Reason for size of ErgoDox outermost columns

[v6ak]

Ergodox outermost columns (i.e., the leftmost column on the left half + the rightmost column on the right half) have traditionally width 1.5 of standard key, except the very bottom row, which is just the basic key size. Why?

This is just a fact I neither love nor hate, but still, I'd like to know the background and if it is reasonable to replicate this or not.

 I have tried to look for ErgoDox design decisions, but I haven't found that.

[Findecanor]

It was officially inspired by the Key64, and the Kinesis contoured which had 1.25u keys in the outer columns, and no 1u key in the outer/bottom corners.
Why Dox instead chose 1.5u keys, I don't know: maybe just because keycaps were easier to find in (close to) the correct profiles. On his personal prototypes he used 1.5u Backspace, 1.5u Control and 1.5u Tab keys, with only the home row's profile being wrong.

Maybe Dox will answer himself.

As to science: If a key is not on one of the main finger-columns, then you could apply Fitt's law. It (when applied to keyboards) says that outlying keys are easier to hit if they are larger.

[dusan]

According to my own (unscientific, unpublished) study, there is a large variation (taken over individuals) of the exact position where the right hand pinky hit the Shift key on the standard keyboard. Somebody hits the left edge while somebody else hits the middle of the key. People tend to hit the gap between it and the upper key, i.e, the Enter.

(The right edge of the right Shift key on the standard keyboard is also hit frequently, but by the index finger in navigation+editation mode. So it does not apply in case of mini boards.)

Therefore, if we can afford column-staggered layout in a design for the mass, the 1.5u width and 0.5u vertical stagger for the outermost column, like the Ergodox, is pretty reasonable.

[Dox]

It was mainly to make it easier to hit those keys. The pinky is not the most precise finger and having bigger keys there makes it easier to hit.

I am convinced that it would be possible to get used to having 1u keys there and it might be easier to find fitting profile keycaps but that's not the way I went with 8 years ago when the keycap choice was much more limited.
There was nothing like the ortholinear kits of today available that had the matching profiles in 1u width.

[Findecanor]

BTW. There are revisions of the ErgoDox PCB into which you could fit 1u keys in the outer columns. Those are not supported by all plates, and you would have to make a custom case unless the plate is the top of the case.

bpiphany had added them in the first public PCB design (maybe because I had asked him to...), but he had made an error in the component footprint, so switches did not originally fit without having to modify the PCB. That error has been corrected in a later revision but there might be vendors out there who sell the older revision still.

[v6ak]

Thank you for the comments. Now, only one question remains: Why the very bottom row is not 1.5U?

[Dox]

It's easier to fill the bottom row with 1u as the columns are staggered. 1.5u just doesn't line up with the columns.