The Minuum keyboard project: mobile keyboards reimagined | Hacker News Reader

The Minuum keyboard project: mobile keyboards reimagined

indiegogo.com

198 pointsneonkiwi86 comments

My (irrational) fear is that if this becomes a modest success, there will be an aqui-hiring with much talk of it appearing in some upcoming edition of Blackberry or Windows or whatever, and then nothing will be heard from them again.

I think I need a little more sunshine, I must be low on Vitamin D.

Once something like this becomes somewhat popular, even if it's killed by acquisition, there will be clones popping up everywhere. I have Swype on my N9 (once considered state-of-the art input innovation) and I remember anticipating it being on every smartphone. That didn't happen, but now you can find dozens of clones in the Android market.

That's how technology moves. Ground-breaking technology becomes a commodity, often laying the foundation for bigger innovations to take place.

"...but now you can find dozens of clones in the Android market."

Android 4.2.2 comes with a swipe keyboard by default. I have it on my Nexus 7--I don't use it all the time, but it's useful when I am tired or I want to type longer sentences.

I really enjoyed Swype early on, but I never figured out what their plan for it was - it seemed to be in perpetual beta, when I'd have happily been paying for it as an app. The weird installer shenanigans and insistence on internet access finally killed it for me, well before them getting acquired.

It's sad in a way - I also thought it'd be on every phone, but it looks now like that's more likely to happen via the same functionality being built into other keyboards rather than Swype doing it themselves.

Unless there are patents involved.

I seem to remember some one-handed keyboard projects (hardware and software) were killed because of patent encumbrance.

(http://blog.monstuff.com/archives/000021.html)

Vaguely reminds me of Dave Mackay's (sadly now neglected) Dasher research project[1], in that both can be used to enter text with input devices that only have a single axis.

Dasher's seems like it fits better with continuous/'analogue' input methods, though (it uses inherently continuous input, Minuum's just using the analogue entry systems to simulate a single-row keyboard).

[1] home: http://www.inference.phy.cam.ac.uk/dasher/

gif showing the concept: http://www.inference.phy.cam.ac.uk/dasher/images/largedasher...

play store: https://play.google.com/store/apps/details?id=dasher.android

Dasher is really cool, but it suffers from two major problems:

1. It's inherently slow. It just takes a long time to get from one letter to the next, even when you're following along the expected path.

2. It's exceptionally distracting to see alternate expected words pop up along your way as you type, and I find it tends to derail my line of thought.

Dasher is really cool - it's easily the most radical take on keyboards I've seen.

Sometimes I just turn it on full speed and swipe at a random angle to the right and let it fly through the infinite dictionary, watching a deterministic gibberish story unfold.

I love dasher. It was very helpful for me in writing emails when I was unable to use one hand. One handed keyboards were actually slower once I got used to both of them. It just doesn't work well for coding, but it's great for other English.

Video: Loud music, lots of bass, dude is a quiet talker. I couldn't make out everything he was saying.

Also: Touch typing on the full size iPad is extremely pleasant, and not bad on the mini. Screen real estate is quickly reclaimed when you quit typing. It's nice to have the same text input interface between your laptop and tablet.

Video: silly beard, silly sweater, and typewriter. I couldn't focus on anything else.

Video: crazy awesomeness starting at 1:30 blew my mind so hard I passed out, and due to unconsciousness did not hear the CTA at 2:10.

You are making me laugh. As Minuum's marketing guy, though, it hurts my feelings that you missed the "CTA." --Mark Thompson

They did not mix the audio correctly. So many people miss out on mixing and end up with all audio fighting to be heard. If you're not going to invest in a sound engineer to mix your audio, at least learn about using a simple compressor with side chain fed from the voice channel; it'll let people hear what you're trying to say even with loud music in the background.

I really appreciate this comment.... I always wondered why it's so hard to make music quiet behind voice tracks. But the words you used don't make any sense to me and they're not that easy to Google. I did poke around a bit though, and I think what you're talking about is referred to as "ducking"?

For those who want to apply this technique, it's built into GarageBand: http://www.youtube.com/watch?v=ulfKubMdxPA

Yeah, dude (Will Walmsley, CEO and lead designer) mixed it around dawn. Next time we'll hire a sound person. --Mark Thompson

> Touch typing on the full size iPad is extremely pleasant.

I don't find it especially pleasant. The keyboard takes up more than half of the screen, and typing quickly on it is much more difficult than any regular keyboard (though much better than any phone keyboard naturally). I hooked up on a Bluetooth keyboard to it once and felt an incredible relief at the effortlessness compared to on-screen typing. There's much that can be improved.

I personally hate auto-correction, false corrections are highly annoying.

I use Messagease on my phone. It took me around a week to get to 30 WPM. With two fingers it's possible to have around 50 WPM.

If you are looking for alternative keyboards, I recommend to try it:

https://play.google.com/store/apps/details?id=com.exideas.me...

The current default auto-correct for the Samsung Galaxy S2 is highly annoying as it actually omits some of the alternatives and you can't switch between languages anymore.

It just pisses me off immensely.

Interesting, because I kinda like the keyboard on the SGS3. It can do touch-typing, markov-chain word proposals, swiping, text recognition input (I don't use that one personally, but when I tried it was surprisingly decent).

Actually, from all the Samsung stuff on the device this one is by far the best. The rest is meh or useless or downright pointless, yet I like the keyboard.

Only gripe is that it uses the same dictionary for all languages. Which can be useful, at times.

Huh, MessageEase looks interesting. Wish I'd known about it.

But I just bought a new Motorola Photon Q with slide-out physical keyboard. I love it already. It seems much faster than Swype or SwiftKey (subjectively, at least; I haven't attempted speed measurements).

Hmm decided to try MeassagEase out. It's a neat little thing. It's the first touch keyboard that I would consider punching code in with. Most tablet keyboards require too many button presses to get to the symbols and lack the deliberate nature this seems to have. Would be nice if it had some predictive text for when I'm just chatting but oh well.

I tried it, too. It works great. Check the add-ons available, one appears to be an English word prediction pack.

Looks interesting, but the music is mixed waaaay too high. I have to really strain to hear his voice over the background music.

That does concern one when it's a video regarding a user interface project, doesn't it? :)

I think the ideas quickly presented near the end of the video (~1:30 onwards) are the most exciting part of this project. Right now, text entry using a video game controller isn't exactly pleasant, and there are no shortage of use cases for technologies like Google Glass where voice recognition wouldn't be appropriate. This area is ripe for innovation; good luck Minuum!

Valve has a nice solution for that: http://williamleeks.com/blog/the-lotus-keyboard/

Alternate keyboards seem to die quickly.

Frogpad was a one handed keyboard.

The Twiddler was another which is probably a good fit for Google Glass. (http://youtu.be/gS8wGePlknM?t=1m19s) (http://www.youtube.com/watch?v=PzgJR8uAQvc) -- but note that Twiddler2 was about $200.

You forgot the AlphaGrip[1]!

[1] http://www.alphagrips.com/

To summarize without having to watch 2 minutes of video: squash three rows of querty keyboard onto one row and auto-correct.

I love the concept, especially if it degrades gracefully when one needs extra precision (ie, when typing things like names that are guaranteed to fail autocomplete). Looking forward to trying it out.

This is also yet another reason for bailing out of the iOS ecosystem. Hope this finds its way to Ubuntu Mobile! :)

My problem with this (as a non-native english speaker) is that such solutions often only support latin based languages and cyrillic is completely left out. This will probably make it impossible to write, as auto-complete will not work well.

Otherwise, it looks interesting, but I am a bit skeptical. good luck to the dev team

Most non English languages don't work with these. My main language speaker count is 1 miljon. I am not going to get auto-correction for this anywhere soon(assuming it uses it). And my language happens to have some special letters, so without changing keyboard, you can't just switch vocabulary(we bypass it by using numbers, but this makes text non computer readable).

I'm on my phone, so I can't watch the video.

My problem with keyboards that rely too heavily on autocomplete is that it makes it too hard to use the web. Email addresses, domain names, and passwords aren't going to be in the autocomplete dictionary.

This doesn't seem to solve any of those issues, except to also provide granular access to keys (just in a much more cramped setting).

Would http://www.youtube.com/watch?v=lzYCZm-ces0 be of any help to you? I'm not familiar with how iPhones open various URLs, but I can open that in the YouTube app on my Android.

The dictionaries on non-English languages are also usually quite abysmal.

If the autocorrect is good this could actually be pretty neat. I found that because you stuck with QWERTY I could "eye type" on it pretty easily, unlike some other concept keyboards.

This should really be implemented as a feature in a full-size keyboard. Like a button with a down arrow that causes the keyboard to shrink into "one dimension" when you want the screen space. Or if 1D must be the default, it should at least be able to expand to full size when a lot of typing is necessary.

This project reminds me of Vernor Vinge's "Rainbows End" and ensemble coding.

I personally consider this book a must read for every aspiring entrepreneur in software and hardware.

Is there a version of the video without the music? I can barely hear what he's saying.

Incidentally, I really love the soundtrack. Does anyone know if this is just stock music from somewhere?

Indeed, the audio is much louder than the voice.

You guys should build it if you think it is worth building, why do you need backing for software keyboard ?!

I am not backing this and it is getting ridiculous. And video, there should be Bootstrap for Kickstarter videos, a template you can just add to.

I would be happy to try this out when it comes out. I expect it will not work for me as I am bilingual and those keyboards only work well for english. But good luck, don't spend all the money at one place.

why do you need backing for software keyboard ?!

It takes time to make software. Backing gives you the money to dedicate time.

i can imagine it takes time to build something like this, but there are a lot of people building cool stuff without the money (open source is all about it).

In my opinion there is no reason why they couldn't start with a first beta for android (payed app) and interrate from there.

If more people like it, they'll buy it.

Kickstarter/etc. is just marketing, i hate the development of this "strategy" for software.

Again, building software takes time and money. Much of open source has been "subsidized" by people either working on getting a degree, or by their employer when they build and release something that solves a problem for the company.

Kickstarters and other similar crowdfunding platforms are a great way to gauge interest in a concept before starting to build it. If thousands of people are willing to pay for it before it exists, then even more are likely to pay for the finished application.

The video says they have been working on Minuum for two years. What have they been doing? They clearly have usable prototypes, as seen in the video.

I just skimmed through the project description, but it seems like they're hiring a developer and don't code it themselves. If you're hiring somebody to build this with great attention to details, I can imagine those 100k being quite reasonable. Consider internationalisation as well (perhaps not for Chinese, but to different languages with Latin writing).

If you think that software is something you build in your spare time between your full-time job, feeding your kids, and sleeping, then by all means maintain this opinion. I think the point of a financing campaign is so that the developer has the freedom to take this on as their full time job. Seems reasonable to me.

I have 3 kids, I have projects I don't have enough time to develop, I think that is normal.

Honestly, and I really wish you all best, but I think this is abuse of what kickstarter and indiegogo were for.

Sorry, but that's just ridiculous. Not only do KickStarter and IndieGogo have countless other software-only projects, they have many other projects with few costs other than the creators' time, like digital comic books and music.

If you feel that's not worthy of backing, fine, but I'd hardly call it "abuse".

Not to be unfair, but my point at least was that it isn't okay. This is what a large chunk of us do for a living. What difference does it make by whom or by what mechanism our work is compensated?

Interesting approach, at first glance I was not really convinced, but seeing the possible applications in wearable electronics, this might be the ideal solution.

There is a layout on Greek phone keyboards (not smartphones) in which every letter possibility is divided in 3 letter segments of the alphabet (i.e. one segment is ABC (or ΑΒΓ) another DEF (ΔΕΖ) and so on and so forth). Normally, you press the key as many times as the letter you wish to write (i.e. one click for A, two for B, three for C) or in this different layout you click once for a letter that exists in the word in the series that it is shown then another series in which the next letter exists etc. The phone then creates a word out of those possible letters and spits out the closest result, with no fail.

For example, you want to write the word 'and', you press the sequence 'ABC', then 'MNO' and lastly 'DEF'. Then it creates the word 'and'.

I haven't found something like that in American phones, albeit I use an iPhone now. It's more intuitive than it sounds and it makes texting a lot faster.

If anybody can bring more input about this happening in English phone keyboards is welcome to do so. (It also might have a name that I am not aware of)

This feature was often referred to as "T9."

More info: http://en.wikipedia.org/wiki/T9_(predictive_text)

This sounds like http://en.wikipedia.org/wiki/T9_(predictive_text) which was quite popular in the US before smartphones took over.

On Android if you enable the numeric keypad as input method (instead of the standard keyboard), it will behave like you describe -- a variant of T9.

I'm not convinced for smartphones, but the tail end of the video shows some real promise. The watch and game controller in particular.

It is interesting that they are using QWERTY. Im curious if thats a decision based on trying to gain acceptance. It seems that if you're looking to really reinvent user input, theres likely a more ideal layout.

I think the choice of QWERTY is purely for acceptance reasons. Everybody can air-type on an imaginary keyboard, using a new layout would be a massive barrier to entry.

For the part with typing for your watch and glasses, I think MYO + ASETNIOP, enabling typing on any surface (such as discreetly on your thighs) with zero interface, will be much more practical.

I'm excited about MYO too, and thanks for bringing ASETNIOP to my attention.

But I think "practical" is the wrong word for the advantages of both of those technologies (separately and also in combination). "Powerful" seems more to-the-point.

Myo requires specialized hardware. ASETNIOP requires specialized training. Minuum doesn't require the former (beyond an accelerometer), and almost none of the latter.

I think Minuum would combine well with Myo.

Links for the curious+lazy:

https://getmyo.com/

http://asetniop.com/

ASETNIOP's got a learning curve, but a MUCH higher ceiling. I'm seeing 37 wpm quoted as the top speed for Minuum, whereas here's a demo of an average speed of 65 wpm sustained over three minutes: http://www.youtube.com/watch?v=nygdYinzpmk

They should sell this on Cydia to replace the iOS default keyboard.

It looks pretty interesting but I do not really have a problem with the default keyboard.

Did swype ever get acquired? this looks interesting, and could be a useful acquisition target for samsung/apple/google

Swype was acquired by Nuance. Nuance has quite the portfolio for input entry whether it be by speech or typing.

Reinventing text entry/editing is something I've always wanted to do/see done. Wish there were more solutions that weren't based off of QWERTY, though. I realize it's been the standard since the typewriter, but if we really want to see huge gains in speed and efficiency, we should shift away from it.

I was always surprised that Apple did not acquire Swype. It is a fantastic input method, the kind of "magic" people expect from Apple. Now that the Google has implemented gesture typing for Android's stock keyboard, I assume Swype must not have patents strong enough to block imitators. Maybe that's why Swype was not acquired by Apple or Google.

Looks like it will be painful to get used to, but we said that about the QWERTY keyboard and look where we are today.

I don't even think so. It's just QWERTY without Y axis, after all. I guess that you'll be accustomed to it after a couple of minutes of typing.

What's hard is adapting to entirely different layouts such as Dvorak (which, by the way, could probably increase autocorrect accuracy greatly).

This is likely true for those of us who have already become accustomed to typing using QWERTY, but there are many who are just learning or have yet to learn to type, most notably children.

If at some point an input method that is faster and more efficient than QWERTY gains real traction and is taught or exposed to our youth, I'd expect a shift away from what's become convention.

This is by no means something that will just happen overnight, but I'd be a little saddened if in the last years of my life, QWERTY remains as the dominant method of text input. It's certainly a good solution, but no where near an optimal one.

The "gaining real traction" part is exactly the issue. Of course, we'd all love our governments to force everyone to learn Dvorak or something similar, but usually, no authority worries about our keyboard layouts, and people just use what others use.

If you're interested in alternative keyboards, take a look at GKOS. It's a thumb based chorded keyboard.

http://gkos.com/gkos/index-gkos-com.html

I think the keys can be grouped in way that you rarely will need to press the same key twice in a row. The typing speed could be increased and would be less words options for each typed sequence. It's just a though...

Sure, but that will require users to get used to a different layout. Dvorak (http://en.wikipedia.org/wiki/Dvorak_Simplified_Keyboard) does this. It isn't very popular though, as it's few people want to spend time relearning how to touch type.

Comparing touch typing to touch keyboards is quite a stretch, pardon the pun.

How is that? Somebody who doesn't know how to touch type is usually having a pretty hard time locating the keys on a QWERTY keyboard. Tactile typing (I just made that name up) doesn't require exact knowledge of the key placement as you need it when touch typing, but you still need some kind of muscle memory if you want to achieve a reasonable speed.

This has some potentially amazing possibilities for accessible devices. I hope they get some people with various limbs and digits missing for testing.

(The music drowns out the voice! That's a shame.)

Tests will show how correct it will be. May I suggest to also try swype with this?

I like swype, and the new swiftkey flow (basically swype + swiftkey), but even with 2D querty it's frustrating how often either of those misinterprets what I'm trying to enter.

If it got still less info -- particularly as this is being developed with far fewer developer resources than Swype and SwiftKey have at their disposal... -- I don't have much hope for it.

I want this in Ubuntu phone.

I think the most interesting use of this would be typing via eye-tracking.

Off topic: This was filmed in the Bahen building at U of T! Go DGP!

Ugh, just give me back my candybar qwerty already!

Yet, it's still in QWERTY.

That's the point. Users don't need to learn a new keyboard layout.

If they could jimmy something up for the Windows phone, I'd be very happy. Its a great idea and aesthetically it looks cool.

>Fact

>People have difficulty typing on mobile devices. (Your experience may vary.)

Uh.

Reply on news.ycombinator.com