I teach tech design history, and one of the key stories I cover is the development of the original iPhone keyboard by Ken Kocienda. Reading about it in his book "Creative Selection" is great, but I wanted my students (and now you!) to actually feel this step in the process.
So, I built a web simulator of the "Blob Keyboard", Kocienda's very first attempt at a touchscreen keyboard that actually works, from September 2005:
Try the Blob Keyboard: https://juliendorra.github.io/blob-keyboard-simulator/blob-k...
- Tap for the middle letter
- Swipe left or right for the side letters
More on the github repo: https://github.com/juliendorra/blob-keyboard-simulator
The Blob Keyboard prototype emerged during a UX crisis for iPhone team (their software keyboard just didn't work at all, fingers being too big, and the Newton failure loomed over them), highlighting how innovation is rarely a straight path. It was developed on a tethered touchscreen display codenamed "Wallaby".
To make this simulator as authentic as possible, I referenced images from Kocienda's book and even got direct feedback and guidance from Ken Kocienda himself on Bluesky.
What to expect (or… what not to expect):
This is a reconstruction of a very early prototype with limitations reflecting that specific moment. The goal was to test first if typing with accuracy was even possible, as all the rest was moot if it failed!
It's NOT QWERTY: They were still hoping to get us out of QWERTY, but then familiarity won.
No Backspace: You can't delete.
No Cursor Movement: The text field is just a simple display.
No Caps or Numbers: Only lowercase letters.
No Smooth Animations: Keys just "pop" instantly when pressed. Kocienda noted that your eye fills in the gaps, giving a sense of movement.
Best Experience:
While it works with a mouse/trackpad on desktop, it's designed for touchscreens to better replicate the original Wallaby hardware interaction. Use it on your phone!
This project aims to provide a tangible glimpse into a turning point moment in iPhone development and the iterative nature of design. It's like stepping back in time and trying out that early demo on Kocienda's desk.
I would love to hear your reactions and thoughts on experiencing this piece of UI history! What other significant prototype do you wish you could experience?
At the time you basically had two choices for touchscreens: resistive or capacitive. Resistive was "obviously" the way to go because it was far more accurate. Choosing capacitive was inspired -- when used with stubby fingers the accuracy problem was moot, and it allowed multi-touch.
Just before the iPhone came out I was fairly confident I knew what the future was. It was now possible to create a phone with the horsepower to run a real web browsers. 800x600 pixel screens were available which would display normal web pages nicely, and a resistive touch screen with a stylus would make them useful.
Then the iPhone came out. 320x480 screen meant normal web pages wouldn't display properly, inaccurate touchscreen meant tap targets needed to be increased massively. Why would anybody buy an iPhone which didn't allow you to install apps, and the web was unusable because it required rewriting every page since existing pages were unusable. Instead you could buy a phone which allowed you to install apps and which allowed you to usably access the web. Obviously the iPhone would be a failure. :)
So really the capacitive screen drove the interactions. Input first, just like the mouse on Macintosh or the stylus on Newton, everything then flows from there.
On the web browser, I disagree with you (sorry!), the killer app of the iPhone was that Safari was the same Safari, with the same capacity and rendering, than on desktop.
It was completely new. Yes, you had to double tap on complex, non responsive websites, but every single (non-flash) site would render the same.
My 640x480 HTC Universal with a plastic keyboard felt antiquated compared to the 320x480 iPhone, especially starting with the 3GS
There was lots of speculation about this starting in 2002 when “Inkwell” handwriting recognition showed up in Mac OS X Jagwire:
https://en.wikipedia.org/wiki/Inkwell_(Macintosh)
https://www.macworld.com/article/155597/wacom.html
https://www.oreilly.com/library/view/mac-os-x/0596004605/ch0...
And related patent filings go as far back as 2000: https://patents.google.com/patent/US7564995B1/en
I suppose GP’s point is that the vast majority of websites in 2008 were “desktop”-only.
Also tapping links was very easy
Opera Mobile existed at that time, though.
As the iPhone OS was Mac OS at its core, Safari was exactly the same engines, and that was a quite novel and enticing promise: the real web in your pocket (minus proprietary plugins like Flash)
The web experience and usability of the original windows phone and blackberries was terrible. Nokia had a Swiss Army knife that didn’t sell or translate well to the iPhone/Android future.
I’m not sure what you mean by that. The whole point of the demo was it rendered the New York Times website just as it rendered in desktop Safari. In contrast to, say, Windows Mobile which would butcher the rendering. If the touch targets were too small, well, through the wonders of multitouch, pinch to zoom in.
the web was unusable because it required rewriting every page since existing pages were unusable
You’re remembering a much different iPhone than I am. Are you sure not confusing it with WinMo 5/6? Because then we’d be in some agreement.
But but on any mainstream phones. The phone I reference didn't exist. It could have and should have. If it did, perhaps Nokia would have survived.
From a technical perspective, yes. You are basically talking about the Nokia N800 but with a cellphone modem and a bit of effort spent shrinking the bezels down.
But from a product design perspective, I suspect it was impossible to make that leap. We are talking about the point when cellphones were at their very smallest. The 1st gen iphone with it's 3.5" display was considered to be large for a phone. Nobody thought mainstream users would be happy pocketing a phone with a "massive" 4.13" display.
And Nokia were only happy excluding the keyboard from the N800 because it was considered to be a content consumption device. At that time, smartphones were regarded as productivity devices (for email) and the physical keyboard was essential, which would have bulked out the device (See N810).
I don't think we could have gotten to today's large smartphones without first creating a viable browsing experience on an iphone sized display.
> Nobody thought mainstream users would be happy pocketing a phone with a "massive" 4.13" display.
Yet it was exceedingly obvious there was a very profitable sizeable niche of users that were willing to do so.
And it shouldn't have taken very much imagination to realize that "web in the pocket" was useful in 2008, and would quickly become much more useful in 2009, 2010 etc as the population of people with the web in their pocket grew and companies started to serve the market.
The big problem was that all of those phone companies were hardware companies. Putting Firefox in a phone was a challenge beyond them. Microsoft could have and should have done it, but they were dysfunctional at the time.
I guess phones getting new features via updates was fairly uncommon at the time though.
It's pretty intuitive because Japanese kana is a syllabary that's organized by their starting consonant and one of five possible vowels in the Gojuon system: https://en.wikipedia.org/wiki/Goj%C5%ABon
I never got used to it but people who use it swear by it. Google even made a mechanical version for an April Fools a few years back: https://youtu.be/5LI1PysAlkU
I'm native English but native Japanese are super fast. It's like watching a speed cuber
I really like their idea to make the touch keyboard work well by increasing the (invisible) "padding" area for the key most likely to be typed next. So obvious in hindsight but demos like your's show part of the journey.
There are rumors that in the beginning they tried the ipod wheel as the user input interface.
While working on my device, that idea is super motivating. I know where i am headed and have done a lot of work and really got something very, very interesting already. But a few parts are yet not clear at all. But, i am definitely further than our "clickwheel" stage. Not yet at the "secret padding" stage for some input ideas, but still.
In case you are free and interested to loosely talk about ideas or feedback you have for a very novel device (and interface), please let me know how to reach out. Just because you made this demo, i feel we'd get a long & i can learn a lot from you. And this might be a very interesting challenge for someone like you. Anyway, great work!
Feel free to contact me on LinkedIn! I'd be happy to chat
[1] https://github.com/Julow/Unexpected-Keyboard
Was hoping the LLM boom would help us get sane autocorrect to help bridge the gap but so far that hasn't happened either.
The web was and still is a very bad hypertext system (no bidirectional links, no versioning, no integrated index, no transclusion, etc.) and we are still paying this debt! But this was also the an incredible hypertext system, permission-less, decentralized, and that made it a success.
I feel that the iPhone keyboard was the same, it had to appears (including in screenshots) innocuous and "without quality" to win over users.
I have a few friends that use gesture-based keyboards similar to this, and I myself use gestures to type diacritics and punctuation. So this idea is still alive after almost two decades, just not mainstream.
The worst parts are: no upper case and the fact that there's no connection between what you typed and the keyboard once you submit.
They stopped the whole 15 developers iPhone OS team and asked all developers to only build keyboard prototypes until they all would have something worth of a demo. As far as I know it's unheard of in the history of OS development!
Ken Kocienda was not in charge of the keyboard at all originally, he stumbled upon one working solution, and that's really what I found interesting in this prototype: it's a step in the iterative process.
I'd be interested to try a full-featured Android keyboard like this.
This was about 15 years ago, but then the app and company disappeared. I can’t find any trace of the keyboard or company on the web today. Perhaps they were sunk by some patent issues.
So Swype probably reinvented the idea independently (as others did too I think)
Apple did patent some swipe behaviors for keyboards at the time of the first iPhone: https://patents.google.com/patent/US8059101B2/en
I didn’t know about the history of the iOS keyboard, and this was a great insight.
An ideal touchscreen text input method should embrace the strengths of the medium. Touchscreens allow swipes and gestures, so use that to your advantage! Like in the OP, yes. I myself experimented with a similar concept, except it was a 3x4 grid of square buttons that each can input one of 5 characters, one for tap and the other four for each of the directions, inspired by Japanese keyboards. I tried to use it for a day, it worked, but it felt somewhat unnatural. Maybe I haven't given myself enough time to get used to it, or maybe it was a terrible idea to begin with, or maybe the character layout wasn't optimal (I arranged them alphabetically).
Another idea I have is to take the 8pen[1] concept but add more lines, make it 6 or 8, so it's practical for Russian which has 33 letters. Haven't tested this one yet.
I feel like touchscreen text input is a very under-researched area for a world where nearly everyone owns a smartphone.
[1] https://www.youtube.com/watch?v=q3OuCR0EpGo
Most tech people really don't understand how valuable "not learning shit" is to a product's success
And people somehow learned to type on phone keypads before. No one complained about it once they figured it out in the first hours after getting their first phone.
On android, on google keyboard, I can pick layouts like Azerty, Dvorak, Colemak, and ClearFlow. There's also plenty of novel replacements for conventional keyboard style hit boxes using various gestures.
These things are readily available. I've never seen them in the wild.
It's not like these things don't exist. You can switch over right now. The reason you aren't ... that's the learning thing.
That's why it's totally different to learn about an UX and actually experiment it, we can notice these things.
But, all in all, I think you would probably get used to it long term!
The QWERTY layout was just a fluke, not the super optimized layout we retcon it to be.
Even its designer later designed different layouts! But the particular typewriter it was designed for had a big commercial success, so there was marketing convergence. It's of course wholly unadapted to typing on glass!
Having texted with T9 this didn’t feel too alien. Just the accept button slowing me down.
Each key on the keyboard was a rocker switch. Push left for q. Push right for w, etc.
What do you mean by "widen the margins"?
In any case, Ken Kocienda validated this reproduction's behavior as very close to his original blob keyboard (see bluesky posts between him and me)
The goal is not to make it better than the original prototype! But to the contrary, to give you the same experience that pushed Kocienda to explore further and design the right user experience.