bosonie
commented
5 years ago Oh wow! I didn't even know there was this possibility of passing .ion files to Siesta instead of Pseudopotential and basis. Good to know! To create .ion families is a good idea, although we have to remind our-self and to the users that the definition of "good basis set" is system dependent. I mean that if a basis choice has been tested on the Mg crystal element, it is not guaranteed to work well for calculations on MgO. To facilitate basis set testing with aiida is something I have in the back of my mind since long time and I would see this as a priority compered to the issue described here, but that's just my opinion.
As an alternative to using a pseudopotential+basis-set-spec pair, we could use pre-packaged .ion files as input to calculations.
This has some benefits for "production calculations" and also adds some more flexibility to the users. For example, a user of an old version of Siesta could still benefit from PSML pseudopotentials by using an .ion file produced by a modern version.
Production of "ion files" might be optional in a calculation, be it a full Siesta calculation or a simpler one with a "gen-basis" program, for example.
Sets of .ion files could be grouped into 'families'.
More metadata than is currently available in an .ion file could be added to the node.