On 12 Jan 2026, at 14:37, Roy Mendelssohn - NOAA Federal via
R-package-devel <r-package-devel at r-project.org> wrote:
For what it is worth in my case as far as I can tell the behaviors of
both devtools::submit_cran and the cran machine were quote reasonable,
even if it caused me a little pain. The block Uwe describes sounds like
fail2ban or the like, and given the junk we see hitting our services it
is a reasonable thing to run. I had the day wrong when I thought cran was
back up. devtools on submission said ti couldn?t connect, It is not
unreasonable as a submitter, thinking the machine was back up, to wait a
bit, and try again, and maybe do so a few times throughout the day.
That is indeed what I did. Then to be sure it wasn?t a devtools problem I
went to the web page and that is when it seemed to block me.
So my point is everything to my mind behaved reasonably. The only thing
I could suggest, and I have no idea of how much work this would entail,
and I am not looking to make work for people, is perhaps a banner on the
webpage when the service is down,
That?s the thing - there *was* a red banner that said no submissions are
accepted so everyone using the official CRAN way was fine, it?s the
back-door access by devtools that caused the problems. The issue is that
once devtools got you blocked, it was too late so you couldn?t go back and
see that you were not supposed to use it since your IP was already on the
blacklist even for the official front-end. Clearly, devtools needs to be
fixed to make sure it?s not submitting things blindly when it?s not
supposed - probably with some help from the CRAN admin team to make sure
there is a well-defiled path for this.