By contrast, SwiftKey is world class.
For now. So was Skype...
Microsoft has bought the British software company SwiftKey, the company we once described as "the UK's stealthiest startup". It's best known for its predictive software keyboard, and the deal is worth $250m, the Financial Times reports. SwiftKey was founded in August 2008 by two Cambridge graduates and was perfectly placed to …
.. until they boll*x this one up and I have to uninstall it. Bought and paid for because it addressed an irritating issue that the stock keyboard had in my browser of choice (The stock keyboard would not auto-space in a web page's dialog box.)
Time to look at detaching it from the market and disabling future updates.
"Turn off auto updates and keep it."
And save a copy of the APK file just in case.
I like ES File Explorer but it started getting bloaty, bloaty, BLOATY!!!!
Luckily I found a site with older versions and downloaded one from before it got BLOATY
Now I take more care. I don't like BLOAT!!!! it makes me ANGRY!!!!
Did you boys miss the bit about these employees joining MS Research? How did you miss the importance of Swiftkey's people's first class AI skills? Whatever you may think about MS as a commercial outfit their research division is world class in the fields they cover.
So world class that it seems to have produced virtually nothing that they have made any money from!
All I can think of off hand, is the Kinect, and Xbox in general hasn't really made any money.
We buy the best people to stifle their creativity, as we can't risk cannibalising our legacy software.
All I can say is Poor Sods! Remember what happened to Nokia.
"Xbox in general hasn't really made any money."
Xbox has made several billion profit over the massive investment costs, and they have nearly 50 million Xbox Live Gold subscribers. Surface tablets are turning over a billion a quarter and growing rapidly. Holovision looks like the dogs bollocks too...
Xbox is only profitable if they don't count the original one they lost $6 billion on. They're still in the red on Xbox overall, and with the most recent one losing to PS4 in the marketplace - and consoles in general not as big of a deal in this generation thanks to mobile gaming - Microsoft is unlikely to ever reach breakeven on the overall Xbox product line.
Microsoft intended Xbox to be their wedge into the living room that they could leverage into bigger things, but it has remained almost exclusively a gaming platform. I suppose there may be a small number who use that silly HDMI passthru feature to try to integrate it with their TV viewing, but while there was a lot of initial speculation they'd turn that into something bigger it has gone nowhere.
If they hadn't been wasting their time trying to beat Sony in the living room, they might have stood a better chance of noticing Apple and Google beating them in mobile before it was too late.
>So world class that it seems to have produced virtually nothing that they have made any money from!
Its job isn't to produce anything - it's to stop the people who might produce something working somewhere else.
Think how good MS's balance sheet would be if they had hired Linus Torvalds to work on some obscure internal OS project that never saw daylight.
>They bought the Israeli company that developed Kinect, kit and kaboodle. MS practically did nothing but limit that device's utility to a single platform.
They licensed the sensor but the SDK was all them and roundly caned PrimeSense's (since bought kit and kaboodle by Apple BTW). MS have won pretty much every prize in computer vision research over the past few years - but they do a little bit of other stuff too:
http://research.microsoft.com/apps/catalog/default.aspx?t=publications
>The software was written by Rare in the UK. Microsoft bought them, too.
No idea where you got that from, this paper was widely publicised and details the MS Research Labs work the tracking system was built on.
http://research.microsoft.com/pubs/145347/BodyPartRecognition.pdf
MS' problems are legion, but they aren't down to lack of original research and mind-boggling engineering talent.
Yes. I was was incorrect and imprecise in my previous comment here (same goes to the other commentard who pointed out the same). MS bought rights to the device, not the entire company. My rhetoric remains the same (MS acquired someone else's junk and their RnD did NOTHING), but I did say a thing that was literally wrong.
MS Research does a lot of long term basic research (basic as in not having an immediate commercial application). Their work takes years to pay off. One example of that work that is beginning to pay off is in the field of real time translation (as has been reported several times here at El Reg). I repeat, whatever you may think about MS' commercial operation; how many major companies would put as much money as they do into research that may not pay off at all, or may only pay off a decade or two down the line? The problem with any MS-thread currently is that, regardless of what the topic might be, it immediately becomes infested by large numbers of anti-Redmond hatebois whose only objective is to excrete large amounts of steaming ordure on the thread. There is plenty that MS deserve very strong criticism for but the utter incapacity of some to distinguish between issues is a royal pain in the backside for those of us who are actually interested in a worthwhile on topic discussion.
MS research seems to be a black hole these days. Things go in and nowt comes out (Apart from the horrid Metro disaster, just my opinion though).
These guys have probably made enough to retire so they may spend some time twiddling their thumbs before leaving MS, starting up anothe biz. Rinse and repeat anyone?
"How did you miss the importance of Swiftkey's people's first class AI skills?"
I don't know how I missed it, but are there any decent backgrounders out there?
To this untrained obsever, there's not that much cleverness in offering a list of dictionary (and other) words, ordered by probability of use, with the list being further subsetted each time there's more keyboard input. But I must be missing something, perhaps because I hated T9 with a vengeance. What am I missing?
""What am I missing?"
Probably any experience with the app, I'd say."
No doubt. My current style (as I say, perhaps due to my historic incompatibility with T9 on phones, and my preference for e.g. Nokia E71 - oh for a modern equivalent) is that short stuff is done in full with Android standard kbd, longer stuff is either done somewhere with a decent real keyboard, or if sufficiently urgent is done using (Android/Moto G) voice recognition and a few clean-up edits if necessary.
"Try out the free version before MS b0rk it."
I see where you're coming from, and largely agree. But I'm not sure I want to get addicted to something that MS are about to break, when I already have nearly-adequate alternatives. I've got history there, going back as far as when MS bought Nextbase (the Autoroute folks), and then Microsofted Autoroute to the extent that the reviews of Autoroute 97 were so bad that e.g. they had to be physically omitted from one hardcopy magazine (AFTER the front covers were printed, with Autoroute on there). Here's another sample review, one that did make it past the editors and lawyers (from Jon Honeyball):
http://windowsitpro.com/windows/nt-europe-01-mar-1997
"But I'm not sure I want to get addicted to something that MS are about to break, when I already have nearly-adequate alternatives."
Upvoted.
Agree with both sides of this. To be honest, SwiftKey is maybe the best of a bad bunch. It is flawed in some respects but is still WAAAAAAAAAAAAY better than T9 (which I liked when it was a thing that was remotely relevant).
@everyone else - Should point put that I moved to SwiftKey after moving to a non-Google version of Android and missed their pretty good keyboard. If you're reading this and on Google Android then this doesn't concern you. Move along.
You're missing the fact that it does do really decent prediction, putting 3 possibilities just above the keyboard. It's a lot easier to tap a couple letters then a correctly spelled 15 character word than laboriously tap it completely out on the keyboard.
There's a lot of copycat apps now, but they were the first with decent prediction, not just "trawl something semi-random from the dictionary"
It learns the words you use often, *and* in what order, and also learns new words.
"Microsoft are porting the Swype keyboard to IOS and Maldroid too..."
There is already Swype on iOS.. and they don't require you register on a web site to use their _purchased_ keyboard.
Swiftkey on iOS is "free" but requires a web registration to use.. so they can personalize your predictions or something.
Their privacy policy claims that they *definitely* don't read what you type on your device. Hmmmm...
I really wish that iOS and Android both would offer a security sandbox where an application needs to ask for TCP/IP as a permission, instead of getting it by default whether it needs it or not.
Am so not interested in this development.
"As MS' predictive keyboard on Winphone is pretty dammed good. There most be more to this than meets the eye."
The "more" to this is that winphone is basically a dead platform. Microsoft are diversifying into platforms which are still in popular use.
If they start acquiring other android apps like Nova Launcher then it might also hint that they intend to produce an Android powered Lumia at some point.
Yes, I would have taken the money and run. Licensing deals - otherwise known as 'hoping the stock Google keyboard doesn't get a bit better - and selling skins to people prepared to pay to have their keyboard look a bit different do not a sustainable business make.
Makes for a nice dividend, due to appear in our accounts in late April. The principal VC backers are among those raising more funds right now, if you want your share of the action in the next story.
It probably is the right time for the founders to cash in: there's only so long a techie wants to work on a particular project before moving on. But (Ian 55) not sustainable? Isn't that comment premised on the supposition that they stand still and lack the imagination that built a successful product in the first place? In which case, it applies throughout the tech industry: you can't stand still.
"what comes out of the mouths of da yoof of today, because to my ears it could just about sound like a language"
See also: a well known local contributor. Outpourings have apparently plausible grammar but rarely have detectable meaning (partial opposite of contributions from da yoof).
Another great app gets sucked up by the MotherShip, ruined and spat out again. I'm expecting the UI to get progressively more shiny as the functionality gets worse, whilst the MotherShip demands more personal information and insists we sign in to every damn account we have for it too work at all.