Yes, good. After reading the comments on their image retention limits [0] talking about simply pulling the images all the time to keep them fresh, this seems like a reasonable response. I'll repeat what I wrote there [1]: If people really think this is a problem, they'd contribute a non-abusive solution. Writing cron jobs to pull periodically in order to artificially reset the timer is abusive. Non-abusive solutions…
This is a great idea in concept, but in practice very challenging.
RUN curl https://www.random.org/integers/?num=1&min=1&max=99999
Docker will cache this after the first invocation. The build is not reproducible. Now what?
Replace "curl random.org" with "nondeterministic and really expensive code build/model training/etc operation".
> extending docker push and pull to allow discovery from different sources that use different protocols like IPFS, TahoeLAFS, or filesharing hosts
This is great, if you can solve the image integrity/trust issues therein, which should be just some signing/merkle tree work.