On 13.01.2024 09:58, Daniel Sahlberg wrote:
Since there wasn't any replies to this and I think the code was working fine in all my tests, I comitted as r1915214. Although I finally decided to solve the spurious revert messages in a different way, see a separate followup/committ e-mail.
I think you just made the code more complicated for no good reason. The situation you described and fixed applies only to working copies that are shared by different users.
You also have to ask yourself what happens when your user runs an "svn update" in the working copy where it's not the owner of the files and when your primary group doesn't have write access to the containing directory. In other words, I think you just fixed an edge case that has lots of other problems, not least that without using at least 'chgrp', the working copy will be either useless for you or quickly corrupted for others. I'm not even going to try to think about what happens to SQLite WAL files. Your exhaustive tests don't seem to cover these cases.
In practice, such working copies tend to be used by prefixing all svn commands with 'sudo -u wc-owner' or starting with 'sudo -u wc-owner -s'. That's a solution that always works and doesn't require more code in svn that someone has to maintain and that only solves one very specific edge case.
-- Brane
