ka9q
commented
5 years ago On 12/6/18 15:49, Jan Schiefer wrote:
Phil manages his code in repositories stored on his personal web site. We need an automatic mechanism to mirror code changes into this repository.
The tentative plan at this stage is to fetch changes periodically using a cron job and adding them to a branch in this repository. Merging will be manual initially, as we don't have any CI builds or test suites yet.
If github has some sort of bare-bones direct rsync interface I haven't heard about yet, I would be willing to use it to push over updates. The more offsite backups I have, the better...
The great thing about rsync, besides being so versatile and easy to use, is that (permissions aside) it always succeeds in creating a complete, up-to-date remote copy regardless of its previous state, if it even exists. Nor does it pollute my local copy with ever-increasing amounts of state from the beginning of the universe.
To me, engineering (especially programming) is a big game of Tetris. I have to keep throwing away old stuff before I get overwhelmed by the complexity that otherwise increases monotonically. When I can't, I know something is wrong and it's time to rethink my approach.
I often get more satisfaction from being able to throw away something I wrote as obsolete than when I originally wrote it and made it work.
Phil
Phil manages his code in repositories stored on his personal web site. We need an automatic mechanism to mirror code changes into this repository.
The tentative plan at this stage is to fetch changes periodically using a cron job and adding them to a branch in this repository. Merging will be manual initially, as we don't have any CI builds or test suites yet.