"opt-out" questions
"opt-out" requests have been advertised for several days now. Supposedly they will be honored with the next "patch release".
When is that? There are close to 4000 opt-outs pending, with seemingly not a single one having been worked on since the method was advertised.
Additionally, how is illegally copied material handled? (see #5) Will all previous versions of this dataset be removed with the next release to avoid reproducing them?
Think: repositories that were removed per DMCA since they were scraped, repositories that were included despite that being explicitly against the license, repositories from people that have chosen to "opt out", and so on.
Additionally, how are individual contributions handled? If you're adding an "opt-out" mechanism, why does it seemingly only apply to the repository level.
Personally I would want every single commit authored by me or one of my associated identities to be entirely, permanently, and retroactively removed from this dataset.
Some of those commits cannot be traced back to me by any easy method other than manually filtering through my old data, since they were made with incorrect or varying identities in git, which are not publicly linked to me in any way. What mechanism can you offer to apply this "opt-out" for that?
Did nobody think to ask these questions before publicizing a multi-terabyte blob of data or was it willingly ignored?