WayTools' TextBlade Keyboard to Start Shipping to Test Group Users Next Week

[dabigkahuna]

No, that is not what I wrote, though your sentence structure makes it almost impossible to figure out what you are saying. Short version of what I said was:

> From my testing experience, I don't know if I'd say there are major problems they don't seem able to fix. Rather, there are little issues that pop up, that do matter and which, typically, they fix quite fast <

What you describe in your second paragraph is pretty much what we are doing, btw. We find problems, report them, they solve them.

But, as I've said, more problems keep coming up. If you never had such things happen, good for you. Also lucky.

Funny thing, but we have plenty of Treg testers who were quite critical of WayTools about the same way you are. And after working with the device, they suddenly seem to understand.

 

[henry2]

My question is to you of the treg testing group .why has the (company) not come out with a single working product on time to be shipped to the people has paid for there product in 18 months time frame they have not shipped one working unit to a person or persons that have ordered them

 

[dabigkahuna]

My question is to you of the treg testing group .why has the (company) not come out with a single working product on time to be shipped to the people has paid for there product in 18 months time frame they have not shipped one working unit to a person or persons that have ordered them

They have come out with working products. I have one. It is "working" in the sense that I'd rather type with the TB, even though there are things they continue to fix or improve, than any other keyboard I've had.

But as with pretty much all "working products" there is always the question of how good it has to be before they do general shipping. BTW, the treg testers are also customers so that kinda kills the last part of your post.

Now, as it happens, asking a Treg tester why they haven't decided it is "ready" to ship is not something a tester can answer with any authority. We can only report our own experiences and make some basic logical points on what MAY be factors in their decision.

 

[henry2]

You may have one but the rest of us do not have there's in hand do we no because we kept getting told that there a problem .
So fix the problem and get the product out to the people

 

[dabigkahuna]

You may have one but the rest of us do not have there's in hand do we no because we kept getting told that there a problem .
So fix the problem and get the product out to the people

Yep, I and a number (low number) of other CUSTOMERS have a WORKING device. Which is what I was correcting.

No one even said there was "a problem". There have been problemS all along. Some big. Some much smaller. But even small problems, if there are enough of them, can combine to be serious enough not to ship.

IOW, you might have 10 small problems, none of which, by themselves, that would block release. Maybe not any two, or three, etc. But maybe they feel they have to eliminate at least 5 of them.

And hope no new ones show up!

It's kinda easy to look from the outside and act like it should just ship, or that new problems shouldn't be coming up, etc. Again, even the treg testers who were pretty critical changed their tune after using the device. Most of us have seen problems that justify not shipping. And we have also seen how complicated this thing is, which naturally makes it harder to get locked down the way they want.

And all of us want the thing to ship! But I don't know of any who think it should just ship now. Some would say they think it is good enough, but they also seem to understand that it isn't a clear cut decision -that it is up to WT.

 

[Night Spring]

And all of us want the thing to ship! But I don't know of any who think it should just ship now. Some would say they think it is good enough, but they also seem to understand that it isn't a clear cut decision -that it is up to WT.

Every time Apple ships a new product, there are things about them that isn't quite perfect, that gets fixed in the second gen, and people say Apple should have just waited until they fixed all problems before shipping. But if you wait until you clear up all problems, you run the risk of never shipping.

What you describe sounds to me like it could be an okay first gen product -- after all, you prefer it over traditional typing methods despite the many minor problems that crop up. Most early adapters would accept that this is a first gen product and live with minor issues, as long as overall experience is good.

I mean, look at Microsoft. They released the Surface, there were a lot of issues, they came back with the Surface 2, which addressed a lot of the issues but still didn't sell much, then they finally built a successful product with the Surface 3. Could they have waited until they had the Surface 3 before they released their product? Maybe. But if they had taken that attitude, I have a feeling they would still be working on their tablet product without releasing anything.

 

[dabigkahuna]

Yes, Apple products have had problems. But they are also Apple and to a considerable extent till be given time to make fixes. Heck, even if a problem made, say, an iphone 4 a major disaster, Apple could literally afford it. Most companies could not have a big failure if they only have one or two products and especially not if they didn't already have a huge fan base.

Besides, NO ONE, not even WT, is saying they are going to wait till it is perfect. Well, except for some critics who actually don't know. You see, they were saying such things for over a year yet it took me no time at all to find a bunch of problems all by myself that justified not shipping. So those critics tere certainly wrong a year or more ago.

I think if this was an Apple product, they could ship it and get away with any problems. Maybe WT could to. But maybe not. It is their call but either way there is risk.

I think most early adopters who are just into getting the latest and greatest would accept problems. I certainly have. But I suspect a majority of customers got it because it looked really good, but will also expect it to work very well. No way to know how they'd react to the present status. Well, the last status I've had - they've done improvements since my version.

While I think an argument could be made to ship now, the thing is, I think an argument can be made not to ship. So I don't get made at WT for choosing the latter. I'm really hoping we'll get their next update very soon and that it contains some seriously positive news.

The last info I have is that they had redone molds for stands and keycaps to deal with some breaking issues (very rare on the keycaps). I believe they had reached the point where they were ready to do test shots. But we haven't heard anything since. That testing, assuming nothing new came up, should be far along or even done. We need info on that and if testing was successful, learn if anything else needs doing before they start mass production leading to shipping.

 

[dabigkahuna]

WayTools posted a new, very detailed, status page update. Covers lots of stuff with pictures too. I'm only going to post one - fairly long - part here as it seems to be the main issue holding back release:

Inside every TextBlade are 4 computer cores connected in a parallel-processing network. Each KeyBlade contains it's own independent computer core. This is the brain that performs all the machine intelligence to understand the user's finger entries, and determine the intended character.

The core is robotically soldered to a flexible printed circuit board

We have done many large batches of these flex PCB assemblies (many thousands of PCB's per run), and they've been installed into our KeyBlade subassembly inventory. We test all of them at the factory before and after assembly. Every time TextBlade boots up, it also tests all these flex connections to verify that they work properly.

Our TREG customers are using their TextBlades that are built from these different batches.Deploying some from different batches exercises each build run, and provides a further end-to-end check of all our processes.

We saw a few cases where users reported some anomalies on certain keys. Some of these units also showed the test indicator lights. After analyzing them, we found that the copper lines near the core on some units had a microscopic intermittent disconnect, which could interfere with using a key.

What made this difficult to detect is that it did not present at factory testing. In some cases, users would only notice an issue after several weeks of good operation. Latent issues require the most care to find and correct.

After investigation, we traced the source to an assembly procedure where operators were stressing the flex PCB during installation.We've now changed the assembly procedures to prevent it.

We're studying our inventory now to see how many batches may be affected, and expect that to get sorted over the next few weeks. Because it's an intermittent, latent issue, the testing is not as simple as we'd like. We do know that some batches have this issue, and some don't.If necessary, we can make new fab runs to replace all these PCB's preemptively as a precaution.We'll have better data on that after some more evaluation of our inventory.

We're updating our TextBlade hardware inventory with the preventative measures found through TREG testing. The flex PCB corrections may extend through to next month, depending upon what we find.

 

[redshifted]

Will it ever be a product deemed by WT to be good enough to ship to non-TREG customers?
Shipping is already looking like at least October now based on hardware issues and fixes and testing and...

 

[dabigkahuna]

Will it ever be a product deemed by WT to be good enough to ship to non-TREG customers?
Shipping is already looking like at least October now based on hardware issues and fixes and testing and...

I'm not worried about it shipping, though I'm certainly frustrated with delays. This one, being so intermittent, is tough to find out which ones are messed up. It's like then you sometimes hear something wrong in your car engine, but every time you take it into the shop, it doesn't make the sound. Right now I don't know of any other hardware problems, but that doesn't mean that no one else has reported something that I don't know about. So, for now, I'm hoping this is the last big thing.

 

[redshifted]

I'm not worried about it shipping, though I'm certainly frustrated with delays. This one, being so intermittent, is tough to find out which ones are messed up. It's like then you sometimes hear something wrong in your car engine, but every time you take it into the shop, it doesn't make the sound. Right now I don't know of any other hardware problems, but that doesn't mean that no one else has reported something that I don't know about. So, for now, I'm hoping this is the last big thing.

Well, at this point I am worried about it ever shipping. The random cracked keys and the holiday delay on making the molds and testing the molds and on and on... weren't that big a deal. But now the flex PCBs are a bigger problem going forward. I worked in electronic manufacturing a long time ago and these things have a habit of coming back over and over. I think they blamed folks on the assembly line for stressing the flex PCBs during assembly. i'm beginning to wonder about the basic hardware design.

 

[dabigkahuna]

I think they blamed folks on the assembly line for stressing the flex PCBs during assembly. i'm beginning to wonder about the basic hardware design.

Anything is possible, but I doubt it is just blaming the assemblers to cover for a bigger problem. But there is an issue with the design - not that it is fundamentally flawed, but that it will require more care in assembly. As it happens, they posted more detail about it today:

All PCB copper traces, like the flexes in an iPhone, contain some level of imperfection, and redundant zones of copper generally provide some backup that eliminates this risk.

In this case, the fine pitch 4 mil (100) micron traces of copper, the flexibility demands, and the high level of thermal and mechanical excitation for the usage profile all combine to make this assembly less forgiving of overstress during fab. So we have to be more diligent about prevention.

The key part being that there is less forgiveness if they over stress it during fab. Now, one could assume this is a fatal flaw, but we don't actually know this - there is no official amount of redundancy that is required. All others thing being equal, sure, more is better. But then, that argument could "justify" a keyboard that weights 2 pounds just to zero out any possible problem.

If the stress is much higher during assembly than it would be in normal use, then they merely have to reduce the stress at the time of assembly..

There is much more to their post on this - it was far longer and more detailed than we usually see.

 

[redshifted]

Anything is possible, but I doubt it is just blaming the assemblers to cover for a bigger problem. But there is an issue with the design - not that it is fundamentally flawed, but that it will require more care in assembly. As it happens, they posted more detail about it today:

All PCB copper traces, like the flexes in an iPhone, contain some level of imperfection, and redundant zones of copper generally provide some backup that eliminates this risk.

In this case, the fine pitch 4 mil (100) micron traces of copper, the flexibility demands, and the high level of thermal and mechanical excitation for the usage profile all combine to make this assembly less forgiving of overstress during fab. So we have to be more diligent about prevention.

The key part being that there is less forgiveness if they over stress it during fab. Now, one could assume this is a fatal flaw, but we don't actually know this - there is no official amount of redundancy that is required. All others thing being equal, sure, more is better. But then, that argument could "justify" a keyboard that weights 2 pounds just to zero out any possible problem.

If the stress is much higher during assembly than it would be in normal use, then they merely have to reduce the stress at the time of assembly..

There is much more to their post on this - it was far longer and more detailed than we usually see.

From the WT forum referring to the problem of the flex PCBs failing...

"PCB fab is a large, high grade supplier in Taiwan, (supplying us for more than 10 years). Flex PCB fab is good. No issues there.

As posted, issue was with install stress and subsequent slow oxidation where copper lattice is parted.

Passivation is for mechanical protection, post-assembly. Passivation also helps resist oxidation, but doesn't stop it. Cure is to prevent source of stress on original install."


So November?
Maybe?
2016?

 

[dabigkahuna]

Could be November. Personally I'm hoping for September and, since there PCB testing will take "several weeks", I figure there is little chance for August.

Wish I knew, if they decided they just had to redo all of them, how fast that could be done. As I recall, when they do keycaps they do a whole set at once and they said they could do "many thousands" a day. Kinda a vague number, but I'd normally say that would be at least 5 thousand and could be far more than that. At that rate, they could start shipping soon after getting successful test results, but this process may not be as fast as keycaps.

 

[redshifted]

Could be November. Personally I'm hoping for September and, since there PCB testing will take "several weeks", I figure there is little chance for August.

Wish I knew, if they decided they just had to redo all of them, how fast that could be done. As I recall, when they do keycaps they do a whole set at once and they said they could do "many thousands" a day. Kinda a vague number, but I'd normally say that would be at least 5 thousand and could be far more than that. At that rate, they could start shipping soon after getting successful test results, but this process may not be as fast as keycaps.

September?
In a parallel universe maybe .

 

[dabigkahuna]

September?
In a parallel universe maybe .

Well, it's always easy to assume it will be longer considering how many times it has been delayed. But, unless you assume it will never ship, at some point it will actually happen. So my hopes are based on what I see as possible, knowing that certainly doesn't mean it will happen. In this case, we know they expect several weeks of testing. But if that testing comes up good and nothing else is found, then why not September? They'd have a full month - possibly longer if testing is done before the end of August - to get enough things made to ship.

Oh, I expect some other steps - for example, I suspect that they'll send some out for Treg testing (though I believe they said something about this kind of testing is best done in house). But Treg testing may catch any new issue that these changes may cause so I still expect some. That can still get us September shipping.

But no, I won't be surprised if it doesn't. Certainly can't predict how testing will go or if anything new will come up.

 

[Night Spring]

Certainly can't predict how testing will go or if anything new will come up.

Do you even remember how many times you've said this now?

Each time a new issue is found, WayTools says they've found a solution, is testing the solution, and TextBlade will ship if testing goes well and no new issue comes up. Well, so far, a new issue has been found EVERY TIME.

I suppose, one day, it could happen that no new issue is found and TextBlade would actually ship. But if people were taking bets, I'd bet odds would be heavily for yet another new issue cropping up.

 

[dabigkahuna]

Do you even remember how many times you've said this now?

873 times! I just say it because so often people seem to think they know how long something should take or know how many or how long problems should take to solve. I don't agree with them so I'm upfront about my view which leads to my comments about possible shipping. That is, not that I can say with any assuredness when they will ship but only, based on what info we do have, what MIGHT be possible.

Each time a new issue is found, WayTools says they've found a solution, is testing the solution, and TextBlade will ship if testing goes well and no new issue comes up. Well, so far, a new issue has been found EVERY TIME.

True. But then, that applies to pretty much everything - new issues that affect shipping come up right up to the point that they don't come up. The difference is that this has happened a lot more than expected.

I'd bet odds would be heavily for yet another new issue cropping up.

Well, that is supported by the record, but, again, unless the assumption is they will never ship, eventually the pattern will change. Meanwhile, we all get to be frustrated and expect yet another problem.

As I've thought for many months, this new tech approach is having issues. Things that probably wouldn't be a big deal a few years down the road, when there is more experience with it, are continually being discovered now.
[doublepost=1471235728][/doublepost]WT just posted new info (lately they've been MUCH better!):

a specialized flex-carrier fixture is used for new builds so the operators can't bend that area, and the PCB is shielded from any possibility of overstress.

Current inventory is what we're studying now, to determine if we can certify it conclusively, or if we should instead preemptively replace all pcb's as a precaution. That's what we're sorting out now.

 

[Night Spring]

new issues that affect shipping come up right up to the point that they don't come up.

Read:
http://www.folklore.org/StoryView.py?story=Real_Artists_Ship.txt

If it had been up to these engineers to decide whether to ship the Mac, the Mac would have never shipped. But they had Steve Jobs to tell them the Mac is shipping NOW, that the product is good enough even if it's not perfect.

 

[dabigkahuna]

Read:
http://www.folklore.org/StoryView.py?story=Real_Artists_Ship.txt

If it had been up to these engineers to decide whether to ship the Mac, the Mac would have never shipped. But they had Steve Jobs to tell them the Mac is shipping NOW, that the product is good enough even if it's not perfect.

Well, I'll make some points about this.

1.There are also all the stories about Jobs being a perfectionist and not shipping things that weren't good enough to him.

2. The example you cite was also about software. And the example of the formatting bug looked like something most people wouldn't ever see - so the negative feedback would be minimal and could be fixed later.

With the TB, we are stuck on a hardware issue.

 

[Night Spring]

1.There are also all the stories about Jobs being a perfectionist and not shipping things that weren't good enough to him.

But did he ever stop something from shipping AFTER the product had been announced to the public, shipping date set, and orders taken?

2. The example you cite was also about software. And the example of the formatting bug looked like something most people wouldn't ever see - so the negative feedback would be minimal and could be fixed later.

With the TB, we are stuck on a hardware issue.

Why are hardware issues still cropping up at this late date?

 

[dabigkahuna]

But did he ever stop something from shipping AFTER the product had been announced to the public, shipping date set, and orders taken?

Well, the white iphone was delayed a very long time. Don't recall if they took orders though.

But that isn't really the right comparison. The question should be if they ever had a problem come up that, in spite of Jobs' rantings, they still couldn't solve well enough in time. Would they have shipped anyway? I important enough, based on their own standards, I doubt it.

Why are hardware issues still cropping up at this late date?

Simply a fact that they have. That is something I've commented on many times. There is no way to guarantee there will only be X number of problems, especially with new tech. I've seen many people claim that proper planning takes care of that possibility. I don't believe it for a minute. It reduces the chances, but no planning in the world can avoid the unknown.

 

[redshifted]

Well, it's always easy to assume it will be longer considering how many times it has been delayed. But, unless you assume it will never ship, at some point it will actually happen. So my hopes are based on what I see as possible, knowing that certainly doesn't mean it will happen. In this case, we know they expect several weeks of testing. But if that testing comes up good and nothing else is found, then why not September? They'd have a full month - possibly longer if testing is done before the end of August - to get enough things made to ship.

Oh, I expect some other steps - for example, I suspect that they'll send some out for Treg testing (though I believe they said something about this kind of testing is best done in house). But Treg testing may catch any new issue that these changes may cause so I still expect some. That can still get us September shipping.

But no, I won't be surprised if it doesn't. Certainly can't predict how testing will go or if anything new will come up.

Your timeline is a fantasy.
Everything has taken WT much longer than anyone can imagine.
It's not wish fulfillment that gets it shipping.
It's engineers not signing off on everything that gates shipping.

TREG testing at some point is a hindrance to saying it's "good enough to ship".
If it's good enough for 99-90% of users it's good enough.
It seems to be at that point now.
Firmware seems stable.
Hardware? Maybe.

The other problem gating shipment is the oxidation of PCB traces that corrupt the signaling through the device after only a couple of months of use.
It's hard to understand how improper installation (despite "passivization") of flexible PCBs would lead to this.
If oxidation (as WT describes it) is really an issue, then it potentially limits the useful life of the device for many buyers/users.

September isn't happening.
October has a very slim chance.
November has some chance if nothing else crops up in the meantime.
December or January 2017 is the most likely - if it ever ships to regular customers.

If I lived in Hawaii and had a mostly functioning TextBlade, I might have a more optimistic point-of-view at this point. I want this thing to work, as much as anyone. It's becoming hard to see that as a possible reality.

NOT Typed on a TextBlade

 

[dabigkahuna]

Your timeline is a fantasy.

All timelines are fantasies right up to the time one comes true. Besides, as I said, I'm just posting possibilities, usually based on the earliest possibility (since there could be an infinite number of other ones!).

TREG testing at some point is a hindrance to saying it's "good enough to ship".

Certainly possible - or that they actually are catching important issues that should be fixed before shipping. Works both ways.

If it's good enough for 99-90% of users it's good enough.

Uh, no. For one thing, that is up to the company. Second, you really think having 1 in 10 fail for a KNOWN problem is okay? Remember, there will always be other problems. You can just look at the most popular apps and still see that some people give one star with comments like, "Crashes every time, don't buy" while almost everyone else is giving 5 stars. So you are adding a lot to the unavoidable problems that always occur for some people.

It's hard to understand how improper installation (despite "passivization") of flexible PCBs would lead to this.

Well, I read their description. Didn't seem hard to understand to me.

 

[redshifted]

All timelines are fantasies right up to the time one comes true. Besides, as I said, I'm just posting possibilities, usually based on the earliest possibility (since there could be an infinite number of other ones!).

Certainly possible - or that they actually are catching important issues that should be fixed before shipping. Works both ways.

Uh, no. For one thing, that is up to the company. Second, you really think having 1 in 10 fail for a KNOWN problem is okay? Remember, there will always be other problems. You can just look at the most popular apps and still see that some people give one star with comments like, "Crashes every time, don't buy" while almost everyone else is giving 5 stars. So you are adding a lot to the unavoidable problems that always occur for some people.

Well, I read their description. Didn't seem hard to understand to me.

No, September is pure fantasy. October is unlikely. "Earliest possibility" estimates are exactly why folks are continually disappointed when delivery dates slip forward another couple of months. There aren't infinite possibilities in this universe. There are just month-to-month ship/no ship decisions being made. There is however a very specific hard reality of no shipment to paying customers.

At some point they're no longer catching anything important. Catching things sure, but not important things in regard to shipping TextBlades to most customers. This would be the customer feedback phase.

Yes, it's up to the company but that doesn't mean all those problems reflect any real problems for 99% of users. None of the problems that WT has talked about is at the 1 in 10 range. They might turn out to be that bad down the road - so some problems are real problems. Those "crashes everytime" "don't buy" ratings are usually not universal problems. Often they are user errors that require hard-restarts or re-installing the app or talking to the app developer. The apps are usually just fine. You're just playing the anecdotal case vs. the statistical case. You will never have 100% of users happy with anything. This is a weak argument.

I don't think you know enough about electrical or mechanical engineering if you accept what they are saying about this problem. Improperly installing flex PCBs does not cause oxidation of the traces months later - as far as I know. Stress fractures sure, but oxidation? Unless, you happen to live near a volcano .

NOT Typed on a TextBlade