I wasn't sure if this was appropriate in the New Puzzles section so I made a new thread here. Since Mf8 is producing a crazy version of the FTO I finally set down and took care about it's solution

How I solved the regular circle FTO:
I had different ideas on how to approach it. So far my easiest solution that I could come up with is solving the circles first. The point is that setupmoves for circle pieces later on are confusing for me. Besides that, all algs that I could find to solve only circle pieces are long.

Solving the circles first: The pieces where three are connected and that belong to the dot that sjoerd drew behave like trivial corners of a cube that need to be oriented, nothing more. The other pieces do behave like pyraminx edges. Solving those is a bit confusing. There's always three of them moving but there are as many as on a magic octahedron.

Once that is done, I use only Face moves to solve the whole rest. First the tips, just like on a skewb diamond.

Than comes the boring but effective part: I use the normal skewb alg that changes tips (skewb centers) four times. Two times and the tips are back in place and another two times to reorient them. That makes 16 moves. It cycles three edges without touching the circles (not a clean cycle though).

Using this alg as a commutator with one more move in between cycles three "triangles" and goes like this (they aren't really triangles but arc like pieces):
[the 16 moves from above . adjacent face]
That restores the edges but leaves a clean arcthingy cycle.
What do I mean by adjacent face? The face is adjacent to one of the two faces of the 16 moves alg and opposite to the other. I guess it doesn't matter wich one exactly.
In gelatinbrain it's [(UFL' ULB UFL ULB')*4 . URF'] as a commutador. That makes 34 moves.

Really monotone but it works easiest from all I found. If anyone has effective algs for solving the circles last I would love to know. I only have one for a three cycle of the circle pieces on the tips.

Hey I've had your post in the back of my mind for a few days but I've been too busy to work on anything and make a post. When I solved the Circle FTO I did it very inefficiently and I wound up with a single twisted center which was quite awful resolving without breaking other pieces.

The way I'd go about solving the Circle FTO now is:

* Solve edges and orient centers (this is just a super-Dino solve and should take less than 20 moves total)
* Solve corners (use [1,1] commutators)
* Solve center triangles (use [3,1] commutators)
* Solve the circle - and | half-corner pieces under each corner (use [6,1] commutators)

The reason I suggest solving the center triangles before the circle half-corner pieces is that there are 24 centers and only 16 circle half-corner pieces so the long routine isn't so bad.


Here are the routines in Gelatinbrain notation:
Corners [1,1]: [UFL', UBR, UFL, UBR']
Center triangles [[1:1], 1]: [ULB, URF, ULB', DBL'&2, ULB, URF', ULB', DBL&2]
Circle half-corners [1:[1,1],1]: [UBR, ULB, URF, ULB', URF', UBR', DBL'&2, UBR, URF, ULB, URF', ULB', UBR', DBL&2]

During the Dino-edge solve phase if you have a twisted center you can fix it with [UFL, URF, UFL', URF]x2

I tried getting the circle half-corner routine down to a [5,1] without any luck. Maybe Julian has something better?

Sorry if this is not exactly the thread topic, but where can I buy a regular circle FTO?

As of right now I don't think there is a place. It isn't known if the "Crazy FTO" series will be a standard Circle FTO or one of the bandaged "Planet" series.

See:

http://twistypuzzles.com/forum/viewtopic.php?p=264405#p264405

Oh, right. Does that mean that the solving discussion in this thread is because they're solving it virtually?

Yup, Gelatinbrain's Virtual Magic Polyhedra is a wonderful gift to humanity.

OK. Thankyou and apologies for the hijack.

Hijack over.

Continue solving discussion...

Cool. I like the first part. Solving edges first and than using the [1,1] commutator for all corners works really well.
I do have problems with the last commutator though. When paste it into the applet I see cicle pieces from 4 sides beeing cycled. 6 pieces in total are moving in 2 seperate 3 cycles. Isn't it difficult in the end to get those solved?
I didn't go on through the whole solve when I reached this point. I think I can't handle it.
But I gues your first two steps are better in any way.
However I realised that your corner alg is probably just what I used to solve the circle pieces in the corner parts at the beginning of the solve.

I should have been more explicit about what is going on here. The circle pieces cut out of the corners come in pairs. What looks like 2 separate 3-cycles is actually 1 3-cycle. The two circle-wedge pieces across from each other under the corners are attached to each other and can't be separated.

That's why I called them "circle half-corners" and referred to them as "|" and "-". Right under each corner are two pieces that cross each other perpendicularly and form a "+". My 3-cyle moves only 3 of these but since each one has two stickers it looks like two 3-cycles.

DUH! Of course. Sorry.
I kind of after the triangle algs seem to have lost the overview or so. I did notice that two colors are one piece when I solved these pieces in the beginning of my method too.
Yeah that was kind of embarassing.
I will try again.

I can't cycle the circle half-corners quicker than [6,1], but if you reduce the Circle FTO to a Pyraminx then you only need to solve the second half of them that way, not all of them. My reduction solution is illustrated here.

Reduction is more confusing than cycling by piece type, but I'm guessing that familiarity with solving the Circle FTO by reduction could prove useful when tackling some of the Crazy Octahedra planets, if/when they are released.