Remove thorn CactusArchive/ADM from thorn list. This thorn is outdated, and we should updated out test cases instead. It also takes a long time and a lot of memory to compile.
If we want an ADM formulation (which is doubtful since we don't use it ourselves), we should implement one via Kranc.
Keyword:
I commented out CactusArchive/ADM in our standard thorn list. This reduce the number of test cases I could run from 125 to 100.
Should we remove this thorn now?
How many of those tests were tests of the ADM thorn, and how many were just using the ADM thorn to test something else? I would rather not remove test cases. How hard would it be to convert these tests to using the McLachlan BSSN thorn?
Available/runnable tests decreases from 176/125 to 172/100 when disabling thorn ADM.
So there are 4 thorns which test ADM, plus 21 tests which need ADM but are not part of it. I would keep ADM around for now, and eventually convert those 21 tests to use ML_BSSN. We can do this at the same time as we convert the tests which need BSSN_MoL. Actually, would it make more sense to make these tests depend on something more basic, like a WaveToy thorn?
It would, if that is possible for that test case. I am removing 'review' status, as we do have a consensus what should happen, but now someone has to do it.
This was done after ET release 2013_11 "Noether". Specifically on 12/13/2013 http://lists.einsteintoolkit.org/pipermail/users/2013-December/003380.html .
I agree. In fact, I compiled it once with --debug, and it took so much memory my laptop swapped itself to death and had to be rebooted. Definitely should be removed. Don't we have an ADM thorn in McLachlan anyway? If not, I'm pretty sure I have one from a while ago which could be adapted.