🚀 go-pugleaf

Ok, guys, this is probably trivial, but...

I am looking for some software (pref. public domain) that can de-compose
curves that are the sum of a number of closely spaced peaks.  To give more
detail, I'm talking about XPS spectra.(x-ray photoelectron spectroscopy)
There is a gently sloping background with peaks of known width on top of it.
If the peaks are spaced by more than half the width, then separation is easy.
In some cases, the peaks are only spaced about 1/10 of a width apart and 
consequently I need software to help.  The software would have to:

1)     Read in the data as x,y points
  
2)     accept guesses such as peak shape (Gaussian/Lorenzian mix and width),
       number of peaks and their approx positions
     
3)     do some kind of non-linear least-squares fit.

Sorry if this sounds like a brief to a software engineering company!

The whole idea is to take a wobbly bulge, have an educated guess what might 
be there, then allow the software to confirm suspicions by attempting to do 
the fit.  Yes, I can hear the purists saying "write it yourself", but
I'd rather be doing XPS if someone has already written something like this or
if it's an easily available program.  I've already re-invented enough
wheels, thanks!

As far as source code goes, I like Fortran, Basic, Pascal and even C.  For
compiled stuff, I'm happy with Macintosh, AppleIIGS or VAX/VMS.

Thanks in advance.
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
= Chris Vernon,  CSIRO Division of Mineral Products, Perth, Western Australia  =
=  Tel (61) 9 350 5411    Bitnet: Vernon_CF%cc.curtin.edu.au@cunyvm.bitnet     =
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

PS - anyone heard of a company called Surface Software (USA, Canada ?)
I would appreciate their address.