What? Did you not follow the US-Russia space race at all?
Indeed. Why were advances made? Becuase resources were pumped at the problem. Do you prefer resources to be pumped at putting 100s of lasers and nuclear weapons in orbit? And FYI the space race reached its zenith in the late 60s/early 70s. 'Competition between countries' moved on to new things - see how space exploration deteorated in the late 70s and early 80s when, a true barometer of 'competition', the amount of ICBMs and targetted military spending as a % of GDP increased to the maximum.
You seem to have a peculiar understanding of China. The Chinese government is a totalitarian dictatorship which has failed to reform. However they know this - their entire motivation is to keep the population happy to prevent revolt - that's why economic growth and repatriation of income across the country is such a massive priority - people don't revolt when, overall, living standards are advancing rapidly. The 'space race' China is undertaking is just another extension of this - keep the population happy that China is a player on the world stage, make the people happy with the government - the totalitarian dictatorship using the exact same population control methods the 'anti-communist democratic republic' the USA used in the 1960s, it is called patriotism and is indeed a dangerous tool - but there is a different in 'positive patriotism' in celebrating achievements and 'negative patriotism' which is saying 'my country is better/more powerful than yours' - interesting that the 'totalitarian dictatorship' China uses 'positive' tools to manage the population while the 'anti-communist democratic republic' is today using negative ones.
Strange paradox, eh? China can only be seen to seek militarisation of space if you think it is motivated by the same population control methods as the USA is today.
Surpnova asked for donations yes, but eXeem is unrelated to suprnova for all but publicity purposes - it was an existing development project which latched on to the recent publicity surrounding SUprnova (with its apparent approval).
Mentioning mindless clicking and downloading - the story takes you to a website where the eXeemLite software just downloads.
WTF? Computer savvy?
Download a cracked version of some shadowy software - and trust that its 100% legit??? Why not publish MD5 sum of the package which can be verified by expert users willing to take the plunge?
But if good programmers make lots and lots and lots of programs, and despite the 'new code' requirement gradually decreasing as they resuse old modules, they can easily end up at 1million+ lines. Unless they use Perl, where it could go on one really really long line.
Gets frustrating resisting the urge to reply to some of the crap that passes as economics around here:).
Ah ha! Indeed.
Omega really is pretty interesting (it has a really nice spinoff into prospect theory, using it as utility function/method rather than the traditional (and pretty suspect) additive utility). Don't look at the logs, it starts to get hard to understand scales then. But the threshold idea makes sense - if the distribution is asymetric (and not specifiable through skew or kurtosis) - then there is no (known) way to rank a selection of distributions - investors tend to have preferences/limits, so the threshold is modelled on them.
Yeah its being promoted as a hedge fund (and long only too, but less aggressively) analysis tool, which should be founded in some rationality, given the promotion of hedge funds is persistent return (adjusted for style bias), and that the return is a behavioural function of the market. But, as I posted earlier, a decent quantative model all comes down to its specification, which is 100% qualatitive. Despite Omega's difficulty in implementation, I'm pleased the issue of asymetric distributions is getting aired, as Shapre rations and Alpha-Beta calculations are blindly accepted as truth even with the asset managers with the biggest names, what gets promoted is a statistical coincidence at best 95% of times.
Automata are cool to model too, but I've not found much application outside of hypotheticals, I find them most useful in worse case scenarios, extreme events.
Michael Page, a quite-big finance-orientated recruiter/headhunter, wrote a very excellent piece on actually getting work as a quant, which people reading the comments on this story may find useful.
The link if chalmers.se because that's what Google gave me (didn't have the story bookmarked, had to search for it) - sorry don't have an original Michael Page URL. Please survive.
Of course never forget the independent variance-covariance condition, which is quant-101 material, but seems to get forgotten by all finance commentaries.
Relating to non-normal returns (some models use t-stats, but these may be similarly faulty) have you come across 'Omega Metrics' - it is a risk/return tradeoff function that uses the entire underlying distribution without making any assumptions r.e. the shape. Of course this means it is a sample descriptor, not a universal error function which could be parametrically defined (though you could monte-carlo a selection of scenarios defined along some qualatitive basis). Omega is being promoted by 'The Finance Development Centre', I can forward some papers if interested (web link not to hand). Fractals also seem to be having a new lease of life (and these can be parametric) with Mandlebrot's recent book.
Monte Carlo modelling is a technique, which is pretty sweet. The hard part is specifying the relationship and choosing the distribution of the error - but these are human choices. Investing over a period of time will reduce risk, but not necessarily in an optimal way - good model specification and an aggressive strategy has lower risk than time-averaging, but then good model specification is the holy grail, I suppose.
Although I have my own biases of things I 'like' and thing I, well, don't, I find it very useful to keep a loudmouth devil's advocate on my shoulder. That seems to do the trick.
Grandparent post was very interesting. To your specific points:
But the grandparent asked: why can't amiga be used on more common PPC hardware
Amiga OS4 does run on PPC (G3s and G4s are linked) did you bother to read the links before posting? Why didn't it run on PPC from the start? You do realise that was the mid 80s??? The PPC was created in early 90s.
(like Mac)
Well, because the Amiga did run on more common hardware (like Mac), the 68000.
...you register thousands of domains and manage them by bulk?
Talking about unethical slimeballs, what do you think about those guys that bulk buy websites with name like (with the odd spelling mistake) or names easily mistaken for other websites? Or domain campers that auction them off to the highest bidder?
I would be hesitant at C++ or Java. Perl, Python or Riby I would feel more comfortable in (given the concept in a skills test is assessment in a short space of time, a conceptually 'nice' RAD language (that can be programmed quickly as well as designed quickly) would win over stock languages).
However given a choice of language, and a desire to get the job done, did you consider APL or Brainfuck as a method of 1-upmanship?
Although "desire more things" is one the topics on the Economics syllabus in high school, it is in no way necessary. From a purely static standpoint, sometimes the preference is to desire less (e.g. pollution), sometimes it is to desire a certain amount (e.g. 2 children) - these could be defined to maximise emotional happiness ('utility' in econspeak), but that does not mean potential emotional happiness has to be an unbounded function.
The whole idea that 'economics'is about desiring more, is utter rubbish, just as much as citicising 'business' or 'capitalism' and branding that upon 'economics' (remember capitalism, environmentalism or plain nihilism is just as 'economic' as capitalism or communism).
Not that 'economics' is without its faults, namely the assumption of additive utility, something which is at long last starting to get its comuppance (with Kahaneman and Tversky style preferences).
Slashdot Mirror
See here
PS3 will contain the 'cell' chip from IMB, Sony, Panasonic. I think that is a pretty awesome detail.
What? Did you not follow the US-Russia space race at all?
Indeed. Why were advances made? Becuase resources were pumped at the problem. Do you prefer resources to be pumped at putting 100s of lasers and nuclear weapons in orbit? And FYI the space race reached its zenith in the late 60s/early 70s. 'Competition between countries' moved on to new things - see how space exploration deteorated in the late 70s and early 80s when, a true barometer of 'competition', the amount of ICBMs and targetted military spending as a % of GDP increased to the maximum.
You seem to have a peculiar understanding of China. The Chinese government is a totalitarian dictatorship which has failed to reform. However they know this - their entire motivation is to keep the population happy to prevent revolt - that's why economic growth and repatriation of income across the country is such a massive priority - people don't revolt when, overall, living standards are advancing rapidly. The 'space race' China is undertaking is just another extension of this - keep the population happy that China is a player on the world stage, make the people happy with the government - the totalitarian dictatorship using the exact same population control methods the 'anti-communist democratic republic' the USA used in the 1960s, it is called patriotism and is indeed a dangerous tool - but there is a different in 'positive patriotism' in celebrating achievements and 'negative patriotism' which is saying 'my country is better/more powerful than yours' - interesting that the 'totalitarian dictatorship' China uses 'positive' tools to manage the population while the 'anti-communist democratic republic' is today using negative ones.
Strange paradox, eh? China can only be seen to seek militarisation of space if you think it is motivated by the same population control methods as the USA is today.
Co-operation between countries in space exploration is only a good thing. Build up trust, knock down militarisation.
Just got it (eXeemLite) running on WINE under Linux (BYO).
Exactly.
Surpnova asked for donations yes, but eXeem is unrelated to suprnova for all but publicity purposes - it was an existing development project which latched on to the recent publicity surrounding SUprnova (with its apparent approval).
Mentioning mindless clicking and downloading - the story takes you to a website where the eXeemLite software just downloads.
WTF? Computer savvy?
Download a cracked version of some shadowy software - and trust that its 100% legit??? Why not publish MD5 sum of the package which can be verified by expert users willing to take the plunge?
Plus, has anyone tried this on WINE?
Good programmers often don't need a lot of lines.
But if good programmers make lots and lots and lots of programs, and despite the 'new code' requirement gradually decreasing as they resuse old modules, they can easily end up at 1million+ lines. Unless they use Perl, where it could go on one really really long line.
Gets frustrating resisting the urge to reply to some of the crap that passes as economics around here :).
Ah ha! Indeed.
Omega really is pretty interesting (it has a really nice spinoff into prospect theory, using it as utility function/method rather than the traditional (and pretty suspect) additive utility). Don't look at the logs, it starts to get hard to understand scales then. But the threshold idea makes sense - if the distribution is asymetric (and not specifiable through skew or kurtosis) - then there is no (known) way to rank a selection of distributions - investors tend to have preferences/limits, so the threshold is modelled on them.
Yeah its being promoted as a hedge fund (and long only too, but less aggressively) analysis tool, which should be founded in some rationality, given the promotion of hedge funds is persistent return (adjusted for style bias), and that the return is a behavioural function of the market. But, as I posted earlier, a decent quantative model all comes down to its specification, which is 100% qualatitive. Despite Omega's difficulty in implementation, I'm pleased the issue of asymetric distributions is getting aired, as Shapre rations and Alpha-Beta calculations are blindly accepted as truth even with the asset managers with the biggest names, what gets promoted is a statistical coincidence at best 95% of times.
Automata are cool to model too, but I've not found much application outside of hypotheticals, I find them most useful in worse case scenarios, extreme events.
Michael Page, a quite-big finance-orientated recruiter/headhunter, wrote a very excellent piece on actually getting work as a quant, which people reading the comments on this story may find useful.
The link if chalmers.se because that's what Google gave me (didn't have the story bookmarked, had to search for it) - sorry don't have an original Michael Page URL. Please survive.
IAWTP
Of course never forget the independent variance-covariance condition, which is quant-101 material, but seems to get forgotten by all finance commentaries.
Relating to non-normal returns (some models use t-stats, but these may be similarly faulty) have you come across 'Omega Metrics' - it is a risk/return tradeoff function that uses the entire underlying distribution without making any assumptions r.e. the shape. Of course this means it is a sample descriptor, not a universal error function which could be parametrically defined (though you could monte-carlo a selection of scenarios defined along some qualatitive basis). Omega is being promoted by 'The Finance Development Centre', I can forward some papers if interested (web link not to hand). Fractals also seem to be having a new lease of life (and these can be parametric) with Mandlebrot's recent book.
Dude,
Monte Carlo modelling is a technique, which is pretty sweet. The hard part is specifying the relationship and choosing the distribution of the error - but these are human choices. Investing over a period of time will reduce risk, but not necessarily in an optimal way - good model specification and an aggressive strategy has lower risk than time-averaging, but then good model specification is the holy grail, I suppose.
Although I have my own biases of things I 'like' and thing I, well, don't, I find it very useful to keep a loudmouth devil's advocate on my shoulder. That seems to do the trick.
sounds like another name for another linux distribution. screw it.
Grandparent post was very interesting. To your specific points:
But the grandparent asked: why can't amiga be used on more common PPC hardware
Amiga OS4 does run on PPC (G3s and G4s are linked) did you bother to read the links before posting? Why didn't it run on PPC from the start? You do realise that was the mid 80s??? The PPC was created in early 90s.
(like Mac)
Well, because the Amiga did run on more common hardware (like Mac), the 68000.
why dont you throw some water and icing sugar into a bowl (raisins or 100s and 1000s for flavour?), mix together, and splat a spoonful on a tray?
That's what I do, and the diabetes was worth it!
Waaah! Bad made up counterarguments make baby Mohammed cry.
Did you test for normality? What about skew and kurtosis? Are there lower (upper?) partial moments?
Now stop pretending to know statistics.
...you register thousands of domains and manage them by bulk?
Talking about unethical slimeballs, what do you think about those guys that bulk buy websites with name like (with the odd spelling mistake) or names easily mistaken for other websites? Or domain campers that auction them off to the highest bidder?
APL wins. APL2 would be best.
I would be hesitant at C++ or Java. Perl, Python or Riby I would feel more comfortable in (given the concept in a skills test is assessment in a short space of time, a conceptually 'nice' RAD language (that can be programmed quickly as well as designed quickly) would win over stock languages).
However given a choice of language, and a desire to get the job done, did you consider APL or Brainfuck as a method of 1-upmanship?
Is pretty easy to spot - just see someone running at warp speed to the extent they've knifed the other teams 2 seconds after spawn.
If that's your idea of porn... you are a sick individual.
I interpreted it as a typical project management meeting.
...I'm interested. As a developer previously Windows based, but moving to Open Source, I'm reading up on GTK+ and QT and their related GUIs. I found this (useful info, though god knows why they put it in a pic very informative.
"I feel lucky" is nearly never used.
Except when concealing goatse.cx links.
Although "desire more things" is one the topics on the Economics syllabus in high school, it is in no way necessary. From a purely static standpoint, sometimes the preference is to desire less (e.g. pollution), sometimes it is to desire a certain amount (e.g. 2 children) - these could be defined to maximise emotional happiness ('utility' in econspeak), but that does not mean potential emotional happiness has to be an unbounded function.
The whole idea that 'economics'is about desiring more, is utter rubbish, just as much as citicising 'business' or 'capitalism' and branding that upon 'economics' (remember capitalism, environmentalism or plain nihilism is just as 'economic' as capitalism or communism).
Not that 'economics' is without its faults, namely the assumption of additive utility, something which is at long last starting to get its comuppance (with Kahaneman and Tversky style preferences).