Author Topic: ErgoDox extended to 88+ (full set of) keys – “Lekhani”  (Read 19575 times)

0 Members and 1 Guest are viewing this topic.

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« on: Sun, 19 July 2015, 06:33:08 »
I was thinking maybe one of the reasons ergonomic keyboard designs like the ErgoDox and derivatives, despite their obvious benefits, are not more popular is that lots of those designs cut down on the number of keys. (Yes I know the Kinesis, TIKE etc don't compromise on the keys, but the Kinesis's rubber F keys are a blot on an otherwise excellent design, and the TIKE's unreliability re stuck keys and customer service is a big no-no... If it were not for these, I would have already bought one of the two...)

Asking people to re-learn their usage patterns will be definitely met with inertia. While certainly it is to their own benefit to unlearn what are actually bad habits resulting from the usage of a staggered-rows keyboard, asking them to sacrifice keys – either give up the digit row and the symbols on them if they want to use F keys a lot, or asking them to use Fn+ combinations for F keys all the time – is too much IMO. Providing them with the well-known set of keys is important.

Of course, almost nobody uses Scroll Lock and Pause/Break, so those can go out (and put on a higher layer such as Fn).

And preferably, if it can be done without increasing the overall footprint, a few more keys with other programmable usages can be included for convenience.

I initially targeted the 88-key ISO layout (to be more applicable worldwide, and for L/R symmetry) and just removed Scroll Lock and Pause/Break and put in "Undo" and "Save" buttons instead, and replaced one GUI (aka Win/Meta/Cmd) key with Fn/Num Lock, so retained the 88-key count.

Some of the changes I first made to the ErgoDox design was:
  • More staggering a la the Axios.
  • Added the Esc+F+Prt row on top. I keep them adjacent to the digit row to minimize stretching.
  • Directional arrow clusters a la TIKE (just trying to be less surprising – I am personally OK with making those 4 keys flat i.e. in a row).
  • Radial thumb cluster ("TC") a la other designs. Note that these are only 5 keys and all 1.5u which makes them easier to target IMO. The lower three to be accessed by the thumb and the inner two by the index/middle fingers with a wrist rotation for chording. I dropped the sixth least accessible position.
  • Inner column on either hand all 1u-s since otherwise it seems to conflict with the 1.5u-s of the thumb cluster.
  • Detachable halves with joint in middle.
  • Of course, tiltable and tentable. Default tent of 15° or so.

The resulting layout was:
107090-0

Now I realize I can add four more keys without increasing the footprint of the board. So:

  • I have restored the second GUI key and pushed it to the TC where it is probably more useful for chording.
  • I have also split out the Fn/Num key which was retained for simplicity of recognition to more powerful Layer+/− keys. I initially intended for the Fn/Num key to be used by tap/hold behaviour, but possibly not all ErgoDox-like firmwares support that, so this is much better, and power users can put this much better use whereas ordinary users will just retain it for going to the NumLk/Mouse layer and back. This also means that Fn sequences (SysRq, multimedia etc) will have to be handled by one or both of the GUI buttons, which is OK I guess (assuming tap/hold behaviour is not available).
  • I also added a Macro key, which would earlier have to have been handled by some obscure Fn combination. Shift+Macro would be sufficient to start online macro recording I suppose (and hitting Macro ends it).
  • Finally, the "Lng" key corresponding to USB HID code LANG1 would be useful to switch virtual keyboard layouts (or, always, power users can reconfigure it).
More importantly as overall design point modifications:
  • Brought the TCs in closer based on the feedback of the nice people here. Even in the first iteration I had a bit of misgiving about this so I went and fixed it.
  • Removed the radial symmetry of the TC which is considered "overrated" (and which might cause extra work in PCB design, I dunno).
  • Trimmed out the excessive unused space for the board for being lighter and more portable – people will use the available surface or their own palmrests.
  • Removed the joiner in the middle. The assembly should be such (with rubber padding or whatever on the bottom) that the halves don't move around all that easily, thereby obviating the necessity of such a joiner (and the development/production costs associated with it).

So here it is:
107088-1

I initially called this the ErgoMax, but that name seems to be overused by googling for it. I thought of other Ergo* options but got bored with overuse of the "ergo" wrod. I will hence do what I should have done first, give it a simple (internationally-pronounceable) name from my native languages: “lekhani” which means "instrument of writing", commonly used for "pen".

Now I realize I'm a noob amidst people much much more experienced. Just wanted to share my thoughts though and get feedback. I also have totally no knowledge of PCB design etc etc, so this is just the overall layout I'm posting just because I can dream of what I'd like to have! :-) I don't have the knowledge to work on this myself. (So I'm not posting it under the "making stuff together" forum.)

Apparently Zustiur is making something very very similar, and I hope he finds success. Hopefully somebody finds this design interesting enough to work on it. I'm open to design changes as seen above.

If such a design becomes publicly available, I think it would be easier to convince many people to buy it, rather than the ErgoDox which forces too much change that not everyone would be willing to put in. I think people should be met half-way and should not be deprived of the opportunity to save their hands just because they wanted the extra 12 keys (ISO 88 − ErgoDox 76)...
« Last Edit: Sat, 01 August 2015, 07:09:55 by jamadagni »

Online Findecanor

  • Posts: 4233
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: ErgoMax
« Reply #1 on: Sun, 19 July 2015, 07:05:44 »
Dude, there is already a vintage keyboard from Maxi Switch which is called ErgoMax.
I had stumbled into this thread expecting to read about it.
« Last Edit: Sun, 19 July 2015, 07:08:13 by Findecanor »
Smplesh dar Ventöl pööl pööl
Daily driver: Phantom (Lubed Cherry MX Clear, Lasered Cherry PBT keycaps with Row A. Plastic "Frankencase". Custom firmware, Swedish layout)

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoMax
« Reply #2 on: Sun, 19 July 2015, 07:23:57 »
Oh -- I never googled for that word. Just pulled it out of thin air. I am not trying to market anything. Probably I should change the thread title...

Offline nick2253

  • Posts: 7
Re: ErgoMax
« Reply #3 on: Tue, 21 July 2015, 14:53:23 »
Dude, there is already a vintage keyboard from Maxi Switch which is called ErgoMax.
I had stumbled into this thread expecting to read about it.

IIRC, Maxi never manufactured the Ergo Max.  They did have their Ergo Master keyboard.

This thread discusses it:  https://geekhack.org/index.php?topic=12625.5

@jamadagni: Since it took me about 30 mins of Googling to come up with that thread, and your thread came up right away when I searched for "ergomax keyboard", I'd say you're safe to use it.  I don't think you'll be creating too much confusion.

I agree with your assessment about the number of keys.  It is one of the biggest sticking points I have with ergo keyboards is the lack of proper F keys and whatnot.  88-keys seems to be a good compromise.

Offline Heliobb

  • Posts: 94
Re: ErgoDox extended to full 88 keys – my take
« Reply #4 on: Tue, 21 July 2015, 15:23:32 »
If such a design becomes publicly available, I think it would be easier to convince many people to buy it, rather than the ErgoDox which forces too much change that not everyone would be willing to put in. I think people should be met half-way and should not be deprived of the opportunity to save their hands just because they wanted the extra 12 keys...

The ergodox force to too much change ?

Have you tried to print your layout to test it ?
Novatouch TKL - Leopold FC660C - PBT my life.

Offline nothing4me

  • Posts: 28
Re: ErgoDox extended to full 88 keys – my take
« Reply #5 on: Tue, 21 July 2015, 22:12:12 »
I feel like CTRL + ALT keys would be a little hard to reach.

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #6 on: Tue, 21 July 2015, 22:28:30 »
Personally I think that more keys are inevitably worse for ergonomics and usability, but you’re probably right that some people will reject keyboards without piles of unnecessary keys.

Anyhow, along the same lines as your idea here, but with IMO some practical advantages/refinements (in particular your thumb keys are way out of position), here was my proposal, on the left (on the right is a smaller version lacking F keys etc.):



From this thread https://geekhack.org/index.php?topic=62848.150

Compared to a standard keyboard:


Compared to an ergodox:

« Last Edit: Tue, 21 July 2015, 22:31:56 by jacobolus »

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to full 88 keys – my take
« Reply #7 on: Wed, 22 July 2015, 05:54:16 »
@nick2253: thanks for the reassurance, but nevertheless googling for just "ErgoMax" turns out too many websites/products. As I said, I'm not trying to market anything, so any other name is fine...

@jacobolus: I actually read in some detail your alternate proposals for the thumb cluster and I think you really have put some serious thought into it! But some issues I have with your proposals (which I should probably mention in that thread) are:
  • the thumb cluster is not radial,
  • I'm not sure what your idea was being placing the F keys radially in twos. I can understand that some might prefer to not have the F keys right next to the number ones like I propose, but placing them radially is exactly useful how? (I hope I didn't miss any explanation of this on that thread.)

BTW I think that compared to 105 key boards, 88 key boards certainly don't have "piles" of unnecessary keys. It's true that much more minimalism might be practical for real geeks (as seen by designs such as the Atreus etc), but to the common person, it requires seriously relearning one's usage patterns. It's difficult already to get people to learn non-QWERTY virtual layouts, and asking them to learn totally different physical layouts and hence re-learn the positions/usage patterns of numbers/symbols/F keys too would be too much. (Especially, having to choose from between F keys vs number/symbol keys convenience.)

As I said, I myself am probably getting an ErgoDox or Axios (since I don't know enough to make myself my own custom design) and will be learning something like the layout here but in thinking more about others, I feel if such an 88-key design were to be actually produced, it would benefit more people.

BTW what is your problem with the thumb cluster I propose? Other than that it probably should be a bit closer to the center of its radial symmetry?

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #8 on: Wed, 22 July 2015, 16:03:34 »
> the thumb cluster is not radial

You should build some prototypes and try them. I built about 20 different versions out of acrylic (mostly not wired up), and many more paper printouts and incremental clay experiments, and my conclusion was that “radial” thumb keys are overrated.

In particular, once they swivel a thumb far enough down, many people find it difficult/uncomfortable to press a key while still maintaining flexibility in the other fingers. On the larger of those example pictures I linked, the furthest-away thumb key is actually a bit uncomfortable, but the goal there was to have at least as many keys in most of the same places as the Ergodox. As I said, I think having more than about 30 keys on each hand is overkill, and I find the smaller version to be very effective.

If you do like radial thumb keys though, you might consider pre-ordering a keyboard.io. They have a pretty solid design, which they have put a lot of thought and work into.

> I'm not sure what your idea was being placing the F keys radially in twos.

The main idea behind the pairs of keys is that you can relatively easily target each individual key without accidentally pressing the wrong one; one possibility would be to use each of your three main fingers for one pair of keys (you don’t want to reach the pinky up that far). Having some shape to keys which require whole-hand reaching helps dramatically with locating them quickly by touch. This is the same idea as putting the arrows on a standard keyboard in an inverted T shape with gaps around it, or putting F keys in groups of four. However, in my opinion the standard F keys are in overly large groups. It’s difficult to e.g. target F6 or F7 accurately on a standard keyboard

If you place the keys directly adjacent to the number keys with no structure or gaps you will get errors and overall it will be slower.

Really though my own preference would be to scrap the F keys altogether, and put their functions on a layer.

> I think that compared to 105 key boards, 88 key boards certainly don't have "piles" of unnecessary keys.

Well they have at least 20 keys more than any reasonable keyboard should have. And they have about 40 keys more than a minimal but perfectly usable keyboard. The extra keys make them less efficient and more error-prone, though a bit more learnable and closer to the (horrible) standard keyboard design. For people coming from a standard keyboard, it might be a reasonable trade-off to keep the unnecessary keys.

> what is your problem with the thumb cluster I propose?

I think you should build it and try it out! It’ll be a fun exercise, and you’ll learn a lot about how your hand works. Even if you just print your design out on paper, you can get some idea of how it would work.

For anyone from small to medium-large hands, only about 1 of your proposed thumb keys is in a reachable spot, and none of the keys would be fun to use as a spacebar. For someone with large hands, more of the keys might be in reach, but they’re still not in very practical spots, IMO.
« Last Edit: Wed, 22 July 2015, 16:16:50 by jacobolus »

Offline nick2253

  • Posts: 7
Re: ErgoDox extended to full 88 keys – my take
« Reply #9 on: Wed, 22 July 2015, 16:44:50 »
> I think that compared to 105 key boards, 88 key boards certainly don't have "piles" of unnecessary keys.

Well they have at least 20 keys more than any reasonable keyboard should have. And they have about 40 keys more than a minimal but perfectly usable keyboard. The extra keys make them less efficient and more error-prone, though a bit more learnable and closer to the (horrible) standard keyboard design. For people coming from a standard keyboard, it might be a reasonable trade-off to keep the unnecessary keys.

I regularly see this kind of minimalism here, and I don't know what the obsession with it is.

If you are going truly minimal, you can build a keyboard with 9 keys, and have everything be chorded.  If that sounds ridiculous to you, it's for the same reason that ~50 key keyboards sound ridiculous to the rest of the public:  the memorization required to operate such a keyboard is beyond the desires or the time of most people.

Already, most keyboards have at least two, if not five or more layers:  normal, shift, Fn, Alt Gr, Alt Gr + Shift, not to mention dead keys.  These layers are difficult to use, unless the keys are labeled.  Most people have failed to commit the normal layer to memory, much less every single punctuation mark in the shift layer and the myriad of various characters found under the Alt Gr layer.  Unless you can produce keycaps with labels for an additional layer, you're not going to get much traction.

Furthermore, the F keys are widely used on modern computers:  many programs use the F keys for functional reasons, and many laptops use the F keys for basic functions, like volume, brightness, etc.  If you remove those keys, suddenly, you have to introduce a third layer (normal, shift, special) to have those.  You call a keyboard without F keys "reasonable" and "perfectly usable."  But, for me, and anyone using my employer's line-of-business software, a keyboard without the F keys would be terrible.  At least 10% of the keypresses I make daily are an F key.  Putting them in a layer is a terrible choice.

The biggest thing that puzzles me about this is that the downsides of adding more keys is remarkably small:  just a few more switches and keycaps, and a slightly larger/heavier keyboard.  On the plus side, not only do you have more keys for programming, macros, usability, etc, there's also that the marginal cost of more switches and keycaps could be more than offset by the possible volume savings we would benefit from if an ergo mechanical keyboard went mainstream.

I'm not saying that there isn't a place for ultra-minimalist keyboards.  But, my point is that, in my mind, the obsession with minimalist keyboards is doing the hobby more harm than good.

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #10 on: Wed, 22 July 2015, 17:41:25 »
If you are going truly minimal, you can build a keyboard with 9 keys, and have everything be chorded.  If that sounds ridiculous to you, it's for the same reason that ~50 key keyboards sound ridiculous to the rest of the public:
It doesn’t sound ridiculous, I think chorded keyboards have their place. A 9-key keyboard is going to be fairly inefficient (probably max of like 40–50 wpm), but a chorded keyboard with like 25–30 keys works pretty well, at the expense of a very steep learning curve. Indeed, stenography keyboards are by far the most efficient text input devices in wide use.

A 50–60-key keyboard is not that similar to a chording board though, and I don’t think your analogy makes much sense.

Quote
Already, most keyboards have at least two, if not five or more layers:  normal, shift, Fn, Alt Gr, Alt Gr + Shift, not to mention dead keys.  These layers are difficult to use, unless the keys are labeled.
The primary reason these layers are difficult to use is that the layers are terribly designed and the modifier keys are in awful awkward places. In particular pressing two modifiers at once is impossible to do with speed and comfort.

If the physical keyboard design is improved, layers (both the standard layers and additional user-defined layers) are much less problematic.

Quote
Furthermore, the F keys are widely used on modern computers:
You can’t have concern for learnability and novice users in one paragraph, and then turn around and defend F keys in the next paragraph. F keys are absolutely terrible for novice users.

Quote
The biggest thing that puzzles me about this is that the downsides of adding more keys is remarkably small:

The downside for additional keys is a huge efficiency and error hit, as well as wasted desk space and worse ergonomics for pointing devices, for essentially zero benefits.

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to full 88 keys – my take
« Reply #11 on: Wed, 22 July 2015, 20:14:05 »
Hi @jacobolus – thanks for that very clear explanation. However, don't you think that once a person starts to test with models on their own fingers, it will differ from person to person? What is needed for a layout suitable for a majority number of people is perhaps some sort of proper university study (backed with sufficient grants) to reach a wider section of the public, but then even that probably will differ across peoples such as Indian, Caucasian, African, and so on...

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #12 on: Wed, 22 July 2015, 22:04:06 »
don't you think that once a person starts to test with models on their own fingers, it will differ from person to person?
Absolutely. Which is why you need to make a prototype and then get a bunch of different people to try it.

It’s impossible to make something that works for everyone, but you can hit a good percentage of the population with a careful design.

Offline nick2253

  • Posts: 7
Re: ErgoDox extended to full 88 keys – my take
« Reply #13 on: Wed, 22 July 2015, 23:14:26 »
It doesn’t sound ridiculous, I think chorded keyboards have their place. A 9-key keyboard is going to be fairly inefficient (probably max of like 40–50 wpm), but a chorded keyboard with like 25–30 keys works pretty well, at the expense of a very steep learning curve. Indeed, stenography keyboards are by far the most efficient text input devices in wide use.

A 50–60-key keyboard is not that similar to a chording board though, and I don’t think your analogy makes much sense.

The requirement to become familiar with layers and modifier keys makes it exactly like a chording board, but just to a much smaller degree.  Layer button + 1 = F1 is a chord.

No disagreement that chorded keyboards are the most efficient, but, like you said, they require a steep learning curve.  Expecting casual users to put in hours of learning for their daily use keyboard is ridiculous.

Quote
The primary reason these layers are difficult to use is that the layers are terribly designed and the modifier keys are in awful awkward places. In particular pressing two modifiers at once is impossible to do with speed and comfort.

If the physical keyboard design is improved, layers (both the standard layers and additional user-defined layers) are much less problematic.

You still don't overcome the difficulty of users being able to successfully identify the keys.  And even an improved layout does not change the fact that you require modifiers.  Instead, a keyboard with more keys would actually solve this problem, by avoiding overload for uncommonly used characters.

Quote
You can’t have concern for learnability and novice users in one paragraph, and then turn around and defend F keys in the next paragraph. F keys are absolutely terrible for novice users.

Can you not accept that extra keys benefit both novice and experienced users?

How are the F keys terrible?  Just because you assert that, it's not the case.  We train novice users on the F keys all the time, and it makes interaction with our line-of-business software easy.  And just because someone may be an experienced computer user, doesn't mean they are an accomplished keyboardist.  For just one example, quite a few programmers have poor typing skills, but depend on the F keys for daily interaction with their software.  They may not be pressing the keys all the time, but I can assure they are widely used.

Quote
The downside for additional keys is a huge efficiency and error hit, as well as wasted desk space and worse ergonomics for pointing devices, for essentially zero benefits.

I've clearly demonstrated a number of benefits, so your "zero benefits" claim is completely disingenuous.

How does adding additional keys make a huge efficiency hit, when instead of one key press, I'm now forced to make at least two?  How does adding additional keys make a huge error hit, when now I'm forced to hunt and peck through layers, since no labeled keycaps are available?

How does adding keys to the top or bottom of a keyboard make worse ergonomics for pointing devices?  How much desk space does one or two extra rows of keys actually take up, especially keys added below the keyboard?

Your reasons for opposing extra keys are dogmatic.  Like I said, I believe such minimalist keyboards have a place, and if that works for you, that's fantastic.  But they do not work for the vast majority of the general public, and they obviously don't work for a number of users here, what with the outpouring of support for keyboard projects like the Axios (98 keys).  Just because full-size keyboards don't work for you, doesn't mean you have to oppose them for others.

Online user 18

  • * Senior Moderator
  • Posts: 2125
  • Location: Ontario, Canada
Re: ErgoDox extended to full 88 keys – my take
« Reply #14 on: Wed, 22 July 2015, 23:59:44 »
BTW I think that compared to 105 key boards, 88 key boards certainly don't have "piles" of unnecessary keys. It's true that much more minimalism might be practical for real geeks (as seen by designs such as the Atreus etc), but to the common person, it requires seriously relearning one's usage patterns. It's difficult already to get people to learn non-QWERTY virtual layouts, and asking them to learn totally different physical layouts and hence re-learn the positions/usage patterns of numbers/symbols/F keys too would be too much. (Especially, having to choose from between F keys vs number/symbol keys convenience.)

There's a lot going on in this thread that's over my head -- I use ANSI layout boards and like it -- but I wanted to share an experience of my own relevant to this particular point.

Learning a different physical layout is actually easier for me than learning a new keymap for an existing layout. Once I have developed and established muscle memory for a particular action (character, sequence, chord), I associate it with the environment and the tool I'm using. Switching to a different physical layout helps me break some of those connections, and I can focus on developing a new set of associations for the new physical context. If I were to try and learn a new layout (e.g. Colemak), I would find it to be difficult because I need to fight my old associations (qwerty) as well as develop new ones. Starting from scratch is better than starting in a hole, so to speak.

For example, the different laptops I use have different physical layouts and different function layers, as well as being physically different machines. I've associated the physical presence and layout of each laptop to its particular layout, and very rarely confuse layouts. Finding the 'end' key for example, requires three different movements on either of my laptops or my desktop. One laptop locates the key on the function layer. I don't frequently find myself pressing the wrong key (and I use home/end a lot).

I would argue that a nontraditional reduced layout might actually be easier to get used to, as a user wouldn't be fighting muscle memory in trying to learn the new system.

Maybe this particular method of association isn't common (I definitely haven't done any research or formal study), but I know for certain that it's not entirely uncommon. Just food for thought.
Please PM me if you are waiting on classifieds approval or have a question about the classifieds rules. | geekhack Terms of Service

Max Nighthawk x8 (MX Brown) | CM QFR (MX Blue) | CM QFR (MX Clear) | RK-9000 (MX Red) | Model M 1391401 | Model M SSK 1370475 | CM Novatouch | G80-8113 (MX Clear) | 60% (85g MX Blue) | Whitefox Aria (MX Clear) | CL-LX (MX Clear) | Mira SE (MX Clear)
Avatar by ashdenej

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #15 on: Thu, 23 July 2015, 00:45:08 »
No disagreement that chorded keyboards are the most efficient, but, like you said, they require a steep learning curve.  Expecting casual users to put in hours of learning for their daily use keyboard is ridiculous.
A “casual user” with lots of existing computer experience who wants to save himself from RSI but not spend effort learning something new should probably get something like a Matias ErgoPro, and be done with it. Anything remotely similar to an Ergodox is not for him.

By contrast, a “power user” who plans to invest a bunch of effort in keyboards, such as learning a new physical layout, customizing her layers, etc., should spring for the 50–70 key board, and then put some thought into the setup. She’ll type circles around the other guy.

Quote
You still don't overcome the difficulty of users being able to successfully identify the keys.
Identifying the keys sucks on absolutely every keyboard for every non-trivial piece of software. No keyboard comes with every software application’s shortcuts printed on the legends, so there are a bunch of arbitrary abstract relations to learn one way or another.

There’s nothing logical about any of the use of F keys in existing software. Putting a shortcut on e.g. spade symbol key + Q isn’t inherently any more or less reasonable than putting it on Alt + F4.

On the other hand, every time you have to reach for a key, and especially every time you have to look down at the keyboard to figure out which key to press, you end up causing a massive human performance hit.

Quote
Can you not accept that extra keys benefit both novice and experienced users?
No, I can’t. Speaking from direct personal experience using keyboards and watching many other people use keyboards, F keys are basically terrible for all levels. There’s a reason Apple, Google, and Microsoft demoted the F keys to be the second layer on their Macbook, Chromebook, and Surface type cover keyboards, and put functions like adjusting screen brightness and media controls on those keys instead: the F keys suck for casual users, and only a tiny percentage of people ever use them for anything.

Quote
We train novice users on the F keys all the time, and it makes interaction with our line-of-business software easy.
That’s a problem with your line-of-business software, but doesn’t really have too much bearing on keyboard enthusiasts’ new designs. If someone had a keyboard without physical F keys where to send the same key event to the computer used a layer, it would not be especially easier or harder to train them. (Frankly, someone with a custom keyboard who carefully designed their own layout can probably just train themselves on your keyboard shortcuts, if you hand them a piece of paper with the shortcuts listed on it. I doubt such people make up any appreciable fraction of your customers.)

If you want keyboard shortcut design feedback, you could start a different thread, and I’m sure several geekhackers would be willing to offer advice.

Quote
How does adding additional keys make a huge efficiency hit, when instead of one key press, I'm now forced to make at least two?
Because every time you need to move your whole hand off the home row, you need to do 2 separate re-orientations, each of which causes a time delay and interrupts your train of thought. I recommend you actually try timing this in a real-world experiment. Time your current keystroke in context the middle of standard typing, then set up a key binding which can be easily reached without any hand movement or contortion, give yourself a few days or a week to learn it, and then time yourself again. I promise you’ll see an improvement.

Examples of keys which cause slowdowns all the time on a standard keyboard: delete, escape, arrows, the “navigation cluster”, various multi-modifier shortcuts, F keys.

If you redesign the keyboard to move these functions to be within reach (as an easy example, by moving the delete/backspace key to be one half of a split spacebar) it makes a dramatic improvement.

Similarly, putting some kind of pointing device directly adjacent to the home hand position makes a big improvement when switching between mousing and typing. IBM did internal experiments showing that each time a pointing stick is uses instead of a separate mouse, it saves seconds of time.

If you’re willing to put some time and effort in, I highly recommend building or buying a column-staggered keyboard with sufficient thumb keys such as an Ergodox, keyboard.io, Kinesis Advantage, or similar, and then re-locating numbers onto a layer (e.g. in a numpad-like arrangement) near the home row instead of using the usual number row. After some practice, you’ll see very noticeable improvements in fluency typing mixed text and numbers compared to using the top row or using a separate numpad. Likewise, if you’re a programmer, I strongly recommend moving common symbols (parenthesis, arithmetic operators, etc.) to layers and positioning them in easy reach. If you commonly use F keys, put those on a layer, perhaps under the same keys as the corresponding numbers to make it easy to remember which is which.

If you set things up carefully, you’ll end up with an overall much more fluent experience. You’ll improve your speed, reduce your error rate, reduce the number of interruptions to your train of thought, keep yourself more comfortable and stave off RSI, and so on. If you give yourself as an end user the ability to add more shortcuts whenever you need them (ideally via real-time macro recording, but it could even be in some offline config step), you’ll give yourself a platform on which you can keep improving constantly, instead of hitting the low skill ceiling of standard keyboards.

Quote
How does adding keys to the top or bottom of a keyboard make worse ergonomics for pointing devices?
That was with reference to the “piles of unnecessary keys” on a standard keyboard. That includes the arrow keys, the forwards delete/home/end/pgup/pgdn keys, and the number pad. F keys don’t have much impact on external pointing devices.

Quote
Your reasons for opposing extra keys are dogmatic.
I’m just telling you how to make the best general-purpose computer keyboard for a “power user” who is willing to invest time and effort, based on my own extensive experience and research; I’m not commenting at all on how to make the best selling product, how to build the best point of sale keyboard or flight simulator input device or stenography keyboard, or how to make life easy for someone who wants to stick to what they’re used to.

Anyhow, type with whatever kind of input device you want; no skin off my back.
« Last Edit: Thu, 23 July 2015, 01:15:47 by jacobolus »

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to full 88 keys – my take
« Reply #16 on: Thu, 23 July 2015, 01:00:07 »
The ergodox force to too much change ?
Have you tried to print your layout to test it ?
I am not talking about the position of the keys. That part about the ErgoDox is mostly fine. My point was mainly about the number of keys, which I thought would be obvious from the title...

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #17 on: Thu, 23 July 2015, 02:06:29 »
jamadagni: By the way, sorry if we’re slightly de-railing your thread. There’s nothing illegitimate about wanting to have a keyboard with F keys and arrows, even if it doesn’t match my own preferences. :-)

Offline nick2253

  • Posts: 7
Re: ErgoDox extended to full 88 keys – my take
« Reply #18 on: Thu, 23 July 2015, 09:57:01 »
No, I can’t. Speaking from direct personal experience using keyboards and watching many other people use keyboards, F keys are basically terrible for all levels. There’s a reason Apple, Google, and Microsoft demoted the F keys to be the second layer on their Macbook, Chromebook, and Surface type cover keyboards, and put functions like adjusting screen brightness and media controls on those keys instead: the F keys suck for casual users, and only a tiny percentage of people ever use them for anything.

I've demonstrated the benefits of the F keys.  The vendors you cite have demonstrated the benefits of the F keys.  If you get rid of those keys entirely, then you take away that benefit.

If you can't accept the preponderance of evidence that extra keys add value for users, then I don't see why you're even here in this thread.  The entire premise is to expand the ErgoDox to make it *more* useful, and here you are arguing against that with anecdotes and platitudes.  You seem to think that people use keyboards in exactly one way with a very particular workflow, and that just clearly isn't the case.

Offline vvp

  • Posts: 780
Re: ErgoDox extended to full 88 keys – my take
« Reply #19 on: Thu, 23 July 2015, 15:43:35 »
You do not understand each other.

@nick2253: jacobolus is talking about "power" users who touch type and do not ever look at the keyboard, those users do not want to move hands from home row since that slows them down, so they opt for chording instead

@jacobolus: nick2253 is talking about user who hunt and peck keys, they do not need to worry about loosing home row since they do not have any, pressing an F-key is probably easier for them than chord (one less key to hunt for)

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #20 on: Thu, 23 July 2015, 20:34:59 »
I've demonstrated the benefits of the F keys.
I think you aren’t reading me or are misunderstanding. You really haven’t demonstrated anything. To “demonstrate”, you need to actually run the experiment both ways, and compare the outcomes. In all of my tests (though admittedly they are mostly small sample sizes and heavily skewed toward observation rather than large-scale data collection), the F keys are empirically less efficient, more error prone, and less comfortable. If I had a few million dollars to set up a real research group, I could prove this in a more satisfying way. If you know any very rich people who want to advance the understanding of keyboard ergonomics and will write me a check, let me know.

All you’ve done though is state your own personal preferences, without any evidence in support, and without even any convincing reasoning.

Quote
If you can't accept the preponderance of evidence that extra keys add value for users,
The existence of a design feature in some human creation is very very weak evidence of “added value” w/r/t modern needs (actually not just human creations; see vestigiality). Most of the design features of keyboards were created by a tiny handful of people (e.g. Sholes 150 years ago who made the original QWERTY layout, or various engineers inside IBM, or DEC, or wherever in the 60s–80s, who were operating under severe pressure to ship features relevant to very narrow specific criteria which were applicable for the specific device they were shipping but are no longer relevant today – for example keyboards today have more CPU power than supercomputers of that era), then replicated across millions of devices, and never subjected to any kind of scrutiny.

Application developers and even operating system developers only barely ever thought about the design of the keyboard per se, instead using its available features as handed to them, and basing most of their keyboard shortcut design decisions on past systems people were already familiar with, but rarely if ever trying to think their designs through logically from first principles.

There are many aspects of our society with similar path dependency and lock-in effects. Decisions lose their context, and 2000 years later a large group of people refuses to eat some particular type of animal because their holy book says so. Etc. Probably had at least some reason at the time. Today, it’s just a cargo cult.
« Last Edit: Fri, 24 July 2015, 03:56:46 by jacobolus »

Offline Zustiur

  • Posts: 198
Re: ErgoDox extended to full 88 keys – my take
« Reply #21 on: Fri, 24 July 2015, 00:07:04 »
As I was reading the original post, I thought 'I better mention that I'm working on an ergodox variation with f keys'. Glad to see you already found it. I should get the time to wire up the prototype soon.

For what it's worth; I use F keys at work a lot, and I'm a very proficient typist, even doing symbols and F keys without looking down. Yes each time it means moving my hand, but I'm not convinced that doing so is a truly bad thing.

The argument against use of  F keys appears to assume that those keystrokes are occurring in the middle of large bodies of text. I can't speak for anyone else but that is not the case for me.
 I'll explain better when I'm not posting from my phone.

Offline vvp

  • Posts: 780
Re: ErgoDox extended to full 88 keys – my take
« Reply #22 on: Fri, 24 July 2015, 04:00:34 »
For what it's worth; I use F keys at work a lot, and I'm a very proficient typist, even doing symbols and F keys without looking down. Yes each time it means moving my hand, but I'm not convinced that doing so is a truly bad thing.
Well, it costs you more time an energy. And I would say it is harder to learn than learning some chords. Whether it is bad or good is a personal preference.
You are actually only a very weak example of love for extra keys (as opposed to chording). Check out this post by kbdfr (tiproman):
http://deskthority.net/keyboards-f2/how-many-rows-should-a-keyboard-have-t11096.html#p242298
He is using a 247 key keyboard still arguing that it is better to have special keys instead of chords for rare actions.

Edit: The only way I imagine how pressing an F-key can cost you less time or energy than a chord is if you typically press them when your hands are not already on the keyboard (maybe because you are using a mouse etc).
« Last Edit: Fri, 24 July 2015, 04:11:02 by vvp »

Offline Zustiur

  • Posts: 198
Re: ErgoDox extended to full 88 keys – my take
« Reply #23 on: Fri, 24 July 2015, 04:31:15 »
Quote from: vvp
Well, it costs you more time an energy.
Yep. I totally does. When it's a single keystroke we're talking about. But when you have to hit a series of F keys?...

To continue what I was saying (not to disagree, simply to explain):

Many of the times that I need to use an F key, I'm not actually doing anything else with the keyboard, so I'd be moving my hand anyway. For example, if I'm debugging some code, F5 starts the code, F9 sets a break point and F8 steps through line by line. If I'm doing that I'll tap F9 once, then F5, then a combination of F8 and F5 depending on what I'm testing and how the code is behaving. At no point while doing this task do I need to type on any other keys. Clearly in this situation, I'm better served with dedicated keys I can rest my fingers on and tap as I watch the screen. Cording with Fn+5 and Fn+8 would be an unnecessary burden here.

On my ErgoDox, the Fn key may not be terribly well placed. It's the middle key of the 5 in the bottom row on the left. Fn+5 is a difficult hand flex. Reaching up and tapping a dedicated F5 would take less energy and be easier on my hand.

The other one I use a lot is F2, to enter a cell in Excel or Access or to rename a file/folder. Whether that is better with or without a Fn layer remains to be seen, and is one of the things that I will be deliberately testing when I get my prototype up and running. I suspect for that task Fn+2 will be better than the dedicated F2, but we'll wait and see. Certainly having Fn+neiu (Colemak right hand) has become my favoured arrow key system (though I'm not convinced that it's better all of the time). However... work constraints may cause problems there too. I noticed at work today, in many situations, when my left hand goes to the mouse, my right goes automatically to the nav cluster. Bad application design 'For The Lose'. Note that application design is rarely in the hands of those who have to use the application.

My other argument for keeping dedicated keys and not going too minimalist is this:
Variety of movement is good for you. With the 'perfect' minimalist keyboard your hands never have to leave home row... which is both good and bad. Good because frequent contortions caused by 104 key qwerty boards are bad for you, but bad because staying perfectly still all day is also bad for you.

So my stance is; reduction of hand movements, yes. Total elimination, no. Ergomonics, as we all hopefully realize by now, is extremely person and task dependent.

While writing this my brain was ticking over. With the way I rest my palms and place fingers on the home row, there's some dead space my fingers can reach below the Ctrl row. I think having the alpha keys one row up and the number shifted to the Z row position might actually be a better design... For me, it would then be possible to hit all rows without having to lift/move my hand. I'll experiment with that at some point.

Offline vvp

  • Posts: 780
Re: ErgoDox extended to full 88 keys – my take
« Reply #24 on: Fri, 24 July 2015, 04:59:08 »
Zustiur:
Yes, some movement outside typing is good. Though I prefer when keyboard is not forcing me to move.
Actually I do not mind additional keys on the keyboards, provided that the keyboard firmware allows me to remap/macro them to more comfortable positions/chords.

Try to place all the modifiers (Crtl, Alt, Win, Shift) and also the LayerShift key on the thumb clusters. I like it quite a lot. Maybe you will like it too.

Post a picture of your keyboard when you are done. I was not satisfied with ergodox/kinesis myself and rather made my own version:
http://deskthority.net/workshop-f7/katy-keyboard-or-k80cs-key80-contoured-split-t8524-60.html#p222016

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to full 88 keys – my take
« Reply #25 on: Fri, 24 July 2015, 12:02:50 »
... On the plus side, not only do you have more keys for programming, macros, usability, etc, there's also that the marginal cost of more switches and keycaps could be more than offset by the possible volume savings we would benefit from if an ergo mechanical keyboard went mainstream.

Ahh you put my dream into words! :-)

jamadagni: By the way, sorry if we’re slightly de-railing your thread.

@jacobolus: No issues at all. I'm happy to have (albeit indirectly) initiated such avid discussion, which after all is the point of this great forum! I can see you feel passionately about the issue and admire it. I also think that actually your and nick2253's arguments don't conflict with each other since you both make the case for different keyboard structures for different kinds of users...

> the thumb cluster is not radial
You should build some prototypes and try them. I built about 20 different versions out of acrylic (mostly not wired up), and many more paper printouts and incremental clay experiments, and my conclusion was that “radial” thumb keys are overrated. In particular, once they swivel a thumb far enough down, many people find it difficult/uncomfortable to press a key while still maintaining flexibility in the other fingers.

What exactly do you mean by "far enough down"? You are talking about a TC that's coplanar with the rest of the keys, no? The way I figure it, the thumbs should normally rest somewhere above (in 3D space) and between the Space/Enter/Bksp/Del keys and it's only for Shift that they would have to slightly angle out...

Quote from: jacobolus
If you do like radial thumb keys though, you might consider pre-ordering a keyboard.io. They have a pretty solid design, which they have put a lot of thought and work into.

Their base cost is a bit too high and the butterfly shape is too non-sober for my taste. If there were an option to avoid the costly maple woodwork and the butterfly, it would be better. But even then, the non-standard shaped keys mean I can't replace them from elsewhere. As it is, ErgoDox EZ presents a much better option on many fronts: soberness, standard keys, lesser base cost, (tilt and tent kit,) tried and tested multiple firmware options...

Quote from: jacobolus
> what is your problem with the thumb cluster I propose?
I think you should build it and try it out! It’ll be a fun exercise, and you’ll learn a lot about how your hand works. Even if you just print your design out on paper, you can get some idea of how it would work.
You're right. I've been busy with other things. I should at least prototype on paper and not just theorize.

Quote
For anyone from small to medium-large hands, only about 1 of your proposed thumb keys is in a reachable spot, and none of the keys would be fun to use as a spacebar. For someone with large hands, more of the keys might be in reach, but they’re still not in very practical spots, IMO.

I have large hands, but I did say I suspect the keys are too high up away from their centre of symmetry. As I also said, I only intend for the bottom three keys to be accessible to the thumb and the upper two are just for chording by the index/middle fingers with the wrist slightly turned inwards. I don't think the upper row TC keys are accessible directly by the thumb in any design and certainly one could not do multiple chords (more than one of Ctrl/Alt/Shift) with them...

Quote from: jacobolus
There’s a reason Apple, Google, and Microsoft demoted the F keys to be the second layer on their Macbook, Chromebook, and Surface type cover keyboards, and put functions like adjusting screen brightness and media controls on those keys instead: the F keys suck for casual users, and only a tiny percentage of people ever use them for anything.

So whether you use it as generic F keys or you use it as multimedia/specific function keys, the point is, the extra keys are useful for ordinary people whereas a power user like you could certainly do (or even be more efficient) with less.

@Zustiur: Nice point about lesser used keys making for change for the hand from its normal pattern to break over-repetition.

Offline nomaded

  • Posts: 197
  • Location: Andover, MA
Re: ErgoDox extended to full 88 keys – my take
« Reply #26 on: Fri, 24 July 2015, 12:57:31 »
On my ErgoDox, the Fn key may not be terribly well placed. It's the middle key of the 5 in the bottom row on the left. Fn+5 is a difficult hand flex. Reaching up and tapping a dedicated F5 would take less energy and be easier on my hand.

With a bit of trial and error, and since I knew I wanted to move most of the symbols out of the number row so they're easier to hit, I had always planned on heavily using a "symbol" layer. Because of this, on the ErgoDox, I set the symbol layer toggle to be the closest 2u key in the thumb cluster. I also have the F keys on the symbol layer, mapped to the number row (with F11 and F12 directly below F1 and F2, on the Q and W keys, in Qwerty). I find the F keys to be very easy to hit while touchtyping, for example by holding the layer key with my left thumb, I can reach up to hit F2 with my left ring finger. I have this symbol layer toggle on both sides of the ErgoDox.

Quote
While writing this my brain was ticking over. With the way I rest my palms and place fingers on the home row, there's some dead space my fingers can reach below the Ctrl row. I think having the alpha keys one row up and the number shifted to the Z row position might actually be a better design... For me, it would then be possible to hit all rows without having to lift/move my hand. I'll experiment with that at some point.

Huh. This is an interesting idea. This layout should be pretty easy to get used to, especially with DSA, or all Row3 SA keycaps. It might actually make hitting the all the keys in the thumb cluster easier. I may need to give it a try.
Dvorak
ErgoDox fullhand (MX Clears) w/Nuclear Green Data SA || Infinity ErgoDox (Zealios 78g tactile) w/SA Retro || Atreus62 (MX Clears) w/Chocolatier || TECK 209 (MX Browns) || TouchStream ST
Kensington Slimblade Trackball || Logitech Cordless Optical Trackman || Apple Magic Trackpad
Current Dvorak-based ErgoDox layout || Current Dvorak-based TECK layout

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #27 on: Fri, 24 July 2015, 12:58:34 »
So whether you use it as generic F keys or you use it as multimedia/specific function keys, the point is, the extra keys are useful for ordinary people whereas a power user like you could certainly do (or even be more efficient) with less.
Not really. More like, they needed to include the F keys for backwards compatibility reasons, so they tried to come up with something to do with them that wasn’t completely useless.

Keyboard design for a mass-market computer (or especially for a mass-market laptop) is really hard. You basically can’t change anything by more than tiny nudges, even if the standard design is terrible, or people will start complaining.

The way keyboards and software have interacted over time is: keyboard gets some new buttons stuck on it, then software authors start using those buttons for something, locking them in for future hardware. Repeat for a few decades up through the mid-1980s, and you end up with a keyboard with a huge number of buttons, enough so that it’s no longer practical to add any new ones, but the ones that are there can’t be removed or moved around without causing complaints.

There’s almost no incentive for a computer vendor to design a better keyboard from scratch, because it would lose all the training and network effects of the standard design.

This is the lucky thing about working on keyboards in the 2010s: we now can easily build firmware that gives the end user full control over the behavior of the keyboard. Someone selling a mass-market keyboard is going to fix the behavior of the keyboard up front and thus is stuck on a layout from 1985, but a hobbyist or someone selling to hobbyists can make any shape and include any number of keys, and leave it up to the user to decide what each key should do.
« Last Edit: Fri, 24 July 2015, 13:00:07 by jacobolus »

Online Findecanor

  • Posts: 4233
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: ErgoDox extended to full 88 keys – my take
« Reply #28 on: Fri, 24 July 2015, 13:26:20 »
I think the thumb keys need to be placed closer to the finger keys. One of the most common complaints about the ErgoDox was the distance to the thumb key cluster.

It is going to be interesting to see if the two rows of thumb keys are going to work. Do you have any plans for using keys of different heights?
Smplesh dar Ventöl pööl pööl
Daily driver: Phantom (Lubed Cherry MX Clear, Lasered Cherry PBT keycaps with Row A. Plastic "Frankencase". Custom firmware, Swedish layout)

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to full 88 keys – my take
« Reply #29 on: Fri, 24 July 2015, 21:20:45 »
I think the thumb keys need to be placed closer to the finger keys. One of the most common complaints about the ErgoDox was the distance to the thumb key cluster.
Yes I know – I did try some measurements on my own fingers before the placement but they are longer than normal so I'll have to pull those in.

I also think that the excessive space below meant for the palm to rest on can be cut off for compactness/weight. I'll post a slightly modified version later on.

Quote
It is going to be interesting to see if the two rows of thumb keys are going to work.

As I said, the higher row is only meant for chording keys where the wrist has to turn inside a bit to enable the index/middle fingers to press+hold Ctrl and Alt with the thumb on the Shift in the lower row. I don't think the higher row in the TC is actually accessible by the thumb, which is why I was pointing out over in the Axios thread that the downward TC with 6 keys doesn't really seem to help much (and it probably increases costs including a separate PCB section)...

BTW I am not aware of the Meta/Win/Cmd key being used in combination with Shift/Alt/Ctrl in common software – does anyone have any info on that?
Quote
Do you have any plans for using keys of different heights?
By height you mean profiles right? I have totally no idea about profiles (yet). As I said, my contribution is only regarding the overall positioning of the keys. Jacobolus has done lots of thinking about profiles AFAICS...

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to full 88 keys – my take
« Reply #30 on: Fri, 24 July 2015, 23:18:41 »
BTW I am not aware of the Meta/Win/Cmd key being used in combination with Shift/Alt/Ctrl in common software – does anyone have any info on that?
On a Mac, {Command + Shift + letter} and {Command + Option + letter} combinations are very common. {Control + Shift + letter} are also used in various places. For navigating around text boxes, {Option + Shift + arrow} combinations are used. Many applications even have {Command + Shift + Option + letter} combinations. Some applications I frequently use also have {Command + Control + letter} combinations and even {Command + Option + Control + letter} combinations.

In general, these multiple-modifier shortcuts are quite terrible on a standard keyboard layout, because the modifier keys were not originally placed where they could easily be used together.

One thing I really enjoyed when I tried it for a month several months ago was putting the modifier keys on the top row, instead of numbers, with each finger dedicated to one modifier so it is easy to hold them as a chord, with the idea being to use the opposite-hand modifiers for any {modifier(s) + letter} shortcut. Something along the lines of:


(On this layout, use thumb shifts for routine capital letters, and use top row pinky shifts for multiple-modifier shortcuts.)

Note, this layout has 52 keys overall, and easily outperforms any standard full-sized keyboard in terms of typing speed/fluency, accuracy, and comfort, while still being quite simple and newbie-friendly. In addition, it’s faster for typing numbers than either a dedicated numpad or top row number keys, while making it a breeze to mix numbers and text. This is pretty minimal as layouts with extra thumb keys go, including massive space and delete keys for the benefit of folks with various hand shapes, and keeping all the keys aligned to a grid.

A more power-user focused logical layout could cram far more functionality into about same number of keys, and might have a slightly different shape for the thumb section with an extra key or two and a smaller spacebar/delete.
« Last Edit: Fri, 24 July 2015, 23:52:43 by jacobolus »

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #31 on: Sun, 26 July 2015, 03:19:12 »
I updated the original post with some changes based on feedback and some rethinking. There are now four more keys for a total of 92 with no more footprint cost. I've explained the other changes in the OP.

@jacobolus and others: Is this TC better?

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #32 on: Sun, 26 July 2015, 03:41:25 »
@jacobolus and others: Is this TC better?
You really need to make some prototypes.

Start with a paper print out. Next step it up to keycaps from a sacrificial rubber dome board pressed into modeling clay. Once you’re happy with that, try getting some laser cut acrylic switchplates made and test with real keyswitches. When you’re satisfied enough to try it live, find someone with a soldering iron you can borrow, get some diodes, and wire your prototype together.

Right now about half of your thumb keys are still not in very convenient spots IMO, but maybe our hands have different shapes and they work great for you.

Offline vvp

  • Posts: 780
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #33 on: Sun, 26 July 2015, 09:13:07 »
I would use 1u keycaps for outer pinky columns and top thumb cluster rows. They will be easier to source and they do not need to be wide. It will be easy to hit the column just next to your pinky home column. And you do not need more than 1u keycaps for top thumb cluster rows since if your thumbs are longer then they can overhang above the keycap.

Use taller keycaps for top thumb cluster row (compared to the keycaps in the bottom thumb cluster row). It looks like you still can move your thumb cluster nearer to the keywell a bit. And you can move the inner keywell columns lower (nearer to the thumb cluster).

As you have it now, the most far away key in the top thumb cluster row will be hard to reach (without wrist move). Make a mockup and check it out.

Here is an example of my latest layout:
106572-0
It is contoured but you can compensate for missing Z-dimension by increasing stagger and keycaps. You also dropped the bottom ergodox row which is a good idea for a flat design. You also added navigation clusters and F-row which may be appealing to some people (although it is not appealing to me since they are already too far away from the home position).

Offline conandy

  • Posts: 43
  • Location: Denver, Colorado
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #34 on: Wed, 29 July 2015, 01:21:03 »
Hey, guys.  I followed jamadagni over here from the Axios thread.  The keyboard you propose, jamadagni, is almost exactly like the one I was designing on paper if I were to make my own.  I am also in favor of the F-keys, both from habit and the fact that CAD software tends to use them a lot.  I am sure I could learn to use modifiers and chords to overcome the loss of F-keys, but having them doesn’t hurt anything.  If anything, you can learn to chord the easy-reach keys (as has been beaten to death by other folks in this thread) and just turn the F-keys into media controls, etc.  Having the buttons doesn’t hurt anything, as long as they put to good use. 

I do agree, though, that using the F-keys is a hindrance to touch typing and it does slow one down drastically.  I have to look down at the keyboard to use them.  They are placed horribly on my TEK (my stopgap keyboard…). 

In my experimenting, I printed several layouts on paper, and used sticky tack to stick keycaps onto the sheet to get a feel of the layout.  I found that I could really only easily make use of 3 thumb keys, arranged radially like you have proposed.  I did find, too, that the radial layout isn’t quite as important in actual use as it seems from a logical standpoint regarding thumb movement.  Having large enough keys located in the right locations was important though.  For the record, I have small hands and tendon damage in my thumbs, so long stretches with my thumbs is uncomfortable and painfull.

I got to go to the show-and-tell here in Denver for the keyboardio a couple weeks ago.  I was once again assured that more than 3 thumb keys is probably too much for me.  They are using four 1x keys arranged radially.   My thumbs didn’t like having to be held into such a tight pattern, and it was very easy for me to get my thumb mis-placed.  A big “space” or “backspace” on each hand, plus a modifier to reach for in either direction seems ideal.  But that is just me.

I think keyboardio is really onto something with the palm keys, though.  Although the prototype keyboardio had the palm key poorly placed, I felt I could really take advantage of it if it were placed correctly, had the right profile height, and had enough resistance to avoid nuisance pressing.  I was dubious before testing it out.  Now I think I am sold on the idea.

All that said, I think jacobolus is onto something with his 52 key proposed board.  That layout intrigues the heck out of me. 

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #35 on: Sat, 01 August 2015, 03:48:46 »
Hey, guys.  I followed jamadagni over here from the Axios thread.  The keyboard you propose, jamadagni, is almost exactly like the one I was designing on paper if I were to make my own.

Glad to hear! :-)

Quote
I do agree, though, that using the F-keys is a hindrance to touch typing and it does slow one down drastically.  I have to look down at the keyboard to use them.
Hmm, but F keys don't really occur in the flow of extended text input. I mean you don't really do "hello my name is<F4> what's yours" unless you've macro-mapped F4 to your name, in which case you could well map Macro + a home key for that purpose.

Re having to look down at them – I find that on a non-ergonomic regular keyboard (or laptop keyboard) most of the time I don't need to do it, but sometimes I do falter. Obviously this is a point where people's experience seriously varies.

Do you think having them adjacently above the number row would be a problem for locating the individual keys?

Jacobolus does have a point above where he proposes that it might be easier if they are placed in two-s, as he originally proposed over here.

However I suspect that even without such a radical reorientation/repositioning, if they are staggered sufficiently and moved away from the number row, it would be quite sufficient:
107086-0

Quote
They are placed horribly on my TEK (my stopgap keyboard…). 
Is it the fact that they are arranged linearly along the top of the board whereas the main keys are laid out separately at an angle convenient to the hands? Golly I never thought of that when I was considering buying the TIKE... That does seem to be a blooper...

Quote
I think keyboardio is really onto something with the palm keys, though.  Although the prototype keyboardio had the palm key poorly placed, I felt I could really take advantage of it if it were placed correctly, had the right profile height, and had enough resistance to avoid nuisance pressing.  I was dubious before testing it out.  Now I think I am sold on the idea.
Hmm... With such a palm key, even the ErgoDox might actually be usable...

Quote
All that said, I think jacobolus is onto something with his 52 key proposed board.  That layout intrigues the heck out of me.
You mean the one with modifiers up on top. Yep, most of his layouts do. He has put a lot of thought into them. Just, (and to get back to the thread's raison d'être,) such radical and minimalistic designs suit neither all scripts (specifically other than Latin) nor all kinds of users...
« Last Edit: Sat, 01 August 2015, 07:06:44 by jamadagni »

Offline conandy

  • Posts: 43
  • Location: Denver, Colorado
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #36 on: Sun, 02 August 2015, 00:41:54 »
Quote

Do you think having them adjacently above the number row would be a problem for locating the individual keys?

Jacobolus does have a point above where he proposes that it might be easier if they are placed in two-s, as he originally proposed over here.

However I suspect that even without such a radical reorientation/repositioning, if they are staggered sufficiently and moved away from the number row, it would be quite sufficient

My concept matches your OP, except I'd keep the F-keys directly above the corresponding numbered keys, and then put F11 and F12 above the extra index finger keys in the middle.  My thought was that that was the easiest to learn to incorporate F-keys into touch typing, or at least something close to touch typing them.   To be honest, I seriously considered putting all 12 F-keys on the left hand side, in two rows of 6, simply because my ideal world would be one where I don't have to remove my right hand from the mouse during normal CAD software use.  That is radical, I know.   I honestly don't know why the clusters of 2 would be better, or worse.  I just don't think it would lend itself as easily to touch typing. 

Quote
Is it the fact that they are arranged linearly along the top of the board whereas the main keys are laid out separately at an angle convenient to the hands? Golly I never thought of that when I was considering buying the TIKE... That does seem to be a blooper...

I don't know if it's a blooper, or that they just assumed that touch typing the F-keys isn't practical, so why make it a logical extension of the key layout?  The TEK in so many ways ended up being a half-assed attempt at a good keyboard.  I like the idea of more keys for the index finger, and fewer for the pinky, but they completely dropped the ball on thumb keys.  I could live with the bad F-key layout if they had just given me a couple of thumb modifier keys.  I say that without ever having actually USED thumb clusters, beyond the shared center Enter key on the TEK, which, by the way, requires a bit more of a stretch to hit with either thumb than I like.

Offline Zustiur

  • Posts: 198
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #37 on: Sun, 02 August 2015, 01:41:03 »
Hmm. On the topic of F keys. I didn't realize until using my prototype that I rely heavily on the Fkeys being in clusters of 4 so that I can find the right one without looking. You wouldn't believe how many times I hit F3 when I wanted F2. I eventually gave up and looked down to realize what the problem was. I'm bound to get used to it, but it's given me cause to reassess how I arrange those keys.

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #38 on: Sun, 02 August 2015, 01:42:03 »
To be honest, I seriously considered putting all 12 F-keys on the left hand side, in two rows of 6, simply because my ideal world would be one where I don't have to remove my right hand from the mouse during normal CAD software use.  That is radical, I know.
Sounds reasonable. Better though would be to design a layout specifically for CAD use, where you move the functions you need to all be within reach of your left hand without any significant required hand movement. If you have a few thumb keys, you can add several layers.

If you need to relocate your hand by moving your whole forearm to hit an F key, as happens on a standard keyboard or the type you’re proposing, then that causes two noticeable pauses, first while you locate the F key, and second when you need to reorient your hand in “home” position. These not only take time, but also slightly interrupt your train of thought. The same is true for the standard keyboard locations of backspace, escape, and several number-row keys. Backspace is the worst offender though, as making mistakes already causes an interruption as it is. Moving backspace to be easy to reach makes a huge improvement.

If you have full control over the keyboard firmware, you can make a multi-layered CAD-specific keyboard layout which only needs ~25–30 discrete physical keys on your left hand, and gives you whatever kind of logical organization you want, making it efficient and comfortable to reach all the functions you need.

Quote
I honestly don't know why the clusters of 2 would be better, or worse.  I just don't think it would lend itself as easily to touch typing.
Grouping keys makes it easier to find keys by touch alone, and gives a buffer around keys you need to reach for, reducing mistakes. If far-away keys you need to reach for are all right next to each other, then when you try to press one quickly you are more likely to miss and press the wrong key.

Quote
I don't know if it's a blooper, or that they just assumed that touch typing the F-keys isn't practical,
It isn’t practical to “touch type” the F keys at the same speed as letter or number keys in any keyboard arrangement I’ve ever seen, unless they’re placed near the home row on a layer.
« Last Edit: Sun, 02 August 2015, 15:58:08 by jacobolus »

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #39 on: Sun, 02 August 2015, 06:44:38 »
@conandy: Re the F11 and F12, I get your point, but as I said in my OP, I didn't want things to be more surprising for ordinary users than they have to be, especially since people are used (?) to have F6 on the left side of the keyboard and certainly F11 and F12 on the right end, which has to be relearnt if F11 and F12 were repositioned... If it were anything like a real bad habit or bad design, like Jacobolus' point about Backspace, then we could reorder, but somehow this doesn't seem like that.

For ErgoDox-type boards without a dedicated F row, it would certainly be meaningful to map the number keys to the F keys one to one, but when we have dedicated F keys, is that so important? For instance, with such F keys on the common keyboard, I don't think people really have any mental association of them with the number keys...

Of course, any such keyboard design if actually produced for consumption [hopefully! :)] would be programmable and you could relocate as per your wish...

@jacobolus: Great point about CAD layer. Layers FTW!

Offline conandy

  • Posts: 43
  • Location: Denver, Colorado
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #40 on: Sun, 09 August 2015, 17:24:48 »
You all may be slowly convincing me to rethink my love of the F-keys.  Or at least my thoughts about not putting them on function layers and chording them.  And a CAD specific layout makes a lot of sense.

Offline Zustiur

  • Posts: 198
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #41 on: Sat, 15 August 2015, 18:32:00 »
One of the key things to remember about using layers is that sometimes you need to use chorded keys on a regular keyboard too. This means that when you put numbers, symbols or F# keys on a layer, you'll sometimes have to press more than two keys at a time. It is therefore necessary to do a bit of mental double checking to ensure you haven't made a necessary combination impossible to press.
For example, in when doing documentation in MS Word, I make a lot of use of Shift+F3 as this cycles a word through lower case, First Letter Caps and FULL CAPS without having to retype that word. On my current ErgoDox layout, that's a bit difficult to do as it's Shift+Fn+3 which is really awkward because Fn is the middle key on the bottom row. A worse (but unlikely) example would be - if you put Tab on a secondary layer. Ctrl+Shift+Tab is a common chord, as is Alt+Shift+Tab. Adding Fn to that mix could be a real problem.

In short what I'm saying is, check that you can still press all of your lovely layer keys while also holding Ctrl, Alt, Win and/or shift. This is where it really helps to have modifiers on both sides of the board.

Offline vvp

  • Posts: 780
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #42 on: Sun, 16 August 2015, 05:42:27 »
Just put all the modifiers (and also the LayerShifts) to the thumb clusters and make them symmetric (so that all modifiers are available on both sides).
Other options are thumb clusters optimized for pressing multiple keys at once (oobly) or easy support for on-the-fly macros and you can trigger more complicated chords using simple ones.

Offline user888

  • Posts: 42
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #43 on: Thu, 20 August 2015, 03:28:03 »
Without reading the complete discussion in this thread, I really like your design as I think it fixes the thumbcluster of the Ergodox:


Offline conandy

  • Posts: 43
  • Location: Denver, Colorado
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #44 on: Fri, 21 August 2015, 10:50:40 »
Just noticed something on your proposed layout that I thought I would mention.  I know a lot of folks (self included) have been trying for a more staggered pinky column of keys (speaking purely of the left handc for this discussion, the A-column) because the stagger on Ergodox, TEK, etc. doesn't seem like enough.  When I tried a sample layout with extreme stagger on the A-column (-12.7 mm relative to the S-column), I found that typing the Z key often resulted in my pinky rolling under and me hitting the key with the tip of my fingernail.  So even though a lot of stagger on the pinky column seems to make sense from the perspective or reducing stretch to the 1 key, don't underestimate the negative impact that stagger has on being able to comfortably hit the Z key. 

Maybe more practice and hand training would have alleviated this, but it is something to consider as you play with layouts and key column stagger.

I think the pinky A-column on the ergodox drops about 3.175 mm relative to the S-column (based on the published CAD layouts).  May be closer to 4 mm on the TEK.  I found much more than this to be hard to work with, but would think as much as 6 or 7 mm wouldn't be outrageous.  12mm + though, caused me problems.  I can't tell from your pics, but it looks like your proposed stagger for that column is in the 12 mm range. 

In an ideal world the pinky column keys would actually be smaller keys with tighter spacing.  Hard to do with the keyswitches and caps available, though.

Offline jacobolus

  • Posts: 3639
  • Location: San Francisco, CA
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #45 on: Fri, 21 August 2015, 21:02:21 »
Just noticed something on your proposed layout that I thought I would mention.  I know a lot of folks (self included) have been trying for a more staggered pinky column of keys (speaking purely of the left handc for this discussion, the A-column) because the stagger on Ergodox, TEK, etc. doesn't seem like enough.  When I tried a sample layout with extreme stagger on the A-column (-12.7 mm relative to the S-column), I found that typing the Z key often resulted in my pinky rolling under and me hitting the key with the tip of my fingernail.  So even though a lot of stagger on the pinky column seems to make sense from the perspective or reducing stretch to the 1 key, don't underestimate the negative impact that stagger has on being able to comfortably hit the Z key.
Note that this depends a lot on how you hold your hand relative to the board, which is also affected by the tent angle, amount the two sides are turned inward, how far the keyboard is from your body and you hold your arms, etc.

On a standard keyboard, there’s going to be at least a bit of angle between the axes of the key grid and the axis along the forearm/fingers. For a column-staggered keyboard, this angle can be greatly reduced, but that takes practice, and people’s natural inclination is to somewhat preserve that angle when moving to a column-staggered board. The more the hand is aligned with the keyboard grid axes, the more stagger there should be between ring finger and pinky columns.

Additionally, if the height of the keycaps is the same across fingers like on an Ergodox (as compared to e.g. the Maltron where the pinky keys are higher), the hands will be rotated slightly about the forearm axis (“supinated”) compared to the angle of the keyboard. On a column staggered board not quite as much as on a standard row-staggered keyboard probably, but still at least a bit. This has the unfortunate side effect of forcing the index finger to have a somewhat flexed first joint when reaching to the side or especially diagonally, which makes pressing those keys weaker and more tiring overall. If the wrist is pronated an amount such that it is aligned with the keyboard, then with uniform height keys the pinky finger faces the same problem. So unless there’s some height difference between columns, there’s always a bit of a trade-off here. In general it’s probably advantageous on most keyboards for people to privilege the pinky by leaving their wrist turned slightly relative to the keyboard, even though it makes the strike direction slightly misaligned with the switch axis for every keystroke and reduces index finger flexibility/strength.

Quote
I think the pinky A-column on the ergodox drops about 3.175 mm relative to the S-column (based on the published CAD layouts).  May be closer to 4 mm on the TEK.  I found much more than this to be hard to work with, but would think as much as 6 or 7 mm wouldn't be outrageous.  12mm + though, caused me problems.  I can't tell from your pics, but it looks like your proposed stagger for that column is in the 12 mm range.
[Side note: it’s a bit easier to follow IMO if you stick to multiples of 19.05mm = 3/4 inches = 1 "keyboard unit".]

For myself, I’ve found that I like a stagger of 3/4u = 9/16" = 14.3 mm between ring finger and pinky column pretty well, but several other people I tested found that that was too much. For designs aimed at a broader audience, I settled on 1/2u = 3/8" = 9.5mm instead, and most people I tested seemed to find that to be pretty effective. 1/2u is definitely an improvement over the 1/4u or less that keyboards like the Ergodox use, and it allows your hand to stay significantly more aligned with the keyboard than it can be on the Ergodox. (On an Ergodox, you have to hold your hands in a quite similar manner to a standard keyboard, and it’s hard to benefit from as much tenting and turn as you can with more aggressively staggered keyboards.)

I also found that there was a negative effect, if the hands are roughly aligned with the keyboard, of having any stagger at all between two columns of index finger keys. If you see a design with stagger between the two main index finger columns that’s a strong indication that the position of the hand is supposed to be at an angle to the direction of the columns on that keyboard.

Quote
In an ideal world the pinky column keys would actually be smaller keys with tighter spacing.  Hard to do with the keyswitches and caps available, though.
Agreed. In an ideal world all of the keys would be somewhat smaller, with tighter spacing especially in the direction along the columns, and perhaps not all the same size, so the columns could be slightly splayed to align with the directions of finger extension. The keys would be at different heights for different fingers (slightly lower for the middle finger, noticeably raised for the pinky), and there would be a substantial height step to further-out rows, to make them as easy to reach as possible.

In an ideal world the pinkies would each be responsible for about 4–6 keys, the middle and ring fingers responsible for 3–4 keys each, the index finger responsible for about 5–8, and the thumb responsible for about 4–7. In total the keyboard would have 40–60 keys overall, and be shaped to very comfortably fit a hand in relaxed resting position.

Offline conandy

  • Posts: 43
  • Location: Denver, Colorado
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #46 on: Fri, 21 August 2015, 23:16:12 »
how much stagger (in keyboard units, or mm) are you showing in this layout between middle finger column and the ring and index finger columns? 

Offline Architect

  • Posts: 254
  • TECK Lover
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #47 on: Sun, 23 August 2015, 08:49:31 »
FWIW I think the design a smart idea. My main complaint about the dox has been the lack of a full set of keys (primarily Function), never heard a good reason why they were dropped.
TECK 209 Blank Keys; Leopold Number Pad; X-Keys Professional; X-Keys 84.

Offline jamadagni

  • Thread Starter
  • Posts: 54
  • Location: 11 N 78 E
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #48 on: Wed, 26 August 2015, 23:03:31 »
how much stagger (in keyboard units, or mm) are you showing in this layout between middle finger column and the ring and index finger columns?
This is what I did, and IIRC I based it on the Axios:

Keep S, F, J, L columns immobile;
move D and K columns: ¼ key up;
move all columns between F and J: ¼ key down;
move all columns outside S and L: ¾ key down.

Offline jayvee

  • Posts: 5
  • Location: Nashik India
Re: ErgoDox extended to 88+ (full set of) keys – “Lekhani”
« Reply #49 on: Tue, 09 October 2018, 06:42:18 »
Hi @jamadagni, I am wondering how far did you reach with your design. Your design looks interesting.