In what way does the ACLU not defend people's freedom of association?
Obviously this particular Limbaugh thing was trumped up, although I'd think Florida would be particularly friendly toward him. But that's exactly the point; the ACLU does defend people from trumped up charges. When someone's popular, they seldom need that kind of defense.
For the generic right wing complaints about the ACLU: who was it who represented Rush Limbaugh recently when Florida tried to get too aggressive with his medical records?
Sorry, this is one that Eric Raymond should have researched a bit more. Not because the interface he's talking about is any good, but because he's firing at the wrong target, as others have pointed out. I wouldn't expect someone who doesn't know Linux to figure this out, but Eric should have been able to tell the difference between a Red Hat hack and CUPS proper (at least the localhost:631 web interface).
While I haven't used it myself, the number of complaints about it on the linuxprinting.org forums (vs. the lack of complaints about Mandrake, SuSE, etc. in this regard) suggests that there's a problem. From my standpoint this is a real nuisance, since a lot of the people blame Gimp-Print for their problems (reasonably enough from their perspective -- I don't blame them for that). However, ESR should know better, and should be able to pick his targets more accurately.
Canon, which does no small digital business, continues to introduce new film cameras -- even low-end SLR's, which would seem to be the most vulnerable to competition from digital. They've recently introduced the Rebel K2 and Rebel G II which are both lower-end versions of the Rebel Ti (Canon has used the Rebel name in the US for about 13 years for their entry-level SLR). Evidently the Rebel Ti was getting just a bit too high-end for comfort. These are all film cameras, by the way.
As others have noted, Kodak getting out of film cameras means nothing. APS has been a well-deserved failure, and Kodak really hasn't built any interesting 35 mm cameras lately.
(I just got a Rebel Digital, which is based on the Rebel Ti body. It's a much, much better camera than my first SLR, a Rebel XS. It's more solid, has better controls, a metal lens mount, much shorter shutter lag and faster drive and in some ways a better autofocus system than my EOS 1N, their previous top of the line prior to the 1V. The controls are still deliberately dumbed down so that they don't completely destroy the market for the Elan, but both the film and digital versions of this camera are very innovative indeed.)
I admit that I haven't in a while, but this is excellent news.
When I was young, there were none of the Star Wars kits and such. It took real imagination to devise and build interesting toys from parts.
My wife has bought a few of the Star Wars type kits for her nephew. There's nothing much to it -- read the directions, figure out which piece is which, put them all together, and that's it. Pay your $20*/5, spend 30 minutes throwing the pieces together (of course, I do all the work), nephew spends 15 minutes with it, and it's done for.
This took some real courage on their part, rebelling against the entertainment-retail complex. Hopefully there are still kids and parents who can think for themselves and don't have to live lives of tie-ins to someone else's creation.
If they really want to do something different, an interface kit to Tinkertoy or Erector Set would be great. Then people could combine these three classics to build new toys combining all of their strengths. That is, if they still make Erector Set any more.
If escputil -n produced gaps, the problem was with the printer. It sends a very simple remote mode command to the printer that tells it to "print your nozzle check". The nozzle check is printed from firmware. I say this as the project lead for Gimp-Print and the author of escputil (which is part of Gimp-Print).
If you're having trouble configuring printers, my advice is to install CUPS, Gimp-Print 4.2.5, and use the localhost:631 interface to the browser.
The problem is that every moment spent looking at the speedometer is time spent not paying attention to the road. The absolute speed you're traveling at has little impact on safety except for the consequences if you do get into an accident, or if it's too fast for the conditions (which a static speed limit doesn't tell you). What's usually more important is the speed you're doing relative to surrounding traffic.
If the signal to the driver is distracting, it may distract his attention at exactly the wrong moment and cause an accident. A loud beep or flashing light suggests an alarm of an immediate hazard, which this isn't.
I think that power requirement is soon going to be more of an issue than storage capacity for these things, since battery technology just isn't advancing that fast. I've thought for a while that the ideal digital camera would use an 80 GB laptop drive, but that the problem there is the power requirement. If CF uses a lot less power (and I don't know if it does), that could be the real advantage, even if the storage is somewhat less.
An EOS 1Ds is 11 Mpixels; that works out to somewhere between 33 and 66 Mbytes/image (for raw images), depending upon the bit depth. A 4 GB card works out to roughly 64 or 128 images, which isn't even close to a day's shooting. For the (current) price, that probably doesn't yield enough of an advantage over film capacity per roll for most cases, although for underwater photography (where you can't change the roll) that might be enough to matter.
If you're going to use lower resolution, though (and most purposes don't require the full 11 Mpixels), then this kind of capacity makes a big difference. I suspect this would make a lot of sense for sports photographers and other journalists, who typically don't need absolute top quality but who would love to shoot hundreds of shots without having to reload. Event photographers (wedding photographers) could use this for candids (which is really a kind of journalism in a lot of cases), where getting the shot is more important than absolute resolution.
I still like the idea of an 80 GB laptop drive for this purpose, but it does have practicality issues -- power, and to some extent size. Journalists could carry around a couple of these 4 GB things, and swap one out to the laptop to unload while using the second one. For my preferred shooting style (landscape photography), though, this isn't enough capacity, although 80 GB would be.
It would surely help raise the standards of discussion if opponents of the GPL would actually take the time to read and understand it (and start by understanding basic properties of copyright, to boot). Knocking down straw men is a weak form of argument. IANAL, but this case is not going to hinge on the validity of the GPL. The case that may hold water for SCO will hinge upon whether IBM redistributed code that they did not have a right to, which is quite independent of the license under which it was (purportedly illegally) redistributed.
The GPL does not say "someone somewhere combined the two, and so now my license applies to both". It can't, legally, since someone can't agree to something that he does not have a legal right to agree to. If you combine source code that you have the right to with source code that you don't, you don't have the right to apply any license at all to it (or indeed, even to do it at all). So it's completely irrelevant whether the first piece of code is GPL, BSD, public domain, community source, shared source, or whatnot, if you don't have a legal right (through license or possession of copyright) to the second piece of code, you can't legally combine the two. All of this is spelled out very clearly in section 7 (see http://www.fsf.org/licenses/gpl.txt).
SCO's public claims are vague and confusing (IMHO intentionally so), but the most far-reaching claims seem to be that all modern operating systems embody concepts from UNIX, and that they are therefore all derivatives of UNIX, and therefore SCO owns all of them. If that were to be taken literally on its face, it would imply that everything that Linus, Alan, Microsoft, et al. have written is actually copyright SCO, which is sheer nonsense. For one, "derivative" is being used in a confusing way here. From a copyright perspective, "derivative" has a very specific meaning, and since concepts aren't subject to copyright (only expression is), the use of "concepts" from System V doesn't mean that anything was actually copied in a copyright sense. Secondly, copyright violation in the form of distribution of a derived work doesn't mean that the copyright automatically belongs to the author of the first work. The copyrights to various parts of Linux still belong to whoever wrote them, including (hypothetically) SCO. If multiple people own the copyrights to a work, then they all have to agree in order for it to be legally distributed.
Presumably, what SCO wants to convince people of is that they are going to sue the creators of Linux, and for damages insist that the copyrights to the entire source to Linux be turned over to them. Whether this is something they would have a snowball's chance in hell of succeeding in is something a copyright lawyer would have to answer, but from what I've read a copyright holder has a duty to warn someone of an inadvertent violation so that they can correct it. SCO has done nothing of the sort.
SCO further weakens their position by continuing to distribute Linux from their web site. Whatever they may or may not hope to win in court, they incontrovertibly do not hold the copyrights to all of Linux, so they only have a right to distribute it under the terms of its license (the GPL). So at the very least, they're implicitly warranting that they can legally distribute whatever that is under the terms of the GPL, and if they went for the nuclear option they would surely be countersued for copyright violation by Linus and other major copyright holders. The only way I can see they would have out of *that* would be to represent that everything they were distributing from their web site as Linux was in fact legal, and so someone could start freely from there (and incidentally more or less destroy their own case). So they're in a bit of a catch-22 (of their own making) -- admit that what they're distributing is legal, and that therefore they have no case, or admit that it's illegal, in which case they're wide open to copyright violation themselves.
Depends on the design point for the amplifier. If you need 20 watts most of the time (which with typical speakers in a typical residential environment is really quite loud), but you occasionally need very brief spikes to 100 watts, it's cheaper and just about as effective to build an amp that can only do 40-50 watts, but has some large caps that allow maybe 10-100 ms bursts of 150-200 watts. It will be smaller and run cooler, to boot.
NAD did this about 20 years ago (for all I know they still do). I have a couple of NAD amps that sound quite good, even at only 40-50 watts/channel (continuous RMS).
if they would be even halfway creative and use the ROM on the chip for something useful, such as storing color or viscosity information about the inks. This would enable companies to periodically reformulate their inks (to increase longevity, saturation, or even just due to manufacturing variation) while preserving perfect matching from cartridge to cartridge. They could also license this to third parties to enable the sale of quadtone, hextone, or true six color inks (instead of the usual light cyan and light magenta, the other two inks might be orange and blue, or spot colors). This would add real value, but instead this technology is being used for the singularly unimaginative use of vendor lock-in. It might not be good for third-party drivers (such as Gimp-Print), but if printer vendors used these chips for useful purposes, it actually would benefit users.
People actually should be careful about third-party cartridges and refill kits; some of these are very bad, and if you're not careful with refill kits you can cause problems either by introducing air bubbles or debris. Some printers (Canon and HP that I know of) include the print head with the cartridge, and the head isn't designed for a very long life; the quality will probably degrade after a few refills. Epson printers use a long-life head technology, but the flip side is that if you damage the heads, you're either looking at an expensive repair or a new printer. Refill kits are also messy. However, that really should be for the user to decide.
Lawyers are human, too. Mistakes do happen. When I read our purchase & sale, I found about 3 errors, one or two of which were potentially significant. These were things that our lawyer -- not the seller's -- had missed, too.
Apparently, reading all the documents at the closing is considered excessively time consuming. Big deal; it's a huge purchase, and it's always possible that there will be mistakes. Read everything.
There absolutely is a market for GPL'ed software. Just ask Red Hat, SuSE, Mandrake, et al. They're very much in the business of selling shrink wrap software.
We can argue back and forth all day about whether they're really selling "service" (saving you the effort of downloading, or sharing, or integrating, or whatnot), but at the end of the day, they are selling a product. Maybe they're not making as much revenue as Microsoft, but that's not the question at hand. There is a market for GPL software, even if it's less lucrative on the seller side.
Rather than having window.open() return a null handle, have it return a real handle, but simply don't create the window. Better yet, have it optionally load the contents of the window, so the remote site never even knows that the window simply was never popped up.
In the short run, there's no problem. In the longer run, if you choose not to use proprietary software, you may have difficulty interoperating with other people if you're "locked out".
Considering how expensive a BitKeeper license is, the choice is either to use BitKeeper free or contribute to CVS or other alternatives. $5800 or whatever I saw quoted for a single RTU is prohibitive for most people.
A freedom that can only be enjoyed upon payment of a substantial sum of money isn't a freedom. It's a privilege, which is very different.
Again, the license *doesn't* simply say "you can't use BitKeeper (no-cost version) to assist in developing a competitor". It's much more pernicious. And since you can only use the latest version of BitKeeper in this way, it can be relicensed at any time with no recourse.
There's no such thing as "objective evaluation". To a coarse approximation, there's a number of factors, and weights are applied to each factor. The choice of weights is hardly "objective".
Just for example, there's the factor of "making money for the creator of the software". RMS, needless to say, gives this a very low weight indeed. There's also the factor of "what preserves maximum control for the creator", to which he presumably assigns zero or negative weight. There's also the matter of "what preserves maximum choice and freedom for the end user", which RMS gives a very high weight.
RMS does try to evaluate which license actually does effectively preserve maximum choice and freedom for the end user, and that choice is not always the GPL. In the case of Ogg Vorbis, he even recommended use of a BSD-style license.
RMS has a very clear goal that trumps all others, preservation of freedom of the user, and within that framework he's willing to consider different strategies and tactics. He isn't willing to compromise that goal.
Note that there's also the matter of what's better in the short vs. the long term. In the short term, using BitKeeper may make life easier for kernel development; in the longer term, it may have serious side effects that may be hard to get out of (lock-in).
Both cirrus and cirrocumulus are formed from tiny ice crystals (cirrocumulus is when small convective cells form within the cloud). A lot of cirrus is actually of convective origin, the blowoff from thunderstorms.
Vorticity actually means rotation (turning) of the air, not lifting. However, positive (cyclonic) vorticity near the surface is associated with lifting of the air. That doesn't necessarily mean dramatic convective lifting that produces thunderstorms, but air will be lifted in those regions.
Hail forms in thunderstorms with strong updrafts. These updrafts are needed to keep the hailstones from simply falling out of the cloud. In order to get giant hailstones, the updraft has to be very strong indeed.
A 35 pound chunk of ice is on the order of a 10-inch cube (if it's really solid). That's much bigger than any hailstone ever recorded; the biggest known was about 7 inches long and 4-5 inches on the short axis; hailstones are rarely that solid, and it probably didn't weigh more than 5 pounds or thereabouts. I think it fell in Coffeyville, Kansas during a severe thunderstorm (surely a supercell, with a very persistent, rotating updraft, that may well have spawned a tornado). This isn't rocket science; think about what the terminal velocity of a 35 pound chunk of ice is, and the updraft has to be close to that in velocity.
Any situation where the air is rising that violently is either going to have a very obvious cloud (read: thunderstorm) associated with it; the heat released by the condensation of the water makes the air more buoyant, and hence increases the lift. A dry thermal isn't going to approach that kind of velocity, and even if it did, it would be...well...dry.
What all of that basically amounts to is...well...that dog don't hunt. The only realistic source of that kind of thing is water being dumped by airplanes at high enough altitudes so that it has time to freeze before reaching the surface. A chunk of ice that big should be easy to save long enough to be examined; they should give a few samples to NOAA, say, and let them inspect it.
35 mm today is probably at least as capable as medium format was 20 years ago, and people were perfectly satisfied with medium format back then. How many people really care about a 16x20 enlargement of a wedding shot, anyway? I've made numerous 16x24's and some 20x30's from 35mm (the latter, believe it or not, was from ISO 800 film, and my best 16x24 was from ISO 400 film with a Tamrom 28-200, hand held -- not the world's sharpest lens), and while they certainly aren't tack sharp, they're perfectly good when viewed from more than a foot or two away. These shots were taken about 5 years ago, and film has improved since then.
Besides which, the prevailing wisdom is that portraits of women are best done with a soft filter, anyway (I absolutely hate to sound sexist like that, but the cultural norm is that any roughness on a woman's skin is bad, while on a man, it's if anything desirable). A lot of wedding photographers go further and use fog filters for a really dreamy look. Granted, there's a very real difference between out of focus and a soft filter -- a soft filter will give a perfectly sharp print, but with a halo surrounding the highlights.
I'm not entirely convinced that 6 Mp is really enough yet, although Canon's coming out with the EOS 1Ds very soon, which will have 11 Mp. However, it's probably good enough for most purposes. Keep in mind that a digital camera, or fast 35 mm camera (such as the EOS 1V, the film version of the same platform that the 1D is based on) is a lot more lithe than a medium format rig. This is nothing to sneer at; the shot that was taken is always better than the one that was missed.
Slashdot Mirror
In what way does the ACLU not defend people's freedom of association?
Obviously this particular Limbaugh thing was trumped up, although I'd think Florida would be particularly friendly toward him. But that's exactly the point; the ACLU does defend people from trumped up charges. When someone's popular, they seldom need that kind of defense.
Before you accuse the ACLU of being anti-religion, read this article, about the ACLU representing a high school student whose quote from a bible verse was censored from her yearbook. The ACLU takes both parts of the establishment clause seriously, both the one forbidding establishment of religion and the one prohibiting free exercise of it.
For the generic right wing complaints about the ACLU: who was it who represented Rush Limbaugh recently when Florida tried to get too aggressive with his medical records?
Sorry, this is one that Eric Raymond should have researched a bit more. Not because the interface he's talking about is any good, but because he's firing at the wrong target, as others have pointed out. I wouldn't expect someone who doesn't know Linux to figure this out, but Eric should have been able to tell the difference between a Red Hat hack and CUPS proper (at least the localhost:631 web interface).
While I haven't used it myself, the number of complaints about it on the linuxprinting.org forums (vs. the lack of complaints about Mandrake, SuSE, etc. in this regard) suggests that there's a problem. From my standpoint this is a real nuisance, since a lot of the people blame Gimp-Print for their problems (reasonably enough from their perspective -- I don't blame them for that). However, ESR should know better, and should be able to pick his targets more accurately.
Canon, which does no small digital business, continues to introduce new film cameras -- even low-end SLR's, which would seem to be the most vulnerable to competition from digital. They've recently introduced the Rebel K2 and Rebel G II which are both lower-end versions of the Rebel Ti (Canon has used the Rebel name in the US for about 13 years for their entry-level SLR). Evidently the Rebel Ti was getting just a bit too high-end for comfort. These are all film cameras, by the way.
As others have noted, Kodak getting out of film cameras means nothing. APS has been a well-deserved failure, and Kodak really hasn't built any interesting 35 mm cameras lately.
(I just got a Rebel Digital, which is based on the Rebel Ti body. It's a much, much better camera than my first SLR, a Rebel XS. It's more solid, has better controls, a metal lens mount, much shorter shutter lag and faster drive and in some ways a better autofocus system than my EOS 1N, their previous top of the line prior to the 1V. The controls are still deliberately dumbed down so that they don't completely destroy the market for the Elan, but both the film and digital versions of this camera are very innovative indeed.)
I admit that I haven't in a while, but this is excellent news.
When I was young, there were none of the Star Wars kits and such. It took real imagination to devise and build interesting toys from parts.
My wife has bought a few of the Star Wars type kits for her nephew. There's nothing much to it -- read the directions, figure out which piece is which, put them all together, and that's it. Pay your $20*/5, spend 30 minutes throwing the pieces together (of course, I do all the work), nephew spends 15 minutes with it, and it's done for.
This took some real courage on their part, rebelling against the entertainment-retail complex. Hopefully there are still kids and parents who can think for themselves and don't have to live lives of tie-ins to someone else's creation.
If they really want to do something different, an interface kit to Tinkertoy or Erector Set would be great. Then people could combine these three classics to build new toys combining all of their strengths. That is, if they still make Erector Set any more.
Those would lend themselves very well to movies. Better yet, they could actually do sequels based on the later robot novels.
I'd like to see Sourceforge work on real improvements rather than this. The bug tracking system, in particular, is an absolute travesty.
If escputil -n produced gaps, the problem was with the printer. It sends a very simple remote mode command to the printer that tells it to "print your nozzle check". The nozzle check is printed from firmware. I say this as the project lead for Gimp-Print and the author of escputil (which is part of Gimp-Print).
If you're having trouble configuring printers, my advice is to install CUPS, Gimp-Print 4.2.5, and use the localhost:631 interface to the browser.
The problem is that every moment spent looking at the speedometer is time spent not paying attention to the road. The absolute speed you're traveling at has little impact on safety except for the consequences if you do get into an accident, or if it's too fast for the conditions (which a static speed limit doesn't tell you). What's usually more important is the speed you're doing relative to surrounding traffic.
If the signal to the driver is distracting, it may distract his attention at exactly the wrong moment and cause an accident. A loud beep or flashing light suggests an alarm of an immediate hazard, which this isn't.
No, I don't have Windows handy, so the built-in Access ODBC driver won't help me. I'm looking for a pure Linux solution to dumping out a .mdb file.
Will that import directly from a file? I don't have Access available to me at all now, all I have is the .mdb file.
So a bit off-topic, but is there anything around that can import from an Access .mdb file?
I think that power requirement is soon going to be more of an issue than storage capacity for these things, since battery technology just isn't advancing that fast. I've thought for a while that the ideal digital camera would use an 80 GB laptop drive, but that the problem there is the power requirement. If CF uses a lot less power (and I don't know if it does), that could be the real advantage, even if the storage is somewhat less.
An EOS 1Ds is 11 Mpixels; that works out to somewhere between 33 and 66 Mbytes/image (for raw images), depending upon the bit depth. A 4 GB card works out to roughly 64 or 128 images, which isn't even close to a day's shooting. For the (current) price, that probably doesn't yield enough of an advantage over film capacity per roll for most cases, although for underwater photography (where you can't change the roll) that might be enough to matter.
If you're going to use lower resolution, though (and most purposes don't require the full 11 Mpixels), then this kind of capacity makes a big difference. I suspect this would make a lot of sense for sports photographers and other journalists, who typically don't need absolute top quality but who would love to shoot hundreds of shots without having to reload. Event photographers (wedding photographers) could use this for candids (which is really a kind of journalism in a lot of cases), where getting the shot is more important than absolute resolution.
I still like the idea of an 80 GB laptop drive for this purpose, but it does have practicality issues -- power, and to some extent size. Journalists could carry around a couple of these 4 GB things, and swap one out to the laptop to unload while using the second one. For my preferred shooting style (landscape photography), though, this isn't enough capacity, although 80 GB would be.
It would surely help raise the standards of discussion if opponents of the GPL would actually take the time to read and understand it (and start by understanding basic properties of copyright, to boot). Knocking down straw men is a weak form of argument. IANAL, but this case is not going to hinge on the validity of the GPL. The case that may hold water for SCO will hinge upon whether IBM redistributed code that they did not have a right to, which is quite independent of the license under which it was (purportedly illegally) redistributed.
The GPL does not say "someone somewhere combined the two, and so now my license applies to both". It can't, legally, since someone can't agree to something that he does not have a legal right to agree to. If you combine source code that you have the right to with source code that you don't, you don't have the right to apply any license at all to it (or indeed, even to do it at all). So it's completely irrelevant whether the first piece of code is GPL, BSD, public domain, community source, shared source, or whatnot, if you don't have a legal right (through license or possession of copyright) to the second piece of code, you can't legally combine the two. All of this is spelled out very clearly in section 7 (see http://www.fsf.org/licenses/gpl.txt).
SCO's public claims are vague and confusing (IMHO intentionally so), but the most far-reaching claims seem to be that all modern operating systems embody concepts from UNIX, and that they are therefore all derivatives of UNIX, and therefore SCO owns all of them. If that were to be taken literally on its face, it would imply that everything that Linus, Alan, Microsoft, et al. have written is actually copyright SCO, which is sheer nonsense. For one, "derivative" is being used in a confusing way here. From a copyright perspective, "derivative" has a very specific meaning, and since concepts aren't subject to copyright (only expression is), the use of "concepts" from System V doesn't mean that anything was actually copied in a copyright sense. Secondly, copyright violation in the form of distribution of a derived work doesn't mean that the copyright automatically belongs to the author of the first work. The copyrights to various parts of Linux still belong to whoever wrote them, including (hypothetically) SCO. If multiple people own the copyrights to a work, then they all have to agree in order for it to be legally distributed.
Presumably, what SCO wants to convince people of is that they are going to sue the creators of Linux, and for damages insist that the copyrights to the entire source to Linux be turned over to them. Whether this is something they would have a snowball's chance in hell of succeeding in is something a copyright lawyer would have to answer, but from what I've read a copyright holder has a duty to warn someone of an inadvertent violation so that they can correct it. SCO has done nothing of the sort.
SCO further weakens their position by continuing to distribute Linux from their web site. Whatever they may or may not hope to win in court, they incontrovertibly do not hold the copyrights to all of Linux, so they only have a right to distribute it under the terms of its license (the GPL). So at the very least, they're implicitly warranting that they can legally distribute whatever that is under the terms of the GPL, and if they went for the nuclear option they would surely be countersued for copyright violation by Linus and other major copyright holders. The only way I can see they would have out of *that* would be to represent that everything they were distributing from their web site as Linux was in fact legal, and so someone could start freely from there (and incidentally more or less destroy their own case). So they're in a bit of a catch-22 (of their own making) -- admit that what they're distributing is legal, and that therefore they have no case, or admit that it's illegal, in which case they're wide open to copyright violation themselves.
I've yet to see anything t
Depends on the design point for the amplifier. If you need 20 watts most of the time (which with typical speakers in a typical residential environment is really quite loud), but you occasionally need very brief spikes to 100 watts, it's cheaper and just about as effective to build an amp that can only do 40-50 watts, but has some large caps that allow maybe 10-100 ms bursts of 150-200 watts. It will be smaller and run cooler, to boot.
NAD did this about 20 years ago (for all I know they still do). I have a couple of NAD amps that sound quite good, even at only 40-50 watts/channel (continuous RMS).
if they would be even halfway creative and use the ROM on the chip for something useful, such as storing color or viscosity information about the inks. This would enable companies to periodically reformulate their inks (to increase longevity, saturation, or even just due to manufacturing variation) while preserving perfect matching from cartridge to cartridge. They could also license this to third parties to enable the sale of quadtone, hextone, or true six color inks (instead of the usual light cyan and light magenta, the other two inks might be orange and blue, or spot colors). This would add real value, but instead this technology is being used for the singularly unimaginative use of vendor lock-in. It might not be good for third-party drivers (such as Gimp-Print), but if printer vendors used these chips for useful purposes, it actually would benefit users.
People actually should be careful about third-party cartridges and refill kits; some of these are very bad, and if you're not careful with refill kits you can cause problems either by introducing air bubbles or debris. Some printers (Canon and HP that I know of) include the print head with the cartridge, and the head isn't designed for a very long life; the quality will probably degrade after a few refills. Epson printers use a long-life head technology, but the flip side is that if you damage the heads, you're either looking at an expensive repair or a new printer. Refill kits are also messy. However, that really should be for the user to decide.
Lawyers are human, too. Mistakes do happen. When I read our purchase & sale, I found about 3 errors, one or two of which were potentially significant. These were things that our lawyer -- not the seller's -- had missed, too.
Apparently, reading all the documents at the closing is considered excessively time consuming. Big deal; it's a huge purchase, and it's always possible that there will be mistakes. Read everything.
There absolutely is a market for GPL'ed software. Just ask Red Hat, SuSE, Mandrake, et al. They're very much in the business of selling shrink wrap software.
We can argue back and forth all day about whether they're really selling "service" (saving you the effort of downloading, or sharing, or integrating, or whatnot), but at the end of the day, they are selling a product. Maybe they're not making as much revenue as Microsoft, but that's not the question at hand. There is a market for GPL software, even if it's less lucrative on the seller side.
Rather than having window.open() return a null handle, have it return a real handle, but simply don't create the window. Better yet, have it optionally load the contents of the window, so the remote site never even knows that the window simply was never popped up.
In the short run, there's no problem. In the longer run, if you choose not to use proprietary software, you may have difficulty interoperating with other people if you're "locked out".
Considering how expensive a BitKeeper license is, the choice is either to use BitKeeper free or contribute to CVS or other alternatives. $5800 or whatever I saw quoted for a single RTU is prohibitive for most people.
A freedom that can only be enjoyed upon payment of a substantial sum of money isn't a freedom. It's a privilege, which is very different.
Again, the license *doesn't* simply say "you can't use BitKeeper (no-cost version) to assist in developing a competitor". It's much more pernicious. And since you can only use the latest version of BitKeeper in this way, it can be relicensed at any time with no recourse.
There's no such thing as "objective evaluation". To a coarse approximation, there's a number of factors, and weights are applied to each factor. The choice of weights is hardly "objective".
Just for example, there's the factor of "making money for the creator of the software". RMS, needless to say, gives this a very low weight indeed. There's also the factor of "what preserves maximum control for the creator", to which he presumably assigns zero or negative weight. There's also the matter of "what preserves maximum choice and freedom for the end user", which RMS gives a very high weight.
RMS does try to evaluate which license actually does effectively preserve maximum choice and freedom for the end user, and that choice is not always the GPL. In the case of Ogg Vorbis, he even recommended use of a BSD-style license.
RMS has a very clear goal that trumps all others, preservation of freedom of the user, and within that framework he's willing to consider different strategies and tactics. He isn't willing to compromise that goal.
Note that there's also the matter of what's better in the short vs. the long term. In the short term, using BitKeeper may make life easier for kernel development; in the longer term, it may have serious side effects that may be hard to get out of (lock-in).
Geologist != meteorologist.
Both cirrus and cirrocumulus are formed from tiny ice crystals (cirrocumulus is when small convective cells form within the cloud). A lot of cirrus is actually of convective origin, the blowoff from thunderstorms.
Vorticity actually means rotation (turning) of the air, not lifting. However, positive (cyclonic) vorticity near the surface is associated with lifting of the air. That doesn't necessarily mean dramatic convective lifting that produces thunderstorms, but air will be lifted in those regions.
Hail forms in thunderstorms with strong updrafts. These updrafts are needed to keep the hailstones from simply falling out of the cloud. In order to get giant hailstones, the updraft has to be very strong indeed.
A 35 pound chunk of ice is on the order of a 10-inch cube (if it's really solid). That's much bigger than any hailstone ever recorded; the biggest known was about 7 inches long and 4-5 inches on the short axis; hailstones are rarely that solid, and it probably didn't weigh more than 5 pounds or thereabouts. I think it fell in Coffeyville, Kansas during a severe thunderstorm (surely a supercell, with a very persistent, rotating updraft, that may well have spawned a tornado). This isn't rocket science; think about what the terminal velocity of a 35 pound chunk of ice is, and the updraft has to be close to that in velocity.
Any situation where the air is rising that violently is either going to have a very obvious cloud (read: thunderstorm) associated with it; the heat released by the condensation of the water makes the air more buoyant, and hence increases the lift. A dry thermal isn't going to approach that kind of velocity, and even if it did, it would be...well...dry.
What all of that basically amounts to is...well...that dog don't hunt. The only realistic source of that kind of thing is water being dumped by airplanes at high enough altitudes so that it has time to freeze before reaching the surface. A chunk of ice that big should be easy to save long enough to be examined; they should give a few samples to NOAA, say, and let them inspect it.
35 mm today is probably at least as capable as medium format was 20 years ago, and people were perfectly satisfied with medium format back then. How many people really care about a 16x20 enlargement of a wedding shot, anyway? I've made numerous 16x24's and some 20x30's from 35mm (the latter, believe it or not, was from ISO 800 film, and my best 16x24 was from ISO 400 film with a Tamrom 28-200, hand held -- not the world's sharpest lens), and while they certainly aren't tack sharp, they're perfectly good when viewed from more than a foot or two away. These shots were taken about 5 years ago, and film has improved since then.
Besides which, the prevailing wisdom is that portraits of women are best done with a soft filter, anyway (I absolutely hate to sound sexist like that, but the cultural norm is that any roughness on a woman's skin is bad, while on a man, it's if anything desirable). A lot of wedding photographers go further and use fog filters for a really dreamy look. Granted, there's a very real difference between out of focus and a soft filter -- a soft filter will give a perfectly sharp print, but with a halo surrounding the highlights.
I'm not entirely convinced that 6 Mp is really enough yet, although Canon's coming out with the EOS 1Ds very soon, which will have 11 Mp. However, it's probably good enough for most purposes. Keep in mind that a digital camera, or fast 35 mm camera (such as the EOS 1V, the film version of the same platform that the 1D is based on) is a lot more lithe than a medium format rig. This is nothing to sneer at; the shot that was taken is always better than the one that was missed.