Canary VS oxeylyzer

https://github.com/Apsu/Canary I don't understand the sections on row stagger and angle mod. Please could someone explain in simple terms where to rest my fingers and which finger to use for each key. Thank you so much.

Keyboard layouts that are worth checking out are Canary, Graphite and Semimak. From what I understand they have all been made in the last two years. I'm biased in favor of Canary as it's very similar to Colemak which made the switch easy for me and I find it really comfortable, but the only thing that really matters is your personal experience when actually typing on the layout, not all the statistics that some people obsess over endlessly.

At https://github.com/Apsu/Canary, it's stated that Canary has 55% of all trigrams as rolls.

Try Canary layout https://github.com/Apsu/Canary It is very similar to Colemak DH and having E and U at different fingers. But, E and O is now at the same finger, isn’t that an issue for French typing?

Romak also has some very nice attributes, as do poqtea and uciea. I also find Canary to have a lot of merit.

My current layout, Canary, wouldn't be a great fit for this board, as with the reduction of six keys from a 34-key layout, six letters would be need to be under either a combo or layer, which I'm not too comfortable with.

I use hrm with canary layout, which puts the most used keys on the home row, so that means that I tap on the hrm keys the most. And I have pretty much 0 mistakes.

The homepages for Semimak JQ and Sturdy include thoughts by their respective creators on the layouts' strengths and weaknesses, though surely with a little bias =) The Canary homepage has detailed description of its Colemak-inspired design.

I use Canary ortho https://github.com/Apsu/Canary

Quick note about ortho vs column stagger vs row stagger when it comes to layouts: they generally don't make a difference, though sometimes you still want to make a change. This is most often with regards to upper + lower inner index vs howerow inner index, like on Canary (note the difference between the indexes of the ortho and the regular variant).

I downloaded https://github.com/O-X-E-Y/oxeylyzer with the idea of adding hebrew, but turns out it's already there! thanks @O_X_E_Y

As a corpus, I'm using https://github.com/O-X-E-Y/oxeylyzer/blob/main/static/language_data/english.json, filtered to just the letters. (And all punctuation-including trigrams account for just 5.2% of all trigrams, so removing these is too small to explain the discrepancy in my calculations.)

O-X-E-Y/oxeylyzer

Unfortunately, they are real. Source Source 2 Source 3 Source 4

My own, oxeylyzer. I just threw in a Ukrainian corpus (the mixed 1M from here), messed around with what should be on the main 3x10 and kinda let it rip with some different weights tbh