Were you thinking of something you could request at any time, or a daily/weekly backup available to the public?

Depending on how much of the data you want at once (everything or something narrower like just the comments for a particular question) there may be some overlap with work on the requirements for a Codidact API. If you want to add an answer there too, you're still early as implementing it hasn't started yet.

Not suggesting a duplication - I can imagine both an API and data dump.

Thanks trichoplax‭, regular backup available to the public

Quarterly would probably be enough at the current scale codidact operates at. Monthly might be nice once post volume increases, but anything more than that is arguably unnecessary regardless of volume, especially considering the computational requirements increase as database size does -- especially for compression, and especially if you can't multithread it. For the record, Wikipedia (who has an insanely good data dump system¹) does 1-2 per month with no real upper limit, with one per month as the minimum and two per month as a usual practice.

1. seriously, poke around their data dump site and documentation - they even include logs for each site. This would be absolute overkill for an initial implementation of course, but worth keeping in mind for Some Point:tm: in the future ↩

If we're being futuristic and all, we might even get to include images in the dumps! :O

Note that an image data dump would be significantly larger due to compression stuff, so an image data dump has very different considersrions, and drastically higher upload requirements

Good point. I guess even if we had an image data dump, we'd want to make sure that the much smaller version without images is still available for people who don't need the images and have a limited download speed / monthly limit.

trichoplax‭ yes, can release separately, eg https://github.com/Franck-Dernoncourt/Stack-Exchange-Image-Dataset