Karl Kosack

in case it's useful as a good (or bad) example, we defined a custom Path trait here: https://github.com/cta-observatory/ctapipe/blob/2e3a5744a1fe87f21f89571f7a05754183dc85b1/ctapipe/core/traits.py#L82

the points you are plotting don't seem to be correct (at least they look shifted wrt the publication). These are what they should look like:   It's likely that...

You may also be able to get away with far less bins if you choose a polar coordinate system for this instead of DETX/DETY, like (R, phi), since to first...

I suggest using the terminology `true_x` rather than `mc_x` which is jargon.

> why it was decided to use binary table extensions with a single row instead of image hdus for the irfs. I think also it's a question of binning -...

Just a comment about the units/quantities: it may be a bit more "VO" than FITS, but UCDs (uniform content descriptors) are designed exactly for that purpose: they allow you to...

Might consider just using the keywords that OFWG already recommends: The OFWG recommends that the following set of keywords be used to for such purposes in each extension: * HDUCLASS...

I agree with Jurgen. You don't need an HDUCLASS in the primary. Couldn't you just identify the file type by which HDUs are present? There's also the hierarchical grouping standard...

Does the per-telescope deadtime really matter? Isn't it the overall deadtime as measured by the SWAT (central trigger software) the more important quantity, which is a sort of "average" or...

Just a comment: it is looking very likely that with CTA we will drop the concept of "runs", meaning that in a particular observation there may be events that are...