Marco Caceci, Dr. Rer. Nat.

Software: 

fitting curves to data...

I used it to fit four unknowns to curves of twenty points, and it took hours on CP/M and 4 MHz Z-80:

"Fitting Curves to Data: the Simplex Algorithm is the Answer".
M.S. Caceci and W.P. Cacheris.
Byte Magazine 9(5), 340-362 (1984).

then faster computers came around and I fitted 1 K points to 16 K data points, for example in spectral deconvolution, using 6 MHz PC ATs:

"A Simple, Performant Diode Array Spectrophotometer".
M.S. Caceci.
Computers and Chemistry, 13 (1) 33-38 (1989).

then faster computers came around, and I used the bootstrap and the jackknife algorithm to multiply the experimental points and play statistics, on 80 MHz or so PC clones. It still took hours...

"Estimating Error Limits in Parametric Curve Fitting".
M.S. Caceci.
Analytical Chemisty 61 (20) 2324-2327 (1989).

"Simplex, Jackknife, Bootstrap and Montecarlo: Four Algorithms for Non-Linear Curve-Fitting and Confidence Interval Estimation Implemented as PASCAL Procedures".
M.S. Caceci.
Note Technique CEA-SESD/89.38 (April 90).

The code can be downloaded from here, but I make no guarantees:

sjbm.zip (size is 32 KB and it includes sources, executables and test data!)

I have the full text of the documentation (some 100 pages), but its on paper only: call me if you need it: cacio@hotmail.com

Hardware:

For optical spectroscopy, I designed and produced a number of diode array spectrophotometers.

This is the front panel of the latest one, for taking spectra at up to 300 C and 300 Bars, now operating at the Polytechnical University of Barcelona. It is meant to be user friendly, and is controlled by a PC.

For the Autonomous University of Barcelona, I built a fiber-optic spectrofluorimeter: here is a typical spectrum: