Serving the Quantitative Finance Community

 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 9th, 2017, 4:39 pm

It sounds like you don't typically create 8-bit, 16-bit, and 32-bit versions of your code then -- only if the specific requirements call for different bit-depth versions. So what is the software version of a pilot plant? Isn't the "start coding now" approach like the laboratory test -> pilot plant -> full-scale strategy?

(BTW, what's interesting about Ariane is that "good" risk management actually caused it. They left in some old code from the older Ariane model rather than risk unintended consequences from removal of the code. But the new rocket's higher speed caused a float->int overflow and literal crash.)
 
User avatar
Cuchulainn
Topic Author
Posts: 20254
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 10th, 2017, 1:40 pm

Prototype
A software prototype is usually code to demonstrate the feasibility of one or more features (system requirements) as a guide for project management.  Which features to include in the prototype are based on certain criteria:
. Choose a feature based on time to develop. “All programmers are optimists” (Brooks)
. Develop the most urgent/critical feature.
. Develop a prototype to show the core process and then throw it away.
 
Fred Brooks said:
“The management question, therefore, is not whether to build a pilot system and throw it away. You will do that. The only question is whether to plan in advance to build a throwaway, or to promise to deliver the throwaway to customers.”
 
Managers don’t like starting again, so the promising prototype code base evolves to the production system. Maintainability issues become one of the major concerns.
 
User avatar
Cuchulainn
Topic Author
Posts: 20254
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 10th, 2017, 1:49 pm

It sounds like you don't typically create 8-bit, 16-bit, and 32-bit versions of your code then -- only if the specific requirements call for different bit-depth versions. 

I didn't say that exactly. Just use C++ template types and std::numeric_limits<T> ? each with its own max, epsilon, ulp etc.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 10th, 2017, 3:35 pm

It sounds like you don't typically create 8-bit, 16-bit, and 32-bit versions of your code then -- only if the specific requirements call for different bit-depth versions. 

I didn't say that exactly. Just use C++ template types and std::numeric_limits<T> ? each with its own max, epsilon, ulp etc.
And that just further illustrates the difference between software and hardware.

The 2,000,000 gallon/day plant is not 10,000 gallon/day plant with a few declarations or data types changed for greater bit-depth.

Yet the deeper issue remains that C++ may be much better language than English for developing, prototyping, and scaling a software system. Sketching the software in the wrong language may be worse than starting to code immediately. Although perhaps this is a personal preference issue because some people might be native English speakers, need to write software in English first, and then translate English to C++.
 
User avatar
Cuchulainn
Topic Author
Posts: 20254
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 10th, 2017, 7:37 pm

I'm not even talking about C++. It's not germane to the discussion IMO. :Let's scope by asking how software projects view prototypes.

//
I vaguely remember RT systems being modeled using Statecharts (pure software) and then throwing them over the partitiion to the hardware dept.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Re: Automatic Reflex and Sapir-Whorf Theory in Programming

December 10th, 2017, 9:09 pm

Given that the thread is on Sapir-Worth, the issue of language choice seems front and center, no?

Maybe there is no one correct language choice for prototypes. The choice may be driven by the nature of risk in the project and the availability of specialized tools (e.g., statecharts for RT systems) to design for that particular risk.