[Fsatom-comm] List of codes

[Gilles Zerah]

> So, I have for you a few questions, and possible answers :

>
>
> I. Why is the list of codes important to us ?
>
> Possible answers :
> I.a. Because we want to promote their use, in the case
>   they are Libre Software
>   => we should put most of our effort on Libre Software,
>   and have a description of these software
> I.b. Because we want a list of developers, in order
>   to set up workgroups for file standardisation, or
>   testing, or for settting up workshops or tutorials
>   => then, all software should be welcome, the description
>   can be rough, but we need
>   to set up a corresponding set of mailing adresses
> I.c Anything else ??

This is a basic database for the organization of the project.
We need name of peoples, and as much information as possible.
I guess non libre must be included, but with less details (a link).
For many reasons, the main one being that we will be able to update
the list with much more input from open source (well I hope...)

>
>
> II. Why not rely on links to existing lists
> of codes (like the http://france.fsfeurope.org/science/chimie.en.html ).
>
> Possible answers :
> II.a. We need to have some control on the list to achieve
> our goal. Existing lists will help us to maintain ours,
> but it is not a substitute to it.
> II.b. This is part of the visibility of FSAtom.
> II.c. On the contrary, let us join the
>   http://france.fsfeurope.org/science/chimie.en.html
>   effort, and propose a list of "Atomic-scale simulation software",
>   that would include some of the chemistry software, but also
>   solid-state physics, biology simulation, pseudopotentials ...
>   Then, our Web page would mention that we maintain this page
>   for the FSF.

I agree on a,b,c. Are they mutually exclusive?

>
>
> III. How to classify the codes ?
>
> Possible answers :
> III.a. According to theoretical framework, then to licence
>   - Molecular Dynamics/Force Fields
>   - Tight binding
>   - Quantum Chemistry : Hartree Fock/MP/CC/CI
>   - Density Functional
>   - Pseudopotential generators
>   - Quantum Monte Carlo
>   - Many-Body Perturbation Theory (GW, Bethe-Salpether)
>   - Miscellaneous ?!
> III.b. According to licence (Libre/Proprietary)
>   then, to theoretical framework ?
> III.c. According to something else than theoretical framework ?

I suggest a) (Mainly for point IV below) b is then an attribute (so to speak)

>
>
> IV. How to maintain ?
>
> Possible answers :
> IV.a. When set up, wait for voluntary contributions.
>   Need also to check periodically whether the links are
>   still active. Every 6 months ?
> IV.b. Allow for voluntary contributions, but the checking
>   also include a survey of the Web. Every 6 months ?
> IV.c. Can we find a voluntary ?

c doing a and b would be the best. But I do not volunteer.
If we cannot satisfy c, could we use the Wiki?
That is, we divide the work according to domains and we need to find
c1, c2, c3 ...c8 one for each domain, and each ci fills its portion.
Is a)  automatable?
We need a c0 to make a frame, with the list of the theoretical frameworks
and a name of voluntary for each.
Then I could take a part.

>
>
>

All the best,

Gilles