Workflow for fetching new Gerrit patchsets

[jakubsmolaga]

Hi, I recently started using jj at work and I think I might be a little bit confused about dealing with Gerrit. Here is my current workflow:

1. Colleague pushes a new version of a change
2. I run git fetch ssh://<...> refs/changes/<...> && git checkout FETCH_HEAD
3. The change appears in jj but it has a different jj change id
4. Rebase all children of the old change onto the new change
5. Abandon the old change

Is this the correct way to do this? If not, what would be the idiomatic way of getting a new patchset? Thanks :)

[jabielecki]

Tough. I have a related question, which seems even simpler. When:

• I look at a Gerrit view of my change, and
• I know that the change is a result of my own jj gerrit upload
• ...how can I switch my jj back to it: jj edit <what>?

I've made an extremely ugly browser plugin to alter Gerrit html (i.e. not server-side but browser-side). How do others deal with it? Appending a real git trailer seems somewhat superfluous, given there's already a git header for storing exactly that. It's just the git header is not shown in Gerrit UI.

🤔

Screenshot from the https://addons.mozilla.org/en-US/firefox/addon/gerrit-jj/ - I only access Gerrit with Firefox, so it's just enough to scratch my own itch.