[gmx-developers] New analysis library?

[Erik Lindahl]

On Sunday, Feb 2, 2003, at 05:58 US/Pacific, David wrote:

> Dear All,
>
> I am considering to modularize the gromacs analysis tools, that is make
> a monster library that contains all functionality, and main programs
> that only call a single function. This function should the contain all
> the command line parameters
>
> i.e. the main program for e.g. g_rdf would look like:
>
> CopyRight
> parse_common_args()
> analyse_rdf(file1, file2, etc..  option1, option2 etc.)
> thanx()
>
> If we have a library with clearly defined functions it will be very 
> easy
> to make a (portable) graphical frontend, using e.g Python/Tcl.
>
> Anyone think this is a silly/useless idea? Other criticism?
>

I think it's a great idea, and as you say - it will make it much easier 
for us or
other people to write simple frontends to Gnome/KDE/Tcl/Aqua/etc.

One idea I've been thinking of is to use some sort of callback function
for logging and other output. I.e., when you call a program you also 
provide
a function which will be called to write the output. This way it is 
easy to
have everything in windows and avoid the command line completely.

The separation between gmxlib and mdlib is also somewhat artifical 
right now -
both of them depend on the other. Why don't we reorganize things into 
two
libraries:

1. libgmx.so - "core" simulation, I/O & manipulation routines that we 
expect to
     use in lots of programs.

2. libgmxutil.so (or whatever name you want) - all routines used for 
viewers,
     analysis tools, etc.


> I would also like to sollicit your help: it would actually be quite
> simple: move all the code in a certain program to a function, and pass
> all the right parameters to it.
>

If it's not finished when I'm through with the parallelization, I can 
try to do the
same with the mdrun and other kernel programs.

Cheers,

Erik