These ten software innovations transformed scientific research

[frauenfelder]

Originally published at: https://boingboing.net/2021/01/20/these-ten-software-innovations-transformed-scientific-research.html

…

[buffaloBill]

Before we had good, fast, cheap general purpose processors, we used a thing called a Surface Acoustic Wave (SAW) device to do the frequency decomposition. Those suckers were expensive.

[apenzott]

OK, where are the other nine innovations?

EDIT:
Here is the full list (from the linked article):

Fortran compiler
Fast Fourier transform
Biological databases
General circulation model of the climate
BLAS
NIH Image / ImageJ / Fiji
BLAST
arXiv
Python Notebook / Jupyter
AlexNet

[d_r]

I’m not sure Jupityr has earned a place there yet. Maybe Mathematica, or Matlab. (Or Linpack, from which Matlab was born.)

One of my professors in grad school liked to remark that Fourier’s work paid for all mathematicians who ever lived, or would ever live.

[willmore]

You used an accelerometer for that task? Wouldn’t a simple microphone have been sufficient?

Couldn’t you also have just used an o’scope to measure the frequency?

FFT is wonderful and I remember writing code for it in BASIC on my Commodore 64. I’m pretty sure the book I used to guide me was Numerical Methods.

[markfrauenfelder]

We didn’t have a mic handy, and the sound the drive made when struck wasn’t very loud. I don’t remember if the scope had FFT built-in.

[ejeffrey]

@willmore

You used an accelerometer for that task? Wouldn’t a simple microphone have been sufficient?

I can’t speak for Mark’s application, but accelerometers are usually way better for this type of application. A microphone will definitely tell you the instantaneous frequency at which the drive is spinning, but it won’t necessarily be dramatically louder / stronger when the baseplate is vibrating vs. not. The motor makes noise regardless. An accelerometer tells you the information you really want to know which is whether the plate is actually vibrating. Sometimes the vibration mode will couple effectively to the air and generate a clear audible signature but it isn’t guaranteed.

[willmore]

Sorry, I keep forgetting the differences in background. You can measure frequency with a normal scope by just measuring the time between zero crossings and doing a reciprical. You don’t need anything fancy.

@ejeffrey I have to take issue with what you say. I microphone will pick up a signal more reliably than an accelerometer if the latter is placed poorly. With the accelerometer, you have to guess what mode the baseplate is vibrating in and place the tool appropriately. A microphone is much more forgiving as it can easily be moved around and thereby compensate for oscillation mode.