@mhchem the triple bond there has the same problem :(

@pentavalentcarbon woah, that fixes it!

I wonder what's the overall pros/cons of using the different renders

@orthocresol thank you

Any computational lovers: has anyone ever experimented with code generators to form discrete expressions for recurrence relations?

I'm sure the lack of for and if statements will reduce the FLOP count but it makes it more difficult to extract intermediate terms via recurrences

@obackhouse Some of your favorite programs use machine-generated code. libint (Psi4, ORCA) generates files for Obara-Saika RRs, then compiles them. Q-Chem has MGC for MD-PRISM and HGP-PRISM.

I think it's fully unrolled in Q-Chem.

ooh I see. I'm trying out some generated code for evaluating the Hermitian auxiliary integrals, along with a Boys lookup table

The higher order angular momentum terms get very bulky, though.