Posts by Arthur Dent
67 posts • joined Friday 20th August 2010 16:08 GMT
Looks a bit incoherent to me
It's fine by me if Kroess tries to enforce some honesty in advertising by ISPs. In fact if she does that so that we can have a sensible market for ISP services it will be an excellent thing. However a single figure for bit rate would not be a reasonable way of describing the service, since their is contention. It has to be somethink like a maximum figure, an mean figure, a median figure, and a minimum figure where the figures are to be measured over say a 1 second interval. And there has to be an equally good description of the upload speeds. However, her rhetoric seems to indicate that's she's been taken in by the ridiculous claims of the network neutrality advocates that the development of new network facilities which are charged differently will destroy the internet; so she may have the opposite effect to that she probably intends, and contribute to a contraction of network capacity as operators pull out because it becomes impossible for them to make a profit. I hope I am wrong on this, but the rhetoric of "neutrality" is there plain to see and if it prevails it will severly damage the internet.
But even forcing the ISPs to state what they will deliver is fraught with technical difficulties. It would also be necessary to specify where the contracted bandwidth will reach - sitting here I have far more bandwidth and far lower latency to Madrid than I do to Delhi, and there is exactly nothing that my ISP (Telefonica) can do about that the bandwidth to Dehli - networks between here and there just can't handle it. So it's quite possible that the description of the service to be provided becomes far too complex for the average consumer (so many Mb/s to there, so many kb/s to there, and so on) unless the ISP goes for a lowest common denominator and says something like "I can sell you 100kb/s down and 100kb/s up because that's the most that can be achieved on the worst internet connection from here" even though it's vastly less that what I might get (and indeed need) on most of the sites that I visit - and then of course the ISP would be perfectly correct to throttle me down to 100kb/s always; and the ISP can't offer a 500kb/s service at any price if the rules are put together badly, because it will be "regulated" out of existence for failing to deliver what it offered because the connection to Chitral never achieves that. And all the people baying about having a single simple number that specifies what you get and the ISP being required to provide that number all the time wherever you are downloading from are heading straight for badly constructed rules.
Actually...
"Read a proper and unbiased article somewhere on the amount of _RADIOACTIVE_ material which is dumped into the atmosphere by a coal burning plant without proper filters (which is the case in China). That is besides all other pollutants."
That's a good idea.
About 6 decades ago Otto Frisch wrote a short piece on the safety of coal burning plants, pointing out the very severe pollution with radioactive material that they generate. At the time he held the Jacksonian Chair of Natural Philiosophy at Cambridge University, and a few years earlier he was one of the authors of the Frisch-Peierls memorandum (which described how to use conventional explosives to obtain criticality at lower mass, and the effects the resultant fission explosion would have, including a good description of the fallout) so he should probably be regarded as fairly well qualified on topics like radioactive pollution. Maybe Lomax could read his "On the safety of coal burning power stations" (I think that was the title) if he can find a copy and the words aren't too long for him.
This is nonsense
The ECJ in its ruling is clearly stating that the decision of the European Parliament and the European Commission that the directive concerned would have national opt-outs not limited by time is something it is empowered to overrule - in other words, the legislators don't decide what the law is (neither the elected ones in the European Parliament no the appointed ones in the Commission), and no matter what the text of the law says the ECJ can say it means something completely different.
How did we ever get here? It doesn't even matter whether the decision is a good one or a bad one in the sense that this discrimination should not or should be permitted, what matters is that this court has arrogated to itself the right to make new law whichdiirectly contradicts the laws put in place by the legislative bodies empowered by the various treaties to determine what the law shall be.
Cherry picking data
<quote>1998 was an unusually hot year. Can you tell me what the trend is like if you chose 1997 or 1999 as your reference point?</quote>
Read the article, jonathanb, and you'll see why no-one can give you a meaningful answer using 1997 or 1999 as the point of division.
1998 was the last year of the recent upward trend of the mean close to surface temperatures on the 24.5N parallel in the Atlantic. This steady increase (which lasted about 42 years) was followed by a decline and in the following 8 years that area of ocean lost more that half the temperature increase gained in the preceeding 42 years. You can't campare behaviour on two sides of a turning point by taking something other than the turning point as your base point, it would make no sense.
It's quite clear that the rapid decrease was not caused by global warming, but was part of the ordinary short term variation in ocean temperatures in that area caused by short term wind variations leading to slightly different current patterns. That's pointed out in the article quite straightforwardly.
All the comments about how this relates to global warming and slagging off the article as being written from a NNCC-sceptic POV are bullshit, and your reference to "cherry picking" is probably the stupidest.
Posted in BA flights website grounded by snow
"nonsens isNonsense" is NOT nonsense
Alister, this is not somethiong that happens once per generation at Heathrow. Last year (2009) flights were cancelled because of snow at Heathrow in February and in December. It happened also in February 2008. In recent years the average number of periods of flight cancellations at Heathrow caused by snow is more than one per year. So basing your ideas about the economics of having clearing equipment available on some crackpot theory that it happens only once in fifty years is not "common sense", a trait which you appear to be sadly lacking, but pure nonsense.
Amusing!
It's amusing in a way to see how so many people rant about net neutrality without having a clue what it's about. It's already been said, but let's say it again: net neutrality is not about not capping bandwidth (as a scientist I deplore the use of the term "bandwidth" to mean what it's being used to mean here, but if I use the proper term most of the clueless will not understand me). It is about maintaining a "one size fits all" myth together with a ludicrous charging model linked to an outrageously unfair concept of "fair charging". Even the (true in the real world) idea that real-time interactive work has different latency requirements for its communications than does background bulk file sharing seems to be anathema to advocates of this spurious neutrality, as does the idea that people should be able to buy higher quality by paying for it. I strongly recommend Bob Briscoe's paper on fairness (http://bobbriscoe.net/projects/2020comms/refb/fair_ccr.pdf) for one view of an alternative to so-called "neutrality", and Frank Kelly's numerous papers on related topics for a solid grounding in the science of shared networks (a list of his papers can be found at http://www.statslab.cam.ac.uk/~frank/PAPERS/ One particular paper, "Explicit Congestion Control: charging, fairness and admission management", written in collaboration with Gaurav Raina, is perhaps a good starting point in the net neutrality context: http://www.statslab.cam.ac.uk/~frank/PAPERS/ecc.pdf) - and I hope that some of those who have commented above will take the trouble to educate themselves by looking at some of this scientific literature rather than the propaganda spewed out on this topic by the likes of Google and the BBC; but the tone of many of the comments suggests that this is a forlorn hope.
Posted in No wonder CompSci grads are unemployed
Re: On the button (AC 1/11/2010 12:58 GMT)
<quote>Personally, I don't care what languages who have or what DBs you have used. But you must have more than one language and know when to use different ones; and at least have good SQL. Know that and I can teach you how our shop works. But without that you are a drone.<unquote>
What barbarous nonsense. Ted Codd studied maths and knew nothing of SQL when he invented the relational model for databases without which SQL would never have never existed. I was a lot younger than Ted, but was doing research in databases and information retrieval long before SQL was invented. SQL is just another language (actally a badly screwed up version of Ted Codd's idea for a relational calculus based language) and saying that someone who doesn't know that particular language is a "drone" for that reason is total garbage (and I say this as someone whose last two technical director/VP level jobs were based partly on my SQL expertise, not as someone who wants to claim SQL doesn't matter because they don't know it).
<quote>We used to have an education system that was the envy of the world. Where did it all go wrong?<unquote>
When they began to let idiots who think some particular computer language is an essential part of CS eductaion have some influence? (Unfortunately that really has happened, and at a large number of Universties that language is Basic, and - o tempora, o mores - at an even larger number it is Java; but the real bad news is that it's yet more often C++.)
Re: storage
It seems clear that at least one anonymous coward thinks that high tide is simultaneous everywhere on the GB coast. since it hasn't occurred to him that 10 tidal generators could be built in positions such that no two have a slack time within an hour of each other - so only one would having a slack time at any given time.
re: he's dead
@Jimmy Floyd: To understand what "Viva Franco" actually means you could try using a decent Spanish dictionary.
The 22nd edition of the Diccionario de la Lengua Española of La Real Academia Española gives twelve meanings for the verb "vivir", of which one (the seventh) is specifically applicable after death/ It has nothing to do with life after death or rising from the grave or anything of that sort, as this dictionary definition is "Mantenerse o durar en la fama o en la memoria después de muerto" - now just take the secon person singular imperative of that and you have the meaning or "Viva" in "Viva Franco".
You can check this at
http://buscon.rae.es/draeI/SrvltGUIBusUsual?TIPO_HTML=2&TIPO_BUS=3&LEMA=vivir
re: RIPA trumps *everything
No it doesn't.
If the authorities have in their possession an encrypted copy of what would have been, did they not have that copy, a privileged document and acquired it in such a manner that the acquisition did not breach privilege, then they can require an unencrypted copy (or in some circumstances a decryption key) to be provided.
That doesn't enable them to acquire a privileged document, just to acquire a plain text copy of a document that properly and legally came to be in their possession in encrypted form without any breach of privilage.
Unless I've misunderstood it all completely; I'm not a lawyer.
Re: Don't trust them
MS updates my Outlook spam filter quite often. Have you perhaps disabled Microsoft Update on the machines it doesn't get updated on?
There is a real MS update screwup but that is with root certs - treated as an optional update so it won't happen automatically.
Why not...?
"If we're including the Dr tag, why not just refer to him by his proper first name, James?
That's right, his actual name is James Brown. "
It may surprise you to know, Mr Isurrenderall, that in a large part of Scotland and some of Northern England the name normally used is the middle name, not the first. If that does surprise you, please evaluate whether you are an ignorant and arrogrant bigot with neither knowledge nor respect for cultures other than your own provincial prejudices. If it doesn't surprise you, what reason did you have for posting nonsense?
Anyway, everyone knows that Gordon Brown is a failed Chancellor, a failed Prime Minister, a failed Party Leader, and a widely reviled and disliked character. So let's not call him James Brown - doing so might let him escape his deserver reputation. There you go, I've answered your stupid question twice for you now.
Software patents in the UK
Why does anyone pretend that the UK (or the EEC) doesn't permit software patents? We've had them for years. I used to work for a large UK company where software developers were encouraged to patent their ideas for out employer's benefit (with the help of some people very competent at formulating claims) by a bonus for every patent application that passed the company's internal checks. We were even forbidden to publish in technical journals until we could put a "patent applied for" footnote on the abstract. My unit produced a lot of patent applications, almost all resultingg in granted patents. These were for real innovation, the result of intensive research, with no prior art discovered in a serious search before the applications were filed - but although they didn't display all the faults of the US system they were still, for all practical purposes, software patents. I would be very surprised if the experience of anyone who has done serious computing research in the UK (whether in industry or in a university) is any different.
Frankly, I wish we didn't have software patents (despite having collected some bonuses for applying for them) because I know that bureaucracies will generally not be able to distinguish a new invention in software from a trivial application of well-known principles, and neither will the courts. Unfortunately internatiopnal treaties require us to enforce utter nonsense patents approved in the US. Maybe what we should be doing is requiring non-European patents to pass a review considering prior art and inventive step and non-obviousness before they can be enforced in Europe - I suspect most of the bad US software patents would become unenforceable here, although some would inevitably slip through the checks.
Radioactive decay
<quote>Radioactive decay happens at constant speed_ and it's a process which don't give a damn if it's happening in nature or in a nuclear reactor.<unquote>
God in heavens! I have nevr seen a better demonstration that a little knowledge is a dangerous thing than that piece of total drivel!
Have you heard or moderating nuclear reactors - ie controlling the chance of a neutron released by the fission of one nucleus to cause nother nucleus to split? This is changing the decay rate, and is how the therrmal output of a reactor is controlled. Do you not realise that a fission bomb is just a fission reactor in which that chance has got too close to 1, and that the decay takes place very rapidly in a bomb? Heaven help us, learn a little about physics and nuclear engineering before posting and then perhaps you won't post such blatant nonsense.
Re "However"
<quote>The RAF found it easier to drop two Grand Slam bombs (total 8 tons of Torpex) rather than lift a spare* municipal swimming pool into low orbit.</quote>
8 tons of Torpex sounds more like either two Tallboys or a single Grand Slam than two Grand Slams (Grand Slam was the 22000 lb camouflet bomb; Tallboy was the 12000 lb version).
Posted in More choosing maths A-Level
Whatever happened to
good old-fashioned S-levels? The new A* appears to be a dumbed down version of these.
Over the last 40 years, the maths syllabi for A-level have lost a lot of topics and some of the comments above seem to suggest that the syllabus for A level additional maths has become identical to that for A level maths, which is appalling if true. The maths syllabus has gained just one topic (trivial set theory) to compensate for all those it has lost. I think the narrowing down began about 35 or 40 years ago, but the hard data I have is the papers from the 50s and early 60s (that we used for practise in the A-level courses) and papers from around 1990 (which I used for excercises for a couple of kids I was helping) which contained a much narrower range of questions than had been the case in the 50s and 60s and had quite a few questions questions (particularly on calculus) that were the sort of thing that could have turned up on O-level Maths or O-level Add Maths papers in the 60s.
For example there appears to be nothing on numerical methods now, but at a bog-standard comprehensive in the early 60s I was taught various numerical techniques for solving various problems, and taught to use a rotary pinwheel calculator to excercise those techniques and to discover in practise that the good methods not only were much more efficient than the bad ones but also much more accurate (less rounding error). There's nothing like that in schools today (a modern course would of course need to use a computer rather than a mechanical calculator).
