A cloud is not a still.

And Reverse Osmosis water purifiers are far more complex than water stills.  Both are designed.  Both are irreducibly complex.  Design has the capability of and often does dramatically increase complexity and information content (countless terabytes is now common) in designed systems of lasting utility.  There are zero examples of material processes doing the same beyond the capability of what blind search is capable of accomplishing (which is less than 500 bits of information).

This is the fundamental problem with your evolutionary narrative.  You have not shown how any materialistic process can or does accomplish what you claim for evolution.  Then when we observe evolutionary processes in action it confirms this problem.  Every observed example of evolution falls well below the 500 bit limit.  Observations confirm that evolution does not proceed fast enough and lacks the required processes that would enable it to generate new biological information to build new form and function.  Can you offer even one example of a materialistic process generating more than 500 bits of information?  NO.  Yet the presumed evolution of ape to man requires on the order of 300-600 megabytes of new information.  Design is quite capable of accomplishing this kind of change.  Direct observation of evolutionary processes strongly suggests that evolution is not, and the hard sciences of information theory chemistry and probability tells us why.

RF = proof that a little learning is a dangerous thing.

Here's something fun: http://planetquest.jpl.nasa.gov/science/finding_life.cfm

I think we can safely say that Specified Complexity is one of the hallmarks of ID.  Complexity is an idea in Information Theory, but SC is specific to ID.

Complexity can be measured, it has a formula and tells us a little bit about the amount of information in something.  It is an objective measure, and therefore tells us nothing of the information of, say, a sentence like "Methinks its a weasel". It can only tell you the odds of those particular letters coming up and how much information is in the sentence based on the odds of those characters.

In other words, "oahofwfu hos o asfaho" has equal information.

Not according to ID.  ID calls the first one SPECIFICALLY complex and then makes the jump to DNA and says it is a coded language that means something.


So, here is the challenge to any ID follower:  How much specified information is in DNA?  Not information (as according to Shannon) but the SC variety?

Since there whole "possibilty theory" rests on the specified information in things, we have a right to ask.




Just so people don't think I'm making this up, this is the definition from the ID'iot's at ISCID:

http://www.iscid.org/encyclopedia/Shannon_Information



So, how much SC is in DNA?  I want an objective number.

Another problem is that Information (in the Shannon sense) is not the same as what Dembski calls Specified Information (or SC).  Demsbki has not come up with a way to measure the subjective determination of "its looks designed".

In cryptography the National Research Council has set 10^94 as the Universal Probability Bound to secure encryption schemes against random search attacks. For more information see, Kenneth Dam and Herbert Lin in "Cryptography's Role in Securing the Information Society", Washington National Press, 1996.


The genetic code is encoded deterministic information and is measured as such.


If so, we would a) recognize the resulting increase in new information and b) Observe these things occurring in the real world and in the lab.



Unless there is some mechanism shown to retain them we would go with observed mutation rates in non-functional sequences.



In experimentation and actual observation mutation and selection falls far short of "enough".  We observe that it occurs far to slowly and accomplishes far too little.


You are free to presume the new genetic information is modified from pre-existing information.

You can't find when you don't look.  It is not a stretch at all to apply the methods of probability to genetic studies.  Nearly ever evaluation in genetic and protein source research included concepts from probability.  Just like in cryptography and other probability studies, there are ways to manage uncertainties to get meaningful results.  protein function is far more closely related to shape and chemic affinities than to changing environments and one can evaluate the potential ratio of workable protein combinations based on these characteristics.  When one does this they find the entire range of potentially workable proteins and not just the ones that may be suitable for the more limited environmental conditions you describe.  Even here the ratio of workable to unworkable proteins is in the order of 10^-70 and higher.



All available evidence suggests that it is not self-organized, it seems silly to presume it is.


The error rate is less than 1 in 100,000,000 duplications and except for in the case of reproduction, the feedback redundancy and repair mechanisms drop that rate to very close to zero.  This is far fewer errors than human language.  It is far fewer failures than computer systems which have failure rates in the low millions.  Your quotes seem like prior commitments from people who are looking to explain the realities away.

The modest mutation rates seem like a clever way to introduce minor variation in non-critical systems to achieve a degree of differences from one individual to the next.  It clearly is quite proficient at this task.
 

This is not an example of a regulatory gene alteration.  You have not provided the evidence that shows that original function was retained (I asked for it twice, still nothing).  It fit well within the range of probabilities for single point mutations and it involved a single protein enzyme with no binding sites and no required interaction to other protein systems. 

This is an example of the kinds of things evolution can do and it fits well within the range of observed evolutionary processes which we have seen can accomplish one and two step processes (if you ever produce the data, this would be one of a few known two step processes) but we don't have even one example of a four step evolutionary pathway.  We have had an opportunity to observe 10^30 organisms in the 50 years of genetic studies by now and we still don't know of one four step evolutionary pathway.  In the mammalian line there are estimated to be fewer than 10^18 total organisms ever to have existed and yet the number of required evolutionary steps number in the billions and billions.  In the lab, the observed rates are coming in billions and billions and billions and billions of times too slowly to account for the diversity we observe.




No, more complexity overall in that it is different and unique.



Sure. but the odds of generating useful proteins from random non-expressed areas of the genome is astronomical given that the ratio of useful to useless proteins is 10^-70 or so as indicated by Axe in 2003.  It is the reality of these kinds of numbers that explain why evolutionary processes are quite proficient at the occasional point changes and two step processes but cannot accomplish the 10-20 step processes required to generate a useful new protein in a reasonable period of time.  HIV and other retroviruses even with huge mutation rates and fast replication rates would need billions and billions of years just to produce one new protein from a non-expressed area.