Hi José,
Thanks for the file, always nice to have that extra information.
I presume that the “Surface” option is your target distribution.
There is no option to sort random clones based on their Y position first, at least not natively. (Edit: Here I was wrong, see post #3/edit) This could be done with an Python script. (Not my scope here, but in short: get the Matrix of each clone with its ID in a list, sort that list based on the P.Y for example, then use the position in the list to request actions.) That would be perhaps a nice entry in the Tutorial Suggestion Forum.
Please have a look at file below:
Scene file
https://www.amazon.com/clouddrive/share/CyZgQRFjARvRIv1dzQ0NkU3TjWGnajJ5GuS2LbLanTX
This works like the Thinking Particle Node PBlurb. It worked like this: two groups of particles with a random amount of particles, as well as randomly placed, move from surface A to surface B. The “trick” was here have two set ups, and to blend the particles while they are kind of in the middle of their path from A to B. In other words, one group starts invisible, where the other group turns on the end invisible. In the file above, I have not implemented the “blend” to keep it simple. The blend could be done with an gradient in the Alpha channel and the gradient starts with object A and ends with object B (3D linear).
I used the uvw space of the clones here, the v is like a sub Y space, and so I triggered them. Since that would create only a sudden jump of the clones, the Delay Effector smoothes the travel. The Delay Effector is responsible for the duration of the traveling.
All the best