Complete ZBLL alg set on Wiki Algorithm Database

[Cride5]

Hi All, as some of you may know eastamazonantidote kindly generated algs over the Christmas holidays covering the entire ZBLL set. With permission, I've uploaded them to the new algorithm database on the Wiki here:
http://www.speedsolving.com/wiki/ind...3&group=ZBLL-T

There are few omissions in the data. Most notably of which there as a Pi COLL case missing (12 algs), but that will be sorted soon. Once they've been generated this will be (I think) the first complete ZBLL alg collection online, all thanks to eastamazonantidote.

If anyone is in the process of generating ZBLL algs themselves and want to contribute, the database allows anyone to upload their algorithms. They will automatically be verified and added to the correct case. To upload an algorithm just click on the set it belongs to and then click on the 'add' link at the top.

The database itself is a fairly early release, and there are a number of features which could be implemented. If you have any feature requests, or discover any bugs/problems please let me know.

Again, huge credit to eastamazonantidote for this. Generating that many algs is a massive effort, and it was very well done!

Enjoy..

[eastamazonantidote]

Hi All, as some of you may know eastamazonantidote kindly generated algs over the Christmas holidays covering the entire ZBLL set. With permission, I've uploaded them to the new algorithm database on the Wiki here:
http://www.speedsolving.com/wiki/ind...3&group=ZBLL-T

There are few omissions in the data. Most notably of which there as a Pi COLL case missing (12 algs), but that will be sorted soon. Once they've been generated this will be (I think) the first complete ZBLL alg collection online, all thanks to eastamazonantidote.

If anyone is in the process of generating ZBLL algs themselves and want to contribute, the database allows anyone to upload their algorithms. They will automatically be verified and added to the correct case. To upload an algorithm just click on the set it belongs to and then click on the 'add' link at the top.

The database itself is a fairly early release, and there are a number of features which could be implemented. If you have any feature requests, or discover any bugs/problems please let me know.

Again, huge credit to eastamazonantidote for this. Generating that many algs is a massive effort, and it was very well done!

Enjoy..

Thanks for all the credit, but you've given me too much. You probably put in much more work than I. You had to put it online. I put it into tables with tons of shortcuts. GREAT WORK!!!

I have fixed every issue you alerted me about, so that Pi set is done. Just had to mirror over FB...

[jazzthief81]

This is great work from both of you! I've been getting a bit into ZZ lately and this makes me want to practice it even more.

Also Conrad, I think your ZZ tutorial is one of the best cubing pages ever made and a textbook example of how to make a good solution method description.

[Cride5]

Thanks for the feedback, its appreciated

Remaining problems have now been fixed, and all omissions taken care of. The set is now complete with an alg for all 472 ZBLL cases (493 inc PLL)..

Now its just a case of learning 'em all

[miniGOINGS]

Hey Cride, I would reaaaaaaly like to help make the Wiki better, is there anything that needs to be done?

[Cride5]

Hey Cride, I would reaaaaaaly like to help make the Wiki better, is there anything that needs to be done?

Certainly! If you're knowledgeable about a particular topic, or have algs to contribute, its all of value. A good place to start looking is the list of stub pages here. They're all in need of a little care and attention

[miniGOINGS]

Awesome, I still have yet to make an account but I might get to work on that.

[jazzthief81]

I've added some algorithms to the database, namely all cases that can be solved with a combination of 2 Sunes™:

• 26 cases can be solved with two Sunes™ with a U move in between (15 moves)
• 6 cases can be solved with two Sunes™ that cancel 1 move (13 moves)
• 4 cases can be solved with two Sunes™ that cancel 3 moves a.k.a. double Sune™ (11 moves)

I was surprised to see that quite a few were shorter than the ones already in the database. It seems the current ones are optimized for a fixed starting orientation, is that right?

I also noticed that when you add an algorithm with a leading U move it gets added to the algorithm and the move mount, but if you omit the leading U move, the script automatically adds a y rotation and it doesn't count as an extra move.

Would it be possible to reprocess the existing algorithms so that the leading U moves get converted to a rotation? Also the trailing U moves shouldn't be counted really but they do now. That way the move counts are a bit more representative and comparable.

[Blake4512]

That must of taken a long time to do.
Great job =D

[eastamazonantidote]

I've added some algorithms to the database, namely all cases that can be solved with a combination of 2 Sunes™:

• 26 cases can be solved with two Sunes™ with a U move in between (15 moves) Didn't keep track of this. Probably should've...
• 6 cases can be solved with two Sunes™ that cancel 1 move (13 moves) I tried to get these. One and mirror for T, one and mirror for U, but the other two I don't think I had up.
• 4 cases can be solved with two Sunes™ that cancel 3 moves a.k.a. double Sune™ (11 moves) I think I got those already. These are the H set, correct?

I was surprised to see that quite a few were shorter than the ones already in the database. It seems the current ones are optimized for a fixed starting orientation, is that right?

I also noticed that when you add an algorithm with a leading U move it gets added to the algorithm and the move mount, but if you omit the leading U move, the script automatically adds a y rotation and it doesn't count as an extra move.

Would it be possible to reprocess the existing algorithms so that the leading U moves get converted to a rotation? Also the trailing U moves shouldn't be counted really but they do now. That way the move counts are a bit more representative and comparable.

Excellent work. Yeah, I had issues with the optimal solver. I set it up with the corner in URF correctly placed and worked off of that, so there were some issues. For the most part, I let Cube Explorer run about 4 moves too long before making a decision. But by Sune and Anti-Sune, I was exhausted. Even now I struggle to find time. However, your additions are excellent and I will also add them to my printable pages (at some point...). I didn't include AUF after the algorithm becuase I use Harris/Baum recognition, which is independent of the AUF. It also makes the algs seem shorter

As for the leading AUF, that is a huge debate. In the end, I decided that after getting the cube into the COLL recognition position, you wouldn't want to perform a whole cube rotation. Thus I went with U/U' turns. I still wonder about this one. On the optimal sheet (I'll get it out sometime you guys!), these will definitely be cube rotations because that saves moves. But for speed, I decided U and U' moves were more acceptable.