## Description From https://gitlab.com/gitlab-org/gitlab-ce/issues/59023#note_160196362 > We already store the initial SHA of the last known commit in the IDE. Which is either the commit the IDE is loaded at or the last commit done in the IDE. > > > should we also pass that SHA (instead of the branch name), when loading the tree and specific blobs? > > Yeah good point :thinking: We currently just request the file content from `/PROJET_NAMESPACE/raw/BRANCH_NAME/FILE_PATH` so I guess we would need to do `/PROJET_NAMESPACE/raw/LAST_COMMIT_SHA/FILE_PATH` instead Let's update our requests to the Web IDE to use the commit sha instead of the branch name. ## Why? It's possible that the branch is updated while the IDE is open. This would lead to strange behavior if a file is loaded and we pull the most recent changes (but the rest of the content is based on my current Web IDE HEAD).