And on the other hand, some downloaded movies immediately equate to increased revenue by providing publicity and making people want to watch in the theaters; compare the studies that have found that P2P music downloaders buy more CDs and official music downloads.
Seems to me that the best we can do is to say that we don't know but that it's probably about a wash.
I nominate having to run a suite of sold-separately (or at least downloaded-separately) programs to watch for viruses, detect spyware, replace the apparently-not-good-enough included firewall... This is a workaround for broken DOSisms that's made itself into and entire industry.
Yes, this probably could have been handled better, but many home NAT boxes are so broken that they ought to be recalled. I'm actually a bit surprised that there's not a WRT54GL equivalent with well-behaved 2.6 wireless drivers that ships with an open-source configuration interface; it seems that it'd be cheaper than constantly reinventing the wheel.
Passing on the low-hanging fruit, it seems to me that this is pretty much exactly the kind of thing that happens all the time at the packet layer. WRED, for example, selectively drops packets even when buffers aren't full and the network is still theoretically operating under capacity so that the next TCP connection doesn't bring down the uplink. How is the Microsoft strategy qualitatively different?
I pick up the rock, I drop it, the rock falls to the ground. I pick it up and drop it again, and again it falls to the ground. Etc.
Until the time when you drop the rock, it hits the ground, and it flies a hundred feet overhead. That's about what happened in the gold-foil experiment, and it goes to show that what you think you know may be entirely useful, just like Newtonian mechanics... but you may not understand fully what's happening.
It could even be argued that there's no inherent reason why these should require a reboot, but as a pragmatic issue there does need to be some method of cleaning up running processes and the like, if only because real-world systems have bugs and give rise to things such as zombie processes.
Reconciling these two is actually an open research question I'm working on; any insights are welcome. My current idea is to have a mechanism to wipe the process table on boot instead of restoring it, which should be about the same as a reboot on FS-based systems.
The other extreme point of view is that modern systems should only have virtual memory and, instead of having an explicit file system, treat mass storage as a level-4 cache. In fact, systems that support mmap(2) do this partially.
The idea here is that modern memory management is actually pretty good, and that it's best to let the OS decide what to keep in RAM and what to swap out, so that issues like prefetching can be handled transparently.
Perhaps so, but if the programmer is aware that his code will be available for future customers to see, that provides a pretty strong incentive not to churn out crapware. This works in most of the rest of the economy, and it's hard to believe on faith, as so many seem to, that it can't work for software. I doubt that all of the programmers hired to work on Apache or MySQL always feel pumped about writing whatever regression test needs doing, but even if they had no personal pride in their work, there would be the external incentive.
Perhaps because you want a globally-distributed DNS system, similar to Akamai's content system? Because you need to be able to do automatic provisioning of devices that are then going to wander around the country and need dynamic DNS? Or maybe you need DNS-level failover for your distributed servers.
(I run tinydns myself, but there are plenty of reasons to hire someone else to do jobs like DNS.)
Although I am one of those who is appalled by the fact that Congress can get away with writing laws about nearly anything by waving their hands and yelling "commerce clause", it really seems that a law like this is just asking to get smacked down at the federal leve.
The ACM chapter at my university was in a very similar spot when I was there (a couple of years ago). We tried to help provide advanced tech-support (at that time, having a USB floppy drive and knowing how to use ddrescue saved quite a few term papers), campus-wide educational opportunities (hour-or-so presentations aimed either at computer people or the general public), and local FOSS mirrors.
We also realized that, like you, we'd essentially been running pilot projects of technologies the university later deployed campus-wide (such as wireless access). We sat down with the IT staff and worked out a plan whereby we'd keep them up to date on what projects we were playing with and look into new technology at their request (such as 802.1x from our SOS), and they'd give us some room to experiment on the main network.
Among other projects, we ran an IPv6 gateway (SixXS tunnel) until I graduated that provided native IPv6 access to anyone in the science building (and with lots of Macs, Linux boxes, and Solaris machines, it was used pretty heavily), and we ran a pilot project with three different types of thin clients running from the same server for comparison and evaluation.
I blame the copyright holders for sticking their fingers in their ears and screaming about "protection" when what they mean is demanding perpetual control over how anyone ever uses the material.
Donovan's comment makes me curious as to whether he understands networking at all. BT data transfers happen over TCP, which naturally contends with other TCP connections and ends up sharing available capacity more or less equally among connections. The fact that BT traffic is heaviest when other usage is lowest is therefore a truism--those BT connections go faster when they're not competing for bandwidth.
I actually wrote a report on this topic a few months ago. With a wire-in-Pringles-can antenna and stock WRT54G on one end, we were able to maintain usable (5.5Mbps) link at 550m (~13dB gain over the stock). A structured cantenna pair should get you u to 2-3km, and if you're interested the WRT's can run off a car battery+solar panel.
Minor correction: copyright *transfers* have to be in writing, but I believe that unwritten *licenses* are OK. (Of course, if it's important, you should always get it in writing, etc., etc.)
I wouldn't call myself a "fanboy", but I've used Linux exclusively on my own computers for nearly 10 years (and recommend Linux and Mac OS X to friends). Right now, I'm working on a massive IT upgrade project for a large company, and whether we end up going with Dell, Sun, HP, or somebody else, we certainly won't be building our new boxes ourselves; it's just not cost-effective for most businesses.
I configured an E520/E520n to identical specs, with the one exception that the Windows version came with a Core 2 Duo (1.86GHz) and the n Series a Pentium D (2.80GHz, and Linux would've used the 64-bit processor). The n Series came out to $959 and the regular to $1089, a difference of $130 for Windows and the Core 2.
If you can demonstrate that you wrote a work independently of someone else, that person's copyright doesn't apply to you. I think you could rather easily demonstrate independent creation of an empty file.;-)
This is the reason I'm always very careful to maintain the distinction between "memory" (RAM) and "space" (secondary storage). Non-geeks may not always understand just from the terms what the difference is, but I've found that most people can grasp the analogy between "memory" (things I have in my head) vs. "space" (things I can't remember but wrote down and put in my desk for next week).
If we're getting grammar-Nazi, saying that $FOO is comprised of $BAR is nonsensical (expand your definition of comprise). Rather, $FOO comprises $BAR and is composed of $BAR.
Slashdot Mirror
And on the other hand, some downloaded movies immediately equate to increased revenue by providing publicity and making people want to watch in the theaters; compare the studies that have found that P2P music downloaders buy more CDs and official music downloads.
Seems to me that the best we can do is to say that we don't know but that it's probably about a wash.
I nominate having to run a suite of sold-separately (or at least downloaded-separately) programs to watch for viruses, detect spyware, replace the apparently-not-good-enough included firewall... This is a workaround for broken DOSisms that's made itself into and entire industry.
Yes, this probably could have been handled better, but many home NAT boxes are so broken that they ought to be recalled. I'm actually a bit surprised that there's not a WRT54GL equivalent with well-behaved 2.6 wireless drivers that ships with an open-source configuration interface; it seems that it'd be cheaper than constantly reinventing the wheel.
Xenu I'll take as a given. I'll need some evidence for Beelzebub's involvement.
Passing on the low-hanging fruit, it seems to me that this is pretty much exactly the kind of thing that happens all the time at the packet layer. WRED, for example, selectively drops packets even when buffers aren't full and the network is still theoretically operating under capacity so that the next TCP connection doesn't bring down the uplink. How is the Microsoft strategy qualitatively different?
I pick up the rock, I drop it, the rock falls to the ground. I pick it up and drop it again, and again it falls to the ground. Etc.
Until the time when you drop the rock, it hits the ground, and it flies a hundred feet overhead. That's about what happened in the gold-foil experiment, and it goes to show that what you think you know may be entirely useful, just like Newtonian mechanics... but you may not understand fully what's happening.
It could even be argued that there's no inherent reason why these should require a reboot, but as a pragmatic issue there does need to be some method of cleaning up running processes and the like, if only because real-world systems have bugs and give rise to things such as zombie processes.
Reconciling these two is actually an open research question I'm working on; any insights are welcome. My current idea is to have a mechanism to wipe the process table on boot instead of restoring it, which should be about the same as a reboot on FS-based systems.
The other extreme point of view is that modern systems should only have virtual memory and, instead of having an explicit file system, treat mass storage as a level-4 cache. In fact, systems that support mmap(2) do this partially.
The idea here is that modern memory management is actually pretty good, and that it's best to let the OS decide what to keep in RAM and what to swap out, so that issues like prefetching can be handled transparently.
Perhaps so, but if the programmer is aware that his code will be available for future customers to see, that provides a pretty strong incentive not to churn out crapware. This works in most of the rest of the economy, and it's hard to believe on faith, as so many seem to, that it can't work for software. I doubt that all of the programmers hired to work on Apache or MySQL always feel pumped about writing whatever regression test needs doing, but even if they had no personal pride in their work, there would be the external incentive.
(I run tinydns myself, but there are plenty of reasons to hire someone else to do jobs like DNS.)
Although I am one of those who is appalled by the fact that Congress can get away with writing laws about nearly anything by waving their hands and yelling "commerce clause", it really seems that a law like this is just asking to get smacked down at the federal leve.
The ACM chapter at my university was in a very similar spot when I was there (a couple of years ago). We tried to help provide advanced tech-support (at that time, having a USB floppy drive and knowing how to use ddrescue saved quite a few term papers), campus-wide educational opportunities (hour-or-so presentations aimed either at computer people or the general public), and local FOSS mirrors.
We also realized that, like you, we'd essentially been running pilot projects of technologies the university later deployed campus-wide (such as wireless access). We sat down with the IT staff and worked out a plan whereby we'd keep them up to date on what projects we were playing with and look into new technology at their request (such as 802.1x from our SOS), and they'd give us some room to experiment on the main network.
Among other projects, we ran an IPv6 gateway (SixXS tunnel) until I graduated that provided native IPv6 access to anyone in the science building (and with lots of Macs, Linux boxes, and Solaris machines, it was used pretty heavily), and we ran a pilot project with three different types of thin clients running from the same server for comparison and evaluation.
I blame the copyright holders for sticking their fingers in their ears and screaming about "protection" when what they mean is demanding perpetual control over how anyone ever uses the material.
Donovan's comment makes me curious as to whether he understands networking at all. BT data transfers happen over TCP, which naturally contends with other TCP connections and ends up sharing available capacity more or less equally among connections. The fact that BT traffic is heaviest when other usage is lowest is therefore a truism--those BT connections go faster when they're not competing for bandwidth.
I actually wrote a report on this topic a few months ago. With a wire-in-Pringles-can antenna and stock WRT54G on one end, we were able to maintain usable (5.5Mbps) link at 550m (~13dB gain over the stock). A structured cantenna pair should get you u to 2-3km, and if you're interested the WRT's can run off a car battery+solar panel.
In particular, I'd suggest using PGP. It supports authentication of the data and is easy to use for individual files.
Ask the recipient to provide a public key over a known-secure channel (such as by courier).
I don't skip commercials... MythTV does it for me!
I'm surprised no one's said it yet...
But somebody get those hooligans off my LAN!
To an extent, seems like the security concern of ensuring that sensitive information wasn't leaked prevailed...
Minor correction: copyright *transfers* have to be in writing, but I believe that unwritten *licenses* are OK. (Of course, if it's important, you should always get it in writing, etc., etc.)
I wouldn't call myself a "fanboy", but I've used Linux exclusively on my own computers for nearly 10 years (and recommend Linux and Mac OS X to friends). Right now, I'm working on a massive IT upgrade project for a large company, and whether we end up going with Dell, Sun, HP, or somebody else, we certainly won't be building our new boxes ourselves; it's just not cost-effective for most businesses.
I configured an E520/E520n to identical specs, with the one exception that the Windows version came with a Core 2 Duo (1.86GHz) and the n Series a Pentium D (2.80GHz, and Linux would've used the 64-bit processor). The n Series came out to $959 and the regular to $1089, a difference of $130 for Windows and the Core 2.
If you can demonstrate that you wrote a work independently of someone else, that person's copyright doesn't apply to you. I think you could rather easily demonstrate independent creation of an empty file. ;-)
This is the reason I'm always very careful to maintain the distinction between "memory" (RAM) and "space" (secondary storage). Non-geeks may not always understand just from the terms what the difference is, but I've found that most people can grasp the analogy between "memory" (things I have in my head) vs. "space" (things I can't remember but wrote down and put in my desk for next week).
If we're getting grammar-Nazi, saying that $FOO is comprised of $BAR is nonsensical (expand your definition of comprise). Rather, $FOO comprises $BAR and is composed of $BAR.