Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 23rd, 2016, 6:32 pm

Not much inheritance here.
Last edited by Cuchulainn on March 22nd, 2016, 11:00 pm, edited 1 time in total.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 25th, 2016, 3:47 pm

Hmmm.. Actually, it's organized 100% by inheritance. All of the taxonomic categories are based on relatedness and the presumption of a common ancestor (the base class of dinosaurs). Admittedly, hierarchy can change as new fossils or modern-day genetic analyses shift the inheritance structure to regroup species that may be more or less closely related to others.
Last edited by Traden4Alpha on March 24th, 2016, 11:00 pm, edited 1 time in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 25th, 2016, 7:09 pm

Maybe we have become brainwashed by Linneaus and his descendants to think this way.Example: Animal that can fly (one aspect across all domains). e.g.Most birdsSome fishpterodactylBurritto BrothersOr classify 'dangerous animals'. Its when these things come in contact with other things is what is less 1-dimensional.
Last edited by Cuchulainn on March 24th, 2016, 11:00 pm, edited 1 time in total.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 25th, 2016, 9:06 pm

The taxonomic/Darwinian approach to defining inheritance structures is the clearly superior approach on both logical and empirical grounds. From a maintenance perspective, the Darwinian approach encompasses the ultimate in long-term maintainability (with hundreds of millions of years of use and millions of successful subclass-generation events to prove it ). Organisms of any kind can readily spawn new subclasses of other kinds of related organisms. DNA is the ultimate in reusable and overloaded code.Maybe the core flaw that causes people to build fragile or incorrect class-subclass trees is that they focus too much on superficial external characteristics such as "can it fly" not realizing that flight may be accomplished in many internal ways (e.g., different configurations of bone, muscle, body mass distribution, and metabolism). It's the internal similarities (e.g., DNA patterns) more so than the external similarities that lead to robust inheritance diagrams.
 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 27th, 2016, 7:52 am

QuoteOriginally posted by: Traden4AlphaThe taxonomic/Darwinian approach to defining inheritance structures is the clearly superior approach on both logical and empirical grounds. From a maintenance perspective, the Darwinian approach encompasses the ultimate in long-term maintainability (with hundreds of millions of years of use and millions of successful subclass-generation events to prove it ). Organisms of any kind can readily spawn new subclasses of other kinds of related organisms. DNA is the ultimate in reusable and overloaded code.Maybe the core flaw that causes people to build fragile or incorrect class-subclass trees is that they focus too much on superficial external characteristics such as "can it fly" not realizing that flight may be accomplished in many internal ways (e.g., different configurations of bone, muscle, body mass distribution, and metabolism). It's the internal similarities (e.g., DNA patterns) more so than the external similarities that lead to robust inheritance diagrams.I don't agree. Darwin believed in predeterminism and DNA is hard-coded.So, I think you are referring to parametric polymorphism. Nothing wrong with that, its just not very flexible.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 27th, 2016, 9:31 am

QuoteOriginally posted by: CuchulainnQuoteOriginally posted by: Traden4AlphaThe taxonomic/Darwinian approach to defining inheritance structures is the clearly superior approach on both logical and empirical grounds. From a maintenance perspective, the Darwinian approach encompasses the ultimate in long-term maintainability (with hundreds of millions of years of use and millions of successful subclass-generation events to prove it ). Organisms of any kind can readily spawn new subclasses of other kinds of related organisms. DNA is the ultimate in reusable and overloaded code.Maybe the core flaw that causes people to build fragile or incorrect class-subclass trees is that they focus too much on superficial external characteristics such as "can it fly" not realizing that flight may be accomplished in many internal ways (e.g., different configurations of bone, muscle, body mass distribution, and metabolism). It's the internal similarities (e.g., DNA patterns) more so than the external similarities that lead to robust inheritance diagrams.I don't agree. Darwin believed in predeterminism and DNA is hard-coded.So, I think you are referring to parametric polymorphism. Nothing wrong with that, its just not very flexible.Not sure what Darwin's thoughts on predeterminism have to do with how orinthologists classify birds based on evolutionary inheritance.DNA is far less hard coded than C++ code. The copy process for DNA is far more error prone than the copy process for C++ files. And almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries.
 
User avatar
Polter
Posts: 1
Joined: April 29th, 2008, 4:55 pm

Why is software not organised like the way bird watchers do?

March 27th, 2016, 4:59 pm

Everything is Miscellaneous!QuoteDavid Weinberger's new book covers the breakdown of the established order of ordering. He explains how methods of categorization designed for physical objects fail when we can instead put things in multiple categories at once, and search them in many ways. This is no dry book on taxonomy, but has the insight and wit you'd expect from the author of The Cluetrain Manifesto, Small Pieces Loosely Joined, and a former writer for Woody Allen.
Last edited by Polter on March 26th, 2016, 11:00 pm, edited 1 time in total.
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Why is software not organised like the way bird watchers do?

March 27th, 2016, 6:31 pm

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteOriginally posted by: Traden4AlphaThe taxonomic/Darwinian approach to defining inheritance structures is the clearly superior approach on both logical and empirical grounds. From a maintenance perspective, the Darwinian approach encompasses the ultimate in long-term maintainability (with hundreds of millions of years of use and millions of successful subclass-generation events to prove it ). Organisms of any kind can readily spawn new subclasses of other kinds of related organisms. DNA is the ultimate in reusable and overloaded code.Maybe the core flaw that causes people to build fragile or incorrect class-subclass trees is that they focus too much on superficial external characteristics such as "can it fly" not realizing that flight may be accomplished in many internal ways (e.g., different configurations of bone, muscle, body mass distribution, and metabolism). It's the internal similarities (e.g., DNA patterns) more so than the external similarities that lead to robust inheritance diagrams.I don't agree. Darwin believed in predeterminism and DNA is hard-coded.So, I think you are referring to parametric polymorphism. Nothing wrong with that, its just not very flexible.Not sure what Darwin's thoughts on predeterminism have to do with how orinthologists classify birds based on evolutionary inheritance.DNA is far less hard coded than C++ code. The copy process for DNA is far more error prone than the copy process for C++ files. And almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries.Not really. Programs too live in some environment and are vulnerable to assaults coming from there. For example (something everybody knows, but phrased as a response to this convo), in the past there were the popular buffer overflow attacks on the stack, which is now better protected, but the heap safety leaves much to be desired, e.g. redirecting pointers to virtual tables [*] is a relatively simple hack, which makes web applications so vulnerable (another argument for Cuchulainn to advocate using templates... even if they are less convenient and, if aimed at absolute safety, prone to different sort of "mutations", such as programmer errors). Finally, if one's really determined, they can try to overwrite the code area.* https://cseweb.ucsd.edu/~lerner/papers/ndss14.pdf
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Why is software not organised like the way bird watchers do?

March 27th, 2016, 6:49 pm

QuoteOriginally posted by: PolterEverything is Miscellaneous!Everything is miscellaneous, but not everything is ;-P
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 27th, 2016, 10:18 pm

QuoteOriginally posted by: katastrofaQuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteOriginally posted by: Traden4AlphaThe taxonomic/Darwinian approach to defining inheritance structures is the clearly superior approach on both logical and empirical grounds. From a maintenance perspective, the Darwinian approach encompasses the ultimate in long-term maintainability (with hundreds of millions of years of use and millions of successful subclass-generation events to prove it ). Organisms of any kind can readily spawn new subclasses of other kinds of related organisms. DNA is the ultimate in reusable and overloaded code.Maybe the core flaw that causes people to build fragile or incorrect class-subclass trees is that they focus too much on superficial external characteristics such as "can it fly" not realizing that flight may be accomplished in many internal ways (e.g., different configurations of bone, muscle, body mass distribution, and metabolism). It's the internal similarities (e.g., DNA patterns) more so than the external similarities that lead to robust inheritance diagrams.I don't agree. Darwin believed in predeterminism and DNA is hard-coded.So, I think you are referring to parametric polymorphism. Nothing wrong with that, its just not very flexible.Not sure what Darwin's thoughts on predeterminism have to do with how orinthologists classify birds based on evolutionary inheritance.DNA is far less hard coded than C++ code. The copy process for DNA is far more error prone than the copy process for C++ files. And almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries.Not really. Programs too live in some environment and are vulnerable to assaults coming from there. For example (something everybody knows, but phrased as a response to this convo), in the past there were the popular buffer overflow attacks on the stack, which is now better protected, but the heap safety leaves much to be desired, e.g. redirecting pointers to virtual tables [*] is a relatively simple hack, which makes web applications so vulnerable (another argument for Cuchulainn to advocate using templates... even if they are less convenient and, if aimed at absolute safety, prone to different sort of "mutations", such as programmer errors). Finally, if one's really determined, they can try to overwrite the code area.* https://cseweb.ucsd.edu/~lerner/papers/ ... dfExellent points. There's certainly an interesting overlap between biological and software viruses. In both cases, a foreign code chunk gains access to the system and converts it for replication and distribution of the virus.But programs have it easy. DNA copy error rates are on the order o 10^7 times worse than bit error rates in computers but the system is robust enough that it: a) does not matter than often; b) generally get culled if it does. Several hundred million years of virus evolution have created beasties that criminals could only dream about (e.g., the human hepatitis B virus is only about 1,200 -1,500 bytes in size). The average human is carrying around about 100,000 copies of various kinds of virus DNA amounting to about 5-8% of the human genome.
 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 28th, 2016, 11:36 am

QuoteAnd almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries. That's called the fragile base class problem, aka implementation inheritance, which is bad.So, DNA is not all that good.
Last edited by Cuchulainn on March 27th, 2016, 10:00 pm, edited 1 time in total.
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 28th, 2016, 11:59 am

QuoteOriginally posted by: CuchulainnQuoteAnd almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries. That's called the fragile base class problem, aka implementation inheritance, which is bad.So, DNA is not all that good.DNA must be doing something right (or coders must be doing something wrong) because the DNA version of this has worked for hundreds of millions of years and probably around 10^20 to 10^30 individual creatures. You, yourself, are the product of mixing chunks of code from your parents.The base classes are extremely robust in the biological realm.
 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 28th, 2016, 3:20 pm

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteAnd almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries. That's called the fragile base class problem, aka implementation inheritance, which is bad.So, DNA is not all that good.DNA must be doing something right (or coders must be doing something wrong) because the DNA version of this has worked for hundreds of millions of years and probably around 10^20 to 10^30 individual creatures. You, yourself, are the product of mixing chunks of code from your parents.The base classes are extremely robust in the biological realm.You are not unhappy that it _took_ DNA millions of years to get to this stage?
 
User avatar
Traden4Alpha
Posts: 3300
Joined: September 20th, 2002, 8:30 pm

Why is software not organised like the way bird watchers do?

March 28th, 2016, 3:42 pm

QuoteOriginally posted by: CuchulainnQuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteAnd almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries. That's called the fragile base class problem, aka implementation inheritance, which is bad.So, DNA is not all that good.DNA must be doing something right (or coders must be doing something wrong) because the DNA version of this has worked for hundreds of millions of years and probably around 10^20 to 10^30 individual creatures. You, yourself, are the product of mixing chunks of code from your parents.The base classes are extremely robust in the biological realm.You are not unhappy that it _took_ DNA millions of years to get to this stage?What makes you think DNA has not ALWAYs had this property? 47% of the critical genes in a yeast cell can be swapped for the versions in the human cell and they still work. Try swapping code modules between VAX VMS and Windows 10 and see what happens.
Last edited by Traden4Alpha on March 27th, 2016, 10:00 pm, edited 1 time in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 20250
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Why is software not organised like the way bird watchers do?

March 28th, 2016, 6:30 pm

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuoteAnd almost all creatures get 50% of the DNA from each parent so the system is far more flexible than most code libraries. That's called the fragile base class problem, aka implementation inheritance, which is bad.So, DNA is not all that good.DNA must be doing something right (or coders must be doing something wrong) because the DNA version of this has worked for hundreds of millions of years and probably around 10^20 to 10^30 individual creatures. You, yourself, are the product of mixing chunks of code from your parents.The base classes are extremely robust in the biological realm.You are not unhappy that it _took_ DNA millions of years to get to this stage?What makes you think DNA has not ALWAYs had this property? 47% of the critical genes in a yeast cell can be swapped for the versions in the human cell and they still work. Try swapping code modules between VAX VMA and Windows 10 and see what happens.What makes you think DNA was otherwise?yes, Windows does not support TK50 tapes.