@Wrzlprmft It is now! (At least for people of GMT timezone and beyond)

@ff524 I didn't think you'd be so exited about hats. But I could say one thing, when winter bash arrives, I bet you and @jakebeal would be among the first to catch 'em all. ;-)

I wonder if those who wonder about the usage of closed-source software in scientific computations realize how many experiments have been implemented along the years with commercial instruments based on closed-source designs and firmware.

@MassimoOrtolano Right on. I'd say one of the best examples would be MATLAB.

@ÉbeIsaac Yes, but there are also more subtle cases. Think for instance at an instrument like this one: keysight.com/en/pdx-x202208-pn-E5080A/… It's an instrument that probably costs around 150 k$, and, among other things, can be used to characterize materials (e.g., the dielectric constant).

Its operation relies on a complex algorithm for the correction of a number of internal errors. Though the principle of the algorithm is documented in the literature, the specific implementation and possible variations are not. They are proprietary software. Even though the instrument can be calibrated (that is, verified), this operation can be done on a limited number of points, much less than those with which you can verify a software like MATLAB.

But I'm sure that those who use such an instrument in experiments almost never question the internal algorithm. So the main point is not whether a software is proprietary or not, but whether it is reasonably verified or not.