Closed moverton000 closed 3 months ago
Hi Madeleine, a bunch of tests were failing because of a network issue, so I just re-ran those tests which are now passing. The only one left is a "real" failure:
make SETUP=randomwind checking phantom setup runs, creates .setup and .in files with no unspecified user input
Checking randomwind (setup)... OK
compiling phantom with SETUP=randomwind
phantomsetup exists
runs
creates .setup file
creates .in file
phantom exists
FAIL: ./phantom myrun.in fails with error
The test here is that you must be able to compile phantomsetup, run ./phantomsetup myrun, press enter a bunch of times and then run ./phantom myrun.in and have it succeed in replacing the _00000.tmp file with the _00000 file. This is currently failing for some reason with SETUP=randomwind...
Hope that helps!
Hi Daniel, I believe the problem is that SETUP=randomwind uses setup_disc which by default only has one sink, but inject_randomwid requires 2 sink particles. So although they can be used together you have to manually specify 2 sinks in the setup. I don't think the defaults for setup_disc should be changed and since the purpose of this pull request is just to rename inject_asteroidwind.f90 to inject_randomwind.f90, I think the best solution is just to remove the SETUP=randomwind option until inject_randomwind can be updated to handle only one sink particle.
Type of PR: modification to existing code
Description: The inject_asteroidwind file has been changed to no longer use the term "asteroid" since it can be applied to general object with winds. Used as a pull request demo for the phantom users conference 2024.
Testing: Phantom was compiled successfully.
Did you run the bots? no
Did you update relevant documentation in the docs directory? no