Recent Posts

Pages: [1] 2 3 4 ... 10
1
Hacks / Re: Balanced Keyboard Layout
« Last post by iandoug on Today at 06:32:48 AM »
will check again later if modifiers are already counted elsewhere.

Normal keys count up and down (am doing test of ab and DK on QWERTY on my local version of Den1 scoring).

So for ab the left pinky moves 0.8 which is 0.4 up and down.

For DK left pinky moves 7.7 cm which is worrying me a bit... the distance from centre to centre for A -> left shift is about 3cm, so total should be around 6.8 not 7.7

Unless KLA is computing distance as 7.7 - 0.8 = 6.9 / 2 = 3.45 but that seems a bit high on two ANSI keyboards I have here (low-end Gigagbyte 'gaming' with Cherry red switches, and low-end Microsoft model 200).

Perhaps Patrick's original model had slightly different key sizes which made the left shift bigger.
Mmmm he has size of 116 pixels == 116/26.315789 = 4.408 cm ... measured the base of the key and it's just under 4.2 cm
[edit; okay forgot about your modifier penalty which should account for the difference.... seems to be about 2mm]

Either way it does indeed appear to be counting up and down on Shift keys.

So doing it again elsewhere is double penalty...
2
Hacks / pixels per cm
« Last post by iandoug on Today at 01:42:40 AM »
@Den

Can you remember what the outcome was of the discussion re pixels per cm in the keymaps?

We have different values: (this be from code my side, in my current live version).

standard: KB.keyMap.standard.s683_225.pixelsPerCm = 26.315789;
Ergolinear Den: KB.keyMap.ergolinearDen.s683_225.pixelsPerCm = 25.315789;
Ergolinear Matrix (original): KB.keyMap.matrix.s683_225.pixelsPerCm = 26.315789;
Ergolinear Ian: KB.keyMap.ergolinear.s683_225.pixelsPerCm = 26.315789;

I remember asking Patrick about it and he said something about the Ergodox having smaller squares but the px/cm is same as standard above.

Pat's original code had 26 for standard and European and 25 for Ergodox.

So now I'm confused...
I'm sure the difference in px/cm affects the distance calculations and thus the scoring.

3
Hacks / Re: PanGalactic Keyboard Layout
« Last post by iandoug on May 27, 2017, 01:09:00 PM »
And you thought Asian languages were complex.

Put this on a keyboard (or in IPA, for that matter...)

http://www.bbc.com/future/story/20170525-the-people-who-speak-in-whistles
4
Hacks / bug
« Last post by iandoug on May 26, 2017, 12:54:52 PM »
Decided to have a go at rewriting/porting analyzer.js to PHP, so it can run headless and hopefully faster for bulk testing. And not gobble up my memory.
If it works will look at doing it in Ada since I want to learn Ada and then it will run even fasterer. :-)

Anyway think there's a bug here:

        if (char2KeyMap[16].errors.length > 1) {
            analysis.errors.push("Fatal Error: AltGr key not set correctly.");
            return analysis;
        }

which I think should be
        if (char2KeyMap[-18].errors.length > 1) {
            analysis.errors.push("Fatal Error: AltGr key not set correctly.");
            return analysis;
        }

since 16 is used above for the Shift keys... and the GUI assigns hex 12 (== dec 18) to AltGr....
It's the same in original version. Presumably a typo by Patrick.
5
Hacks / Re: Balanced Keyboard Layout
« Last post by Den on May 25, 2017, 04:23:38 PM »
Um, how do you type "N" ?

KLA would like to know :-)

thanks, Ian

telekinetically. in fact, all typing should be done telekinetically.
6
Hacks / Re: Balanced Keyboard Layout
« Last post by iandoug on May 25, 2017, 02:56:41 PM »

- added 3-thumb matrix layouts (L_RN, P_RN, U_RN)

Um, how do you type "N" ?

KLA would like to know :-)

thanks, Ian
7
Hacks / X6.4H dethroned...
« Last post by iandoug on May 25, 2017, 11:56:08 AM »
Played some more.

X6.4H has fallen. Long live the king. Ra rah.

This be on Alice.
8
Hacks / Re: Balanced Keyboard Layout
« Last post by philippe.quesnel on May 25, 2017, 09:27:24 AM »
Somehow stumbled upon the next BEAKL 6 candidate. Good distance and Low same finger.

Code: [Select]
BEAKL 6

qyouj wdrcv
hiea, gtsnp
/'x.z kmlbf
Interesting.
Gave it a quick run and kinda like it.
As usual, I tweaked it a bit, swapped . x for my shifted left hand bottom row access
(I realized yesterday that the US kbd version of Arensito uses this shifted bottom left row thing too  ;))

Here's something that I came up yesterday with AdNW, using an US kbd adaptation of BEAKL 3.4, with shifted bottom left row (with adjusted costs for the shifted row)
Code: [Select]
xncdv -oiujz
wrstg ,aehpk
lfmbq '."y: 
B-V is not oBVious though, (poetic, that's how I realized it actually, typed 'obvious'), I kinda avoid it a bit by hitting the bottom B with my thumb ! I think BEAKL6 might be nicer.
9
Hacks / Re: Balanced Keyboard Layout
« Last post by Den on May 25, 2017, 08:35:32 AM »
Using thumbs for modifiers seems underutilizing them. Mods take up only 2% of key press. That's about the amount I'd allow for each pinky. That means one of the thumb is only doing as much work as the pinky. (The other thumb is for space, which accounts for 16% key press, which is about the right amount of effort for thumb.)

So mods could be assigned to the fingers. But where? Pinkies are okay if within the main 30 key block; possibly even home pinky. This is perfectly fine if we're putting letters on thumb cluster anyway. (Remember that left one key empty when migrating 3 letters to thumbs.)

Next project for Opt.
10
Hacks / Re: Balanced Keyboard Layout
« Last post by iandoug on May 25, 2017, 07:03:04 AM »
Was annoyed with life so I played around with your P_RN some more... this gets a very respectable score at the expense of more same-finger use on left thumb.

I suspect with some tweaking it could beat X6.4H, as I only played with the letters and not the punctuation/brackets.

Gives me some food for thought :-)

Pages: [1] 2 3 4 ... 10