Nick Hodgskin

> I think we should add padding as a kwarg, or remove the dep warning on `boundary` Which of these solutions would you like to see @jbusecke ? Happy to...

Talked with @jbusecke - Add padding as an arg - Internally in the code and rename thing to padding I'm happy to work on this

Also, I'm not 100% sure why a data array for `trajectory` is added after interpolation in the MCVE: ```diff Original dataset: Size: 2kB Dimensions: (trajectory: 10, obs: 10) Dimensions without...

Didn't know the context surrounding this function - agreed that I think it can be removed!

> we now pass a list of KernelParticles to the Kernels, instead of the entire ParticleSet *not a list of `KernelParticle` instances, but a `KernelParticle` instance where the indices is...

@erikvansebille is this the issue most relevant to the KernelParticle discussion? Also, I feel like there was another issue with test cases (repeatedly assigning to variables and that causing broken...

> I feel like there was another issue with test cases (repeatedly assigning to variables and that causing broken state) Found it https://github.com/Parcels-code/Parcels/pull/2199#issuecomment-3285278876

I think this problem runs deeper than a simple refactoring issue. I've have reached out to others to ask what they think - hence I have written this comment here...

> 2\. explain to users what the key changes are between v3 and v4 Many packages publish a migration guide document when doing a major release. I think that this...

To treat the migration guide as the 'source of truth' (and minimise duplication), we could do something along the lines of ```` ```{tip} The behaviour of ... has changed between...