My problem is that Australian discourse on this was really fucked up for a while. People were arguing about everything except the actual problem.
The point of AHRC mediation is precisely so that cases like this avoid going to court and nobody gets too inconvenienced if a case is meritless. problem is that The AHRC dropped the ball, making the process itself the punishment.
Contrary to what lucm said, the system didn't work. But more to the point, the system didn't work as designed.
Nobody (apart from Andrew Bolt's fanboys) thinks the law is bad. The students were legally in the right and legally vindicated. It's just that the law was administered badly, and again, Australia isn't exactly alone here.
Australia's justice system is almost as adversarial as America's, and the EFF did not turn up to court to defend itself. As such, the argument that any order was unenforceable was never presented to the court. GEMSA probably only had to show that the EFF publishes in Australia by virtue of the fact that their web site is visible in Australia.
This may seem like affrontery on behalf of the court, but the fact is they only considered one side of the argument.
Just in case it isn't clear, I think the EFF did the right thing by fighting this in a US court and not an Australian court. I'm merely pointing out that this sort of ruling is inevitable if one party doesn't defend itself, and it's not necessarily the court's fault.
All analogies break down at some point. In this case, though, the DMV, while not exactly endorsing you, is certifying that you have satisfied certain minimal competence requirements to drive.
To be fair, judging by some of the other responses, they are extremely minimal by world standards. So the analogy is probably better than I thought it was.
They are certifying that you have demonstrated a certain level of competence as a driver. That is not an endorsement but it's somewhat analogous to it.
That's a very misleading figure because it includes short-term visas for students and business people. In 2015-2016, the number of immigrants was 189,770 under the migration program plus 17,555 under the humanitarian program. Net migration is around the 178,000 mark.
But even this is beside the point. The problem isn't that we let in so many people, the problem is that we shove them into capital cities. There is no housing crisis in regional centres. But there isn't enough infrastructure either.
It's no more an endorsement of the person than a driver's license is an endorsement of them by the DMV.
You don't have to pass a driving test to get a licence in America? You learn something every day.
In my country, a driver's licence is proof not only of your identity, but also that you have demonstrated some level of knowledge of the road rules and some level of competence in operating a 2 tonne death machine in a public area.
That's not to say that Pseudonyum has not stored all of those 60 languages as abstract concepts, letting auto-complete fill in the gaps.
After about 20 or so they start to run together, it's true.
But if it helps, here's how I justify that claim.
I speak Java. I know Java well enough that I could write, say, a version 7 compiler in Java without any difficulty. What I don't know: Swing, Spring, Ant, Maven... all that stuff.
So while I can say with a good conscience "I speak Java", what I can't say with a good conscience is "I am a Java programmer".
It's also worth pointing out that a dozen of those languages are not really something you can put on a CV. For example, I can speak, and I can articulate most of the differences between, a few of the David Turner-family languages (Miranda, Miracula, Orwell, etc). All of them except Haskell are now extinct.
I agree. I happen to agree that paper ballots are superior for the foreseeable future, but the anonymity issue is not insurmountable in a technical sense. In principle.
There's been a lot of work in non-interactive zero-knowledge proofs recently that can help with the anonymity issue.It's nowhere near ready for prime-time in an election, but good enough for a second-tier cryptocurrency.
It helps that Haskell code is also less ambiguous than Python, so incorrect indentation tends to cause compile-time errors. I have seen too many Python programs where indentation was the only factor separating a correct program from one with valid syntax but incorrect runtime behavior.
Oh, yes, that's a good point. Whitespace mistakes in Haskell are almost always caught by the compiler.
If I am understanding you correctly. You say that C is simpler and faster at certain tasks.
Well I am saying that, but this isn't my central point. My central point is that efficient use of resources (whether it's real-time performance, hardware needs, power consumption, programmer time, or whatever metric is important) is always important.
I speak over 60 programming languages at last count, to varying levels of fluency. Python is not on the list. I tried to learn it, I tried to find something good about it, but I found very quickly that Python taught me nothing useful about programming, and encouraged (and in some cases enforced) poor programming habits, and I spent so much time fighting the language that I gave up.
Even Erlang never faught me. Hell, even JavaScript never faught me to the same extent that Python did.
I would not choose Python for live deployed code. I do not object to it being part of a system, as a "better Perl" or "better shell script". I wouldn't even necessarily be against it being used as the "soft layer" of a scripted system as long as the system didn't spend very much time in the soft layer. But I would be very careful to avoid Python's considerable problems.
Here are Python's main advantages:
1. There are lots of Python programmers available to hire. There is no guarantee as to their quality, but they are there.
2. There are lots of third-party Python libraries available to use. Again, there is no guarantee as to their quality, but they are there. (On the minus side, most third-party Python libraries are the enemies of robust software and kill any hope you had to use static analysis tools on your software.)
3. Programmer performance, measured as the number of deployable well-tested features produced per unit time, is not too bad. It's not even close to the top, but it's not too bad.
And... that's pretty much it, as far as I can tell.
On point #3, I suspect that a lot of Python programmers feel more productive than they actually are. In a statically-typed language, there is a phase that programmers go through where they have finished writing some code, and now need to get the compiler to accept it. This feels unproductive, in the sense that you know what you meant and now you just need to convince the compiler.
For a good type system, this is actually one of the most productive parts of programming, because you are fixing bugs that the compiler found for you. Hunting down a tricky bug feels productive in a way that fixing a compiler-checked type error does not, even though they are actually the same thing, the main difference being that the compiler found the bug instead of a customer.
Slashdot Mirror
And that makes it right ?
Nobody claimed that.
My problem is that Australian discourse on this was really fucked up for a while. People were arguing about everything except the actual problem.
The point of AHRC mediation is precisely so that cases like this avoid going to court and nobody gets too inconvenienced if a case is meritless. problem is that The AHRC dropped the ball, making the process itself the punishment.
Contrary to what lucm said, the system didn't work. But more to the point, the system didn't work as designed.
This sums up my feelings about the whole clusterfuck, except that I think the AHRC's incompetence is the problem, not 18C itself.
Like that never happens in the UK or the USA?
Nobody (apart from Andrew Bolt's fanboys) thinks the law is bad. The students were legally in the right and legally vindicated. It's just that the law was administered badly, and again, Australia isn't exactly alone here.
Never heard of it, mate. But steak sounds good.
What, you mean the Monkee or Bowie? They're both dead, mate.
Besides, they'd have to give up their citizenship. We take Section 44(i) seriously these days.
In this case, South Australian defamation law.
Australia's justice system is almost as adversarial as America's, and the EFF did not turn up to court to defend itself. As such, the argument that any order was unenforceable was never presented to the court. GEMSA probably only had to show that the EFF publishes in Australia by virtue of the fact that their web site is visible in Australia.
This may seem like affrontery on behalf of the court, but the fact is they only considered one side of the argument.
Just in case it isn't clear, I think the EFF did the right thing by fighting this in a US court and not an Australian court. I'm merely pointing out that this sort of ruling is inevitable if one party doesn't defend itself, and it's not necessarily the court's fault.
There is no dark side of the Moon.
Really. Matter of fact, it's all dark.
Do I need to bring out the chart? The chart is pretty good.
True Scotsmen against Scotsmen in this case.
UN-altered REPRODUCTION and DISSEMINATION of this IMPORTANT Information is ENCOURAGED, ESPECIALLY to COMPUTER BULLETIN BOARDS.
I doubt there are tens of thousands of Antifa accounts on Twitter, even including the fake Russian ones, let alone "spewing hate".
Forget tens of thousands, I would be surprised if there were three verified antifa accounts in existence.
All analogies break down at some point. In this case, though, the DMV, while not exactly endorsing you, is certifying that you have satisfied certain minimal competence requirements to drive.
To be fair, judging by some of the other responses, they are extremely minimal by world standards. So the analogy is probably better than I thought it was.
They are certifying that you have demonstrated a certain level of competence as a driver. That is not an endorsement but it's somewhat analogous to it.
We let in nearly 300,000 people a year, [...]
That's a very misleading figure because it includes short-term visas for students and business people. In 2015-2016, the number of immigrants was 189,770 under the migration program plus 17,555 under the humanitarian program. Net migration is around the 178,000 mark.
But even this is beside the point. The problem isn't that we let in so many people, the problem is that we shove them into capital cities. There is no housing crisis in regional centres. But there isn't enough infrastructure either.
It's no more an endorsement of the person than a driver's license is an endorsement of them by the DMV.
You don't have to pass a driving test to get a licence in America? You learn something every day.
In my country, a driver's licence is proof not only of your identity, but also that you have demonstrated some level of knowledge of the road rules and some level of competence in operating a 2 tonne death machine in a public area.
That's not to say that Pseudonyum has not stored all of those 60 languages as abstract concepts, letting auto-complete fill in the gaps.
After about 20 or so they start to run together, it's true.
But if it helps, here's how I justify that claim.
I speak Java. I know Java well enough that I could write, say, a version 7 compiler in Java without any difficulty. What I don't know: Swing, Spring, Ant, Maven... all that stuff.
So while I can say with a good conscience "I speak Java", what I can't say with a good conscience is "I am a Java programmer".
It's also worth pointing out that a dozen of those languages are not really something you can put on a CV. For example, I can speak, and I can articulate most of the differences between, a few of the David Turner-family languages (Miranda, Miracula, Orwell, etc). All of them except Haskell are now extinct.
And it's D-ah-S, not D-oh-S for your operating system.
Not in an Australian accent, it isn't.
I never used CompuServe (difficult in this country) but I thought of them recently when I debated how to pronounce "GIF".
I agree. I happen to agree that paper ballots are superior for the foreseeable future, but the anonymity issue is not insurmountable in a technical sense. In principle.
There's been a lot of work in non-interactive zero-knowledge proofs recently that can help with the anonymity issue.It's nowhere near ready for prime-time in an election, but good enough for a second-tier cryptocurrency.
It helps that Haskell code is also less ambiguous than Python, so incorrect indentation tends to cause compile-time errors. I have seen too many Python programs where indentation was the only factor separating a correct program from one with valid syntax but incorrect runtime behavior.
Oh, yes, that's a good point. Whitespace mistakes in Haskell are almost always caught by the compiler.
Yes. I still use Perl as my go-to "little script to do one thing" language, but I wouldn't condemn anyone who preferred Python for that task.
More like "a plague o' both your houses". There are a lot of things to dislike about Python, but most of the criticisms of it are deeply ignorant.
If I am understanding you correctly. You say that C is simpler and faster at certain tasks.
Well I am saying that, but this isn't my central point. My central point is that efficient use of resources (whether it's real-time performance, hardware needs, power consumption, programmer time, or whatever metric is important) is always important.
I speak over 60 programming languages at last count, to varying levels of fluency. Python is not on the list. I tried to learn it, I tried to find something good about it, but I found very quickly that Python taught me nothing useful about programming, and encouraged (and in some cases enforced) poor programming habits, and I spent so much time fighting the language that I gave up.
Even Erlang never faught me. Hell, even JavaScript never faught me to the same extent that Python did.
I would not choose Python for live deployed code. I do not object to it being part of a system, as a "better Perl" or "better shell script". I wouldn't even necessarily be against it being used as the "soft layer" of a scripted system as long as the system didn't spend very much time in the soft layer. But I would be very careful to avoid Python's considerable problems.
Here are Python's main advantages:
1. There are lots of Python programmers available to hire. There is no guarantee as to their quality, but they are there.
2. There are lots of third-party Python libraries available to use. Again, there is no guarantee as to their quality, but they are there. (On the minus side, most third-party Python libraries are the enemies of robust software and kill any hope you had to use static analysis tools on your software.)
3. Programmer performance, measured as the number of deployable well-tested features produced per unit time, is not too bad. It's not even close to the top, but it's not too bad.
And... that's pretty much it, as far as I can tell.
On point #3, I suspect that a lot of Python programmers feel more productive than they actually are. In a statically-typed language, there is a phase that programmers go through where they have finished writing some code, and now need to get the compiler to accept it. This feels unproductive, in the sense that you know what you meant and now you just need to convince the compiler.
For a good type system, this is actually one of the most productive parts of programming, because you are fixing bugs that the compiler found for you. Hunting down a tricky bug feels productive in a way that fixing a compiler-checked type error does not, even though they are actually the same thing, the main difference being that the compiler found the bug instead of a customer.
Well, an alternative fact.
We would presumably have to spend something other than a penny, yes.
I hereby patent the unblockchain.
Very true.