New port: science/simint: Obara-Saika (OS) method of calculating electron repulsion integrals

[yuri]

Summary

Port to support psi4 quantum chemistry package.

[yuri]

.

[tcberner]

science/simint/Makefile

24

SSE is present in bsd.options.desc.mk

SSE_DESC?=              Use SSE optimized routines

29

why not use '\' to break lines?

Unless you're adding comments to the items, I don't see much points in defining them separately.

[yuri]

science/simint/Makefile

24

In the case of this port, SSE means SSE3, not just any SSE. There are many levels of SSE that various processor families implement.

The generic version of the message would be confusing in this case.

[yuri]

.

[mat]

science/simint/Makefile

24

Then maybe make the option SSE3, so that it says what it means.

[yuri]

Renamed SSE option to SSE3.

[tobik]

science/simint/Makefile

28

Why change the framework and not just use something like

.for opt in ${OPTIONS_SINGLE_VECTOR}
${opt}_VARS=    SIMINT_VECTOR=${opt}
.endfor

CMAKE_ARGS=    -DSIMINT_VECTOR:STRING=${SIMINT_VECTOR:S/3//:tl} \

[yuri]

science/simint/Makefile

28

Why change the framework and not just use something like

This is possible. However, OPTIONS_${okind}_${opt}_SELECTED is a very logical and consistently defined feature in itself. It can easily be useful in other contexts to simplify such 3-liners into 1-liners. The implementation in the framework looks quite elegant to me, and is only 9 lines.

[tobik]

science/simint/Makefile

28

I don't know about that.

What I do know is that this review has been open for almost two months now. You probably could commit it now or very soon, if this didn't require the change in D13078 and portmgr@ approval :)

[yuri]

science/simint/Makefile

28

Many reviews have been open for a long time.
There is nothing that says that this is the reason.