TextBlade Update: An Inside Look at Production Troubles and New Info on Shipping Estimates

[Night Spring]

But it still seems to me that a regular keyboard on a computer or a phone or tablet would be very fast unless you just weren't used to it. After all, I type over 60 wpm or 300 characters a minute. That comes out to 5 letters a second, minimum.

You are right that once you learn the key layout well enough to touch type, it's fast, no matter which layout you use, English or kana. What slows me down when typing with a physical keyboard is converting from kana to Chinese characters. In order to select one of the many possibilities that come up, you have to go to the mouse or the arrow keys. Typing on the touch screen on iOS, the selections are right there, at the top of the keyboard, or in the case of iPad, you type with your right hand and select with your left, without having to shift either hand.

Oh, I'm curious, what would the consonants and vowels they use (in English). In Hawaiian, we only have 12 letters: AEIOU and HKLMNPW

Japanese has five vowels, (a i u e o) and I think nine consonants (KSTNHMYRW). KST can be voiced, which is indicated by adding dots to the kana. "ka" is か. "ga" is が. H with dots is B, and with round dot is P. はばぱ ＝ ha, ba, and pa.

Very funny about many names starting with K! I think we get similar situations in Japan where foreigners mix up the vowels in words, and end up speaking complete nonsense.

 

[dabigkahuna]

Another question from one of my Japanese typists:

They wondered why you don't just type using romaji on the phone. It's all Greek to me, but little by little I'll put the pieces together!

 

[Night Spring]

Another question from one of my Japanese typists:

They wondered why you don't just type using romaji on the phone. It's all Greek to me, but little by little I'll put the pieces together!

Romaji is Japanified way of saying."Roman letters". And I'm horrible at typing on the iPhone, period. I like the kana keyboard better on the iPhone because each key is much larger, so easier to hit my target.

 

[dabigkahuna]

Well, I guess my next question, concerning those Chinese characters, is how many would you deal with (I have a feeling it is one of those "thousands" issues!)? But if it limited, you could type the regular letters as is and then use the green layer (there are other layers too) to access a number of common special characters pretty easily. Heck, I don't know how Japanese deals with capital letters, but if you don't have them the same way we do, then you could have some on the green layer, some in place of capitals (I'll have to check what is possible there as at one time I had an issue with some customization I tried when replacing capitals or something. Can't recall), and even green with shift.
[doublepost=1477117251][/doublepost]

Romaji is Japanified way of saying."Roman letters". And I'm horrible at typing on the iPhone, period. I like the kana keyboard better on the iPhone because each key is much larger, so easier to hit my target.

Ah, we may have gone in different directions about the focus of this - my way was about whether the TB would be as good or better than the on screen keyboard for Japanese. I had learned earlier today that kana was better because of the larger targets IF you are using an onscreen keyboard. But, of course, if you are using the TB, that ceases to be an issue - or am I missing something? I can certainly imagine a problem if you are simply used to kana and not used to a regular keyboard. That's why I was asking about what you used on a computer.

 

[Night Spring]

Chinese characters are in the thousands. I think in order to read the newspaper, you need to know about 5000, and the largest dictionary has way over 10,000 characters. So no amount of layers will suffice!

 

[dabigkahuna]

Chinese characters are in the thousands. I think in order to read the newspaper, you need to know about 5000, and the largest dictionary has way over 10,000 characters. So no amount of layers will suffice!

That's kinda what I thought, unfortunately. So the phone keyboard, using context, does contain thousands?

I'm trying to think of a solution and can imagine something where you type with the TB but still get suggestions. There would be the issue of avoiding having to grab the mouse to select one from any list. That may be possible by using a modifier for keyboard control. For example, if you held the space bar or some other special control key(s) and then other keys would move the cursor whatever direction you needed to select the suggestion you want.

Most TREG users have found the editing method on the TB much easier than the old way and this would be sort of like moving a cursor when editing - except you are moving to select a suggestion. If the number of suggestions are limited to a certain number (let's say 4), then you could handle it two ways - using an editing approach to go up and down the list (I assume vertically listed suggestions) or hitting one of 4 keys on the opposite hand to select directly.

However, at this point I don't think the TB is set up to handle this particular situation and it may require Apple to allow certain things as well since they'd be accessing Apple's suggestion list.

 

[Night Spring]

my way was about whether the TB would be as good or better than the on screen keyboard for Japanese.

Having never seen a textblade in person, it's hard for me to say if it'd be better than the on screen keyboard for Japanese, and other Japanese typists would probably have different opinions, because they work differently than I do.

As I said somewhere else in this discussion, the main thing that slows me down is selecting the right Chinese characters once I type the kana -- either directly or through roman characters. iOS word prediction learns as you type, so sometimes I can type entire sentences just by typing the first letter of each word. If I type ohayou gozaimasu (good morning) often, just typing "o" will bring up the entire phraseおはようございます. For some reason, the predictive learning on PC/Macs I've used have never been as good. Again, other Japanese users may have access to PC software that is as good as iOS. I've never bothered using anything other than what comes with the computer OS.

 

[dabigkahuna]

I did just find out that you can hit the space bar on the TB to call up the Chinese characters on the phone.

He said he needs to double check how he selects it so he's getting it out to test so I should hear soon.
[doublepost=1477118997][/doublepost]Here's what he said:

you press space to bring up the suggestions and then press space to cycle between suggestions and then enter to select one of the suggestions.

So better than having to grab a mouse or having to reach out to tap a selection.

 

[Night Spring]

So the phone keyboard, using context, does contain thousands?

Think of it more like doing a dictionary search. You type "alp" and you get a list of words that begin with "alp," including alphabet, alps, alpine, etc.

So you type "shin" and you get しん, then you get a list of common words beginning with that sound.


You can expand the list.


Or you can type more letters to narrow down the list.

 

[dabigkahuna]

This is getting beyond my knowledge - or logic - base! IOW, it's really hard for me to evaluate these situations without actually doing them myself - and I don't think I'm going to take up typing in Japanese or Chinese!

But I did pass on your example to a person who does and this is his response, which may or may not be useful:

Yeah, the prediction works too. You select it the same way -- by pressing space bar or using the arrows

Maybe some day you'll get to try it and can tell us how it works!

 

[Night Spring]

I did just find out that you can hit the space bar on the TB to call up the Chinese characters on the phone.

He said he needs to double check how he selects it so he's getting it out to test so I should hear soon.
[doublepost=1477118997][/doublepost]Here's what he said:

you press space to bring up the suggestions and then press space to cycle between suggestions and then enter to select one of the suggestions.

So better than having to grab a mouse or having to reach out to tap a selection.

Using space to bring up the list is the same as what you do with conventional keyboards. Cycling through the list by pressing the space keys many times -- I think that should also work with conventional,keyboards, but I just prefer to hit the arrow keys, maybe because I don't trust myself to not hit the space key too many times and move past the selection I want. Or maybe hitting the arrow keys or using the mouse brings up all selections at once while hitting the space key shows only one selection at a time. I don't quite remember, it's been a while since I typed Japanese on something other than my iPad! But there was something awkward about using spsace to cycle through the selections.
[doublepost=1477119812][/doublepost]

This is getting beyond my knowledge - or logic - base! IOW, it's really hard for me to evaluate these situations without actually doing them myself - and I don't think I'm going to take up typing in Japanese or Chinese!

But I did pass on your example to a person who does and this is his response, which may or may not be useful:

Yeah, the prediction works too. You select it the same way -- by pressing space bar or using the arrows

Maybe some day you'll get to try it and can tell us how it works!

All I can say is I used to do this on a computer, then I started using iOS, and I like iOS better. Has this person used iOS?

 

[dabigkahuna]

Apparently arrow keys work too. On TB it is part of the editing commands. You don't move your fingers away to arrow keys though. You hold down the F and G key to activate edit mode and then use the right hand on regular keys which now act like arrows in a logical position. You stay on or right next to the home row at all times. Actually really easy after a few days of adjustment. I don't know how the other edit movement keys would work in this situation. I mean, you can move up and down by line or left and right by character. But you can also jump left and right by word, or move by line, or by paragraph, or by page, or to the beginning or end of the whole document. At least that's how it works in a document. I don't know if all of these work in this selection process.

 

[Night Spring]

Well, to me, textblade is a variety of the physical keyboard, however innovative it's input method might be. And as I think I've mentioned before in the other thread, since I got my 12.9 iPad, I'm so satisfied with how the on screen keyboard is working for me, I've quite lost interest in physical keyboards.

Do you use iOS or Android? If you do, do you use predictive text?

 

[dabigkahuna]

Well, to me, textblade is a variety of the physical keyboard, however innovative it's input method might be. And as I think I've mentioned before in the other thread, since I got my 12.9 iPad, I'm so satisfied with how the on screen keyboard is working for me, I've quite lost interest in physical keyboards.

Do you use iOS or Android? If you do, do you use predictive text?

I've heard good things about that big ipad virtual keyboard. BTW, have you ever tested yourself on typing speed - there are lots of places online you can do it. I've mentioned I knew of one person who could type on a screen really fast (I think it was 80 wpm which amazes me). I have a theory that, once you hit a certain speed, most people are quite satisfied. Real slot speeds typically mean you aren't thinking about it right and that affects your comfort in structuring sentences, etc. I think for kids they said that something over 20 wpm meant they were functioning decently. For me, while I hate to be under 60 - mostly because it just bugs me personally - I think most people who are over 40 wpm feel comfortable and are quite satisfied. So, sure, nice to be faster, but I don't think you have to be really fast to get most of the benefits speed offers. That decent speed is sufficient to let your brain work well. Faster than that may mostly mean nothing more than you finish work sooner, but don't necessarily have other benefits.

I use iOS - besides my iPhone 6s, I bought the new iPad 9.7" Pro. I rarely use predictive text and I'm not a good screen typist! I simply don't practice that and don't type much on them anyway to get experience. Only rarely use predictive text because my eyes tend to be focusing on the keypad and what I see printed on screen. About the only time I even look at predictive text is when I know I just mistyped something and figure there is a good chance that the correct spelling is listed so I can just tap it. And some words I know will auto-correct so I just keep going with those. I'm amazing at how often I'll try to type "the" and instead type "hte".

 

[Night Spring]

I haven't tested myself in a while, but I'm probably quite slow. I have a physical disability that affects muscular control, so I'm never going to be fast, no matter what form of input I use. I was interested in textblade from a portability standpoint, not because I thought it would let me type any faster.

Anyway, Japanese input is basically predictive text on steroids. When typing Japanese, I'm staring at the predictive suggestions more than the keys or the document I'm typing. I think iOS works for me because the keys and predictive suggestions are pretty much in the same space, whereas when typing on a computer, I have to switch between looking at the keyboard and the screen. Perhaps if I were a better touch typist, or better at "translating" between kana and roman letters, it'd be diffferent. But I find I often have to look down at the keys to check what I'm typing because I'm unsure how to "spell" a kana, and then I become unsure of where the keys are... so onscreen works better, because I can stare at the keys as I type, and at the same time, I can see the suggestions come up without moving my focus.

 

[dabigkahuna]

Ah, well, any special physical issues is going to really complicate things. For me, sometimes when typing fast my fingers just go spastic occasionally. And sometimes the fingers have a little quiver to them - sometimes old age can be really annoying!

Well, if portability is the main attraction, it would be hard - I think impossible - to beat the TB. And it has the bonus of just feeling like a great keyboard as a bonus! Not to mention all the customization stuff, use with up to 6 devices, etc!

You said there are 50 kanas. It seems to me that that could be a nice thing. I mean, when I learned touch typing, the problem at first that I was reading letters, actually saying them in my head. It was when I was thinking in words instead of letters that I got much faster. Even patterns of words. So with just 50 kanas, I could see setting up custom typing tests to focus on those 50 things. Like what I do with the TB to solve letters I tend to miss. Sometimes when typing on a particular site that lists the words you missed and what you typed instead, I'll take those words and put into a text file, gradually in some sort of sentence form as much as possible - it will be nonsense sentences, but try to get nouns, verbs, adjectives, etc, in a fairly normal order so my brain isn't getting confused just because of a surprising pattern! Then I practice typing that a lot. Periodically I create a new list.

Of course, a lot of this stuff is dependent on how often you type. I type a lot so anything I get better at is more likely to be retained simply because I type a lot. If it was more rare, all that work may only help me temporarily and then be lost again.

 

[Night Spring]

Well, if portability is the main attraction, it would be hard - I think impossible - to beat the TB.

That is true if you have to have a physical keyboard. If you can manage with the on screen keyboard, well, nothing beats "nothing" in terms of portability.

The thing about typing Japanese is, you HAVE to pause quite often, like at least at the end of each clause, to convert from kana to Chinese characters and check that the conversion went well. So achieving rapid typing speed like English speed typists do is meaningless. An unconverted string of only kana is like an English sentence typed with no punctuation or spaces between words. Itwouldbeliketryingtoreadthis.

I did a lot more typing of both Japanese and English when I was in school, and I believe I was much faster then than I am now. Now, I'm not as fast, but it's good enough for the things I do.

 

[dabigkahuna]

I think I'm glad I don't have to type in Japanese or Chinese - sounds way to complicated for me. I don't mind too much if I'm personally not typing super fast (though I would be pleased to be able to do 100 wpm) as long as my speed is decent and it is simply my own ability that limits me. But I'd find it really irritating to be slow because of the system, be it hardware or language issues.

BTW, we are, slowly, getting info from testers who have the new units (new PCB, new keycaps). I only know of two so far, but at least that confirms that is happening.

 

[Rolanbek]

I think I'm glad I don't have to type in Japanese or Chinese - sounds way to complicated for me. I don't mind too much if I'm personally not typing super fast (though I would be pleased to be able to do 100 wpm) as long as my speed is decent and it is simply my own ability that limits me. But I'd find it really irritating to be slow because of the system, be it hardware or language issues.

BTW, we are, slowly, getting info from testers who have the new units (new PCB, new keycaps). I only know of two so far, but at least that confirms that is happening.

WT only confirmed Keycaps in public. Twice.

No confirmations from WT of new PCB's in TREG units yet. One tester states they have had Sensor changes confirmed.

If WT can get off their high horse to confirm Keycaps, and sensors (in private, not public) they can confirm new PCB's with recipients.

But, they have yet to do so.

Even TREG users seem hesitant to claim absolute certainty regarding PCB status. Which is fun.

R

 

[Night Spring]

I think I'm glad I don't have to type in Japanese or Chinese - sounds way to complicated for me. I don't mind too much if I'm personally not typing super fast (though I would be pleased to be able to do 100 wpm) as long as my speed is decent and it is simply my own ability that limits me. But I'd find it really irritating to be slow because of the system, be it hardware or language issues.

It is a complicated writing system, but it has its advantages. Like we can read Chinese texts from 3000 years ago with about a month's training, whereas if you wanted to read Latin from the Roman Empire, well...

 

[dabigkahuna]

It is a complicated writing system, but it has its advantages. Like we can read Chinese texts from 3000 years ago with about a month's training, whereas if you wanted to read Latin from the Roman Empire, well...

As it happens, I studied Latin two years in high school!

 

[dabigkahuna]

New, short, update last night:

>we are right now settling some detail points reported by a few users. They relate to charging and occasional swaps of characters. We're analyzing now to see what's needed to prevent them, and will update this week after we've worked out the causes and solutions.

I've seen some of this, but the most recent update I got, at least for the amount of time I had with it, seemed to get all the characters right. Even the non-typical events where I could force an error. An example of that is that I could hit a keycap on a certain character over and over again and, eventually, some would give me a different character (but still one of those on the same keycap). Long ago, my first experience with it, it may take me 40 or more hits in a row to get an error. With a later update it would take 80 or more. Then it got to 120 or more. How rapid you hit would make a difference and any pause would let things reset. In my more careful tests, I would be hitting pretty fast - like 320 hits a minute or faster.

Well, the last update I had, which was experimental at the time so only I and a couple others got it, I was hitting it in 3-400 times before just giving up because no error ever showed. I went through every key, some more than once. So it seemed that problem was gone.

There have been other cases of wrong characters that don't seem to be the same thing. Like for a little while, before an update fixed it, I was getting a lot of ENTER keys rather than P (again, they are both on the same keycap). After the last update, that stopped. However, that's when the problem with charging came up. I tried charging when it was under 10% and when I got up the next morning it still showed 10%. Not good. I wasn't worried since it should be no big deal to solve since it worked with all the other versions.

I downgraded the space blade section (the firmware is divided into 3 parts - one for each section - just so I could recharge again. About 2 minutes after I started the recharge, I looked at it and it was already up to 30% and over the next two minutes it got to 100%. That made me think it had been properly charging even with the new firmware, but the problem was in how it was displayed.

Based on how long it has lasted since them, I'm pretty sure I'm right. WayTools also feels that is the actual problem. So this issue seems like no big deal.

We know there is one even newer firmware version, obtained when a brand new treg tester picked up his device at WT headquarters today and he reported a higher number than anyone else has. Shouldn't be long before other testers can get it.

It's difficult to evaluate the issue of the accuracy of the characters because there are a number of different hardware and firmware versions out there. WT wants it that way so they can compare, but we testers often aren't sure exactly what hardware versions we have. For example, I know I have a version with at least most of the recent pcb board changes. But I do NOT know if I have the version there the boards were assembled using a method that avoids any stress. I also know I don't have the new keycaps. Some have older versions than I do. A small number of people we know have the latest stuff. They are the ones whose reports will be most important.

We can compare firmware easily though. But firmware that might work on some hardware could still have an issue with another piece of hardware. For example, when I got mine and had the P vs Enter key problem, I learned that my hardware sent a stronger signal about the keys. That's good, but they had to tweak the firmware to properly handle the stronger signal on certain keys.

As you can see from what I quoted above, there should be another update coming this week. And, yes, such things often take longer than they say, but I certainly wouldn't get upset about it unless nothing comes out by 3:30 AM California time (that is, Tuesday morning). That is the most typical update deadline because they review tests on Monday.

 

[dabigkahuna]

I'm getting a new TB today with all the latest stuff (new pcb board assembled in a way to eliminate any stress during assembly, new stronger keycaps).

Even when things are testing well on a TB I have, there is always the question of whether it also works well on the latest version - the version more likely to be an eventual shipping version. Or if I find a problem, does the newest hardware already solve it? So having the latest is something I've looked forward to.

At the same time, it makes sense for WT to keep multiple versions in the hands of testers. Since they get all the test results, they can compare and that lets them more easily find what causes a problem.

 

[redshifted]

I'm out .
I've cancelled my order from 3/2014.
WT has refunded my money to my CC.
If they ever start shipping product, I'll reorder I suppose.
Good luck to all the folks still waiting!

 

[dabigkahuna]

I received a new TextBlade today - latest hardware updates all applied. Just begun testing - the brute force stuff where I hit the same key at least 300 times rapidly. This is a test that, in the past, could eventually result in a different character on the same keycap to be sent. So far I haven't had that happen.