• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Stenographic Square-1 Controls - The Future of Virtual Squan!

OreKehStrah

Member
Joined
May 24, 2019
Messages
755
Hey all,

Just wanted to share the latest project I've been working on recently that I finished yesterday, which is stenographic square-1 controls for virtual square-1.
Here's my video demo of this, and if you want to read more about the idea behind it, and how this can be used feel free to read more!


and here's a link to the Plover Squan Control Dictionary I wrote:

The motivation/problem:

So to give context to this, I want to explain the motivations behind this as well as the problem this aims to solve.

I recently started using the virtual cube on in trainyu, found at: https://tao-yu.github.io/Alg-Trainer/ (go check it out, Tao did great work with this!), and used it to learn some ZBLL. I was shocked at how drastically it boosted my alg learning rate. I was able to learn 36 ZBLLs in roughly an hour and retained them in my memory for months without drilling before I started to forget any from not using them in solves often (got busy with classes).

More recently I've gotten back into square-1 and started learning the Lin method, and started wondering if it would be possible to get the same explosive increase in the learning rate by using a virtual squan to train algs on. This is where the problems start.

I realized I can repurpose https://www.cubedb.net/?puzzle=SQ1 (created by Gil Zussman, go check this site out and donate if you can!) for the actual virtual cube, which left one pretty large problem still: How do you control a virtual squan with a keyboard with fast, intuitive, and precise controls?

This seemed to be the common killer of virtual square-1 since a lot of versions used slow, imprecise controls, such as using one key to move the U layer in increments of (1,0) and the same for the D layer, and a key for the / . I thought that there had to be a better way and I found it through the application of stenography!

Stenographic Controls:

For those unaware of what stenography is, it is a system of "typing" designed to keep up with human speech by using chorded input. For example, instead of typing out the word "coffee" out letter-by-letter, you would press "KOEF" on a stenography machine all at the same time and it would output coffee. (Note: this is a very quick run down of it, but you should get the idea.)

Typically this was a skill behind a large economic barrier, but now steno is something anyone can learn using their normal keyboard and a piece of free software called Plover that takes keyboard input and reads it as stenographic input.

This can be applied to controlling a virtual squan very quickly and precisely!
You use the left hand to control the U layer and the right hand to control the D layer outputs. If you need both layers, just press multiple keys at once! So clean and simple!

Edit: by "this" I mean the idea of chorded inputs by repurposing Plover

There are some quirks that had to be worked around with how Plover and steno works but these are the controls I set up:

F = (1,0)/
D = (2,0)/
S = (3,0)/
A = (4,0)/
DF = (5,0)/
SD = (6,0)/
R = (-1,0)/
E = (-2,0)/
W = (-3,0)/
ER = (-4,0)/
WE =(-5,0)/

The same pattern is mirrored for the right hand to control the D layer starting with J. This is to keep the hands around the home row and help minimize and systematize the hand/finger movements. To get any UD layer combo, simply press the U layer combo and the D layer combo at once. For example, (-1,-1)/ is given by pressing RU at the same time. Also, you might have noticed that the output ends with "/" which can be truncated using a thumb key with any combo to get the desired output without slicing.

(Future) Uses/Applications:

This is essentially a usable proof-of-concept more than anything since it repurposes two pieces of software not originally meant for this use case. However, it does work and I learned squan V perm just by testing this earlier.

I think it would be really cool to see an virtual squan alg trainer as well as solver designed with this idea in mind in the future. I think this could really facilitate the alg learning process since you wouldn't need to constantly solve and scramble a real squan, which can be really annoying and boring. This also opens the doors for really fast virtual squan solving, so maybe people in the future will start solving virtual squan like they do virtual 3x3s.

anyway, thanks for checking this out! I'd love to hear what you guys think! Definitely give this a try since it's completely free to do so!

Again major thanks to Gil for having written speedcubedb and squandb as this project/proof-of-concept would have taken way more time without them!
 
Last edited:

abunickabhi

Member
Joined
Jan 9, 2014
Messages
3,347
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Hey all,

Just wanted to share the latest project I've been working on recently that I finished yesterday, which is stenographic square-1 controls for virtual square-1.
Here's my video demo of this, and if you want to read more about the idea behind it, and how this can be used feel free to read more!


and here's a link to the Plover Squan Control Dictionary I wrote:

The motivation/problem:

So to give context to this, I want to explain the motivations behind this as well as the problem this aims to solve.

I recently started using the virtual cube on in trainyu, found at: https://tao-yu.github.io/Alg-Trainer/ (go check it out, Tao did great work with this!), and used it to learn some ZBLL. I was shocked at how drastically it boosted my alg learning rate. I was able to learn 36 ZBLLs in roughly an hour and retained them in my memory for months without drilling before I started to forget any from not using them in solves often (got busy with classes).

More recently I've gotten back into square-1 and started learning the Lin method, and started wondering if it would be possible to get the same explosive increase in the learning rate by using a virtual squan to train algs on. This is where the problems start.

I realized I can repurpose https://www.cubedb.net/?puzzle=SQ1 (created by Gil Zussman, go check this site out and donate if you can!) for the actual virtual cube, which left one pretty large problem still: How do you control a virtual squan with a keyboard with fast, intuitive, and precise controls?

This seemed to be the common killer of virtual square-1 since a lot of versions used slow, imprecise controls, such as using one key to move the U layer in increments of (1,0) and the same for the D layer, and a key for the / . I thought that there had to be a better way and I found it through the application of stenography!

Stenographic Controls:

For those unaware of what stenography is, it is a system of "typing" designed to keep up with human speech by using chorded input. For example, instead of typing out the word "coffee" out letter-by-letter, you would press "KOEF" on a stenography machine all at the same time and it would output coffee. (Note: this is a very quick run down of it, but you should get the idea.)

Typically this was a skill behind a large economic barrier, but now steno is something anyone can learn using their normal keyboard and a piece of free software called Plover that takes keyboard input and reads it as stenographic input.

This can be applied to controlling a virtual squan very quickly and precisely!
You use the left hand to control the U layer and the right hand to control the D layer outputs. If you need both layers, just press multiple keys at once! So clean and simple!

Edit: by "this" I mean the idea of chorded inputs by repurposing Plover

There are some quirks that had to be worked around with how Plover and steno works but these are the controls I set up:

F = (1,0)/
D = (2,0)/
S = (3,0)/
A = (4,0)/
DF = (5,0)/
SD = (6,0)/
R = (-1,0)/
E = (-2,0)/
W = (-3,0)/
ER = (-4,0)/
WE =(-5,0)/

The same pattern is mirrored for the right hand to control the D layer starting with J. This is to keep the hands around the home row and help minimize and systematize the hand/finger movements. To get any UD layer combo, simply press the U layer combo and the D layer combo at once. For example, (-1,-1)/ is given by pressing RU at the same time. Also, you might have noticed that the output ends with "/" which can be truncated using a thumb key with any combo to get the desired output without slicing.

(Future) Uses/Applications:

This is essentially a usable proof-of-concept more than anything since it repurposes two pieces of software not originally meant for this use case. However, it does work and I learned squan V perm just by testing this earlier.

I think it would be really cool to see an virtual squan alg trainer as well as solver designed with this idea in mind in the future. I think this could really facilitate the alg learning process since you wouldn't need to constantly solve and scramble a real squan, which can be really annoying and boring. This also opens the doors for really fast virtual squan solving, so maybe people in the future will start solving virtual squan like they do virtual 3x3s.

anyway, thanks for checking this out! I'd love to hear what you guys think! Definitely give this a try since it's completely free to do so!

Again major thanks to Gil for having written speedcubedb and squandb as this project/proof-of-concept would have taken way more time without them!
This is super big. Thanks for taking up this project. All squaners will be benefited.
 

Tao Yu

Member
Joined
Jun 29, 2011
Messages
1,165
Location
Ireland
WCA
2012YUTA01
YouTube
Visit Channel
This is really cool! I had stumbled upon plover at some point and was somewhat interested in learning it, but never ended up doing so.

I am just curious, have you seen Ben Whitmore's virtual Square-1 solves? I'm fairly sure his sim just uses the same controls as normal 3x3 sim controls for RUD moves (though I'm not 100% sure because he wrote it himself) and gets really fast solves with it. I don't know enough about square-1 to have any thoughts myself, but it would seem to me that the question of whether stuff like plover will become widely adopted would depend on how it compares to the old approach.

.
 

OreKehStrah

Member
Joined
May 24, 2019
Messages
755
This is really cool! I had stumbled upon plover at some point and was somewhat interested in learning it, but never ended up doing so.

I am just curious, have you seen Ben Whitmore's virtual Square-1 solves? I'm fairly sure his sim just uses the same controls as normal 3x3 sim controls for RUD moves (though I'm not 100% sure because he wrote it himself) and gets really fast solves with it. I don't know enough about square-1 to have any thoughts myself, but it would seem to me that the question of whether stuff like plover will become widely adopted would depend on how it compares to the old approach.

.
I have! Someone actually suggested I check it out earlier today. The controls seem to use keys to move the U and D layers in the smallest possible increment legal for whatever the state the puzzle is in iirc. It’s better than most systems I’ve seen, but still not as precise as having combos to do the exact move you want instantly. I really hope chorded controls are adopted natively into some simulators so installing Plover and a special dictionary isn’t needed.
 
Top