3 ms·

Getting over the code delusion (2010)

20 points by kkoncevicius 5y ago

lmilcin 5y ago
If you think about it, two completely different programs running on your PC share most of their code. You just have to be willing to recognize that the operating systems, libraries, even CPU microcode and structure are all part of the program. If you think about the program as totality of everything that makes it work within some closed structure (like PC case) then it is easy to see how most of the basic mechanisms, infrastructure, can be shared and yet the program as a whole can produce completely different results.
- k__ 5y ago
  Yes. If you look how much of an ape also exist in a human, this shouldn't come as a surprise.
  - aliswe 5y ago
    did sir read the article
  - dTal 5y ago
    That's not a very striking example - Humans are apes, by any reasonable phylogeny. More striking would be how apes share 60% of their DNA with the bananas we so love to eat!
- kuratkull 5y ago
  "share most of your code" is very much false in most cases. you are exaggerating by a huge margin for any serious piece of software
  - lmilcin 5y ago
    Am I? How many serious applications have more code than the operating system, all libraries it uses, all extra tools running there to make it work/usable/maintainable, all CPU design, and so on? Even if you just look at the Linux kernel there aren't many applications that have more code than it has. If you add all standard library and basic OS and additional tools like SSH and likes that you typically have installed then I am not even sure there exists any single application that would have more code than that.
    - aliswe 5y ago
      i think this analogy seems very good its also interesting that a water flea had more genes than us, and something like a worm had about the same ...
TTPrograms 5y ago
I think the thesis should actually be the opposite. Computer science tells us that determining properties of the result of executing arbitrary code without actually executing it is extremely difficult (e.g. halting problem, static analyses etc.). So there's no reason to think "DNA is code" implies "we can predict phenotype from simple analyses of DNA inferred statistically".
nanomonkey 5y ago
Anyone interested in how much more complex evolution is over just genetic material alone, should pick up a copy of the book "Evolution in Four Dimensions"[0]. It goes into how genetic, epigenetic (or non-DNA cellular transmission of traits), behavioral, and symbolic (transmission through language and other forms of symbolic communication) all contribute to a species complexity and ability to evolve. [0]https://mitpress.mit.edu/books/evolution-four-dimensions https://mitpress.mit.edu/books/evolution-four-dimensions
SamPatt 5y ago
Wow. A humbling article. I've been reading a lot about evolutionary biology and genetics and fancied I had a good enough understanding of how DNA worked physically. Apparently not!

20 points by kkoncevicius 5y ago

If you think about it, two completely different programs running on your PC share most of their code. You just have to be willing to recognize that the operating systems, libraries, even CPU microcode and structure are all part of the program. If you think about the program as totality of everything that makes it work within some closed structure (like PC case) then it is easy to see how most of the basic mechanisms, infrastructure, can be shared and yet the program as a whole can produce completely different results.

k__ 5y ago
Yes. If you look how much of an ape also exist in a human, this shouldn't come as a surprise.
- aliswe 5y ago
  did sir read the article
- dTal 5y ago
  That's not a very striking example - Humans are apes, by any reasonable phylogeny. More striking would be how apes share 60% of their DNA with the bananas we so love to eat!
kuratkull 5y ago
"share most of your code" is very much false in most cases. you are exaggerating by a huge margin for any serious piece of software
- lmilcin 5y ago
  Am I? How many serious applications have more code than the operating system, all libraries it uses, all extra tools running there to make it work/usable/maintainable, all CPU design, and so on? Even if you just look at the Linux kernel there aren't many applications that have more code than it has. If you add all standard library and basic OS and additional tools like SSH and likes that you typically have installed then I am not even sure there exists any single application that would have more code than that.
  - aliswe 5y ago
    i think this analogy seems very good its also interesting that a water flea had more genes than us, and something like a worm had about the same ...

TTPrograms 5y ago

I think the thesis should actually be the opposite. Computer science tells us that determining properties of the result of executing arbitrary code without actually executing it is extremely difficult (e.g. halting problem, static analyses etc.). So there's no reason to think "DNA is code" implies "we can predict phenotype from simple analyses of DNA inferred statistically".

nanomonkey 5y ago

Anyone interested in how much more complex evolution is over just genetic material alone, should pick up a copy of the book "Evolution in Four Dimensions"[0]. It goes into how genetic, epigenetic (or non-DNA cellular transmission of traits), behavioral, and symbolic (transmission through language and other forms of symbolic communication) all contribute to a species complexity and ability to evolve. [0]https://mitpress.mit.edu/books/evolution-four-dimensions https://mitpress.mit.edu/books/evolution-four-dimensions

SamPatt 5y ago

Wow. A humbling article. I've been reading a lot about evolutionary biology and genetics and fancied I had a good enough understanding of how DNA worked physically. Apparently not!