Manual focus on an SLR depends on the lens of course, and some cheap lenses have a pretty lame focus ring, but on a good lens you auto-focus in much the same way you did in the '60s. Grab the ring and turn until the picture looks right.
On Canon cameras there's a little slider switch at the base of the lens clearly marked AF|MF, if the slider is at AF (towards the front of the lens) it'll focus for you, either when you half-press the shutter, or when you press the optional focus button on the back. (If you don't want the focusing to happen when you shoot.) If you slide the switch to MF you have to do all the focusing yourself, but if you wish, the camera will blink the auto-focus points when it thinks you've done it right. Good for a newbie, harmless for everyone else.
I've got a Digital Rebel, Canon's most hand-holding SLR, and it has a fully-functional manual mode, where you adjust aperture and shutter manually, and priority modes for both, as well as program, which is auto but where you can override the camera's choice. For instance, in program it'll pick the settings and you can spin the handy dial with your index finger to change the shutter/f-stop combo it picked while keeping the same exposure, or press the exposure compensation button and spin the same dial to increase or decrease the exposure.
Being able to modify the metered settings, upping the shutter speed in trade for a wider fstop, means I don't use the shutter/aperture priority modes as much. I only use shutter priority when I know that anything below a 1/100 will be blurry, in a car for instance, and I'm willing to sacrifice a properly exposed shot in order to just get the shot.
I was taking pictures within minutes of pulling the camera out of the box, despite never having used a Canon SLR before. The bells and whistles are there, but they aren't required.
The Canon EOS 300D (Digital Rebel) has already matched any 35mm film that any of us are going to buy, and that's assuming you can get the negative scanned at more than 4000dpi and actually get the full resolution, if you can see it around the grain. It's $900 USD.
The lack of noise in digital SLRs until you get to around ISO800 also makes the output look a lot nicer. If you routinely print larger than 20x30 you might find very sharp film has more detail, but at anything smaller it's not even a contest for modern digital SLRs.
There isn't skill in taking fifty pictures of a single view. The skill comes from this though, because you gain a very real understanding of what a stop under and over look like, what changing your focal length does to your DoF, and what f-stops are and why they do what they do.
If you're doing this on film you don't get as many test shots, or instant feedback.
Don't get me wrong, I like my digital SLR more than my digital rangefinder, but I can't imagine going back to film, no matter what the camera body was.
The only way to become a good photographer is to take pictures, digital facilitates this.
Oh, and you don't just focus of the whole shot on the screen, you zoom up to 10x closer and scroll around getting a very good idea of what the shot looked like, as well as if people had their eyes open.
I just sold my old digital. A Canon Powershot G2 that was 18-months old. I'd taken 19,500 photos with it. The equivalent of 540 rolls of film. About $6k worth by your calculations.
Before I'd gotten it I'd shot maybe five rolls of film. I was always wondering if a shot was worth it ("Girlfriend's cute, but I already got a picture last week.") Then, they'd sit around months or years until I got around to getting them developed. I didn't know anything about photography because what little feedback I had was so far after the shot.
Digital changed it all. I got a nice camera with all the manual overrides and full manual mode and I just started taking pictures. Lots of pictures.
I'd agree about cost, but I don't think there's anything good to be said about learning with film.
Film is expensive per shot. You're less likely to take pictures if every time you press the shutter you pay. You're even less likely to take five or ten shots of the same thing in order to experiment.
Even if you do, you need to have a log book (unless you're using a modern APS camera which tags the film with the settings - I think) in which you enter all the shot details, shutter, aperture, lens used, etc. Otherwise when you do get them developed - a week later for most people, or even a few hours later if you have your own darkroom, you won't remember why shot four of the group of ten is the best.
With digital you take the picture and get instant feedback, with histograms and blown highlights meters. If you don't like the result you change a setting and try again. Not only do you learn through the instant feedback, but a month later when you look at the photo on the computer and wonder why it looks the way it does you can check the exif data and see what it says. All your settings are recorded.
If you buy a digital SLR you'll also get amazingly low noise (the digital equivalent of grain). ISO 200 on my camera (Canon EOS 300D) is noise free, ISO 800 is barely noticable except in shadows, and ISO 1600 is acceptable. But, more than that, you can switch from ISO 100 to 1600 with the spin of a dial, giving you much more flexibility than with film.
You can also white-balance, producing shots where white objects look white, under vastly different lights. For white-balance and ISO in film you need to rewind, eject, and mark the spool, then load another roll and jump to the next empty frame. You then need to carry enough rolls for any occasion and keep track of your sessions as they cross rolls. A pain if you want timely development.
Most modern SLRs with auto-focus are blinding fast. Just today I fired off a four-shot burst at a crow that flew about ten feet over my head. At my first shot it was about 25' out, at the third shot is was straight above and half the distance, at the fourth shot it was already getting farther away.
This was with the Canon EOS 300D (Digital Rebel).
My old Canon Powershot G2 was a nice camera, but its autofocus was slow. With the rebel I can focus and shoot in about the same time (125ms) as the G2 took to shoot from focused and metered.
All you need to do is keep the focus point over the subject and your shots will be clear. And, if you can't, you turn off AF and do it yourself. With a manual camera you can't turn AF on, so you'd miss all of the fast-moving shots.
So maybe we should arrange a bounty for people who decompile and reverse-engineer closed-source drivers and post the intimate details about the hardware.
Not only would this let people write drivers for Linux based on the Windows drivers, but it would make hardware companies see that there's no value is keeping secrets... Surely ATI has looked at NVidia's drivers and vice-versa, it's perfectly legal.
We aren't saying you shouldn't set a schedule, just that you shouldn't set the schedule before you see the specs, unless you're allowed to prioritize the specs to meet the fixed timeline.
But, every company wants to set a release date that works for marketing (before next christmas) and then tell you what it needs to do.
Of course that's broken. They need to tell you what it needs to do and when they want it, then you prune the specs to achieve this.
Give her a stick (unmarked dowel would work) and a pencil. Tell her to measure the length of the microwave (for a hypothetical new cupboard or something). Now, if she takes this stick with her to the kitchen store, she's got a pretty accurate idea of the size. So she orders a cabinet. They don't have any tape-measures, but they've got another stick, they lay it down next to her stick and put a mark on theirs at roughly the same place. Then they send the stick with a description of the product, to the guy who builds it, and he marks his own stick off of theirs, and sends their stick back. Etc.
At every step, each stick is being marked from the stick before. Errors accumulate and eventually the microwave might not fit the cabinet.
But, if she'd used a ruler and read out the next-largest measurement, she'd be able to tell someone the number very accurately. 84cm is still 84cm, even after being written down a few times. It works over the phone, etc.
The stick, and pencial mark, is analog. It's based on taking a measurement of something and trying to convey that based on the size of the stick, or the voltage of a signal. If each sampling and reproduction isn't perfect (and it can't be), error is introduced.
The measurement from the tape is digital. Discrete signals that aren't measured by size and as such, are less error prone. That 84cm is the same, written in huge block betters with a wide marker, or neatly lettered with a fine-point pen by an architect.
If someone receives a stick with a slightly blurry line where do they mark the next stick? But if someone receives a piece of paper with 84cm written on it and the 4 is a bit smudged, they simply erase it and write a neat 4 over it, re-enforcing the signal so to speak. Unless it's so corrupted that they can't read anything, each step will be made without errors.
Binary is just the fairly obvious next step. What two signals are the hardest to mistake? Direct opposites. On/Off, Hot/Cold, Black/White. If you have other signals, there's more chance of error.
But they need to have a rough idea of how to use Linux, or MS will just ignore them. To get an idea, they need to study how they'd go about it and the pros and cons. At this point, they've already made the first step towards switching, learning about alternatives.
But, in the end, even if they don't switch they let everyone else know that MS will give you the licenses at a drastically reduced price to keep you. So now everyone is going to be investigating and threatening to switch, and in doing so they'll learn about the benefits.
So when MS says no, when their bottom-line won't allow them to give any more sweet deals, these companies will already be partly ready to switch.
The main reason people still run WinNT is because it costs to upgrade. It's a terrible pain to use. With Linux upgrades are of course free and people will upgrade when a feature they want is available. Few users run 2.2 kernels, even on old machine, because 2.4 offered a lot of desktop-level fixes.
Modern Linuxes can be stripped down. I know a guy with a 486-50 Notebook with 32mb of ram. He installed Mandrake, I think, and simply put in an old windows manager. It's modular so he can choose. He might want the recent kernels and tools of a later distro, but not want KDE/Gnome. With Windows, you take the whole package and all you can do is tweak some of the graphics settings.
And as for planned obsolescence, yeah, Linux distros intend for users to upgrade. But they don't have lock-in, you can either download their distro, or another one, if you don't want to pay them for an upgrade. That means that plan new features and reasons people will want to upgrade, they don't plan to make the old system stop working well because of new file formats. If they did that, people would switch all right - to another distro.
I find that Linux is friendlier on old hardware because I can upgrade to a new OS, or a new component, and a new GUI isn't required. I could take an old 486 that's been running as a router for years and slap iptables onto it without having to upgrade the whole OS. As such, it's not going to increase the hardware requirements.
Ditto with a user machine. If they only need a few updates, or a new tool or two, it never requires a complete upgrade and then nothing but the new tool is different, all the old apps run exactly the same.
XP is fairly lightweight, compared to 2k, so it's closer to this, but when I had to upgrade an old Win98 print-server to Win2k because of a networking issue the performance was terrible. With Linux I'd have just installed a specific module.
Yeah. Like in a game of cards, you can reverse what you're doing before the next guy starts his move. At the point that he's actually done something based on your move, you're screwed.
If they ran up while you were reading the sales agreement and demanded an extra $5k, that's one thing. To demand it after you've all signed, that's not acceptable.
Ditto here, where people had made trades based on the mistake. IMHO, everything goes (ditch the whole hand) or you keep going and the person who made the mistake takes a loss.
Did the buy price pop up and indicate that it was because of a bug, or did it look like if someone had decided to bail on a stock because they had decided it was a bad deal.
The stock market is all about arbitrary prices of stocks that rarely have anything to do with the value of the companies they supposedly represent.
Besides, if they haven't tossed Ken Lay in jail for the rest of his life and taken away every single possession and auctioned it, what sort of message are they sending? That the best way to get rich is through outright fraud?
But it is a wildcard law. Just not versus other corporations who can afford to fight back.
I could easily see some Adobe-like company claiming that by decoding the XORed traffic on their drive controller that you defeated a copy protection feature, even if the device is a microwave or something, and bury you under paperwork unless you agree to settle for $10k and a gag order.
I mean, that's pretty much a template for a future Slashdot submission right there.
But of course, when entities will millions of dollars for legal fees can't simply pick on the poor, they at least invent slightly better legal justification.
Insightful? How about retarded? How else can you write a post directly contrary to all evidence?
A product using ext3 wouldn't have to be open sourced, any more than a product running on Linux. Any changes made to the filesystem would, but it's highly unlikely that you're going to have so grand an idea for a filesystem that your product hinges on it, and then have to implement it on top of someone else's filesystem.
Besides, using GPLed components basically prevents patent issues. By intentionally releasing something that requires you to agree that it is patent encumbered you pretty much give implicit free licenses to any patents that you may have on that code. Otherwise you didn't honor the contract you entered into with the original author of the software.
What would you prefer? Having to open source some tiny filesystem component you added to ext3, or having to pay Microsoft up to $250,000?
Most explosives contain their own oxygen (bound weakly in some molecule). Explosives combust so quickly (by definition, if the speed is below a certain point it's not "explosive") that they'd never get oxygen to the explosive and they'd put themselves out.
Many things that are only explosive when mixed with air are like this. Gasoline will (usually, don't try this at home) just burn if a bucket of it is lit, but if it has time to evaporate and fill the room with fumes, it'll blow up.
As an example, think of a bullet. It's in a fairly air-tight barrel, with a hunk of lead between it and the source of oxygen. Bullets will fire in space even. No air needed. If they didn't contain everything needed for their reaction, they'd be a lot less impressive.
Yeah. If people investigated monthly plans they'd realize that they're about the same price, once you figure in the phone purchase, they let you cancel at any time, and as the company has an incentive to keep you they tend to be much nicer to deal with.
I switched from Telus (a Canadian version of a baby bell) to Fido and got incredible rates, deals (free FidoFido calls and text messages, free evenings, etc) and a company that wasn't rude when I had to call them. All because I bought an unlocked phone that could be transfered to their competition. Lock-in is bad when Microsoft does it, it's also bad when the phone company tries it.
Now if only we had number portability, we'd be set.
These people are abusing any and all common resources they can find, as well as attacking anyone who tries to get in their way.
Why shouldn't they be killed? What possible reason could there be to justify their ongoing existence at the cost of everyone around them? It's not like they don't know they're hurting people - they're specifically setting out to do so.
I'm willing to give everyone the benefit of the doubt, in the beginning, but I don't see why it's unreasonable to think that people could abuse their neighbors long enough for that assumption to change to them having to prove their worth.
Previously used books were sold direct-to-consumers, but in a local area only. Now they're sold, often to resellers, globally. These resellers handle the searching and internet buying aspects and deliver the books to consumers.
So yes, they did add a middle-man, but they also increased the usefulness of the service. (Especially since the major buyers at library sales tend to be resellers anyways.)
But now, it's an open system. The online auctions don't discriminate. If you're willing to offer more than a reseller, you'll get the book. Eventually we'll solve e-commerce (to the point where grandmothers are comfortable purchasing products online) and the last use of the reseller will be gone, because the system of making a book available globally will be in place.
You must not read the newspapers. While the parent poster's definition of epidemic might be a bit low, there are too many home invasions in the GVRD.
This sort of thing does happen because the criminals don't think there's much risk. They usually target older people and have a gun or knives. The victims are suprised and weaponless, if they can't call 911 quickly enough they're in trouble.
I'm not saying that guns would make this better, but consider that an 80-year old woman with a gun is the equal of a young thug with a gun, but with knives, or bats, or any improvised weaponry, she's far more likely to lose. Furthermore, gunshot wounds are far more fatal than knife wounds, a thug is risking his life going against someone with a gun, but probably only risking mild injury attacking the average person in their home.
This is why you specifically construct small elements that are provable, and string them together. You don't have to prove anything about a randomly created program.
You don't sit down and attempt to formally prove that Doom3 is "correct", you work on small pieces of it and then the logic that calls those pieces, etc.
If you assume a routine can only take certain inputs and produce certain outputs, use asserts (in production code, not just for testing) and if nothing else, at least catch your errors early before they cascade.
Yeah. If you want your business to score highly in the results, have something there worth searching for. Host forums and convince people to review the products you sell, or do something that differentiates between you and just another retailer.
Like Amazon for example. Their user reviews and such make their site a good one to look at when considering a product, and well, since I'm there anyways...
Slashdot Mirror
Manual focus on an SLR depends on the lens of course, and some cheap lenses have a pretty lame focus ring, but on a good lens you auto-focus in much the same way you did in the '60s. Grab the ring and turn until the picture looks right.
On Canon cameras there's a little slider switch at the base of the lens clearly marked AF|MF, if the slider is at AF (towards the front of the lens) it'll focus for you, either when you half-press the shutter, or when you press the optional focus button on the back. (If you don't want the focusing to happen when you shoot.) If you slide the switch to MF you have to do all the focusing yourself, but if you wish, the camera will blink the auto-focus points when it thinks you've done it right. Good for a newbie, harmless for everyone else.
I've got a Digital Rebel, Canon's most hand-holding SLR, and it has a fully-functional manual mode, where you adjust aperture and shutter manually, and priority modes for both, as well as program, which is auto but where you can override the camera's choice. For instance, in program it'll pick the settings and you can spin the handy dial with your index finger to change the shutter/f-stop combo it picked while keeping the same exposure, or press the exposure compensation button and spin the same dial to increase or decrease the exposure.
Being able to modify the metered settings, upping the shutter speed in trade for a wider fstop, means I don't use the shutter/aperture priority modes as much. I only use shutter priority when I know that anything below a 1/100 will be blurry, in a car for instance, and I'm willing to sacrifice a properly exposed shot in order to just get the shot.
I was taking pictures within minutes of pulling the camera out of the box, despite never having used a Canon SLR before. The bells and whistles are there, but they aren't required.
The Canon EOS 300D (Digital Rebel) has already matched any 35mm film that any of us are going to buy, and that's assuming you can get the negative scanned at more than 4000dpi and actually get the full resolution, if you can see it around the grain. It's $900 USD.
The lack of noise in digital SLRs until you get to around ISO800 also makes the output look a lot nicer. If you routinely print larger than 20x30 you might find very sharp film has more detail, but at anything smaller it's not even a contest for modern digital SLRs.
There isn't skill in taking fifty pictures of a single view. The skill comes from this though, because you gain a very real understanding of what a stop under and over look like, what changing your focal length does to your DoF, and what f-stops are and why they do what they do.
If you're doing this on film you don't get as many test shots, or instant feedback.
Don't get me wrong, I like my digital SLR more than my digital rangefinder, but I can't imagine going back to film, no matter what the camera body was.
The only way to become a good photographer is to take pictures, digital facilitates this.
Oh, and you don't just focus of the whole shot on the screen, you zoom up to 10x closer and scroll around getting a very good idea of what the shot looked like, as well as if people had their eyes open.
I just sold my old digital. A Canon Powershot G2 that was 18-months old. I'd taken 19,500 photos with it. The equivalent of 540 rolls of film. About $6k worth by your calculations.
Before I'd gotten it I'd shot maybe five rolls of film. I was always wondering if a shot was worth it ("Girlfriend's cute, but I already got a picture last week.") Then, they'd sit around months or years until I got around to getting them developed. I didn't know anything about photography because what little feedback I had was so far after the shot.
Digital changed it all. I got a nice camera with all the manual overrides and full manual mode and I just started taking pictures. Lots of pictures.
I'd agree about cost, but I don't think there's anything good to be said about learning with film.
Film is expensive per shot. You're less likely to take pictures if every time you press the shutter you pay. You're even less likely to take five or ten shots of the same thing in order to experiment.
Even if you do, you need to have a log book (unless you're using a modern APS camera which tags the film with the settings - I think) in which you enter all the shot details, shutter, aperture, lens used, etc. Otherwise when you do get them developed - a week later for most people, or even a few hours later if you have your own darkroom, you won't remember why shot four of the group of ten is the best.
With digital you take the picture and get instant feedback, with histograms and blown highlights meters. If you don't like the result you change a setting and try again. Not only do you learn through the instant feedback, but a month later when you look at the photo on the computer and wonder why it looks the way it does you can check the exif data and see what it says. All your settings are recorded.
If you buy a digital SLR you'll also get amazingly low noise (the digital equivalent of grain). ISO 200 on my camera (Canon EOS 300D) is noise free, ISO 800 is barely noticable except in shadows, and ISO 1600 is acceptable. But, more than that, you can switch from ISO 100 to 1600 with the spin of a dial, giving you much more flexibility than with film.
You can also white-balance, producing shots where white objects look white, under vastly different lights. For white-balance and ISO in film you need to rewind, eject, and mark the spool, then load another roll and jump to the next empty frame. You then need to carry enough rolls for any occasion and keep track of your sessions as they cross rolls. A pain if you want timely development.
Most modern SLRs with auto-focus are blinding fast. Just today I fired off a four-shot burst at a crow that flew about ten feet over my head. At my first shot it was about 25' out, at the third shot is was straight above and half the distance, at the fourth shot it was already getting farther away.
This was with the Canon EOS 300D (Digital Rebel).
My old Canon Powershot G2 was a nice camera, but its autofocus was slow. With the rebel I can focus and shoot in about the same time (125ms) as the G2 took to shoot from focused and metered.
All you need to do is keep the focus point over the subject and your shots will be clear. And, if you can't, you turn off AF and do it yourself. With a manual camera you can't turn AF on, so you'd miss all of the fast-moving shots.
So maybe we should arrange a bounty for people who decompile and reverse-engineer closed-source drivers and post the intimate details about the hardware.
Not only would this let people write drivers for Linux based on the Windows drivers, but it would make hardware companies see that there's no value is keeping secrets... Surely ATI has looked at NVidia's drivers and vice-versa, it's perfectly legal.
We aren't saying you shouldn't set a schedule, just that you shouldn't set the schedule before you see the specs, unless you're allowed to prioritize the specs to meet the fixed timeline.
But, every company wants to set a release date that works for marketing (before next christmas) and then tell you what it needs to do.
Of course that's broken. They need to tell you what it needs to do and when they want it, then you prune the specs to achieve this.
Here's an analogy that's sure to work.
Give her a stick (unmarked dowel would work) and a pencil. Tell her to measure the length of the microwave (for a hypothetical new cupboard or something). Now, if she takes this stick with her to the kitchen store, she's got a pretty accurate idea of the size. So she orders a cabinet. They don't have any tape-measures, but they've got another stick, they lay it down next to her stick and put a mark on theirs at roughly the same place. Then they send the stick with a description of the product, to the guy who builds it, and he marks his own stick off of theirs, and sends their stick back. Etc.
At every step, each stick is being marked from the stick before. Errors accumulate and eventually the microwave might not fit the cabinet.
But, if she'd used a ruler and read out the next-largest measurement, she'd be able to tell someone the number very accurately. 84cm is still 84cm, even after being written down a few times. It works over the phone, etc.
The stick, and pencial mark, is analog. It's based on taking a measurement of something and trying to convey that based on the size of the stick, or the voltage of a signal. If each sampling and reproduction isn't perfect (and it can't be), error is introduced.
The measurement from the tape is digital. Discrete signals that aren't measured by size and as such, are less error prone. That 84cm is the same, written in huge block betters with a wide marker, or neatly lettered with a fine-point pen by an architect.
If someone receives a stick with a slightly blurry line where do they mark the next stick? But if someone receives a piece of paper with 84cm written on it and the 4 is a bit smudged, they simply erase it and write a neat 4 over it, re-enforcing the signal so to speak. Unless it's so corrupted that they can't read anything, each step will be made without errors.
Binary is just the fairly obvious next step. What two signals are the hardest to mistake? Direct opposites. On/Off, Hot/Cold, Black/White. If you have other signals, there's more chance of error.
But they need to have a rough idea of how to use Linux, or MS will just ignore them. To get an idea, they need to study how they'd go about it and the pros and cons. At this point, they've already made the first step towards switching, learning about alternatives.
But, in the end, even if they don't switch they let everyone else know that MS will give you the licenses at a drastically reduced price to keep you. So now everyone is going to be investigating and threatening to switch, and in doing so they'll learn about the benefits.
So when MS says no, when their bottom-line won't allow them to give any more sweet deals, these companies will already be partly ready to switch.
The main reason people still run WinNT is because it costs to upgrade. It's a terrible pain to use. With Linux upgrades are of course free and people will upgrade when a feature they want is available. Few users run 2.2 kernels, even on old machine, because 2.4 offered a lot of desktop-level fixes.
Modern Linuxes can be stripped down. I know a guy with a 486-50 Notebook with 32mb of ram. He installed Mandrake, I think, and simply put in an old windows manager. It's modular so he can choose. He might want the recent kernels and tools of a later distro, but not want KDE/Gnome. With Windows, you take the whole package and all you can do is tweak some of the graphics settings.
And as for planned obsolescence, yeah, Linux distros intend for users to upgrade. But they don't have lock-in, you can either download their distro, or another one, if you don't want to pay them for an upgrade. That means that plan new features and reasons people will want to upgrade, they don't plan to make the old system stop working well because of new file formats. If they did that, people would switch all right - to another distro.
I find that Linux is friendlier on old hardware because I can upgrade to a new OS, or a new component, and a new GUI isn't required. I could take an old 486 that's been running as a router for years and slap iptables onto it without having to upgrade the whole OS. As such, it's not going to increase the hardware requirements.
Ditto with a user machine. If they only need a few updates, or a new tool or two, it never requires a complete upgrade and then nothing but the new tool is different, all the old apps run exactly the same.
XP is fairly lightweight, compared to 2k, so it's closer to this, but when I had to upgrade an old Win98 print-server to Win2k because of a networking issue the performance was terrible. With Linux I'd have just installed a specific module.
Yeah. Like in a game of cards, you can reverse what you're doing before the next guy starts his move. At the point that he's actually done something based on your move, you're screwed.
If they ran up while you were reading the sales agreement and demanded an extra $5k, that's one thing. To demand it after you've all signed, that's not acceptable.
Ditto here, where people had made trades based on the mistake. IMHO, everything goes (ditch the whole hand) or you keep going and the person who made the mistake takes a loss.
Did the buy price pop up and indicate that it was because of a bug, or did it look like if someone had decided to bail on a stock because they had decided it was a bad deal.
The stock market is all about arbitrary prices of stocks that rarely have anything to do with the value of the companies they supposedly represent.
Besides, if they haven't tossed Ken Lay in jail for the rest of his life and taken away every single possession and auctioned it, what sort of message are they sending? That the best way to get rich is through outright fraud?
But it is a wildcard law. Just not versus other corporations who can afford to fight back.
I could easily see some Adobe-like company claiming that by decoding the XORed traffic on their drive controller that you defeated a copy protection feature, even if the device is a microwave or something, and bury you under paperwork unless you agree to settle for $10k and a gag order.
I mean, that's pretty much a template for a future Slashdot submission right there.
But of course, when entities will millions of dollars for legal fees can't simply pick on the poor, they at least invent slightly better legal justification.
Insightful? How about retarded? How else can you write a post directly contrary to all evidence?
A product using ext3 wouldn't have to be open sourced, any more than a product running on Linux. Any changes made to the filesystem would, but it's highly unlikely that you're going to have so grand an idea for a filesystem that your product hinges on it, and then have to implement it on top of someone else's filesystem.
Besides, using GPLed components basically prevents patent issues. By intentionally releasing something that requires you to agree that it is patent encumbered you pretty much give implicit free licenses to any patents that you may have on that code. Otherwise you didn't honor the contract you entered into with the original author of the software.
What would you prefer? Having to open source some tiny filesystem component you added to ext3, or having to pay Microsoft up to $250,000?
Most explosives contain their own oxygen (bound weakly in some molecule). Explosives combust so quickly (by definition, if the speed is below a certain point it's not "explosive") that they'd never get oxygen to the explosive and they'd put themselves out.
Many things that are only explosive when mixed with air are like this. Gasoline will (usually, don't try this at home) just burn if a bucket of it is lit, but if it has time to evaporate and fill the room with fumes, it'll blow up.
As an example, think of a bullet. It's in a fairly air-tight barrel, with a hunk of lead between it and the source of oxygen. Bullets will fire in space even. No air needed. If they didn't contain everything needed for their reaction, they'd be a lot less impressive.
Yeah. If people investigated monthly plans they'd realize that they're about the same price, once you figure in the phone purchase, they let you cancel at any time, and as the company has an incentive to keep you they tend to be much nicer to deal with.
I switched from Telus (a Canadian version of a baby bell) to Fido and got incredible rates, deals (free FidoFido calls and text messages, free evenings, etc) and a company that wasn't rude when I had to call them. All because I bought an unlocked phone that could be transfered to their competition. Lock-in is bad when Microsoft does it, it's also bad when the phone company tries it.
Now if only we had number portability, we'd be set.
These people are abusing any and all common resources they can find, as well as attacking anyone who tries to get in their way.
Why shouldn't they be killed? What possible reason could there be to justify their ongoing existence at the cost of everyone around them? It's not like they don't know they're hurting people - they're specifically setting out to do so.
I'm willing to give everyone the benefit of the doubt, in the beginning, but I don't see why it's unreasonable to think that people could abuse their neighbors long enough for that assumption to change to them having to prove their worth.
Previously used books were sold direct-to-consumers, but in a local area only. Now they're sold, often to resellers, globally. These resellers handle the searching and internet buying aspects and deliver the books to consumers.
So yes, they did add a middle-man, but they also increased the usefulness of the service. (Especially since the major buyers at library sales tend to be resellers anyways.)
But now, it's an open system. The online auctions don't discriminate. If you're willing to offer more than a reseller, you'll get the book. Eventually we'll solve e-commerce (to the point where grandmothers are comfortable purchasing products online) and the last use of the reseller will be gone, because the system of making a book available globally will be in place.
Should they be lost to history? No. But they should be sorted so that you don't stumble across them looking for modern advice.
What is the category when the book is history, instead of just detailing history?
That's a sad truth? Seems only fair after MS has put so many other companies out of business with their crimes.
You must not read the newspapers. While the parent poster's definition of epidemic might be a bit low, there are too many home invasions in the GVRD.
This sort of thing does happen because the criminals don't think there's much risk. They usually target older people and have a gun or knives. The victims are suprised and weaponless, if they can't call 911 quickly enough they're in trouble.
I'm not saying that guns would make this better, but consider that an 80-year old woman with a gun is the equal of a young thug with a gun, but with knives, or bats, or any improvised weaponry, she's far more likely to lose. Furthermore, gunshot wounds are far more fatal than knife wounds, a thug is risking his life going against someone with a gun, but probably only risking mild injury attacking the average person in their home.
This is why you specifically construct small elements that are provable, and string them together. You don't have to prove anything about a randomly created program.
You don't sit down and attempt to formally prove that Doom3 is "correct", you work on small pieces of it and then the logic that calls those pieces, etc.
If you assume a routine can only take certain inputs and produce certain outputs, use asserts (in production code, not just for testing) and if nothing else, at least catch your errors early before they cascade.
Yeah. If you want your business to score highly in the results, have something there worth searching for. Host forums and convince people to review the products you sell, or do something that differentiates between you and just another retailer.
Like Amazon for example. Their user reviews and such make their site a good one to look at when considering a product, and well, since I'm there anyways...