Deprecating and ending support for software

[standage]

In my most recent run of this workshop, during the discussion on hosting (and specifically link rot associated with departmental servers), one of the attendees contended that this is not actually a bad thing. People move on, project maintainers no longer have time to put in to a project, and the software is going to be irrelevant in a few years anyway. He said something like "there's only so much you can do with 0% time on a project."

I sympathize but totally disagree with him. Career progress, changing affiliations, and eventual obsolescence of most software is no excuse to not archive responsibly. Perhaps we should add something to address this: to make sure that there is should not be an expectation that someone maintains software indefinitely after funding runs out, but that we still can and should do what we can to preserve the work that was done.

[standage]

For example: some scientists explicitly mark software packages as deprecated, without explicitly closing them down or restricting access, or letting them dissipate into the ether.