Comment by user_7832 | Hacker News Reader

Comment by user_7832 | Hacker News Reader

Slightly off-topic, but when Swype was initially released I was absolutely blown away by how good it was. I remember swiping in the general periphery of the correct letters, and I do not know if the algorithm smoothened out the input points, used the context of previous words or both, but the keyboard was borderline reading my mind. And trust me, I tried to be as lazy as reasonably possible while swiping to get it to predict wrong words, but it was a fight I would almost always lose, albeit very happily.

My question is: why does swiping now feel so much more inaccurate? As this is HN I'd be very happy if someone from Google/Microsoft (or hopefully even Nuance) could shed some light. Was it "the big players forced the small and innovate folks out and then bureaucratized it like every commerical software because the competition was killed"? Or was it something else? I find it hard to believe that the predictive text accuracy has increased at all, if not gone down, when comparing Swype to GBoard/Swiftkey.

PS: Swype is/was the name of the keyboard app by Nuance, I'm not referring to swipe-typing in general in the 1st paragraph.

I worked at Swype and then Nuance after the acquisition.

My experience with Google keyboard has been very good. The main difference I've noticed over the years is that phones got bigger. Swype was magical on a 3 inch screen but not nearly as good on a 5 inch. If I remember right the core would take the exact touch points less seriously on a small screen but on a big screen usage might go back and forth between very accurate and sloppy.

If I had to speculate somewhat, it's possible that newer keyboards take a very principled approach to the code for input recognition. I didn't work on the input matching part at Swype but it was often discussed with great fear, because it was a couple giant functions with tons of conditionals. It had been built up over I think a few years early on but without much documentation.

The other thing is that there was a race to the bottom on price, which later led to much smaller teams working on keyboards. There just isn't enough financial incentive to put many resources towards it, compared to something like speech recognition.

Thank you so much for your comment! (I indeed am continually impressed by HackerNews's capabilities in this regard.)

> If I remember right the core would take the exact touch points less seriously on a small screen but on a big screen usage might go back and forth between very accurate and sloppy.

Do you mean that the points taken would fluctuate between more and less accurately/precise only the screens, or was the point selection consistent but the results poor?

Re: your guess about principled approach, that's kinda funny and ironic how a poorly-written bunch of code can be better than something very precise.

> The other thing is that there was a race to the bottom on price

Slightly ignorant as a customer but iirc while Swype had a paid option it also had a free option like the competition then (and now). Would this imply that Keyboard companies need external funding/or be bought over by MS etc?

I had the same experience. It was absolute sorcery the first time I used it. I was able to type without looking and would be confident that the result was correct. Using the Google keyboard now is still pretty good, but I routinely have to correct words.

I also have two languages installed now and it will sometimes type a completely nonsensical sentence in the wrong language. I can't say for sure if this is better or worse though, I only knew one language back when I used Swype.

Same experience here. When I used Swype a decade? ago I thought it was fantastic.

Now it seems to cause me more time correcting it than it saves.

And for some damn reason, the MS version at least (I know of no better multilingual swipe keyboard for iOS than that, so any recommendations would be great) constantly chooses the word "toy" instead of "you". So much so that I put a manual autocorrect entry in to stop it, which is of course frustrating anytime I'm talking about what my dog is doing or that I have several projects of a non-serious nature in process.

Are you using the old(er) versions of the app, or did you upgrade it to whatever was the latest? If you're on Android you could likely go back to an older version.

I think GBoard supports multilingual typing on iOS as well, though I can't comment on how good it is.

Amen to "toy" for "you".i would imagine in conversation, for anyone older than 5 years old, "you" is at least 20 times more likely to be the intent than "toy". Why can't their score weighting take that into consideration

"toy" for "you" is so irritating on ios, but here's something else I get, I swipe "can" and it puts in "can't", almost every time.

I used Swype on release, then not for years, until picking it up again recently, when I found that I still loved the input method but made tons of errors. The lack of accuracy had me thinking I'd just gotten old and lost fine motor control -- I couldn't imagine that a single-function piece of software had gotten that much worse.

It's an interesting difference from other software: with almost anything else, there's a UI that helps evaluate whether the software degraded. Swype doesn't have any such contextual clue. I'd imagine this problem will pop up a lot in the future with various user-assist, AI-based tech.

> Swype doesn't have any such contextual clue.

Detecting when a user deletes a word and replaces it with a word that was previously in the options list seems like it would be a fairly strong signal.

If you saw an update and then were immediately exposed to the new behavior, maybe so. But for someone who used it for maybe a couple years, stopped and then came back years later, it's extremely difficult to realize that a quiet feature like this is new.

Contrast that with, say, any Microsoft Office product where coming back a few years later is like returning to a city you used to know and finding all the old buildings you knew razed to the ground and an entirely new street layout.

I use an old APK of Swype, because it's still the unsurpassed state of the art for swipe typing. I don't know why, either, but it's not just you.

Do the old APKs still work? Is there no server verification that they do? I remember trying to get the APK but there were quite a few hoops to jump through and something went wrong.

Did you try getting them off APKMirror? They're pretty good.

Thanks, I did, and it works as well! The only issue is... well... it's not very accurate in predicting words (at least this version). It's funding* and putting other words. What version of Swype do you use?

At the time of posting, I didn't use any version, but I use APKMirror on a regular basis. I currently have the Swype+Dragon Trial on my telephone to try it out, though.

I still use Swype as well. I'd installed it from the Play store prior to it being discontinued.

It still works fine, but hasn't had any updates in 2 years. I've also noticed that if I open the Play store link (below) in incognito mode, it says "not found", but if I'm logged into my Google account, it brings up the app's info.

https://play.google.com/store/apps/details?id=com.nuance.swy...

I've tried switching to Gboard a number of times, but its swipe typing is way less accurate. Swype also has a bunch of gestures that are very convenient, like one to toggle the current word between UPPER, lower, and Title case, if you swipe from apostrophe to s, it'll insert "'s", and it's trivially easy to add new words to its dictionary. Overall, I'm not looking forward to when I eventually have to give up using Swype.

So it's not just me! I find myself far less accurate on Swift key, or other variations, than I did many years ago on Swype - to the point I subconsciously reverted back to poking without even noticing it as the swiping became so bad. And my phones got bigger so I'm pretty sure my actual accuracy is higher...

I had installed Swift Key after a year or two of Swype and it was a marked improvement. Many years later, there were days where it wouldn't accurately handle a single sentence without error or correctly predict a single word. Updates seemed to clear the learned history and reset its accuracy. The Microsoft acquisition seemed to kill any forward movement on it.

I switched, after much anxiety, to GBoard. It feels much like SK used to feel. It isn't awesome that it's a Google keyboard, but at least it's accurate and consistent.

I kind of agree but kind of disagree.

I find GBoard to be incredibly good, more or less as good as I remember Swype being. (I initially switched because GBoard supported Hebrew and Swype didn't, not because of accuracy, but didn't find it worse at least.)

I find the native Apple swipe interface to be terrible. It doesn't even come close to being as good as GBoard. And I use it fairly often, because if the Apple keyboard comes up I'll usually try to use it for 2-3 words, before getting frustrated and switching.

I have no idea why.

That's an interesting perspective. I personally haven't found Apple's keyboard to be bad (it's a bit over-aggressive, but I don't find it incorrect very often, it just censors everything to a duck). But the multilingual thing is an important point - in my very limited testing of Hinglish (English+Hindi) Swype used to be very pro-Hindi, even when I wanted to use an English word. But I don't remember comparing it to GBoard then (not sure if it even supported Hinglish back then). Google's team has certainly put a lot of effort into local Indian languages from what I've seen (including ads if I'm not mistaken).

Oddly, I find Apple's Keyboard almost completely superior to GBoard (the only exception being the number row option). However I use it on an iPad and not a phone so that may also be a factor.

I have good experience with Apple's keyboard for accuracy, but I do get frustrated routinely that it won't let me swear.

To be clear, I have to explicitly switch language. It doesn't do anything automatically (which makes sense since it's a different alphabet).

But my hit rate with Gboard is staggeringly higher than apple keyboard. I wonder if we're using it differently, which is what is causing the difference in experience here.

> I wonder if we're using it differently, which is what is causing the difference in experience here.

Yep I think this is true. I use the Apple Keyboard only on my iPad (my phone is android), while if I had to guess you use an iPhone with a much smaller screen/keyboard. Everyone's real life mileage varies, as the saying goes.

I wonder if they've trained all their AI ~~novels~~ models ~~in~~ on people's ~~inbuilt~~ input where mistakes aren't corrected, reinforcing incorrect ~~Freddie's~~ ~~fuses~~ guesses.

All mistakes there were genuinely made.

No way they made that rookie mistake... training on your own predictions leads to a useless feedback loop that only exacerbates errors.

What about a GPT-3-based autocorrect/Swype predictor?

Yeah it would be so amazing to have a FOSS swipe typing keyboard. I still use Google KB for this with internet access limited.

I tried some options but they weren't up to scratch especially with multi language

What about whatever Bayesian or Markov chain systems they used to use. At least they worked.

Curiously, Apple’s swiping on iPhone got every word of that right. I don’t know to what degree I’ve individually trained it, though.

On a semi related note, there is some speculation out there that iOS 13 changed how the native iOS keyboard predicts words and has gotten much worse.

Wow, I thought it was just me!

I have had a similar experience, and just wrote it off as being a different device or something.

My guess would be simply that we’re all typing on phones an order of magnitude more than we were then, and we only notice when it doesn’t work.

Could also be that now our expectations are higher than they were back in the days?

Reply on news.ycombinator.com