If you flip a coin, you have a 50% chance of heads or tails. That is 'random shit'. If you count the number of heads, let's say X heads, as you roll then that is a stochastic process. If you flip N times, the probability of having X = N/2 goes to one as N goes to infinity. So, while the process arises from 'random shit', the answer is NOT RANDOM.
This happens in a lot of cases, including Monte Carlo integration (another stochastic process) and statistical thermodynamics.
So go stfu if you don't understand the difference. Other people do.
Remember when Yahoo, Altavista, MSN, etc etc had their crappy search engines running? What did they base their search algorithm on? It was text/string batching mostly...heavily investing in lists and relevance factors. What does Google do differently?
They use advanced mathematics to predict the likelihood of matching. What exactly do they do?
Google represents the Web as a Markov chain, where each website is a node and each link increases the unidirectional transition probability of moving from the origin node to the destination node. You can find the 'limiting probability distribution' of a Markov chain relatively easily, even though Google has billions of nodes and trillions? of links.
The limiting probability distribution gives you the relative probability that, web surfing through the links (ie. performing a random walk on the chain), you will eventually settle on a particular website. This is Google's original innovation and technology.
Unfortunately, this type of mathematics is usually only taught in CS graduate school. It should be taught sooner, but it requires getting into a branch of mathematics called 'stochastic processes'.
Chaperones are proteins that help other proteins fold through protein-protein interactions. They're very common in mammalian cells and are also used as sensors for heat shock or other environmental stresses that cause widespread protein denaturing.
A prion is a protein that can induce a conformational change in the structure of a molecule of its own type. It's called an autocatalytic reaction.
So you might get something like Prion-inactivated + Prion-activated --> 2 Prion-activated.
The hypthesis is that the fibers responsible for Alziehmers form as the prions become activated and polymerize. The fibers then disrupt the synapses between neurons.
RFID signals broadcast from advertisements will be used in conjunction with the RFID tags now secretly being placed in your contact lenses to permit 24 hour total penis enlargement awareness.
Yes, like I'd vote for a guy who keeps sending me something I hate.
A better strategy would be to use pop ads to attack your opponents (leaving out your name completely). You get angry at the pop up ad and, of course, irrationally blame the guy who the pop up ad criticizes.
By the McCain-Feingold (Feinstein?) Act, that would be illegal. But that never stopped spammers, did it?
I'm not an editor at a peer-reviewed journal, but here's my two cents:
The whole process of selecting and managing the reviewers, sending out the pre-prints, editing them (because there's always mistakes), and making sure the text and graphics are set well does, in fact, cost money. The publisher makes a profit, of course, but I don't think it's as much as you would expect, given their prices. They're just not very efficient (and why be efficient if you're the major publisher and have a near-monopoly)?
So Elsevier just needs actual competition, but the barrier to entry in the science publication business is extremely high:
Besides actually publishing the journals (which any media company could probably do), you have to make extensive contacts with scientists in the respective field (so they can review papers) and convince those scientists that your journal is good enough to publish in.
The journals exist in a heirarchy of prestige. No one wants to publish in...'Bioinformatics' when they can publish in 'Biophysical journal' or 'Journal of Chemical Physics'. 'Chaos' was originally the bottom of the heap, but they've been steadily rising in notoriety.
So the journal is supported by the people who publish in it...not the company who owns it.
Oh, so you want to talk about statistical mechanics, do ya? I was sticking to classical, but I'll bring out the big guns.
And 'ergodics' has nothing to do with the Gibbs relation.
Entropy is the number the different possible microstates a system may exist in.
S = k Ln(Omega), where Omega is the microcanonical ensemble.
which was derived by Boltzmann and ridiculed by people like you. He committed suicide because of the pressure, but don't worry, you're too stupid to inflict damage upon my ego.
And, to show that you have no idea what you just said, the ergodic principle states that the system inhabits EVERY state in the phase space EQUALLY. Not this 'nature of time spent in local regions of the microstate phase space'. Wtf?
If you're talking 'local regions' as in energy states of the phase space, then we can switch over to the canonical ensemble, such as
E, Energy of the system = -dlnZ/dB where Z is the canonical partition, Int(exp(-HB)) over all energy levels
and
E, Entropy of the system = k[ln Z - BdLnZ/dB] + So
And wtf is 'quantum collapse during measurement'. It's not a quantum wave function...it's a partition ensemble (which usually does not take into account electronic states). Wtf are you talking about?
That's not the 2nd law. That's not even close. That's a Gibb's (-Duham) relation.
The 2nd law states that, for a closed system, the amount of available energy must never increase, or equivalently, that the entropy of the system must increase. Actually, that's the corallory of the 2nd law, which was originally phrased in a much less useful way.
So, fyi, when 'experts' say that the necessary energy is equal to "4000 Nagasaki bombs" or an explosion worth 40 kilotons of TNT, they mean the the amount of energy released when TNT explodes, ie. its heat of combustion.
So we're talking about enthalpic release of heat. I don't know what you think you're talking about, but I know what I'm talking about.
And I was being facetious when I was comparing the total power output of the sun to the ludicrously stupid numbers coming out of your mouth.
And, btw, solar energy can't be converted into a carbon source. You can convert carbon dioxide into carbohydrates via photosynthesis, but there's no conversion from energy to matter...only work performed.
And unless you state the 2nd law of thermodynamics, I won't reply anymore.
God, more number mongering. Here's a number to suck on:
The sun produces 1x10^15 Watts of energy.
The heat of combustion of TNT is 15 kJ/gram. By your numbers (4 tons TNT per year per acre required) and 2379400204 acres in the entire United States, we would need to harness just 1.9 x 10^-10 % of the sun's energy per acre or 0.453% for growing farmstuff in _every_ acre of the U.S (notice that only 2% of the entire US is used for farmland, however).
Of course, by your numbers, the temperature of the Earth (especially Iowa in your example), given that it absorbed that much energy in a year, would be catastrophic itself. Considering that I'm still alive and not burnt to a crisp, I'm still going to hold by my statement that you are
Still Verifiably Talking Out Of Your Ass.
I bet that expert of yours included the absorbance of solar energy by the plants as 'energy consumption'.
Or, considering your 2nd or 3rd comment, that even though we burn 400 years worth of fossil fuel every year, that the fossile fuel has been accumulating for hundreds of thousands of years and that we still have enough fossil fuel for (at least) the next 100 years.
I agree that we should use more solar or wind energy sources, but only for economic and political benefits. We've got enough oil for quite a long time.
And I don't need a web quote, I've got common sense. Unlike your expert.
Population Pressure? The average birth rate in the US is about 2.7 or so. Our population just hit 300 million and there's more than enough food (and places to grow food).
It really seems like you're talking out of your ass.
And I completely disagree that there's a fresh water shortage. Rainfall in the US has not decreased in the past decade and I know of no environmentalist who has even claimed that global warming would decrease rainfall (it might actually increase it!). Rainfall gives you fresh water (yes, Captain Obvious has spoken!).
Now, as for Israel & the West Bank, they don't get a whole lot of rainfall..mostly because it's a DESERT.
And my profession has spent decades studying the best way to purify water. Those pharmaceutical and manufacturing plants commonly purify their water outflows so much they are often MORE pure than the water inflow.
And, yes, it does require energy to perform, but it doesn't have to be from burning fossil fuels. They usually use electricity so a significant fraction of it comes from nuclear reactors.
Now, go rant on nuclear reactors for all the good it'll do ya, rofl.
"1.7 metric tons of material are consumed by making one PC"
Bullshit! What are we doing, fusion? The 1.5 metric tons of water doesn't disappear. It gets recycled in one way or another. Yeah, the fabrication process is very chemical intensive, but the big manufacturers (Intel, AMD) have strict environmental policies. They recycle where they get, purify their outflows, and use as little material as possible.
Both for cost-cutting sake and environmental law sake.
So that 1.5 metric tons of water is reused over and over and over in making each PC. The actual specific waste per PC should be measured as the material that leaves the manufacturing factory per day (as waste) divided by the number of pieces of hardware it made that day.
For computer geeks, you guys are really stupid. That is, unless your PC weights 1.7 metric tons.
Do you go to a University or live near one? Go to your local University library (or if they allow online access..) and take a look at yesterday's issue of Science.
It has articles on both the microreactor and the human embryonic stem cells that were cloned.
I heard about Wolfram's book a while back, picked it up, and read it. Well, 90% of it. I do agree with some of the harsh remarks directed towards the patently unacademic (read, un-peer reviewed) method in which Wolfram is purporting his (supposedly) new ideas, but I think he's proposing something even worse:
His "New Kind of Science" is based purely on supposition and example..by example by example. This isn't science.
Observation is not scientific proof, it is only the first step in attempting to _explain why_ something occurs. I read a whole lot of "Here, if we make a cellular automaton use it looks a whole lot like ". Even Wolfram admits that among the millions of combinations of rules, something interesting is bound to happen. That this interesting automaton should hope to _explain_ something physically important is simply supposition, not proof that the physical world obeys some underlying automaton.
When Wolfram can devise an automaton that quantitatively _predicts_ the behavior of a physical system, then I will be impressed. Oh, with a caveat. The prediction shouldn't be a trivial one, where an already existing quantitative description exists that is simply discretized into an automaton and declared "new". (I've seen this done and it isn't impressive.)
...forget that bearing right could mean anything from making a turn that is less than 90 degrees right to turning the wheel an inch to the right while driving straight.
...successfully navigate yourself to BLAH BLAH Ave. South when you had typed in BLAH BLAH Ave. North. Congratulations! You are now lost in downtown! (And..no, those two streets could be miles apart and unconnected! Evil steet namers should die.)
You follow the directions given by Mapquest and...end up going down a one-way street....forget that bearing right could mean anything from making a turn that is less than 90 degrees right to turning the wheel an inch to the right while driving straight....successfully navigate yourself to BLAH BLAH Ave. South when you had typed in BLAH BLAH Ave. North. Congratulations! You are now lost in downtown! (And..no, those two streets could be miles apart and unconnected! Evil steet namers should die.)
The Israelis that run the website are not hiding in the mountains somewhere in Pakistan. The fact that the Bush Administration put the website on the list is political...they're sending a message that they will consider certain Israeli organizations to be terrorists if they behave like terrorist Palestinian organizations.
The people who run this website may or may not have family in the U.S or might want to visit the U.S in the future. Cutting them off from that access is a punishment, certainly.
But will this work for any fundamentalist Islamic terrorist website? Hardly.
This has nothing to do with free speech and everything to do with accountability of ones' actions on the Internet. Anyone is free to spread hatred on the Internet. The Bush Admin is saying that, whenever possible, we don't want those people running the websites to be any friends of ours.
I played the original Sojourn like in 1992. I tried Duris after they split (bad move?), but it wasn't the same (and neither was Toril).
I was "Warony". Cleric.
FYI, the lead designer of EQ, Brad McQuaid, played Sojourn aka "Aradune" or something such. Ranger. EQ felt exactly like Sojourn...making it highly successful and extremely boring for someone who played the real thing.
But, if you code in a sequence that is 30 nucleotides long as "locations" you are almost gaurenteed that the 30 nucleotide sequence is UNIQUE to the DNA.
But, of course, DNA is just floating around and you don't know exactly where the DNA really is...only that if you want to target Gene X on Chromosome Y you need a XXXXXXXXXXXXXX sequence to probe.
I think what you're saying is that DNA "addresses" are less precise than circuit addresses. They probably are... but they are also more flexible in information storage and retrieval.
There are proteins that activate all genes that contain XXXXXXXX sequence. There are proteins that bind to YYYYYYYY sequence, cut it, and invert it. (now that's a real switch!)
There are sequences that flank the "data storage" area of DNA (heterochromatin), which may be removed transiently to transfer information back and forth from the 'inactive' to the 'active' section of chromatin DNA.
I agree that 'DNA computers' will never be faster than silicon circuity, but (as you said) they are parallel and _analog_ circuits. More useful for signal processing than doing arithmatic. (Think of our brains and how they are different from silicon computers...)
Organisms can also interact with the real world through chemical reactions on the 'nano' scale. A computer can't do that.:)
Slashdot Mirror
Now now, just because your school can't get into the top ten, it's no reason to get snippy.
;)
Are you joking? My department is rated #1 in the country for the chemical engineering. (Ok, honestly, it's a tie between us, UC Berkeley, and MIT).
And John Q. Random doesn't give a shit about Microsoft. You're thinking of Sumgion Slashdot.
Because they're different. I can explain:
If you flip a coin, you have a 50% chance of heads or tails. That is 'random shit'. If you count the number of heads, let's say X heads, as you roll then that is a stochastic process. If you flip N times, the probability of having X = N/2 goes to one as N goes to infinity. So, while the process arises from 'random shit', the answer is NOT RANDOM.
This happens in a lot of cases, including Monte Carlo integration (another stochastic process) and statistical thermodynamics.
So go stfu if you don't understand the difference. Other people do.
Remember when Yahoo, Altavista, MSN, etc etc had their crappy search engines running? What did they base their search algorithm on? It was text/string batching mostly...heavily investing in lists and relevance factors. What does Google do differently?
They use advanced mathematics to predict the likelihood of matching. What exactly do they do?
Google represents the Web as a Markov chain, where each website is a node and each link increases the unidirectional transition probability of moving from the origin node to the destination node. You can find the 'limiting probability distribution' of a Markov chain relatively easily, even though Google has billions of nodes and trillions? of links.
The limiting probability distribution gives you the relative probability that, web surfing through the links (ie. performing a random walk on the chain), you will eventually settle on a particular website. This is Google's original innovation and technology.
Unfortunately, this type of mathematics is usually only taught in CS graduate school. It should be taught sooner, but it requires getting into a branch of mathematics called 'stochastic processes'.
Chaperones are proteins that help other proteins fold through protein-protein interactions. They're very common in mammalian cells and are also used as sensors for heat shock or other environmental stresses that cause widespread protein denaturing.
A prion is a protein that can induce a conformational change in the structure of a molecule of its own type. It's called an autocatalytic reaction.
So you might get something like
Prion-inactivated + Prion-activated --> 2 Prion-activated.
The hypthesis is that the fibers responsible for Alziehmers form as the prions become activated and polymerize. The fibers then disrupt the synapses between neurons.
My two bits.
RFID signals broadcast from advertisements will be used in conjunction with the RFID tags now secretly being placed in your contact lenses to permit 24 hour total penis enlargement awareness.
Funny. Very funny.
Yes, like I'd vote for a guy who keeps sending me something I hate.
A better strategy would be to use pop ads to attack your opponents (leaving out your name completely). You get angry at the pop up ad and, of course, irrationally blame the guy who the pop up ad criticizes.
By the McCain-Feingold (Feinstein?) Act, that would be illegal. But that never stopped spammers, did it?
I'm not an editor at a peer-reviewed journal, but here's my two cents:
...'Bioinformatics' when they can publish in 'Biophysical journal' or 'Journal of Chemical Physics'. 'Chaos' was originally the bottom of the heap, but they've been steadily rising in notoriety.
The whole process of selecting and managing the reviewers, sending out the pre-prints, editing them (because there's always mistakes), and making sure the text and graphics are set well does, in fact, cost money. The publisher makes a profit, of course, but I don't think it's as much as you would expect, given their prices. They're just not very efficient (and why be efficient if you're the major publisher and have a near-monopoly)?
So Elsevier just needs actual competition, but the barrier to entry in the science publication business is extremely high:
Besides actually publishing the journals (which any media company could probably do), you have to make extensive contacts with scientists in the respective field (so they can review papers) and convince those scientists that your journal is good enough to publish in.
The journals exist in a heirarchy of prestige. No one wants to publish in
So the journal is supported by the people who publish in it...not the company who owns it.
E, Entropy of the system = k[ln Z - BdLnZ/dB] + So
-->
S, Entropy of the system = k[ln Z - BdLnZ/dB] + So
Oh, so you want to talk about statistical mechanics, do ya? I was sticking to classical, but I'll bring out the big guns.
And 'ergodics' has nothing to do with the Gibbs relation.
Entropy is the number the different possible microstates a system may exist in.
S = k Ln(Omega),
where Omega is the microcanonical ensemble.
which was derived by Boltzmann and ridiculed by people like you. He committed suicide because of the pressure, but don't worry, you're too stupid to inflict damage upon my ego.
And, to show that you have no idea what you just said, the ergodic principle states that the system inhabits EVERY state in the phase space EQUALLY. Not this 'nature of time spent in local regions of the microstate phase space'. Wtf?
If you're talking 'local regions' as in energy states of the phase space, then we can switch over to the canonical ensemble, such as
E, Energy of the system = -dlnZ/dB
where Z is the canonical partition, Int(exp(-HB)) over all energy levels
and
E, Entropy of the system = k[ln Z - BdLnZ/dB] + So
And wtf is 'quantum collapse during measurement'.
It's not a quantum wave function...it's a partition ensemble (which usually does not take into account electronic states). Wtf are you talking about?
Ok, now I'm bored of you. B'bye.
Uh, no.
That's not the 2nd law. That's not even close. That's a Gibb's (-Duham) relation.
The 2nd law states that, for a closed system, the amount of available energy must never increase, or equivalently, that the entropy of the system must increase. Actually, that's the corallory of the 2nd law, which was originally phrased in a much less useful way.
I'm not replying anymore. You failed my test.
Wow, the Slashdot banter. Gotta love it.
So, fyi, when 'experts' say that the necessary energy is equal to "4000 Nagasaki bombs" or an explosion worth 40 kilotons of TNT, they mean the the amount of energy released when TNT explodes, ie. its heat of combustion.
So we're talking about enthalpic release of heat.
I don't know what you think you're talking about, but I know what I'm talking about.
And I was being facetious when I was comparing the total power output of the sun to the ludicrously stupid numbers coming out of your mouth.
And, btw, solar energy can't be converted into a carbon source. You can convert carbon dioxide into carbohydrates via photosynthesis, but there's no conversion from energy to matter...only work performed.
And unless you state the 2nd law of thermodynamics, I won't reply anymore.
God, more number mongering. Here's a number to suck on:
The sun produces 1x10^15 Watts of energy.
The heat of combustion of TNT is 15 kJ/gram.
By your numbers (4 tons TNT per year per acre required) and 2379400204 acres in the entire United States, we would need to harness just
1.9 x 10^-10 % of the sun's energy per acre or
0.453% for growing farmstuff in _every_ acre of the U.S (notice that only 2% of the entire US is used for farmland, however).
Of course, by your numbers, the temperature of the Earth (especially Iowa in your example), given that it absorbed that much energy in a year, would be catastrophic itself. Considering that I'm still alive and not burnt to a crisp, I'm still going to hold by my statement that you are
Still Verifiably Talking Out Of Your Ass.
I bet that expert of yours included the absorbance of solar energy by the plants as 'energy consumption'.
Or, considering your 2nd or 3rd comment, that even though we burn 400 years worth of fossil fuel every year, that the fossile fuel has been accumulating for hundreds of thousands of years and that we still have enough fossil fuel for (at least) the next 100 years.
I agree that we should use more solar or wind energy sources, but only for economic and political benefits. We've got enough oil for quite a long time.
And I don't need a web quote, I've got common sense. Unlike your expert.
Population Pressure? The average birth rate in the US is about 2.7 or so. Our population just hit 300 million and there's more than enough food (and places to grow food).
It really seems like you're talking out of your ass.
Unless the female pelvis enlarges to account for the increased skull size.
Also note that our intelligence is not directly proportional to the simple size of our skull.
It's got a lot more to do with the types of neuron connections that are possible than with the bulk quantity of neurons in general.
IANAN: I Am Not A NeuroScientist
The parent poster is better described as a "factoid partisan". Someone who uses web quotes from other partisans to support their views.
Yeah, a forecast from you is as good as gold, rofl.
The problem with your argument is that these so-called experts have no idea about what they're talking about. Let me explain:
"And how do you recycle that water?", you ask? Generally, some of the ways of doing it are...
1) Evaporation & Condensation
2) Filtering
3) Biological catalysis (popular with sewage)
And I completely disagree that there's a fresh water shortage. Rainfall in the US has not decreased in the past decade and I know of no environmentalist who has even claimed that global warming would decrease rainfall (it might actually increase it!). Rainfall gives you fresh water (yes, Captain Obvious has spoken!).
Now, as for Israel & the West Bank, they don't get a whole lot of rainfall..mostly because it's a DESERT.
And my profession has spent decades studying the best way to purify water. Those pharmaceutical and manufacturing plants commonly purify their water outflows so much they are often MORE pure than the water inflow.
And, yes, it does require energy to perform, but it doesn't have to be from burning fossil fuels. They usually use electricity so a significant fraction of it comes from nuclear reactors.
Now, go rant on nuclear reactors for all the good it'll do ya, rofl.
"1.7 metric tons of material are consumed by making one PC"
Bullshit! What are we doing, fusion? The 1.5 metric tons of water doesn't disappear. It gets recycled in one way or another. Yeah, the fabrication process is very chemical intensive, but the big manufacturers (Intel, AMD) have strict environmental policies. They recycle where they get, purify their outflows, and use as little material as possible.
Both for cost-cutting sake and environmental law sake.
So that 1.5 metric tons of water is reused over and over and over in making each PC. The actual specific waste per PC should be measured as the material that leaves the manufacturing factory per day (as waste) divided by the number of pieces of hardware it made that day.
For computer geeks, you guys are really stupid.
That is, unless your PC weights 1.7 metric tons.
Duh?
Do you go to a University or live near one?
Go to your local University library (or if they allow online access..) and take a look at yesterday's issue of Science.
It has articles on both the microreactor and the human embryonic stem cells that were cloned.
I heard about Wolfram's book a while back, picked it up, and read it. Well, 90% of it. I do agree with some of the harsh remarks directed towards the patently unacademic (read, un-peer reviewed) method in which Wolfram is purporting his (supposedly) new ideas, but I think he's proposing something even worse:
His "New Kind of Science" is based purely on supposition and example..by example by example. This isn't science.
Observation is not scientific proof, it is only the first step in attempting to _explain why_ something occurs. I read a whole lot of "Here, if we make a cellular automaton use it looks a whole lot like ". Even Wolfram admits that among the millions of combinations of rules, something interesting is bound to happen. That this interesting automaton should hope to _explain_ something physically important is simply supposition, not proof that the physical world obeys some underlying automaton.
When Wolfram can devise an automaton that quantitatively _predicts_ the behavior of a physical system, then I will be impressed. Oh, with a caveat. The prediction shouldn't be a trivial one, where an already existing quantitative description exists that is simply discretized into an automaton and declared "new". (I've seen this done and it isn't impressive.)
Salis
A Cable Modem, through Time Warner, costs $40 a month.
In the suburbs, the cost for a cable modem + basic cable is around $60.
How many times has this happened to you?
You follow the directions given by Mapquest and
Feel free to add your own.
How many times has this happened to you?
...end up going down a one-way street. ...forget that bearing right could mean anything from making a turn that is less than 90 degrees right to turning the wheel an inch to the right while driving straight. ...successfully navigate yourself to BLAH BLAH Ave. South when you had typed in BLAH BLAH Ave. North. Congratulations! You are now lost in downtown! (And..no, those two streets could be miles apart and unconnected! Evil steet namers should die.)
You follow the directions given by Mapquest and
Feel free to add your own.
The Israelis that run the website are not hiding in the mountains somewhere in Pakistan. The fact that the Bush Administration put the website on the list is political...they're sending a message that they will consider certain Israeli organizations to be terrorists if they behave like terrorist Palestinian organizations.
The people who run this website may or may not have family in the U.S or might want to visit the U.S in the future. Cutting them off from that access is a punishment, certainly.
But will this work for any fundamentalist Islamic terrorist website? Hardly.
This has nothing to do with free speech and everything to do with accountability of ones' actions on the Internet. Anyone is free to spread hatred on the Internet. The Bush Admin is saying that, whenever possible, we don't want those people running the websites to be any friends of ours.
Hah!
I played the original Sojourn like in 1992. I tried Duris after they split (bad move?), but it wasn't the same (and neither was Toril).
I was "Warony". Cleric.
FYI, the lead designer of EQ, Brad McQuaid, played Sojourn aka "Aradune" or something such. Ranger. EQ felt exactly like Sojourn...making it highly successful and extremely boring for someone who played the real thing.
But, if you code in a sequence that is 30 nucleotides long as "locations" you are almost gaurenteed that the 30 nucleotide sequence is UNIQUE to the DNA.
:)
But, of course, DNA is just floating around and you don't know exactly where the DNA really is...only that if you want to target Gene X on Chromosome Y you need a XXXXXXXXXXXXXX sequence to probe.
I think what you're saying is that DNA "addresses" are less precise than circuit addresses. They probably are... but they are also more flexible in information storage and retrieval.
There are proteins that activate all genes that contain XXXXXXXX sequence. There are proteins that bind to YYYYYYYY sequence, cut it, and invert it. (now that's a real switch!)
There are sequences that flank the "data storage" area of DNA (heterochromatin), which may be removed transiently to transfer information back and forth from the 'inactive' to the 'active' section of chromatin DNA.
I agree that 'DNA computers' will never be faster than silicon circuity, but (as you said) they are parallel and _analog_ circuits. More useful for signal processing than doing arithmatic. (Think of our brains and how they are different from silicon computers...)
Organisms can also interact with the real world through chemical reactions on the 'nano' scale. A computer can't do that.