David van der Spoel

@baoilleach I will be happy to help to debug this, or reimplement from scratch if you think that may be helpful, but I will need some pointers. Thanks!

> I ran obenergy in the debugger with breakpoint set -f kekulize.cpp -l 88. Basically, the routines in kekulize.cpp is called for the .log file not for the .sdf file....

> but in mdlformat: > 545 if (order == 4) { > 546 flag |= OBBond::Aromatic; > 547 order = 1; > 548 needs_kekulization = true; > 549 } >...

Curly braces for constructor calls are fine since at least C++11 and have some advantages: https://stackoverflow.com/questions/15396124/calling-constructor-with-braces

Most patches seem OK at a glance, although it is difficult to overview the if statement change. As regards whitespace changes, in principle I support an automatic cleanup and uniformisation...

@FMeinicke thanks for the instruction. Unfortunately NextCloud insist on autoupdating, but even without that a downgrade did not seem to help in my case. More tips appreciated!

Hm, the test failure seems to be related to code I haven't touched. Any suggestion on how to debug this?

Of course it was easy to find the official Martini 3.0 repository, but bow I found that the small molecules file there is different from the one provided here. So...

Thanks for your quick answer. I realise the gradient is provided by my function, but I find that either the function value converges to 0 before the gradients are sufficiently...

I'm using the standard double precision, the code is unmodified. What happens when the gradients get small is that the change in function value is even smaller. As stated in...