Can I run fossilize manually? about fossilize

[kakra]

You probably want to skip all the fd related options, those are created by the Steam client to communicate with the processes. Also, maybe skip quiet-slave while testing and look for obvious problems.

[letharion]

Oh I see, that explains the missing file descriptor. There were plenty of options passed, so I didn't notice them there. I should of course have realized that's how the missing FD was passed in.

Dropping --shmem-fd --control-fd --quite-slave fossilize does indeed run.
It yields loads of

Fossilize WARN: Did not replay blob (tag: 7, hash: 0x31a65b0417b7c55f). See previous logs for context.
Fossilize WARN: Referenced Pipeline layout 1443983c1d9866f5, but it does not exist.
This can be expected when replaying an archive from Steam.
If replaying just the application cache, make sure to replay together with the common cache, as application cache can depend on common cache.

But maybe that's not an issue, as it says This can be expected. I'll try this out for a while, gonna need some updates to the game to see if it works well.

[kakra]

@letharion Be aware that IO/CPU pressure auto-adjust may not work when running directly from command line or cron... I think --master-process did the trick for enabling background mode but I'm not sure. If auto-adjust works, you may want to bump up the thread count.

That's for replaying fossils, but how do you record fossils manually though?