Split ergonomic keyboard project

[justcallmecrash]

I'm pretty sure it's just for the PCBs... BYOeverythingelse. To do cases would add 6-8 more months or more to the project... I'm sure there will be an aluminum case GB at some point.

[bisl]

I'm sure there will be an aluminum case GB at some point.

Oof. I certainly hope it's somewhat soon after the PCB buy. If people were going to buy their own cases, are the specs available from the ones already made (for submission to shapeways or wherever)?

I'm new to the community (and indeed it seems as if I arrived right at the end of this project) but I'd really like to get my hands on one of these; I was right about to buy a Kinesis but I held off (if temporarily) because this seemed way better.

[OrangeJewce]

The Original Post on GH explicitly states the following will be provided in the GB:

"Full CNC anodized aluminium case and plate, [and] PCB."

I think that the organizers want to maintain one large order for the PCBs instead of two smaller ones so that everybody benefits from a cost reduction. I furthermore implore people to stop perpetuation the rumor there will be a PCB GB prior to one which includes cases.

Case prototype has already been produced so no, it will not take 6-8 months for the case, more like however long for manufacturing. This project has already been going on for close to a year now, I don't think there has been any announced release date as of yet, and I have not heard anything regarding production dates as this impacts my worfklow for the UI part of this project. DIY keyboards notoriously take a lot of time. Us on the software team still have a lot of work to do so that you guys can easily program the boards to take whatever layout your heart desires.

Cheers,

EDIT: added source material and terseness.
EDIT2: Fixed a mistake.

[justcallmecrash]

This feels so close I'd hate to add another 6 months if cases had to be included.

I'd rather have kit without case and have a group buy for cases after. Or the layered plate option with the kit.

Sorry, I was going on the information above because I thought it was from someone inside the dev group who had it on good authority that it would add that kind of time to the project.

[OrangeJewce]

As per page 3, you can see a completed prototype of the case. I am not sure what the progress is exactly on the aluminum version, but given that diagrams have been made, I can only presume that, it won't take too long.

No worries, there hasn't been a lot of information coming out from us really, and since dox is the one who has all the cards with respect to the hardware, I cannot say for sure when the GB should officially go live.

Cheers,

[possum]

The PCB files will be published, but at least I think that it should wait until the first round of a group buy has been completed. I don't think it is in anyone's interest to have people getting impatient starting small runs here and there by themselves. The latest design isn't even tested yet.

Or to have them copied and already in the shops before it's finished...

Ok cool, i'd definitely be in for a pcb buy then. Who wouldn't want one of these?

EDIT: just saw the clarifications from OrangeJewce re group buys. will shut it and wait.

[bisl]

"Full CNC anodized aluminium case and plate, [and] PCB."

Thanks!

EDIT: added source material and terseness.

Sorry!

[dirge]

This feels so close I'd hate to add another 6 months if cases had to be included.

I'd rather have kit without case and have a group buy for cases after. Or the layered plate option with the kit.

Sorry, I was going on the information above because I thought it was from someone inside the dev group who had it on good authority that it would add that kind of time to the project.

Flattered, no I'm just an interested forum member. Dox and biphany/prinsvallium are the main guys on this.

[dirge]

Antone suggested accommodating for Alps as well as cherry yet? >:)

Also that sheet steel case may be heavy and European members may pay lots on postage. Just worth considering.

[OrangeJewce]

Antone suggested accommodating for Alps as well as cherry yet? >:)

Also that sheet steel case may be heavy and European members may pay lots on postage. Just worth considering.

Sheet steel? Where is that coming from. AFAIK the GB won't be offering that as an option, only aluminum (or nothing maybe? Dox hasn't confirmed this).

Cheers,

[dirge]

Got the metal mixed up. But if I could get it locally and cheap steel would be nice

[dirge]

Thanks guys!
Ceramic is possible, as well as gold plated stainless steel.

shapeways.PNG

Knew I wasn't going mad, must have been remembering dox and dorkvaders comment about layered steel. Gold plated stainless steel does sound nice and expensive

[bisl]

The way I read that, it'd be like $350ish per plate...and there are like 11 plates? So just short of $4k for the whole case?

Totally baller, but until I pay off my mortgage I think aluminum will do just fine.

[OrangeJewce]

The way I read that, it'd be like $350ish per plate...and there are like 11 plates? So just short of $4k for the whole case?

Totally baller, but until I pay off my mortgage I think aluminum will do just fine.

Lol! Seriously I won't be able to afford more than one with the amount of money required...which is really bad since I am unsure these will ever be made more than once (in any significant quantity).

Cheers,

[damorgue]

I am interested in one. I may not be interested in the case if buying without it is possible. It depends on which route you take with it.

[dox]

The PCBs will probably be under 50$ for the pair. The case situation is still pending. I'm currently using a shapeways case that cost ~275$ (because it's angled) and a flat shapeways case cost ~200$. dorkvader is working on a aluminium plate laminate case.

[OrangeJewce]

Thanks for the update Dox! That seems pretty reasonable. Under $50 is really a great price. $250 per kit is pretty reasonable IMO.

I should probably update regarding the UI in case anyone is interested: right now I am waiting for some simple middleware to be written and a relatively stable version of the firmware to be branched so I can work on the backend of the UI. There is still a lot of work to do, but progress is being made. I'd like to remind everyone the UI will be Windows only!

Cheers,

[justcallmecrash]

I don't mind the UI being Windows only, so long as the layout is stored inside the board itself. I have access to a couple Win machines, but I primarily use GNU/Linux, so if it won't be statically programmable, I may have to pass. :-/

[bpiphany]

I don't think the idea is to have drivers that will tell the OS what the keyboard is talking about. The software OJ is talking about is just to alter the layout on the chip, I am pretty sure..

Running Linux (like anyone sane) I don't think you will mind altering some layout c or h file, recompiling and re-flashing the controller yourself ;D

[dirge]

(like anyone sane)

MCSE, I wear the badge of insanity...

[justcallmecrash]

(like anyone sane)

MCSE, I wear the badge of insanity...

And you likely make a decent wage wearing it.

I don't mind mucking about in the firmware, I just want to be sure I can set it up to do Dvorak natively and set up the second layer the way I need it. I can still do it in Windows, just wanted to be sure the changes would be static from station to station.

[ondast]

Looking at the current code, it shouldn't be too hard to set up a different layout with some interesting layers or even make a function that can swap layouts.

[OrangeJewce]

All,

The UI will have some presets built in when I release it, and Dvorak will be one of them. The way that this will work is the following:

When you compile the firmware, you get a .hex file which stores the mappings, layers, etc. The UI will modify and save a new .hex for each edit you make, while always preserving the default ANSI layout in case you brick your board (you never know). Then you can flash the new layout to the ErgoDox from the UI. You will have access to all the layers and keys, for you VI users that want to swap esc and capslock.

crash: Once you've flashed the layout onto the board, it won't change between workstations, because the layout will be on the Teensy micro controller.

I vetoed the idea of drivers for the simple reason that it would be insane to do as a one person project. There is just too much variation in hardware to create drivers for each platform. Therefore, changes you make will likely be affected by changes made on the OS level.

bipaphany and I have had this discussion in the past, and whilst we agree that drivers are the better solution, the reality is that doing so is untenable at this time. Let me know if you guys have anymore questions.

Cheers,

[OrangeJewce]

I'd also like to add that there is no difference between the method I'm using and recompiling the firmware manually, because you still won't have the Teensy-OS communication a driver would afford you. The differences will be what flags you set in your C compiler, and the physical hardware you compiled on. This may or may not have any performance implications.

Cheers,

[ic07]

I don't mind mucking about in the firmware, I just want to be sure I can set it up to do Dvorak natively and set up the second layer the way I need it. I can still do it in Windows, just wanted to be sure the changes would be static from station to station.

As OJ said, the UI will will use firmware level remapping (via some nice binary manipulation of the .hex file locally, and then a reflash of the controller). If you have a programming background (which from your comments it sounds like you do) remapping without the UI should be easy too :) - I tried to make it intuitive... - don't know that anyone but me and Dox have messed with it yet.

I'm planning to release the firmware with a Dvorak layout as an option in the source as well, if I have time. If you have input on how the Dvorak layout should look, please post it :) I haven't personally seen any work on one yet.

[ashleydev]

Nice work!

I'd love to own 1 or 2 of these.

Are initial orders going to be out this year?

[Tafryn]

I took a stab at a dvorak layout, and here it is. It's mostly inspired by the Kinesis Advantage with some HHKB and TE accents. I'm not sure if the firmware supports layers or not, but I drew up a function layer just in case it does. The capability to lock one or both sides of the board to the function layer (perhaps with a double tap of the Func key) would be pretty great. However, I'm not sure if that's possible while still maintaining modifier-like behavior on a single press of the Func key. Either way, the main layout has pretty much everything you'd need without the function layer (except for the F-keys).

ergodox-dvorak.png (928 KiB) Viewed 5550 times

ergodox-dvorak-func-layer.png (893.33 KiB) Viewed 5550 times

[justcallmecrash]

I took a stab at a dvorak layout, and here it is. It's mostly inspired by the Kinesis Advantage with some HHKB and TE accents. I'm not sure if the firmware supports layers or not, but I drew up a function layer just in case it does. The capability to lock one or both sides of the board to the function layer (perhaps with a double tap of the Func key) would be pretty great. However, I'm not sure if that's possible while still maintaining modifier-like behavior on a single press of the Func key. Either way, the main layout has pretty much everything you'd need without the function layer (except for the F-keys).

I LOVE it! Awesome layout.

[ic07]

I took a stab at a dvorak layout, and here it is. It's mostly inspired by the Kinesis Advantage with some HHKB and TE accents. I'm not sure if the firmware supports layers or not, but I drew up a function layer just in case it does. The capability to lock one or both sides of the board to the function layer (perhaps with a double tap of the Func key) would be pretty great. However, I'm not sure if that's possible while still maintaining modifier-like behavior on a single press of the Func key. Either way, the main layout has pretty much everything you'd need without the function layer (except for the F-keys).

Thanks! :) . There's a few things I'm working on first, but I'll get it in there. I also posted it on the github page.

The firmware does support layers, and layer locking, and it'd be possible to implement locking just one side. Locking based on double tap though would take enabling the timer, and I don't think I'll have time to play with that part of the Teensy for a little while.

[Tafryn]

Thanks! . There's a few things I'm working on first, but I'll get it in there. I also posted it on the github page.

The firmware does support layers, and layer locking, and it'd be possible to implement locking just one side. Locking based on double tap though would take enabling the timer, and I don't think I'll have time to play with that part of the Teensy for a little while.

How about having <Left Func> and <Right Func>. Both could act as modifiers when holding them and pressing any other normal key. Locking a side to the function layer could be accomplished by holding one function key and pressing the other (e.g. Holding down <Left Func> and pressing <Right Func> would lock the right side to the function layer and vice versa). Unlocking could be accomplished by repeating the locking key-combination. That way the behavior of the function keys stays consistent regardless of which layer either side is in.

The behavior of a <*Func> + <Normal Key> keypress on a function locked side is still somewhat unclear. It could act like a keypress on the default layer, or it could act like a keypress on the function layer. For example, lock the right side to the function layer with <Left Func>+<Right Func> then press <Right Func>+<F10>. Should the resulting keypress be "8" or "F10"?