Why don't other programming languages have ranges?

[Walter Bright]

BCS wrote:
> Hello Walter,
> 
>> BCS wrote:
>>
>>> Every engineering discipline I have any experience with gets a heck
>>> of a lot closer to producing formal proofs of correctness than
>>> programing.
>>>
>> Mechanical engineering designs also tend to be a lot simpler than
>> programs, although the environment they work in is far more complex.
>> Modeling for the design analysis also takes a very simplified view of
>> the actual design, justified by taking the worst case. For example,
>> the strength calculations are done for the weakest cross section, and
>> are not bothered with for the obviously stronger sections.
> 
> Now days they just jump to using finite element and compute everything.

I still see calcs submitted for approval that are done by hand on paper.

If you want to see real seat of the pants engineering, look at one of those hot 
rod shows like Musclecar. I don't think those guys have ever even seen a calculator.


>> Furthermore, after a while a good mechanical engineer develops a
>> "feel" for things that is pretty darned accurate. Going through the
>> analysis is a backup
> 
> No, the analysis is mandated, by code if not law.

Not much. Even for buildings, only a few critical spots need checking. This is 
possible because building structures are usually way over-designed, because it's 
cheap and convenient to do so. Where every gram counts, like in a spacecraft, 
everything is analyzed.

I once had a fire hydrant installed on my property. The city required an 
engineering analysis, which ran to quite a stack of paper. After approval, the 
workers came by to install it. They never looked at the analysis, or even the 
drawings, they just dug up the water main and stuck a hydrant on it with a 
specialized tool they had. Done in an hour or so.

The "engineering analysis" was a freakin' joke.