Just like in our DNA...

[Will Ware]

Charles G Waldman (cgw at fnal.gov) wrote:
> Michael Vanier writes:
>  > With all due respect to Guido, who is usually the most low-key no-bullshit
>  > person on the planet, the statement about DNA is a beautiful example of
>  > marketing-speak....
> Do you know anything about so-called "Junk DNA"?  To me, Guido's quote
> seems like a perfectly apt analogy for what one of my former
> co-workers liked to call "surgical code adhesions".

An analogy to junk DNA would be my interpretation as well. The idea with
junk DNA is this: Our DNA (and that of every other living creature) is
a long string of chemical units called base pairs. The DNA sits inside
the nucleus of our cells. Periodically, a portion of the string is
copied to a substring called messenger RNA. The messenger RNA leaves
the nucleus and gets picked up by a gadget called a ribosome. The
ribosome translates the RNA string into a string of amino acids, also
known as a protein. As the protein is built, it starts folding itself
into a characteristic shape. This entire process (which is vastly
complicated when looked at in detail) is called protein synthesis.
Proteins do everything in the cell; they build the walls and internal
structures, they handle all the metabolism (actually, there is a complex
role played by a gadget called a mitochondrion, which has its own
system of DNA), and they build new ribosomes, and several thousand
other tasks.

In the DNA, there are markers for "beginning of amino acid string" and
"end of amino acid string". Each amino acid, or each such marker, is
coded by a sequence of three base pairs. There are, however, large
substrings within the DNA of any species that lie outside any pair of
beginning and end markers. That is called "junk DNA" because it is
never translated to a protein. There are various conjectures about it.
The most obvious is that it does nothing at all, and has accumulated
over billions of years of evolution and is never removed because there
is no process of DNA garbage collection. Another is that, while junk
DNA doesn't synthesize proteins, it is useful because it gives
mechanical stability to the double helix. I suppose you could conjecture
that it offers a little protection of the non-junk DNA from mutation
by radioactive particles, though that seems like a stretch.

Getting back to the analogy with software: In any very large, very old
software system, there will be a number of what are called kludges.
These are pieces of software that were put together in a hurry to
satisfy some short-term demand (often the ire of some attention-span-
challenged manager) and do not conform to whatever overall organization
was intended for the software. Or in a very large body of software, there
may be several different notions of overall organization, done by several
different people who came and went over the years without the chance to
talk to one another. There are often large pieces of code which are
permanently unused, but which are never discarded from the source code
database because they might some day be useful, or the aforementioned
manager is afraid of discarding any intellectual property, regardless
how useless it is.

In both cases, you have information being propogated into the future
which does not actually influence the behavior of the system. There is
a small (but presumably acceptable) cost associated with this propogation,
and in both cases, risk aversion keeps it around. It looks like an awfully
good analogy to me.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com