I Uninstalled Fleksy Today, and Want to Tell you Why

Since its release, I’ve been using Fleksy, an app for iOS.

Long, long ago, when I used a Nokia device with a number pad, over time I became very fast with text input, thanks to T9 predictive texting. The more I used my iPhone, the faster I became entering text with the virtual keyboard, but I never approached the speeds I did with the Nokia devices.

Fleksy made a massive difference. The genius of Fleksy is that you type without giving it much thought. You type roughly where you think you need to, and Fleksy’s prediction algorithm does a good job of working out what it is you intend. Sometimes it may present a series of words, but it’s easy to flick through those, then move onto the next word you wish to type.

Initially, Fleksy was targeted exclusively at the blind market, and had to be purchased from the iOS App Store. Over time, the business model has changed. The app is now free, an Android version is available, and the company is targeting a much wider, mainstream market.

I applaud this.

As I’ve blogged previously here, there are many examples of technologies originally designed for the blind that have moved onto the mainstream, including the LP record and the desktop scanner. Given the socio-economic status of our community, a free app allows the technology to get into the hands of more people. I was pleased to have been able to buy the app initially, help its development along in some small way through the purchase, and ultimately see it freely available.

Fleksy’s innovation continues. An SDK means the keyboard is being integrated in third-party apps now. This is a work-around to Apple’s sandbox approach which prevents an installed app from providing functions system-wide, unless they are doing something exposed by an API provided by Apple.

Last week, Fleksy dropped a bombshell. They announced they would have two versions in the store. One version would be targeted specifically at VoiceOver users, and the other would be for the mainstream market. Blind people using the built-in VoiceOver screen reader for iOS were encouraged to delete the current Fleksy app they had been using since its inception, and download this new one, based on an older version of the code.

My initial reaction was that what counts is that we have the best access we can to the tools we need. We are a market with unique requirements, and if it takes a specific app to meet those requirements, so be it.

My view was not by any means shared universally. A number of people with credentials in the accessibility sector took to Twitter to voice their disapproval, claiming that with this decision, Fleksy was segregating us.

When the topic was raised initially, I felt considerable sympathy for Fleksy, believing them to be victims of a massive over-reaction.

Yesterday, Fleksy released an update to the original Fleksy app, which is now the app designed for mainstream users. It was then that the consequences of this decision really became apparent. The app we once used is inaccessible.

In response to the push-back, Fleksy has said that the blindness-specific app is only a temporary work-around, and they pointed out that blind people require some tweaking of the prediction algorithm because we use it a bit differently from the way it is used by a sighted person.

Having had time to consider the arguments from others, I now believe my initial decision to shrug my shoulders and not worry about this decision was the wrong one.

At present, I’m giving a lot of thought to how we encourage iOS app developers to ensure their apps are accessible. Having thought it through more carefully, I’m now deeply troubled by the signal this approach sends to the developer community. Let’s remember, Fleksy are brilliant at marketing their product. I read a lot of tech publications, and they are everywhere.

A lot of people know Fleksy’s out there. It’s sending totally the wrong signal to app developers when they read that Fleksy has a special VoiceOver-friendly version of their app. It’s counterproductive because it reinforces the notion that accessibility is hard, that considerable lengths have to be gone to, that hoops have to be jumped through, in order to make an app VoiceOver friendly. I can’t condone this approach and I don’t believe any of us in the blind community should condone it either. The people are great, the app is great, but this is the wrong decision and it needs to be reversed as a matter of the highest priority.

Consequently, I’ve uninstalled Fleksy, and it will remain off my devices unless or until they return to one app, universally accessible to all. If they intend getting around to it sometime, then may I respectfully suggest they drop whatever else it is they are doing and do it now.

If they need a setting in the app to invoke specific behaviours, fine. If they need to detect the presence of VoiceOver and give us a different experience, OK. Just don’t tell the world at large that blind people need an app separate from everyone else. It does our accessibility cause no good at all, and indeed, it potentially does it active harm.

I congratulate Fleksy on all they’ve achieved, but this is a colossal mistake, and there are times when as consumers, we should respond when a boundary has been overstepped. This is one such time.


  1. Jay

    I agree 100 percent, especially because there is freely available documentation for implementing VO. If that wasn’t the case, I would cut them some slack. Look at the quick response from some game developers when contacted. Some are more than happy to assist and with in weeks if not days, a game is greatly improved if not completely playable. There is no excuse here. If the developers were at least honest with us from the get go, perhaps the backlash would be less.

  2. Lachlan

    Hi. I too have upgraded to the latest version of Fleksy in hopes they would address a few bugs I’ve been having, namely the space and backspace commands stop working at random, this has been happening over the past few days. I wish I hadn’t upgraded now, I agree that we don’t need a different version of an app, there should just be one version of an app that works for everyone, VoiceOver users or not. I too hope that Fleksy reverts back to one app for all. The way Fleksy works changes a bit when VoiceOver is turned on, even in the older version. When VoiceOver is off, you see keys for thinks like shift, space, delete and enter, those keys disappear when VoiceOver is switched on, but that’s just fine with me, it doesn’t bother me. I hope they go back to the previous approach. I’ll keep using Fleksy VO because I absolutely hate typing on the standard default keyboard with VoiceOver, it’s way too slow for me and it’s frustrating, so I need Fleksy very badly, but I don’t like there new approach, I’ll probably try sending them feedback on the issue.

  3. mitchell

    I did not now about that app an y wood thay do sucha thing and can you work that app

  4. Pamela Francis

    I wholeheartedly agree. I thought I heard somewhere that this company was using us as a blind community as a test market to see if in fact their app could work and would work. I too uninstalled Fleksy. as an app that was revered for inclusion within general technology as it is progressing, then to stab us in the back as a blind community, is totally wrong and regressive. This move will encourage other developers to give us mediocre versions of their app because they figure, we as a blind community will settle for whatever we can get. how are we supposed to believe another developer in the future who may genuinely hold our interests at heart after the move that was made with this app? We do not need a different prediction engine, our language is no different because we can’t see. That statement was a direct response to the interview with the Ceretec team. Within that interview it was clearly stated we as a blind community need a different prediction engine then our cited peers. G A R B A G E!!! your issues with your SDK, should not have been made public to the extent that it separated the app from the community it was originally supposedly, designed for..
    You accepted the accolades given to you by tech companies along with organizations and websites that were geared for and designed for the blind community. your clarification statement on your blog stating the VO app was only temporary, was enough for me along with others with any sense to realize that we were being fed a line of crap.
    Had you taken the time to include everyone in your SDK to begin with, you would not have needed that clarification. We as a sector of the community, are no one’s guinea pig. Until and unless general accessibility is being included in to one app , I will have no more to do with you, your company, your app or anything that may succeed it. I see you as a traitor to us as a community who helped get you off the ground.

  5. Robert Jenkins

    This blog was just spot on. There really isn’t much to say. It is rather simple to understand.

    What we want is a reversal of what they have done. That is that we want just one app. One app that is universally accepted and works for all.

    Sure, I understand. They have worked hard on this project. They have done everything they could to make this work for the VoiceOver. I am really appreciative of that. What you fail to understand is why we feel this way. It may not be your intent to do this but it did happen.

    Sure there may have been a lot of technical problems and such. However, I really don’t want to hear what happened. That is not important to me. What happened has already happened.

    You say this is temporary. Well, then. Here is my question. What is temporary?

    I think for us to be able to move on and to disuse this problem is the answer. Saying anything other than it is in our strategy that we will have this incorporated in the future. To me that just means you don’t know. It is for all I know some mystical Unicorn timeframe. Plus, it tells us you currently have this at your lowest priority. Otherwise, if it wasn’t in the low priority. You would have some idea of when this could happen. Say 3 or 4 months perhaps? Don’t need to know the exact date but we would be more comfortable if we know a ballpark idea and that this will happen. Not if we can but rather we will make this happen.

    There is no need for any interviews and lots of questions and answers. There just need to be one question and one answer. Perhaps, along with some things about that you are sorry. It was never your intention to hurt us like this. We understand how we would feel this way and that it does look bad on us in what we have done. We will now take this our highest priority to fix this. You can expect that this will be one universal app that you all had enjoyed. Please forgive us and we can expect this to be released within blah blah months.

    Simple and to the point.

  6. Cat

    while I agree what they did was wrong, I don’t think I am going to remove the app from either my phone or my touch. why? because it helps me type. Yes, I know I should practice, and I have, however I just can’t type as fast as I can with Fleksy. I have been trying for two or more years, and touch typing just never came to me. I will, however send feedback because they do need to hear our voices and I do agree what they did was a bad move. they could at least say sorry to us and explain why they did what they did, and tell us when the apps will be one again. I haven’t seen any of this talked about, and I know they are on apple vis, and they post to twitter.

  7. Ken Downey

    I just want to know how our predictive text needs are different from those of sighted people. I’d like an example or three.
    I upgraded to iOS7 recently, and the only reason I did it was because Fleksy dropped support for iOS6. I was happy with iOS6 and had no intention of upgrading until I had to choose between 6 and Fleksy, and Fleksy won out. As far as productivity, I lose a lot more time trying to navigate the web with the terrible bugginess of iOS7 than I gain using Fleksy, but that’s not their fault. I guess that what I’m getting at is this overeager desire to make the SDK and have it work for everything. First they left iOS6 users behind, then they left us behind. Okay, so we have a VO version, but why do we even need a version that is so different? With VO turned off, editional keys are visible, like shift and delete, but we can work around those if we need to. Also, as I said earlier, I don’t believe our predictive needs are that much different than those of our sighted counterparts.
    Finally, though I don’t feel that it’s as dramatic as being stabbed in the back, I can understand why people might feel that way. I know a few people who were upset when the app went free after they paid $15 for it. They said, even back then, that the company was using blind people to get started. Well, even if that’s true, at least someone was smart enough to acknowlege our existance, something most companies don’t even bother doing. I do hope they drop the VO version and reintegrate us, and soon. I also hope that they create a keyboard for jailbroken devices, so we can use it in all our favorite apps. That would be sweet!
    Regardless of what they do, I’ll still use their product as long as it’s accessible. It’s way faster than touch typing, more accurate than dictation, and, of course, the vO version is still accessible. They could have just left us in the dust, but they didn’t.

  8. DigitalDarragh

    I said this over a week ago. Instead of writing on Twitter, I responded to the blog post on their website. However, my comment was moderated and never published.

    The moment they released a second app, I knew that this wasn’t temporary. It takes too much work to fork development streams. You don’t do it without giving it serious thought and serious resources.

    The day the released the Fleksy app on Android without Talkback support I knew something was going very wrong.

    I’m annoyed. I payed for this app because of their commitment to voice over over users at the time. Now that they are getting more interest from other users, I feel that they’ve dumped their commitment to one side until it suits them again.

  9. Kevan

    In response to an earlier comment, I don’t think Fleksy will ever make a keyboard for jailbroken devices… They’ve told me before that Apple disapproves of such behavior if an app is going to be in the Appstore it must follow their strict and annoying rules.
    I’d be pretty excited though if a hacker was to make a keyboard with Fleksy’s algorithm to use on Jailbroken devices, and put it in Cydia! I’d jailbreak my phone in a heartbeat and probably wouldn’t mind paying a small price for it LOL.

  10. pvagner

    I agree with Daragh. They are promising accessibility for their android port for about a year now without doing something usefull at all. So I have uninstalled it from my android a while ago and I will uninstall from the iphone as well.

  11. Wayne Merritt

    While I don’t agree with your decision or the decision of Fleksy to put out a separate app, I do respect your decision. Separate is not equal, but it is better than the company denying the issue or saying “deal with it, “like many other larger companies would. Example, consider the basic view in gmail and how that is never updated by Google, but the standard view is updated quite often. I haven’t stopped using Gmail though. Besides like another person said, I can type faster and better with Fleksy VO.

  12. Dean

    Time for a different opinion, even more different than Wayne’s.

    I want to analyze the question of how to respond to the Fleksy change from two angles, difference and effectiveness.

    I am different. Besides all sorts of other differences, I am blind, hence I use different tools and take different approaches. those differences are not going away, and I’m not concerned that they do.

    Clearly, some differences need to be minimized or eliminated, as they can keep people from performing necessary life activities. Using a different version of Fleksy than my sighted neighbor does not any of these effects. After all, I already use my iPhone differently, with the screen dark and using different gestures. No neighbor or prospective employer is going to notice if I use Fleksy-VO or don’t.

    There are good reasons to urge developers to build in accessibility from the ground up, but if this company wants, temporarily or permanently, to maintain two apps, I am not concerned about this as long as I have equal functionality, which I apparently now do. I will have different things to say in the quite possible event that they implement some useful functionality in the main app and don’t give it to me. Until then, the company has taken the needs of the blind into consideration, and has in addition gone to some trouble to explain their philosophy and actions. So this is a difference that does not concern me.

    Let’s look at the effectiveness angle. I am a big believer in using the best tool for the job. this is not always a politically correct or popular viewpoint. For instance, I make heavy use of a stand-alone braille device, (I think of it as a tablet, but it falls under the label of notetaker.) Some would call it a “blind ghetto” device and be critical, but for what I do, it is the best tool. Maybe I’ll amend that assessment if text processing and text searching on a mobile device is ever equal to what I have, but I don’t see that happening any time soon.

    Because I do use that somewhat maligned standalone braille device with my iPhone, I find that it, along with dictation, pretty much eliminates the need for me to use Fleksy. But if I needed to use it, if it were the best tool in the box, I would use the current free VO version. If the Fleksy people were to be linked to human trafficking operations, or child labor, or Naziism, I would have to rethink this attitude, but thus far nothing of the like has come to light.

    Let’s look at effectiveness from a second standpoint. If I uninstall Fleksy, what exactly does that accomplish? If I have heretofore found it to be the best tool I can use for writing, I am hurting myself and nobody else, since the uninstallation will be completely unnoticed. And if I set out to discourage others from using this free app, I may be doing them a disservice if it is, as many have found, the best tool out there.

    A final thought on effectiveness: We know we have to pick our battles. So is this really the battle to pick? I mean, are we spoiled or what? In your technological life, is there not at least one product that would really help you, perhaps enabling you to do something you cannot now do? If there is, and obviously there is, why are we fretting about Fleksy? I suggest it is because we know how to and can, not that it is the best use of our advocacy time and energy.

    • Jacques

      Good points Dean, the most sensible arguments I’ve read on this thread thus far.

  13. David Csercsics

    I removed Fleksy as well. I’m not that slow at the standard keyboard and the only reason for installing fleksy is that siri is terrible at texting for me. But I don’t like that they’re handing older code to voiceover users that is just silly. I don’t think they need different algorithms for word prediction for voiceover either. How many ways can you tap a keyboard with 1 finger?
    Thanks for sharing this with us.

  14. John Greer

    Let’s turn the table for a moment. I have asked this question of some people, but I will ask it again. How many blind product companies make products for the sighted as well as the blind? The answer I most often get back is, what sighted person would want to buy blind products? The other answer I get back is they can’t make products look good enough to appeal to a sighted consumer. The other answer I get is, they don’t know how. All answers in my opinion are wrong, yet blind product companies continue to only make products for the blind. If it is important to blind people that sighted companies make products for both blind and sighted, why can’t blind companies do the same? Do they not try? Do they not hire a sighted visual designer to make a product that looks better than a huge black brick? My aproach to advocacy is that if we want the world to change for us, we have to learn to be the example we want to set. Throwing away a product that works for you, be it considered ghetto or not is not the answer, nor is throwing away something simply because they provided a blind version along-side a product that is not blind friendly. I will grant that programmers are not the best at PR, but they did explain their intent. Let me translate. They split their app into two versions because they said they were going to make changes to it that would continue to break Voice Over support as they added features. So they came to the decision to split the app into a Voice Over version and a non Voice Over version so that we had a stable free version that would continue to work as they tested things with the other version. Keep this in mind, they didn’t have to do this, they are not beholden to only cater to blind people, yet they still did. The next thing they mentioned is they wanted to make an SDK of their prediction engine. Keep in mind that the prediction engine is not the entire app. It is the engine that figures out what a person may be typing as they do so. It is similar to the T9 prediction engine people like Jonathn and I used back in the Nokia days. I have also heard complaints that apps that have used the Flecksi SDK are not accessible. That is not the fault of the Flecksi prediction engine SDK, that is the fault of the other app developers using the SDK and not using Apple’s accessibility SDK. To be clearer, there are two different software development kits, Flecksi’s SDK and Apple’s accessibility SDK. They are not the same and both have to be used for a keyboard to work like Flecksi’s alternative keyboard. Now, the other issue, through partnering, they did find a way to make the app free for us. They also did not have to do that, but they found a way, while still having a business plan in mind of selling their SDK. That is certainly more effort for making things affordable than I’ve seen from a few blind only product companies. They have also found a way to offer both a blind product and a sighted product, which is more effort than I’ve seen from any blind only product company. Think about what I have said here, and if it is a tool that still works for you, use it, and we’ll see whether they make it into one app again some day, but keep in mind, again, they are not beholden to market to only blind people. LOL but what do I know.

  15. John Greer

    Oh, one more small point to make. A previous post about the ghetto product basically came to the conclusion to use what works for you. However, this post seems to suggest that making a different Voice Over app makes it segregated into a ghetto product whether it still works for you or not. An odd bit of double speak.

  16. BackDaws

    I’ve just uninstalled it but not for the same reasons but because it’s gone really shit

Comments are closed.