[Octopus-devel] [Octopus-notify] svn commit: r2497 - trunk/src by appel

[alberto.castro at tddft.org]

> > Maybe I am saying nonsense, but have you checked carefully that you can't
> > introduce this feature by making use of the lasers block? You know, the
> > less variables, the better...
>
> In general, yes I agree. The less variables, the better ...
>
>
> > The "lasers" is what we have been using for
> > time-dependent potentials until now.
>
> ... but I have a problem calling moving/breathing quantum dots/rings, or
> electron pumps something like a laser. Since the time and spatial dependence
> of such potentials can be arbitrary one can come up with other interesting
> cases that have nothing to do with a "laser" in the classical sense.
>
> So, I'm not sure how to deal with this. Perhaps one could think the other
> way around: td laser potentials are just a special class of time-dependent
> potentials and we could remove a bunch of laser variables/parameters.

Yes, maybe I  agree with that approach; in my opinion, it would be better
to clean a little bit the lasers feature, and allow for reading spatial parts
from a string (currently it may read spatial parts from a file), and also
arbitrary time functions. We could also rename it to "TDPotential" or
something like that. This way we would have things more cleanly separated;
the epot does not know about time, and the lasers do.

Well, I don't know, maybe we should think this a bit more. It is not a
functionality issue, it can be done in many ways, it is just a matter of
having variables that are "intuitive", so that they are easier to learn
and remember what they are for.

Al.