I've heard a lot of economists and "experts" (and the IMF - who are neither;) ) say a very different thing during the Asian financial crisis in 1997. Especially the western ones.
They were saying "no bailouts".
I figure they wanted stuff to go bust so they could buy it all up cheap and thus gain more power and wealth.
"It's truly pathetic that MS Office 2k3 runs faster under vmware+xp than OO does natively in linux."
Yeah, but I've mentioned stuff like that before and I got modded troll for it.
Hopefully 3.0 will be faster, I use OOo on Linux at work and it takes _ages_ to start.
If they get it right, maybe a lot of companies might actually switch from MSO 2K3 to OOo instead of going to MSO 2007 - since switching to MSO 2007 will require massive retraining/relearning, perhaps more than even switching from MSO2K3 to OOo.
A good design for complexity does NOT require _exposing_ the complexity by default.
In my opinion programming is a form of compression. Decision compression- rather than zillions of if-then-else, you compress that to recursion, functions, classes etc. A good programmer knows which bits to keep uncompressed and which bits to compress so that the overall behaviour is easy to change when the Boss wants it changed AND more importantly so that compressed result is compatible with other decent programmers.
A good programming language has good defaults in its "huffman codebook" for the expected problem domains, to make it easier for the programmer to achieve good compression with less "human CPU" time:).
IMO, saying there is one perfect language is like saying there is one perfect compression algorithm.
But a programming language that has "poor compression" (need to write a lot) or "very high brain power or memory requirements" (need to think/remember a lot for stuff trivial in other languages) is crap, unless it has redeeming features like extremely high runtime performance.
Some XML config files for programs have more lines of code than an equivalent program written in popular higher level languages.
If it's easier to figure out how to modify a server written in Lisp/Python/Perl to do what you want than to modify some fancy config file for a similar server, then something is wrong somewhere.
A sufficiently "advanced" XML config file resembles Lisp after you convert tags to parentheses:).
I'd say _also_ learn perl if just because of CPAN alone. To me the best form of code reuse is being able to reuse someone else's code that's better than yours:).
While languages like Lisp are very powerful - in that smart programmers don't have to write so much to do a lot, the problem for stupid and lazy people like me is I still have to think and write a lot more since there isn't as much great stuff out there to reuse (yes I know the less people using a language = less code out there to reuse, and that is sometimes not the fault of the language).
It's the difference between being able to recursively build a house atom by atom exactly the way you want it, and being able build one using mostly prefab components. The latter is likely to be faster for stupid and lazy people like me:).
The boss/customer usually doesn't care if each door hinge is perfectly macro-crafted - he's probably going to change his mind on where the doors are going to be anyway...
In short: Lisp is powerful - for the code you write (expressiveness). C is powerful - for the code you write (low level control + performance). Perl is powerful - for the code you _didn't_ write (laziness).:)
Itanium? People laughed and started calling it the Itanic. Many still do, some might also call it an EPIC failure. It's only good for very niche applications. You're usually better off using an x86 (from AMD or Intel) or IBM POWER.
The x86 is still king. It may be as ugly as a pig with a rocket strapped on, but it still flies faster than those elegant RISC eagles.
While IBM's POWER stuff might be faster, it sure doesn't like a RISC anymore - definitely not a very "reduced instruction set":). My bet is complex instructions that do a lot will keep winning as long as memory bandwidth and latency are constraints.
The Real Thing is too slow and will be too slow. The way I see it, everyone will still have to use "tricks".
Carmack has had a good track record of figuring out nifty tricks that current popular tech can achieve or at least the popular near cutting edge tech - I remember just barely managing to play the first Doom on a 386SX, it sure looked a lot better than the other stuff out there.
He used tricks for commander keen, wolf 3d, doom (2D game with some 3D), and so on.
Carmack's engines tend to do pretty decent FPS for the visuals you get (maybe with the exception of Doom 3, I can't remember whether that one had a good balance or not - because I stuck with my Ti4200:) ).
Stuff like Crysis looks great, but you'll need an 8800GT just to get decent frames per sec at 1680 x 1050, and not so many months back (before the 8800GT and the ATI/AMD 3870 etc) something that fast would have cost way too much money for >=99% of the people. Now maybe it's too much for 90-95% of the people.
That said, nowadays with the current tech, if you have good enough artists and textures, they can cover over a lot of graphics engine inadequacies. If you compare crysis in vista dx10 very high quality vs the hacked XP dx9 very high quality, while there are shadow differences, the textures make the XP version look almost as good. Also for games like TF2 where the graphics style is not so much on realism, I don't really see it being a minus if done well.
In the end one may no longer need the latest and greatest tricks. The "bog standard crap" might end up good enough. Lots of people are still playing Counterstrike, Starcraft, Warcraft, Sims etc and they don't appear to care about cutting edge graphics quality.
We present our candidates: 1) George W Bush 2) Tony Blair 3) Simon Cowell
Get your phones ready to vote.
send <candidate number> to 33333 for return trip send <candidate number> to 44444 for "one way".
50p per vote, you can vote as many times as you want.
Alternatively visit votethemoff.theplanet.com to vote on line:).
Conditions and terms apply. Not all candidates will accept the wonderful opportunity to go to space, in which case the organizers will have a good laugh all the way to the bank.
Visit the Hall of Fame at votedoff.theplanet.com to see our past winners.
(note this is not for real - there's no such site or show yet:) )
"So what is the point of sending people to different planets?"
Well I've been proposing an election or reality TV show where you get to vote people off the planet (Earth).
Choices are 1-way, or return trip.
There should be no problem getting the USD20 million or so it takes to send someone to space - just have one of those SMS voting thingy. I wouldn't mind putting a fair amount of money on some candidates even if they politely decline their tickets. It'll be worth the interviews and questions;).
My colleague actually registered voteofftheplanet.com or something but I think it's now being squatted on by one of the usual bunch.
Space elevators would be in a geostationary orbit.
They can't easily sync with medium or lower orbits - any stuff lower down will be whizzing past the tether (if not colliding with it).
Geostationary orbit - at the centre of mass = 3km/s. Tether dangling down to the altitude of a medium or low orbit will be travelling at 3km/s * (6400km+ altitude)/35000km = 0.6 to 1km/s (at surface it's about 0.4km/sec).
low orbit satellite = 8km/s medium orbit = slower than 8km/s but faster than 3km/s and definitely a lot faster than 1km/sec
Slowing down something travelling at 5km/sec so it can hook up with a tether travelling at 1km/sec is not trivial. After that how are you going to speed it back up again so that it can ferry stuff from medium orbit to 35000km up?
If there is a storm cloud at a convenient location you could fire the laser to create an ionized path from cloud to target and make it look like lightning did it. Problem is lightning is not very good at killing so if people are nearby the target could be resuscitated.
Anyway, it's more of a "learning thing" - lots of people claim that the space elevator thing is a good idea. The idea has its merits IF it can actually be done safely and it actually works out cheaper than rockets in practice (rather than just theory:) ).
Doing tethered satellites would be good practice - I doubt anybody will be able to get a space elevator right first time.
Not sure we're ready for such things politically - such satellites (and elevators) would be extremely vulnerable to hostiles.
Well I think they are doing things wrong. They keep talking about travelling to Mars etc when what they should do is focus on building much better space stations. Once you have a space station with artificial "gravity", decent radiation shielding, and all the other good stuff so that astronauts can live on it for years without suffering so much like the russian astronauts, then you can talk about travelling. In fact people might then prefer to travel to the asteroid belt instead - get raw materials for building more space stations without having to spend lots of energy fighting a gravity well.
They might also want to try out tethered satellites. Instead of a full space elevator right from the start, try suspending the "comms/sensor" bits of the satellite closer to earth, with the counter weight at the other end (solar panels etc), so that the satellite is still in geostationary orbit, but you have much better comms latencies. I suspect some people are willing to pay a premium for lower latency sat comms. If they can't even do such satellites then I think trying for a space elevator is silly.
Because the company probably wouldn't have that money anymore, it's mostly gone to the bosses.
So the company goes bust and the bosses live happily ever after.
They might even be able to do it over and over again, I'm sure there are plenty of tricks:).
Whereas if you jail the bosses for massive fraud it is a deterrent, since for the X years they are in jail they can't enjoy the stuff they bought or could buy. Everyone has a limited lifespan, the richer you are the less you want to spend your life in jail.
So once the journals change their rules, submitters will be allowed to advertise their stuff on Wikipedia?
Then you might see on pages many citations/references leading to a "please pay $$$" journal.
I vote for 6.28318531...
on
Happy Pi Day
·
· Score: 1
The current pi _is_ a bit arbitrary.
In my opinion using the ratio of circumference to diameter is stupid. I've found that strange since the first I learnt of pi.
Doing the ratio of circumference to radius would have made more sense - the radius and circumference are more related to the same centre point of a circle.
If you see lots of 2 * pi in formulas that's more evidence people got it wrong.
6.28318531... would have been a better number than the existing 3.141....
Slashdot Mirror
I've heard a lot of economists and "experts" (and the IMF - who are neither ;) ) say a very different thing during the Asian financial crisis in 1997. Especially the western ones.
They were saying "no bailouts".
I figure they wanted stuff to go bust so they could buy it all up cheap and thus gain more power and wealth.
Hypocracy - the rule or power through hypocrisy.
"It's truly pathetic that MS Office 2k3 runs faster under vmware+xp than OO does natively in linux."
Yeah, but I've mentioned stuff like that before and I got modded troll for it.
Hopefully 3.0 will be faster, I use OOo on Linux at work and it takes _ages_ to start.
If they get it right, maybe a lot of companies might actually switch from MSO 2K3 to OOo instead of going to MSO 2007 - since switching to MSO 2007 will require massive retraining/relearning, perhaps more than even switching from MSO2K3 to OOo.
How much is my left little finger worth?
Don't get the wrong idea, I'm quite attached to it.
So you'll have to prise it from my cold dead hands (or over my dead body)...
Oh wait...
Yeah. My brain is not working.
:).
Anyway point is if players think the Casino isn't fair and there are too many players cheating they might go elsewhere.
Yes I know a stockmarket isn't like a casino. The top stockmarkets don't operate for 24 hours all year
A good design for complexity does NOT require _exposing_ the complexity by default.
:).
In my opinion programming is a form of compression. Decision compression- rather than zillions of if-then-else, you compress that to recursion, functions, classes etc. A good programmer knows which bits to keep uncompressed and which bits to compress so that the overall behaviour is easy to change when the Boss wants it changed AND more importantly so that compressed result is compatible with other decent programmers.
A good programming language has good defaults in its "huffman codebook" for the expected problem domains, to make it easier for the programmer to achieve good compression with less "human CPU" time
IMO, saying there is one perfect language is like saying there is one perfect compression algorithm.
But a programming language that has "poor compression" (need to write a lot) or "very high brain power or memory requirements" (need to think/remember a lot for stuff trivial in other languages) is crap, unless it has redeeming features like extremely high runtime performance.
Don't forget the XML config file part...
:).
Some XML config files for programs have more lines of code than an equivalent program written in popular higher level languages.
If it's easier to figure out how to modify a server written in Lisp/Python/Perl to do what you want than to modify some fancy config file for a similar server, then something is wrong somewhere.
A sufficiently "advanced" XML config file resembles Lisp after you convert tags to parentheses
How about when people in effect write Lisp-like interpreters in Java, and then proceed to write Lisp-like programs in XML configuration files?
:).
I'm kidding. Kind of I think
You want the Casino (or a favored player) to be able to know what the next cards are before other people at the table?
If there are other Casinos around, nobody will want to play at your "Milton Friedman approved" Casino.
I'd say _also_ learn perl if just because of CPAN alone. To me the best form of code reuse is being able to reuse someone else's code that's better than yours :).
:).
:)
While languages like Lisp are very powerful - in that smart programmers don't have to write so much to do a lot, the problem for stupid and lazy people like me is I still have to think and write a lot more since there isn't as much great stuff out there to reuse (yes I know the less people using a language = less code out there to reuse, and that is sometimes not the fault of the language).
It's the difference between being able to recursively build a house atom by atom exactly the way you want it, and being able build one using mostly prefab components. The latter is likely to be faster for stupid and lazy people like me
The boss/customer usually doesn't care if each door hinge is perfectly macro-crafted - he's probably going to change his mind on where the doors are going to be anyway...
In short:
Lisp is powerful - for the code you write (expressiveness).
C is powerful - for the code you write (low level control + performance).
Perl is powerful - for the code you _didn't_ write (laziness).
Itanium? People laughed and started calling it the Itanic. Many still do, some might also call it an EPIC failure. It's only good for very niche applications. You're usually better off using an x86 (from AMD or Intel) or IBM POWER.
:). My bet is complex instructions that do a lot will keep winning as long as memory bandwidth and latency are constraints.
The x86 is still king. It may be as ugly as a pig with a rocket strapped on, but it still flies faster than those elegant RISC eagles.
While IBM's POWER stuff might be faster, it sure doesn't like a RISC anymore - definitely not a very "reduced instruction set"
The Real Thing is too slow and will be too slow. The way I see it, everyone will still have to use "tricks".
:) ).
Carmack has had a good track record of figuring out nifty tricks that current popular tech can achieve or at least the popular near cutting edge tech - I remember just barely managing to play the first Doom on a 386SX, it sure looked a lot better than the other stuff out there.
He used tricks for commander keen, wolf 3d, doom (2D game with some 3D), and so on.
Carmack's engines tend to do pretty decent FPS for the visuals you get (maybe with the exception of Doom 3, I can't remember whether that one had a good balance or not - because I stuck with my Ti4200
Stuff like Crysis looks great, but you'll need an 8800GT just to get decent frames per sec at 1680 x 1050, and not so many months back (before the 8800GT and the ATI/AMD 3870 etc) something that fast would have cost way too much money for >=99% of the people. Now maybe it's too much for 90-95% of the people.
That said, nowadays with the current tech, if you have good enough artists and textures, they can cover over a lot of graphics engine inadequacies. If you compare crysis in vista dx10 very high quality vs the hacked XP dx9 very high quality, while there are shadow differences, the textures make the XP version look almost as good. Also for games like TF2 where the graphics style is not so much on realism, I don't really see it being a minus if done well.
In the end one may no longer need the latest and greatest tricks. The "bog standard crap" might end up good enough. Lots of people are still playing Counterstrike, Starcraft, Warcraft, Sims etc and they don't appear to care about cutting edge graphics quality.
Reality Show?
:).
:) )
How about "Vote them off the planet!".
We present our candidates:
1) George W Bush
2) Tony Blair
3) Simon Cowell
Get your phones ready to vote.
send <candidate number> to 33333 for return trip
send <candidate number> to 44444 for "one way".
50p per vote, you can vote as many times as you want.
Alternatively visit votethemoff.theplanet.com to vote on line
Conditions and terms apply. Not all candidates will accept the wonderful opportunity to go to space, in which case the organizers will have a good laugh all the way to the bank.
Visit the Hall of Fame at votedoff.theplanet.com to see our past winners.
(note this is not for real - there's no such site or show yet
Well whose money did he use to get there?
That said, I might be very tempted to pay $$$ to send various politicians to Mars. It might be a good thing for everyone else.
"So what is the point of sending people to different planets?"
;).
Well I've been proposing an election or reality TV show where you get to vote people off the planet (Earth).
Choices are 1-way, or return trip.
There should be no problem getting the USD20 million or so it takes to send someone to space - just have one of those SMS voting thingy. I wouldn't mind putting a fair amount of money on some candidates even if they politely decline their tickets. It'll be worth the interviews and questions
My colleague actually registered voteofftheplanet.com or something but I think it's now being squatted on by one of the usual bunch.
I don't understand your suggestion.
Space elevators would be in a geostationary orbit.
They can't easily sync with medium or lower orbits - any stuff lower down will be whizzing past the tether (if not colliding with it).
Geostationary orbit - at the centre of mass = 3km/s.
Tether dangling down to the altitude of a medium or low orbit will be travelling at 3km/s * (6400km+ altitude)/35000km = 0.6 to 1km/s (at surface it's about 0.4km/sec).
low orbit satellite = 8km/s
medium orbit = slower than 8km/s but faster than 3km/s and definitely a lot faster than 1km/sec
Slowing down something travelling at 5km/sec so it can hook up with a tether travelling at 1km/sec is not trivial. After that how are you going to speed it back up again so that it can ferry stuff from medium orbit to 35000km up?
I'm the tech guy who doesn't (currently) work in IT.
:).
;).
Once the graphics card failed on my machine (Dell, with Nvidia vidcards with those faulty capacitors) it took IT _weeks_ to get it replaced.
I only need one port open - port 443. HTTP Proxy is fine just let me do the CONNECT a.b.c.d
My choice of "Fav App"? vmware server
If there is a storm cloud at a convenient location you could fire the laser to create an ionized path from cloud to target and make it look like lightning did it. Problem is lightning is not very good at killing so if people are nearby the target could be resuscitated.
Yeah where they stock useful books like Despotism for Dummies.
Probably not.
:) ).
Anyway, it's more of a "learning thing" - lots of people claim that the space elevator thing is a good idea. The idea has its merits IF it can actually be done safely and it actually works out cheaper than rockets in practice (rather than just theory
Doing tethered satellites would be good practice - I doubt anybody will be able to get a space elevator right first time.
Not sure we're ready for such things politically - such satellites (and elevators) would be extremely vulnerable to hostiles.
Well I think they are doing things wrong. They keep talking about travelling to Mars etc when what they should do is focus on building much better space stations. Once you have a space station with artificial "gravity", decent radiation shielding, and all the other good stuff so that astronauts can live on it for years without suffering so much like the russian astronauts, then you can talk about travelling. In fact people might then prefer to travel to the asteroid belt instead - get raw materials for building more space stations without having to spend lots of energy fighting a gravity well.
They might also want to try out tethered satellites. Instead of a full space elevator right from the start, try suspending the "comms/sensor" bits of the satellite closer to earth, with the counter weight at the other end (solar panels etc), so that the satellite is still in geostationary orbit, but you have much better comms latencies. I suspect some people are willing to pay a premium for lower latency sat comms. If they can't even do such satellites then I think trying for a space elevator is silly.
"How is that not a deterrent?"
:).
Because the company probably wouldn't have that money anymore, it's mostly gone to the bosses.
So the company goes bust and the bosses live happily ever after.
They might even be able to do it over and over again, I'm sure there are plenty of tricks
Whereas if you jail the bosses for massive fraud it is a deterrent, since for the X years they are in jail they can't enjoy the stuff they bought or could buy. Everyone has a limited lifespan, the richer you are the less you want to spend your life in jail.
So once the journals change their rules, submitters will be allowed to advertise their stuff on Wikipedia?
Then you might see on pages many citations/references leading to a "please pay $$$" journal.
The current pi _is_ a bit arbitrary.
In my opinion using the ratio of circumference to diameter is stupid. I've found that strange since the first I learnt of pi.
Doing the ratio of circumference to radius would have made more sense - the radius and circumference are more related to the same centre point of a circle.
If you see lots of 2 * pi in formulas that's more evidence people got it wrong.
6.28318531... would have been a better number than the existing 3.141....
I believe you got the second part of the quote wrong. But yes he was saying a similar thing overall.
Well it's a fancy way of killing someone.
:)
Maybe someone one day might invent a disintegrator that isn't actually one