Saturday 18 June 2011

Jerry McFarland's Quadlock 1

Quadlock 1
Jerry McFarland is one of the most amazing puzzle makers out there! He mostly makes burr puzzles using equipment he has developed himself (including a computerised automatic notching machine!!) The quality of his finished work is nothing short of stunning. He is the actual manufacturer of Bill Cutler's puzzles but he also has designed quite a few of his own. I have been in email contact with him for a few months now as we swap back and forth ideas for puzzles (and tales of excessive spending by spouses - yes, my wife has as much of a shoe and handbag problem as I do with puzzles!!)

I was very disappointed that at the last BaxterWeb auction one of his original Quadlock puzzles went for rather a lot more than I could afford. He told me that he was retooling his jig to make them again using smaller stock. I asked if I could have one when it became available and he, like the gentleman he is, was true to his word and let me know about a month later when one was finished. It arrived this week and despite being pretty ill with a very virulent bout of food poisoning, I couldn't resist picking it up and having a go!! Jerry describes it as "a 19 piece difficult to take apart flattened cube." The previous version was 3.5" x 3.5" x 2.6" and the new one is 2.9" x 2.9" x 2.2" so it is still a nicely substantial puzzle and made from Walnut, Mahogany and Maple all polished to a beautiful sheen as always. The fit of all the pieces is fabulous. When I showed it to the missus even she admired how beautiful it looked!

When you first pick it up there is no clue at all how it will open. The 4 walnut central pieces are the only moving parts and they each can move in either direction and don't seem to interact with each other. I was fairly mystified for about 10 minutes whilst trying as many possible combinations of positions and pushing everything else I could. Apart from the occasional very slight movement there was nothing shifting significantly. At this point I decided to read the leaflet he sent with it!! Jerry always sends very nice puzzle documentation. I didn't want to read the solution so I kept my eyes off the pictures and just read a few sentences. On page 3 he describes "using a lock-picking technique" and I stopped reading there.

Now, as a card carrying physician, I can honestly say that I have never picked a lock before! But locks fascinate me, so I have learned just enough about them to know what Jerry means. I went back to the  puzzle and had a small play about. After another 3 or 4 minutes I had a breakthrough and something moved. This allowed the removal of the first 3 pieces. It was made harder by the fact that one of the squares is a false piece made to match the others - it is just a thin tile rather that a whole stick and the grain even matches the other pieces!! Sneaky that!! Once those 3 pieces were out, I struggled to find the next piece removal but on sliding sticks back and forth it soon became clear where the next pair came out. After this it came apart into 2 separate assemblies and then rapidly to pieces.

Quadlock 1 pieces - notice the fake end caps with grain!
I had paid some attention to the order but deliberately chose to muddle them up. As long as you understand the process, reassembly is not too challenging. It is, however, quite easy to get a few pieces the wrong way around - it looks OK but when trying to insert the next, it won't slide far enough to let it happen. A bit of investigation revealed the problem and I remedied it. The reassembly took about 15 minutes the first time.

Jerry doesn't have time to make Burrtools files for all his puzzles so I promised that I would make one for him. I absolutely love this aspect of puzzling so didn't mind doing this for him. It really took some doing, though!! Initially I worked on a 16x16x12 grid because I thought that he had made ¼ piece notches. This was fairly time consuming and, of course, when I reached the last 2 pieces to be entered I discovered that these had ⅛ notches!! OMG!! For a horrible moment I thought I was going to have to start again!! Luckily Burrtools has a button to double up the pieces in every dimension. So I ended up with a 32x32x24 grid which was real fun to add the colours to(!) but at least I could get these last 2 pieces done. After colouring and double and triple checking I set it to extract a solution and...

Burrtools crashed on me!!!!

I tried on a different computer. There was no warning, it just died on me! I got the standard Mac OS error report which is gibberish to the non-programmer (I have not programmed since my days of BBC Basic or 6809 Assembly language in 1983!) Jerry tried the file on his computer and it also failed. I sent off an email to Andreas Röver just to let him know and he responded within 2 hours!! Amazing! Apparently it fails due to the shear size of the puzzle and I didn't have enough memory - with 4Gb and shutting down all other apps he produced a file with the solution(s). If you want a copy of the file then just let me know and I can email it to you. Unbeknownst to Jerry there are 12 other solutions - they are only slightly different from the original but require different lock-picking positions.

In summary this is not a particularly difficult puzzle but it is absolutely beautiful (as well as beautifully made) and involves a very different opening technique to everything else I have seen. Jerry is starting to make a few of these and they should be available from his site in July. Buy them - you will not be disappointed!


  1. I'm surprised, as BT has worked for me on some quite large files. I have one here on a 51x51x21 grid. Bur I suppose memory consumption is related to the size of the solve tree, not necessarily the grid size.

  2. I'm not sure whether that's the case, George. I would expect that the size of the solve tree only causes a very long duration of solution. The app actually falls over before getting that far - it crashes during the "optimising pieces" stage. Andreas said it was a problem with available memory - I onl have 2Gig. He struggled with 4! I will email you a copy to try for yourself if you like.

  3. That's OK, I have 4GB as well and I'm sure Andreas knows best ...

  4. Actually, I think you should be fine defining this puzzle on a 16x16x12 grid. It is true that two of the lock picking pieces are on a finer grid, but because of how the solution works, the coarse grid should capture the workings of the puzzle. Also, I think the "extra 12 solutions" are just the same solution with the lock picking pieces moving in a different order, so really just the same solution.

    By the way I purchased this puzzle for myself and it is fantastic. Excellent workmanship!!

  5. You are right, George. It will work with the coarser grid but then the 2 asymmetric lock picking pieces become symetrical or very offset indeed. This is due to having far more limited choices about where a notch goes.

    I haven't properly explored the alternative solutions. Some are due to order but it des appear that the position changes too (the heights of the lock tumblers is slightly different in some). It is hard to really tell because my computer struggles with the file.