Dear Job,
On 9 Feb 2022, at 10:45, Job Snijders via routing-wg <routing-wg@ripe.net> wrote:
Congratulations on this accomplishment and achieving this milestone!
Thank you.
Am I right in assuming that - going forward - commits won't be squashed (more than needed)? I imagine it'll be educational for the community to be able to follow the train of thought and storyline of future developments.
We have chosen to squash our internal commits on publication. The main reason for doing this is that the public repository is a completely different tree from our internal code tree. While we could transplant (rebase) individual commits to the new tree, we cannot transplant merge commits. Hence we would risk losing changes in merge commits (often resolution to merge conflicts) and break the published code. We do see value in giving the community access to the code as well as the commits that lead to it. Therefore we add the titles and internal refs of all included commits in the body of the squashed commit. An example of this is in [1]. We do release frequently. Therefore most patches between releases are small and contain a limited number of changes. These patches are often incremental and show the direction we are following with development. We hope this helps build an understanding of why we choose this publication scheme. [1] - https://github.com/RIPE-NCC/rpki-core/commit/2d83b6aa197cffe7974391dbafb439e... Kind regards, Bart Bakker RIPE NCC