Beefy Boxes and Bandwidth Generously Provided by pair Networks
Clear questions and runnable code
get the best and fastest answer
 
PerlMonks  

Re: (OT) Evolutionary Design??

by Elgon (Curate)
on Oct 30, 2003 at 18:33 UTC ( [id://303350]=note: print w/replies, xml ) Need Help??


in reply to (OT) Evolutionary Design??

Anonymonk,

A very interesting and impassioned post, much of which I agree with, however I gently take issue with a couple of points:

  • It is entirely possible to develop and 'grow' algorithms through a process which models evolution through natural selection. This has been demonstrated albeit at a simple level here in the monastery. A search should turn up the relevant threads. I know of no actual production code which has been developed using this method, so calling it a development methodology is a tad premature.
  • It could be argued that any development methodology where independent teams develop similar modules of code in parallel (population) and then compare their code (variation) against predetermined criteria before reworking the code in the light of the lessons learned (evolution) bears a likeness to the biological process, the major difference being that the fitness criteria are not determiined by the environment but by the observer.
  • I would disagree slightly with your point about evolution merely producing better reproducers: This is the generalised result of the types of evolutionary pressure found in nature, however if the criteria used to judge the fitness of an algorith or module are correctly chosen, this does not necessarily need to be the result. As a further point, I would add that Word is the dominant member of the population because of the pressures of the market, which have been manipulated, quite apart from the software development process.

Though provoking, none the less and I wholeheartedly agree with your crusade against the marketroid crud which can be used to sell software.

Elgon

Please, if this node offends you, re-read it. Think for a bit. I am almost certainly not trying to offend you. Remember - Please never take anything I do or say seriously.

Replies are listed 'Best First'.
Re: Re: (OT) Evolutionary Design??
by Anonymous Monk on Oct 30, 2003 at 18:48 UTC
    It is entirely possible to develop and 'grow' algorithms through a process which models evolution through natural selection. This has been demonstrated albeit at a simple level here in the monastery. A search should turn up the relevant threads. I know of no actual production code which has been developed using this method, so calling it a development methodology is a tad premature.
    I totally neglected, though intended, to bring up evolutionary programming in the sense of genetic algorithms and the like. These do try to mimic important aspects of biological evolution and are, IMHO, quite properly associated with biological evolutionary concepts. Note, read some papers in genetic algorithms and methodology and the researchers nearly always discuss the particular mechanisms they are trying to model, and how their models differ from models in evolutionary biology.

    But no one in those fields is even suggesting that genetic algorithms could be usefully applied in the design and implementation of even a simple text editor. At least not at the present state of the field.

    I was only referring to the use of evolutionary principles in the larger software design and development methodologies. That is an important clarification, thanks!

Log In?
Username:
Password:

What's my password?
Create A New User
Domain Nodelet?
Node Status?
node history
Node Type: note [id://303350]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this?Last hourOther CB clients
Other Users?
Others pondering the Monastery: (9)
As of 2024-04-18 16:47 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    No recent polls found