Re: [Axiom-developer] Re: AMS Notices: Open Source Mathematical Software

[Michel Lavaud]

M. Edward (Ed) Borasky a écrit :
> > If the plane I'm flying is built based on simulations with commercial 
> > mathematical software tools, whose accuracy is guaranteed in the 
> > usual way, i.e. no guarantee at all except refund for the price of 
> > the software whatever consequences and it is forbidden to get the 
> > source code to check if it is correct -  then I will for sure take 
> > the next plane, if it has been built with free Open Source software :-)
>
> [snip]
>
> I've heard this argument before -- it's fallacious on a number of 
> levels, and I don't have time to dig into it right now.
Ah dear, you win, I confess I am unable to refute your argument. So, 
after closed source programs, we have now closed source arguments! Very 
clever. Can I buy a licence ? (Ok, just a joke:-))
> But I want to remind people that: 1. Aircraft used to be designed with 
> slide rules and mechanical desk calculators. The equations involved 
> are "open source" in the sense that everyone who is a professional 
> aeronautical engineer learns them in college, knows them intimately. 
> What today's computers allow us to do is build larger and more complex 
> aviation systems that are more economical on fuel.
Yes of course, I don't deny the usefulness of computers for aviation. As 
for "open source" equations : we inherited the old traditional 
scientific way of not selling knowledge. In the new framework of 
so-called "economy of knowledge" (which is, in my opinion, an oxymoron, 
but that's another story), that promote to put property rights on 
knowledge, this will not be the case any more. That's one of my points : 
the trend (i.e. the derivative) is that the situation will go worse, 
i.e. less and less "open source" equations, if we scientists do not stop 
this trend by realizing that selling scientific software and more 
generally selling knowledge is "tuer la poule aux oeufs d'or" (don't 
know in English : "kill the hen with golden eggs"?)

> 2. Very few aircraft crashes are caused by design flaws of any kind, 
> and even fewer by incorrect software. Human error at the time of the 
> flight and sabotage/terrorism/military actions are the two main causes 
> of aircraft crashes. The only really blatant example of a design flaw 
> causing aircraft crashes I can remember was the DeHavilland Comet. 
> That was not a software flaw as far as I know -- I'm not even sure 
> scientific computers were available outside of the military when the 
> Comet was designed, and they would have been on the scale of a Von 
> Neumann/IAS machine, or maybe an IBM 704, if they were.
Yes, OK : in the times when computers were inexistant, I agree it is 
highly improbable that plane crashes were caused by sofware errors :-) 
However, in the times when they existed and were used, I would bet that 
most numerical computations for planes were made in Fortran, and Fortran 
is the exception :that confirms the rule : there are many free libraries 
of subroutines in this language, and some (if not all ?) commercial 
libraries of subroutines are sold with the source code. But maybe I'm 
wrong ?

Best wishes,