IANAL either, but I don't think you have to be one to see why the court ruled the way it did.
In the opinion, the judges cite Michigan law, which states that (a) "child sexually abusive material" includes [a long list of things, as well as] reproductions or copies of photos, videos, etc., and (b) it is a crime to make "child sexually abusive material."
So... since the defendant was making reproductions of the material, he was, under Michigan law, "making child sexually abusive material."
In another form, the reasoning goes like this:
Brian Lee Hill makes reproductions of child sexually abusive material.
Under Michigan law, reproductions of child sexually abusive material are child sexually abusive material.
Therefore, Brian Lee Hill makes child sexually abusive material.
The legislature is apparently stuck in the world of physical artifacts--they assume you have to make something you can hold in order for you to 'make' something. They ignore the question of incidental copying. But wasn't Hill making reproductions of the material when he downloaded it through memory onto his hard drive? Didn't he make more copies when he caused the data to be copied to his computer's video card memory? And what about his ISP? Didn't they, too, make reproductions of the material in their routers? If I were a Michigan ISP, I'd talk to my state representative to get this sorted out. Quickly. Unless they're protected by some common carrier law, they might be guilty also.
This is more of a matter of formatting posts after they've been edited, but how about putting the year after the date in the "Posted by" line of an article? Especially when the article is from a previous year.
If the year isn't included because of space limitations, maybe the year could replace the hour & minute after a few days or months.
Yes, I know you can look at an article's URL to get the year. But is that an interface idea you'd like to see gain popularity?
Also, speaking of years, how about publish dates for books reviewed on Slashdot?
...and for the people who buy it. See the example from CNET of an HP memory controller. (http://news.com.com/2300-1006_3-5887476-4.html) The artwork flaked off and shorted out other parts of the chip. Note, in particular, how small the art details are compared to the line widths on the chip. There's a reason they make the lines that wide, and not thinner--because they won't stay put if they're thinner.
With the cost of developing a chip as high as it is, these little attempts at humor can be very expensive. When this article re-appears in 2007, let's hope all the examples are the same old ones.
It's permitted to access robots.txt, and once you do you can figure there's a website there.
Can't you tell there's a web server there just by trying to open a TCP connection on port 80? If so, there's no need to download any files.
In any case, I was figuring that Netcraft would use spiders to populate their Big List of Hostnames. One way to find out if there exists a hostname "xyz.acme.org" is to go to www.acme.org and spider that site for hostnames. That's where I was figuring someone might have ignored the rules in a robots.txt or two.
I'm willing to be disabused of this assumption with an explanation of how they come up with the hostnames list.
...basically it amounts to "EU and UN say 'Give us the root servers"
Outside of a small, relatively powerless group of our fellow geeks in Europe, I doubt that many people are concerned with DNS.
It's more likely that the parties behind the challenge are interested in control of content. Consider who's leading the charge: the EU, which is just plain pissed at the US (and from where there have been several court cases challenging content on US servers that violated EU/EU-member laws,) Brazil, which is pissed at ICANN, and Iran, Pakistan, and China, who are pissed at free speech.
The Bush administration (and maybe Clinton's, too) missed the chance to spin-off Internet governance to a strong, independent private organization. Now the struggle for control is between a state and an international organization. The lone state is at a disadvantage anytime that's the way a bureaucratic conflict is structured.
However, if even 10% of this revenue went to chip production, at $40 per chip we are looking at 90 million chips, give or take. Did they ship this many? Perhaps.
IANAA (I am not an accountant) but...
A quick peek at Intel's recent 10Q SEC filing tells us that they spent about $15B on "cost of sales" in the last year. That is close to half of net revenue. "Cost of sales" are expenses directly related to production. That includes only expensable items that go into making things (silicon, electricity, salary, Intel Inside stickers, etc.) It does not include capital investment (really, really clean buildings with fancy ovens & stuff in them) or R&D or marketing or general & administrative expenses.
Now, assuming that the money that Intel spends on things unrelated to chip manufacturing is a relatively small share of total expenses...
If the $40 figure is a true average, then Intel made about 350M chips in the last year. Worldwide, there were about 50 million PCs shipped in the last quarter. Round that off to 200M for the year. If Intel made 350M chips, then some of them were probably something other than CPUs.
What's Intel's market share? Something like 80%? It's a little higher in notebook PCs, but still they shouldn't have been making more than 160M CPUs in the last 12 months. Ignoring all the other stuff Intel makes, that's closer to $100 per chip, on average.
In any case, averaging across all chip production would severely underestimate the marginal cost for a chip like a Pentium 4. And if the $40 figure really does apply to Pentium 4s, then there's a large chunk of the company's expenses unexplained by CPU production. Those integrated video chips they put on cheap PC motherboards must be tough to make! Or could the cost of making CPUs really represent only 40% of Intel's expenses? ($40 x 160M)/$15B = 43%
Bottom line: you'd have to know an awful lot about the innards of Intel to correctly allocate costs across all the stuff they make. I don't know how you could have confidence in any number an outsider calculated for average cost-per-chip.
Intel's R&D expenses, by the way, are about 30% of their 'cost of sales' expenses--somewhere in the neighborhood of $5B in the last year. That is, about $1 for each $6 or $7 of net revenue goes to R&D.
Slashdot Mirror
Is anyone besides me seeing an ad for Adobe Acrobat 8 on the page for this story?
I'm suing for wrongful admission. Someone with valid credentials would have sent me packing.
IHTFpdf
In the opinion, the judges cite Michigan law, which states that (a) "child sexually abusive material" includes [a long list of things, as well as] reproductions or copies of photos, videos, etc., and (b) it is a crime to make "child sexually abusive material."
So... since the defendant was making reproductions of the material, he was, under Michigan law, "making child sexually abusive material."
In another form, the reasoning goes like this:
The legislature is apparently stuck in the world of physical artifacts--they assume you have to make something you can hold in order for you to 'make' something. They ignore the question of incidental copying. But wasn't Hill making reproductions of the material when he downloaded it through memory onto his hard drive? Didn't he make more copies when he caused the data to be copied to his computer's video card memory? And what about his ISP? Didn't they, too, make reproductions of the material in their routers? If I were a Michigan ISP, I'd talk to my state representative to get this sorted out. Quickly. Unless they're protected by some common carrier law, they might be guilty also.
Now, how do I set the preferences for book reviews?
If the year isn't included because of space limitations, maybe the year could replace the hour & minute after a few days or months.
Yes, I know you can look at an article's URL to get the year. But is that an interface idea you'd like to see gain popularity?
Also, speaking of years, how about publish dates for books reviewed on Slashdot?
With the cost of developing a chip as high as it is, these little attempts at humor can be very expensive. When this article re-appears in 2007, let's hope all the examples are the same old ones.
Can't you tell there's a web server there just by trying to open a TCP connection on port 80? If so, there's no need to download any files.
In any case, I was figuring that Netcraft would use spiders to populate their Big List of Hostnames. One way to find out if there exists a hostname "xyz.acme.org" is to go to www.acme.org and spider that site for hostnames. That's where I was figuring someone might have ignored the rules in a robots.txt or two.
I'm willing to be disabused of this assumption with an explanation of how they come up with the hostnames list.
That's the one.
How many?
I'd like to know--how many spiders.txt files do you have to ignore to come up with numbers like this?
Outside of a small, relatively powerless group of our fellow geeks in Europe, I doubt that many people are concerned with DNS.
It's more likely that the parties behind the challenge are interested in control of content. Consider who's leading the charge: the EU, which is just plain pissed at the US (and from where there have been several court cases challenging content on US servers that violated EU/EU-member laws,) Brazil, which is pissed at ICANN, and Iran, Pakistan, and China, who are pissed at free speech.
The Bush administration (and maybe Clinton's, too) missed the chance to spin-off Internet governance to a strong, independent private organization. Now the struggle for control is between a state and an international organization. The lone state is at a disadvantage anytime that's the way a bureaucratic conflict is structured.
IANAA (I am not an accountant) but...
A quick peek at Intel's recent 10Q SEC filing tells us that they spent about $15B on "cost of sales" in the last year. That is close to half of net revenue. "Cost of sales" are expenses directly related to production. That includes only expensable items that go into making things (silicon, electricity, salary, Intel Inside stickers, etc.) It does not include capital investment (really, really clean buildings with fancy ovens & stuff in them) or R&D or marketing or general & administrative expenses.
Now, assuming that the money that Intel spends on things unrelated to chip manufacturing is a relatively small share of total expenses... If the $40 figure is a true average, then Intel made about 350M chips in the last year. Worldwide, there were about 50 million PCs shipped in the last quarter. Round that off to 200M for the year. If Intel made 350M chips, then some of them were probably something other than CPUs.
What's Intel's market share? Something like 80%? It's a little higher in notebook PCs, but still they shouldn't have been making more than 160M CPUs in the last 12 months. Ignoring all the other stuff Intel makes, that's closer to $100 per chip, on average.
In any case, averaging across all chip production would severely underestimate the marginal cost for a chip like a Pentium 4. And if the $40 figure really does apply to Pentium 4s, then there's a large chunk of the company's expenses unexplained by CPU production. Those integrated video chips they put on cheap PC motherboards must be tough to make! Or could the cost of making CPUs really represent only 40% of Intel's expenses? ($40 x 160M)/$15B = 43%
Bottom line: you'd have to know an awful lot about the innards of Intel to correctly allocate costs across all the stuff they make. I don't know how you could have confidence in any number an outsider calculated for average cost-per-chip.
Intel's R&D expenses, by the way, are about 30% of their 'cost of sales' expenses--somewhere in the neighborhood of $5B in the last year. That is, about $1 for each $6 or $7 of net revenue goes to R&D.
FYI, here's Intel's latest 10Q: http://edgar.sec.gov/Archives/edgar/data/50863/000 119312505159285/d10q.htm