There is one project on Language Forge that has had support requests several times for merge problems. At least once, the only problem was that creating the merge commit took too long. Chorus has a default timeout of 15 minutes for local operations (hg commit, hg merge, and so on) and on the most recent Send/Receive of that project (2025-06-16), the merge commit took 23 minutes and 3 seconds to complete. It failed the first time, but I was able to make it succeed by setting the CHORUS_LOCAL_TIMEOUT_SECONDS variable in /usr/share/lfmerge/7000072/environ. That edit was live on the container, however, and will be overwritten the next time the lfmerge container is deployed.

We should edit the environ file in this repo and set CHORUS_LOCAL_TIMEOUT_SECONDS to a higher number than 15 minutes. I suggest 60 minutes, 3600 seconds, as a good compromise between allowing large projects to continue merging while not allowing a runaway merge process to lock up the server forever. That way the next time lfmerge is deployed, this large project will continue to Send/Receive successfully.