On 05/12/2017 12:49 PM, Andreas Schwab wrote: > On Mai 12 2017, Rob Landley <r...@landley.net> wrote: > >> Last I checked I couldn't just "git push" the fullhist tree to >> git.kernel.org because git graft didn't propagate right. > > Perhaps you could recreate them with git replace --graft. That creates > replace objects that can be pushed and fetched. (They are stored in > refs/replace, and must be pushed/fetched explicitly.)
It's the "must be pushed/fetched explicitly" part that I couldn't figure out back when I tried it. I inherited this tree from somebody who made it. I noticed its existence because lwn.net covered it, and then 6 months later it had vanished without trace (as so many things do). I reproduced it from the build script (if you can't reproduce the experiment from initial starting conditions, it's not science), went "look, cool thing", and hosted a copy with an occasional repaint. I would be _thrilled_ to hand it off to somebody who knows what they're doing with git. I'm just unusually interested in computer history and the preservation thereof. (https://landley.net/history/mirror). Rob