It's remarkable how much that gets reported turns out to be unquestionably false. We hear reports of cows with accents, and it turns out that all of the quotes are from people who were mostly directly contradicting the story, but said a few things that could be totally rewritten to suggest support for it. We hear about school districts allowing text message slang in exams, and it turns out that this is entirely reporters extrapolating from schools not flunking students who make spelling mistakes when writing correct answers. We hear that women use 3 times as many words as men, and this turns out to be directly contradicted by every study actually carried out, from major academic studies to curious people handing tape recorders to a pair of people and then counting.
I hardly think it's surprising any more that successful completion of a project is reported as the project being scrapped. It's almost surprising that we didn't get reports on the recent election of the Republicans keeping control of both houses of Congress for another two months.
I think the real problem is that Firefox doesn't make enough use of the extension mechanism. It's clearly sufficiently powerful to do all sorts of interesting things with, but the default install of Firefox doesn't have any extensions at all. You can change a lot of the appearance of Firefox with themes, but it doesn't come with any themes, either (the "default theme" thing is actually not handled as a theme, but is just the built-in defaults for things, which is why you can't install 1.5's default theme on 2.0).
What they should do is split most of the functionality into extensions and all of the appearance into themes, and ship those extensions and themes with the browser. Then you'd be able to choose exactly which features you want from each version. And, since extensions can be updated independantly of the rest of the browser, it would be no problem if there was a bug in deciding that a word had been typed completely for the inline spellchecker; they'd release a new version of the spellchecker whenever they fixed it.
I think the complaint is that Firefox 1.5 is better than Firefox 2.0. If I try out a new Honda, and it's less pleasant to drive than my old Honda, I stick with my current car, and I worry about what I'll do if it breaks down.
All of your items are good reasons to use Firefox 1.5. Firefox 2.0 doesn't add anything I find compelling, and it adds non-removable junk to the navigation toolbar. It's a legitimate worry that I'll eventually have to go to a worse browser than Firefox 1.5 to have the latest security updates, because the Firefox team will only support 2.0 and their trademark policies prohibit other people from fixing things.
The only problem with a manager that's good at what the workers are doing is if the boss doesn't want to quit doing what he's good at. This is a problem because a manager has no attention to devote to work (being the person who filters useless distractions) and little time (having the full-time managing job and twice as many meetings as anybody else). So, no matter how good a techie the person is, the output is lousy, ill-considered, and never ready.
So the answer is really that a manager can't really be a techie, but a manager should be a former techie.
Also, a manager only needs a good approximation (but it can't be a bad approximation) of skill in the field. There's a level of skill where you can't solve a problem yourself, but you know whether it's fundamentally impossible, impractical, or just difficult, and that's ideal for a manager. (The worst thing is if the manager knows the problem can be solved, but nobody else on the team is good enough to solve it and the manager is too busy.)
Of course there's also the Peter Principle; there are plenty of cases of techies without any management skill at all promoted to management positions on the basis of seniority and great technical skill, such that they don't have the necessary skills for their actual job, are too valuable to let go or demote, and don't have the time to do the work they are better at than anyone else.
There are a bunch of things that make this less useful for doing weird things:
You can't send a message back in time. You can only receive a message from the future. That is, you can only send a message back in time to a point where you had arranged to get it. It's like an box that you take stuff out of before you put it in; things go back in time to the point where you took stuff out, not to any other time. So there's no issue with the fact that we're not getting messages from the future; the time before the time machine is invented is inaccessible.
You can't tell what it says in the past. This is where quantum is weird. Basically, what happens is that person A receives the message, which is a series of dots to put in a picture. It looks like random static. Then person B sends the message, which consists of choosing, for each dot, "bell" or "bars". Then they talk to each other, and they find that if you look at only the "bell" dots, the picture is a bell, and if you look at the "bars" dots, it's a set of bars. Since all of the data is collected by A before B chooses, they have to come to the conclusion that something really weird is going on, and the choice later clearly affects the data that was already written down. But they can only come to this conclusion after the experiment is over; before the message is sent, the received message can't be interpreted, although all of the observations can be taken.
This of it like this magic trick: the audience gets a deck of cards with a variety of backs which they examine in detail. A volunteer on stage shuffled a second deck of cards, writes down a few numbers between 1 and 52, and draws the cards with the given numbers (i.e., for 10, draws the 10th card in the shuffled deck). When the volunteer announces the set of names, they all turn out to have the same backs in the audience's deck. The volunteer chose freely, the deck was really random, and the audience saw the fronts and backs of all of the cards in their deck before the choice was made. If the trick is repeated with fresh decks, it always works. We have to conclude that the volunteer is affecting the construction of the deck in the past, but we're only impressed after it's all over, and we have no idea what the volunteer is going to choose in advance. Even if we agree on a set of numbers to pick if the stock market goes up and a different set to pick if it goes down, we can't tell by looking at the audience's deck which it will be, but the trick still works.
But that didn't involve any code whose copyright Sun held. It was stopped because Sun has a trademark on Java which they license only to implementations which pass certain conformance tests (which Microsoft's obviously didn't). It would have made no difference to that case if Sun's JDK were available under the GPL or even under the BSD license, because the complaint wasn't over the copyright on the implementation.
As far as I can tell, the trademark restrictions are still in place, so you can take Sun's JDK and create a derived work that behaves differently, but you can't call it Java unless it still passes the test suite. I expect there to be a number of research and special-purpose derivatives of the Sun JDK before long, but I expect them to have names that don't use the trademark and actually not be "Java" (like C++ isn't C). I also expect people to improve the Sun JVM in a variety of ways that don't change the language (e.g., inline ArrayList.get in methods that call get on a List, with a check on the class of the List object in the native code that the JIT compiler produces, or inline final wrapper methods at call sites, to avoid the function call overheads). I also expect the non-"Java" derivatives to be a useful source of JSRs, and to make the whole JSR process more efficient, because people with proposals can implement them without a huge amount of work making an otherwise functional implementation.
That's why they're using a closed GSM module. It's like using a cell modem with your laptop, except that it's internal. The carrier doesn't care what you say to your cell modem, because the device is certified to behave appropriately regardless. And the GSM modules are high volume, because a wide variety of applications use them, and the modules don't get new versions all the time to add new UI bells and whistles.
So you're right that part of the device is locked down and tamper-proof, but it's running on an entirely separate chipset (probably only connected by a serial port, actually).
You can't really call Lieberman a Democrat, because he was elected without the Democrat nomination. He was running an an independant. On the other hand, he claims he's going to caucus with the Democrats, so he counts as a Democrat for the purposes of control of the Senate.
As far as getting anything done, it's going to be 2 years of getting bills vetoed, most likely. For that matter, it's likely to be a few weeks of passing all the bills on the Democrat agenda, getting them all vetoed, followed by 2 years of investigations and hearings once all of the other things that Congress can do have been done.
We use paper ballots, with markers and Scantronic machines. It's a bit like a large-print SAT. You mark your ballot, then feed it into the machine, which scans it, spits it back out if it's not sufficiently clear and valid, and otherwise counts it and also keeps it in the box. It would be pretty difficult to end up with a ballot which was incorrect but valid (since there are large keep-out areas between where you make marks), and it's not hard to make the correct mark (you need to fill in a broken line with a marker), and the machine is going to reject anything that a human wouldn't read unambiguously the way the machine reads it.
What makes it a good system is the digital discipline: there's a lot of separation between valid states, and the transmitting end has a much narrower valid range than the receiving end does. In order from the ballot to make it out of the voter's hands, it must be very clear; if it gets into the box, it's considered valid and counts for whatever it's close to.
Daniel Brandt is against Wikipedia's portrayal of him not because of it being unflattering (it is, in my opinion, if anything oddly sympathetic to his position, despite his position being that it shouldn't exist at all), but because of his privacy concerns. He's a privacy activist with a particular focus on the actions of information organizing sites, and so he's not unexpectedly against the existance of unauthorized widely-available detailed biographies. He's gone so far as to complain about CIA and NSA websites using cookies, so it's not surprising that he wouldn't be happy about a vast conspiracy to produce reports on unwilling individuals, regardless of the merits of the reports.
I'm not surprised that they decided to make them available again. I'm surprised, though that they didn't leave them off YouTube and instead put them up officially on Google Video. I'd guess they have better encodings of their videos than are on youtube (at least based on how the copies of Weird Al's videos that he put up himself are so much better than the copies put up by other people). And if they put it on Google Video, they could probably work out a deal similar to this one and get some of the revenue from advertizing by the hosting site.
It's just a bit odd for a copyright holder to specificly permit somebody to distribute a copy of something acquired from an unlicensed third party.
They probably really delete clips that the users delete, but only disable clips they're requested to take down, in case it turns out that the entity requesting it doesn't have the authority or something.
The part of the genome that doesn't code for proteins is hardly "junk". It's full of binding sites for regulatory molecules, sections that don't code for anything but help the molecule fold and unfold without breaking.
It's also plausible that the viral sections you mention are kept so that the immune system can produce antibodies for them, by treating those sections of the genome as deactivated virus. This would give the organism as advantage to having those sections deactivated over not having them at all if there is active virus of those sorts out there somewhere.
For that matter, all sorts of weird stuff going on with the immune system. As far as anyone can tell, it uses an enzyme set that does double-stranded break repair for some purpose such that if that set is not functioning, the immune system destroys the organism. Perhaps some of the regions that don't code for proteins are used as a certificate of authenticity for the organism's own cells, demonstrating that those cells have the same random strings that the immune system does. If all of the genome coded for useful proteins, the password wouldn't be secret enough.
It doesn't happen in chess or go, because these have simple rules and two players. But in games with more than two players and complicated rules (as opposed to complexity arising from the non-confluence of the game tree), it's reasonably likely for a novice playing with experts to pick a strategy which is less effective in general, but which avoids direct conflict with the strategies the other players are using.
I once played a game of Rail Baron against a bunch of people who'd been playing it for years. They used the generally advantageous strategy of each buying up all of the railroads in some area, which forces other players to pay to go there. I bought one railroad in each area, and made all my money going places on my own rails. Evidently, their strategy wins if only one person does that and everybody else does mine (since the monopolist will manage to shut out almost everybody, and make a lot of money from them), and it works well if everybody is doing it (you win if you get a good area), but it totally fails if everybody but one person does it (the odd player has a lot of income, while the rest of the players mostly trade money around). And if two people use my strategy, it fails, because they each get into only half of the areas, and the fees for the other areas eat up their profits.
So there is an advantage in some games to using a non-optimal strategy if everybody else is using the same other strategy. Of course, this depends on social dynamics more than pure strategy, which is what chess and go focus on. But business (back to the actual topic) is largely social dynamics.
Re:Definitely has uses but..
on
Oracle Linux?
·
· Score: 1
I was mostly thinking of OCFS, which is only really useful if you're running a clustered Oracle installation, but is important for that one case. Back in 2002 when I last used Oracle, we got a production server which was two Dell machines, running Red Hat, running Oracle, and I don't think we ever managed to get the two machines working together, because Red Hat tech support had no idea how it was supposed to work and Oracle didn't do support for the OS.
I suspect they actually added it because some people occasionally quit their browsers because they're logging out or shutting down their computers. In fact, I think that they probably didn't worry too much about the memory issue in the first place because most people don't run their browsers continuously for weeks.
Oracle is the wrong company to provide competition in the Linux market. Their background is entirely in databases and database-centric middleware. The only thing they can sensibly do with a Linux distribution is support it and configure it optimally for running a database on. Their publicity has been badly misdirected, because they're setting themselves against non-competitors with it. Their target market is really: people who buy SQL Server on Windows to get a complete database server preinstalled with a single tech support contact; people who get frustrated with Oracle at some point in installing Red Hat on the weird database hardware configuration they've bought or in installing Oracle on the Red Hat they've not configured correctly; and people who fiugre that, if they're installing Red Hat, they might as well use one of the databases that it comes with. They're maligning other vendors' offerings, but their real hook is that nobody at your company knows how to get Oracle and Red Hat to work together as well as they're supposed to, but Oracle Linux is already set up that way.
The only way to judge whether this is legitimate is to see whether sites that do fraudulent things (get traffic from mistyped domain names, send out "renewal" requests to non-customers, etc) are able to get these certificates. If Verisign is able to make sure that sites that do these things or have a history of doing them can't get certificates, then maybe they'll mean more than current SSL certificates.
Of course, there are technical issues with a PKI system without trusted root certificates, so it might not work even then.
It'd depend a lot on how it's done. Popular software makers shouldn't install Firefox with their stuff, and Mozilla shouldn't encourage them to. On the other hand, Firefox (at least on Linux) runs perfectly well not installed. (Like, you can untar it in a random subdirectory of your home directory and run it, and you don't even need to have it in your path.) So popular software makers could include Firefox with their stuff and use it uninstalled to view URLs. Then, if people notice that browsing the web is better through Quicken than using the system default browser, they can be told to install Firefox. (And the software maker doesn't have to work around IE rendering bugs for pages accessed through the application.)
I remember when Weblogic got support for EJB 2.0 (I think; it was a while ago now). They were much more prompt about that. They had it implemented and in production use before the standard was even done getting major changes. When we switched (for cost reasons) to JBoss, almost all of the porting effort was in porting from the intermediate version of the standard that Weblogic had implemented to the actual released version.
People confuse the CD containing software with the program on the CD. But they expect anything usb-storage to be a place they can put stuff, not a program. If anything, connecting an iPod should run iTunes (or some other appropriate program for manipulating the inert contents of the iPod), not anything on the iPod.
Slashdot Mirror
It's remarkable how much that gets reported turns out to be unquestionably false. We hear reports of cows with accents, and it turns out that all of the quotes are from people who were mostly directly contradicting the story, but said a few things that could be totally rewritten to suggest support for it. We hear about school districts allowing text message slang in exams, and it turns out that this is entirely reporters extrapolating from schools not flunking students who make spelling mistakes when writing correct answers. We hear that women use 3 times as many words as men, and this turns out to be directly contradicted by every study actually carried out, from major academic studies to curious people handing tape recorders to a pair of people and then counting.
I hardly think it's surprising any more that successful completion of a project is reported as the project being scrapped. It's almost surprising that we didn't get reports on the recent election of the Republicans keeping control of both houses of Congress for another two months.
I think the real problem is that Firefox doesn't make enough use of the extension mechanism. It's clearly sufficiently powerful to do all sorts of interesting things with, but the default install of Firefox doesn't have any extensions at all. You can change a lot of the appearance of Firefox with themes, but it doesn't come with any themes, either (the "default theme" thing is actually not handled as a theme, but is just the built-in defaults for things, which is why you can't install 1.5's default theme on 2.0).
What they should do is split most of the functionality into extensions and all of the appearance into themes, and ship those extensions and themes with the browser. Then you'd be able to choose exactly which features you want from each version. And, since extensions can be updated independantly of the rest of the browser, it would be no problem if there was a bug in deciding that a word had been typed completely for the inline spellchecker; they'd release a new version of the spellchecker whenever they fixed it.
I think the complaint is that Firefox 1.5 is better than Firefox 2.0. If I try out a new Honda, and it's less pleasant to drive than my old Honda, I stick with my current car, and I worry about what I'll do if it breaks down.
All of your items are good reasons to use Firefox 1.5. Firefox 2.0 doesn't add anything I find compelling, and it adds non-removable junk to the navigation toolbar. It's a legitimate worry that I'll eventually have to go to a worse browser than Firefox 1.5 to have the latest security updates, because the Firefox team will only support 2.0 and their trademark policies prohibit other people from fixing things.
The only problem with a manager that's good at what the workers are doing is if the boss doesn't want to quit doing what he's good at. This is a problem because a manager has no attention to devote to work (being the person who filters useless distractions) and little time (having the full-time managing job and twice as many meetings as anybody else). So, no matter how good a techie the person is, the output is lousy, ill-considered, and never ready.
So the answer is really that a manager can't really be a techie, but a manager should be a former techie.
Also, a manager only needs a good approximation (but it can't be a bad approximation) of skill in the field. There's a level of skill where you can't solve a problem yourself, but you know whether it's fundamentally impossible, impractical, or just difficult, and that's ideal for a manager. (The worst thing is if the manager knows the problem can be solved, but nobody else on the team is good enough to solve it and the manager is too busy.)
Of course there's also the Peter Principle; there are plenty of cases of techies without any management skill at all promoted to management positions on the basis of seniority and great technical skill, such that they don't have the necessary skills for their actual job, are too valuable to let go or demote, and don't have the time to do the work they are better at than anyone else.
There are a bunch of things that make this less useful for doing weird things:
You can't send a message back in time. You can only receive a message from the future. That is, you can only send a message back in time to a point where you had arranged to get it. It's like an box that you take stuff out of before you put it in; things go back in time to the point where you took stuff out, not to any other time. So there's no issue with the fact that we're not getting messages from the future; the time before the time machine is invented is inaccessible.
You can't tell what it says in the past. This is where quantum is weird. Basically, what happens is that person A receives the message, which is a series of dots to put in a picture. It looks like random static. Then person B sends the message, which consists of choosing, for each dot, "bell" or "bars". Then they talk to each other, and they find that if you look at only the "bell" dots, the picture is a bell, and if you look at the "bars" dots, it's a set of bars. Since all of the data is collected by A before B chooses, they have to come to the conclusion that something really weird is going on, and the choice later clearly affects the data that was already written down. But they can only come to this conclusion after the experiment is over; before the message is sent, the received message can't be interpreted, although all of the observations can be taken.
This of it like this magic trick: the audience gets a deck of cards with a variety of backs which they examine in detail. A volunteer on stage shuffled a second deck of cards, writes down a few numbers between 1 and 52, and draws the cards with the given numbers (i.e., for 10, draws the 10th card in the shuffled deck). When the volunteer announces the set of names, they all turn out to have the same backs in the audience's deck. The volunteer chose freely, the deck was really random, and the audience saw the fronts and backs of all of the cards in their deck before the choice was made. If the trick is repeated with fresh decks, it always works. We have to conclude that the volunteer is affecting the construction of the deck in the past, but we're only impressed after it's all over, and we have no idea what the volunteer is going to choose in advance. Even if we agree on a set of numbers to pick if the stock market goes up and a different set to pick if it goes down, we can't tell by looking at the audience's deck which it will be, but the trick still works.
But that didn't involve any code whose copyright Sun held. It was stopped because Sun has a trademark on Java which they license only to implementations which pass certain conformance tests (which Microsoft's obviously didn't). It would have made no difference to that case if Sun's JDK were available under the GPL or even under the BSD license, because the complaint wasn't over the copyright on the implementation.
As far as I can tell, the trademark restrictions are still in place, so you can take Sun's JDK and create a derived work that behaves differently, but you can't call it Java unless it still passes the test suite. I expect there to be a number of research and special-purpose derivatives of the Sun JDK before long, but I expect them to have names that don't use the trademark and actually not be "Java" (like C++ isn't C). I also expect people to improve the Sun JVM in a variety of ways that don't change the language (e.g., inline ArrayList.get in methods that call get on a List, with a check on the class of the List object in the native code that the JIT compiler produces, or inline final wrapper methods at call sites, to avoid the function call overheads). I also expect the non-"Java" derivatives to be a useful source of JSRs, and to make the whole JSR process more efficient, because people with proposals can implement them without a huge amount of work making an otherwise functional implementation.
Shouldn't elections be determined when the ballots are cast, not when they're counted? Or maybe when voters decide how they're going to vote?
That's why they're using a closed GSM module. It's like using a cell modem with your laptop, except that it's internal. The carrier doesn't care what you say to your cell modem, because the device is certified to behave appropriately regardless. And the GSM modules are high volume, because a wide variety of applications use them, and the modules don't get new versions all the time to add new UI bells and whistles.
So you're right that part of the device is locked down and tamper-proof, but it's running on an entirely separate chipset (probably only connected by a serial port, actually).
You can't really call Lieberman a Democrat, because he was elected without the Democrat nomination. He was running an an independant. On the other hand, he claims he's going to caucus with the Democrats, so he counts as a Democrat for the purposes of control of the Senate.
As far as getting anything done, it's going to be 2 years of getting bills vetoed, most likely. For that matter, it's likely to be a few weeks of passing all the bills on the Democrat agenda, getting them all vetoed, followed by 2 years of investigations and hearings once all of the other things that Congress can do have been done.
We use paper ballots, with markers and Scantronic machines. It's a bit like a large-print SAT. You mark your ballot, then feed it into the machine, which scans it, spits it back out if it's not sufficiently clear and valid, and otherwise counts it and also keeps it in the box. It would be pretty difficult to end up with a ballot which was incorrect but valid (since there are large keep-out areas between where you make marks), and it's not hard to make the correct mark (you need to fill in a broken line with a marker), and the machine is going to reject anything that a human wouldn't read unambiguously the way the machine reads it.
What makes it a good system is the digital discipline: there's a lot of separation between valid states, and the transmitting end has a much narrower valid range than the receiving end does. In order from the ballot to make it out of the voter's hands, it must be very clear; if it gets into the box, it's considered valid and counts for whatever it's close to.
Daniel Brandt is against Wikipedia's portrayal of him not because of it being unflattering (it is, in my opinion, if anything oddly sympathetic to his position, despite his position being that it shouldn't exist at all), but because of his privacy concerns. He's a privacy activist with a particular focus on the actions of information organizing sites, and so he's not unexpectedly against the existance of unauthorized widely-available detailed biographies. He's gone so far as to complain about CIA and NSA websites using cookies, so it's not surprising that he wouldn't be happy about a vast conspiracy to produce reports on unwilling individuals, regardless of the merits of the reports.
I'm not surprised that they decided to make them available again. I'm surprised, though that they didn't leave them off YouTube and instead put them up officially on Google Video. I'd guess they have better encodings of their videos than are on youtube (at least based on how the copies of Weird Al's videos that he put up himself are so much better than the copies put up by other people). And if they put it on Google Video, they could probably work out a deal similar to this one and get some of the revenue from advertizing by the hosting site.
It's just a bit odd for a copyright holder to specificly permit somebody to distribute a copy of something acquired from an unlicensed third party.
They probably really delete clips that the users delete, but only disable clips they're requested to take down, in case it turns out that the entity requesting it doesn't have the authority or something.
The part of the genome that doesn't code for proteins is hardly "junk". It's full of binding sites for regulatory molecules, sections that don't code for anything but help the molecule fold and unfold without breaking.
It's also plausible that the viral sections you mention are kept so that the immune system can produce antibodies for them, by treating those sections of the genome as deactivated virus. This would give the organism as advantage to having those sections deactivated over not having them at all if there is active virus of those sorts out there somewhere.
For that matter, all sorts of weird stuff going on with the immune system. As far as anyone can tell, it uses an enzyme set that does double-stranded break repair for some purpose such that if that set is not functioning, the immune system destroys the organism. Perhaps some of the regions that don't code for proteins are used as a certificate of authenticity for the organism's own cells, demonstrating that those cells have the same random strings that the immune system does. If all of the genome coded for useful proteins, the password wouldn't be secret enough.
It doesn't happen in chess or go, because these have simple rules and two players. But in games with more than two players and complicated rules (as opposed to complexity arising from the non-confluence of the game tree), it's reasonably likely for a novice playing with experts to pick a strategy which is less effective in general, but which avoids direct conflict with the strategies the other players are using.
I once played a game of Rail Baron against a bunch of people who'd been playing it for years. They used the generally advantageous strategy of each buying up all of the railroads in some area, which forces other players to pay to go there. I bought one railroad in each area, and made all my money going places on my own rails. Evidently, their strategy wins if only one person does that and everybody else does mine (since the monopolist will manage to shut out almost everybody, and make a lot of money from them), and it works well if everybody is doing it (you win if you get a good area), but it totally fails if everybody but one person does it (the odd player has a lot of income, while the rest of the players mostly trade money around). And if two people use my strategy, it fails, because they each get into only half of the areas, and the fees for the other areas eat up their profits.
So there is an advantage in some games to using a non-optimal strategy if everybody else is using the same other strategy. Of course, this depends on social dynamics more than pure strategy, which is what chess and go focus on. But business (back to the actual topic) is largely social dynamics.
I was mostly thinking of OCFS, which is only really useful if you're running a clustered Oracle installation, but is important for that one case. Back in 2002 when I last used Oracle, we got a production server which was two Dell machines, running Red Hat, running Oracle, and I don't think we ever managed to get the two machines working together, because Red Hat tech support had no idea how it was supposed to work and Oracle didn't do support for the OS.
I suspect they actually added it because some people occasionally quit their browsers because they're logging out or shutting down their computers. In fact, I think that they probably didn't worry too much about the memory issue in the first place because most people don't run their browsers continuously for weeks.
Yeah, I ought to have included those links. Ah, well, at least a few people got the joke.
Oracle is the wrong company to provide competition in the Linux market. Their background is entirely in databases and database-centric middleware. The only thing they can sensibly do with a Linux distribution is support it and configure it optimally for running a database on. Their publicity has been badly misdirected, because they're setting themselves against non-competitors with it. Their target market is really: people who buy SQL Server on Windows to get a complete database server preinstalled with a single tech support contact; people who get frustrated with Oracle at some point in installing Red Hat on the weird database hardware configuration they've bought or in installing Oracle on the Red Hat they've not configured correctly; and people who fiugre that, if they're installing Red Hat, they might as well use one of the databases that it comes with. They're maligning other vendors' offerings, but their real hook is that nobody at your company knows how to get Oracle and Red Hat to work together as well as they're supposed to, but Oracle Linux is already set up that way.
The only way to judge whether this is legitimate is to see whether sites that do fraudulent things (get traffic from mistyped domain names, send out "renewal" requests to non-customers, etc) are able to get these certificates. If Verisign is able to make sure that sites that do these things or have a history of doing them can't get certificates, then maybe they'll mean more than current SSL certificates.
Of course, there are technical issues with a PKI system without trusted root certificates, so it might not work even then.
It'd depend a lot on how it's done. Popular software makers shouldn't install Firefox with their stuff, and Mozilla shouldn't encourage them to. On the other hand, Firefox (at least on Linux) runs perfectly well not installed. (Like, you can untar it in a random subdirectory of your home directory and run it, and you don't even need to have it in your path.) So popular software makers could include Firefox with their stuff and use it uninstalled to view URLs. Then, if people notice that browsing the web is better through Quicken than using the system default browser, they can be told to install Firefox. (And the software maker doesn't have to work around IE rendering bugs for pages accessed through the application.)
Legend has it that it won't matter, because if it's using too much memory, you can just restart it without losing what you're in the middle of.
I remember when Weblogic got support for EJB 2.0 (I think; it was a while ago now). They were much more prompt about that. They had it implemented and in production use before the standard was even done getting major changes. When we switched (for cost reasons) to JBoss, almost all of the porting effort was in porting from the intermediate version of the standard that Weblogic had implemented to the actual released version.
But does it run... yes.
People confuse the CD containing software with the program on the CD. But they expect anything usb-storage to be a place they can put stuff, not a program. If anything, connecting an iPod should run iTunes (or some other appropriate program for manipulating the inert contents of the iPod), not anything on the iPod.