The goal is to be able to slide your palm about 1 Inch in all directions...

If that is the goal then the lower part must be bigger by 1 inch in all directions compared to the upper part. The pictures do not indicate that at all.

Also the picture indicates that the ball sits in the upper bearing depression:
1) if the depression is simple spherical cutout then you will get sliding friction there and it is likely to be too much for smooth motion (even if it is made from a low friction material like e.g. teflon)
2) if the depression is a groove (shown from the direction of the groove ... therefore seen on the picture as a simple spherical cutout) then it will allow easy movement only in one direction (the direction of the groove)

If you want to have a friction less feeling with balls then the balls must be able to move as well. E.g. to move the pad to the right by one inch the ball must be able to move to the right by half an inch! This is true for any direction.
You do not have any indication in your pictures that you allow for this.

Look .e.g. at this exploded view of a linear bearing:

Notice that the bearing balls are there in "squished" loops. The balls mus be in loops since one side of the loop provides the return path for the moving balls. The other side of the loop actually is the working area. This is the area where balls provide friction less movement between the carriage a the rail. They provide the friction less movement by rolling between the rail and the carriage surface. But because of this rolling they are also moving along the loop. Therefore they need the return path. The balls are exiting the working area of the loop ... returning through the return path  and entering the working area on the other side.

Balls move through the working area without friction since they are rolling. This is also the area where the balls are under load (possibly high forces between carriage and rail). The balls are not moving without friction in the return path. But the friction is very low there since there is no load on the balls.

Your design does not address the moving ball problem at all. That is the reason I pointed out other options (magnetic levitation, ...).

MW Very Harry

Its not Harry Potter themed, its medieval themed, with elements from FMA - Fullmetal Alchemist...

The new draft keyboard label design has been completed!

While the current keyboard label was based on the IBM 4704 keyboards, the new one is based on the 1984-1986 IBM Model F and Model M labels (now that the 4704's are no longer the main project boards), but with some changes to make the font slightly thicker, etc.  Just like with the keycap characters, these characters were drawn based off the originals, but this time I made it into a font as needed for label creation.  It is my understanding that the originals were made not with a dot matrix, laser, or offset printer as was likely the case with other IBM keyboard labels, but instead with some kind of line printer, impact printer, or chain printer which involved the striking of the metal key into the paper label through a ribbon (which is why many of the characters were missing sections of ink or slightly varied in their thickness and spacing).  Some of my original IBM Model F122 labels even have typos, probably from a key striking twice (a date code of MAY844 instead of MAY84 for example).  I can't find any records of the IBM line printers that could handle large printing of about 1/4" line height, so my guess is that another company's printers were used.  The keyboard labels will be covered with a slightly oversized clear plastic label protector as IBM did for this series of labels.

You can't request the new label, replace your existing label, or check to see which label you are getting; there are still many old labels that are already printed and will be used first.  For simplicity sake the new designs will also be used with the F62/F77 boards.  The box labels will stay with the same design.

Please consider adding R1 pipe to the alert kit in order to support tsangan layout. Thank you!