Post #2050774

@drj For data we have things like zenodo, which are there for the long term presevation of results and now you can link a github code repository to zenodo and it should get long term preservation. Obviously this further cements github, which is a bad thing, but just trying to think of what the current "best" practice is: For example, we did this repo here as required for a publication that just came out: https://zenodo.org/records/18712812 It's a fairly crap rendering of the repo - no README or such, but at least I trust Zenodo more than github to preserve the method implementation, and to be frank also trust it more than university IT (in the UK often now a microsoft shop), unless it was a problem handed to the library, who actually understand long term archival requirements. Now that would be glorious. Perhaps I should just print it and bind it and get an ISBN for the code, then the library system would have to store it :D