Hi Andy,

Data like this is definitely very valuable. I'll shy away from the philosophical debate about openness/freedom, and just comment on the subscription itself. When I first came across this, I must admit that I didn't find USD1500 "shockingly small" (e.g. it's quite a bit more than an individual academic licence for MATLAB). However (at risk of opening up another debate) I think what you are offering is *far* more valuable than what journals offer in exchange for their gold open access article processing charges (which are often twice that price), and if funding bodies are happy to pay these (in the UK, some seem positively enthusiastic to part with their cash) then I think they should be more than happy to pay your subscription.

I wonder though, whether you could offer more flexible schemes. E.g. not everyone will need 60+ scans for what they have in mind, and many potential users would only be interested in either end of your 5 - 96 year age range, and in these cases, it might seem bad value to pay for data that won't be used. How much do you think you would need to charge on a per scan basis? As another alternative, what if someone only wants e.g. left temporal lobe regions, would 60 subjects with only those labels be cheaper?

If I've understood correctly, the subscription is an annual charge. This seems a slightly weird model to me, in that I would try to do whatever I wanted in the first year and then stop the subscription. Even if I could benefit from more scans, I don't think the value of a second year would be the same as that of the first. Similarly, if I subscribe this year, let the subscription lapse next year, but then a few years later (e.g. after you have added some new feature/data that I really want, like hippocampal subfields, brainstem nuclei, etc.) decide to pay again, could there be some kind of reduced-fee "upgrade" option?

I hope those thoughts are of some interest, they're somewhat off-the-cuff, so I apologise if any of them don't withstand more careful consideration!

All the best,
Ged