Archive DOI not validated

[sneakers-the-rat]

Re:

• https://github.com/openjournals/joss-papers/pull/5158
• https://github.com/openjournals/joss/issues/1324

@sdruskat found two papers which had incorrect archive DOIs (nice catch!),

• https://github.com/openjournals/joss-reviews/issues/4684#issuecomment-1267210460 looks like a copy paste error
• https://github.com/openjournals/joss-reviews/issues/5883#issuecomment-1848671737 looks like a lil confusion set in <3

The responder is here: https://github.com/openjournals/buffy/blob/joss/app/responders/openjournals/set_archive_responder.rb

and there is already a valid_doi_value method - it seems like we should probably at least check that the link doesn't 404, right? Would additional validation be useful or overengineering? Since we allow any archive to be used, and the archive might also not share the title with with the reviewed work, it seems like it might be hard to validate that "this archive link is for sure the right one"

some ideas:

• query eg. crossref for the archive metadata, prompt with a warning when the title is >threshold different than the submitted work title
• add an item to post-review editorial/author checklist that is "ensure that the archive link is the right thing"

thoughts? i'd be happy to PR whatever we decide here :)

[sdruskat]

Thanks for pinging me, @sneakers-the-rat.

I agree that doing some checking of DOIs would be great.

Point in case, I just found that the software_archive for https://joss.theoj.org/papers/10.21105/joss.00041 leads to a paper about a species of bee 🐝 from 1910, written in French 😆. This is as far as it gets from a software archive I believe 😉. (The correct DOI would have been https://doi.org/10.5281/zenodo.61965).