While working on this question, I asked the submitter for their data, so I could play with it, and the answer was: fine, how can I share it?

I have no idea :p

Looking on SO, I saw this question, but it looks directed for people working together (as opposed to opening it on the Internet for anyone to download), and this one which is regarding making data available for an app.

This other question from meta is a request for a feature on Open Data.SO. It lists some options, but still nothing definitive.

Coming from an IT background, my first reaction for a smaller dataset would be to create a gist and upload the file. Not optimal, but it would make the data available.

How do you share data for a question?

If possible, different options for small files (several Kb to a few Mb), medium (several Mb to a few Gb) and big (several Gb) would be welcome.

Please note the question is different from this one, as I'm not asking for sharing within the site, nor where to find a specific dataset.

  • $\begingroup$ So you are asking where to go for free data hosting? $\endgroup$
    – Stephen Rauch Mod
    Oct 16, 2019 at 4:46
  • $\begingroup$ I think so, @Stephen. Perhaps specialized free data hosting, perhaps just free data hosting. By specialized, I mean akin to what github is for code or pastebin for text-only data: is there any service specific for dataset sharing? If not, what's the best general data sharing service for the purposes of sharing data related to questions in SO? $\endgroup$ Oct 16, 2019 at 11:47
  • $\begingroup$ You want to be clear whether "share data" means a snippet to reproduce an issue, or the whole dataset. Also whether that is KB, MB, GB, TB, etc. Should it be uncompressed or compressed (and which format)? Also (since people will likely use third-party sites and cloud-hosting) what should be our expectation about ephemerality: is it better to have a small KB snippet that will still be there in 1 month/year/decade, or a 1GB dataset that is inaccessible when the question is answered? And are there issues with licensing/(CC-SA forced) relicensing? (are users aware they're open-sourcing it?) $\endgroup$
    – smci
    Jun 13 at 21:57
  • $\begingroup$ If the dataset disappearing makes the question irreproducible, then what's our expectation about MCVE? And if a dataset is proprietary and can't be made public, is it ok for users to post "Contact me if you need access to the data"? $\endgroup$
    – smci
    Jun 13 at 21:58


You must log in to answer this question.

Browse other questions tagged .