[guylhem]

IMHO, you should start by offering to host the paper native source (whether latex or doc - I'm not a huge fan of PDF, especially when A4 printing issues arise)

Then, putting the matching data online would be a worthy 2nd step.

I have a "long term project" to put that on my own website but never did. Maybe one day when I have more time.

For some simulations I did, uploading the sourcecode and the various part so that my results can be replicated on the same page as the paper would be great - but it can get big (like in GB) and I'm not sure it would be cost effective for you to do that.

I'd even like to upload a VM with all the tools setup and ready (free software only so I'm fine) and the queries making my results as batch files along with other premade queries, but that would be even worse - the overhead of a distribution is big.

Yet making results more reproductible would be a great goal.

Good luck with your project.

[RichardPrice]

Thanks for these ideas. We host whatever version the user uploads; sometimes that is a .doc or latex file, but mostly it's a pdf.

Thanks for your perspective on data-sharing - interesting ideas.