In the genetic code that life has used up to now, there are 64 possible triplet combinations of the four nucleotide letters; these genetic "words" are called codons. Each codon either codes for an amino acid or tells the cell to stop making a protein chain. Now Chin's team have created 256 blank four-letter codons that can be assigned to amino acids that don't even exist yet.
To achieve this, the team had to redesign three pieces of the cellular machinery that make proteins. But they didn't stop there. The team went on to prove their new genetic code works by assigning two "unnatural" amino acids to their quadruplet codons, and incorporated them into a protein chain.
But last month there was this: Horizontal and vertical: The evolution of evolution
The essence of the genetic code is that sequences of three consecutive bases, known as codons, correspond to specific amino acids. [...] There are 64 codons in total and 20 amino acids, which means that the code has some redundancy, with multiple codons specifying the same amino acid.
This code is universal, shared by all organisms, and biologists have long known that it has remarkable properties. In the early 1960s, for example, Woese himself pointed out that one reason for the code's deep tolerance for errors was that similar codons specify either the same amino acid or two with similar chemical properties. Hence, a mutation of a single base, while changing a codon, will tend to have little effect on the properties of the protein being produced.
In 1991, geneticists David Haig and Lawrence Hurst at the University of Oxford went further, showing that the code's level of error tolerance is truly remarkable. They studied the error tolerance of an enormous number of hypothetical genetic codes, all built from the same base pairs but with codons associated randomly with amino acids. They found that the actual code is around one in a million in terms of how good it is at error mitigation. "The actual genetic code," says Goldenfeld, "stands out like a sore thumb as being the best possible."
And all of this really reminds me of this: MOS Technology 6502, Bugs and quirks:
The original 6502 and its NMOS derivatives are noted for having a variety of undocumented instructions, which vary from one chip manufacturer to the next. The 6502's instruction decoding is implemented in a hardwired logic array (similar to a programmable logic array) which is only defined for 151 of the 256 available opcodes. The remaining 105 trigger strange and hard-to-predict actions (e.g., immediately crashing the processor, performing several valid instructions at once, or simply doing nothing at all).
Eastern House Software developed the "Trap65", a device that plugged between the processor and its socket to convert (trap) unimplemented opcodes into BRK (software interrupt) instructions. Some programmers utilized this feature to extend the 6502's instruction set by providing functionality for the unimplemented opcodes with specially written software intercepted at the BRK instruction's 0xFFFE vector.