Slashdot Mirror


User: Hal_Porter

Hal_Porter's activity in the archive.

Stories
0
Comments
8,852
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 8,852

  1. Re:Twisting words to make Iran appear a threat on Iran Launches Payload into Space · · Score: 1

    Your comments about Iranian politicians needing to make sure they watch what they say rather reminds me of the recent comments of a Muslim cleric here in Australia who said that women in revealing clothing were like meat left outside with cats prowling. He tried to explain it away by saying he was just referring to the risk, but virtually everyone took his comments as callously blaming rape victims for their attacker's crime.

    Well rape isn't wrong if you're an Islamic fundamentalist, look at the rules on slave girls -

    http://www.answering-islam.de/Main/Authors/Arlands on/women_slaves.htm

    Islamic morality only applies to Muslims, it specifically excludes anyone else. Which is something to bear in mind when dealing with Islamic politicians. Most people think rape is wrong no matter the religion of the parties involved, the same with slavery.

    Apart from the moral situation, the rapist and rape victim are both embedded in a legal system in which the rapist is in the wrong no matter what the victim is wearing. The police will catch him and lock him up.

    But between countries there isn't really any legal system and certainly no police. Each country thus operates in a odd amoral environment, similar to the prisoner's dilemma. So not only is aggression not punished, it's not in itself wrong. E.g. consider Israel's attack on Osirak. Given Saddam later behavior, including an unprovoked attack on Israel during the first Gulf War, most people consider it to not be unreasonable. There are countless historical situations where preemptive attack would have been justified.

    So to sum up making it rational for your opponent to attack you now by making him think you will attack him when you get nuclear weapons later (the Iran case) is completely different from wearing a short skirt and giving some pervert a bogus justification for molesting you (the rape victim one).

    Yes, all politicians should think about what they say, but when it comes to the question of possible US-Israeli aggression against Iran, it would be wrong to suggest in any way that the victim (or even the leaders of the victims) are to blame in a significant sense. I remind you again that Iran does not have nuclear weapons, and the Supreme Leader has issued a solemn fatwa condemning such weapons.

    Well for an oil rich country, they sure seem keen on refining nuclear material.

    And you realize that he doesn't have to tell the truth, don't you? Muslims can lie all they want to unbelievers.

    http://www.answering-islam.de/Main////Gilchrist/Mu hammad/enemies.html

    I realize anyone can lie, but no other religion codifies the importance of deceiving your enemies.

  2. Re:Twisting words to make Iran appear a threat on Iran Launches Payload into Space · · Score: 1

    I think it's inevitable that Iran will get nukes. And given that, Iranian politicians needs to be careful about what they say. Because if they aren't, Israel might decide that a preemptive attack is the only way they can stop them.

    The same applies to the US of course. If US politicians regularly hinted they were planning to nuke some other nuclear power, and the US got attacked they would be mostly responsible. And by and large, the US establishment seems to understand this concept. People that advocated using nuclear weapons in other conflicts like Korea and Vietnam were disowned pretty quickly.

    Basically, once you enter the grown up world of having moderately effective weapons, trash talk about your potential opponents is not a very good idea. It doesn't matter if you're talking about the US, Israel or Iran, the same thing applies. It's the reason that the US/USSR has doctrines of Strategic Ambiguity on things that were hard to answer. At the very least, it's important not to allow your opponent a window where a preemptive attack looks like a good option.

    Now quite possibly subtlety like this is beyond a rabid little theocracy like Iran, in which case they probably won't survive a transition to being a nuclear power. But that's kind of shame, since most of the expat Iranians I've met seem like ok people. Incidentally, for all your comments about Iran being a democracy, none of them thought it was anything like one. Even if it were, it doesn't change the fact that Ahmadinejad says some things which Israel might react to a way that's disasterous for Iran.

  3. Re:Iranian HIV prevention: better than cure ? on Iran Launches Payload into Space · · Score: 1

    Well if Ahmadinejad isn't speaking for Iran, why does the Supreme Leader not just point this out publically. It's quite possible that what you say is true, and Ahmadinejad is just playing to his home base. But if you were Israel, how would you know? Maybe the whole thing is complicated triple bluff and Iran really will nuke you once they can. In which case you should get them first.

    In any case, the fact that the Iranian establishment let Ahmadinejad say the things he keeps saying is incredibly irresponsible, given that they don't have diplomatic relations or any back channel to reassure Israel that he's not speaking for Iran. It seems like they have got so used to being impotent militarily that they don't understand the implications that having nuclear weapons has on what you can let people say publically.

  4. Re:Iranian HIV prevention: better than cure ? on Iran Launches Payload into Space · · Score: 1

    Actually, it wouldn't. Iran doesn't have enough nukes to totally destroy Israel on a first strike. What's more likely is that they'd make a lot of noise about lauching without actually doing it, and trigger an Israeli strike. But the Israelis are unlikely to wipe out Israel, because that would be genocide which they being founded by genocide survivors won't do. So it's more likely that Iran would use the limited arsenal they have and Israel would make a limited response, or possibly preempt them. Probably at some point it would turn back into a conventional war.

    Lots of people would die, but it wouldn't make the region more peaceful. More likely, it would cause the current cold war between Israel and Iran to heat up.

  5. Re:Why wouldn't they? on Old Islamic Tile Patterns Show Modern Math Insight · · Score: 1

    Did you read about this is Islam for Dhimmis?

  6. Re:The Catholic Church happened. on Old Islamic Tile Patterns Show Modern Math Insight · · Score: 1

    Don't you mean "The same year Colonia was discovered by Cristobal Colon".

  7. Re:Why wouldn't they? on Old Islamic Tile Patterns Show Modern Math Insight · · Score: 1

    It's a relative thing. Medieval Islamic societies had more scientists than medieval societies in the west. But that's like saying you have more scientists than the Taliban, quite literally. It's important to note that medieval islam was by no means enlightened by todays's standards. It was still possible in those societies to get killed for blasphemy, or just for displeasing the Caliph. Then in the west the enlightenment totally reformatted society - it was suddenly possible for people to ignore religion without fear of getting killed. Nothing like that happened in the Islamic world, in fact the fundamentalists and secular dictators have managed to roll back what civilisation existed there. And the roots of the enlightenment, a powerful middle class that demands a law based society still don't exist in the Islamic world now.

    You can see in Iraq for example, that there isn't sufficient domestic pressure to sustain a liberal democracy after the US/UK leave. It will go back to being a tyranny, and most people will see that as being preferable to civil war.

  8. Re:Far more software layers exist today on 4 GB May Be Vista's RAM Sweet Spot · · Score: 1

    Except the some cases are things like using an array, or calling a method, or doing any math. And if you compare Eclipse and Visual Studio, or uTorrent and Azureus you can see that Java is slow even outside the lab.

  9. Re:Bust the buster? on Ex-judge Gets 27 Months on Evidence From Hacked PC · · Score: 1

    I dunno, TV tells me it's not as clear cut as that. For example, if law enforcement had been left to Commisioner Gordon, The Joker would have been able to poison Gotham City. Luckily, Batman performed an extra-Judicial search of his hideout in the abandoned cement factory, summary execution on some of his henchmen, assault and kidnapping on the Joker, and left him trussed up outside police headquarters.

    Or in 24, if Jack Bauer weren't there to torture the terrorists until they told him where the nuclear weapons were hidden, many innocent people would die.

  10. Re:OS X Intel? on Visual Basic on GNU/Linux · · Score: 1

    One extra pipeline stage to do the translation from x86 into one or more uops? At least, that's how Jazelle is. Modern x86s are more superpipelined than Arm, so maybe a couple of extra stages. So a mispredicted branch is more costly, but then again you have industrial strength branch prediction in a modern x86 anyway. Probably compared to the superscalar logic it's not too significant. And most of the die space is used by the cache anyway, so increasing the core size is probably not an issue.

    Someone asked the Athlon 64 team leader, and he said something like "I don't know, but it's worth it"

  11. Re:Tag suggestion... on Scientists Make Quantum Encryption Breakthrough · · Score: -1, Offtopic

    proofreadyourfuckingheadlines

  12. Re:Far more software layers exist today on 4 GB May Be Vista's RAM Sweet Spot · · Score: 1

    There was a paper about this a few years ago, which I can't find now. They showed some math code in both C/C++ and and Java. There were a bunch of things that C compilers can do, like inlining, common subexpression elmination and so on that you can't do in Java because the language mandates dynamic binding, array bounds checking and so on.

    There are cases in C++, where functions are non virtual and all the math is on built in types and where the compiler can generate really good assembler code. Java is higher level, and at that point Java native compilers couldn't do those sort of things because the high level features of the language get in the way.

    I think it's a really hard problem actually. If you look at the heroic effort Arm have put into Java performance - adding support for some Java instructions natively, profiling and generating native code and then adding instructions to Thumb2 so that the native code can do bounds checking with minimal penalty, and the fact that computationally intensive code is still slower in Java, you start to see that the additional features that make Java code safer have a significant cost in terms of performance.

  13. Re:Better link on UK's Blair Dismisses Online Anti ID-Card Petition · · Score: 1

    How does that work for online purchases? (Or mail order, telephone orders, etc)

    They hold the ID card up to the camera on their 3G mobiles. No seriously, they don't show a card unless they're in a store. They have secret numbers on the back of the card, but there isn't an ID card check. Mind you, if you shop online they'll probably ask for a personal number at some point, and that is a key into the national ID database. Of course, if someone steals your wallet, they can use your card online. Hopefully there's some check on delivery addresses though.

    But no security is 100% foolproof. That doesn't mean you should give up on the idea. PIN number + ID card should cut down on unskilled credit card fraud.

    Oh please. The UK dealt with a terrorist threat for decades without having to resort to ID cards, and the July 7th bombers in London were all British citizens.

    We had internment though, and shoot to kill too, so I guess you're happy with those. Actually the UK did have an ID card system in WWII.

    But I can't take Islamic terrorism seriously, those guys have managed to kill a pathetically small number of people in the US/UK since 9/11. It's more illegal immigrants and criminals I'm concerned about. It's not good if the UK has a reputation as being the easiest place in Europe to survive in the grey economy.

  14. Re:Far more software layers exist today on 4 GB May Be Vista's RAM Sweet Spot · · Score: 1
    But kernel32.dll and user32.dll are usually really thin wrappers around a syscall to kernel mode

    e.g.
    http://www.microsoft.com/msj/archive/S413.aspx

    In KERNEL32.DLL, the code for PulseEvent begins like this:
     

    PulseEvent proc
    PUSH 00
    PUSH DWORD PTR [ESP+08]
    CALL DWORD PTR [NtPulseEvent]
     
    All KERNEL32.DLL does is grab the single parameter off the stack and pass it as a parameter to an NTDLL.DLL function. In NTDLL.DLL, the code for NtPulseEvent looks like this:

     

    NtPulseEvent proc
    MOV EAX,0000005C
    LEA EDX,[ESP+04]
    INT 2E
    RET 0008
     


    Pretty much all of Win32 is in Win32k.sys the kernel mode driver.
  15. Re:Far more software layers exist today on 4 GB May Be Vista's RAM Sweet Spot · · Score: 1

    Today a program has far more software layers than the programs of yesteryear. A .NET or Java procedure call, for example, has to go through the virtual machine, the virtual machine through the native DLLs, the native DLLs through Win32, Win32 through the NTOS translation layer, the NTOS layer through the drivers, the drivers through the kernel, the kernel through the HAL, the HAL through the drivers again (the vxd part)


    The HAL sits on the hardware directly, there isn't anything underneath it. And most of the time Win32 GUI stuff is just a trap into kernel mode. File operations are pipelined by the IO manager. There are no VxDs. Though in a strange kind of way, future hypervisor OSs will have something analogous to VxD code running so that multiple OSs can believe they are in control of shared hardware.

    But Java and .Net are serious overhead compared to running the code natively. Even if you could JIT the code to native code that was as efficient as C++ which you can't yet, you still need load the JIT and a load of libraries in memory.

    Why is Windows 3.1 so fast? because the program calls do not have to go through more than 2 layers of APIs. Win32 in Windows 3.1 is directly executed on the hardware.


    Not really. Any file access will go through the Windows API (note not Win32) into the kernel. Lots of things require calls into Dos or the Bios. Every time Dos, the Bios or the Windows kernel does any IO access or POPF, an exception will happen and VxD code will run. The VxD code is there to virtualize hardware, so that several Dos applications, or Windows ones using the Dos and Bios can believe they are in control of hardware which is in fact shared between them.

    Plus the processor is running in 16 bit protected mode and doing lots of loads to segment registers which are painfully slow. E.g. anything bigger than 64K like a bitmap accessed by C code probably needs a segment register load before every memory load which is a huge overhead. The GDI code to access the screen is hacked to stop this effect, but any other bitmap access by a third party app will likely be very slow indeed.

    If you have enough Ram to stop it thrashing, an NT based OS will be faster at IO, and faster at multimedia stuff.
  16. Re:Better link on UK's Blair Dismisses Online Anti ID-Card Petition · · Score: 1

    You'll need a biometric passport to get a visa free trip to the US, and people have to show an id card when they use a credit card in Sweden. There's no reason that credit card companies couldn't do the same thing if they wanted to in the UK if the ID cards could be made compulsory.

    Seriously, I don't see the problem. Having an ID which everyone has is common in most countries, and allows businesses to check who they are dealing with.

    It also makes it hard to survive in the country if you are an illegal immigrant. The sloppy way the UK worked, where it's perfectly possible to enter the country illegally and survive untracably in the grey economy wasn't very good back before Islamic terrorism, and it's a total liability now. I remember sitting on a train in Sweden listening to two Nigerians talking about how much easier it was to work illegally in the UK than in Sweden since the UK doesn't have an ID card system. But those guys are by no means the worst case illegal immigrant in these post 9/11 days. The idea behind ID cards is to make it impossible for people to live in this part of society. It works pretty well in Sweden - you'd be hard pressed to survive their without an ID number.

    Nice touch quoting Gilligan by the way. So the government have sexed up the "£1.7 billion" figure eh? Not too convinced by his article myself. Still, it's good to see he takes the trouble to find people who agree with his conspiracy theories these days, rather than lying about what his anonymous sources have said. That's almost real journalism.

  17. Re:Pig parts? on Regrowing Lost Body Parts Getting Closer All the Time · · Score: 1

    Richard Dawkins isn't a cult leader, he's the One True God ;-)

  18. Re:On a general level... on How Jobs Played Hardball In iPhone Birth · · Score: 2, Interesting

    But what about people that only want to use free and open source software? How are they supposed to play DirectX games? Unless Microsoft are forced to document their APIs in a manner acceptable to the Free Software movement (i.e. no fees or NDAs), how are other OSs supposed to implement them?

  19. Re:what certain scenarios? on OLPC Has Kill-Switch Theft Deterrent · · Score: 1

    They're shipping them prepaid to countries where the authorities are highly corrupt and the laptops are highly in demand in the US. You'd expect criminals in those countries to buy the shipments off the authorities and sell them on ebay for as much as possible back in the US. If they have a kill switch it should discourage this I guess, assuming someone high up in the country gives a shit about what's happened.

  20. Re:That crap in Suse 10.1 sucks monkey nuts on The Future of Packaging Software in Linux · · Score: 1

    Gentoo will never be mainstream until it supports a user friendly installation method like Vista's MSI files.

  21. Re:Please take care of Linus on Godwin's Law Invoked in Linus/Gnome Spat · · Score: 1

    So I don't have autism?! Then why am I so odd? Shit.

    <Sits rocking in his basement>

  22. Re:An Old Canard . . . on Stallman Convinces Cuba to Switch to Open Source · · Score: 1

    How do you know all this stuff is true though? In the cold war the CIA believed that the GDR had a larger economy than France. Maybe they're overly concerned about not underestimating their enemies, like with WMD. Certainly, once the cold war ended, all Eastern Block countries turned out to be in a much weaker state than everyone expected. It's not too hard for a closed country to fool outsiders, just make sure you only invite people who are sympathetic, and only show them the best hospitals. Seriously ill people can be forceably locked up too, like they do with AIDS patients. Most importantly, if there are any problems with healthcare, ordinary people can't do anything about them.

  23. Re:further development of an existing technology on Camera Phones Read Hidden Messages in Print · · Score: 3, Informative

    It's called QR code

    http://en.wikipedia.org/wiki/QR_Code

    The idea is that you print them on business cards, and people can scan your name and phone number into their phone quickly. Kind of useful in Japan where you end up with piles of business cards quite quickly.

  24. Re:Sudoku: The np-easy version of Traveling Salesm on Quantum Computer Demoed, Plays Sudoku · · Score: 1

    No, no. A true Sith Master would write an ActiveX control using ATL and script it from Visual Basic for Applications. It's cross platform you see, it can run from Word AND Excel.

    I have to admit, I kind of like ATL. Writing controls that can be used from VBA is an exercise in masochism though.

  25. Re:obligatory on Quantum Computer Demoed, Plays Sudoku · · Score: 1

    No, and it probably won't run any existing OS or language.

    You can load an N bit register with all the possible values of that register at the same time, and then do operations on all the possible values in parallel. But the operations would presumably not be anything like the bitwise operations in a classical computer.

    Here's an example

    http://en.wikipedia.org/wiki/Shor's_algorithm#Quan tum_part:_Period-finding_subroutine