Hi Nick,

The way Sparky expects files in directories with fixed names (Save,
Projects, ...) is horrible. That was developed in the early 1990s
before Python was part of Sparky so it is in the C++ code. Some small
modifications to the C++ could ease the pain but it would require
recompiling Sparky. That requires having the right versions of Python
and Tcl/Tk and it uses an unconventional build method briefly described
in the manual:

http://www.cgl.ucsf.edu/home/sparky/manual/install.html#Compile

The easiest modification would be to session.cc changing the
CheckDirectories() routine to always return true. Then you would never
be asked to create the standard directories.

Tom


nickfitzkee wrote:
>
>
>
> Hi,
>
> I love sparky, but I find it's default directory organization in
> SPARKYHOME cumbersome, especially for multiple sets of data. Currently,
> whenever I load files, I have to navigate through the system path to the
> local directories where I store processed spectra, assignments, etc.
> This can be a lot of navigating, and Sparky doesn't always consistently
> remember the last working directory--so I'll typically have to navigate
> repeatedly if I want to open several spectra, assginments, etc.
>
> Alternatively, I could set SPARKYHOME to the current working directory
> (CWD, '.'), but then I'm prompted to create the home directories every
> time I run sparky from a new location. Even then, I'm still stuck with
> extra directories that I'd need to delete later.
>
> Is there any way to set all three SPARKYHOME directories (Projects,
> Lists, Saves) to the CWD so that I don't have to bother with its data
> store anymore? I'm comfortable modifying the python source, but I
> haven't found where this information may be stored there.
>
> Thanks,
> Nick
>
>