the clearest and easiest to read.
Initially we tried preserving the indentation of the original design by insetting the stack in a frame.
Trouble was, it meant that we were losing a tremendous amount of space for people to play with in the context of the reblog. Particularly on small screens, where space is already at a premium, we didn’t want to cut into space for creative responses with images, text, or literally whatever.
At the same time, we played around with including avatars next to commentary. Not only did it further clarify who said what, but it gave people a better chance of being recognized for adding something thoughtful or funny to a conversation.
In the end, we combined our favorite parts of the two of these, hoisting the avatar and username up above the commentary. It allowed us to include a little more interesting detail about the who-said-what, and give people as much space as possible to say anything they like. It also actually reduced the height of very long chains.
Ultimately, creative discourse is one of the things we love most about Tumblr. It’s a big part of the reason we’re all here. The mechanic that fuels that—posts being literally passed from blog to blog—isn’t going anywhere.
With the latest update, we wanted to open up discourse more: simply removing visual barriers allowed interesting, weird, and funny conversations from dying out before they even get rolling.
A great post from the Ministry of Design about reblogs. Yes, Virginia, staff does think about this stuff. Deeply. Or derply, depending on your point of view.