Hacker News new | past | comments | ask | show | jobs | submit login

This reminds me of how we used to type on pre-smart phone era phones.

Eg https://phonesdata.com/files/models/Nokia-E51-807.jpg

The funny thing with those little handsets was, with a bit practice you could actually get very quick at typing. Quicker even than on QWERTY keyboards on touch screen phones (at least you had tactile buttons).




Maybe it’s just nostalgia, but I also remember that typing on those keyboards was way more reliable. Even with or without T9. I probably typed a little slower than on a virtual keyboard but at a more predictable pace. On my iPhone, I totally rely on autocorrect and I lose a lot of time if it get a big word wrong.

Typing on those phones required some practice to get muscle memory but it was accessible to barely anyone that had to use it. Even people who use computer keyboard all their life but are still watching they keycaps succeeded to type SMSs with the phone in their pockets.

I’m sad that we are getting nothing that stands between the old closed feature phones and the mini computers we call smartphones.


Typically this was T9: https://en.wikipedia.org/wiki/T9_(predictive_text)

I still miss being able to compose messages in my pocket...


Speed typers did not use T9. A friend of mine used to compete in texting.


This isn't surprising given that T9 is predictive and a typing contest would probably not coincide with your typical typing habits, though I am very skeptical that typing by pressing successive numbers on a number pad is faster than T9 for general texting purposes, at least after you've trained it. Perhaps I'm underestimating the speed of which these people could press buttons in succession, just it seems like depending on the word you're pressing significantly more buttons to type the same thing.


But was that because it was bad or because it was slow on those old microprocessors? T9 would fly on a modern smartphone chip


Neither, T9 was pretty reliable and fast. That was the best method for most people.

But if your goal was to be above the average speed, the probability, even low, of false prediction by the algorithm would make you lose « a lot » of time while with no T9 your error span is scoped to one character.

Plus, for the people who typed without seeing the screen, you couldn’t take the risk to accidentally send the wrong word that would totally change the meaning of your message while an error on one character could be understood by your recipient.

But for the average user who just wanted acceptable speed, reliability and used its phone while watching the screen, T9 was the best for them.


speed of the lookup was not a limiting factor and felt instantaneous. the problem was more short words for which there were too many suggestions and a bad or non existing ranking system and missing words in the dictionary. very similar to swipe keyboards but more precise and faster


I had no idea it was a named technology, though I probably shouldn’t have been surprised. Thanks for sharing that link.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: