my understanding of inflation is that the expansion of space was superluminal
... for a very brief period, on the order of 1e-33 seconds.
During that period, the universe expanded by a factor of on the order of 1e26.
What had been causally connected beforehand became causally disconnected as a result.
The expansion became subluminal, leaving (from the perspective of any individual particle) islands of causal connectivity bounded by the particle horizon. The particle horizon is the present largest comoving distance from which light could have reached the observer. Today, here, that's ~78 Mly in all directions.
In GR here is also a related event horizon, which is the largest comoving distance from which a photon can ever reach an observer at some (any!) future time.
The ultimate end of the universe (crunch, asymptotic fade, big rip, false vacuum and other possibilities) has a huge impact on how the particle horizon evolves over large timescales, and whether there *is* a cosmological event horizon for an observer, and what its characteristics are (especially those that can be tested by observational cosmologists).
That being said if you could recommend a good graduate level cosmology text I'd be grateful.
Guth, Alan H. (1997). The Inflationary Universe. London: Vintage Press (Random House). ISBN 0-09-995950-X.
Although Cosmic Inflation has been evolving in the past almost ten years, the book is absorbing and gives a good overview of not only CI but cosmology in general, outlining the various cosmological problems and ways in which they have tackled. The maths are there, but there are plenty of appendices that help out people with varying exposures to cosmology, astronomy, and particle physics. As a particle physicist you will enjoy the parts that focus on the Higgs field and GUT, I imagine.
The importance of GUT work to cosmologists cannot be overstated, especially for inflationists; the large scale structure of the universe in CI is driven by gravitation as it falls out. The quark-gluon plasma's characteristics as it cools is also essential, and there are tight constraints upon these if cosmic inflation is true.
What if the small group of cosmologists pushing the "gravity is different at large distances" is correct?
Uh... well, gravitation and inertia can be firmly in the weird category (well, weirder), but most mainstream cosmology relies upon GR since it has routinely passed every unambiguous test put to it. A refinement of GR might be useful, but an overthrowing of it would raise lots of questions about why we see what we see through telescopes, and why we have been good at predicting what we will see when we look in different ways.
Returning to what triggered this (expansion and inflation), with respect to CI, there are a number of forthcoming experiments (SDSS, Planck Surveyor and some 21cm tomology) that will test CI predictions about the CMBR. So far predictions of the simpler CI models have held up remarkably well, with the key being a strong prediction of an adiabatic, nearly scale-invariant, Gaussian random field with more power at longer wavelengths. The characteristics of the field are critical in order for CI to explain that the large scale structure of the universe are caused by the gravitational collapse of perturbations during the inflationary period that were formed by QM fluctuations. If the field is not what is predicted, CI is false, and this in turn opens the door for "beyond GR" theories to answer the Big Bang cosmological problems.
So log yourself in to a 10.4.8 system with the latest developer kit (xcode-2.4) and try building something with the "-m64" flag that needs to link to e.g. -ltermcap. Oops!
There are several important Frameworks that don't support 64 bit (either architecture).
You also can't mix and match architectures when doing dynamic loading. This is especially annoying since there can be an enormous performance increase associated with a 64-bit data model in some cases. (gmp, for example, tests for and uses -m64 because of this).
Finally, lots of opensource stuff that will build on 10.4.8 has 64-bit-mode (LP64) bugs, and autoconf/libtool both really really suck when it comes to multilibrary/fat/multiarch support.:-|
So, yes, you can build programs that are 64 bit. You can even link to libsystem, some other standard libraries and Frameworks when you do so, and you won't hit syscall "gotchas" unless you try really really hard. However, you are likely to find yourself very restricted in terms of what you can do in 64 bit code, because of components that are NOT 64 bit (yet).
However, what is fairly common (if not quite pervasive) is the generation of 64 bit instructions (and the use of 64 bit doubleword GPRs) for the PPC970 (G5) which can help 32-bit-mode (ILP32) performance, particularly if it uses 64-bit maths (on the long long datatype). As a result, there is often little gain in moving from ILP32 to LP64 on the G5.
x86-64, on the other hand, with the extra registers and instructions, makes LP64 attractive purely for performance reasons, but the same "-m64" vs (default) "-m32" issues apply.
Breaking Net Neutrality violates the End to End Principal.
Principle. The "principal" would be Dave Clark.
"Net Neutrality" and the End-to-End principle are essentially orthogonal.
The Clark/Seltzer/Reed argument was that an open-loop congestion-avoiding datagram mode protocol would perform better as a generic bulk transport than any admission-controlled or per-hop-error-corrected one. Except for unusual corner cases, the argument is generallly considered to be right, and is the basis for TCP and other RFC-2001-style congstion-avoiding transports.
The unusual corner cases which are relevant today are where the open control loop is too long (time-wise) for any retransmitted data to be usefully fresh upon arrival at the ultimate destination, and where there are large numbers of receivers.
In the past -- especially in the original End-to-end argument advanced in 1981 -- there was a great deal of argument about the value of offloading processing from the end systems to intermediate systems, consolidating instead of distributing processing. Distribution has better scaling properties in large networks, and full consolidation was not possible in any event, as even in fully circuit-switched networks the end systems perform the same computational tasks (validating data, recovering from corruption, controlling admission/emission) as between any of the intermediate systems along the path.
End-to-end argues that these computational tasks are only needed at the end systems, and not in any of the intermediate systems. While the "only" part is debated from time to time, the Internet has scaled extremely well surviving not only large increases in the number of end systems and total traffic, but also huge and sudden fluctuations in traffic patterns. In the mid-1990s, there was concern about the then dominant traffic pattern, namely very short TCP connections which would never exit slow start (in this case, HTTP instead of FTP). Earlier there was concern about very bursty, transactional TCP connections (NNTP, for example). Now, the usual mode is TCP bulk transfers of about 9MBytes between entirely arbitrary end systems (file sharing and other P2P apps), while the current concern is for applications whose data become stale within a round trip time between source and destination (VOIP, live (rather than recorded) TV), which is inherently bursty, and which has to navigate paths which can be probed at any moment by a long-lived TCP bulk transfer seeking to reach equilibrium (this involves bottleneck bandwidth exploration and will change the time it takes for traffic to exit the bottleneck's queue).
There are a number of areas where TCP is a poor fit. Some of these problems go beyond the Internet however: go to a sports bar with multiple screens, especially one which is receiving multiple transmissions of the same live sporting event, e.g. standard definition broadcast over the air and high definition over digital cable or satellite... compare clocks. Alternatively, try to have a phone conversation with someone in central Kyrgyzstan from your U.S. mobile phone during periods of maximum solar noise and count the number of person-to-person retransmissions and back-offs ("Say again? I didn't hear that because of the static." and "Oops, go ahead and talk.")
A variety of approaches wherein the network service offering is altered to accomodate these situations exist (and many more are proposed). A handful of end-to-end approaches also exist for many of these problems, are deployed, and are working reasonably well (some aren't easy or beautiful though). Finally, there are hybrid approaches wherein end-to-end RFC-2001-like protocols are carried over links which deal with local congestion and corruption without the participation of the end systems.
End-to-end isn't really concerned about homogeneity of the intermediate systems as much as it is about distributing congestion avoidance and recovery from data corruption to the end
Nothing in the news article, or in the commentary written on the WoW board is suggestive of impropriety on the part of any of the police or CBSA officers.
I am a little sceptical of people who claim to have an argument with officials in any situation remotely as stressful as this one, but even if we accept the dialogue as accurate reporting rather than coloured by artistic licence, the worst accusation that can be levelled is that the police (likely Ottawa-Carleton regulars, or duty Mounties) asked an inappropriate question ("if you were drunk..."), which would best be followed up with his or her boss, in writing.
The Charter regulates the activity of everyone involved in the government of Canada (s.32(1)(a)) except in unusual circumstances backed by statute that survives court scrutiny. The Federal Court (which enforces the Charter per s.25) generally anticipates the Supreme Court of Canada's take on these exceptions, which I think you describe reasonably.
The (federal) loopholes in the Charter in general are only interesting with respect to legislation passed by Parliament, and this includes the gigantic one in section 1, which you describe. (It's s.1 of the Canadian Charter of Rights and Freedoms (citation as per s.34), not the Constitution Act, 1982 (Schedule B, Part I), which it also is, and isn't, simultaneously. The Constitution is fickle about citations mainly because much of it originated as Acts of the UK Parliament, including the Canada Act, 1982 (UK). Ugh.).
The most relevant sections of the Charter have universal applicability ("Everyone has the right..." ss.7-10,12) and unlimited situational applicability ("Any person charged" s.11; "A witness..." s.13; "A party or witness..." s.14) and importantly s.15(1) "Every individual is equal before and under the law and has the right to equal protection and equal benefit of the law without discrimination...". This applies to non-Canadians and Canadians equally, and it also applies outside Canada, with respect to Canadian officials performing government activities.
When you are in Canada physically, you are in Canada legally. "Sterile" areas and flights are still bound by Canadian law.
There is a singular lack of obvious grounds of action (or complaint, except for rudeness) against any Canadian official in the story recounted by the WoW poster.
The general policy of "land, isolate, interview, interview again" makes overreaction likely, however, and sadly the policy is the responsibility of Stockwell Day, the minister for Public Safety (Mon-Sat only).
Let us hope, for Tony Blairs sake, that the terror plot is not a case of "muslim buys DVD".
Let us hope for everyone else's sake that Tony Blair (authoritarian, self-righteous, strong supporter of the Bush Administration, and surrounded by people who encourage public hysteria) is thrown out of office sooner rather than later. Whether he exits "gracefully", is turfed out by his own party, or is defeated at an election triggered by yet another lost vote in the House of Commons matters little. Whatever's fastest, please. (And hope he takes John Reid with him).
No, they do not have complete authority. The Canadian Charter of Rights and Freedoms s.2 starts with:
2. Everyone has the following fundamental freedoms: a) freedom of conscience and religion; b) freedom of thought, belief, opinion and expression, including freedom of the press and other media of communication; c) freedom of peaceful assembly; and d) freedom of association.
Everyone. Not "everyone in Canada", not "every Canadian". The universality of the Charter was chosen to be clear and unquestionable, and it repeats itself throughout the sections headed with "... Rights".
7. Everyone has the right to life, liberty and security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice.
8. Everyone has the right to be secure against unreasonable search or seizure.
9. Everyone has the right not to be arbitrarily detained or imprisoned.
10. Everyone has the right on arrest or detention...
11. Any person charged with an offence has the right...
12. Everyone has the right not to be subjected to any cruel and unusual treatment or punishment...
etc.
Note that I left out section 6: "6. (1) Every citizen of Canada has the right to enter, remain in and leave Canada."
Immigration officials with the Canadian Border Service Agency may exclude non-Canadians from Canada under statute law and regulation in accordance with the Charter.) No statute, regulation, or administrative act by the government is legal if it contravenes the universal rights to due process guaranteed everyone by the Charter. It is not legal to deprive non-Canadians of their rights, and s.24 remedies are regularly meted out by the courts in such cases.
There is a detailed set of administrative processes and judicial reviews involved when excluding or removing people from Canada. (Here is an overview).
Note that even persons constrained by statute law from appealing such an order may also resort to the Federal Court in two main (and very different) ways. However, CBSA/CIC will always be very careful about exclusion and removal order processes.
CBSA agents may also make arrests. There is an overview here. Note that even Canadian citizens may be arrested and detained.
Arrests and detention are serious in Canada, and the arresting officer and all his superiors are open to a variety of civil suits, section 24 remedies, and even criminal charges, if they do not follow the appropriate processes and are not especially conscious of the Charter. A number of prominent lawyers offer pro bono services in many such cases.
Since Canada is country with many immigrants and international travellers, especially concentrated in major cities, a variety of media organizations tend to be interested in the activities of the CBSA, especially given who the current Public Safety minister http://en.wikipedia.org/wiki/Stockwell_Day >idiot happens to be.
Canada is certianly not perfect, and the CBSA can be unfriendly especially at busy times, however "don't fool around" should not be confused with "you are liable to be shot", and "I want to get this over with too" is not the same as "I can abuse you if you don't cooperate".
You are under-informed. While a market for cheap connectivity for residential users who do not suck down gigs and gigs of traffic is developing -- one popular approach is "prime time" download limits (or extra charges for heavy daytime downloaders) and unlimited overnight/weekend downloads -- there are many competing ISPs. A tiny bit of research would have found you (for example) Bulldog and ntl, which offers unlimited downloading at rates not much higher than their competitors whose offerings come with download limits. A tiny bit more research would indicate that the couple pounds a month savings in exchange for download limits is worth it to many -- perhaps most -- users. High-speed Internet access is useful for many things beyond pure bulk downloading.
As a result, ISPs can reduce some of their operating costs and thus lower their prices, if they buy their long-haul/international connectivity on a $/peak Mbps/month basis (which is common). Bulldog is one of a handful of UK ISPs who operate a substantial international network on their own, and whose costs are not directly influenced by traffic loads, and who can therefore offer cheaper and less-capped services to atypically heavy users.
I wish, for one second that libtards would THINK
It has nothing to do with libtards. This is simply how the market has developed. Many customers are price-sensitive; there is sufficient competition that many suppliers are cost-sensitive (margins are important!); many suppliers' costs are directly influenced by the 95th percentile 5-minute-average traffic loads delivered from their international carrier(s), because the wholesale/long-haul market has developed that way over the years. Caps and limits control those costs.
However, different suppliers have dramatically different cost structures, and so there are a range of offerings that one can pick and choose from. Including unlimited download multimegabit per second broadband for prices competitive with what you are paying.
Here's a reasonably well-maintained price list that illustrates much of the variety of packages available in the market, including a variety of usage caps and no-cap services. It was easy to find a standard offering directly comparable to yours (you claim a much higher upload rate than is listed here, however this is not a firm cap rather than an estimate of the expected statmux loss on cable plant, which differs from asymmetrical DSL).
However, again, the UK market is expensive compared to other national markets in the European Union. The incumbent was very good at playing politics and ended up with an early set of changes in its regulatory environment, and a series of tame regulators. This has changed recently, and market price erosion has been substantial. Hopefully people will be able to tell you about cheaper and cheaper prices still over the course of the next months, since otherwise the artificial market approach would be failing.
Only Sweden, Finland, and Norway have lower average population densities. I don't see this article extolling the virtues of their telco competition....
Sweden was the first country in the EU to begin telecomms deregulation, in 1981 (mobile telephony), 1986 (cable television), 1991 (Internet connectivity and long haul leased lines) and 1997 (full service voice telephony).
Sweden has by far the most well-developed last-mile services. In the three major cities (Stockholm, Gothenburg, Malmö) it is common to have a fast ethernet drop from one's apartment to a service cupboard with a switch that has 8 other fast ethernet ports (for your immediate neighbours) and a gigabit ethernet or other high-speed optical connection (IEEE 802.17 is popular, as is Packet-over-SDH) to an ISP. In minor cities there is still vigorous competition for high single-digit Mbps connections, and double-digit Mbps ones are usually available. Even in most of Sweden's rural areas (which are very sparsely populuated), Mbps or near-Mbps connections can be delivered reasonably affordably.
Stockholm was a major focal point for Internet connectivity in Europe throughout the 1990s -- a commercial ISP and a consortium of the Nordic countries' national research networks shared what were always the fastest connections across the Atlantic ocean: multiples of 2Mbps in the early 1990s, the first ever trans-atlantic 34Mbps terrestrial connection, the first ever trans-atlantic 155Mbps connection, and so forth, increasing every two years or so. Stockholm was the focus in large part because the regulatory environment encouraged price erosion on the Swedish half-circuit. At the time, European half-circuits were often 2-50x (yes, really) more expensive than the cost of the U.S. end of an international leased line, while pricing in Sweden could be cheaper than in the U.S. in some cases.
Sweden's lead on costs, and Stockholm as a hub of business and technology, encouraged many foreign telecomms companies to begin offering services in the city. As a result, there were no fewer than four competing companies undertaking city-wide fibre deployments, which meant major streets being restricted by a lane or two here and there on a regular basis, worsening the city's traffic congestion. The city itself solved this problem neatly by forming a city agency (Stokab) which laid down fibre and ducting in the city as a matter of course, whenever other works (water and sewer and electricity and gas line repairs) would be digging up a road anyway. Moreover, Stokab offered fibre between any two locations in the city on a cost-recovery basis to anyone including the incumbent telecomms and cable companies. This was a huge success for both telecommunications development (ISPs routinely offering fibre-to-the-basement of small apartment buildings, gigabit inter-office connections, etc.). The Stokab approach has since spread a bit beyond Stockholm.
The other three continental Nordic countries caught on eventually, Finland leading the rest of the EU, and Denmark and Norway (the latter is not in the EU) lagging behind somewhat.
Short of government intervention forcing the issue, the only thing that will cause consumers to see better internet service is the introduction of a disruptive force. That means adding new competitors.
Yes. You are exactly right. Adding new competitors is the key. That is exactly what deregulation is trying to do in Europe.
In the mid-1990s, the European Commission (notably then-DGIV (now the Competition Directorate) and then-DGXIII (now the Information Society and Media Directorate)) began developing the deregulation of telecommunications services in the European Union.
Rather than the U.S.-style policy of focusing on immediate benefit to consumer (notably falling prices or offering of new services), the approach of the Commission was to establish artificial markets in which barriers to e
They only kill one strain of bacteria. Will consumers (and meat packers!) get a false sense of security, get sloppy, and wind up with some different strain of bacteria poisoning the meat?
There are lots of bacteria happy to colonize meat. Most of the ones that cause human (or pet) diseases are readily controlled by careful packaging, storage, preparation/handling and cooking. Cook your food thoroughly and wash your hands carefully before eating, and you likely will never run into food poisoning by any of them.
Listeria monocytogenes is an exceptionally hardy bacterium for one that does not form spores. It will grow in packed refrigerated meats, for example. It will survive freezing, and can survive cooking. (It's also very tolerant of dessication.) It readily causes gastrointestinal troubles for healthy people, and can cause serious symptoms. Rarely, it will prove fatal.
You are right that L. mono. has plenty of opportunities to adapt to environmental pressures. Viral infection is one of these -- "that which L. mono. survives, makes it stronger". However, bacteriophages co-evolve with their target bacteria. Their replication and assembly mechanisms are error-prone, and typically billions of copies are made before the infected cell lyses. This leads to even greater opportunties for genetic changes than the bacteria have.
Most evolved mechanisms of resistance are energy intensive, so resistant organisms typically don't overwhelm unresistant organisms in the wild.
There is, however, a possibility of a locally resistant strain arising in e.g. a meat-packing plant making routine use of the bacteriophage spray.
This will typically be noticed through inspection. The approach to this situation is straightforward: find a virus that infects the new strain, and breed it (by letting the infection run its course through a culture of the new strain). "Finding" may involve simply looking around the meat packing plant, or more simply (and perhaps just as effectively), turning to ponds and soils wherein L. mono. is naturally found. Scooping up a bunch of the surrounding scum is liable to net some viruses which can infect the new strain.
Irradiation seems better. It will rearrange some molecules...
Phage treatments remain active for a longer duration than a flash irradiation, and do not change the nature of the food medium they are applied to. Gamma irradiation denatures food material as well as bacterial material.
Gamma irradiation does not worry me, but some people appreciate more conservative approaches to pathogen control.
Finally, given the differences in difficulty of preparation and delivery of the anti-listeria agents, phage treatments are almost certainly much less expensive than using ionizing radiation, especially factoring in the various regulatory realms involved.
Phages aren't genetically engineered (now). They are obtained from the same environment that the target bacterium inhabits. Viral infections are natural for bacteria like listeria. The method of preparing phage treatments is simple: find and take an infected population (from the wild, or from the meat packing plant) and mix it with an ordinary bacterial culture, let the viral infection run its course, then extract the viral particles. Breeding techniques and test automation can make the "find and take" process more efficient.
As other people have noted, viruses also adapt to selection pressures (and quickly, since the failure rate and scale of virion manufacture is large), so viruses are likely to catch up with any but the most dramatically runaway acquired immunity in its target population.
Introducing lots of listeriophages into the environment might make Listeria monocytogenes less common, or press it into virtual extinction? Oh noes! What a horror. We shouldn't make listeriosis (at least locally) extinct, because the mostly domesticated 37ish mammalian species (and 15ish birds) it infects, including humans, might undergo a population explosion?
I don't really see the downside other than fewer sick animals.
Unfortunately, the propagation mechanism is very local -- sprayed phages will get carried around by droplets and as dust. There are listeria reservoirs in places far away and well isolated from western meat packing plants. Unlike the bacteria, the phages cannot replicate themselves -- they can only replicate in the presence of susceptible bacteria. If they're ingested by grazing animals or flying birds or the like, they probably will not survive passage through the digestive system, or attack by the animal's own immune system. Thus transportation to an area with as-yet-uninfected bacteria is unlikely. L. mono. is very hardy for a nonsporulating bacterium (that's why it's dangerous) is much more hardy than the several viruses that prey upon it.
What are the chances that we will be surprised by a newspaper article decrying the death of 100 elderly because they had (sprayed) luncheon meat in which very rare but virus-immune bacteria had developed (and had chance to develop because standards of hygiene went down and the meat was kept out of the fridge for say 24 hours)
Much less than the chance today of listeriosis acquired by refrigerated (or even frozen) and sealed packed meats that are well protected against e.g. E. coli and a range of other non-sporulating food-colonizing bacteria.
Leaving meat exposed outside of a fridge for 24h will also breed a range of moulds and bacteria, notably salmonella.
Bacteriophages aren't instant, so it's more useful to spray a solution on the meat at a packing plant, than on unwrapped meats.
(So it's still not a good idea to spray on an anti-listeria preparation and re-freeze or refrigerate unwrapped and left-out meats and the like.)
These bacteria grows in e.g. meat and are readily killed by cooking, so Listeria only has a chance when meat is kept uncooked.
those with a weakened immune system or special vulnerabilities can simply take special care
With listeria, it's not so simple: the bacterium also survives considerable heating (sometimes even cooking) and dessication, as well as being able to reproduce in fridge temperature ranges, and survive deep freezing.
Careful storage, cooking and preserving, in other words, can fail to eliminate listeria. Listeriosis produces a range of symptoms, some serious, and while fatalities are rare, you really wouldn't enjoy the gastrointestinal upset it produces in mild cases in an otherwise-healthy individual.
Other people and comments have addressed the co-evolution of phage and bacterium, compared with antibiotic resistance. However, bear in mind that methicillin (the M in MRSA) is
Yoghurt isn't animals (the active ingredients are bacteria in the Lactobacillus genus, mainly L. acidophilus and L. bifidus.
These bacteria, like all others, are unicellular organisms, and so lack bowels (teeming with lots of cells of all sorts within and surrounded by lots of other cells belonging to the animal itself) and blood (teeming with lots of cells belonging to the animal).
On the other hand, there are cognates... in Lactobacilli there are a number of vesicles and nutrient storage structures that might do as a sort of bowel kinda, and there is of course a cytosol, which is sorta like the non-cellular part of animal blood.
Sadly, it is not the crushing (with your teeth, presumably) which will let you spill out their bowels (chewing hard on yoghurt likely won't harm a single bacterium out of the billions in each spoonful), but rather acids and enzymes found much further down your digestive system. Most of them, in fact, will be found in your bowels (and descending colon), and won't really be produced by you per se, but rather by the other microorganisms dwelling there.
L. acidophilus actually mostly helps you spill open much less friendly bacteria so that you can feast upon their "bowels" and "blood".
The phage culture would be added to whatever washing fluid is used to remove fecal matter and other contaminants during the butchering process(es).
The bacteriophage viruses are easy to grow at any lab which can culture the relevant bacteria. The bacteria is the phage "food".
So all that's necessary is shipping out some purified phage solution (or even as a powder) to meat packing plants every so often.
This is an inexpensive way of preventing a number of diseases associated with handling or eating uncooked or improperly cooked meat. Some of these diseases are extremely unpleasant for the victim, even if the number of fatalities a year is low.
(It can also be applied to other foodstuffs which are subject to common infections by microorganisms for whom there is a natural "predator" phage virus).
Bacteriophage viruses also mutate, and usually at a faster rate than the bacteria they attack (this is mostly because duplication of viral RNA sequences and assembly of virion particles are especially error prone processes, and many many copies are made before the bacteria is lysed). Bacteriophage therapy thus has the handy property that any resistance acquired by the target bacterium is almost certain to be met by an equivalent viral adaptation, all occuring within the bacterial growth medium, whether that's a slab of beef or a human patient.
Any mutation in the host bacterium, moreover, is likely to have been seen in its natural environmental reservoirs and thus there is likely already an equivalent reservoir of already-adapted phage.
Procedure: extract phage particles from pond scum (literally, in many cases). Add to culture of resistant bacterium. When the culture is clearly succumbing to a viral infection, extract and purify the phage particles. Inject into patient with resistant bacterium. Repeat as necessary.
The main downside to this approach to acquired resistance to phage therapy is that culturing the resistant strain takes time. Meanwhile, the resistant strain is infecting the patient. If the infection is especially aggressive, or the patient is especially weakened by it, the infection will outrun the preparation of the new viruses. (In that case, use wider-spectrum agents like western antibiotics, hyperbaric oxygenation, debridement, amputation, etc. as necessary).
Using phage therapy in cattle (or on dead beef as a surface treatment) against known strains of known pathogenic (to humans) bacteria is sensible: firstly, you reduce the incidence of sick humans who handle uncooked meat, who eat improperly cooked meat, or who eat cooked meat in which bacteriotoxins have survived; and secondly, you do not breed antibiotic-resistant bacteria on the surface of the cuts of beef.
Treating the surface and near subsufrace of dead meat also has the positive advantage that the dead animal no longer has an immune system to interfere with the phage's attack on its target bacteria (this is also unfortunately why bacterial contamination of untreated dead meat is so common).
But the question is really whether your data can be decorrelated and correlated (or scattered across the wavelengths and gathered up on the receive side) across that many channels.
The important question in packet-based networking is when the last bit of a packet arrives, not when the first one does.
Let's make it 10 000 channels of 1000 modulations per second versus one single channel of 10 000 000 modulations per second, so the aggregate rate for the serial versus parallel link is identical (10 mbps).
If all your L3 traffic is expressable as 10 000 bits per conceptual L2 frame, then in both cases you get 1000 pps; one thousandth of a packet per "blink" on the serial link, one packet per "blink" on the parallel one.
If your L3 traffic is less than 10000 bits (and the link is not a bottleneck) then it takes one blink to send the L3 traffic across the parallel link no matter what the L3 packet size is. This means that your minimum link latency is one millisecond on the parallel link. The serial link's latency is a function of packet size; smaller packets take less time for the last bit to arrive; you save a microsecond per bit.
For packets that are larger, you get into situations where you have, for instance, 10001 bits to send across an otherwise empty link. On the serial channel, this takes you 0.1 microseconds more than a frame of 10000 bits. On the parallel channel, 10001 bits takes a whole millisecond (10 000 microseconds) longer.
This is certainly an effect which can be discovered end-to-end (akin to MTU exploration), and could be adapted to by the transmitter, however there are probably no in-use transfer protocols which deliberately look for step function increases in packet size to delay.
Moreover, when the link is a bottleneck (which is expected at every link across which TCP bulk transfers happen) and variable sized packets, scheduling becomes a problem, if you want to avoid the step-function from being a serious source of jitter. If you have multiple packet flows occupying the queue, you inevitably end up with some packets which will suffer from millisecond jumps in delay for the final bit at the receiver, compared to the serial link. While this is not an intractable problem, minimizing the incidence of partial-packet delay across streams of variable-sized packets is a difficult one, particularly as you scale from megabits/second aggregates to gigabits/second aggregates (one 10GHz channel vs ten thousand megabit/second channels), and as you increase the number (also the variability of number) and end-to-end throughput (also the burstiness) of independent flows.
This will be an important problem for Internet Engineering in the future, as there are difficult price challenges in 40Gbps/optical channel now, and difficult price and physics challenges in anything faster than that. Slashdot readers familiar with megahertz versus cores will probably notice an echo.
Also of interest (and stuck in the realm of routing and control plane research): with your serial channel, a failure anywhere in the path disables the whole 10 Mbps channel. With the parallel channel, you introduce a whole range of partial failures that wipe out one (or a handful) of parallel colours. It may be better for a network's users to continue to use even 1/2 of the 10 000 parallel channels (depending on things like traffic load, how a failure of the whole channel is dealt with at various restoration or routing levels, and so on).
Pseudo-parallel connections are even worse; there are a variety of situations in which the receive side detects simultaneously transmitted pulses of two frequencies at different times...
In short, massively parallel links in theory should be no worse than equivalent serial (or even "lightly" parallel) ones when it comes to goodput; however, we don't know how to do this in practice yet.
So, sadly, for real world Internet traffic, I would not be surprised if a 1 000 000 blink/second serial channel performs better than 10 000 parallel channels of 1000 blink/second each.
Though the senate tries to be the more 'stable' of the branches of government by having 6 year terms, it seems that they are just as concerned about public appearances, if not more so, than the members of the house
This is largely because the members of the House of Representatives have more leeway in, and are much better at, choosing their electors. Senators are elected at the state level, and changing the borders between states is not as easy as changing the borders between the smaller electoral districts within them.
Senators face a state worth of electors every six years; members of the House of Representatives have two years to swap pockets of voters who would vote for another candidate in their district for pockets of voters more inclined to vote for him or her, but who are living in a nearby district controlled by a member of his or her own party who wins by larger majorities, or by a member of the opposite party thus improving the electoral chances of both incumbent members of Congress.
But for significant changes in cell organisation that occur in the same random manner, probabilities are very low, because some mutations need to occur together in order to be non-destructive.
Probabilities are very low, and most mutations are destructive.
The most significant multi-mutation events are found in sexually reproducing organisms during meiosis and fertilization.
The expected result when there is this sort of error is a dead gamete or a dead zygote.
This is very common... malformed or non-viable seeds in plants; malformed or dead eggs in ovipositing animals; miscarriage in viviparous ones.
In humans, for example, early miscarriage has come under quite a bit of study because it turns out to be very common -- the embryo dies and is expelled without the woman ever knowing she was pregnant. A British Medical Journal study is consistent with reports of 11-16% miscarriage rates in known pregnancies. Some early miscarriages are largely asymtomatic, and there are studies underway in women likely to become pregnant that expect a doubling of the known rate.
In [Hogge, W.A. The Clinical Use of Karyotyping Spontaneous Abortions. American Journal of Obstetrics and Gynecology, volume 189, number 2, August 2003, pp 397-402] 70% of studied first trimester miscarriages were caused by chromosomal abnormalities, with the bulk attributable to defects in the sperm or egg cells involved.
So in humans, some 5-10% of pregnancies end because of a cluster of unfavourable germline mutations.
That's around 800 000 terminal mutations per year, globally, in humans alone.
Humans gestate over a period of nine months, and presently are producing less than three offspring per woman. There are species with comparable miscarriage rates who gestate much more quickly and produce many more offspring, and are also more populous than humans, and which consequently get to "test out" a larger variety of random multi-mutation combinations over the course of ten years than humans.
So yes, the probabilities are very low, but the number of tries is very high, particularly over the course of millions of years.
I am saying that if we rewind and play, replicating the perfection that is nature will be virtually impossible.
What we know of nature is that it isn't a perfect environment for life, it's actually pretty harsh and filled with serious hazards. However, life has adapted successfully to the harshness, when considering all life, rather than individual species (many of which have become extinct).
"Successfully" means living things are still producing viable offspring right now.
A "rewind and play" of the environment with the same biomass at the time of the first photosynthesizing autotrophs, likely would lead to different species than we have now, but there would be environmental niches readily occupied by land plants, aquatic heterotrophs, swimmers, aquatic predators, land grazers, taller hardier plants, land predators, and so forth.
On this scale, probabilities are so chaotic, the question of whether oak trees, carp, thylacines, or human beings might arise is a philosophical one, along the lines of the cosmological anthropic principle.
2) The changes need to be beneficial for the organism with respect to others that lack this particular mutation.
No, heritable changes simply need to be non-harmful, as per your point (1).
If the changes are propagated to offspring which in turn propagate the changes, the change is likely to remain in the population, as long as the changed individuals reproduce at the same rate as the unchanged ones.
In other words, with respect to the production of viable descendants, most random changes are likely to reduce the number of offspring per changed organism (sometimes the reduction is down to zero). Of the remainder of chainges, the majority is likely to make no difference to individual reproductive rate as long as the environment remains much as it is. A tiny minority of changes will provide some advantage over unchanged indivuals in the present environment.
Environmental changes can cause new traits which are neutral to become positive or negative, or to make positive changes neutral or negative, or to make negative changes neutral or positive.
Mutations are simply random. An evaluation of their "constructiveness" (using your term) can only be done retrospectively by counting the prevalence within a population of the inherited mutant trait in a given generation.
What is "constructive" today may be burdensome tomorrow. That is several generations from today in many microbial species.
Simultaneity of mutations does happen sometimes -- DNA replication processes are error-prone, and the error rate varies from species to species. There are a number of types of errors that can cause several simultaneous mutations. Again, the success of these can only be evaluated in hindsight, as with point mutations.
However, successful mutations more likely are spread out over time. Highly developed eyes are good examples.
The cephalopod eye (in octopuses and the like) started as a photosensitive pigment spot on the skin surface of an ancestor. Shine light on it, and it produces and releases hormones, much like shining different frequencies of UV light on a melanocyte in your skin produces and releases melanin. These hormones excite nearby nerve cells. Mutations which tighten the binding (increasing the speed at which changes in light are propagated) between the pigment-containing cells and the nerves would help in avoiding predators or finding prey, for example. In cephalopod ancestors, a gene would influence whether the pigment cells were flat against the surface of the skin, protruding above it like a mole, or sunken below it in a pit. Each of these has advantages -- protrusion enables a wider field of view, invagination enables a detection of directionality. Invagination won. A further set of genes influenced whether the pocket the pigment cells rested in widened like a bowl, or narrowed like a bottleneck. Narrowing won, and as a result a camera obscura eye was the result. A subsequent mutation allowed camera obscura squid ancestors to produce crystallins through a mutation of one of the several genes which produced enzymes. Crystallin production around the opening of the camera obscura eye evolved into a useful lens.
Vertebrate eyes evolved differently: an early ancestor (an early form of sea squirt) produced a crystallin (with some enzymatic uses) within its central nervous system; it also had a primitive light detection system based on pigmented cells also located within its central nervous system. The closer the pigmented cells were to the outer surface of the organism, the better at detecting predators and the like. This was enabled by a series of mutations producing an ever longer chain of nerve cells connecting the pigmented cells to the CNS. Later descendants gained some sight advantages by producing crystallins near the photodetecting cells, as well as by producing different pigments. Eventually a pre-vertebrate ancestor's eye evaginated thanks to a genetic change, with the lens protrudi
Let's tackle this one, since it explains a lot of the "we don't know" answers you'd get to many of the other good questions on your list that evolutionary biologists, paleontologists and so forth are also very keen on answering.
The answer is: they died out, and may be in the fossil record.
The problem is that fossilization is extremely uncommon.
Most dead organisms and their byproducts decompose rather than fossilize. Many organisms that die in conditions where fossilization may occur are unsuitable for fossilization. Inclusion fossils, like insects, plants microbes trapped in fossil resin (amber), are unsuitable for large animals because they are simply too large to be covered in decomposition-resisting resin, or which will largely decompose anyway. Permineralized fossils are extremely useful, but the organism has to be completely covered in suitable sediment very soon after death. This is rare. Partial coverage leads to no fossil, as decomposing agents would ultimately reach the whole organism. Similarly, mould and cast fossils can be useful even just for the impressions within them, but the conditions for their formation are also rare.
Also, different tissues react to fossilization processes in different ways. Harder, more stable tissues, especially polymers, are more likely to be preserved. Softer tissues are much more rarely preserved. It is more common to find a fossil which is simply chitinous plates, bones, or feathers than to find any preservation including flesh or connective tissue.
Different organisms also live and lived in different areas. A large number of organisms may never have been in an area where fossilization was possible. A large number of organisms' fossils are in probably in inaccessible places (under modern-day ice caps, in modern-day deep water, under lots of desert sand).
Transitional species which did not die in the right places (currently known fossil fields) at the right times (when fossilization was possible) and under the right circumstances, with the right "body" parts, in the right microbial company (mostly obligate aerobes and non-decomposing anaerobes), and in places where people have thought to look and have the budget to dig in, are obviously unlikely to be discovered.
Even among those whose fossils are plausibly "missing", there will be transitional species which had short-lived or small populations (or both). Fossils do not give much detail about internal morphology or fine differences in biochemistry. So there will also be transitional and intermediate species which are indistinguishable via examinations of fossils from their bracketing forms.
Gaps certainly exist in the fossil record. This means data useful for answering many of your questions are not available. Other data may be found, and they should concur with predictions about the "missing" fossils.
That said, there are plenty of predicted transitional and intermediate species which have been found in the fossil record which corroborate and illustrate macroevolution. Horses and whales are well known families with plenty of fossil evidence of adaptations to known changes in their environment, and there are a number of significant fossils that illustrate major changes in vertebrates which are consistent with common descent. There are fossils which are consistent with the current theory of reptile->bird evolution, and fortunately feathers and scales and things which look like fluffy and striated scales permineralize well, as do bones, leaving fossils of great detail. Recent finds (reported on slashdot even!) include gliding reptiles and non-flying bird-like animals that most likely used their not-quite-wings for stability while running (like ostriches).
A number of your other questions cannot be answered in the fossil record, but there are theories which predict fossils which are likely to be discovered. These theories are becoming in
D'oh! What you've described is Darwinian, the clue being "environmental stressor", which is synonymous with natural selection. More specifically, you're describing the modern evolutionary synthesis.
Selection pressures result in some genes passing into a population. You've described exactly that, and roughly produced a biological mechanism that is part of the process.
Unfortunately you use "animal" rather than "organism" (plants, bacteria and the rest evolve too!) and are too focused on protein folding, whereas differences in whether or not a protein is expressed are much more important in e.g. the evolution of bacterial pathogens. For example, with respect to resistance to the penicillin family of antibiotics, one gene expresses beta-lactamase, another produces penicillinase. Express the gene, gain resistance. Don't express the gene, get wiped out by common antibiotics. To be fair, a third (and less common) mode of resistance produces a variant penicillin binding protein with a reduced affinity for beta-lactam rings, and the chemical variation in the PBP will inevitably produce a protein with a different geometry. The variant PBP bacteria rapidly fall out of populations not actively stressed by beta-lactam antibiotics, as the variant PBPs dramatically increases the energy cost to the bacteriam of building its cell wall.
Chaperone proteins? These are mostly expressed in prokaryotes as a heat shock defence and in eukaryotes to a lesser extent in endosymbionts (mitochondria et al), endoplastic reticuli and the cytosol, where they assist in transport and in protein degradation. Removing chaperonins in prokaryotes leads to an organism easily cooked to death at relatively low temperatures. Turning them off in eukaryotes exposes the cell to protein aggregation diseases, or kills the mitochondria (and thus the cell). The most obvious way of turning chaperonins off is to deprive them of ATP, which is another good way of killing a cell.
Known chaperone proteins are described by genes which are governed via Mendelian inheritance.
There are people doing research in the area of conformational changes as an amplifier or attenuator of expression of phenotype. There are also people doing research in the area of non-genetic heritable traits, just as there were people studying the trait inheritance mechanism in endosymbionts. To argue that they have supplanted genetic heritability is a stretch. These broad groups may be uncovering a new, parallel trait propagation mechanism, akin to endosymbiont DNA versus nuclear DNA. However, most of the work seems focused on rapid expression of genotype mutations in conditions of environmental stress, and while terms like "Lamarckism" get thrown around, that does not seem to be what these researchers are reporting. To suggest that this sort of work is trying to "kill" natural selection just seems wrong.
I'm still looking for an example where speciation lead to one group having a different number of chromosomes than another group.
Let's start with differences in ploidy. It happens often in some plants, it's readily observable, and it's an important factor in plant speciation (natural and through breeding).
Wheat is straightforward: there are wheat species that are diploid (Einkorn wheat), tetraploid (durum), and hexaploid (bread wheat). The last evolved in farm fields, and the middle evolved in the wild and is the result of a non-human-influenced hybridization of two diploid wild grasses.
Apple species, tulip species and lilly species vary in their ploidy as well.
With respect to changes in chromosomal number other than changes in ploidy, during meiosis, homologous pairs can fail to segregate properly (non-disjunction), leading to monosomy (where one of the chromosomal pair is missing in a normally diploid organism) or trisomy (where there is an additional chromosome attached to the chromosomal pair, again in a normally diploid organism). These are not exceptionally important factors in speciation, as aneuploidy rarely results in reproductive advantage for the organism affected (and often is disadvantageous).
A pair of individuals with heritable monosomy may produce viable offspring that are missing a full chromosomal pair (or euploid "set" in a normally non-diploid organism). This is more likely in organisms with many or very small chromosomes, or in amphiploid organisms which are cytochemically of lower ploidy than they are in terms of reproduction (e.g. cotton, which has four sets of chromosomes, but behaves like a strict diploid).
So heritable change in the number of chromosomes is observed in the wild and in the lab (or field), and while this usually creates significant phenotypal differences when it happens, the offspring between pairs of similarly mutated organisms is -- occasionally -- viable.
Multiply "occasionally" by many generations and the result is that the possibility of two species with a common ancestor may have different numbers of chromosomes is entirely plausible.
Finally, mutations within a chromosome are simply much less dramatic (in terms of being likely to help or harm (or neither) an organism) and are more likely to be caused by environmental factors than mutations involving the addition or deletion of entire chromosomes, and so the former will be observed in a natural population much more often than the latter.
Evolution is really about mutation, that is where the real fun happens. And mutation is largely random, based on the physical state of the universe at any given time.
Mutation is largely random because DNA and especially RNA are structurally frail and break when exposed to e.g. ionizing radiation, or oxidizers and other mutagens prevalent in the natural environment here on Earth. All cellular organisms have mechanisms which protect against and repair damage to their in-cell DNA. Some of these are passive structural protections, others are active molecules, particularly repair polymerases. Repair polymerases are imperfect (some are highly error-prone), and will not be able to repair all externally-driven changes to genetic material.
Unrepaired changes which are passed down to a cell's descendants, as well as replication errors introduced in the process of spawning descendants, are relevant to evolution; those which prevent a cell from producing descendants in the first place are not. In multicellular organisms, the multicellular descendants are considered instead of those of constituent cells not directly related to producing the multicellular offspring. In such organisms, the mutations are split between heritable and non-heritable mutations. Usually these are described as somatic and germline mutations, however some somatic changes are heritable in some multicellular organisms.
Heritable changes are the mechanism of evolution, and the spread of these changes within a population is the metric (this is usually put as: "evolutionary success is measured by the number of viable offspring produced by an individual organism"). Most heritable changes will not spread through a population, as they reduce the lifetime reproductive ability of the mutant. Usually this is because in the existing environment, the mutation introduces an extra energy burden upon the mutant -- a change in the genes expressing a protein creates a protein which requires more energy to assemble or use. That is energy not available for reproductive purposes.
However, some mutations -- despite increasing the energy requirements of the mutant -- are favourable because (as examples) they enable an organism to produce a toxin that eliminates competing organisms (including non-mutant members of the population), better survive toxins in the environment, survive in a wider range of temperatures, salinities, illumination, and the like. The favourability of each mutation can only be judge retrospectively by examining the genes in a population some time (often generations) after the mutation is first passed into it.
Some mutations completely overtake the population, such that no (or very nearly no) individual is left without the mutation.
Generations are longer in multicellular organisms, especially ones that reproduce sexually, so the overtake can take a longer time on the calendar, even if it takes only about the same number of generations. In these organisms, mutations arise which make the members of the population with the mutation unable to produce viable offspring with members without. Offspring which do not live to reproductive age, or which cannot produce working germ cells, are non-viable. This is one way in which one species can arise from and replace another. There is also a degree of sex selection in animals, such that mutant and non-mutant individuals can produce viable offspring, but do not do so, because the mutant isn't attractive to the non-mutant, or vice-versa, or both.
The number of factors involved render this random event of almost infinite possibility(the chances of getting the same results again are less than can be represented in today's computing)
The environment is a large constraint upon the success of any given mutation. The environment is usually stable across multiple generations of most species.
Parallel mutations have been observed in disconnected (or even to
Every people has the right of self-determination. The EU is an abrogation of that right.
Every person has the right to self-determination. The EU enshrines that right in, among other things, its collective adherence to the Convention for the Protection of Human Rights and Fundamental Freedoms.
It can only justify itself, in the end, through the annihilation of its members
If by "its members" you mean the 25 member-states, then I think that would be a good idea purely on the grounds of subsidiarity, starting with the most obvious nation-states like Belgium, Spain, the United Kingdom, and Sweden. Article 9 of the Treaty establishing a Constitution for Europe (TCE) explicitly extends the principle of subsidiarity downwards to the regional and local level. Two important criteria are also affirmed therein:
* Decisions should be taken as closely as possible to the citizen * The action should secure greater freedoms for the individual
TCE moreover is another move towards an ever-closer union of individuals rather than member-states. This runs contrary to the goals of some of the governments-of-the-day in a few member-states, and a much larger number of opposition parties, but it broadly reflects the evolutionary trend of the EU over the past 20 years.
In short, if the EU strips member-states' national governments of their powers in favour of a federal system enshrining direct representation, the close-to-the-citizen principle, and the autonomy principle, I'm all for it. That this will anger many nationalists, particularly those who want to impose their vision of their culture and society and their "people" on unwilling fellow citizens, makes me very happy indeed.
Well, obviously he didn't get terribly sick, but he was certainly playing with fire.
Actually, the chemical toxicity of uranium is much higher than the radioactive toxicity, and it would take a lot of the stuff to make you sick
There are other hazards than direct toxicity, chemistry-wise. Hahn started with uranium dioxide samples, some of which he pulverized and nitrated. Uranyl nitrate is fun stuff; if mixed with oxidizable fuels and subjected to shock or heat, it will combust (possibly explosively). UO2(NO3)2 is also much more poisonous than UO2.
His breeder reactor was essentially producing 232Th.O2 -> 233U.O2 at low temperatures (so it was a MOX rather than a ceramic). His uranium dioxide compound would be finely divided by the nature of transmutation and crushing his mail-order samples with a hammer. Small grains of uranium dioxide are pyrophoric, so there was a fire risk in air at room temperature. Water would aggravate this.
Other than the uranyl nitrate, the major toxicity risks to him and others was via inhalation, not ingestion, and inhaling a sizable amount of 232Th.O2 ash wouldn't be too healthy for the lungs, despite being a negligible radiation risk and not very toxic from a chemical reactivity standpoint. Nobody loves pneumoconiosis. Uranium dioxide aerosolizes readily and tends to stick around in the lungs, although studies focus on the probably stickier (and certainly less-soluble) ceramic Depleted Uranium aerosols. If UO2 washes out readily, then it would have been excreted through the kidneys as per your referenced URL.
Radiotoxicity in general depends strongly on isotope (and mass).
With respect to the breeding products, the 233Pa decay (beta + gamma, 22d half life) would be a much bigger radiation risk than the 233U it decays into (1.6e5a half life), but inevitably some 232U would have been produced in his pile, and that nuclide has a decay chain with strong gamma emitters (especially 208Tl). You don't want any of that in you at all.
As to mass, we can only guess at that, although Hahn claimed progressively hot GM counter readings (through concrete even) over the course of several weeks, which isn't hard to believe from a checkerboard fuelling layout.
Finally, his radium and americium samples were also pretty hazardous.
As he adjusted the geometry and composition of his pile trying to maximize GM counter readings, and as his "ignition match" neutron source was uncontrolled, and as he was playing around with neutron moderators, this garden shed project could have become one of the nasty nuclear accidents which injured or killed a person through radioactive contamination. It would have "only" been similar to a radiotherapy patient receiving the wrong dose of gamma radiation, however, or perhaps a fire.
However, in the end, he frightened himself early enough that he was never really in the sort of danger that the Radium Girls suffered, let alone Slotin.
Slashdot Mirror
During that period, the universe expanded by a factor of on the order of 1e26.
What had been causally connected beforehand became causally disconnected as a result.
The expansion became subluminal, leaving (from the perspective of any individual particle) islands of causal connectivity bounded by the particle horizon. The particle horizon is the present largest comoving distance from which light could have reached the observer. Today, here, that's ~78 Mly in all directions.
In GR here is also a related event horizon, which is the largest comoving distance from which a photon can ever reach an observer at some (any!) future time.
The ultimate end of the universe (crunch, asymptotic fade, big rip, false vacuum and other possibilities) has a huge impact on how the particle horizon evolves over large timescales, and whether there *is* a cosmological event horizon for an observer, and what its characteristics are (especially those that can be tested by observational cosmologists).
Guth, Alan H. (1997). The Inflationary Universe. London: Vintage Press (Random House). ISBN 0-09-995950-X.
Although Cosmic Inflation has been evolving in the past almost ten years, the book is absorbing and gives a good overview of not only CI but cosmology in general, outlining the various cosmological problems and ways in which they have tackled. The maths are there, but there are plenty of appendices that help out people with varying exposures to cosmology, astronomy, and particle physics. As a particle physicist you will enjoy the parts that focus on the Higgs field and GUT, I imagine.
The importance of GUT work to cosmologists cannot be overstated, especially for inflationists; the large scale structure of the universe in CI is driven by gravitation as it falls out. The quark-gluon plasma's characteristics as it cools is also essential, and there are tight constraints upon these if cosmic inflation is true.
Uh... well, gravitation and inertia can be firmly in the weird category (well, weirder), but most mainstream cosmology relies upon GR since it has routinely passed every unambiguous test put to it. A refinement of GR might be useful, but an overthrowing of it would raise lots of questions about why we see what we see through telescopes, and why we have been good at predicting what we will see when we look in different ways.
Returning to what triggered this (expansion and inflation), with respect to CI, there are a number of forthcoming experiments (SDSS, Planck Surveyor and some 21cm tomology) that will test CI predictions about the CMBR. So far predictions of the simpler CI models have held up remarkably well, with the key being a strong prediction of an adiabatic, nearly scale-invariant, Gaussian random field with more power at longer wavelengths. The characteristics of the field are critical in order for CI to explain that the large scale structure of the universe are caused by the gravitational collapse of perturbations during the inflationary period that were formed by QM fluctuations. If the field is not what is predicted, CI is false, and this in turn opens the door for "beyond GR" theories to answer the Big Bang cosmological problems.
Uh, yeah, right...
:-|
So log yourself in to a 10.4.8 system with the latest developer kit (xcode-2.4) and try building something with the "-m64" flag that needs to link to e.g. -ltermcap. Oops!
There are several important Frameworks that don't support 64 bit (either architecture).
You also can't mix and match architectures when doing dynamic loading. This is especially annoying since there can be an enormous performance increase associated with a 64-bit data model in some cases. (gmp, for example, tests for and uses -m64 because of this).
Finally, lots of opensource stuff that will build on 10.4.8 has 64-bit-mode (LP64) bugs, and autoconf/libtool both really really suck when it comes to multilibrary/fat/multiarch support.
So, yes, you can build programs that are 64 bit. You can even link to libsystem, some other standard libraries and Frameworks when you do so, and you won't hit syscall "gotchas" unless you try really really hard. However, you are likely to find yourself very restricted in terms of what you can do in 64 bit code, because of components that are NOT 64 bit (yet).
However, what is fairly common (if not quite pervasive) is the generation of 64 bit instructions (and the use of 64 bit doubleword GPRs) for the PPC970 (G5) which can help 32-bit-mode (ILP32) performance, particularly if it uses 64-bit maths (on the long long datatype). As a result, there is often little gain in moving from ILP32 to LP64 on the G5.
x86-64, on the other hand, with the extra registers and instructions, makes LP64 attractive purely for performance reasons, but the same "-m64" vs (default) "-m32" issues apply.
Principle. The "principal" would be Dave Clark.
"Net Neutrality" and the End-to-End principle are essentially orthogonal.
The Clark/Seltzer/Reed argument was that an open-loop congestion-avoiding datagram mode protocol would perform better as a generic bulk transport than any admission-controlled or per-hop-error-corrected one. Except for unusual corner cases, the argument is generallly considered to be right, and is the basis for TCP and other RFC-2001-style congstion-avoiding transports.
The unusual corner cases which are relevant today are where the open control loop is too long (time-wise) for any retransmitted data to be usefully fresh upon arrival at the ultimate destination, and where there are large numbers of receivers.
In the past -- especially in the original End-to-end argument advanced in 1981 -- there was a great deal of argument about the value of offloading processing from the end systems to intermediate systems, consolidating instead of distributing processing. Distribution has better scaling properties in large networks, and full consolidation was not possible in any event, as even in fully circuit-switched networks the end systems perform the same computational tasks (validating data, recovering from corruption, controlling admission/emission) as between any of the intermediate systems along the path.
End-to-end argues that these computational tasks are only needed at the end systems, and not in any of the intermediate systems. While the "only" part is debated from time to time, the Internet has scaled extremely well surviving not only large increases in the number of end systems and total traffic, but also huge and sudden fluctuations in traffic patterns. In the mid-1990s, there was concern about the then dominant traffic pattern, namely very short TCP connections which would never exit slow start (in this case, HTTP instead of FTP). Earlier there was concern about very bursty, transactional TCP connections (NNTP, for example). Now, the usual mode is TCP bulk transfers of about 9MBytes between entirely arbitrary end systems (file sharing and other P2P apps), while the current concern is for applications whose data become stale within a round trip time between source and destination (VOIP, live (rather than recorded) TV), which is inherently bursty, and which has to navigate paths which can be probed at any moment by a long-lived TCP bulk transfer seeking to reach equilibrium (this involves bottleneck bandwidth exploration and will change the time it takes for traffic to exit the bottleneck's queue).
There are a number of areas where TCP is a poor fit. Some of these problems go beyond the Internet however: go to a sports bar with multiple screens, especially one which is receiving multiple transmissions of the same live sporting event, e.g. standard definition broadcast over the air and high definition over digital cable or satellite... compare clocks. Alternatively, try to have a phone conversation with someone in central Kyrgyzstan from your U.S. mobile phone during periods of maximum solar noise and count the number of person-to-person retransmissions and back-offs ("Say again? I didn't hear that because of the static." and "Oops, go ahead and talk.")
A variety of approaches wherein the network service offering is altered to accomodate these situations exist (and many more are proposed). A handful of end-to-end approaches also exist for many of these problems, are deployed, and are working reasonably well (some aren't easy or beautiful though). Finally, there are hybrid approaches wherein end-to-end RFC-2001-like protocols are carried over links which deal with local congestion and corruption without the participation of the end systems.
End-to-end isn't really concerned about homogeneity of the intermediate systems as much as it is about distributing congestion avoidance and recovery from data corruption to the end
You might find this table of voter turnouts in the past couple of EP elections interesting.
Nothing in the news article, or in the commentary written on the WoW board is suggestive of impropriety on the part of any of the police or CBSA officers.
I am a little sceptical of people who claim to have an argument with officials in any situation remotely as stressful as this one, but even if we accept the dialogue as accurate reporting rather than coloured by artistic licence, the worst accusation that can be levelled is that the police (likely Ottawa-Carleton regulars, or duty Mounties) asked an inappropriate question ("if you were drunk..."), which would best be followed up with his or her boss, in writing.
The Charter regulates the activity of everyone involved in the government of Canada (s.32(1)(a)) except in unusual circumstances backed by statute that survives court scrutiny. The Federal Court (which enforces the Charter per s.25) generally anticipates the Supreme Court of Canada's take on these exceptions, which I think you describe reasonably.
The (federal) loopholes in the Charter in general are only interesting with respect to legislation passed by Parliament, and this includes the gigantic one in section 1, which you describe. (It's s.1 of the Canadian Charter of Rights and Freedoms (citation as per s.34), not the Constitution Act, 1982 (Schedule B, Part I), which it also is, and isn't, simultaneously. The Constitution is fickle about citations mainly because much of it originated as Acts of the UK Parliament, including the Canada Act, 1982 (UK). Ugh.).
The most relevant sections of the Charter have universal applicability ("Everyone has the right..." ss.7-10,12) and unlimited situational applicability ("Any person charged" s.11; "A witness..." s.13; "A party or witness..." s.14) and importantly s.15(1) "Every individual is equal before and under the law and has the right to equal protection and equal benefit of the law without discrimination...". This applies to non-Canadians and Canadians equally, and it also applies outside Canada, with respect to Canadian officials performing government activities.
When you are in Canada physically, you are in Canada legally. "Sterile" areas and flights are still bound by Canadian law.
There is a singular lack of obvious grounds of action (or complaint, except for rudeness) against any Canadian official in the story recounted by the WoW poster.
The general policy of "land, isolate, interview, interview again" makes overreaction likely, however, and sadly the policy is the responsibility of Stockwell Day, the minister for Public Safety (Mon-Sat only).
Let us hope for everyone else's sake that Tony Blair (authoritarian, self-righteous, strong supporter of the Bush Administration, and surrounded by people who encourage public hysteria) is thrown out of office sooner rather than later. Whether he exits "gracefully", is turfed out by his own party, or is defeated at an election triggered by yet another lost vote in the House of Commons matters little. Whatever's fastest, please. (And hope he takes John Reid with him).
Well you could always try complaining to Stockwell Day, the Minister responsible. Just don't complain on a Sunday. He doesn't work on Sundays.
Everyone. Not "everyone in Canada", not "every Canadian". The universality of the Charter was chosen to be clear and unquestionable, and it repeats itself throughout the sections headed with "... Rights".
Note that I left out section 6: "6. (1) Every citizen of Canada has the right to enter, remain in and leave Canada."
Immigration officials with the Canadian Border Service Agency may exclude non-Canadians from Canada under statute law and regulation in accordance with the Charter.) No statute, regulation, or administrative act by the government is legal if it contravenes the universal rights to due process guaranteed everyone by the Charter. It is not legal to deprive non-Canadians of their rights, and s.24 remedies are regularly meted out by the courts in such cases.
There is a detailed set of administrative processes and judicial reviews involved when excluding or removing people from Canada. (Here is an overview).
Note that even persons constrained by statute law from appealing such an order may also resort to the Federal Court in two main (and very different) ways. However, CBSA/CIC will always be very careful about exclusion and removal order processes.
CBSA agents may also make arrests. There is an overview here. Note that even Canadian citizens may be arrested and detained.
Arrests and detention are serious in Canada, and the arresting officer and all his superiors are open to a variety of civil suits, section 24 remedies, and even criminal charges, if they do not follow the appropriate processes and are not especially conscious of the Charter. A number of prominent lawyers offer pro bono services in many such cases.
Since Canada is country with many immigrants and international travellers, especially concentrated in major cities, a variety of media organizations tend to be interested in the activities of the CBSA, especially given who the current Public Safety minister http://en.wikipedia.org/wiki/Stockwell_Day >idiot happens to be.
Canada is certianly not perfect, and the CBSA can be unfriendly especially at busy times, however "don't fool around" should not be confused with "you are liable to be shot", and "I want to get this over with too" is not the same as "I can abuse you if you don't cooperate".
As a result, ISPs can reduce some of their operating costs and thus lower their prices, if they buy their long-haul/international connectivity on a $/peak Mbps/month basis (which is common). Bulldog is one of a handful of UK ISPs who operate a substantial international network on their own, and whose costs are not directly influenced by traffic loads, and who can therefore offer cheaper and less-capped services to atypically heavy users.
It has nothing to do with libtards. This is simply how the market has developed. Many customers are price-sensitive; there is sufficient competition that many suppliers are cost-sensitive (margins are important!); many suppliers' costs are directly influenced by the 95th percentile 5-minute-average traffic loads delivered from their international carrier(s), because the wholesale/long-haul market has developed that way over the years. Caps and limits control those costs.
However, different suppliers have dramatically different cost structures, and so there are a range of offerings that one can pick and choose from. Including unlimited download multimegabit per second broadband for prices competitive with what you are paying.
Here's a reasonably well-maintained price list that illustrates much of the variety of packages available in the market, including a variety of usage caps and no-cap services. It was easy to find a standard offering directly comparable to yours (you claim a much higher upload rate than is listed here, however this is not a firm cap rather than an estimate of the expected statmux loss on cable plant, which differs from asymmetrical DSL).
However, again, the UK market is expensive compared to other national markets in the European Union. The incumbent was very good at playing politics and ended up with an early set of changes in its regulatory environment, and a series of tame regulators. This has changed recently, and market price erosion has been substantial. Hopefully people will be able to tell you about cheaper and cheaper prices still over the course of the next months, since otherwise the artificial market approach would be failing.
Sweden was the first country in the EU to begin telecomms deregulation, in 1981 (mobile telephony), 1986 (cable television), 1991 (Internet connectivity and long haul leased lines) and 1997 (full service voice telephony).
Sweden has by far the most well-developed last-mile services. In the three major cities (Stockholm, Gothenburg, Malmö) it is common to have a fast ethernet drop from one's apartment to a service cupboard with a switch that has 8 other fast ethernet ports (for your immediate neighbours) and a gigabit ethernet or other high-speed optical connection (IEEE 802.17 is popular, as is Packet-over-SDH) to an ISP. In minor cities there is still vigorous competition for high single-digit Mbps connections, and double-digit Mbps ones are usually available. Even in most of Sweden's rural areas (which are very sparsely populuated), Mbps or near-Mbps connections can be delivered reasonably affordably.
Stockholm was a major focal point for Internet connectivity in Europe throughout the 1990s -- a commercial ISP and a consortium of the Nordic countries' national research networks shared what were always the fastest connections across the Atlantic ocean: multiples of 2Mbps in the early 1990s, the first ever trans-atlantic 34Mbps terrestrial connection, the first ever trans-atlantic 155Mbps connection, and so forth, increasing every two years or so. Stockholm was the focus in large part because the regulatory environment encouraged price erosion on the Swedish half-circuit. At the time, European half-circuits were often 2-50x (yes, really) more expensive than the cost of the U.S. end of an international leased line, while pricing in Sweden could be cheaper than in the U.S. in some cases.
Sweden's lead on costs, and Stockholm as a hub of business and technology, encouraged many foreign telecomms companies to begin offering services in the city. As a result, there were no fewer than four competing companies undertaking city-wide fibre deployments, which meant major streets being restricted by a lane or two here and there on a regular basis, worsening the city's traffic congestion. The city itself solved this problem neatly by forming a city agency (Stokab) which laid down fibre and ducting in the city as a matter of course, whenever other works (water and sewer and electricity and gas line repairs) would be digging up a road anyway. Moreover, Stokab offered fibre between any two locations in the city on a cost-recovery basis to anyone including the incumbent telecomms and cable companies. This was a huge success for both telecommunications development (ISPs routinely offering fibre-to-the-basement of small apartment buildings, gigabit inter-office connections, etc.). The Stokab approach has since spread a bit beyond Stockholm.
The other three continental Nordic countries caught on eventually, Finland leading the rest of the EU, and Denmark and Norway (the latter is not in the EU) lagging behind somewhat.
Yes. You are exactly right. Adding new competitors is the key. That is exactly what deregulation is trying to do in Europe.
In the mid-1990s, the European Commission (notably then-DGIV (now the Competition Directorate) and then-DGXIII (now the Information Society and Media Directorate)) began developing the deregulation of telecommunications services in the European Union.
Rather than the U.S.-style policy of focusing on immediate benefit to consumer (notably falling prices or offering of new services), the approach of the Commission was to establish artificial markets in which barriers to e
There are lots of bacteria happy to colonize meat. Most of the ones that cause human (or pet) diseases are readily controlled by careful packaging, storage, preparation/handling and cooking. Cook your food thoroughly and wash your hands carefully before eating, and you likely will never run into food poisoning by any of them.
Listeria monocytogenes is an exceptionally hardy bacterium for one that does not form spores. It will grow in packed refrigerated meats, for example. It will survive freezing, and can survive cooking. (It's also very tolerant of dessication.) It readily causes gastrointestinal troubles for healthy people, and can cause serious symptoms. Rarely, it will prove fatal.
You are right that L. mono. has plenty of opportunities to adapt to environmental pressures. Viral infection is one of these -- "that which L. mono. survives, makes it stronger". However, bacteriophages co-evolve with their target bacteria. Their replication and assembly mechanisms are error-prone, and typically billions of copies are made before the infected cell lyses. This leads to even greater opportunties for genetic changes than the bacteria have.
Most evolved mechanisms of resistance are energy intensive, so resistant organisms typically don't overwhelm unresistant organisms in the wild.
There is, however, a possibility of a locally resistant strain arising in e.g. a meat-packing plant making routine use of the bacteriophage spray.
This will typically be noticed through inspection. The approach to this situation is straightforward: find a virus that infects the new strain, and breed it (by letting the infection run its course through a culture of the new strain). "Finding" may involve simply looking around the meat packing plant, or more simply (and perhaps just as effectively), turning to ponds and soils wherein L. mono. is naturally found. Scooping up a bunch of the surrounding scum is liable to net some viruses which can infect the new strain.
Phage treatments remain active for a longer duration than a flash irradiation, and do not change the nature of the food medium they are applied to. Gamma irradiation denatures food material as well as bacterial material.
Gamma irradiation does not worry me, but some people appreciate more conservative approaches to pathogen control.
Finally, given the differences in difficulty of preparation and delivery of the anti-listeria agents, phage treatments are almost certainly much less expensive than using ionizing radiation, especially factoring in the various regulatory realms involved.
As other people have noted, viruses also adapt to selection pressures (and quickly, since the failure rate and scale of virion manufacture is large), so viruses are likely to catch up with any but the most dramatically runaway acquired immunity in its target population.
Introducing lots of listeriophages into the environment might make Listeria monocytogenes less common, or press it into virtual extinction? Oh noes! What a horror. We shouldn't make listeriosis (at least locally) extinct, because the mostly domesticated 37ish mammalian species (and 15ish birds) it infects, including humans, might undergo a population explosion?
I don't really see the downside other than fewer sick animals.
Unfortunately, the propagation mechanism is very local -- sprayed phages will get carried around by droplets and as dust. There are listeria reservoirs
in places far away and well isolated from western meat packing plants. Unlike the bacteria, the phages cannot replicate themselves -- they can only replicate in the presence of susceptible bacteria. If they're ingested by grazing animals or flying birds or the like, they probably will not survive passage through the digestive system, or attack by the animal's own immune system. Thus transportation to an area with as-yet-uninfected bacteria is unlikely. L. mono. is very hardy for a nonsporulating bacterium (that's why it's dangerous) is much more hardy than the several viruses that prey upon it.
Much less than the chance today of listeriosis acquired by refrigerated (or even frozen) and sealed packed meats that are well protected against e.g. E. coli and a range of other non-sporulating food-colonizing bacteria.
Leaving meat exposed outside of a fridge for 24h will also breed a range of moulds and bacteria, notably salmonella.
Bacteriophages aren't instant, so it's more useful to spray a solution on the meat at a packing plant, than on unwrapped meats.
(So it's still not a good idea to spray on an anti-listeria preparation and re-freeze or refrigerate unwrapped and left-out meats and the like.)
With listeria, it's not so simple: the bacterium also survives considerable heating (sometimes even cooking) and dessication, as well as being able to reproduce in fridge temperature ranges, and survive deep freezing.
Careful storage, cooking and preserving, in other words, can fail to eliminate listeria. Listeriosis produces a range of symptoms, some serious, and while fatalities are rare, you really wouldn't enjoy the gastrointestinal upset it produces in mild cases in an otherwise-healthy individual.
Other people and comments have addressed the co-evolution of phage and bacterium, compared with antibiotic resistance. However, bear in mind that methicillin (the M in MRSA) is
Yoghurt isn't animals (the active ingredients are bacteria in the Lactobacillus genus, mainly L. acidophilus and L. bifidus.
These bacteria, like all others, are unicellular organisms, and so lack bowels (teeming with lots of cells of all sorts within and surrounded by lots of other cells belonging to the animal itself) and blood (teeming with lots of cells belonging to the animal).
On the other hand, there are cognates... in Lactobacilli there are a number of vesicles and nutrient storage structures that might do as a sort of bowel kinda, and there is of course a cytosol, which is sorta like the non-cellular part of animal blood.
Sadly, it is not the crushing (with your teeth, presumably) which will let you spill out their bowels (chewing hard on yoghurt likely won't harm a single bacterium out of the billions in each spoonful), but rather acids and enzymes found much further down your digestive system. Most of them, in fact, will be found in your bowels (and descending colon), and won't really be produced by you per se, but rather by the other microorganisms dwelling there.
L. acidophilus actually mostly helps you spill open much less friendly bacteria so that you can feast upon their "bowels" and "blood".
Eat more yoghurt! They are your proxy teeth!
It's not exceptionally difficult to implement.
The phage culture would be added to whatever washing fluid is used to remove fecal matter and other contaminants during the butchering process(es).
The bacteriophage viruses are easy to grow at any lab which can culture the relevant bacteria. The bacteria is the phage "food".
So all that's necessary is shipping out some purified phage solution (or even as a powder) to meat packing plants every so often.
This is an inexpensive way of preventing a number of diseases associated with handling or eating uncooked or improperly cooked meat. Some of these diseases are extremely unpleasant for the victim, even if the number of fatalities a year is low.
(It can also be applied to other foodstuffs which are subject to common infections by microorganisms for whom there is a natural "predator" phage virus).
Bacteriophage viruses also mutate, and usually at a faster rate than the bacteria they attack (this is mostly because duplication of viral RNA sequences and assembly of virion particles are especially error prone processes, and many many copies are made before the bacteria is lysed). Bacteriophage therapy thus has the handy property that any resistance acquired by the target bacterium is almost certain to be met by an equivalent viral adaptation, all occuring within the bacterial growth medium, whether that's a slab of beef or a human patient.
Any mutation in the host bacterium, moreover, is likely to have been seen in its natural environmental reservoirs and thus there is likely already an equivalent reservoir of already-adapted phage.
Procedure: extract phage particles from pond scum (literally, in many cases). Add to culture of resistant bacterium. When the culture is clearly succumbing to a viral infection, extract and purify the phage particles. Inject into patient with resistant bacterium. Repeat as necessary.
The main downside to this approach to acquired resistance to phage therapy is that culturing the resistant strain takes time. Meanwhile, the resistant strain is infecting the patient. If the infection is especially aggressive, or the patient is especially weakened by it, the infection will outrun the preparation of the new viruses. (In that case, use wider-spectrum agents like western antibiotics, hyperbaric oxygenation, debridement, amputation, etc. as necessary).
Using phage therapy in cattle (or on dead beef as a surface treatment) against known strains of known pathogenic (to humans) bacteria is sensible: firstly, you reduce the incidence of sick humans who handle uncooked meat, who eat improperly cooked meat, or who eat cooked meat in which bacteriotoxins have survived; and secondly, you do not breed antibiotic-resistant bacteria on the surface of the cuts of beef.
Treating the surface and near subsufrace of dead meat also has the positive advantage that the dead animal no longer has an immune system to interfere with the phage's attack on its target bacteria (this is also unfortunately why bacterial contamination of untreated dead meat is so common).
10000*1000 > 1000000 (by a factor of 10)
But the question is really whether your data can be decorrelated and correlated (or scattered across the wavelengths and gathered up on the receive side) across that many channels.
The important question in packet-based networking is when the last bit of a packet arrives, not when the first one does.
Let's make it 10 000 channels of 1000 modulations per second versus one single channel of 10 000 000 modulations per second, so the aggregate rate for the serial versus parallel link is identical (10 mbps).
If all your L3 traffic is expressable as 10 000 bits per conceptual L2 frame, then in both cases you get 1000 pps; one thousandth of a packet per "blink" on the serial link, one packet per "blink" on the parallel one.
If your L3 traffic is less than 10000 bits (and the link is not a bottleneck) then it takes one blink to send the L3 traffic across the parallel link no matter what the L3 packet size is. This means that your minimum link latency is one millisecond on the parallel link. The serial link's latency is a function of packet size; smaller packets take less time for the last bit to arrive; you save a microsecond per bit.
For packets that are larger, you get into situations where you have, for instance, 10001 bits to send across an otherwise empty link. On the serial channel, this takes you 0.1 microseconds more than a frame of 10000 bits. On the parallel channel, 10001 bits takes a whole millisecond (10 000 microseconds) longer.
This is certainly an effect which can be discovered end-to-end (akin to MTU exploration), and could be adapted to by the transmitter, however there are probably no in-use transfer protocols which deliberately look for step function increases in packet size to delay.
Moreover, when the link is a bottleneck (which is expected at every link across which TCP bulk transfers happen) and variable sized packets, scheduling becomes a problem, if you want to avoid the step-function from being a serious source of jitter. If you have multiple packet flows occupying the queue, you inevitably end up with some packets which will suffer from millisecond jumps in delay for the final bit at the receiver, compared to the serial link. While this is not an intractable problem, minimizing the incidence of partial-packet delay across streams of variable-sized packets is a difficult one, particularly as you scale from megabits/second aggregates to gigabits/second aggregates (one 10GHz channel vs ten thousand megabit/second channels), and as you increase the number (also the variability of number) and end-to-end throughput (also the burstiness) of independent flows.
This will be an important problem for Internet Engineering in the future, as there are difficult price challenges in 40Gbps/optical channel now, and difficult price and physics challenges in anything faster than that. Slashdot readers familiar with megahertz versus cores will probably notice an echo.
Also of interest (and stuck in the realm of routing and control plane research): with your serial channel, a failure anywhere in the path disables the whole 10 Mbps channel. With the parallel channel, you introduce a whole range of partial failures that wipe out one (or a handful) of parallel colours. It may be better for a network's users to continue to use even 1/2 of the 10 000 parallel channels (depending on things like traffic load, how a failure of the whole channel is dealt with at various restoration or routing levels, and so on).
Pseudo-parallel connections are even worse; there are a variety of situations in which the receive side detects simultaneously transmitted pulses of two frequencies at different times...
In short, massively parallel links in theory should be no worse than equivalent serial (or even "lightly" parallel) ones when it comes to goodput; however, we don't know how to do this in practice yet.
So, sadly, for real world Internet traffic, I would not be surprised if a 1 000 000 blink/second serial channel performs better than 10 000 parallel channels of 1000 blink/second each.
This is largely because the members of the House of Representatives have more leeway in, and are much better at, choosing their electors. Senators are elected at the state level, and changing the borders between states is not as easy as changing the borders between the smaller electoral districts within them.
Senators face a state worth of electors every six years; members of the House of Representatives have two years to swap pockets of voters who would vote for another candidate in their district for pockets of voters more inclined to vote for him or her, but who are living in a nearby district controlled by a member of his or her own party who wins by larger majorities, or by a member of the opposite party thus improving the electoral chances of both incumbent members of Congress.
Probabilities are very low, and most mutations are destructive.
The most significant multi-mutation events are found in sexually reproducing organisms during meiosis and fertilization.
The expected result when there is this sort of error is a dead gamete or a dead zygote.
This is very common... malformed or non-viable seeds in plants; malformed or dead eggs in ovipositing animals; miscarriage in viviparous ones.
In humans, for example, early miscarriage has come under quite a bit of study because it turns out to be very common -- the embryo dies and is expelled without the woman ever knowing she was pregnant. A British Medical Journal study is consistent with reports of 11-16% miscarriage rates in known pregnancies. Some early miscarriages are largely asymtomatic, and there are studies underway in women likely to become pregnant that expect a doubling of the known rate.
In [Hogge, W.A. The Clinical Use of Karyotyping Spontaneous Abortions. American Journal of Obstetrics and Gynecology, volume 189, number 2, August 2003, pp 397-402] 70% of studied first trimester miscarriages were caused by chromosomal abnormalities, with the bulk attributable to defects in the sperm or egg cells involved.
So in humans, some 5-10% of pregnancies end because of a cluster of unfavourable germline mutations.
That's around 800 000 terminal mutations per year, globally, in humans alone.
Humans gestate over a period of nine months, and presently are producing less than three offspring per woman. There are species with comparable miscarriage rates who gestate much more quickly and produce many more offspring, and are also more populous than humans, and which consequently get to "test out" a larger variety of random multi-mutation combinations over the course of ten years than humans.
So yes, the probabilities are very low, but the number of tries is very high, particularly over the course of millions of years.
What we know of nature is that it isn't a perfect environment for life, it's actually pretty harsh and filled with serious hazards. However, life has adapted successfully to the harshness, when considering all life, rather than individual species (many of which have become extinct).
"Successfully" means living things are still producing viable offspring right now.
A "rewind and play" of the environment with the same biomass at the time of the first photosynthesizing autotrophs, likely would lead to different species than we have now, but there would be environmental niches readily occupied by land plants, aquatic heterotrophs, swimmers, aquatic predators, land grazers, taller hardier plants, land predators, and so forth.
On this scale, probabilities are so chaotic, the question of whether oak trees, carp, thylacines, or human beings might arise is a philosophical one, along the lines of the cosmological anthropic principle.
No, heritable changes simply need to be non-harmful, as per your point (1).
If the changes are propagated to offspring which in turn propagate the changes, the change is likely to remain in the population, as long as the changed individuals reproduce at the same rate as the unchanged ones.
In other words, with respect to the production of viable descendants, most random changes are likely to reduce the number of offspring per changed organism (sometimes the reduction is down to zero). Of the remainder of chainges, the majority is likely to make no difference to individual reproductive rate as long as the environment remains much as it is. A tiny minority of changes will provide some advantage over unchanged indivuals in the present environment.
Environmental changes can cause new traits which are neutral to become positive or negative, or to make positive changes neutral or negative, or to make negative changes neutral or positive.
Mutations are simply random. An evaluation of their "constructiveness" (using your term) can only be done retrospectively by counting the prevalence within a population of the inherited mutant trait in a given generation.
What is "constructive" today may be burdensome tomorrow. That is several generations from today in many microbial species.
Simultaneity of mutations does happen sometimes -- DNA replication processes are error-prone, and the error rate varies from species to species. There are a number of types of errors that can cause several simultaneous mutations. Again, the success of these can only be evaluated in hindsight, as with point mutations.
However, successful mutations more likely are spread out over time. Highly developed eyes are good examples.
The cephalopod eye (in octopuses and the like) started as a photosensitive pigment spot on the skin surface of an ancestor. Shine light on it, and it produces and releases hormones, much like shining different frequencies of UV light on a melanocyte in your skin produces and releases melanin. These hormones excite nearby nerve cells. Mutations which tighten the binding (increasing the speed at which changes in light are propagated) between the pigment-containing cells and the nerves would help in avoiding predators or finding prey, for example. In cephalopod ancestors, a gene would influence whether the pigment cells were flat against the surface of the skin, protruding above it like a mole, or sunken below it in a pit. Each of these has advantages -- protrusion enables a wider field of view, invagination enables a detection of directionality. Invagination won. A further set of genes influenced whether the pocket the pigment cells rested in widened like a bowl, or narrowed like a bottleneck. Narrowing won, and as a result a camera obscura eye was the result. A subsequent mutation allowed camera obscura squid ancestors to produce crystallins through a mutation of one of the several genes which produced enzymes. Crystallin production around the opening of the camera obscura eye evolved into a useful lens.
Vertebrate eyes evolved differently: an early ancestor (an early form of sea squirt) produced a crystallin (with some enzymatic uses) within its central nervous system; it also had a primitive light detection system based on pigmented cells also located within its central nervous system. The closer the pigmented cells were to the outer surface of the organism, the better at detecting predators and the like. This was enabled by a series of mutations producing an ever longer chain of nerve cells connecting the pigmented cells to the CNS. Later descendants gained some sight advantages by producing crystallins near the photodetecting cells, as well as by producing different pigments. Eventually a pre-vertebrate ancestor's eye evaginated thanks to a genetic change, with the lens protrudi
Let's tackle this one, since it explains a lot of the "we don't know" answers you'd get to many of the other good questions on your list that evolutionary biologists, paleontologists and so forth are also very keen on answering.
The answer is: they died out, and may be in the fossil record.
The problem is that fossilization is extremely uncommon.
Most dead organisms and their byproducts decompose rather than fossilize. Many organisms that die in conditions where fossilization may occur are unsuitable for fossilization. Inclusion fossils, like insects, plants microbes trapped in fossil resin (amber), are unsuitable for large animals because they are simply too large to be covered in decomposition-resisting resin, or which will largely decompose anyway. Permineralized fossils are extremely useful, but the organism has to be completely covered in suitable sediment very soon after death. This is rare. Partial coverage leads to no fossil, as decomposing agents would ultimately reach the whole organism. Similarly, mould and cast fossils can be useful even just for the impressions within them, but the conditions for their formation are also rare.
Also, different tissues react to fossilization processes in different ways. Harder, more stable tissues, especially polymers, are more likely to be preserved. Softer tissues are much more rarely preserved. It is more common to find a fossil which is simply chitinous plates, bones, or feathers than to find any preservation including flesh or connective tissue.
Different organisms also live and lived in different areas. A large number of organisms may never have been in an area where fossilization was possible. A large number of organisms' fossils are in probably in inaccessible places (under modern-day ice caps, in modern-day deep water, under lots of desert sand).
Transitional species which did not die in the right places (currently known fossil fields) at the right times (when fossilization was possible) and under the right circumstances, with the right "body" parts, in the right microbial company (mostly obligate aerobes and non-decomposing anaerobes), and in places where people have thought to look and have the budget to dig in, are obviously unlikely to be discovered.
Even among those whose fossils are plausibly "missing", there will be transitional species which had short-lived or small populations (or both). Fossils do not give much detail about internal morphology or fine differences in biochemistry. So there will also be transitional and intermediate species which are indistinguishable via examinations of fossils from their bracketing forms.
Gaps certainly exist in the fossil record. This means data useful for answering many of your questions are not available. Other data may be found, and they should concur with predictions about the "missing" fossils.
That said, there are plenty of predicted transitional and intermediate species which have been found in the fossil record which corroborate and illustrate macroevolution. Horses and whales are well known families with plenty of fossil evidence of adaptations to known changes in their environment, and there are a number of significant fossils that illustrate major changes in vertebrates which are consistent with common descent. There are fossils which are consistent with the current theory of reptile->bird evolution, and fortunately feathers and scales and things which look like fluffy and striated scales permineralize well, as do bones, leaving fossils of great detail. Recent finds (reported on slashdot even!) include gliding reptiles and non-flying bird-like animals that most likely used their not-quite-wings for stability while running (like ostriches).
A number of your other questions cannot be answered in the fossil record, but there are theories which predict fossils which are likely to be discovered. These theories are becoming in
D'oh! What you've described is Darwinian, the clue being "environmental stressor", which is synonymous with natural selection. More specifically, you're describing the modern evolutionary synthesis.
Selection pressures result in some genes passing into a population. You've described exactly that, and roughly produced a biological mechanism that is part of the process.
Unfortunately you use "animal" rather than "organism" (plants, bacteria and the rest evolve too!) and are too focused on protein folding, whereas differences in whether or not a protein is expressed are much more important in e.g. the evolution of bacterial pathogens. For example, with respect to resistance to the penicillin family of antibiotics, one gene expresses beta-lactamase, another produces penicillinase. Express the gene, gain resistance. Don't express the gene, get wiped out by common antibiotics. To be fair, a third (and less common) mode of resistance produces a variant penicillin binding protein with a reduced affinity for beta-lactam rings, and the chemical variation in the PBP will inevitably produce a protein with a different geometry. The variant PBP bacteria rapidly fall out of populations not actively stressed by beta-lactam antibiotics, as the variant PBPs dramatically increases the energy cost to the bacteriam of building its cell wall.
Chaperone proteins? These are mostly expressed in prokaryotes as a heat shock defence and in eukaryotes to a lesser extent in endosymbionts (mitochondria et al), endoplastic reticuli and the cytosol, where they assist in transport and in protein degradation. Removing chaperonins in prokaryotes leads to an organism easily cooked to death at relatively low temperatures. Turning them off in eukaryotes exposes the cell to protein aggregation diseases, or kills the mitochondria (and thus the cell). The most obvious way of turning chaperonins off is to deprive them of ATP, which is another good way of killing a cell.
Known chaperone proteins are described by genes which are governed via Mendelian inheritance.
There are people doing research in the area of conformational changes as an amplifier or attenuator of expression of phenotype. There are also people doing research in the area of non-genetic heritable traits, just as there were people studying the trait inheritance mechanism in endosymbionts. To argue that they have supplanted genetic heritability is a stretch. These broad groups may be uncovering a new, parallel trait propagation mechanism, akin to endosymbiont DNA versus nuclear DNA. However, most of the work seems focused on rapid expression of genotype mutations in conditions of environmental stress, and while terms like "Lamarckism" get thrown around, that does not seem to be what these researchers are reporting. To suggest that this sort of work is trying to "kill" natural selection just seems wrong.
Let's start with differences in ploidy. It happens often in some plants, it's readily observable, and it's an important factor in plant speciation (natural and through breeding).
Wheat is straightforward: there are wheat species that are diploid (Einkorn wheat), tetraploid (durum), and hexaploid (bread wheat). The last evolved in farm fields, and the middle evolved in the wild and is the result of a non-human-influenced hybridization of two diploid wild grasses.
Apple species, tulip species and lilly species vary in their ploidy as well.
With respect to changes in chromosomal number other than changes in ploidy, during meiosis, homologous pairs can fail to segregate properly (non-disjunction), leading to monosomy (where one of the chromosomal pair is missing in a normally diploid organism) or trisomy (where there is an additional chromosome attached to the chromosomal pair, again in a normally diploid organism). These are not exceptionally important factors in speciation, as aneuploidy rarely results in reproductive advantage for the organism affected (and often is disadvantageous).
A pair of individuals with heritable monosomy may produce viable offspring that are missing a full chromosomal pair (or euploid "set" in a normally non-diploid organism). This is more likely in organisms with many or very small chromosomes, or in amphiploid organisms which are cytochemically of lower ploidy than they are in terms of reproduction (e.g. cotton, which has four sets of chromosomes, but behaves like a strict diploid).
So heritable change in the number of chromosomes is observed in the wild and in the lab (or field), and while this usually creates significant phenotypal differences when it happens, the offspring between pairs of similarly mutated organisms is -- occasionally -- viable.
Multiply "occasionally" by many generations and the result is that the possibility of two species with a common ancestor may have different numbers of chromosomes is entirely plausible.
Finally, mutations within a chromosome are simply much less dramatic (in terms of being likely to help or harm (or neither) an organism) and are more likely to be caused by environmental factors than mutations involving the addition or deletion of entire chromosomes, and so the former will be observed in a natural population much more often than the latter.
Mutation is largely random because DNA and especially RNA are structurally frail and break when exposed to e.g. ionizing radiation, or oxidizers and other mutagens prevalent in the natural environment here on Earth. All cellular organisms have mechanisms which protect against and repair damage to their in-cell DNA. Some of these are passive structural protections, others are active molecules, particularly repair polymerases. Repair polymerases are imperfect (some are highly error-prone), and will not be able to repair all externally-driven changes to genetic material.
Unrepaired changes which are passed down to a cell's descendants, as well as replication errors introduced in the process of spawning descendants, are relevant to evolution; those which prevent a cell from producing descendants in the first place are not. In multicellular organisms, the multicellular descendants are considered instead of those of constituent cells not directly related to producing the multicellular offspring. In such organisms, the mutations are split between heritable and non-heritable mutations. Usually these are described as somatic and germline mutations, however some somatic changes are heritable in some multicellular organisms.
Heritable changes are the mechanism of evolution, and the spread of these changes within a population is the metric (this is usually put as: "evolutionary success is measured by the number of viable offspring produced by an individual organism"). Most heritable changes will not spread through a population, as they reduce the lifetime reproductive ability of the mutant. Usually this is because in the existing environment, the mutation introduces an extra energy burden upon the mutant -- a change in the genes expressing a protein creates a protein which requires more energy to assemble or use. That is energy not available for reproductive purposes.
However, some mutations -- despite increasing the energy requirements of the mutant -- are favourable because (as examples) they enable an organism to produce a toxin that eliminates competing organisms (including non-mutant members of the population), better survive toxins in the environment, survive in a wider range of temperatures, salinities, illumination, and the like. The favourability of each mutation can only be judge retrospectively by examining the genes in a population some time (often generations) after the mutation is first passed into it.
Some mutations completely overtake the population, such that no (or very nearly no) individual is left without the mutation.
Generations are longer in multicellular organisms, especially ones that reproduce sexually, so the overtake can take a longer time on the calendar, even if it takes only about the same number of generations. In these organisms, mutations arise which make the members of the population with the mutation unable to produce viable offspring with members without. Offspring which do not live to reproductive age, or which cannot produce working germ cells, are non-viable. This is one way in which one species can arise from and replace another. There is also a degree of sex selection in animals, such that mutant and non-mutant individuals can produce viable offspring, but do not do so, because the mutant isn't attractive to the non-mutant, or vice-versa, or both.
The environment is a large constraint upon the success of any given mutation. The environment is usually stable across multiple generations of most species.
Parallel mutations have been observed in disconnected (or even to
Every person has the right to self-determination. The EU enshrines that right in, among other things, its collective adherence to the Convention for the Protection of Human Rights and Fundamental Freedoms.
If by "its members" you mean the 25 member-states, then I think that would be a good idea purely on the grounds of subsidiarity, starting with the most obvious nation-states like Belgium, Spain, the United Kingdom, and Sweden. Article 9 of the Treaty establishing a Constitution for Europe (TCE) explicitly extends the principle of subsidiarity downwards to the regional and local level. Two important criteria are also affirmed therein:
* Decisions should be taken as closely as possible to the citizen
* The action should secure greater freedoms for the individual
TCE moreover is another move towards an ever-closer union of individuals rather than member-states. This runs contrary to the goals of some of the governments-of-the-day in a few member-states, and a much larger number of opposition parties, but it broadly reflects the evolutionary trend of the EU over the past 20 years.
In short, if the EU strips member-states' national governments of their powers in favour of a federal system enshrining direct representation, the close-to-the-citizen principle, and the autonomy principle, I'm all for it. That this will anger many nationalists, particularly those who want to impose their vision of their culture and society and their "people" on unwilling fellow citizens, makes me very happy indeed.
There are other hazards than direct toxicity, chemistry-wise. Hahn started with uranium dioxide samples, some of which he pulverized and nitrated. Uranyl nitrate is fun stuff; if mixed with oxidizable fuels and subjected to shock or heat, it will combust (possibly explosively). UO2(NO3)2 is also much more poisonous than UO2.
His breeder reactor was essentially producing 232Th.O2 -> 233U.O2 at low temperatures (so it was a MOX rather than a ceramic). His uranium dioxide compound would be finely divided by the nature of transmutation and crushing his mail-order samples with a hammer. Small grains of uranium dioxide are pyrophoric, so there was a fire risk in air at room temperature. Water would aggravate this.
Other than the uranyl nitrate, the major toxicity risks to him and others was via inhalation, not ingestion, and inhaling a sizable amount of 232Th.O2 ash wouldn't be too healthy for the lungs, despite being a negligible radiation risk and not very toxic from a chemical reactivity standpoint. Nobody loves pneumoconiosis. Uranium dioxide aerosolizes readily and tends to stick around in the lungs, although studies focus on the probably stickier (and certainly less-soluble) ceramic Depleted Uranium aerosols. If UO2 washes out readily, then it would have been excreted through the kidneys as per your referenced URL.
Radiotoxicity in general depends strongly on isotope (and mass).
With respect to the breeding products, the 233Pa decay (beta + gamma, 22d half life) would be a much bigger radiation risk than the 233U it decays into (1.6e5a half life), but inevitably some 232U would have been produced in his pile, and that nuclide has a decay chain with strong gamma emitters (especially 208Tl). You don't want any of that in you at all.
As to mass, we can only guess at that, although Hahn claimed progressively hot GM counter readings (through concrete even) over the course of several weeks, which isn't hard to believe from a checkerboard fuelling layout.
Finally, his radium and americium samples were also pretty hazardous.
As he adjusted the geometry and composition of his pile trying to maximize GM counter readings, and as his "ignition match" neutron source was uncontrolled, and as he was playing around with neutron moderators, this garden shed project could have become one of the nasty nuclear accidents which injured or killed a person through radioactive contamination. It would have "only" been similar to a radiotherapy patient receiving the wrong dose of gamma radiation, however, or perhaps a fire.
However, in the end, he frightened himself early enough that he was never really in the sort of danger that the Radium Girls suffered, let alone Slotin.