Unfortunately, now that the old data have been shown to be badly flawed,
Actually, it was only data for the last month or so that was flawed due to the sensor malfunction. They use the old method of estimation exactly so they can monitor for long-term trends. That way the error stays consistent. The method of estimation is pretty well correlated by actual observations, so it's well understood what its shortcomings are. While not as accurate, it's still better than good enough for long-term trend analysis.
The vendors purchase licenses for a specific number of computers. If the vendor uses two licenses for the same machine (Vista & downgrade to XP), I don't see the vendor profiting in any way. What you did was purchase 2 licenses, and the profits are shared by your vendor as well as M$. Its hard to tell how they share those profits.
You'd only buy one license - the Vista one - because all Windows licenses come with backward rights. You can use a Vista license to install XP or 2k, or even Win95, although you may need to request a key. Volume licenses are no different in this regard.
The vendor makes money because they charge $60 to insert a custom auto-everything-self-installing XP DVD instead of a Vista DVD before pressing the reset button. Same work. $60 more. Profit!
Mandatory car analogy: providing cars that can drive faster than any speed limit makes the car makers and sellers accomplices in every case of excessive speed. Especially sports cars and motorcycles - what purpose can they possibly have other than to break the law?!
Of course, this is ridiculous - people who illegally share files are breaking the law, not the tool and community site vendors. The people who actually do break the law should get the equivalent of a $50 speeding ticket. It doesn't matter what the TPB calls itself, it's just a name and doesn't mean squat. The name Dodge Viper doesn't mean its purpose is to dodge law enforcement and send people to the hospital! It's simply a company and product name and they can call themselves and their products whateverthehell they wish. It doesn't prove intent of anything.
There's no measuring stick with "shadow of doubt" in our legal system, afaik. It's "beyond reasonable doubt", which is the amount of "Hrmm...." that would cause a reasonable person to suspect it might not be true.
It's also important to recognize that "reasonable doubt" is not a mathematical certainty. So even if a screenshot could theoretically have been photoshopped in a conspirary with ISPs to manufacture logs, the fact that none of them have a motive to do so means it's presumed they didn't. Unless demonstrated otherwise.
And yes, innocent people do get convicted. But not usually from manufactured technical evidence and complex conspiracies involving legions, but from either planted evidence or flakey eye witnesses. The latter can simply be a person who saw you shortly after a crime and is so convinced you're guilty that they'll say they actually saw you commit it. This is enough to convict you if you have a prior record or otherwise automatically suspect to begin with.
As said by another poster, "good way for mischievous students to cause mayhem by getting their university's mail servers blacklisted."
It takes a significant number of reports for an IP address to be blacklisted. By the time you've gotten that many messages out you're by definition spamming, and if the university lets you send it without throttling they're permitting its students to spam. The fix is to throttle individual send rates and caps to prevent automated mail senders - whether legit or not. If you run a legit newsletter or some such you should obtain permission. Once the university has fixed its servers so students can't use them for spam it takes about a minute to get themselves delisted.
The end-to-end signature scheme, while a good idea in general, won't help with spam. They'll just sign their spam using the signature of the owners of the machines in their botnet used to send it. This is assuming you don't have to regularly reauthenticate when you send email, to unlock your keys, but that will never be acceptable to end users.
It is, however, a very good idea for banks and others to start signing their mail. The signature should be an SMTP header, and it should be a legal requirement for banks and other trust corporations to use it in their communications. (Why they don't already at least sign their correspondence beats me, but I guess without a standardized header UMAs won't have automatic verification built in, making it a feature practical only to a small number of users.) This would be effective against phishing, although of no use against spam. PGP as usual is massive overkill (internal platform syndrome) for what is a simple problem.
Is there a problem with Microsoft using BSD code in their proprietary products?
No! Especially if they use BSD networking components - because that will mean their stuff will work well enough while staying compatible with the rest of the world. We all know what happens when MS goes off to invent a "better" way...
The real test will be whether the next iteration of this hardware runs this same OS or whether it comes with WinMo/WinCE.
Note to interested applicants: dead-end job. You will soon become a legacy department. You. Alone. Ask yourself what a single product iteration with NetBSD will do for your resume - very little!
Do yourself a favor and find an embedded Linux job instead...
This is actually one of the things I admire about developers in a position to release their code under BSD licenses. The end user is free to do anything they please with the code, including rolling it into a proprietary product, as long as they follow the attribution requirements. As for myself, most of my public code is licensed under the GPL, for various reasons (some being financially related). No one can reasonably argue that BSD-licensed code isn't truly free.
I couldn't agree more. For something of limited commercial value I'll GPL it. That way others can enhance and work on it, and it might bring some value back to me. Or it might give it sufficient exposure to permit a separate commercial license of a private branch; one with only my parts in it. Or it might bring in consulting work - lots of good reasons. If I release it under the BSD license I'm much less likely to benefit financially; it might end up used more, but I'm not convinced it will bring a corresponding bump in contract work. I take pride in crafting good software, and the better it's designed the easier it is for a more junior developer to pick it up and run with it, and this works against a BSD license but for a GPL-with-commercial-on-the-side setup. Of course, in the other direction I love the BSD license and public domain since I can use those for product work. But so can anyone else, so the value add is the value of my enhancements. For trivial changes I'll just submit it for consideration by the maintainers, but often it's such a specialized one-off that they're not interested. The big value is in not having to roll the same basic infrastructure and plumbing stuff from scratch - usually pretty easy stuff but tedious and time consuming to retype and get right. Again. Like say sha1 - just grab the public domain Reid version.
I don't consider the GPL as free as BSD for my own code - that's why I use it!
I always wondered how that cherry-pick arrangement came to pass, if any of you know, please respond because that would perhaps enlighten us as to Charter's financial woes.
It was broken into many small pieces that were sold off, and then reconsolidated itself into the current arrangement? Just speculating - but this is how it usually plays out.
The catch is it'll only be for very brief periods, then traffic shaping will kick in and throttle you back. Basically, it'll be for sitting and browsing the web or reading email - where latency, round-trip time, and rendering speed are dominant once you get beyond even a few megabit/s. If you can't actually sustain it for bulk downloads (of your favorite ISOs only of course), then what's the point. Given the practices of the cable industry in the past and even right now, without further evidence to the contrary I'd write this off as more of the same BS. It's the all-you-can-eat scheme where they give you tiny plates, only two trips to the buffet, and charge $15 for a beer (normally $20 but you get the special bundle discount) to go with.
I know a lot of people are all big on plug-in electric cars, but what do those of us not fortunate enough to have integral garages with outlets in them? I don't know what the percentages are, but I'm assuming there's a lot of average Joe's like me who, even if we own our own homes, have to park on the street wherever we can find parking. Are they going to put outlets in the sidewalks for me?
You know, you can call an electrician and have them install a circuit...
Yes, you'd need very high current - industrial power. Which means it's only relevant when retrofitting gas stations and other commercial charging locations. For home it hard matters since you can leave the car to charge while doing something else. It's mostly important for the gas-station-equivalent.
His mention of range was to point out that it's ridiculous to have a 1,000 HP motor when actually using all 1,000 HP would drain the batteries so fast you'd never get anywhere. Of course it would go farther if you only used a tiny fraction of the motor's capability, which is almost certainly what the 150 mi value is based on.
But you really only use the capacity of any car engine a fraction of the time - to start from lights, to accelerate to highway speed, etc. The rest of the time you coast at a fraction of capacity. And for this electric motors are fantastic - power out follows power in very closely, unlike a large-displacement combustion engine. And most of the time you wouldn't use the full 1000 hp, in fact anything over 300 hp likely makes no difference whatsoever in normal traffic. With a composite body you can still accelerate and brake like a motorcycle. But it might be fun to have on track days, and since it doesn't affect the everyday driving negatively, if you don't mind the expense - why not! (And yes, I do understand that acceleration is a function of gearing, while top speed is a function of hp.)
The rear wheel could be completely inactive, and just a caster. Turning can be done by braking and driving the front wheels - to enter a turn brake the left and continue driving the right; as the vehicle picks up rotational momentum the tail will come around; then to exit the turn reverse - brake the outside and drive the inside. Think Segway with a tail.
MS Research does a lot of good research, and this is no exception. Stuff like this allows us (humanity) to explore what makes something musical, why we enjoy musing, and so on. This is all good stuff. The problem really is when research organizations are burdened with a requirement for projects to result in marketable products and revenue pull; that's when you get silly products like this. Clearly this would be best off open sourced and shared, it likely has no future as a proprietary product. I'm sure the researchers themselves would totally agree, they just can't openly express this sentiment for political reasons.
I use Dvorak because I have CRPS from damage to the brachial plexus. Not only can I type with a lot less movement (except for the letter f), which means a lot less pain, but I'm much faster than I ever was with qwerty.
An alternative conclusion is that what might give you relief is the change in movement patterns inherent to any change in layout, rather than the specifics of Dvorak. This also doesn't necessarily mean your problem won't resurface and you won't eventually benefit from a switch back to qwerty! The problem with these sort of X-hurts-and-when-I-switched-to-Y-it-stopped is that they're, well, kind of obvious.
I almost unconsciously hit C-a when I want to go to the beginning of the line, which often results in a surprise when using a Windows machine:P
The most annoying to me used to be when going to a DOS machine and rediscovering that C-p still doesn't go to the previous line. Just like last time, it would start printing...
I'd love to see a completely new layout (incl number of keys) that are more compatible with the major languages that uses roughly the same characters.
Yes! I type mostly C++, so would like to see [ ] { } ( ) * & ! = + - ? : ; ' " placed more prominently on the keyboard, with special keys for && and ||. ^ % are rarely used, and $ # @ | can stay where they are. The control key goes where PC103 normally places caps lock (thankfully remappable on OS X), and caps lock can be removed completely. The numeric keyboard can be completely removed since that's where the mouse should sit for those of us who are right handed. Arrows, page keys, and all that crap can go as well - with property control and meta keys all of that just wastes desk space, and removing it would allow bringing the mouse in closer to the keyboard.
I could deal with a little awkwardness in the.1% of my usage that constitutes rambling geeky posts on slashdot and other English language...
But then, I don't think a change of layout would appreciably affect my productivity so is mostly a matter of academic curiosity.
more importantly the most common letter pairs will have one on each side of the keyboard so that when you press a letter with one hand your next key will be under the other hand. This makes typing a lot easier.
Maybe easier, but for speed what matters is that you type with your fingers, not your hands. What's important for speed is that common pairs are not under the same finger, and when under the same hand within normal finger dexterity - like not on top and bottom rows under adjacent fingers.
Of course this is mainly a big deal if you're a professional typist, but is there even such a job these days? As a programmer, for me, what matters is that typing is internalized so it doesn't interfere with the thought process. One could argue this is different from a typist who presumably can pay their full attention to the typing process itself. For me speed is unimportant because I bottleneck around thinking, not typing - I can almost always type faster than I think. To me consistent layout is far more important than whether a professional typist can accomplish 90 or 100 wpm!
isn't centralization the heart of source code management
Not necessarily. Consider a common case: - Project A works on a significant feature (say a new file system) - As part of their work, they significantly restructure some related part (say how a fs ties into the kernel) and update other parts of the source to match - Project B works on a different feature (say overhauling the interrupt thread implementation on SMP systems) - Project B wants to update the same related parts (say how kernel modules, including file systems, tie into the kernel) - Project A is already done, and is scheduled to get on the train before project B, so it's natural for B to integrate portions of project A and then track project A's fixes and updates to these portions up to when A integrates into the trunk
This situation is handled very cleanly by distributed systems like git and teamworks. As B selectively merges parts of A it picks up the change log. With svn when you diffpatch across from the A branch to the B branch you lose changelogs, there will be no record that these changes came from A but they'll appear independently in B. When A integrates to trunk if B simply tracks these it will appear as if B edited trunk. This is an inaccurate history. With p4 I get a headache just thinking about it.
And unlike CO2, water tends to precipitate out of the atmosphere rather than hang around for decades.
Yeah, so this half an inch times the earth's surface will quickly come back down over a few relatively localized areas. Nice. I just hope it's not where *I* live.
Let management care about... well, whatever it is these guys were up to.
Getting ahead on the career ladder and growing their fiefdoms. Which has more to do with getting the credit for what's good and making sure someone else is stuck with the bag of fail. Which, predictably, results in a dysfunctional organization that can't tell ass from head.
Slashdot Mirror
Unfortunately, now that the old data have been shown to be badly flawed,
Actually, it was only data for the last month or so that was flawed due to the sensor malfunction. They use the old method of estimation exactly so they can monitor for long-term trends. That way the error stays consistent. The method of estimation is pretty well correlated by actual observations, so it's well understood what its shortcomings are. While not as accurate, it's still better than good enough for long-term trend analysis.
The vendors purchase licenses for a specific number of computers. If the vendor uses two licenses for the same machine (Vista & downgrade to XP), I don't see the vendor profiting in any way. What you did was purchase 2 licenses, and the profits are shared by your vendor as well as M$. Its hard to tell how they share those profits.
You'd only buy one license - the Vista one - because all Windows licenses come with backward rights. You can use a Vista license to install XP or 2k, or even Win95, although you may need to request a key. Volume licenses are no different in this regard.
The vendor makes money because they charge $60 to insert a custom auto-everything-self-installing XP DVD instead of a Vista DVD before pressing the reset button. Same work. $60 more. Profit!
Mandatory car analogy: providing cars that can drive faster than any speed limit makes the car makers and sellers accomplices in every case of excessive speed. Especially sports cars and motorcycles - what purpose can they possibly have other than to break the law?!
Of course, this is ridiculous - people who illegally share files are breaking the law, not the tool and community site vendors. The people who actually do break the law should get the equivalent of a $50 speeding ticket. It doesn't matter what the TPB calls itself, it's just a name and doesn't mean squat. The name Dodge Viper doesn't mean its purpose is to dodge law enforcement and send people to the hospital! It's simply a company and product name and they can call themselves and their products whateverthehell they wish. It doesn't prove intent of anything.
There's no measuring stick with "shadow of doubt" in our legal system, afaik. It's "beyond reasonable doubt", which is the amount of "Hrmm...." that would cause a reasonable person to suspect it might not be true.
It's also important to recognize that "reasonable doubt" is not a mathematical certainty. So even if a screenshot could theoretically have been photoshopped in a conspirary with ISPs to manufacture logs, the fact that none of them have a motive to do so means it's presumed they didn't. Unless demonstrated otherwise.
And yes, innocent people do get convicted. But not usually from manufactured technical evidence and complex conspiracies involving legions, but from either planted evidence or flakey eye witnesses. The latter can simply be a person who saw you shortly after a crime and is so convinced you're guilty that they'll say they actually saw you commit it. This is enough to convict you if you have a prior record or otherwise automatically suspect to begin with.
Okay. Is someone with a doctoral degree in structural mechanics thereby qualified to comment on quantum gravity?
No, but if 82% of structural engineers agree a structure is unsafe I find that worth paying attention to.
As said by another poster, "good way for mischievous students to cause mayhem by getting their university's mail servers blacklisted."
It takes a significant number of reports for an IP address to be blacklisted. By the time you've gotten that many messages out you're by definition spamming, and if the university lets you send it without throttling they're permitting its students to spam. The fix is to throttle individual send rates and caps to prevent automated mail senders - whether legit or not. If you run a legit newsletter or some such you should obtain permission. Once the university has fixed its servers so students can't use them for spam it takes about a minute to get themselves delisted.
The end-to-end signature scheme, while a good idea in general, won't help with spam. They'll just sign their spam using the signature of the owners of the machines in their botnet used to send it. This is assuming you don't have to regularly reauthenticate when you send email, to unlock your keys, but that will never be acceptable to end users.
It is, however, a very good idea for banks and others to start signing their mail. The signature should be an SMTP header, and it should be a legal requirement for banks and other trust corporations to use it in their communications. (Why they don't already at least sign their correspondence beats me, but I guess without a standardized header UMAs won't have automatic verification built in, making it a feature practical only to a small number of users.) This would be effective against phishing, although of no use against spam. PGP as usual is massive overkill (internal platform syndrome) for what is a simple problem.
Is there a problem with Microsoft using BSD code in their proprietary products?
No! Especially if they use BSD networking components - because that will mean their stuff will work well enough while staying compatible with the rest of the world. We all know what happens when MS goes off to invent a "better" way...
The real test will be whether the next iteration of this hardware runs this same OS or whether it comes with WinMo/WinCE.
Note to interested applicants: dead-end job. You will soon become a legacy department. You. Alone. Ask yourself what a single product iteration with NetBSD will do for your resume - very little!
Do yourself a favor and find an embedded Linux job instead...
This is actually one of the things I admire about developers in a position to release their code under BSD licenses. The end user is free to do anything they please with the code, including rolling it into a proprietary product, as long as they follow the attribution requirements. As for myself, most of my public code is licensed under the GPL, for various reasons (some being financially related). No one can reasonably argue that BSD-licensed code isn't truly free.
I couldn't agree more. For something of limited commercial value I'll GPL it. That way others can enhance and work on it, and it might bring some value back to me. Or it might give it sufficient exposure to permit a separate commercial license of a private branch; one with only my parts in it. Or it might bring in consulting work - lots of good reasons. If I release it under the BSD license I'm much less likely to benefit financially; it might end up used more, but I'm not convinced it will bring a corresponding bump in contract work. I take pride in crafting good software, and the better it's designed the easier it is for a more junior developer to pick it up and run with it, and this works against a BSD license but for a GPL-with-commercial-on-the-side setup. Of course, in the other direction I love the BSD license and public domain since I can use those for product work. But so can anyone else, so the value add is the value of my enhancements. For trivial changes I'll just submit it for consideration by the maintainers, but often it's such a specialized one-off that they're not interested. The big value is in not having to roll the same basic infrastructure and plumbing stuff from scratch - usually pretty easy stuff but tedious and time consuming to retype and get right. Again. Like say sha1 - just grab the public domain Reid version.
I don't consider the GPL as free as BSD for my own code - that's why I use it!
I always wondered how that cherry-pick arrangement came to pass, if any of you know, please respond because that would perhaps enlighten us as to Charter's financial woes.
It was broken into many small pieces that were sold off, and then reconsolidated itself into the current arrangement? Just speculating - but this is how it usually plays out.
What is the catch behind 60 Mbps service?
The catch is it'll only be for very brief periods, then traffic shaping will kick in and throttle you back. Basically, it'll be for sitting and browsing the web or reading email - where latency, round-trip time, and rendering speed are dominant once you get beyond even a few megabit/s. If you can't actually sustain it for bulk downloads (of your favorite ISOs only of course), then what's the point. Given the practices of the cable industry in the past and even right now, without further evidence to the contrary I'd write this off as more of the same BS. It's the all-you-can-eat scheme where they give you tiny plates, only two trips to the buffet, and charge $15 for a beer (normally $20 but you get the special bundle discount) to go with.
I know a lot of people are all big on plug-in electric cars, but what do those of us not fortunate enough to have integral garages with outlets in them? I don't know what the percentages are, but I'm assuming there's a lot of average Joe's like me who, even if we own our own homes, have to park on the street wherever we can find parking. Are they going to put outlets in the sidewalks for me?
You know, you can call an electrician and have them install a circuit...
Yes, you'd need very high current - industrial power. Which means it's only relevant when retrofitting gas stations and other commercial charging locations. For home it hard matters since you can leave the car to charge while doing something else. It's mostly important for the gas-station-equivalent.
His mention of range was to point out that it's ridiculous to have a 1,000 HP motor when actually using all 1,000 HP would drain the batteries so fast you'd never get anywhere. Of course it would go farther if you only used a tiny fraction of the motor's capability, which is almost certainly what the 150 mi value is based on.
But you really only use the capacity of any car engine a fraction of the time - to start from lights, to accelerate to highway speed, etc. The rest of the time you coast at a fraction of capacity. And for this electric motors are fantastic - power out follows power in very closely, unlike a large-displacement combustion engine. And most of the time you wouldn't use the full 1000 hp, in fact anything over 300 hp likely makes no difference whatsoever in normal traffic. With a composite body you can still accelerate and brake like a motorcycle. But it might be fun to have on track days, and since it doesn't affect the everyday driving negatively, if you don't mind the expense - why not! (And yes, I do understand that acceleration is a function of gearing, while top speed is a function of hp.)
The rear wheel could be completely inactive, and just a caster. Turning can be done by braking and driving the front wheels - to enter a turn brake the left and continue driving the right; as the vehicle picks up rotational momentum the tail will come around; then to exit the turn reverse - brake the outside and drive the inside. Think Segway with a tail.
There's a public demo up at Microsoft Research
MS Research does a lot of good research, and this is no exception. Stuff like this allows us (humanity) to explore what makes something musical, why we enjoy musing, and so on. This is all good stuff. The problem really is when research organizations are burdened with a requirement for projects to result in marketable products and revenue pull; that's when you get silly products like this. Clearly this would be best off open sourced and shared, it likely has no future as a proprietary product. I'm sure the researchers themselves would totally agree, they just can't openly express this sentiment for political reasons.
I use Dvorak because I have CRPS from damage to the brachial plexus. Not only can I type with a lot less movement (except for the letter f), which means a lot less pain, but I'm much faster than I ever was with qwerty.
An alternative conclusion is that what might give you relief is the change in movement patterns inherent to any change in layout, rather than the specifics of Dvorak. This also doesn't necessarily mean your problem won't resurface and you won't eventually benefit from a switch back to qwerty! The problem with these sort of X-hurts-and-when-I-switched-to-Y-it-stopped is that they're, well, kind of obvious.
I almost unconsciously hit C-a when I want to go to the beginning of the line, which often results in a surprise when using a Windows machine :P
The most annoying to me used to be when going to a DOS machine and rediscovering that C-p still doesn't go to the previous line. Just like last time, it would start printing...
I'd love to see a completely new layout (incl number of keys) that are more compatible with the major languages that uses roughly the same characters.
Yes! I type mostly C++, so would like to see [ ] { } ( ) * & ! = + - ? : ; ' " placed more prominently on the keyboard, with special keys for && and ||. ^ % are rarely used, and $ # @ | can stay where they are. The control key goes where PC103 normally places caps lock (thankfully remappable on OS X), and caps lock can be removed completely. The numeric keyboard can be completely removed since that's where the mouse should sit for those of us who are right handed. Arrows, page keys, and all that crap can go as well - with property control and meta keys all of that just wastes desk space, and removing it would allow bringing the mouse in closer to the keyboard.
I could deal with a little awkwardness in the .1% of my usage that constitutes rambling geeky posts on slashdot and other English language...
But then, I don't think a change of layout would appreciably affect my productivity so is mostly a matter of academic curiosity.
more importantly the most common letter pairs will have one on each side of the keyboard so that when you press a letter with one hand your next key will be under the other hand. This makes typing a lot easier.
Maybe easier, but for speed what matters is that you type with your fingers, not your hands. What's important for speed is that common pairs are not under the same finger, and when under the same hand within normal finger dexterity - like not on top and bottom rows under adjacent fingers.
Of course this is mainly a big deal if you're a professional typist, but is there even such a job these days? As a programmer, for me, what matters is that typing is internalized so it doesn't interfere with the thought process. One could argue this is different from a typist who presumably can pay their full attention to the typing process itself. For me speed is unimportant because I bottleneck around thinking, not typing - I can almost always type faster than I think. To me consistent layout is far more important than whether a professional typist can accomplish 90 or 100 wpm!
s/is already done/is already done with the shared part/g
isn't centralization the heart of source code management
Not necessarily. Consider a common case:
- Project A works on a significant feature (say a new file system)
- As part of their work, they significantly restructure some related part (say how a fs ties into the kernel) and update other parts of the source to match
- Project B works on a different feature (say overhauling the interrupt thread implementation on SMP systems)
- Project B wants to update the same related parts (say how kernel modules, including file systems, tie into the kernel)
- Project A is already done, and is scheduled to get on the train before project B, so it's natural for B to integrate portions of project A and then track project A's fixes and updates to these portions up to when A integrates into the trunk
This situation is handled very cleanly by distributed systems like git and teamworks. As B selectively merges parts of A it picks up the change log. With svn when you diffpatch across from the A branch to the B branch you lose changelogs, there will be no record that these changes came from A but they'll appear independently in B. When A integrates to trunk if B simply tracks these it will appear as if B edited trunk. This is an inaccurate history. With p4 I get a headache just thinking about it.
And unlike CO2, water tends to precipitate out of the atmosphere rather than hang around for decades.
Yeah, so this half an inch times the earth's surface will quickly come back down over a few relatively localized areas. Nice. I just hope it's not where *I* live.
Let management care about... well, whatever it is these guys were up to.
Getting ahead on the career ladder and growing their fiefdoms. Which has more to do with getting the credit for what's good and making sure someone else is stuck with the bag of fail. Which, predictably, results in a dysfunctional organization that can't tell ass from head.