Amar Takhar commented on a discussion: https://gitlab.rtems.org/rtems/rtos/rtems/-/issues/5294#note_126175 Yes this is how FreeBSD has been doing it for 40+ years now and it works amazingly well for them you can see the entire history. We don't update often enough for it to e a problem and what's gained by having an internal, linear history is massive. Also you can create a new branch from the tag and cherry pick the previous commits in. If it's a struggle that means you have too many local changes and either need to upstream, rethink the changes or separate them into their own file. Updating 3rd party code is only painful if you've made it so. As a policy we won't use submodules in RTEMS and only sparingly elsewhere when they are helper repos. They cause more issues than they solve. Having all the code local makes far more sense than having to reach out to the internet constantly to check for updates. There are several other reasons why it doesn't work well for us. With gitlab when we do update / mirror all our contrib code to https://gitlab.rtems.org/contrib/ We can reference the commit from our local instance which will always match the upstream hash since the repos are read-only. GitLab is fantastic at making these kinds of connections so we also gain nothing by using submodules. Thanks for the input it's great to have these conversations. -- View it on GitLab: https://gitlab.rtems.org/rtems/rtos/rtems/-/issues/5294#note_126175 You're receiving this email because of your account on gitlab.rtems.org.
_______________________________________________ bugs mailing list bugs@rtems.org http://lists.rtems.org/mailman/listinfo/bugs
