"The Global Hawk is the first UAV to be certified by the FAA to file its own flight plans and use civilian air corridors in the United States with no advance notice.[21] This potentially paves the way for a revolution in unmanned flight, including that of remotely piloted cargo or passenger airliners."
and
"On April 24, 2001 a Global Hawk flew non-stop from Edwards in the US to RAAF Base Edinburgh in Australia, making history by being the first pilotless aircraft to cross the Pacific Ocean. The flight took 22 hours, and set a world record for absolute distance flown by a UAV, 13,219.86 kilometers (8,214.44 mi).[27][28]"
Counting the ballots and the preference distribution is by hand. Having scrutineered booths for my party the last 5 or 6 federal elections and many state elections in two states I've some familiarity with it. There are also some details you've got slightly wrong.
- just make a "1". well that doesn't work where preference distribution is compulsory (e.g. federal elections). in that situation you must number all the squares (bar one) starting at your first preference and decreasing order of preference. So a ballot with three candidates and just one marked "1" is an informal vote in that case. Only in some state elections, and council elections, is voting optional preferential (i.e. where you can "exhaust your preferences").
- naturally the number of marked ballots is tallied with the number of marked voters, before the votes are even counted. after the votes are counted, the numbers are again tallied. in fact this often takes the longer than actually just counting the votes.
- Registration. There are certainly flaws with electoral registration in Australia. Howard's mob made some last minute measures just before they were thrashed in the last election in this regard many of which served to disenfranchise sections of the electorate. Generally however these is little registration fraud.
- If you vote multiple times, you will be caught, because the AEC checks for it (they can obviously tell if you've been marked on the roll in several different booths!). And they use a machine to check who has been marked off (have a gander at the way the worker marks the roll next time, it's made for machine reading).
But then I'd say problems with the registration system are quite distinct from problems with the ballot.
Speaking as a party member & voting scrutineer, we are not allowed to touch the ballot boxes or the ballots and can only watch the process occur. I find it totally ludicrous that in the USA party-political hacks are allowed to organise and run the vote! No wonder you get the "Florida situation". The USA needs to get the voting system out of the hands of the people who compete in the elections as its first priority.
Paper ballot systems are the only really secure voting system - provided various other provisos are met (e.g. voter registration and ballot box security). I don't understand the obsession for electronic voting at all.
I participate regularly in elections involving both compulsory and optional preferential voting (aka "instant runoff voting") and we can count all the votes and distribute the preferences and produce the numbers for a voting booth inside a couple of hours. Including arguments about "is this a valid vote" (i.e. preferences legibly distributed according to the rules for that election). Party workers cannot touch the ballots. The overall results can be easily known an hour or two later and we're all at post-election events within 3 or 4 hours of the polls closing and the general shape of the results are known.
So why you need to introduce hackable, unreliable, electronic machinery into this process is completely beyond me. In fact if they tried to introduce this into my elections I would vigourously oppose it at every turn.
A comment below asks about the disabled. That's what a pre-poll postal ballot is for. That allows inclusion for people who can't make it to the voting booth on election day as well. And they are trivial to secure.
The biggest challenge I have is finding a technically competent recruiter; many I've spoken to are fine so long as you repeat buzz words, but if you try to explain anything more complex, their eyes glaze over...
Recruiters here are extremely variable. Some are really shit and others quite good. Still, every decent *permanent* position (as opposed to a contract) I've ever got in 15 years of programming work was direct to the advertising employer and not a recruiter.
Most recruiters are nothing more than not-very-competent pattern matchers (customer says java hibernate, resume says java hibernate, is match send to customer). In my experience, avoid Hays especially. They are woeful. Try to steer to the specialist IT recruiters, but you'll still find the field extremely variable. They are rarely, if not ever, technical. But if you find one that works for you and you like contract work, stick with them as much as possible.
If you are looking for permanent work my advice is to put an automated email daily job alert on a couple of the career boards (e.g. seek, mycareer, careerone, etc), get to know all your relevant user & special interest groups, and use google to find companies that use the technology or business space you work in or would like to work in and have a look through their web site to find if they have jobs going.
Read the spec. Page 20, section 1.2 of the 2.4 specification. While Tomcat or some other containers might let you spawn threads the specification makes it abundantly clear that another container may actively prohibit it if the container vendor chooses.
Hence, threading in servlets is non-portable and not to be trusted or relied upon.
The language appears to be loosening as in previous version the warning was much more general. Also see that threads you spawn are still subject to the lifecycle of the service method on the servlet. I would not recommend generally to block client i/o on the successfull return of some threads. If I needed threading, parallel processing, I would consider using message queues to send requests to MDBs which would perform the logic independent of the servlet's lifecycle. Which is not the same thing as just spawning threads.
The servlet still runs in 'a container' so your last paragraph about the 'container doing it for you' contradicts the first.
And you can, in fact, spawn threads inside an EJB too. By which I mean the container doesn't have to stop you from doing it even though it violates the specification. I have seen it done in an application I once inherited and had to redesign. While the container (JBoss) was perfectly happy to let you do it, and it appeared to work ok at least under minimal load, the actual behaviour is non-deterministic.
At the top of the article the author says "Computers are very rational, and people are abstract;"
Programmer joke: if people are "abstract" how come I keep seeing so many instances of them. Maybe they are subclasses?
Anyway it's completely trite. And untrue. Computers are algorithmic. Humans can be rational, which is usually defined as 'capable of exercising reason'.
Unless, of course the author means rational as in mathematics, as in a rational number (i.e. a number that can be represented as a fraction). But in this definition, the author is even more wrong; computers are of course binary machines.
This is just the sort of faulty reasoning that makes me stop reading articles. Quite aside from that first sentence !!! from this single example, perhaps we can conclude (erroneously) that people aren't abstract, they are illiterate. At least in this instance.
OK, so it only costs $0.40 to make an album? You come up with a way to record an album for 40 cents and you will be hero to every musician in the land.
Listen, I agree that when you buy music you buy a licence to listen to that music no matter what the medium; e.g. on cd, off your ipod, in your car, etc. And you have a licence to edit and sequence your collection in any way you see fit, for your own personal pleasure. And I just do exactly that, the RIAA or the MPAA or ARIA hasn't bothered me.
But the manufacturing cost of the cd itself is a small part of the overall cost. It's like saying a software download time is the only expense in creating software.
Brian Wilson mixed 'Pet Sounds' in mono because he didn't think people would set up their stereos right and didn't want to leave his masterpiece sounding crap because someone didn't know how to place stereo speakers. It's a still a masterpiece. Adding surrround sound or even stereo (like on the modern cd version that has stereo and mono versions) actually adds completely nothing. In my opinion of course (on Slashdot, as if).
Surround is cool for movies where the channel use has a mostly standardised 'language' already (center=dialogue, front=music/fx, rear=fx,.1=LF effects), I find a lot of musicians will go overboard with the whacky panning tricks but in 10 years time this will seem really dated. A lot of other people are really just using it to add (or capture) room ambience to the recordings, which seems to me to not really be a killer reason to go surround. I am so far, underwhelmed.
i don't know much about law, but i do suspect that a computer can't be a 'human' intelligence. if it never had a human bodily manifestation, no matter how intelligent or self aware or conscious it may be, it isn't of a human variety. i think that machine consciousness will be radically different to the human variety.
Also maybe worrying about the architecture you are using if you are putting jdbc directly into your JSPs!!! The whole article is whack; small projects done by inexperienced staff and with no maintainability issues are the best ones to do with scripts; database access in JSPs is just silly architecture; etc. It's like saying, "I discovered that using a Ferrari to drive over a small bug on your driveway is an incredibly inefficient way to crush bugs. Therefore, Ferraris are bad cars."
"First, the optimizing compiler compiles byte code down to native assembler. Ok, no big deal here, so do the C/C++ compilers and if that were the end of the story... The interesting part comes with the realization that the native code produced by the optimizing compiler will have undergone optimizations based on the current runtime conditions, something that the C/C++ static compliers are incapable of doing. Thus, the native code that is produced by the JIT may actually be more highly optimized for the task it is performing. "
even the whole concept of a band playing and recording their own songs as a coherent unit was more or less popularised by the beatles. before them it was usually just singers singing the songs of the ir producers or others backed up by interchangeable band members.
lennon/mccartney are the mozart of the 20th century. whatever you think of their music (or mozart) in 200 years time it's them who'll be remembered as the great artists of that era (along with a handful of others) and not, say, fred durst.
It amazes me how people bitch endlessly about spam, but when a company filters it (and inevitably messes up trying) you freak out. It is not like AOL is trying to charge you for sending to someone on AOL.
Unfortunately AOL doesn't just block spam. What they are doing is tantamount to arresting every black guy in town because ONE black guy did a stick-up at the 7-11, and they are too lazy or stupid to investigate the crime properly.
They block the IP of my mail server, because it is on a dynamically-assigned broadband IP block of one of their competitors. DESPITE the fact the competitor (Telstra) is very vigilant in shutting down anyone who breaks the terms of their service agreement, (i.e. spammers).
My view is that it is simply anti-competitive behaviour. They like closed systems; that's where they came from. I would bet at least half the motivation for shutting off email from servers in the Telstra broadband networks is that AOL is simply pissed off they can't break into the Australian market as well as they would like and they simply want to inconvienience the customers of their major local competitor.
I'm not going to feign the usual "oh it's microsoft and it must be bad". Actually the guy said some interesting things about how to go about sorting out informational posts in newsgroups using the sociology of the posters. It's an interesting approach and the information presented in the interview made me want to know more about it. ((and before you say anothing I'm a Java programmer so normally I'm very MS-suspicious)).
The Netscan tool looks interesting, if that sort of toolset was built into newsgroup programs or the Google newsgroups site it would be very nice. Now, if only MS had a decent newsgroup program in the first place...
I have always said that MS is more vunerable than, say, Oracle. Replacing desktops and productivity apps occurs in 18 to 36 month cycles anyway.
On the other hand, databases, and the other critical enterprise applications which rely on them, take 18 to 36 months _to_ replace.
Your MIS is going to cast a much more critical eye over the decision to replace the enterprise layer of the business than the access-points (i.e. the computers on the worker's desks) to that enterprise layer. And that access layer is frequently nowadays nothing but a web browser over a secured network. Such as the project I am currently working on for a client, 'Business transformation project', taking the mission-critical apps - the core business - off an OS/390 mainframe onto a n-Tier J2EE based architecture using Websphere, DB2, and Webmethods. It is very much in the front of the client's minds, has been for quite a while, and will be for quite a while more. The success or otherwise is totally critical to the enterprise in a way that 'the desktop' never will be.
Oracle (and SAP, and J2EE, etc) is far more 'mission critical' than Windows. That's why MS wants to get their dirty fingers into a piece of that pie - harder to dislodge from it.
oh yeah.. at my current client site I can't get ssh access outside, and subsequently, I have to use webmail to read email at work, which is a right royal PITA.
This company is the major credit reference supplier in Aus/NZ. Identity fraud is a major issue for them, too. Here is their tips page about protecting yourself from identity fraud.
http://www.baycorpadvantage.com/personal_informa ti on/prevention.asp?CountryID=1&UserTypeID=11
I think the interesting information is "Studies show identity fraud victims typically know the person who uses, or tries to use, their identity."
Slashdot Mirror
ahhh ... Global Hawk anybody? http://en.wikipedia.org/wiki/RQ-4_Global_Hawk
"The Global Hawk is the first UAV to be certified by the FAA to file its own flight plans and use civilian air corridors in the United States with no advance notice.[21] This potentially paves the way for a revolution in unmanned flight, including that of remotely piloted cargo or passenger airliners."
and
"On April 24, 2001 a Global Hawk flew non-stop from Edwards in the US to RAAF Base Edinburgh in Australia, making history by being the first pilotless aircraft to cross the Pacific Ocean. The flight took 22 hours, and set a world record for absolute distance flown by a UAV, 13,219.86 kilometers (8,214.44 mi).[27][28]"
Counting the ballots and the preference distribution is by hand. Having scrutineered booths for my party the last 5 or 6 federal elections and many state elections in two states I've some familiarity with it. There are also some details you've got slightly wrong.
- just make a "1". well that doesn't work where preference distribution is compulsory (e.g. federal elections). in that situation you must number all the squares (bar one) starting at your first preference and decreasing order of preference. So a ballot with three candidates and just one marked "1" is an informal vote in that case. Only in some state elections, and council elections, is voting optional preferential (i.e. where you can "exhaust your preferences").
- naturally the number of marked ballots is tallied with the number of marked voters, before the votes are even counted. after the votes are counted, the numbers are again tallied. in fact this often takes the longer than actually just counting the votes.
- Registration. There are certainly flaws with electoral registration in Australia. Howard's mob made some last minute measures just before they were thrashed in the last election in this regard many of which served to disenfranchise sections of the electorate. Generally however these is little registration fraud.
- If you vote multiple times, you will be caught, because the AEC checks for it (they can obviously tell if you've been marked on the roll in several different booths!). And they use a machine to check who has been marked off (have a gander at the way the worker marks the roll next time, it's made for machine reading).
But then I'd say problems with the registration system are quite distinct from problems with the ballot.
Speaking as a party member & voting scrutineer, we are not allowed to touch the ballot boxes or the ballots and can only watch the process occur. I find it totally ludicrous that in the USA party-political hacks are allowed to organise and run the vote! No wonder you get the "Florida situation". The USA needs to get the voting system out of the hands of the people who compete in the elections as its first priority.
Paper ballot systems are the only really secure voting system - provided various other provisos are met (e.g. voter registration and ballot box security). I don't understand the obsession for electronic voting at all.
I participate regularly in elections involving both compulsory and optional preferential voting (aka "instant runoff voting") and we can count all the votes and distribute the preferences and produce the numbers for a voting booth inside a couple of hours. Including arguments about "is this a valid vote" (i.e. preferences legibly distributed according to the rules for that election). Party workers cannot touch the ballots. The overall results can be easily known an hour or two later and we're all at post-election events within 3 or 4 hours of the polls closing and the general shape of the results are known.
So why you need to introduce hackable, unreliable, electronic machinery into this process is completely beyond me. In fact if they tried to introduce this into my elections I would vigourously oppose it at every turn.
A comment below asks about the disabled. That's what a pre-poll postal ballot is for. That allows inclusion for people who can't make it to the voting booth on election day as well. And they are trivial to secure.
The biggest challenge I have is finding a technically competent recruiter; many I've spoken to are fine so long as you repeat buzz words, but if you try to explain anything more complex, their eyes glaze over...
Recruiters here are extremely variable. Some are really shit and others quite good. Still, every decent *permanent* position (as opposed to a contract) I've ever got in 15 years of programming work was direct to the advertising employer and not a recruiter.Most recruiters are nothing more than not-very-competent pattern matchers (customer says java hibernate, resume says java hibernate, is match send to customer). In my experience, avoid Hays especially. They are woeful. Try to steer to the specialist IT recruiters, but you'll still find the field extremely variable. They are rarely, if not ever, technical. But if you find one that works for you and you like contract work, stick with them as much as possible.
If you are looking for permanent work my advice is to put an automated email daily job alert on a couple of the career boards (e.g. seek, mycareer, careerone, etc), get to know all your relevant user & special interest groups, and use google to find companies that use the technology or business space you work in or would like to work in and have a look through their web site to find if they have jobs going.
Read the spec. Page 20, section 1.2 of the 2.4 specification. While Tomcat or some other containers might let you spawn threads the specification makes it abundantly clear that another container may actively prohibit it if the container vendor chooses.
Hence, threading in servlets is non-portable and not to be trusted or relied upon.
The language appears to be loosening as in previous version the warning was much more general. Also see that threads you spawn are still subject to the lifecycle of the service method on the servlet. I would not recommend generally to block client i/o on the successfull return of some threads. If I needed threading, parallel processing, I would consider using message queues to send requests to MDBs which would perform the logic independent of the servlet's lifecycle. Which is not the same thing as just spawning threads.
The servlet still runs in 'a container' so your last paragraph about the 'container doing it for you' contradicts the first.
And you can, in fact, spawn threads inside an EJB too. By which I mean the container doesn't have to stop you from doing it even though it violates the specification. I have seen it done in an application I once inherited and had to redesign. While the container (JBoss) was perfectly happy to let you do it, and it appeared to work ok at least under minimal load, the actual behaviour is non-deterministic.
With a running app, ... you can spawn threads, etc.
Not in J2EE you can't. Read the servlet spec.
Otherwise I'd agree with you.
Well I stopped reading after I got over the bit about computers being "rational".
At the top of the article the author says "Computers are very rational, and people are abstract;"
Programmer joke: if people are "abstract" how come I keep seeing so many instances of them. Maybe they are subclasses?
Anyway it's completely trite. And untrue. Computers are algorithmic. Humans can be rational, which is usually defined as 'capable of exercising reason'.
Unless, of course the author means rational as in mathematics, as in a rational number (i.e. a number that can be represented as a fraction). But in this definition, the author is even more wrong; computers are of course binary machines.
This is just the sort of faulty reasoning that makes me stop reading articles. Quite aside from that first sentence !!! from this single example, perhaps we can conclude (erroneously) that people aren't abstract, they are illiterate. At least in this instance.
OK, so it only costs $0.40 to make an album? You come up with a way to record an album for 40 cents and you will be hero to every musician in the land.
Listen, I agree that when you buy music you buy a licence to listen to that music no matter what the medium; e.g. on cd, off your ipod, in your car, etc. And you have a licence to edit and sequence your collection in any way you see fit, for your own personal pleasure. And I just do exactly that, the RIAA or the MPAA or ARIA hasn't bothered me.
But the manufacturing cost of the cd itself is a small part of the overall cost. It's like saying a software download time is the only expense in creating software.
RFC 1149 - Standard for the transmission of IP datagrams on avian carriers.
David Waitzman 1 April 1990.
RFC 1149 from google's cache
Brian Wilson mixed 'Pet Sounds' in mono because he didn't think people would set up their stereos right and didn't want to leave his masterpiece sounding crap because someone didn't know how to place stereo speakers. It's a still a masterpiece. Adding surrround sound or even stereo (like on the modern cd version that has stereo and mono versions) actually adds completely nothing. In my opinion of course (on Slashdot, as if).
.1=LF effects), I find a lot of musicians will go overboard with the whacky panning tricks but in 10 years time this will seem really dated. A lot of other people are really just using it to add (or capture) room ambience to the recordings, which seems to me to not really be a killer reason to go surround. I am so far, underwhelmed.
Surround is cool for movies where the channel use has a mostly standardised 'language' already (center=dialogue, front=music/fx, rear=fx,
The CIA World Factbook. Published every year and online.
may be what's needed is an interface which has it's controls built into the arms of your chair, not hovering out in front of you somewhere.
i don't know much about law, but i do suspect that a computer can't be a 'human' intelligence. if it never had a human bodily manifestation, no matter how intelligent or self aware or conscious it may be, it isn't of a human variety. i think that machine consciousness will be radically different to the human variety.
Also maybe worrying about the architecture you are using if you are putting jdbc directly into your JSPs!!! The whole article is whack; small projects done by inexperienced staff and with no maintainability issues are the best ones to do with scripts; database access in JSPs is just silly architecture; etc. It's like saying, "I discovered that using a Ferrari to drive over a small bug on your driveway is an incredibly inefficient way to crush bugs. Therefore, Ferraris are bad cars."
Java is not necessarily slower than other languages. It's true for Java 1.1, but most Java 1.3 + 1.4 implementations are way faster than that.
u p0 33.shtml
In fact, when used with a JIT compiler, it can be faster than an optimized C++ program. To quote from a roundup at javaperformancetuning.com;
http://www.javaperformancetuning.com/news/round
"First, the optimizing compiler compiles byte code down to native assembler. Ok, no big deal here, so do the C/C++ compilers and if that were the end of the story... The interesting part comes with the realization that the native code produced by the optimizing compiler will have undergone optimizations based on the current runtime conditions, something that the C/C++ static compliers are incapable of doing. Thus, the native code that is produced by the JIT may actually be more highly optimized for the task it is performing. "
regs,
even the whole concept of a band playing and recording their own songs as a coherent unit was more or less popularised by the beatles. before them it was usually just singers singing the songs of the ir producers or others backed up by interchangeable band members.
lennon/mccartney are the mozart of the 20th century. whatever you think of their music (or mozart) in 200 years time it's them who'll be remembered as the great artists of that era (along with a handful of others) and not, say, fred durst.
Well, OK, in their usual state of torpor but they will shut down people who use broadband to send spam.
AOL is blocking all dynamically assigned IP blocks without concern whether there is spam coming from then or whether the ISP is 'responsible' or not.
Unfortunately AOL doesn't just block spam. What they are doing is tantamount to arresting every black guy in town because ONE black guy did a stick-up at the 7-11, and they are too lazy or stupid to investigate the crime properly.
They block the IP of my mail server, because it is on a dynamically-assigned broadband IP block of one of their competitors. DESPITE the fact the competitor (Telstra) is very vigilant in shutting down anyone who breaks the terms of their service agreement, (i.e. spammers).
My view is that it is simply anti-competitive behaviour. They like closed systems; that's where they came from. I would bet at least half the motivation for shutting off email from servers in the Telstra broadband networks is that AOL is simply pissed off they can't break into the Australian market as well as they would like and they simply want to inconvienience the customers of their major local competitor.
I'm not going to feign the usual "oh it's microsoft and it must be bad". Actually the guy said some interesting things about how to go about sorting out informational posts in newsgroups using the sociology of the posters. It's an interesting approach and the information presented in the interview made me want to know more about it. ((and before you say anothing I'm a Java programmer so normally I'm very MS-suspicious)).
...
The Netscan tool looks interesting, if that sort of toolset was built into newsgroup programs or the Google newsgroups site it would be very nice. Now, if only MS had a decent newsgroup program in the first place
I have always said that MS is more vunerable than, say, Oracle. Replacing desktops and productivity apps occurs in 18 to 36 month cycles anyway.
On the other hand, databases, and the other critical enterprise applications which rely on them, take 18 to 36 months _to_ replace.
Your MIS is going to cast a much more critical eye over the decision to replace the enterprise layer of the business than the access-points (i.e. the computers on the worker's desks) to that enterprise layer. And that access layer is frequently nowadays nothing but a web browser over a secured network. Such as the project I am currently working on for a client, 'Business transformation project', taking the mission-critical apps - the core business - off an OS/390 mainframe onto a n-Tier J2EE based architecture using Websphere, DB2, and Webmethods. It is very much in the front of the client's minds, has been for quite a while, and will be for quite a while more. The success or otherwise is totally critical to the enterprise in a way that 'the desktop' never will be.
Oracle (and SAP, and J2EE, etc) is far more 'mission critical' than Windows. That's why MS wants to get their dirty fingers into a piece of that pie - harder to dislodge from it.
oh yeah .. at my current client site I can't get ssh access outside, and subsequently, I have to use webmail to read email at work, which is a right royal PITA.
Here it is again as a hyperlink;
Baycorp Advantage
This company is the major credit reference supplier in Aus/NZ. Identity fraud is a major issue for them, too. Here is their tips page about protecting yourself from identity fraud.
a ti on/prevention.asp?CountryID=1&UserTypeID=11
http://www.baycorpadvantage.com/personal_inform
I think the interesting information is "Studies show identity fraud victims typically know the person who uses, or tries to use, their identity."