Sam Yates

Just saw the closing; I agree. This should be outside the scope of the core library; ultimately we could provide something in the extension library, or maybe even in the...

I'm not really convinced that we need to modify the C++ API. It's flexible enough to support any number of variations in the Python API, and was designed to support...

Just to expand a bit, the idea with probe return values is that they constitute a contract between the cell type and the sampler: the sampler expects to get the...

An alternative would be to have git enforce a LF-only ending on scripts such as `git-source-id` — would this avoid the issue? Using [GitHub's line ending config docs](https://docs.github.com/en/get-started/getting-started-with-git/configuring-git-to-handle-line-endings) as a...

My feelings about this: * The dt provided to `simulation::run` should be regarded as firstly a maximum dt and secondly as being up to cell group implementations to interpret in...

My chief reason to believe that the performance would be poorer is that we pay the cost of extra kernel calls for each copy of the mechanism, and each copy...

We don't have any data on the frequency of such models, unfortunately. @bcumming brought up an example of spatially variant leak current reversal potential in chat, but I don't know...

Would this be something that might better live in `arborio/`? Especially if down the track we're looking towards pretty-printed representations of e.g. morphologies too.

I've always been against it as you know, primarily because it adversely affects readability of mathematical expressions. But I also acknowledge that my opinion should carry little weight given that...

Ahah! I will take a closer look at uncrustify — my concerns may well now be moot.