alg.cubing.net

[Lucas Garron]

It only mirrors the Moves, not the scramble. This should be corrected.

The jury's still out on the right behaviour, but I've actually changed my mind and implemented mirroring both together for now. Try mirroring the WR, for example.

what about a check box so you can disable live updating? it runs really slowly for me because of that

I still don't want to disable this, because there are quit few interrelated parts, but I've implemented "debouncing".
If you make one change, it will happen immediately. If you quickly do a few more, if will wait until you've finished.

it doesnt mirror the moves for me though

If it still doesn't work, could you post your OS/browser?
If you can access the console, I could also use any error messages you get.

 

[Petro Leum]

If it still doesn't work, could you post your OS/browser?
If you can access the console, I could also use any error messages you get.

it does work by now (miraculously without having changed anything)

 

[Renslay]

Am I right that there is a new feature? I mean, when the cursor is before a turn, in the cube it shows that turn with a slight movement.

Since I noticed this update, the site (the update of the animation) became really slow. I have to wait half or 1 second for every movement I just type in, which is really annoying. Especially when I type a longer sequence, it can freeze for seconds.

(I have Windows 7, Chrome 35.0.1916.153 m)

 

[Lucas Garron]

Am I right that there is a new feature? I mean, when the cursor is before a turn, in the cube it shows that turn with a slight movement.

I suppose it's a feature. This was a tiny thing I added when I restored some of the move highlighting tracking.

Since I noticed this update, the site (the update of the animation) became really slow. I have to wait half or 1 second for every movement I just type in, which is really annoying.

This was sad to read, because one recent change was supposed to reduce the amount of slow things in basically every situation. :-(

I've tried fix that a bit more. Nothing should get updated until 1 second after you've stopped typing.
Let me know if that's not the case, or if it's still really slow for some other reason.

 

[10461394944000]

adding a new line doesn't automatically add a pause anymore, is this a feature or a bug? I preferred it with the pause.

 

[Lucas Garron]

adding a new line doesn't automatically add a pause anymore, is this a feature or a bug? I preferred it with the pause.

More of a bug. I didn't even realize anyone noticed it was a feature before. I'll add it back in.

EDIT: Fixed. It'll highlight the end of the line while pausing, until I think of something better.

 

[Christopher Mowla]

I'm going to link the 4x4x4 parity algorithms wiki page to alg.cubing.net, but I just wanted to know what the parameter names are for controlling whether the hint stickers and hollow cube options are turned on or off. (I want to have hint stickers turned off when an algorithm link is clicked, as I don't think they are necessary for 4x4x4 algorithms).

 

[Lucas Garron]

I'm going to link the 4x4x4 parity algorithms wiki page to alg.cubing.net, but I just wanted to know what the parameter names are for controlling whether the hint stickers and hollow cube options are turned on or off. (I want to have hint stickers turned off when an algorithm link is clicked, as I don't think they are necessary for 4x4x4 algorithms).

Not yet. I want to find a way to let everyone choose personal settings for these (possibly depending on the cube size), but this is not the case for now.

If you want "stage" settings (like PLL or OLL stickerings), I can implement them soon. Feel free to post suggestions; if it's very clear how to implement it, I should be able to do so soon.

 

[Christopher Mowla]

Not yet. I want to find a way to let everyone choose personal settings for these (possibly depending on the cube size), but this is not the case for now.

Oh, okay. I'll wait. I thought that I was a little behind with linking template Alg4 to alg.cubing.net, so I mentioned it.

I plan to change the notation of all algorithms on that page to SiGN, and I have made the decision to write all algorithms from now on in SiGN notation (the stubborn cmowla has been broken!), using r as Rw and not 2r (since that is what the majority of people agree with, I can't dispute it). So just let me know when it's time, and I will modify all algorithms on that page so that you don't have to.

If you want "stage" settings (like PLL or OLL stickerings), I can implement them soon. Feel free to post suggestions suggestions; if it's very clear how to implement it, I should be able to do so soon.

Perhaps PLL and OLL stickerings would be eventually nice to be implemented for those who would possibly want it for 4x4x4 PLL parity cases, 4x4x4 OLLs and K4 OLLs, but I don't think it should be a priority right now.

As far as suggestions in general, I think the default speed is too fast. Can the default speed for x1 be changed at least to the current .5? I can't see how the current speed setting is suitable for the majority of users. In addition, as I mentioned, a parameter option (in the url code) to turn hint stickers and hollow on and off should be made available.

However everything else looks great. I also noticed that your program is on cube20.org now. I'm glad to see that.

 

[Lucas Garron]

General stuff: I've changed the progress bar and redid the right panels, so everything should look a little bit nicer now. Still not great, but hopefully it works better for everyone.

I've also moved metrics to the right of the header for the solve/alg/moves, although I still need to make sure it displays in the correct location at all times.

I plan to change the notation of all algorithms on that page to SiGN, and I have made the decision to write all algorithms from now on in SiGN notation (the stubborn cmowla has been broken!), using r as Rw and not 2r (since that is what the majority of people agree with, I can't dispute it).

Awesome! I know you weren't a big fan, but I'm glad to see everyone converging on a standard.
Any time you want to change over is probably good.

As far as suggestions in general, I think the default speed is too fast. Can the default speed for x1 be changed at least to the current .5? I can't see how the current speed setting is suitable for the majority of users.

The original speed was a balance between going slow enough to keep up with, but fast enough to feel "real-time". But I've thought about it, and I think you're right; it makes sense for the default to be slower. I've slowed it down to be half the original speed.

For actual "real-time" reconstructions, I will have a different timing mechanism anyhow.

In addition, as I mentioned, a parameter option (in the url code) to turn hint stickers and hollow on and off should be made available.

Don't hold your breath on this one. For most cases, I believe that this should be a user preference. I don't want to make something a URL parameter if it isn't really needed.

Some day I will probably implement user preferences, but it isn't a high priority.
(If someone is interested, it should be easy to take the source and modify the checkboxes to be persisted to localStorage. But 1) I'd like this to be integrated with angular.js, and 2) I don't want this to confuse/annoy users any more than the current setup.)

 

[Renslay]

A few things:

1a) The move counts are originally in the line of the Solve line. However, when I scroll the window, it keeps its relative position. It's a bit weird. (Or did you mean this as a problem when you said "although I still need to make sure it displays in the correct location at all times" ?)

1b) The scroll bar overlaps the last bracelet of the move counts. Should move it little to the left. (I use Chrome 35.0.1916.153 m, Win7 if that counts.)

2) The right panel has a default background color (the color behind the text "alg.cubing.net [beta]". This color changes in the header lines ("Setup" text line, "Solves" text line, etc). This should not be a problem on its own, however...
The place where you actually type texts (the actual setups and the solves below the header lines) has also background color.
This confuses me all the time, because normally (in a web page for example) the place where you enter texts has the different color (like the search field). I always click on the "setup" and the "moves" textline to enter my moves, because I always have the impression those are the fields where I should type, because of the colors. Moreover, the dark, thin line between the two colors also make the impression that the text line "Setup" and "Moves" are a bit deeper, making the false assumption "click here to type".

I know there is a text "click here to add (...)", but this still bothers me a little. Moreover, I am not sure what would be a good solution. Maybe put a white box where you can type, similar to the scroll lists in the settings (size of cube, Color scheme, etc). Also maybe you should put a triangle (pointing down) in the right side of the scroll list, showing they are actual scroll lists.

EDIT:
3) If I click on the "image", an image of the cube appears. But when I click it on again, it does not disappear (I suspects it would.)

~~~~~~~~~~~~~~~~~~~~~~~

And finally, let me clear that I love this little(?) site, I use it daily, and I am extremely grateful for all the efforts you put in there to make it better and better! :tu

 

[Lucas Garron]

1a) The move counts are originally in the line of the Solve line. However, when I scroll the window, it keeps its relative position. It's a bit weird. (Or did you mean this as a problem when you said "although I still need to make sure it displays in the correct location at all times" ?)

Yeah. This and 1b were caused by the same thing. Your post encouraged me to fix this sooner rather than later, and it turned out to be very simple. Let me know if you still have problems after a hard refresh.

2) The right panel has a default background color (the color behind the text "alg.cubing.net [beta]". This color changes in the header lines ("Setup" text line, "Solves" text line, etc). This should not be a problem on its own, however...

I'm not completely sure what to do about this. I've put in a lot of affordances on the text areas, including a white color when you hover or type, and highlighting the moves/alg/solve textarea by default.
However, when you're not manipulating them, I *do* want the textarea to blend passively into the background.

Does anyone else have similar issues? If it affects multiple people, then 1) I'll definitely want to consider changing something, and 2) I might know better what to change.

3) If I click on the "image", an image of the cube appears. But when I click it on again, it does not disappear (I suspects it would.)

Every time you click it, you get an image of the current state. For example, try clicking repeatedly while the alg is animating.

I suppose I could add a button to hide it, but it's not a priority. (You can just refresh the page to reset anything you tinkered with.)
Anyone is welcome to add that button, though. ;-)

 

[Renslay]

Issue 1a) and 1b) are fixed, thank you!

Issue 2) is just a minor thing. I just described it; newcomers might find that strange (if not, my apology). But I won't complain if you decide to keep it this way.

Issue 3) is again a minor thing. I wasn't sure if "the image won't disappear" is a feature or a bug. I just assumed that clicking once generates an image, and clicking second times hides it (clicking third times re-generates, etc.) If you think about it, it might be a bit weird that some feature appears after pressing a button, and it won't disappear unless you reload the whole page. But again, this is a very minor thing, and honestly, I never used that button before.

Thank you again for all the effort and creating this marvelous toolkit!

 

[Christopher Mowla]

Lucas,

I noticed that you didn't implement a move to move just the central slice with a letter m.

Obviously right now we can just use a coefficient in front of the letters R or L. \( \left( \frac{n+1}{2} \right)\text{R} \)

Do you intend to leave it like this?

 

[Lucas Garron]

Lucas,

I noticed that you didn't implement a move to move just the central slice with a letter m.

Obviously right now we can just use a coefficient in front of the letters R or L. \( \left( \frac{n+1}{2} \right)\text{R} \)

Do you intend to leave it like this?

M slices on big cubes are unclear.

I think an uppercase letter should always be a single slice by default, but that only works well for odd cubes..
qq thinks M should be everything but the outer layers, so that uppercase 3x3x3 algs will still work on a big cube (with solved centers and edges, e.g. after reduction). I somewhat agree that outer-layer algs shouldn't need to use centered slice moves in lower case to keep the correct semantic meaning for big cubes.

Personally, I think the latter is a good use for m, but it's unclear for the time being. There are also no "popular" conventions, so we'd just be picking something based on what we *think* people want or find useful... and cubing is full of mistakes where someone made a choice that seemed sensible based on the available information, but might not be so good in retrospect. (My least favorite: the direction of M, allowing +2 for misalignments, and allowing inspection time in competitions.)

Since there is nothing obviously correct, I'm keeping things as canonical as possible on alg.cubing.net. I've got too many things to do – resolving this is not a priority, but I want it to have a good solution some day.

 

[Hypocrism]

M slices on big cubes are unclear.

I think an uppercase letter should always be a single slice by default, but that only works well for odd cubes..
qq thinks M should be everything but the outer layers, so that uppercase 3x3x3 algs will still work on a big cube (with solved centers and edges, e.g. after reduction). I somewhat agree that outer-layer algs shouldn't need to use centered slice moves in lower case to keep the correct semantic meaning for big cubes.

Personally, I think the latter is a good use for m, but it's unclear for the time being. There are also no "popular" conventions, so we'd just be picking something based on what we *think* people want or find useful... and cubing is full of mistakes where someone made a choice that seemed sensible based on the available information, but might not be so good in retrospect. (My least favorite: the direction of M, allowing +2 for misalignments, and allowing inspection time in competitions.)

Since there is nothing obviously correct, I'm keeping things as canonical as possible on alg.cubing.net. I've got too many things to do – resolving this is not a priority, but I want it to have a good solution some day.

Just out of curiosity, why don't you agree with inspection time?

 

[Renslay]

I have another idea:

1) A button that transforms the movement rotations into outer layer only rotations, removing cube turns also.
So, before:
x y M U2 x M L2 S Rw

After:
U D' L2 U D' D2 F' B R (z2 x)

I don't think it would take much effort, effectively you have to cascade the cube rotations through (and changing the middle layer turns and wide turns), for examle the steps inside:
x y M U2 x M L2 S Rw
x y M U2 x M L2 S L x
x y M U2 x M L2 F' B z L x
x y M U2 x M L2 F' B D z x
x y M U2 x R L' x' L2 F' B D z x
x y M U2 x R L' L2 x' F' B D z x
x y M U2 x R L' L2 U' x' B D z x
x y M U2 x R L' L2 U' D x' D z x
x y M U2 x R L' L2 U' D F x' z x
x y M U2 R x L' L2 U' D F x' z x
...
U D' L2 U D' D2 F' B R (/* bunch of cube rotations here */)
U D' L2 U D' D2 F' B R (z2 x)

 

[Lucas Garron]

I have another idea:

Yep: also see Petro Leum's post for a more general idea.

The current traversal mechanism should be powerful enough, and I maintain that anyone who's willing to get their hands a little bit dirty in Javascript can implement this easily, if they want to see it sooner. (alg.js is even in its own project as of two days ago.)

(A slight problem is deciding what to do with commutators: I don't see a great way of "removing rotations" from [R, U z] without expanding it. Expanding it is definitely an easy way to do it, though.)

 

[inteli5]

a few suggestion

hi, Lucas, Thank you so much for this awesome job. I have been using it for a couple of month.
I have a few suggestions and issues to the project. All my comments are based on the current webpage on alg.cubing.net

(1) parameter in URL: please add speed and hint sticker.
(2) Let users have an interface to define the stage, through a file or through a URL parameter like stage=111111111222222222333333333444444444555555555666666666. So that we can demonstrate algs of cross etc. For larger cubes, we can do it through a text file.
(3) when I change the color scheme to custom and type in my color scheme. it does not change anything.
(4) can you add a move like SR in superset ENG notation. SR=R+L' SU=U+D' etc, if letter S is used for slice move. we can do another letter like Q. or if two move is included in {} and they do not conflict you can let them move together.

 

[cubernya]

hi, Lucas, Thank you so much for this awesome job. I have been using it for a couple of month.
I have a few suggestions and issues to the project. All my comments are based on the current webpage on alg.cubing.net

(1) parameter in URL: please add speed and hint sticker.
(2) Let users have an interface to define the stage, through a file or through a URL parameter like stage=111111111222222222333333333444444444555555555666666666. So that we can demonstrate algs of cross etc. For larger cubes, we can do it through a text file.
(3) when I change the color scheme to custom and type in my color scheme. it does not change anything.
(4) can you add a move like SR in superset ENG notation. SR=R+L' SU=U+D' etc, if letter S is used for slice move. we can do another letter like Q. or if two move is included in {} and they do not conflict you can let them move together.

2) You can add comments at the end of a line by adding // before the comment
4) I've never seen this done. ATM exists, but that isn't used to often (and is not supported in alg.cubing.net). And yes, S is used for a slice
I also believe he wants to make it so if two moves are on the same axis, they turn at the same time.